diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,78371 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9999329863515536,
+  "eval_steps": 500,
+  "global_step": 11191,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 8.9351531261867e-05,
+      "grad_norm": 0.7860376238822937,
+      "learning_rate": 2.976190476190476e-07,
+      "loss": 1.8548,
+      "step": 1
+    },
+    {
+      "epoch": 0.000178703062523734,
+      "grad_norm": 0.5748365521430969,
+      "learning_rate": 5.952380952380952e-07,
+      "loss": 1.8254,
+      "step": 2
+    },
+    {
+      "epoch": 0.000268054593785601,
+      "grad_norm": 0.8621116876602173,
+      "learning_rate": 8.928571428571428e-07,
+      "loss": 1.8688,
+      "step": 3
+    },
+    {
+      "epoch": 0.000357406125047468,
+      "grad_norm": 0.6630846261978149,
+      "learning_rate": 1.1904761904761904e-06,
+      "loss": 1.8504,
+      "step": 4
+    },
+    {
+      "epoch": 0.000446757656309335,
+      "grad_norm": 0.7190170884132385,
+      "learning_rate": 1.4880952380952381e-06,
+      "loss": 1.8197,
+      "step": 5
+    },
+    {
+      "epoch": 0.000536109187571202,
+      "grad_norm": 0.7450002431869507,
+      "learning_rate": 1.7857142857142857e-06,
+      "loss": 1.8078,
+      "step": 6
+    },
+    {
+      "epoch": 0.000625460718833069,
+      "grad_norm": 0.903852641582489,
+      "learning_rate": 2.0833333333333334e-06,
+      "loss": 1.9407,
+      "step": 7
+    },
+    {
+      "epoch": 0.000714812250094936,
+      "grad_norm": 0.8060555458068848,
+      "learning_rate": 2.3809523809523808e-06,
+      "loss": 1.7434,
+      "step": 8
+    },
+    {
+      "epoch": 0.000804163781356803,
+      "grad_norm": 1.1107157468795776,
+      "learning_rate": 2.6785714285714285e-06,
+      "loss": 1.9593,
+      "step": 9
+    },
+    {
+      "epoch": 0.00089351531261867,
+      "grad_norm": 0.7935706973075867,
+      "learning_rate": 2.9761904761904763e-06,
+      "loss": 1.8844,
+      "step": 10
+    },
+    {
+      "epoch": 0.000982866843880537,
+      "grad_norm": 0.825862467288971,
+      "learning_rate": 3.273809523809524e-06,
+      "loss": 1.8664,
+      "step": 11
+    },
+    {
+      "epoch": 0.001072218375142404,
+      "grad_norm": 0.8665326833724976,
+      "learning_rate": 3.5714285714285714e-06,
+      "loss": 1.8477,
+      "step": 12
+    },
+    {
+      "epoch": 0.001161569906404271,
+      "grad_norm": 0.7275350689888,
+      "learning_rate": 3.869047619047619e-06,
+      "loss": 1.862,
+      "step": 13
+    },
+    {
+      "epoch": 0.001250921437666138,
+      "grad_norm": 0.9508900046348572,
+      "learning_rate": 4.166666666666667e-06,
+      "loss": 1.8362,
+      "step": 14
+    },
+    {
+      "epoch": 0.001340272968928005,
+      "grad_norm": 0.9345436096191406,
+      "learning_rate": 4.464285714285715e-06,
+      "loss": 1.9073,
+      "step": 15
+    },
+    {
+      "epoch": 0.001429624500189872,
+      "grad_norm": 0.8254266977310181,
+      "learning_rate": 4.7619047619047615e-06,
+      "loss": 1.8047,
+      "step": 16
+    },
+    {
+      "epoch": 0.001518976031451739,
+      "grad_norm": 0.8926609754562378,
+      "learning_rate": 5.05952380952381e-06,
+      "loss": 1.8944,
+      "step": 17
+    },
+    {
+      "epoch": 0.001608327562713606,
+      "grad_norm": 0.8045806288719177,
+      "learning_rate": 5.357142857142857e-06,
+      "loss": 1.8269,
+      "step": 18
+    },
+    {
+      "epoch": 0.001697679093975473,
+      "grad_norm": 0.7670357823371887,
+      "learning_rate": 5.654761904761905e-06,
+      "loss": 1.7618,
+      "step": 19
+    },
+    {
+      "epoch": 0.00178703062523734,
+      "grad_norm": 0.8601244688034058,
+      "learning_rate": 5.9523809523809525e-06,
+      "loss": 1.8391,
+      "step": 20
+    },
+    {
+      "epoch": 0.001876382156499207,
+      "grad_norm": 0.9491680264472961,
+      "learning_rate": 6.25e-06,
+      "loss": 1.8631,
+      "step": 21
+    },
+    {
+      "epoch": 0.001965733687761074,
+      "grad_norm": 0.8884113430976868,
+      "learning_rate": 6.547619047619048e-06,
+      "loss": 1.786,
+      "step": 22
+    },
+    {
+      "epoch": 0.002055085219022941,
+      "grad_norm": 0.8622804880142212,
+      "learning_rate": 6.845238095238096e-06,
+      "loss": 1.7653,
+      "step": 23
+    },
+    {
+      "epoch": 0.002144436750284808,
+      "grad_norm": 0.8542262315750122,
+      "learning_rate": 7.142857142857143e-06,
+      "loss": 1.7977,
+      "step": 24
+    },
+    {
+      "epoch": 0.002233788281546675,
+      "grad_norm": 0.6878840923309326,
+      "learning_rate": 7.4404761904761905e-06,
+      "loss": 1.7566,
+      "step": 25
+    },
+    {
+      "epoch": 0.002323139812808542,
+      "grad_norm": 1.2377084493637085,
+      "learning_rate": 7.738095238095238e-06,
+      "loss": 1.7302,
+      "step": 26
+    },
+    {
+      "epoch": 0.002412491344070409,
+      "grad_norm": 0.9201034903526306,
+      "learning_rate": 8.035714285714286e-06,
+      "loss": 1.7652,
+      "step": 27
+    },
+    {
+      "epoch": 0.002501842875332276,
+      "grad_norm": 0.9445549249649048,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.7671,
+      "step": 28
+    },
+    {
+      "epoch": 0.002591194406594143,
+      "grad_norm": 0.7296178340911865,
+      "learning_rate": 8.630952380952381e-06,
+      "loss": 1.6999,
+      "step": 29
+    },
+    {
+      "epoch": 0.00268054593785601,
+      "grad_norm": 0.6263100504875183,
+      "learning_rate": 8.92857142857143e-06,
+      "loss": 1.7825,
+      "step": 30
+    },
+    {
+      "epoch": 0.002769897469117877,
+      "grad_norm": 0.5757091045379639,
+      "learning_rate": 9.226190476190477e-06,
+      "loss": 1.6878,
+      "step": 31
+    },
+    {
+      "epoch": 0.002859249000379744,
+      "grad_norm": 0.5896580219268799,
+      "learning_rate": 9.523809523809523e-06,
+      "loss": 1.658,
+      "step": 32
+    },
+    {
+      "epoch": 0.002948600531641611,
+      "grad_norm": 0.4904349446296692,
+      "learning_rate": 9.821428571428573e-06,
+      "loss": 1.6135,
+      "step": 33
+    },
+    {
+      "epoch": 0.003037952062903478,
+      "grad_norm": 0.628957211971283,
+      "learning_rate": 1.011904761904762e-05,
+      "loss": 1.617,
+      "step": 34
+    },
+    {
+      "epoch": 0.003127303594165345,
+      "grad_norm": 0.5395067930221558,
+      "learning_rate": 1.0416666666666668e-05,
+      "loss": 1.6439,
+      "step": 35
+    },
+    {
+      "epoch": 0.003216655125427212,
+      "grad_norm": 0.4463937282562256,
+      "learning_rate": 1.0714285714285714e-05,
+      "loss": 1.5962,
+      "step": 36
+    },
+    {
+      "epoch": 0.003306006656689079,
+      "grad_norm": 0.4546874463558197,
+      "learning_rate": 1.1011904761904762e-05,
+      "loss": 1.6277,
+      "step": 37
+    },
+    {
+      "epoch": 0.003395358187950946,
+      "grad_norm": 0.45821845531463623,
+      "learning_rate": 1.130952380952381e-05,
+      "loss": 1.6352,
+      "step": 38
+    },
+    {
+      "epoch": 0.003484709719212813,
+      "grad_norm": 0.59076327085495,
+      "learning_rate": 1.1607142857142857e-05,
+      "loss": 1.6098,
+      "step": 39
+    },
+    {
+      "epoch": 0.00357406125047468,
+      "grad_norm": 0.39401739835739136,
+      "learning_rate": 1.1904761904761905e-05,
+      "loss": 1.6404,
+      "step": 40
+    },
+    {
+      "epoch": 0.003663412781736547,
+      "grad_norm": 0.3501727879047394,
+      "learning_rate": 1.2202380952380953e-05,
+      "loss": 1.6279,
+      "step": 41
+    },
+    {
+      "epoch": 0.003752764312998414,
+      "grad_norm": 0.3471863269805908,
+      "learning_rate": 1.25e-05,
+      "loss": 1.5823,
+      "step": 42
+    },
+    {
+      "epoch": 0.003842115844260281,
+      "grad_norm": 0.3801725506782532,
+      "learning_rate": 1.2797619047619047e-05,
+      "loss": 1.5969,
+      "step": 43
+    },
+    {
+      "epoch": 0.003931467375522148,
+      "grad_norm": 0.3517632782459259,
+      "learning_rate": 1.3095238095238096e-05,
+      "loss": 1.6075,
+      "step": 44
+    },
+    {
+      "epoch": 0.004020818906784015,
+      "grad_norm": 0.32643967866897583,
+      "learning_rate": 1.3392857142857144e-05,
+      "loss": 1.537,
+      "step": 45
+    },
+    {
+      "epoch": 0.004110170438045882,
+      "grad_norm": 0.36205053329467773,
+      "learning_rate": 1.3690476190476192e-05,
+      "loss": 1.5646,
+      "step": 46
+    },
+    {
+      "epoch": 0.004199521969307749,
+      "grad_norm": 0.3534471094608307,
+      "learning_rate": 1.398809523809524e-05,
+      "loss": 1.5713,
+      "step": 47
+    },
+    {
+      "epoch": 0.004288873500569616,
+      "grad_norm": 0.349733829498291,
+      "learning_rate": 1.4285714285714285e-05,
+      "loss": 1.5545,
+      "step": 48
+    },
+    {
+      "epoch": 0.004378225031831483,
+      "grad_norm": 0.3604637682437897,
+      "learning_rate": 1.4583333333333335e-05,
+      "loss": 1.569,
+      "step": 49
+    },
+    {
+      "epoch": 0.00446757656309335,
+      "grad_norm": 0.3466920256614685,
+      "learning_rate": 1.4880952380952381e-05,
+      "loss": 1.571,
+      "step": 50
+    },
+    {
+      "epoch": 0.004556928094355217,
+      "grad_norm": 0.3057480752468109,
+      "learning_rate": 1.5178571428571429e-05,
+      "loss": 1.4928,
+      "step": 51
+    },
+    {
+      "epoch": 0.004646279625617084,
+      "grad_norm": 0.3168238401412964,
+      "learning_rate": 1.5476190476190476e-05,
+      "loss": 1.5952,
+      "step": 52
+    },
+    {
+      "epoch": 0.004735631156878951,
+      "grad_norm": 0.33002302050590515,
+      "learning_rate": 1.5773809523809524e-05,
+      "loss": 1.5885,
+      "step": 53
+    },
+    {
+      "epoch": 0.004824982688140818,
+      "grad_norm": 0.35049816966056824,
+      "learning_rate": 1.6071428571428572e-05,
+      "loss": 1.5679,
+      "step": 54
+    },
+    {
+      "epoch": 0.004914334219402685,
+      "grad_norm": 0.398250937461853,
+      "learning_rate": 1.636904761904762e-05,
+      "loss": 1.4671,
+      "step": 55
+    },
+    {
+      "epoch": 0.005003685750664552,
+      "grad_norm": 0.34120944142341614,
+      "learning_rate": 1.6666666666666667e-05,
+      "loss": 1.5189,
+      "step": 56
+    },
+    {
+      "epoch": 0.005093037281926419,
+      "grad_norm": 0.3141593635082245,
+      "learning_rate": 1.6964285714285715e-05,
+      "loss": 1.5387,
+      "step": 57
+    },
+    {
+      "epoch": 0.005182388813188286,
+      "grad_norm": 0.2942512035369873,
+      "learning_rate": 1.7261904761904763e-05,
+      "loss": 1.5823,
+      "step": 58
+    },
+    {
+      "epoch": 0.005271740344450153,
+      "grad_norm": 0.3249220550060272,
+      "learning_rate": 1.755952380952381e-05,
+      "loss": 1.4426,
+      "step": 59
+    },
+    {
+      "epoch": 0.00536109187571202,
+      "grad_norm": 0.3305152356624603,
+      "learning_rate": 1.785714285714286e-05,
+      "loss": 1.4518,
+      "step": 60
+    },
+    {
+      "epoch": 0.005450443406973887,
+      "grad_norm": 0.31893137097358704,
+      "learning_rate": 1.8154761904761906e-05,
+      "loss": 1.4131,
+      "step": 61
+    },
+    {
+      "epoch": 0.005539794938235754,
+      "grad_norm": 0.35763436555862427,
+      "learning_rate": 1.8452380952380954e-05,
+      "loss": 1.4497,
+      "step": 62
+    },
+    {
+      "epoch": 0.005629146469497621,
+      "grad_norm": 0.34308096766471863,
+      "learning_rate": 1.8750000000000002e-05,
+      "loss": 1.445,
+      "step": 63
+    },
+    {
+      "epoch": 0.005718498000759488,
+      "grad_norm": 0.3230386972427368,
+      "learning_rate": 1.9047619047619046e-05,
+      "loss": 1.4297,
+      "step": 64
+    },
+    {
+      "epoch": 0.005807849532021355,
+      "grad_norm": 0.3689243495464325,
+      "learning_rate": 1.9345238095238097e-05,
+      "loss": 1.4096,
+      "step": 65
+    },
+    {
+      "epoch": 0.005897201063283222,
+      "grad_norm": 0.2907058894634247,
+      "learning_rate": 1.9642857142857145e-05,
+      "loss": 1.49,
+      "step": 66
+    },
+    {
+      "epoch": 0.005986552594545089,
+      "grad_norm": 0.3283701539039612,
+      "learning_rate": 1.9940476190476193e-05,
+      "loss": 1.4467,
+      "step": 67
+    },
+    {
+      "epoch": 0.006075904125806956,
+      "grad_norm": 0.29510149359703064,
+      "learning_rate": 2.023809523809524e-05,
+      "loss": 1.4465,
+      "step": 68
+    },
+    {
+      "epoch": 0.006165255657068823,
+      "grad_norm": 0.3028480112552643,
+      "learning_rate": 2.0535714285714285e-05,
+      "loss": 1.5013,
+      "step": 69
+    },
+    {
+      "epoch": 0.00625460718833069,
+      "grad_norm": 0.30100324749946594,
+      "learning_rate": 2.0833333333333336e-05,
+      "loss": 1.3521,
+      "step": 70
+    },
+    {
+      "epoch": 0.006343958719592557,
+      "grad_norm": 0.2899802029132843,
+      "learning_rate": 2.113095238095238e-05,
+      "loss": 1.4652,
+      "step": 71
+    },
+    {
+      "epoch": 0.006433310250854424,
+      "grad_norm": 0.37080496549606323,
+      "learning_rate": 2.1428571428571428e-05,
+      "loss": 1.3875,
+      "step": 72
+    },
+    {
+      "epoch": 0.006522661782116291,
+      "grad_norm": 0.3363150656223297,
+      "learning_rate": 2.172619047619048e-05,
+      "loss": 1.355,
+      "step": 73
+    },
+    {
+      "epoch": 0.006612013313378158,
+      "grad_norm": 0.3124528229236603,
+      "learning_rate": 2.2023809523809524e-05,
+      "loss": 1.3817,
+      "step": 74
+    },
+    {
+      "epoch": 0.006701364844640025,
+      "grad_norm": 0.293714702129364,
+      "learning_rate": 2.2321428571428575e-05,
+      "loss": 1.3825,
+      "step": 75
+    },
+    {
+      "epoch": 0.006790716375901892,
+      "grad_norm": 0.3099130690097809,
+      "learning_rate": 2.261904761904762e-05,
+      "loss": 1.4276,
+      "step": 76
+    },
+    {
+      "epoch": 0.006880067907163759,
+      "grad_norm": 0.32042503356933594,
+      "learning_rate": 2.2916666666666667e-05,
+      "loss": 1.4544,
+      "step": 77
+    },
+    {
+      "epoch": 0.006969419438425626,
+      "grad_norm": 0.3349694311618805,
+      "learning_rate": 2.3214285714285715e-05,
+      "loss": 1.3263,
+      "step": 78
+    },
+    {
+      "epoch": 0.007058770969687493,
+      "grad_norm": 0.3454136848449707,
+      "learning_rate": 2.3511904761904762e-05,
+      "loss": 1.3386,
+      "step": 79
+    },
+    {
+      "epoch": 0.00714812250094936,
+      "grad_norm": 0.30371707677841187,
+      "learning_rate": 2.380952380952381e-05,
+      "loss": 1.3713,
+      "step": 80
+    },
+    {
+      "epoch": 0.007237474032211227,
+      "grad_norm": 0.35178881883621216,
+      "learning_rate": 2.4107142857142858e-05,
+      "loss": 1.2754,
+      "step": 81
+    },
+    {
+      "epoch": 0.007326825563473094,
+      "grad_norm": 0.34617844223976135,
+      "learning_rate": 2.4404761904761906e-05,
+      "loss": 1.4094,
+      "step": 82
+    },
+    {
+      "epoch": 0.007416177094734961,
+      "grad_norm": 0.31313589215278625,
+      "learning_rate": 2.4702380952380953e-05,
+      "loss": 1.3358,
+      "step": 83
+    },
+    {
+      "epoch": 0.007505528625996828,
+      "grad_norm": 0.3196007311344147,
+      "learning_rate": 2.5e-05,
+      "loss": 1.4212,
+      "step": 84
+    },
+    {
+      "epoch": 0.007594880157258695,
+      "grad_norm": 0.3096368610858917,
+      "learning_rate": 2.529761904761905e-05,
+      "loss": 1.408,
+      "step": 85
+    },
+    {
+      "epoch": 0.007684231688520562,
+      "grad_norm": 0.4201064705848694,
+      "learning_rate": 2.5595238095238093e-05,
+      "loss": 1.4166,
+      "step": 86
+    },
+    {
+      "epoch": 0.007773583219782429,
+      "grad_norm": 0.2974705398082733,
+      "learning_rate": 2.5892857142857148e-05,
+      "loss": 1.4298,
+      "step": 87
+    },
+    {
+      "epoch": 0.007862934751044296,
+      "grad_norm": 0.29475000500679016,
+      "learning_rate": 2.6190476190476192e-05,
+      "loss": 1.363,
+      "step": 88
+    },
+    {
+      "epoch": 0.007952286282306162,
+      "grad_norm": 0.29409998655319214,
+      "learning_rate": 2.648809523809524e-05,
+      "loss": 1.3672,
+      "step": 89
+    },
+    {
+      "epoch": 0.00804163781356803,
+      "grad_norm": 0.33392786979675293,
+      "learning_rate": 2.6785714285714288e-05,
+      "loss": 1.3839,
+      "step": 90
+    },
+    {
+      "epoch": 0.008130989344829897,
+      "grad_norm": 0.35475441813468933,
+      "learning_rate": 2.7083333333333332e-05,
+      "loss": 1.3376,
+      "step": 91
+    },
+    {
+      "epoch": 0.008220340876091765,
+      "grad_norm": 0.3339526951313019,
+      "learning_rate": 2.7380952380952383e-05,
+      "loss": 1.3865,
+      "step": 92
+    },
+    {
+      "epoch": 0.00830969240735363,
+      "grad_norm": 0.35839566588401794,
+      "learning_rate": 2.767857142857143e-05,
+      "loss": 1.2942,
+      "step": 93
+    },
+    {
+      "epoch": 0.008399043938615498,
+      "grad_norm": 0.3491029739379883,
+      "learning_rate": 2.797619047619048e-05,
+      "loss": 1.2828,
+      "step": 94
+    },
+    {
+      "epoch": 0.008488395469877365,
+      "grad_norm": 0.3070695400238037,
+      "learning_rate": 2.8273809523809523e-05,
+      "loss": 1.3261,
+      "step": 95
+    },
+    {
+      "epoch": 0.008577747001139231,
+      "grad_norm": 0.3386940360069275,
+      "learning_rate": 2.857142857142857e-05,
+      "loss": 1.3626,
+      "step": 96
+    },
+    {
+      "epoch": 0.008667098532401099,
+      "grad_norm": 0.3212776184082031,
+      "learning_rate": 2.886904761904762e-05,
+      "loss": 1.3482,
+      "step": 97
+    },
+    {
+      "epoch": 0.008756450063662966,
+      "grad_norm": 0.3101659119129181,
+      "learning_rate": 2.916666666666667e-05,
+      "loss": 1.3217,
+      "step": 98
+    },
+    {
+      "epoch": 0.008845801594924834,
+      "grad_norm": 0.3669290542602539,
+      "learning_rate": 2.9464285714285718e-05,
+      "loss": 1.3461,
+      "step": 99
+    },
+    {
+      "epoch": 0.0089351531261867,
+      "grad_norm": 0.3357386291027069,
+      "learning_rate": 2.9761904761904762e-05,
+      "loss": 1.3613,
+      "step": 100
+    },
+    {
+      "epoch": 0.009024504657448567,
+      "grad_norm": 0.3187052607536316,
+      "learning_rate": 3.005952380952381e-05,
+      "loss": 1.2509,
+      "step": 101
+    },
+    {
+      "epoch": 0.009113856188710434,
+      "grad_norm": 0.36679479479789734,
+      "learning_rate": 3.0357142857142857e-05,
+      "loss": 1.3448,
+      "step": 102
+    },
+    {
+      "epoch": 0.0092032077199723,
+      "grad_norm": 0.3677942752838135,
+      "learning_rate": 3.0654761904761905e-05,
+      "loss": 1.3676,
+      "step": 103
+    },
+    {
+      "epoch": 0.009292559251234168,
+      "grad_norm": 0.36349645256996155,
+      "learning_rate": 3.095238095238095e-05,
+      "loss": 1.3115,
+      "step": 104
+    },
+    {
+      "epoch": 0.009381910782496035,
+      "grad_norm": 0.3469372093677521,
+      "learning_rate": 3.125e-05,
+      "loss": 1.3233,
+      "step": 105
+    },
+    {
+      "epoch": 0.009471262313757903,
+      "grad_norm": 0.556049644947052,
+      "learning_rate": 3.154761904761905e-05,
+      "loss": 1.299,
+      "step": 106
+    },
+    {
+      "epoch": 0.009560613845019768,
+      "grad_norm": 0.3648083508014679,
+      "learning_rate": 3.1845238095238096e-05,
+      "loss": 1.3484,
+      "step": 107
+    },
+    {
+      "epoch": 0.009649965376281636,
+      "grad_norm": 0.3116719424724579,
+      "learning_rate": 3.2142857142857144e-05,
+      "loss": 1.3059,
+      "step": 108
+    },
+    {
+      "epoch": 0.009739316907543503,
+      "grad_norm": 0.3249325454235077,
+      "learning_rate": 3.244047619047619e-05,
+      "loss": 1.2851,
+      "step": 109
+    },
+    {
+      "epoch": 0.00982866843880537,
+      "grad_norm": 0.3656911551952362,
+      "learning_rate": 3.273809523809524e-05,
+      "loss": 1.2504,
+      "step": 110
+    },
+    {
+      "epoch": 0.009918019970067237,
+      "grad_norm": 0.3634350895881653,
+      "learning_rate": 3.303571428571429e-05,
+      "loss": 1.4084,
+      "step": 111
+    },
+    {
+      "epoch": 0.010007371501329104,
+      "grad_norm": 0.39990171790122986,
+      "learning_rate": 3.3333333333333335e-05,
+      "loss": 1.3367,
+      "step": 112
+    },
+    {
+      "epoch": 0.010096723032590972,
+      "grad_norm": 0.4223339855670929,
+      "learning_rate": 3.363095238095238e-05,
+      "loss": 1.2866,
+      "step": 113
+    },
+    {
+      "epoch": 0.010186074563852838,
+      "grad_norm": 0.3745315968990326,
+      "learning_rate": 3.392857142857143e-05,
+      "loss": 1.3021,
+      "step": 114
+    },
+    {
+      "epoch": 0.010275426095114705,
+      "grad_norm": 0.4239887297153473,
+      "learning_rate": 3.422619047619048e-05,
+      "loss": 1.2906,
+      "step": 115
+    },
+    {
+      "epoch": 0.010364777626376573,
+      "grad_norm": 0.3679508566856384,
+      "learning_rate": 3.4523809523809526e-05,
+      "loss": 1.2425,
+      "step": 116
+    },
+    {
+      "epoch": 0.010454129157638438,
+      "grad_norm": 0.38168981671333313,
+      "learning_rate": 3.4821428571428574e-05,
+      "loss": 1.213,
+      "step": 117
+    },
+    {
+      "epoch": 0.010543480688900306,
+      "grad_norm": 0.33976343274116516,
+      "learning_rate": 3.511904761904762e-05,
+      "loss": 1.2012,
+      "step": 118
+    },
+    {
+      "epoch": 0.010632832220162173,
+      "grad_norm": 0.3905802369117737,
+      "learning_rate": 3.541666666666667e-05,
+      "loss": 1.2856,
+      "step": 119
+    },
+    {
+      "epoch": 0.01072218375142404,
+      "grad_norm": 0.3847375810146332,
+      "learning_rate": 3.571428571428572e-05,
+      "loss": 1.3329,
+      "step": 120
+    },
+    {
+      "epoch": 0.010811535282685907,
+      "grad_norm": 0.38759222626686096,
+      "learning_rate": 3.6011904761904765e-05,
+      "loss": 1.3153,
+      "step": 121
+    },
+    {
+      "epoch": 0.010900886813947774,
+      "grad_norm": 0.36399710178375244,
+      "learning_rate": 3.630952380952381e-05,
+      "loss": 1.2413,
+      "step": 122
+    },
+    {
+      "epoch": 0.010990238345209642,
+      "grad_norm": 0.390902042388916,
+      "learning_rate": 3.6607142857142853e-05,
+      "loss": 1.2319,
+      "step": 123
+    },
+    {
+      "epoch": 0.011079589876471507,
+      "grad_norm": 0.37981510162353516,
+      "learning_rate": 3.690476190476191e-05,
+      "loss": 1.2977,
+      "step": 124
+    },
+    {
+      "epoch": 0.011168941407733375,
+      "grad_norm": 0.3691035807132721,
+      "learning_rate": 3.7202380952380956e-05,
+      "loss": 1.2927,
+      "step": 125
+    },
+    {
+      "epoch": 0.011258292938995242,
+      "grad_norm": 0.3708125948905945,
+      "learning_rate": 3.7500000000000003e-05,
+      "loss": 1.2989,
+      "step": 126
+    },
+    {
+      "epoch": 0.01134764447025711,
+      "grad_norm": 0.3609068989753723,
+      "learning_rate": 3.779761904761905e-05,
+      "loss": 1.3181,
+      "step": 127
+    },
+    {
+      "epoch": 0.011436996001518976,
+      "grad_norm": 0.4091266095638275,
+      "learning_rate": 3.809523809523809e-05,
+      "loss": 1.2894,
+      "step": 128
+    },
+    {
+      "epoch": 0.011526347532780843,
+      "grad_norm": 0.41847166419029236,
+      "learning_rate": 3.839285714285715e-05,
+      "loss": 1.1952,
+      "step": 129
+    },
+    {
+      "epoch": 0.01161569906404271,
+      "grad_norm": 0.5045241713523865,
+      "learning_rate": 3.8690476190476195e-05,
+      "loss": 1.2808,
+      "step": 130
+    },
+    {
+      "epoch": 0.011705050595304576,
+      "grad_norm": 0.4047856032848358,
+      "learning_rate": 3.898809523809524e-05,
+      "loss": 1.2858,
+      "step": 131
+    },
+    {
+      "epoch": 0.011794402126566444,
+      "grad_norm": 0.4096450209617615,
+      "learning_rate": 3.928571428571429e-05,
+      "loss": 1.3116,
+      "step": 132
+    },
+    {
+      "epoch": 0.011883753657828311,
+      "grad_norm": 0.3629261553287506,
+      "learning_rate": 3.958333333333333e-05,
+      "loss": 1.2968,
+      "step": 133
+    },
+    {
+      "epoch": 0.011973105189090179,
+      "grad_norm": 0.40072211623191833,
+      "learning_rate": 3.9880952380952386e-05,
+      "loss": 1.257,
+      "step": 134
+    },
+    {
+      "epoch": 0.012062456720352045,
+      "grad_norm": 0.40854790806770325,
+      "learning_rate": 4.017857142857143e-05,
+      "loss": 1.2404,
+      "step": 135
+    },
+    {
+      "epoch": 0.012151808251613912,
+      "grad_norm": 0.43366968631744385,
+      "learning_rate": 4.047619047619048e-05,
+      "loss": 1.2719,
+      "step": 136
+    },
+    {
+      "epoch": 0.01224115978287578,
+      "grad_norm": 0.35978153347969055,
+      "learning_rate": 4.077380952380952e-05,
+      "loss": 1.2299,
+      "step": 137
+    },
+    {
+      "epoch": 0.012330511314137645,
+      "grad_norm": 0.3845697343349457,
+      "learning_rate": 4.107142857142857e-05,
+      "loss": 1.2342,
+      "step": 138
+    },
+    {
+      "epoch": 0.012419862845399513,
+      "grad_norm": 0.48100224137306213,
+      "learning_rate": 4.136904761904762e-05,
+      "loss": 1.2568,
+      "step": 139
+    },
+    {
+      "epoch": 0.01250921437666138,
+      "grad_norm": 0.40048471093177795,
+      "learning_rate": 4.166666666666667e-05,
+      "loss": 1.2981,
+      "step": 140
+    },
+    {
+      "epoch": 0.012598565907923248,
+      "grad_norm": 0.46120962500572205,
+      "learning_rate": 4.196428571428572e-05,
+      "loss": 1.2794,
+      "step": 141
+    },
+    {
+      "epoch": 0.012687917439185114,
+      "grad_norm": 0.4530184864997864,
+      "learning_rate": 4.226190476190476e-05,
+      "loss": 1.2063,
+      "step": 142
+    },
+    {
+      "epoch": 0.012777268970446981,
+      "grad_norm": 0.39693745970726013,
+      "learning_rate": 4.255952380952381e-05,
+      "loss": 1.3012,
+      "step": 143
+    },
+    {
+      "epoch": 0.012866620501708849,
+      "grad_norm": 0.40875282883644104,
+      "learning_rate": 4.2857142857142856e-05,
+      "loss": 1.2265,
+      "step": 144
+    },
+    {
+      "epoch": 0.012955972032970714,
+      "grad_norm": 0.4045838415622711,
+      "learning_rate": 4.315476190476191e-05,
+      "loss": 1.2228,
+      "step": 145
+    },
+    {
+      "epoch": 0.013045323564232582,
+      "grad_norm": 0.4586333632469177,
+      "learning_rate": 4.345238095238096e-05,
+      "loss": 1.215,
+      "step": 146
+    },
+    {
+      "epoch": 0.01313467509549445,
+      "grad_norm": 0.4658620059490204,
+      "learning_rate": 4.375e-05,
+      "loss": 1.3559,
+      "step": 147
+    },
+    {
+      "epoch": 0.013224026626756317,
+      "grad_norm": 0.4266040325164795,
+      "learning_rate": 4.404761904761905e-05,
+      "loss": 1.2215,
+      "step": 148
+    },
+    {
+      "epoch": 0.013313378158018183,
+      "grad_norm": 0.39535924792289734,
+      "learning_rate": 4.4345238095238095e-05,
+      "loss": 1.3024,
+      "step": 149
+    },
+    {
+      "epoch": 0.01340272968928005,
+      "grad_norm": 0.4783201515674591,
+      "learning_rate": 4.464285714285715e-05,
+      "loss": 1.2651,
+      "step": 150
+    },
+    {
+      "epoch": 0.013492081220541918,
+      "grad_norm": 0.44912487268447876,
+      "learning_rate": 4.494047619047619e-05,
+      "loss": 1.2107,
+      "step": 151
+    },
+    {
+      "epoch": 0.013581432751803783,
+      "grad_norm": 0.40533971786499023,
+      "learning_rate": 4.523809523809524e-05,
+      "loss": 1.2386,
+      "step": 152
+    },
+    {
+      "epoch": 0.01367078428306565,
+      "grad_norm": 0.47869575023651123,
+      "learning_rate": 4.5535714285714286e-05,
+      "loss": 1.2709,
+      "step": 153
+    },
+    {
+      "epoch": 0.013760135814327518,
+      "grad_norm": 0.45831164717674255,
+      "learning_rate": 4.5833333333333334e-05,
+      "loss": 1.2435,
+      "step": 154
+    },
+    {
+      "epoch": 0.013849487345589386,
+      "grad_norm": 0.5027099847793579,
+      "learning_rate": 4.613095238095239e-05,
+      "loss": 1.2706,
+      "step": 155
+    },
+    {
+      "epoch": 0.013938838876851252,
+      "grad_norm": 0.4425562024116516,
+      "learning_rate": 4.642857142857143e-05,
+      "loss": 1.1868,
+      "step": 156
+    },
+    {
+      "epoch": 0.014028190408113119,
+      "grad_norm": 0.5197780728340149,
+      "learning_rate": 4.672619047619048e-05,
+      "loss": 1.2776,
+      "step": 157
+    },
+    {
+      "epoch": 0.014117541939374987,
+      "grad_norm": 0.4800921380519867,
+      "learning_rate": 4.7023809523809525e-05,
+      "loss": 1.2213,
+      "step": 158
+    },
+    {
+      "epoch": 0.014206893470636852,
+      "grad_norm": 0.4182106852531433,
+      "learning_rate": 4.732142857142857e-05,
+      "loss": 1.2358,
+      "step": 159
+    },
+    {
+      "epoch": 0.01429624500189872,
+      "grad_norm": 0.4423435926437378,
+      "learning_rate": 4.761904761904762e-05,
+      "loss": 1.263,
+      "step": 160
+    },
+    {
+      "epoch": 0.014385596533160587,
+      "grad_norm": 0.5085612535476685,
+      "learning_rate": 4.791666666666667e-05,
+      "loss": 1.206,
+      "step": 161
+    },
+    {
+      "epoch": 0.014474948064422455,
+      "grad_norm": 0.4774646461009979,
+      "learning_rate": 4.8214285714285716e-05,
+      "loss": 1.2295,
+      "step": 162
+    },
+    {
+      "epoch": 0.01456429959568432,
+      "grad_norm": 0.4879673719406128,
+      "learning_rate": 4.8511904761904764e-05,
+      "loss": 1.2492,
+      "step": 163
+    },
+    {
+      "epoch": 0.014653651126946188,
+      "grad_norm": 0.4564456343650818,
+      "learning_rate": 4.880952380952381e-05,
+      "loss": 1.196,
+      "step": 164
+    },
+    {
+      "epoch": 0.014743002658208056,
+      "grad_norm": 0.4189358353614807,
+      "learning_rate": 4.910714285714286e-05,
+      "loss": 1.2245,
+      "step": 165
+    },
+    {
+      "epoch": 0.014832354189469921,
+      "grad_norm": 0.48813703656196594,
+      "learning_rate": 4.940476190476191e-05,
+      "loss": 1.1321,
+      "step": 166
+    },
+    {
+      "epoch": 0.014921705720731789,
+      "grad_norm": 0.5083624124526978,
+      "learning_rate": 4.9702380952380955e-05,
+      "loss": 1.1557,
+      "step": 167
+    },
+    {
+      "epoch": 0.015011057251993656,
+      "grad_norm": 0.4980645179748535,
+      "learning_rate": 5e-05,
+      "loss": 1.2758,
+      "step": 168
+    },
+    {
+      "epoch": 0.015100408783255524,
+      "grad_norm": 0.4460103511810303,
+      "learning_rate": 5.029761904761905e-05,
+      "loss": 1.2135,
+      "step": 169
+    },
+    {
+      "epoch": 0.01518976031451739,
+      "grad_norm": 0.4435397684574127,
+      "learning_rate": 5.05952380952381e-05,
+      "loss": 1.295,
+      "step": 170
+    },
+    {
+      "epoch": 0.015279111845779257,
+      "grad_norm": 0.5360916256904602,
+      "learning_rate": 5.089285714285714e-05,
+      "loss": 1.2055,
+      "step": 171
+    },
+    {
+      "epoch": 0.015368463377041125,
+      "grad_norm": 0.5073984265327454,
+      "learning_rate": 5.119047619047619e-05,
+      "loss": 1.2104,
+      "step": 172
+    },
+    {
+      "epoch": 0.01545781490830299,
+      "grad_norm": 0.4850315451622009,
+      "learning_rate": 5.1488095238095234e-05,
+      "loss": 1.2246,
+      "step": 173
+    },
+    {
+      "epoch": 0.015547166439564858,
+      "grad_norm": 0.46031928062438965,
+      "learning_rate": 5.1785714285714296e-05,
+      "loss": 1.1649,
+      "step": 174
+    },
+    {
+      "epoch": 0.015636517970826724,
+      "grad_norm": 0.4688253104686737,
+      "learning_rate": 5.208333333333334e-05,
+      "loss": 1.2064,
+      "step": 175
+    },
+    {
+      "epoch": 0.015725869502088593,
+      "grad_norm": 0.44692277908325195,
+      "learning_rate": 5.2380952380952384e-05,
+      "loss": 1.2672,
+      "step": 176
+    },
+    {
+      "epoch": 0.01581522103335046,
+      "grad_norm": 0.5310292840003967,
+      "learning_rate": 5.267857142857143e-05,
+      "loss": 1.2706,
+      "step": 177
+    },
+    {
+      "epoch": 0.015904572564612324,
+      "grad_norm": 0.5005099177360535,
+      "learning_rate": 5.297619047619048e-05,
+      "loss": 1.2482,
+      "step": 178
+    },
+    {
+      "epoch": 0.015993924095874194,
+      "grad_norm": 0.45003390312194824,
+      "learning_rate": 5.327380952380953e-05,
+      "loss": 1.25,
+      "step": 179
+    },
+    {
+      "epoch": 0.01608327562713606,
+      "grad_norm": 0.4997723400592804,
+      "learning_rate": 5.3571428571428575e-05,
+      "loss": 1.1774,
+      "step": 180
+    },
+    {
+      "epoch": 0.01617262715839793,
+      "grad_norm": 0.4725038409233093,
+      "learning_rate": 5.3869047619047616e-05,
+      "loss": 1.226,
+      "step": 181
+    },
+    {
+      "epoch": 0.016261978689659794,
+      "grad_norm": 0.47115814685821533,
+      "learning_rate": 5.4166666666666664e-05,
+      "loss": 1.2809,
+      "step": 182
+    },
+    {
+      "epoch": 0.01635133022092166,
+      "grad_norm": 0.4939117133617401,
+      "learning_rate": 5.446428571428571e-05,
+      "loss": 1.2195,
+      "step": 183
+    },
+    {
+      "epoch": 0.01644068175218353,
+      "grad_norm": 0.4947628974914551,
+      "learning_rate": 5.4761904761904766e-05,
+      "loss": 1.1385,
+      "step": 184
+    },
+    {
+      "epoch": 0.016530033283445395,
+      "grad_norm": 0.43742480874061584,
+      "learning_rate": 5.5059523809523814e-05,
+      "loss": 1.2191,
+      "step": 185
+    },
+    {
+      "epoch": 0.01661938481470726,
+      "grad_norm": 0.48737379908561707,
+      "learning_rate": 5.535714285714286e-05,
+      "loss": 1.2033,
+      "step": 186
+    },
+    {
+      "epoch": 0.01670873634596913,
+      "grad_norm": 0.46898481249809265,
+      "learning_rate": 5.565476190476191e-05,
+      "loss": 1.2672,
+      "step": 187
+    },
+    {
+      "epoch": 0.016798087877230996,
+      "grad_norm": 0.5089002251625061,
+      "learning_rate": 5.595238095238096e-05,
+      "loss": 1.1853,
+      "step": 188
+    },
+    {
+      "epoch": 0.01688743940849286,
+      "grad_norm": 0.5069641470909119,
+      "learning_rate": 5.6250000000000005e-05,
+      "loss": 1.1838,
+      "step": 189
+    },
+    {
+      "epoch": 0.01697679093975473,
+      "grad_norm": 0.48622408509254456,
+      "learning_rate": 5.6547619047619046e-05,
+      "loss": 1.2445,
+      "step": 190
+    },
+    {
+      "epoch": 0.017066142471016597,
+      "grad_norm": 0.48085153102874756,
+      "learning_rate": 5.6845238095238094e-05,
+      "loss": 1.1636,
+      "step": 191
+    },
+    {
+      "epoch": 0.017155494002278462,
+      "grad_norm": 0.4827207326889038,
+      "learning_rate": 5.714285714285714e-05,
+      "loss": 1.2002,
+      "step": 192
+    },
+    {
+      "epoch": 0.01724484553354033,
+      "grad_norm": 0.5341354608535767,
+      "learning_rate": 5.744047619047619e-05,
+      "loss": 1.2213,
+      "step": 193
+    },
+    {
+      "epoch": 0.017334197064802197,
+      "grad_norm": 0.4994199872016907,
+      "learning_rate": 5.773809523809524e-05,
+      "loss": 1.1627,
+      "step": 194
+    },
+    {
+      "epoch": 0.017423548596064067,
+      "grad_norm": 0.5125264525413513,
+      "learning_rate": 5.803571428571429e-05,
+      "loss": 1.2425,
+      "step": 195
+    },
+    {
+      "epoch": 0.017512900127325932,
+      "grad_norm": 0.534212589263916,
+      "learning_rate": 5.833333333333334e-05,
+      "loss": 1.1437,
+      "step": 196
+    },
+    {
+      "epoch": 0.017602251658587798,
+      "grad_norm": 0.4822109341621399,
+      "learning_rate": 5.863095238095239e-05,
+      "loss": 1.1876,
+      "step": 197
+    },
+    {
+      "epoch": 0.017691603189849667,
+      "grad_norm": 0.4595533311367035,
+      "learning_rate": 5.8928571428571435e-05,
+      "loss": 1.1951,
+      "step": 198
+    },
+    {
+      "epoch": 0.017780954721111533,
+      "grad_norm": 0.4773840606212616,
+      "learning_rate": 5.922619047619048e-05,
+      "loss": 1.2074,
+      "step": 199
+    },
+    {
+      "epoch": 0.0178703062523734,
+      "grad_norm": 0.4863610565662384,
+      "learning_rate": 5.9523809523809524e-05,
+      "loss": 1.1874,
+      "step": 200
+    },
+    {
+      "epoch": 0.017959657783635268,
+      "grad_norm": 0.5983034372329712,
+      "learning_rate": 5.982142857142857e-05,
+      "loss": 1.1794,
+      "step": 201
+    },
+    {
+      "epoch": 0.018049009314897134,
+      "grad_norm": 0.5169399380683899,
+      "learning_rate": 6.011904761904762e-05,
+      "loss": 1.25,
+      "step": 202
+    },
+    {
+      "epoch": 0.018138360846159,
+      "grad_norm": 0.49807149171829224,
+      "learning_rate": 6.041666666666667e-05,
+      "loss": 1.2069,
+      "step": 203
+    },
+    {
+      "epoch": 0.01822771237742087,
+      "grad_norm": 0.488290399312973,
+      "learning_rate": 6.0714285714285715e-05,
+      "loss": 1.2747,
+      "step": 204
+    },
+    {
+      "epoch": 0.018317063908682735,
+      "grad_norm": 0.49439024925231934,
+      "learning_rate": 6.101190476190477e-05,
+      "loss": 1.2401,
+      "step": 205
+    },
+    {
+      "epoch": 0.0184064154399446,
+      "grad_norm": 0.5511735081672668,
+      "learning_rate": 6.130952380952381e-05,
+      "loss": 1.1785,
+      "step": 206
+    },
+    {
+      "epoch": 0.01849576697120647,
+      "grad_norm": 0.5392850637435913,
+      "learning_rate": 6.160714285714286e-05,
+      "loss": 1.2753,
+      "step": 207
+    },
+    {
+      "epoch": 0.018585118502468335,
+      "grad_norm": 0.4930606484413147,
+      "learning_rate": 6.19047619047619e-05,
+      "loss": 1.2354,
+      "step": 208
+    },
+    {
+      "epoch": 0.018674470033730205,
+      "grad_norm": 0.49877017736434937,
+      "learning_rate": 6.220238095238095e-05,
+      "loss": 1.2199,
+      "step": 209
+    },
+    {
+      "epoch": 0.01876382156499207,
+      "grad_norm": 0.5193923711776733,
+      "learning_rate": 6.25e-05,
+      "loss": 1.1794,
+      "step": 210
+    },
+    {
+      "epoch": 0.018853173096253936,
+      "grad_norm": 0.5702071189880371,
+      "learning_rate": 6.279761904761905e-05,
+      "loss": 1.237,
+      "step": 211
+    },
+    {
+      "epoch": 0.018942524627515805,
+      "grad_norm": 0.542413055896759,
+      "learning_rate": 6.30952380952381e-05,
+      "loss": 1.2469,
+      "step": 212
+    },
+    {
+      "epoch": 0.01903187615877767,
+      "grad_norm": 0.518735945224762,
+      "learning_rate": 6.339285714285714e-05,
+      "loss": 1.1726,
+      "step": 213
+    },
+    {
+      "epoch": 0.019121227690039537,
+      "grad_norm": 0.49173060059547424,
+      "learning_rate": 6.369047619047619e-05,
+      "loss": 1.1974,
+      "step": 214
+    },
+    {
+      "epoch": 0.019210579221301406,
+      "grad_norm": 0.46574926376342773,
+      "learning_rate": 6.398809523809524e-05,
+      "loss": 1.2094,
+      "step": 215
+    },
+    {
+      "epoch": 0.019299930752563272,
+      "grad_norm": 0.5576440691947937,
+      "learning_rate": 6.428571428571429e-05,
+      "loss": 1.1892,
+      "step": 216
+    },
+    {
+      "epoch": 0.019389282283825138,
+      "grad_norm": 0.527204692363739,
+      "learning_rate": 6.458333333333334e-05,
+      "loss": 1.2032,
+      "step": 217
+    },
+    {
+      "epoch": 0.019478633815087007,
+      "grad_norm": 0.5345803499221802,
+      "learning_rate": 6.488095238095238e-05,
+      "loss": 1.1977,
+      "step": 218
+    },
+    {
+      "epoch": 0.019567985346348873,
+      "grad_norm": 0.5545653700828552,
+      "learning_rate": 6.517857142857143e-05,
+      "loss": 1.1988,
+      "step": 219
+    },
+    {
+      "epoch": 0.01965733687761074,
+      "grad_norm": 0.5376014113426208,
+      "learning_rate": 6.547619047619048e-05,
+      "loss": 1.1331,
+      "step": 220
+    },
+    {
+      "epoch": 0.019746688408872608,
+      "grad_norm": 0.5965554714202881,
+      "learning_rate": 6.577380952380953e-05,
+      "loss": 1.1333,
+      "step": 221
+    },
+    {
+      "epoch": 0.019836039940134473,
+      "grad_norm": 0.48786482214927673,
+      "learning_rate": 6.607142857142857e-05,
+      "loss": 1.2119,
+      "step": 222
+    },
+    {
+      "epoch": 0.019925391471396343,
+      "grad_norm": 0.5010102391242981,
+      "learning_rate": 6.636904761904762e-05,
+      "loss": 1.1576,
+      "step": 223
+    },
+    {
+      "epoch": 0.02001474300265821,
+      "grad_norm": 0.5798513293266296,
+      "learning_rate": 6.666666666666667e-05,
+      "loss": 1.2731,
+      "step": 224
+    },
+    {
+      "epoch": 0.020104094533920074,
+      "grad_norm": 0.600602388381958,
+      "learning_rate": 6.696428571428572e-05,
+      "loss": 1.1185,
+      "step": 225
+    },
+    {
+      "epoch": 0.020193446065181943,
+      "grad_norm": 0.5265418291091919,
+      "learning_rate": 6.726190476190477e-05,
+      "loss": 1.2328,
+      "step": 226
+    },
+    {
+      "epoch": 0.02028279759644381,
+      "grad_norm": 0.6028567552566528,
+      "learning_rate": 6.755952380952381e-05,
+      "loss": 1.222,
+      "step": 227
+    },
+    {
+      "epoch": 0.020372149127705675,
+      "grad_norm": 0.5283799171447754,
+      "learning_rate": 6.785714285714286e-05,
+      "loss": 1.2209,
+      "step": 228
+    },
+    {
+      "epoch": 0.020461500658967544,
+      "grad_norm": 0.5363395810127258,
+      "learning_rate": 6.815476190476191e-05,
+      "loss": 1.1309,
+      "step": 229
+    },
+    {
+      "epoch": 0.02055085219022941,
+      "grad_norm": 0.5541881918907166,
+      "learning_rate": 6.845238095238096e-05,
+      "loss": 1.1629,
+      "step": 230
+    },
+    {
+      "epoch": 0.020640203721491276,
+      "grad_norm": 0.5489634275436401,
+      "learning_rate": 6.875e-05,
+      "loss": 1.148,
+      "step": 231
+    },
+    {
+      "epoch": 0.020729555252753145,
+      "grad_norm": 0.582408607006073,
+      "learning_rate": 6.904761904761905e-05,
+      "loss": 1.1505,
+      "step": 232
+    },
+    {
+      "epoch": 0.02081890678401501,
+      "grad_norm": 0.5389354825019836,
+      "learning_rate": 6.93452380952381e-05,
+      "loss": 1.1644,
+      "step": 233
+    },
+    {
+      "epoch": 0.020908258315276877,
+      "grad_norm": 0.6120844483375549,
+      "learning_rate": 6.964285714285715e-05,
+      "loss": 1.1522,
+      "step": 234
+    },
+    {
+      "epoch": 0.020997609846538746,
+      "grad_norm": 0.5017001628875732,
+      "learning_rate": 6.99404761904762e-05,
+      "loss": 1.2153,
+      "step": 235
+    },
+    {
+      "epoch": 0.02108696137780061,
+      "grad_norm": 0.5250133275985718,
+      "learning_rate": 7.023809523809524e-05,
+      "loss": 1.2756,
+      "step": 236
+    },
+    {
+      "epoch": 0.02117631290906248,
+      "grad_norm": 0.5131645202636719,
+      "learning_rate": 7.053571428571429e-05,
+      "loss": 1.1909,
+      "step": 237
+    },
+    {
+      "epoch": 0.021265664440324347,
+      "grad_norm": 0.5468138456344604,
+      "learning_rate": 7.083333333333334e-05,
+      "loss": 1.1209,
+      "step": 238
+    },
+    {
+      "epoch": 0.021355015971586212,
+      "grad_norm": 0.5460432171821594,
+      "learning_rate": 7.113095238095239e-05,
+      "loss": 1.2472,
+      "step": 239
+    },
+    {
+      "epoch": 0.02144436750284808,
+      "grad_norm": 0.5693673491477966,
+      "learning_rate": 7.142857142857143e-05,
+      "loss": 1.1466,
+      "step": 240
+    },
+    {
+      "epoch": 0.021533719034109947,
+      "grad_norm": 0.5243647694587708,
+      "learning_rate": 7.172619047619048e-05,
+      "loss": 1.1108,
+      "step": 241
+    },
+    {
+      "epoch": 0.021623070565371813,
+      "grad_norm": 0.51932692527771,
+      "learning_rate": 7.202380952380953e-05,
+      "loss": 1.2149,
+      "step": 242
+    },
+    {
+      "epoch": 0.021712422096633682,
+      "grad_norm": 0.5603271722793579,
+      "learning_rate": 7.232142857142858e-05,
+      "loss": 1.1534,
+      "step": 243
+    },
+    {
+      "epoch": 0.021801773627895548,
+      "grad_norm": 0.5533414483070374,
+      "learning_rate": 7.261904761904762e-05,
+      "loss": 1.1692,
+      "step": 244
+    },
+    {
+      "epoch": 0.021891125159157414,
+      "grad_norm": 0.5640945434570312,
+      "learning_rate": 7.291666666666667e-05,
+      "loss": 1.191,
+      "step": 245
+    },
+    {
+      "epoch": 0.021980476690419283,
+      "grad_norm": 0.565290629863739,
+      "learning_rate": 7.321428571428571e-05,
+      "loss": 1.1163,
+      "step": 246
+    },
+    {
+      "epoch": 0.02206982822168115,
+      "grad_norm": 0.5642583966255188,
+      "learning_rate": 7.351190476190477e-05,
+      "loss": 1.1447,
+      "step": 247
+    },
+    {
+      "epoch": 0.022159179752943015,
+      "grad_norm": 0.5096793174743652,
+      "learning_rate": 7.380952380952382e-05,
+      "loss": 1.1812,
+      "step": 248
+    },
+    {
+      "epoch": 0.022248531284204884,
+      "grad_norm": 0.49320727586746216,
+      "learning_rate": 7.410714285714286e-05,
+      "loss": 1.1238,
+      "step": 249
+    },
+    {
+      "epoch": 0.02233788281546675,
+      "grad_norm": 0.5757387280464172,
+      "learning_rate": 7.440476190476191e-05,
+      "loss": 1.2047,
+      "step": 250
+    },
+    {
+      "epoch": 0.02242723434672862,
+      "grad_norm": 0.5943542718887329,
+      "learning_rate": 7.470238095238096e-05,
+      "loss": 1.146,
+      "step": 251
+    },
+    {
+      "epoch": 0.022516585877990485,
+      "grad_norm": 0.5858287215232849,
+      "learning_rate": 7.500000000000001e-05,
+      "loss": 1.1685,
+      "step": 252
+    },
+    {
+      "epoch": 0.02260593740925235,
+      "grad_norm": 0.4847318232059479,
+      "learning_rate": 7.529761904761905e-05,
+      "loss": 1.1975,
+      "step": 253
+    },
+    {
+      "epoch": 0.02269528894051422,
+      "grad_norm": 0.5576103925704956,
+      "learning_rate": 7.55952380952381e-05,
+      "loss": 1.158,
+      "step": 254
+    },
+    {
+      "epoch": 0.022784640471776085,
+      "grad_norm": 0.5339264869689941,
+      "learning_rate": 7.589285714285714e-05,
+      "loss": 1.1694,
+      "step": 255
+    },
+    {
+      "epoch": 0.02287399200303795,
+      "grad_norm": 0.5801980495452881,
+      "learning_rate": 7.619047619047618e-05,
+      "loss": 1.1646,
+      "step": 256
+    },
+    {
+      "epoch": 0.02296334353429982,
+      "grad_norm": 0.5421757698059082,
+      "learning_rate": 7.648809523809523e-05,
+      "loss": 1.1546,
+      "step": 257
+    },
+    {
+      "epoch": 0.023052695065561686,
+      "grad_norm": 0.546415388584137,
+      "learning_rate": 7.67857142857143e-05,
+      "loss": 1.2262,
+      "step": 258
+    },
+    {
+      "epoch": 0.023142046596823552,
+      "grad_norm": 0.584190309047699,
+      "learning_rate": 7.708333333333334e-05,
+      "loss": 1.1528,
+      "step": 259
+    },
+    {
+      "epoch": 0.02323139812808542,
+      "grad_norm": 0.5467146635055542,
+      "learning_rate": 7.738095238095239e-05,
+      "loss": 1.1452,
+      "step": 260
+    },
+    {
+      "epoch": 0.023320749659347287,
+      "grad_norm": 0.5554835796356201,
+      "learning_rate": 7.767857142857144e-05,
+      "loss": 1.1936,
+      "step": 261
+    },
+    {
+      "epoch": 0.023410101190609153,
+      "grad_norm": 0.5491194725036621,
+      "learning_rate": 7.797619047619048e-05,
+      "loss": 1.1152,
+      "step": 262
+    },
+    {
+      "epoch": 0.023499452721871022,
+      "grad_norm": 0.5814348459243774,
+      "learning_rate": 7.827380952380953e-05,
+      "loss": 1.1432,
+      "step": 263
+    },
+    {
+      "epoch": 0.023588804253132888,
+      "grad_norm": 0.598092257976532,
+      "learning_rate": 7.857142857142858e-05,
+      "loss": 1.1661,
+      "step": 264
+    },
+    {
+      "epoch": 0.023678155784394753,
+      "grad_norm": 0.6232405304908752,
+      "learning_rate": 7.886904761904761e-05,
+      "loss": 1.1545,
+      "step": 265
+    },
+    {
+      "epoch": 0.023767507315656623,
+      "grad_norm": 0.53800368309021,
+      "learning_rate": 7.916666666666666e-05,
+      "loss": 1.2058,
+      "step": 266
+    },
+    {
+      "epoch": 0.02385685884691849,
+      "grad_norm": 0.5912578105926514,
+      "learning_rate": 7.946428571428571e-05,
+      "loss": 1.1638,
+      "step": 267
+    },
+    {
+      "epoch": 0.023946210378180358,
+      "grad_norm": 0.6035777926445007,
+      "learning_rate": 7.976190476190477e-05,
+      "loss": 1.1325,
+      "step": 268
+    },
+    {
+      "epoch": 0.024035561909442223,
+      "grad_norm": 0.5701507925987244,
+      "learning_rate": 8.005952380952382e-05,
+      "loss": 1.1693,
+      "step": 269
+    },
+    {
+      "epoch": 0.02412491344070409,
+      "grad_norm": 0.9521252512931824,
+      "learning_rate": 8.035714285714287e-05,
+      "loss": 1.1926,
+      "step": 270
+    },
+    {
+      "epoch": 0.02421426497196596,
+      "grad_norm": 0.5579119324684143,
+      "learning_rate": 8.065476190476191e-05,
+      "loss": 1.2114,
+      "step": 271
+    },
+    {
+      "epoch": 0.024303616503227824,
+      "grad_norm": 0.5646568536758423,
+      "learning_rate": 8.095238095238096e-05,
+      "loss": 1.1351,
+      "step": 272
+    },
+    {
+      "epoch": 0.02439296803448969,
+      "grad_norm": 0.5492066740989685,
+      "learning_rate": 8.125000000000001e-05,
+      "loss": 1.187,
+      "step": 273
+    },
+    {
+      "epoch": 0.02448231956575156,
+      "grad_norm": 0.5730810761451721,
+      "learning_rate": 8.154761904761904e-05,
+      "loss": 1.1687,
+      "step": 274
+    },
+    {
+      "epoch": 0.024571671097013425,
+      "grad_norm": 0.5157971978187561,
+      "learning_rate": 8.184523809523809e-05,
+      "loss": 1.1312,
+      "step": 275
+    },
+    {
+      "epoch": 0.02466102262827529,
+      "grad_norm": 0.5865328311920166,
+      "learning_rate": 8.214285714285714e-05,
+      "loss": 1.207,
+      "step": 276
+    },
+    {
+      "epoch": 0.02475037415953716,
+      "grad_norm": 0.47492775321006775,
+      "learning_rate": 8.244047619047619e-05,
+      "loss": 1.1525,
+      "step": 277
+    },
+    {
+      "epoch": 0.024839725690799026,
+      "grad_norm": 0.5046519041061401,
+      "learning_rate": 8.273809523809524e-05,
+      "loss": 1.1959,
+      "step": 278
+    },
+    {
+      "epoch": 0.02492907722206089,
+      "grad_norm": 0.5498790144920349,
+      "learning_rate": 8.30357142857143e-05,
+      "loss": 1.1969,
+      "step": 279
+    },
+    {
+      "epoch": 0.02501842875332276,
+      "grad_norm": 0.5528784394264221,
+      "learning_rate": 8.333333333333334e-05,
+      "loss": 1.0772,
+      "step": 280
+    },
+    {
+      "epoch": 0.025107780284584626,
+      "grad_norm": 0.5229126811027527,
+      "learning_rate": 8.363095238095239e-05,
+      "loss": 1.1485,
+      "step": 281
+    },
+    {
+      "epoch": 0.025197131815846496,
+      "grad_norm": 0.5191675424575806,
+      "learning_rate": 8.392857142857144e-05,
+      "loss": 1.1568,
+      "step": 282
+    },
+    {
+      "epoch": 0.02528648334710836,
+      "grad_norm": 0.5272664427757263,
+      "learning_rate": 8.422619047619049e-05,
+      "loss": 1.1794,
+      "step": 283
+    },
+    {
+      "epoch": 0.025375834878370227,
+      "grad_norm": 0.5842853784561157,
+      "learning_rate": 8.452380952380952e-05,
+      "loss": 1.101,
+      "step": 284
+    },
+    {
+      "epoch": 0.025465186409632096,
+      "grad_norm": 0.501756489276886,
+      "learning_rate": 8.482142857142857e-05,
+      "loss": 1.1639,
+      "step": 285
+    },
+    {
+      "epoch": 0.025554537940893962,
+      "grad_norm": 0.491220623254776,
+      "learning_rate": 8.511904761904762e-05,
+      "loss": 1.2092,
+      "step": 286
+    },
+    {
+      "epoch": 0.025643889472155828,
+      "grad_norm": 0.5032030940055847,
+      "learning_rate": 8.541666666666666e-05,
+      "loss": 1.181,
+      "step": 287
+    },
+    {
+      "epoch": 0.025733241003417697,
+      "grad_norm": 0.5295215249061584,
+      "learning_rate": 8.571428571428571e-05,
+      "loss": 1.1861,
+      "step": 288
+    },
+    {
+      "epoch": 0.025822592534679563,
+      "grad_norm": 0.5146911144256592,
+      "learning_rate": 8.601190476190477e-05,
+      "loss": 1.1854,
+      "step": 289
+    },
+    {
+      "epoch": 0.02591194406594143,
+      "grad_norm": 0.5277708172798157,
+      "learning_rate": 8.630952380952382e-05,
+      "loss": 1.1318,
+      "step": 290
+    },
+    {
+      "epoch": 0.026001295597203298,
+      "grad_norm": 0.5158389210700989,
+      "learning_rate": 8.660714285714287e-05,
+      "loss": 1.1603,
+      "step": 291
+    },
+    {
+      "epoch": 0.026090647128465164,
+      "grad_norm": 0.4982542395591736,
+      "learning_rate": 8.690476190476192e-05,
+      "loss": 1.185,
+      "step": 292
+    },
+    {
+      "epoch": 0.02617999865972703,
+      "grad_norm": 0.5195929408073425,
+      "learning_rate": 8.720238095238095e-05,
+      "loss": 1.2306,
+      "step": 293
+    },
+    {
+      "epoch": 0.0262693501909889,
+      "grad_norm": 0.48286932706832886,
+      "learning_rate": 8.75e-05,
+      "loss": 1.1665,
+      "step": 294
+    },
+    {
+      "epoch": 0.026358701722250764,
+      "grad_norm": 0.5617235898971558,
+      "learning_rate": 8.779761904761905e-05,
+      "loss": 1.1484,
+      "step": 295
+    },
+    {
+      "epoch": 0.026448053253512634,
+      "grad_norm": 0.5541561841964722,
+      "learning_rate": 8.80952380952381e-05,
+      "loss": 0.9718,
+      "step": 296
+    },
+    {
+      "epoch": 0.0265374047847745,
+      "grad_norm": 0.47326454520225525,
+      "learning_rate": 8.839285714285714e-05,
+      "loss": 1.1289,
+      "step": 297
+    },
+    {
+      "epoch": 0.026626756316036365,
+      "grad_norm": 0.5681378841400146,
+      "learning_rate": 8.869047619047619e-05,
+      "loss": 1.1569,
+      "step": 298
+    },
+    {
+      "epoch": 0.026716107847298234,
+      "grad_norm": 0.5514600276947021,
+      "learning_rate": 8.898809523809524e-05,
+      "loss": 1.1495,
+      "step": 299
+    },
+    {
+      "epoch": 0.0268054593785601,
+      "grad_norm": 0.5337786078453064,
+      "learning_rate": 8.92857142857143e-05,
+      "loss": 1.2366,
+      "step": 300
+    },
+    {
+      "epoch": 0.026894810909821966,
+      "grad_norm": 0.5156247019767761,
+      "learning_rate": 8.958333333333335e-05,
+      "loss": 1.1969,
+      "step": 301
+    },
+    {
+      "epoch": 0.026984162441083835,
+      "grad_norm": 0.5061919689178467,
+      "learning_rate": 8.988095238095238e-05,
+      "loss": 1.1633,
+      "step": 302
+    },
+    {
+      "epoch": 0.0270735139723457,
+      "grad_norm": 0.6192631721496582,
+      "learning_rate": 9.017857142857143e-05,
+      "loss": 1.1552,
+      "step": 303
+    },
+    {
+      "epoch": 0.027162865503607567,
+      "grad_norm": 0.5444059371948242,
+      "learning_rate": 9.047619047619048e-05,
+      "loss": 1.1532,
+      "step": 304
+    },
+    {
+      "epoch": 0.027252217034869436,
+      "grad_norm": 0.49790215492248535,
+      "learning_rate": 9.077380952380952e-05,
+      "loss": 1.1339,
+      "step": 305
+    },
+    {
+      "epoch": 0.0273415685661313,
+      "grad_norm": 0.5882278680801392,
+      "learning_rate": 9.107142857142857e-05,
+      "loss": 1.1195,
+      "step": 306
+    },
+    {
+      "epoch": 0.027430920097393167,
+      "grad_norm": 0.5879011154174805,
+      "learning_rate": 9.136904761904762e-05,
+      "loss": 1.1341,
+      "step": 307
+    },
+    {
+      "epoch": 0.027520271628655037,
+      "grad_norm": 0.5458969473838806,
+      "learning_rate": 9.166666666666667e-05,
+      "loss": 1.1498,
+      "step": 308
+    },
+    {
+      "epoch": 0.027609623159916902,
+      "grad_norm": 0.5539296865463257,
+      "learning_rate": 9.196428571428572e-05,
+      "loss": 1.1261,
+      "step": 309
+    },
+    {
+      "epoch": 0.02769897469117877,
+      "grad_norm": 0.508406400680542,
+      "learning_rate": 9.226190476190478e-05,
+      "loss": 1.1318,
+      "step": 310
+    },
+    {
+      "epoch": 0.027788326222440637,
+      "grad_norm": 0.5266230702400208,
+      "learning_rate": 9.255952380952382e-05,
+      "loss": 1.1463,
+      "step": 311
+    },
+    {
+      "epoch": 0.027877677753702503,
+      "grad_norm": 0.5170016288757324,
+      "learning_rate": 9.285714285714286e-05,
+      "loss": 1.1736,
+      "step": 312
+    },
+    {
+      "epoch": 0.027967029284964372,
+      "grad_norm": 0.5622848272323608,
+      "learning_rate": 9.31547619047619e-05,
+      "loss": 1.2113,
+      "step": 313
+    },
+    {
+      "epoch": 0.028056380816226238,
+      "grad_norm": 0.5831321477890015,
+      "learning_rate": 9.345238095238095e-05,
+      "loss": 1.1497,
+      "step": 314
+    },
+    {
+      "epoch": 0.028145732347488104,
+      "grad_norm": 0.46541884541511536,
+      "learning_rate": 9.375e-05,
+      "loss": 1.146,
+      "step": 315
+    },
+    {
+      "epoch": 0.028235083878749973,
+      "grad_norm": 0.5889435410499573,
+      "learning_rate": 9.404761904761905e-05,
+      "loss": 1.1694,
+      "step": 316
+    },
+    {
+      "epoch": 0.02832443541001184,
+      "grad_norm": 0.49173977971076965,
+      "learning_rate": 9.43452380952381e-05,
+      "loss": 1.1854,
+      "step": 317
+    },
+    {
+      "epoch": 0.028413786941273705,
+      "grad_norm": 0.6178017258644104,
+      "learning_rate": 9.464285714285715e-05,
+      "loss": 1.1072,
+      "step": 318
+    },
+    {
+      "epoch": 0.028503138472535574,
+      "grad_norm": 0.5283975005149841,
+      "learning_rate": 9.494047619047619e-05,
+      "loss": 1.0578,
+      "step": 319
+    },
+    {
+      "epoch": 0.02859249000379744,
+      "grad_norm": 0.5168601870536804,
+      "learning_rate": 9.523809523809524e-05,
+      "loss": 1.1003,
+      "step": 320
+    },
+    {
+      "epoch": 0.028681841535059305,
+      "grad_norm": 0.4885217845439911,
+      "learning_rate": 9.553571428571429e-05,
+      "loss": 1.1953,
+      "step": 321
+    },
+    {
+      "epoch": 0.028771193066321175,
+      "grad_norm": 0.4856249690055847,
+      "learning_rate": 9.583333333333334e-05,
+      "loss": 1.2207,
+      "step": 322
+    },
+    {
+      "epoch": 0.02886054459758304,
+      "grad_norm": 0.5465936660766602,
+      "learning_rate": 9.613095238095238e-05,
+      "loss": 1.1668,
+      "step": 323
+    },
+    {
+      "epoch": 0.02894989612884491,
+      "grad_norm": 0.5614081025123596,
+      "learning_rate": 9.642857142857143e-05,
+      "loss": 1.1091,
+      "step": 324
+    },
+    {
+      "epoch": 0.029039247660106775,
+      "grad_norm": 0.5728946328163147,
+      "learning_rate": 9.672619047619048e-05,
+      "loss": 1.1232,
+      "step": 325
+    },
+    {
+      "epoch": 0.02912859919136864,
+      "grad_norm": 0.6166178584098816,
+      "learning_rate": 9.702380952380953e-05,
+      "loss": 1.1756,
+      "step": 326
+    },
+    {
+      "epoch": 0.02921795072263051,
+      "grad_norm": 0.5272330641746521,
+      "learning_rate": 9.732142857142858e-05,
+      "loss": 1.1131,
+      "step": 327
+    },
+    {
+      "epoch": 0.029307302253892376,
+      "grad_norm": 0.4859834909439087,
+      "learning_rate": 9.761904761904762e-05,
+      "loss": 1.2047,
+      "step": 328
+    },
+    {
+      "epoch": 0.029396653785154242,
+      "grad_norm": 0.5186814069747925,
+      "learning_rate": 9.791666666666667e-05,
+      "loss": 1.0997,
+      "step": 329
+    },
+    {
+      "epoch": 0.02948600531641611,
+      "grad_norm": 0.5721623301506042,
+      "learning_rate": 9.821428571428572e-05,
+      "loss": 1.0986,
+      "step": 330
+    },
+    {
+      "epoch": 0.029575356847677977,
+      "grad_norm": 0.5345954895019531,
+      "learning_rate": 9.851190476190477e-05,
+      "loss": 1.1971,
+      "step": 331
+    },
+    {
+      "epoch": 0.029664708378939843,
+      "grad_norm": 0.5651283264160156,
+      "learning_rate": 9.880952380952381e-05,
+      "loss": 1.1753,
+      "step": 332
+    },
+    {
+      "epoch": 0.029754059910201712,
+      "grad_norm": 0.6897709369659424,
+      "learning_rate": 9.910714285714286e-05,
+      "loss": 1.1662,
+      "step": 333
+    },
+    {
+      "epoch": 0.029843411441463578,
+      "grad_norm": 0.5349520444869995,
+      "learning_rate": 9.940476190476191e-05,
+      "loss": 1.1446,
+      "step": 334
+    },
+    {
+      "epoch": 0.029932762972725444,
+      "grad_norm": 0.5659092664718628,
+      "learning_rate": 9.970238095238096e-05,
+      "loss": 1.1273,
+      "step": 335
+    },
+    {
+      "epoch": 0.030022114503987313,
+      "grad_norm": 0.4331722557544708,
+      "learning_rate": 0.0001,
+      "loss": 1.192,
+      "step": 336
+    },
+    {
+      "epoch": 0.03011146603524918,
+      "grad_norm": 0.5557327270507812,
+      "learning_rate": 9.999999790598352e-05,
+      "loss": 1.1184,
+      "step": 337
+    },
+    {
+      "epoch": 0.030200817566511048,
+      "grad_norm": 0.5717477798461914,
+      "learning_rate": 9.999999162393425e-05,
+      "loss": 1.1695,
+      "step": 338
+    },
+    {
+      "epoch": 0.030290169097772913,
+      "grad_norm": 0.6119928359985352,
+      "learning_rate": 9.999998115385273e-05,
+      "loss": 1.1202,
+      "step": 339
+    },
+    {
+      "epoch": 0.03037952062903478,
+      "grad_norm": 0.5220287442207336,
+      "learning_rate": 9.999996649573982e-05,
+      "loss": 1.0898,
+      "step": 340
+    },
+    {
+      "epoch": 0.03046887216029665,
+      "grad_norm": 0.5707947015762329,
+      "learning_rate": 9.999994764959675e-05,
+      "loss": 1.1645,
+      "step": 341
+    },
+    {
+      "epoch": 0.030558223691558514,
+      "grad_norm": 0.5686008930206299,
+      "learning_rate": 9.99999246154251e-05,
+      "loss": 1.1141,
+      "step": 342
+    },
+    {
+      "epoch": 0.03064757522282038,
+      "grad_norm": 0.551203727722168,
+      "learning_rate": 9.999989739322682e-05,
+      "loss": 1.1881,
+      "step": 343
+    },
+    {
+      "epoch": 0.03073692675408225,
+      "grad_norm": 0.5531251430511475,
+      "learning_rate": 9.999986598300417e-05,
+      "loss": 1.0881,
+      "step": 344
+    },
+    {
+      "epoch": 0.030826278285344115,
+      "grad_norm": 0.5076424479484558,
+      "learning_rate": 9.999983038475978e-05,
+      "loss": 1.1175,
+      "step": 345
+    },
+    {
+      "epoch": 0.03091562981660598,
+      "grad_norm": 0.5364487171173096,
+      "learning_rate": 9.999979059849662e-05,
+      "loss": 1.1934,
+      "step": 346
+    },
+    {
+      "epoch": 0.03100498134786785,
+      "grad_norm": 0.557384729385376,
+      "learning_rate": 9.999974662421805e-05,
+      "loss": 1.1748,
+      "step": 347
+    },
+    {
+      "epoch": 0.031094332879129716,
+      "grad_norm": 0.5250005125999451,
+      "learning_rate": 9.999969846192774e-05,
+      "loss": 1.1338,
+      "step": 348
+    },
+    {
+      "epoch": 0.03118368441039158,
+      "grad_norm": 0.4806252121925354,
+      "learning_rate": 9.999964611162974e-05,
+      "loss": 1.1639,
+      "step": 349
+    },
+    {
+      "epoch": 0.03127303594165345,
+      "grad_norm": 0.520580530166626,
+      "learning_rate": 9.99995895733284e-05,
+      "loss": 1.1534,
+      "step": 350
+    },
+    {
+      "epoch": 0.03136238747291532,
+      "grad_norm": 0.5017713308334351,
+      "learning_rate": 9.999952884702848e-05,
+      "loss": 1.1473,
+      "step": 351
+    },
+    {
+      "epoch": 0.031451739004177186,
+      "grad_norm": 0.5286357402801514,
+      "learning_rate": 9.999946393273506e-05,
+      "loss": 1.179,
+      "step": 352
+    },
+    {
+      "epoch": 0.03154109053543905,
+      "grad_norm": 0.5509297251701355,
+      "learning_rate": 9.999939483045359e-05,
+      "loss": 1.1269,
+      "step": 353
+    },
+    {
+      "epoch": 0.03163044206670092,
+      "grad_norm": 0.4697697162628174,
+      "learning_rate": 9.999932154018983e-05,
+      "loss": 1.1569,
+      "step": 354
+    },
+    {
+      "epoch": 0.031719793597962787,
+      "grad_norm": 0.4342000484466553,
+      "learning_rate": 9.999924406194996e-05,
+      "loss": 1.1648,
+      "step": 355
+    },
+    {
+      "epoch": 0.03180914512922465,
+      "grad_norm": 0.5409436225891113,
+      "learning_rate": 9.999916239574043e-05,
+      "loss": 1.0954,
+      "step": 356
+    },
+    {
+      "epoch": 0.03189849666048652,
+      "grad_norm": 0.529708981513977,
+      "learning_rate": 9.99990765415681e-05,
+      "loss": 1.1114,
+      "step": 357
+    },
+    {
+      "epoch": 0.03198784819174839,
+      "grad_norm": 0.5057774186134338,
+      "learning_rate": 9.999898649944016e-05,
+      "loss": 1.1176,
+      "step": 358
+    },
+    {
+      "epoch": 0.032077199723010257,
+      "grad_norm": 0.4980236291885376,
+      "learning_rate": 9.999889226936415e-05,
+      "loss": 1.1574,
+      "step": 359
+    },
+    {
+      "epoch": 0.03216655125427212,
+      "grad_norm": 0.49295690655708313,
+      "learning_rate": 9.999879385134797e-05,
+      "loss": 1.1922,
+      "step": 360
+    },
+    {
+      "epoch": 0.03225590278553399,
+      "grad_norm": 0.4822183847427368,
+      "learning_rate": 9.999869124539984e-05,
+      "loss": 1.0773,
+      "step": 361
+    },
+    {
+      "epoch": 0.03234525431679586,
+      "grad_norm": 0.5717249512672424,
+      "learning_rate": 9.999858445152839e-05,
+      "loss": 1.1254,
+      "step": 362
+    },
+    {
+      "epoch": 0.03243460584805772,
+      "grad_norm": 0.49593716859817505,
+      "learning_rate": 9.999847346974253e-05,
+      "loss": 1.1105,
+      "step": 363
+    },
+    {
+      "epoch": 0.03252395737931959,
+      "grad_norm": 0.5179683566093445,
+      "learning_rate": 9.999835830005158e-05,
+      "loss": 1.0686,
+      "step": 364
+    },
+    {
+      "epoch": 0.03261330891058146,
+      "grad_norm": 0.5655368566513062,
+      "learning_rate": 9.999823894246517e-05,
+      "loss": 1.0889,
+      "step": 365
+    },
+    {
+      "epoch": 0.03270266044184332,
+      "grad_norm": 0.49578356742858887,
+      "learning_rate": 9.999811539699331e-05,
+      "loss": 1.1124,
+      "step": 366
+    },
+    {
+      "epoch": 0.03279201197310519,
+      "grad_norm": 0.5003335475921631,
+      "learning_rate": 9.999798766364634e-05,
+      "loss": 1.1669,
+      "step": 367
+    },
+    {
+      "epoch": 0.03288136350436706,
+      "grad_norm": 0.5258259177207947,
+      "learning_rate": 9.999785574243496e-05,
+      "loss": 1.1333,
+      "step": 368
+    },
+    {
+      "epoch": 0.03297071503562892,
+      "grad_norm": 0.5132459998130798,
+      "learning_rate": 9.999771963337024e-05,
+      "loss": 1.1795,
+      "step": 369
+    },
+    {
+      "epoch": 0.03306006656689079,
+      "grad_norm": 0.48852595686912537,
+      "learning_rate": 9.999757933646354e-05,
+      "loss": 1.127,
+      "step": 370
+    },
+    {
+      "epoch": 0.03314941809815266,
+      "grad_norm": 0.530531108379364,
+      "learning_rate": 9.999743485172666e-05,
+      "loss": 1.1424,
+      "step": 371
+    },
+    {
+      "epoch": 0.03323876962941452,
+      "grad_norm": 0.45526745915412903,
+      "learning_rate": 9.999728617917165e-05,
+      "loss": 1.213,
+      "step": 372
+    },
+    {
+      "epoch": 0.03332812116067639,
+      "grad_norm": 0.5421516299247742,
+      "learning_rate": 9.9997133318811e-05,
+      "loss": 1.1793,
+      "step": 373
+    },
+    {
+      "epoch": 0.03341747269193826,
+      "grad_norm": 0.5866374373435974,
+      "learning_rate": 9.999697627065752e-05,
+      "loss": 1.0692,
+      "step": 374
+    },
+    {
+      "epoch": 0.03350682422320012,
+      "grad_norm": 0.5176992416381836,
+      "learning_rate": 9.999681503472433e-05,
+      "loss": 1.1502,
+      "step": 375
+    },
+    {
+      "epoch": 0.03359617575446199,
+      "grad_norm": 0.5438038110733032,
+      "learning_rate": 9.999664961102495e-05,
+      "loss": 1.1342,
+      "step": 376
+    },
+    {
+      "epoch": 0.03368552728572386,
+      "grad_norm": 0.5158547759056091,
+      "learning_rate": 9.999647999957325e-05,
+      "loss": 1.1954,
+      "step": 377
+    },
+    {
+      "epoch": 0.03377487881698572,
+      "grad_norm": 0.46927282214164734,
+      "learning_rate": 9.999630620038343e-05,
+      "loss": 1.1292,
+      "step": 378
+    },
+    {
+      "epoch": 0.03386423034824759,
+      "grad_norm": 0.5032052397727966,
+      "learning_rate": 9.999612821347003e-05,
+      "loss": 1.108,
+      "step": 379
+    },
+    {
+      "epoch": 0.03395358187950946,
+      "grad_norm": 0.48456865549087524,
+      "learning_rate": 9.999594603884798e-05,
+      "loss": 1.1719,
+      "step": 380
+    },
+    {
+      "epoch": 0.034042933410771324,
+      "grad_norm": 0.49805065989494324,
+      "learning_rate": 9.999575967653252e-05,
+      "loss": 1.2259,
+      "step": 381
+    },
+    {
+      "epoch": 0.03413228494203319,
+      "grad_norm": 0.5582330226898193,
+      "learning_rate": 9.999556912653929e-05,
+      "loss": 1.2235,
+      "step": 382
+    },
+    {
+      "epoch": 0.03422163647329506,
+      "grad_norm": 0.5083693265914917,
+      "learning_rate": 9.999537438888423e-05,
+      "loss": 1.1593,
+      "step": 383
+    },
+    {
+      "epoch": 0.034310988004556925,
+      "grad_norm": 0.49797725677490234,
+      "learning_rate": 9.999517546358364e-05,
+      "loss": 1.1255,
+      "step": 384
+    },
+    {
+      "epoch": 0.034400339535818794,
+      "grad_norm": 0.5025597214698792,
+      "learning_rate": 9.999497235065418e-05,
+      "loss": 1.093,
+      "step": 385
+    },
+    {
+      "epoch": 0.03448969106708066,
+      "grad_norm": 0.49305295944213867,
+      "learning_rate": 9.999476505011289e-05,
+      "loss": 1.1264,
+      "step": 386
+    },
+    {
+      "epoch": 0.034579042598342526,
+      "grad_norm": 0.46636465191841125,
+      "learning_rate": 9.999455356197713e-05,
+      "loss": 1.0769,
+      "step": 387
+    },
+    {
+      "epoch": 0.034668394129604395,
+      "grad_norm": 0.5070281028747559,
+      "learning_rate": 9.999433788626461e-05,
+      "loss": 1.0925,
+      "step": 388
+    },
+    {
+      "epoch": 0.034757745660866264,
+      "grad_norm": 0.5060571432113647,
+      "learning_rate": 9.999411802299339e-05,
+      "loss": 1.2208,
+      "step": 389
+    },
+    {
+      "epoch": 0.03484709719212813,
+      "grad_norm": 0.5558078289031982,
+      "learning_rate": 9.999389397218186e-05,
+      "loss": 1.1395,
+      "step": 390
+    },
+    {
+      "epoch": 0.034936448723389996,
+      "grad_norm": 0.49151378870010376,
+      "learning_rate": 9.999366573384884e-05,
+      "loss": 1.1426,
+      "step": 391
+    },
+    {
+      "epoch": 0.035025800254651865,
+      "grad_norm": 0.4773429036140442,
+      "learning_rate": 9.99934333080134e-05,
+      "loss": 1.1587,
+      "step": 392
+    },
+    {
+      "epoch": 0.035115151785913734,
+      "grad_norm": 0.43944311141967773,
+      "learning_rate": 9.999319669469505e-05,
+      "loss": 1.2301,
+      "step": 393
+    },
+    {
+      "epoch": 0.035204503317175596,
+      "grad_norm": 0.4840754270553589,
+      "learning_rate": 9.999295589391358e-05,
+      "loss": 1.1404,
+      "step": 394
+    },
+    {
+      "epoch": 0.035293854848437466,
+      "grad_norm": 0.49783623218536377,
+      "learning_rate": 9.999271090568918e-05,
+      "loss": 1.0676,
+      "step": 395
+    },
+    {
+      "epoch": 0.035383206379699335,
+      "grad_norm": 0.47931092977523804,
+      "learning_rate": 9.999246173004233e-05,
+      "loss": 1.1528,
+      "step": 396
+    },
+    {
+      "epoch": 0.0354725579109612,
+      "grad_norm": 0.47932660579681396,
+      "learning_rate": 9.999220836699395e-05,
+      "loss": 1.1604,
+      "step": 397
+    },
+    {
+      "epoch": 0.035561909442223066,
+      "grad_norm": 0.5339416861534119,
+      "learning_rate": 9.999195081656522e-05,
+      "loss": 1.097,
+      "step": 398
+    },
+    {
+      "epoch": 0.035651260973484936,
+      "grad_norm": 0.5252090096473694,
+      "learning_rate": 9.999168907877776e-05,
+      "loss": 1.085,
+      "step": 399
+    },
+    {
+      "epoch": 0.0357406125047468,
+      "grad_norm": 0.5167770981788635,
+      "learning_rate": 9.999142315365345e-05,
+      "loss": 1.1563,
+      "step": 400
+    },
+    {
+      "epoch": 0.03582996403600867,
+      "grad_norm": 0.6297523975372314,
+      "learning_rate": 9.999115304121457e-05,
+      "loss": 1.0762,
+      "step": 401
+    },
+    {
+      "epoch": 0.035919315567270536,
+      "grad_norm": 0.5326511859893799,
+      "learning_rate": 9.999087874148379e-05,
+      "loss": 1.1082,
+      "step": 402
+    },
+    {
+      "epoch": 0.0360086670985324,
+      "grad_norm": 0.5279747843742371,
+      "learning_rate": 9.999060025448403e-05,
+      "loss": 1.1412,
+      "step": 403
+    },
+    {
+      "epoch": 0.03609801862979427,
+      "grad_norm": 0.4827171564102173,
+      "learning_rate": 9.999031758023863e-05,
+      "loss": 1.1705,
+      "step": 404
+    },
+    {
+      "epoch": 0.03618737016105614,
+      "grad_norm": 0.5073494911193848,
+      "learning_rate": 9.999003071877129e-05,
+      "loss": 1.1579,
+      "step": 405
+    },
+    {
+      "epoch": 0.036276721692318,
+      "grad_norm": 0.4927610158920288,
+      "learning_rate": 9.9989739670106e-05,
+      "loss": 1.1158,
+      "step": 406
+    },
+    {
+      "epoch": 0.03636607322357987,
+      "grad_norm": 0.45825162529945374,
+      "learning_rate": 9.998944443426719e-05,
+      "loss": 1.1761,
+      "step": 407
+    },
+    {
+      "epoch": 0.03645542475484174,
+      "grad_norm": 0.5363956093788147,
+      "learning_rate": 9.998914501127954e-05,
+      "loss": 1.15,
+      "step": 408
+    },
+    {
+      "epoch": 0.0365447762861036,
+      "grad_norm": 0.5403926968574524,
+      "learning_rate": 9.998884140116816e-05,
+      "loss": 1.0831,
+      "step": 409
+    },
+    {
+      "epoch": 0.03663412781736547,
+      "grad_norm": 0.5043397545814514,
+      "learning_rate": 9.998853360395846e-05,
+      "loss": 1.0479,
+      "step": 410
+    },
+    {
+      "epoch": 0.03672347934862734,
+      "grad_norm": 0.5072253346443176,
+      "learning_rate": 9.998822161967623e-05,
+      "loss": 1.0596,
+      "step": 411
+    },
+    {
+      "epoch": 0.0368128308798892,
+      "grad_norm": 0.4440356492996216,
+      "learning_rate": 9.99879054483476e-05,
+      "loss": 1.1423,
+      "step": 412
+    },
+    {
+      "epoch": 0.03690218241115107,
+      "grad_norm": 0.5258128046989441,
+      "learning_rate": 9.998758508999906e-05,
+      "loss": 1.1072,
+      "step": 413
+    },
+    {
+      "epoch": 0.03699153394241294,
+      "grad_norm": 0.48225632309913635,
+      "learning_rate": 9.998726054465744e-05,
+      "loss": 1.1436,
+      "step": 414
+    },
+    {
+      "epoch": 0.0370808854736748,
+      "grad_norm": 0.49311602115631104,
+      "learning_rate": 9.998693181234992e-05,
+      "loss": 1.0847,
+      "step": 415
+    },
+    {
+      "epoch": 0.03717023700493667,
+      "grad_norm": 0.46585437655448914,
+      "learning_rate": 9.998659889310406e-05,
+      "loss": 1.1562,
+      "step": 416
+    },
+    {
+      "epoch": 0.03725958853619854,
+      "grad_norm": 0.5066165924072266,
+      "learning_rate": 9.99862617869477e-05,
+      "loss": 1.1278,
+      "step": 417
+    },
+    {
+      "epoch": 0.03734894006746041,
+      "grad_norm": 0.4618039131164551,
+      "learning_rate": 9.998592049390911e-05,
+      "loss": 1.1738,
+      "step": 418
+    },
+    {
+      "epoch": 0.03743829159872227,
+      "grad_norm": 0.4884462356567383,
+      "learning_rate": 9.998557501401687e-05,
+      "loss": 1.1826,
+      "step": 419
+    },
+    {
+      "epoch": 0.03752764312998414,
+      "grad_norm": 0.4807124137878418,
+      "learning_rate": 9.99852253472999e-05,
+      "loss": 1.1397,
+      "step": 420
+    },
+    {
+      "epoch": 0.03761699466124601,
+      "grad_norm": 0.4974918067455292,
+      "learning_rate": 9.998487149378752e-05,
+      "loss": 1.1481,
+      "step": 421
+    },
+    {
+      "epoch": 0.03770634619250787,
+      "grad_norm": 0.42765894532203674,
+      "learning_rate": 9.998451345350935e-05,
+      "loss": 1.1967,
+      "step": 422
+    },
+    {
+      "epoch": 0.03779569772376974,
+      "grad_norm": 0.509459912776947,
+      "learning_rate": 9.998415122649537e-05,
+      "loss": 1.1086,
+      "step": 423
+    },
+    {
+      "epoch": 0.03788504925503161,
+      "grad_norm": 0.4979933202266693,
+      "learning_rate": 9.998378481277593e-05,
+      "loss": 1.1003,
+      "step": 424
+    },
+    {
+      "epoch": 0.03797440078629347,
+      "grad_norm": 0.4624142348766327,
+      "learning_rate": 9.998341421238173e-05,
+      "loss": 1.1323,
+      "step": 425
+    },
+    {
+      "epoch": 0.03806375231755534,
+      "grad_norm": 0.4574481248855591,
+      "learning_rate": 9.998303942534382e-05,
+      "loss": 1.1195,
+      "step": 426
+    },
+    {
+      "epoch": 0.03815310384881721,
+      "grad_norm": 0.542635977268219,
+      "learning_rate": 9.998266045169356e-05,
+      "loss": 1.1185,
+      "step": 427
+    },
+    {
+      "epoch": 0.038242455380079074,
+      "grad_norm": 0.48302701115608215,
+      "learning_rate": 9.99822772914627e-05,
+      "loss": 1.13,
+      "step": 428
+    },
+    {
+      "epoch": 0.03833180691134094,
+      "grad_norm": 0.4780206084251404,
+      "learning_rate": 9.998188994468337e-05,
+      "loss": 1.0496,
+      "step": 429
+    },
+    {
+      "epoch": 0.03842115844260281,
+      "grad_norm": 0.5406165719032288,
+      "learning_rate": 9.998149841138797e-05,
+      "loss": 1.0936,
+      "step": 430
+    },
+    {
+      "epoch": 0.038510509973864675,
+      "grad_norm": 0.4520138204097748,
+      "learning_rate": 9.998110269160932e-05,
+      "loss": 1.2128,
+      "step": 431
+    },
+    {
+      "epoch": 0.038599861505126544,
+      "grad_norm": 0.4702879786491394,
+      "learning_rate": 9.998070278538057e-05,
+      "loss": 1.2563,
+      "step": 432
+    },
+    {
+      "epoch": 0.03868921303638841,
+      "grad_norm": 0.5223757028579712,
+      "learning_rate": 9.998029869273518e-05,
+      "loss": 1.091,
+      "step": 433
+    },
+    {
+      "epoch": 0.038778564567650275,
+      "grad_norm": 0.5411685705184937,
+      "learning_rate": 9.997989041370704e-05,
+      "loss": 1.1356,
+      "step": 434
+    },
+    {
+      "epoch": 0.038867916098912145,
+      "grad_norm": 0.48428136110305786,
+      "learning_rate": 9.997947794833034e-05,
+      "loss": 1.2075,
+      "step": 435
+    },
+    {
+      "epoch": 0.038957267630174014,
+      "grad_norm": 0.4824248254299164,
+      "learning_rate": 9.997906129663961e-05,
+      "loss": 1.1918,
+      "step": 436
+    },
+    {
+      "epoch": 0.039046619161435876,
+      "grad_norm": 0.5200150609016418,
+      "learning_rate": 9.997864045866975e-05,
+      "loss": 1.1364,
+      "step": 437
+    },
+    {
+      "epoch": 0.039135970692697745,
+      "grad_norm": 0.5119284391403198,
+      "learning_rate": 9.997821543445602e-05,
+      "loss": 1.1136,
+      "step": 438
+    },
+    {
+      "epoch": 0.039225322223959615,
+      "grad_norm": 0.4953431785106659,
+      "learning_rate": 9.997778622403402e-05,
+      "loss": 1.2217,
+      "step": 439
+    },
+    {
+      "epoch": 0.03931467375522148,
+      "grad_norm": 0.49554356932640076,
+      "learning_rate": 9.997735282743969e-05,
+      "loss": 1.1535,
+      "step": 440
+    },
+    {
+      "epoch": 0.039404025286483346,
+      "grad_norm": 0.5138264894485474,
+      "learning_rate": 9.997691524470936e-05,
+      "loss": 0.9905,
+      "step": 441
+    },
+    {
+      "epoch": 0.039493376817745215,
+      "grad_norm": 0.4627537727355957,
+      "learning_rate": 9.997647347587964e-05,
+      "loss": 1.1075,
+      "step": 442
+    },
+    {
+      "epoch": 0.03958272834900708,
+      "grad_norm": 0.5191687941551208,
+      "learning_rate": 9.997602752098758e-05,
+      "loss": 1.172,
+      "step": 443
+    },
+    {
+      "epoch": 0.03967207988026895,
+      "grad_norm": 0.4805525839328766,
+      "learning_rate": 9.997557738007049e-05,
+      "loss": 1.154,
+      "step": 444
+    },
+    {
+      "epoch": 0.039761431411530816,
+      "grad_norm": 0.5130792260169983,
+      "learning_rate": 9.99751230531661e-05,
+      "loss": 1.0842,
+      "step": 445
+    },
+    {
+      "epoch": 0.039850782942792685,
+      "grad_norm": 0.4610874354839325,
+      "learning_rate": 9.997466454031246e-05,
+      "loss": 1.0973,
+      "step": 446
+    },
+    {
+      "epoch": 0.03994013447405455,
+      "grad_norm": 0.43982231616973877,
+      "learning_rate": 9.997420184154798e-05,
+      "loss": 1.156,
+      "step": 447
+    },
+    {
+      "epoch": 0.04002948600531642,
+      "grad_norm": 0.49081990122795105,
+      "learning_rate": 9.99737349569114e-05,
+      "loss": 1.0796,
+      "step": 448
+    },
+    {
+      "epoch": 0.040118837536578286,
+      "grad_norm": 0.5369452238082886,
+      "learning_rate": 9.997326388644183e-05,
+      "loss": 1.0677,
+      "step": 449
+    },
+    {
+      "epoch": 0.04020818906784015,
+      "grad_norm": 0.4784664213657379,
+      "learning_rate": 9.997278863017874e-05,
+      "loss": 1.0792,
+      "step": 450
+    },
+    {
+      "epoch": 0.04029754059910202,
+      "grad_norm": 0.53533935546875,
+      "learning_rate": 9.997230918816191e-05,
+      "loss": 1.0963,
+      "step": 451
+    },
+    {
+      "epoch": 0.04038689213036389,
+      "grad_norm": 0.5231233239173889,
+      "learning_rate": 9.997182556043155e-05,
+      "loss": 1.1181,
+      "step": 452
+    },
+    {
+      "epoch": 0.04047624366162575,
+      "grad_norm": 0.4834752380847931,
+      "learning_rate": 9.997133774702812e-05,
+      "loss": 1.1174,
+      "step": 453
+    },
+    {
+      "epoch": 0.04056559519288762,
+      "grad_norm": 0.48885536193847656,
+      "learning_rate": 9.997084574799252e-05,
+      "loss": 1.0655,
+      "step": 454
+    },
+    {
+      "epoch": 0.04065494672414949,
+      "grad_norm": 0.4479488730430603,
+      "learning_rate": 9.99703495633659e-05,
+      "loss": 1.1568,
+      "step": 455
+    },
+    {
+      "epoch": 0.04074429825541135,
+      "grad_norm": 0.5108731985092163,
+      "learning_rate": 9.996984919318989e-05,
+      "loss": 1.1712,
+      "step": 456
+    },
+    {
+      "epoch": 0.04083364978667322,
+      "grad_norm": 0.5030118227005005,
+      "learning_rate": 9.996934463750636e-05,
+      "loss": 1.0666,
+      "step": 457
+    },
+    {
+      "epoch": 0.04092300131793509,
+      "grad_norm": 0.5227623581886292,
+      "learning_rate": 9.996883589635757e-05,
+      "loss": 1.1652,
+      "step": 458
+    },
+    {
+      "epoch": 0.04101235284919695,
+      "grad_norm": 0.4760054349899292,
+      "learning_rate": 9.996832296978616e-05,
+      "loss": 1.151,
+      "step": 459
+    },
+    {
+      "epoch": 0.04110170438045882,
+      "grad_norm": 0.42174920439720154,
+      "learning_rate": 9.996780585783508e-05,
+      "loss": 1.1306,
+      "step": 460
+    },
+    {
+      "epoch": 0.04119105591172069,
+      "grad_norm": 0.42818742990493774,
+      "learning_rate": 9.996728456054762e-05,
+      "loss": 1.1775,
+      "step": 461
+    },
+    {
+      "epoch": 0.04128040744298255,
+      "grad_norm": 0.48023608326911926,
+      "learning_rate": 9.996675907796749e-05,
+      "loss": 1.1809,
+      "step": 462
+    },
+    {
+      "epoch": 0.04136975897424442,
+      "grad_norm": 0.45978084206581116,
+      "learning_rate": 9.996622941013867e-05,
+      "loss": 1.2312,
+      "step": 463
+    },
+    {
+      "epoch": 0.04145911050550629,
+      "grad_norm": 0.4810321629047394,
+      "learning_rate": 9.996569555710553e-05,
+      "loss": 1.0946,
+      "step": 464
+    },
+    {
+      "epoch": 0.04154846203676815,
+      "grad_norm": 0.43682757019996643,
+      "learning_rate": 9.996515751891279e-05,
+      "loss": 1.1556,
+      "step": 465
+    },
+    {
+      "epoch": 0.04163781356803002,
+      "grad_norm": 0.4703795313835144,
+      "learning_rate": 9.996461529560553e-05,
+      "loss": 1.1268,
+      "step": 466
+    },
+    {
+      "epoch": 0.04172716509929189,
+      "grad_norm": 0.44790583848953247,
+      "learning_rate": 9.996406888722914e-05,
+      "loss": 1.0907,
+      "step": 467
+    },
+    {
+      "epoch": 0.04181651663055375,
+      "grad_norm": 0.5048056840896606,
+      "learning_rate": 9.996351829382941e-05,
+      "loss": 1.0709,
+      "step": 468
+    },
+    {
+      "epoch": 0.04190586816181562,
+      "grad_norm": 0.4316783845424652,
+      "learning_rate": 9.996296351545244e-05,
+      "loss": 1.1355,
+      "step": 469
+    },
+    {
+      "epoch": 0.04199521969307749,
+      "grad_norm": 0.4551528990268707,
+      "learning_rate": 9.996240455214472e-05,
+      "loss": 1.0943,
+      "step": 470
+    },
+    {
+      "epoch": 0.042084571224339354,
+      "grad_norm": 0.4864242970943451,
+      "learning_rate": 9.996184140395306e-05,
+      "loss": 1.0194,
+      "step": 471
+    },
+    {
+      "epoch": 0.04217392275560122,
+      "grad_norm": 0.5039882659912109,
+      "learning_rate": 9.996127407092462e-05,
+      "loss": 1.1392,
+      "step": 472
+    },
+    {
+      "epoch": 0.04226327428686309,
+      "grad_norm": 0.5886643528938293,
+      "learning_rate": 9.996070255310692e-05,
+      "loss": 1.0797,
+      "step": 473
+    },
+    {
+      "epoch": 0.04235262581812496,
+      "grad_norm": 0.5036435723304749,
+      "learning_rate": 9.996012685054786e-05,
+      "loss": 1.1886,
+      "step": 474
+    },
+    {
+      "epoch": 0.042441977349386824,
+      "grad_norm": 0.4894576370716095,
+      "learning_rate": 9.995954696329562e-05,
+      "loss": 1.1631,
+      "step": 475
+    },
+    {
+      "epoch": 0.04253132888064869,
+      "grad_norm": 0.4920431673526764,
+      "learning_rate": 9.99589628913988e-05,
+      "loss": 1.1382,
+      "step": 476
+    },
+    {
+      "epoch": 0.04262068041191056,
+      "grad_norm": 0.42358994483947754,
+      "learning_rate": 9.995837463490632e-05,
+      "loss": 1.1086,
+      "step": 477
+    },
+    {
+      "epoch": 0.042710031943172425,
+      "grad_norm": 0.4317459464073181,
+      "learning_rate": 9.995778219386744e-05,
+      "loss": 1.132,
+      "step": 478
+    },
+    {
+      "epoch": 0.042799383474434294,
+      "grad_norm": 0.4853540062904358,
+      "learning_rate": 9.995718556833178e-05,
+      "loss": 1.0694,
+      "step": 479
+    },
+    {
+      "epoch": 0.04288873500569616,
+      "grad_norm": 0.49975448846817017,
+      "learning_rate": 9.995658475834933e-05,
+      "loss": 1.1211,
+      "step": 480
+    },
+    {
+      "epoch": 0.042978086536958025,
+      "grad_norm": 0.4265710115432739,
+      "learning_rate": 9.995597976397042e-05,
+      "loss": 1.1266,
+      "step": 481
+    },
+    {
+      "epoch": 0.043067438068219895,
+      "grad_norm": 0.5126653909683228,
+      "learning_rate": 9.995537058524569e-05,
+      "loss": 1.1113,
+      "step": 482
+    },
+    {
+      "epoch": 0.043156789599481764,
+      "grad_norm": 0.5130075812339783,
+      "learning_rate": 9.99547572222262e-05,
+      "loss": 1.1382,
+      "step": 483
+    },
+    {
+      "epoch": 0.043246141130743626,
+      "grad_norm": 0.46531277894973755,
+      "learning_rate": 9.995413967496333e-05,
+      "loss": 1.0806,
+      "step": 484
+    },
+    {
+      "epoch": 0.043335492662005495,
+      "grad_norm": 0.4815559983253479,
+      "learning_rate": 9.995351794350876e-05,
+      "loss": 1.136,
+      "step": 485
+    },
+    {
+      "epoch": 0.043424844193267365,
+      "grad_norm": 0.417111873626709,
+      "learning_rate": 9.99528920279146e-05,
+      "loss": 1.158,
+      "step": 486
+    },
+    {
+      "epoch": 0.04351419572452923,
+      "grad_norm": 0.4570912718772888,
+      "learning_rate": 9.995226192823329e-05,
+      "loss": 1.134,
+      "step": 487
+    },
+    {
+      "epoch": 0.043603547255791096,
+      "grad_norm": 0.5166110992431641,
+      "learning_rate": 9.995162764451758e-05,
+      "loss": 1.1111,
+      "step": 488
+    },
+    {
+      "epoch": 0.043692898787052965,
+      "grad_norm": 0.4371122419834137,
+      "learning_rate": 9.99509891768206e-05,
+      "loss": 1.1687,
+      "step": 489
+    },
+    {
+      "epoch": 0.04378225031831483,
+      "grad_norm": 0.5123320817947388,
+      "learning_rate": 9.995034652519586e-05,
+      "loss": 1.1159,
+      "step": 490
+    },
+    {
+      "epoch": 0.0438716018495767,
+      "grad_norm": 0.5270020961761475,
+      "learning_rate": 9.994969968969715e-05,
+      "loss": 1.0321,
+      "step": 491
+    },
+    {
+      "epoch": 0.043960953380838566,
+      "grad_norm": 0.43343213200569153,
+      "learning_rate": 9.994904867037867e-05,
+      "loss": 1.1311,
+      "step": 492
+    },
+    {
+      "epoch": 0.04405030491210043,
+      "grad_norm": 0.5779858827590942,
+      "learning_rate": 9.994839346729495e-05,
+      "loss": 1.0312,
+      "step": 493
+    },
+    {
+      "epoch": 0.0441396564433623,
+      "grad_norm": 0.4757930338382721,
+      "learning_rate": 9.994773408050084e-05,
+      "loss": 1.1148,
+      "step": 494
+    },
+    {
+      "epoch": 0.04422900797462417,
+      "grad_norm": 0.483888179063797,
+      "learning_rate": 9.994707051005164e-05,
+      "loss": 1.1387,
+      "step": 495
+    },
+    {
+      "epoch": 0.04431835950588603,
+      "grad_norm": 0.4878624677658081,
+      "learning_rate": 9.994640275600285e-05,
+      "loss": 1.1294,
+      "step": 496
+    },
+    {
+      "epoch": 0.0444077110371479,
+      "grad_norm": 0.5229454040527344,
+      "learning_rate": 9.994573081841046e-05,
+      "loss": 1.0685,
+      "step": 497
+    },
+    {
+      "epoch": 0.04449706256840977,
+      "grad_norm": 0.49779602885246277,
+      "learning_rate": 9.994505469733071e-05,
+      "loss": 1.1596,
+      "step": 498
+    },
+    {
+      "epoch": 0.04458641409967163,
+      "grad_norm": 0.5035576224327087,
+      "learning_rate": 9.994437439282027e-05,
+      "loss": 1.1966,
+      "step": 499
+    },
+    {
+      "epoch": 0.0446757656309335,
+      "grad_norm": 0.4735001027584076,
+      "learning_rate": 9.99436899049361e-05,
+      "loss": 1.1213,
+      "step": 500
+    },
+    {
+      "epoch": 0.04476511716219537,
+      "grad_norm": 0.5072652697563171,
+      "learning_rate": 9.994300123373554e-05,
+      "loss": 1.1671,
+      "step": 501
+    },
+    {
+      "epoch": 0.04485446869345724,
+      "grad_norm": 0.4931294620037079,
+      "learning_rate": 9.994230837927627e-05,
+      "loss": 1.1065,
+      "step": 502
+    },
+    {
+      "epoch": 0.0449438202247191,
+      "grad_norm": 0.4530344307422638,
+      "learning_rate": 9.994161134161634e-05,
+      "loss": 1.1528,
+      "step": 503
+    },
+    {
+      "epoch": 0.04503317175598097,
+      "grad_norm": 0.5151768326759338,
+      "learning_rate": 9.99409101208141e-05,
+      "loss": 0.9932,
+      "step": 504
+    },
+    {
+      "epoch": 0.04512252328724284,
+      "grad_norm": 0.5038042068481445,
+      "learning_rate": 9.994020471692833e-05,
+      "loss": 1.1132,
+      "step": 505
+    },
+    {
+      "epoch": 0.0452118748185047,
+      "grad_norm": 0.47841110825538635,
+      "learning_rate": 9.993949513001807e-05,
+      "loss": 1.1563,
+      "step": 506
+    },
+    {
+      "epoch": 0.04530122634976657,
+      "grad_norm": 0.5167407989501953,
+      "learning_rate": 9.993878136014278e-05,
+      "loss": 1.1383,
+      "step": 507
+    },
+    {
+      "epoch": 0.04539057788102844,
+      "grad_norm": 0.5208450555801392,
+      "learning_rate": 9.993806340736225e-05,
+      "loss": 1.1225,
+      "step": 508
+    },
+    {
+      "epoch": 0.0454799294122903,
+      "grad_norm": 0.5085687637329102,
+      "learning_rate": 9.99373412717366e-05,
+      "loss": 1.1747,
+      "step": 509
+    },
+    {
+      "epoch": 0.04556928094355217,
+      "grad_norm": 0.5266485214233398,
+      "learning_rate": 9.993661495332633e-05,
+      "loss": 1.1311,
+      "step": 510
+    },
+    {
+      "epoch": 0.04565863247481404,
+      "grad_norm": 0.5289018154144287,
+      "learning_rate": 9.993588445219227e-05,
+      "loss": 1.114,
+      "step": 511
+    },
+    {
+      "epoch": 0.0457479840060759,
+      "grad_norm": 0.4422353208065033,
+      "learning_rate": 9.99351497683956e-05,
+      "loss": 1.178,
+      "step": 512
+    },
+    {
+      "epoch": 0.04583733553733777,
+      "grad_norm": 0.5575171113014221,
+      "learning_rate": 9.993441090199787e-05,
+      "loss": 1.0279,
+      "step": 513
+    },
+    {
+      "epoch": 0.04592668706859964,
+      "grad_norm": 0.53556227684021,
+      "learning_rate": 9.993366785306097e-05,
+      "loss": 1.1212,
+      "step": 514
+    },
+    {
+      "epoch": 0.0460160385998615,
+      "grad_norm": 0.5543893575668335,
+      "learning_rate": 9.993292062164714e-05,
+      "loss": 1.0113,
+      "step": 515
+    },
+    {
+      "epoch": 0.04610539013112337,
+      "grad_norm": 0.5223544239997864,
+      "learning_rate": 9.993216920781894e-05,
+      "loss": 1.0224,
+      "step": 516
+    },
+    {
+      "epoch": 0.04619474166238524,
+      "grad_norm": 0.5060791969299316,
+      "learning_rate": 9.993141361163935e-05,
+      "loss": 1.1628,
+      "step": 517
+    },
+    {
+      "epoch": 0.046284093193647104,
+      "grad_norm": 0.4724702537059784,
+      "learning_rate": 9.993065383317163e-05,
+      "loss": 1.1255,
+      "step": 518
+    },
+    {
+      "epoch": 0.04637344472490897,
+      "grad_norm": 0.47285956144332886,
+      "learning_rate": 9.992988987247944e-05,
+      "loss": 1.1659,
+      "step": 519
+    },
+    {
+      "epoch": 0.04646279625617084,
+      "grad_norm": 0.4982796013355255,
+      "learning_rate": 9.992912172962674e-05,
+      "loss": 1.1952,
+      "step": 520
+    },
+    {
+      "epoch": 0.046552147787432704,
+      "grad_norm": 0.48061710596084595,
+      "learning_rate": 9.99283494046779e-05,
+      "loss": 1.1388,
+      "step": 521
+    },
+    {
+      "epoch": 0.046641499318694574,
+      "grad_norm": 0.45872175693511963,
+      "learning_rate": 9.99275728976976e-05,
+      "loss": 1.1262,
+      "step": 522
+    },
+    {
+      "epoch": 0.04673085084995644,
+      "grad_norm": 0.4480443000793457,
+      "learning_rate": 9.992679220875088e-05,
+      "loss": 1.1235,
+      "step": 523
+    },
+    {
+      "epoch": 0.046820202381218305,
+      "grad_norm": 0.43175145983695984,
+      "learning_rate": 9.992600733790314e-05,
+      "loss": 1.1185,
+      "step": 524
+    },
+    {
+      "epoch": 0.046909553912480174,
+      "grad_norm": 0.41958916187286377,
+      "learning_rate": 9.99252182852201e-05,
+      "loss": 1.1661,
+      "step": 525
+    },
+    {
+      "epoch": 0.046998905443742044,
+      "grad_norm": 0.4344329237937927,
+      "learning_rate": 9.992442505076787e-05,
+      "loss": 1.1364,
+      "step": 526
+    },
+    {
+      "epoch": 0.047088256975003906,
+      "grad_norm": 0.4704360067844391,
+      "learning_rate": 9.992362763461287e-05,
+      "loss": 1.0588,
+      "step": 527
+    },
+    {
+      "epoch": 0.047177608506265775,
+      "grad_norm": 0.4403749704360962,
+      "learning_rate": 9.992282603682192e-05,
+      "loss": 1.1517,
+      "step": 528
+    },
+    {
+      "epoch": 0.047266960037527644,
+      "grad_norm": 0.4906516969203949,
+      "learning_rate": 9.992202025746215e-05,
+      "loss": 1.0916,
+      "step": 529
+    },
+    {
+      "epoch": 0.04735631156878951,
+      "grad_norm": 0.5123149752616882,
+      "learning_rate": 9.992121029660106e-05,
+      "loss": 1.0794,
+      "step": 530
+    },
+    {
+      "epoch": 0.047445663100051376,
+      "grad_norm": 0.46089616417884827,
+      "learning_rate": 9.992039615430648e-05,
+      "loss": 1.1846,
+      "step": 531
+    },
+    {
+      "epoch": 0.047535014631313245,
+      "grad_norm": 0.5317028760910034,
+      "learning_rate": 9.99195778306466e-05,
+      "loss": 1.1289,
+      "step": 532
+    },
+    {
+      "epoch": 0.047624366162575114,
+      "grad_norm": 0.48291924595832825,
+      "learning_rate": 9.991875532568999e-05,
+      "loss": 1.1427,
+      "step": 533
+    },
+    {
+      "epoch": 0.04771371769383698,
+      "grad_norm": 0.46745729446411133,
+      "learning_rate": 9.991792863950552e-05,
+      "loss": 1.1027,
+      "step": 534
+    },
+    {
+      "epoch": 0.047803069225098846,
+      "grad_norm": 0.4555657207965851,
+      "learning_rate": 9.991709777216242e-05,
+      "loss": 1.0926,
+      "step": 535
+    },
+    {
+      "epoch": 0.047892420756360715,
+      "grad_norm": 0.4779694080352783,
+      "learning_rate": 9.991626272373033e-05,
+      "loss": 1.0919,
+      "step": 536
+    },
+    {
+      "epoch": 0.04798177228762258,
+      "grad_norm": 0.4468933045864105,
+      "learning_rate": 9.991542349427916e-05,
+      "loss": 1.1903,
+      "step": 537
+    },
+    {
+      "epoch": 0.04807112381888445,
+      "grad_norm": 0.5170602202415466,
+      "learning_rate": 9.99145800838792e-05,
+      "loss": 1.0183,
+      "step": 538
+    },
+    {
+      "epoch": 0.048160475350146316,
+      "grad_norm": 0.4570893347263336,
+      "learning_rate": 9.991373249260112e-05,
+      "loss": 1.0834,
+      "step": 539
+    },
+    {
+      "epoch": 0.04824982688140818,
+      "grad_norm": 0.4547278881072998,
+      "learning_rate": 9.99128807205159e-05,
+      "loss": 1.1534,
+      "step": 540
+    },
+    {
+      "epoch": 0.04833917841267005,
+      "grad_norm": 0.47675079107284546,
+      "learning_rate": 9.991202476769488e-05,
+      "loss": 1.0934,
+      "step": 541
+    },
+    {
+      "epoch": 0.04842852994393192,
+      "grad_norm": 0.4400666654109955,
+      "learning_rate": 9.991116463420976e-05,
+      "loss": 1.0891,
+      "step": 542
+    },
+    {
+      "epoch": 0.04851788147519378,
+      "grad_norm": 0.4726406931877136,
+      "learning_rate": 9.99103003201326e-05,
+      "loss": 1.0805,
+      "step": 543
+    },
+    {
+      "epoch": 0.04860723300645565,
+      "grad_norm": 0.484070748090744,
+      "learning_rate": 9.990943182553579e-05,
+      "loss": 1.0694,
+      "step": 544
+    },
+    {
+      "epoch": 0.04869658453771752,
+      "grad_norm": 0.5411748290061951,
+      "learning_rate": 9.990855915049204e-05,
+      "loss": 1.0275,
+      "step": 545
+    },
+    {
+      "epoch": 0.04878593606897938,
+      "grad_norm": 0.46557214856147766,
+      "learning_rate": 9.990768229507447e-05,
+      "loss": 1.2306,
+      "step": 546
+    },
+    {
+      "epoch": 0.04887528760024125,
+      "grad_norm": 0.5048271417617798,
+      "learning_rate": 9.990680125935657e-05,
+      "loss": 1.063,
+      "step": 547
+    },
+    {
+      "epoch": 0.04896463913150312,
+      "grad_norm": 0.4694403409957886,
+      "learning_rate": 9.990591604341206e-05,
+      "loss": 1.117,
+      "step": 548
+    },
+    {
+      "epoch": 0.04905399066276498,
+      "grad_norm": 0.5101834535598755,
+      "learning_rate": 9.990502664731515e-05,
+      "loss": 1.096,
+      "step": 549
+    },
+    {
+      "epoch": 0.04914334219402685,
+      "grad_norm": 0.4212850332260132,
+      "learning_rate": 9.99041330711403e-05,
+      "loss": 1.2322,
+      "step": 550
+    },
+    {
+      "epoch": 0.04923269372528872,
+      "grad_norm": 0.4730430245399475,
+      "learning_rate": 9.990323531496235e-05,
+      "loss": 1.1706,
+      "step": 551
+    },
+    {
+      "epoch": 0.04932204525655058,
+      "grad_norm": 0.4167949855327606,
+      "learning_rate": 9.990233337885652e-05,
+      "loss": 1.1796,
+      "step": 552
+    },
+    {
+      "epoch": 0.04941139678781245,
+      "grad_norm": 0.44390869140625,
+      "learning_rate": 9.990142726289837e-05,
+      "loss": 1.1358,
+      "step": 553
+    },
+    {
+      "epoch": 0.04950074831907432,
+      "grad_norm": 0.47111496329307556,
+      "learning_rate": 9.990051696716375e-05,
+      "loss": 1.1188,
+      "step": 554
+    },
+    {
+      "epoch": 0.04959009985033618,
+      "grad_norm": 0.40082886815071106,
+      "learning_rate": 9.989960249172894e-05,
+      "loss": 1.1225,
+      "step": 555
+    },
+    {
+      "epoch": 0.04967945138159805,
+      "grad_norm": 0.42682352662086487,
+      "learning_rate": 9.989868383667054e-05,
+      "loss": 1.0989,
+      "step": 556
+    },
+    {
+      "epoch": 0.04976880291285992,
+      "grad_norm": 0.45663878321647644,
+      "learning_rate": 9.989776100206548e-05,
+      "loss": 1.1296,
+      "step": 557
+    },
+    {
+      "epoch": 0.04985815444412178,
+      "grad_norm": 0.5189902782440186,
+      "learning_rate": 9.989683398799106e-05,
+      "loss": 1.0356,
+      "step": 558
+    },
+    {
+      "epoch": 0.04994750597538365,
+      "grad_norm": 0.41676539182662964,
+      "learning_rate": 9.989590279452492e-05,
+      "loss": 1.0858,
+      "step": 559
+    },
+    {
+      "epoch": 0.05003685750664552,
+      "grad_norm": 0.47391507029533386,
+      "learning_rate": 9.989496742174509e-05,
+      "loss": 1.1989,
+      "step": 560
+    },
+    {
+      "epoch": 0.05012620903790739,
+      "grad_norm": 0.48219195008277893,
+      "learning_rate": 9.989402786972988e-05,
+      "loss": 1.0994,
+      "step": 561
+    },
+    {
+      "epoch": 0.05021556056916925,
+      "grad_norm": 0.5182317495346069,
+      "learning_rate": 9.989308413855802e-05,
+      "loss": 1.068,
+      "step": 562
+    },
+    {
+      "epoch": 0.05030491210043112,
+      "grad_norm": 0.44291019439697266,
+      "learning_rate": 9.989213622830853e-05,
+      "loss": 1.0893,
+      "step": 563
+    },
+    {
+      "epoch": 0.05039426363169299,
+      "grad_norm": 0.4342930018901825,
+      "learning_rate": 9.989118413906082e-05,
+      "loss": 1.1266,
+      "step": 564
+    },
+    {
+      "epoch": 0.050483615162954854,
+      "grad_norm": 0.4785180985927582,
+      "learning_rate": 9.989022787089463e-05,
+      "loss": 1.1128,
+      "step": 565
+    },
+    {
+      "epoch": 0.05057296669421672,
+      "grad_norm": 0.4673145115375519,
+      "learning_rate": 9.988926742389009e-05,
+      "loss": 1.087,
+      "step": 566
+    },
+    {
+      "epoch": 0.05066231822547859,
+      "grad_norm": 0.42585358023643494,
+      "learning_rate": 9.98883027981276e-05,
+      "loss": 1.1635,
+      "step": 567
+    },
+    {
+      "epoch": 0.050751669756740454,
+      "grad_norm": 0.4596778154373169,
+      "learning_rate": 9.988733399368799e-05,
+      "loss": 1.1818,
+      "step": 568
+    },
+    {
+      "epoch": 0.050841021288002324,
+      "grad_norm": 0.4845656454563141,
+      "learning_rate": 9.988636101065239e-05,
+      "loss": 1.1218,
+      "step": 569
+    },
+    {
+      "epoch": 0.05093037281926419,
+      "grad_norm": 0.4579009413719177,
+      "learning_rate": 9.988538384910231e-05,
+      "loss": 1.0752,
+      "step": 570
+    },
+    {
+      "epoch": 0.051019724350526055,
+      "grad_norm": 0.4427371025085449,
+      "learning_rate": 9.988440250911959e-05,
+      "loss": 1.1323,
+      "step": 571
+    },
+    {
+      "epoch": 0.051109075881787924,
+      "grad_norm": 0.4617055654525757,
+      "learning_rate": 9.988341699078643e-05,
+      "loss": 1.1641,
+      "step": 572
+    },
+    {
+      "epoch": 0.051198427413049793,
+      "grad_norm": 0.4859091639518738,
+      "learning_rate": 9.988242729418538e-05,
+      "loss": 1.0782,
+      "step": 573
+    },
+    {
+      "epoch": 0.051287778944311656,
+      "grad_norm": 0.4534977674484253,
+      "learning_rate": 9.988143341939933e-05,
+      "loss": 1.1194,
+      "step": 574
+    },
+    {
+      "epoch": 0.051377130475573525,
+      "grad_norm": 0.46073511242866516,
+      "learning_rate": 9.988043536651153e-05,
+      "loss": 1.091,
+      "step": 575
+    },
+    {
+      "epoch": 0.051466482006835394,
+      "grad_norm": 0.4721985459327698,
+      "learning_rate": 9.98794331356056e-05,
+      "loss": 1.1074,
+      "step": 576
+    },
+    {
+      "epoch": 0.05155583353809726,
+      "grad_norm": 0.45086175203323364,
+      "learning_rate": 9.987842672676544e-05,
+      "loss": 1.0933,
+      "step": 577
+    },
+    {
+      "epoch": 0.051645185069359126,
+      "grad_norm": 0.4174632728099823,
+      "learning_rate": 9.98774161400754e-05,
+      "loss": 1.1387,
+      "step": 578
+    },
+    {
+      "epoch": 0.051734536600620995,
+      "grad_norm": 0.4772718548774719,
+      "learning_rate": 9.987640137562008e-05,
+      "loss": 1.1551,
+      "step": 579
+    },
+    {
+      "epoch": 0.05182388813188286,
+      "grad_norm": 0.4615418016910553,
+      "learning_rate": 9.987538243348453e-05,
+      "loss": 1.1593,
+      "step": 580
+    },
+    {
+      "epoch": 0.05191323966314473,
+      "grad_norm": 0.48848336935043335,
+      "learning_rate": 9.987435931375406e-05,
+      "loss": 1.1167,
+      "step": 581
+    },
+    {
+      "epoch": 0.052002591194406596,
+      "grad_norm": 0.4644308090209961,
+      "learning_rate": 9.987333201651436e-05,
+      "loss": 1.0925,
+      "step": 582
+    },
+    {
+      "epoch": 0.05209194272566846,
+      "grad_norm": 0.46790382266044617,
+      "learning_rate": 9.98723005418515e-05,
+      "loss": 1.1186,
+      "step": 583
+    },
+    {
+      "epoch": 0.05218129425693033,
+      "grad_norm": 0.4854678213596344,
+      "learning_rate": 9.987126488985188e-05,
+      "loss": 1.1301,
+      "step": 584
+    },
+    {
+      "epoch": 0.0522706457881922,
+      "grad_norm": 0.47811320424079895,
+      "learning_rate": 9.987022506060221e-05,
+      "loss": 1.1431,
+      "step": 585
+    },
+    {
+      "epoch": 0.05235999731945406,
+      "grad_norm": 0.5481966733932495,
+      "learning_rate": 9.986918105418963e-05,
+      "loss": 1.0918,
+      "step": 586
+    },
+    {
+      "epoch": 0.05244934885071593,
+      "grad_norm": 0.4821578562259674,
+      "learning_rate": 9.986813287070158e-05,
+      "loss": 1.0715,
+      "step": 587
+    },
+    {
+      "epoch": 0.0525387003819778,
+      "grad_norm": 0.46433693170547485,
+      "learning_rate": 9.986708051022583e-05,
+      "loss": 1.0865,
+      "step": 588
+    },
+    {
+      "epoch": 0.052628051913239667,
+      "grad_norm": 0.44410815834999084,
+      "learning_rate": 9.986602397285054e-05,
+      "loss": 1.0884,
+      "step": 589
+    },
+    {
+      "epoch": 0.05271740344450153,
+      "grad_norm": 0.4573810398578644,
+      "learning_rate": 9.986496325866422e-05,
+      "loss": 1.1896,
+      "step": 590
+    },
+    {
+      "epoch": 0.0528067549757634,
+      "grad_norm": 0.5605972409248352,
+      "learning_rate": 9.986389836775569e-05,
+      "loss": 1.0038,
+      "step": 591
+    },
+    {
+      "epoch": 0.05289610650702527,
+      "grad_norm": 0.4885507822036743,
+      "learning_rate": 9.986282930021418e-05,
+      "loss": 1.1346,
+      "step": 592
+    },
+    {
+      "epoch": 0.05298545803828713,
+      "grad_norm": 0.4384009838104248,
+      "learning_rate": 9.986175605612921e-05,
+      "loss": 1.1165,
+      "step": 593
+    },
+    {
+      "epoch": 0.053074809569549,
+      "grad_norm": 0.4354461133480072,
+      "learning_rate": 9.986067863559067e-05,
+      "loss": 1.0599,
+      "step": 594
+    },
+    {
+      "epoch": 0.05316416110081087,
+      "grad_norm": 0.503073513507843,
+      "learning_rate": 9.985959703868884e-05,
+      "loss": 1.1166,
+      "step": 595
+    },
+    {
+      "epoch": 0.05325351263207273,
+      "grad_norm": 0.5100207924842834,
+      "learning_rate": 9.985851126551428e-05,
+      "loss": 1.1165,
+      "step": 596
+    },
+    {
+      "epoch": 0.0533428641633346,
+      "grad_norm": 0.5106130838394165,
+      "learning_rate": 9.985742131615794e-05,
+      "loss": 1.0688,
+      "step": 597
+    },
+    {
+      "epoch": 0.05343221569459647,
+      "grad_norm": 0.46070772409439087,
+      "learning_rate": 9.985632719071113e-05,
+      "loss": 1.1599,
+      "step": 598
+    },
+    {
+      "epoch": 0.05352156722585833,
+      "grad_norm": 0.45849481225013733,
+      "learning_rate": 9.985522888926549e-05,
+      "loss": 1.0446,
+      "step": 599
+    },
+    {
+      "epoch": 0.0536109187571202,
+      "grad_norm": 0.4356798827648163,
+      "learning_rate": 9.985412641191301e-05,
+      "loss": 1.1269,
+      "step": 600
+    },
+    {
+      "epoch": 0.05370027028838207,
+      "grad_norm": 0.5090345144271851,
+      "learning_rate": 9.985301975874604e-05,
+      "loss": 1.0689,
+      "step": 601
+    },
+    {
+      "epoch": 0.05378962181964393,
+      "grad_norm": 0.45965901017189026,
+      "learning_rate": 9.985190892985726e-05,
+      "loss": 1.0531,
+      "step": 602
+    },
+    {
+      "epoch": 0.0538789733509058,
+      "grad_norm": 0.45686468482017517,
+      "learning_rate": 9.985079392533974e-05,
+      "loss": 1.0928,
+      "step": 603
+    },
+    {
+      "epoch": 0.05396832488216767,
+      "grad_norm": 0.46496498584747314,
+      "learning_rate": 9.984967474528684e-05,
+      "loss": 0.9913,
+      "step": 604
+    },
+    {
+      "epoch": 0.05405767641342953,
+      "grad_norm": 0.41646531224250793,
+      "learning_rate": 9.984855138979233e-05,
+      "loss": 1.1572,
+      "step": 605
+    },
+    {
+      "epoch": 0.0541470279446914,
+      "grad_norm": 0.46098536252975464,
+      "learning_rate": 9.984742385895029e-05,
+      "loss": 1.0705,
+      "step": 606
+    },
+    {
+      "epoch": 0.05423637947595327,
+      "grad_norm": 0.44525283575057983,
+      "learning_rate": 9.984629215285516e-05,
+      "loss": 1.1466,
+      "step": 607
+    },
+    {
+      "epoch": 0.05432573100721513,
+      "grad_norm": 0.5084784030914307,
+      "learning_rate": 9.984515627160176e-05,
+      "loss": 1.0186,
+      "step": 608
+    },
+    {
+      "epoch": 0.054415082538477,
+      "grad_norm": 0.4281807839870453,
+      "learning_rate": 9.98440162152852e-05,
+      "loss": 1.1293,
+      "step": 609
+    },
+    {
+      "epoch": 0.05450443406973887,
+      "grad_norm": 0.44532129168510437,
+      "learning_rate": 9.984287198400098e-05,
+      "loss": 1.1377,
+      "step": 610
+    },
+    {
+      "epoch": 0.054593785601000734,
+      "grad_norm": 0.5224083662033081,
+      "learning_rate": 9.984172357784495e-05,
+      "loss": 1.1273,
+      "step": 611
+    },
+    {
+      "epoch": 0.0546831371322626,
+      "grad_norm": 0.5680818557739258,
+      "learning_rate": 9.984057099691329e-05,
+      "loss": 1.1502,
+      "step": 612
+    },
+    {
+      "epoch": 0.05477248866352447,
+      "grad_norm": 0.46198832988739014,
+      "learning_rate": 9.983941424130255e-05,
+      "loss": 1.1101,
+      "step": 613
+    },
+    {
+      "epoch": 0.054861840194786335,
+      "grad_norm": 0.4595877528190613,
+      "learning_rate": 9.983825331110961e-05,
+      "loss": 1.1089,
+      "step": 614
+    },
+    {
+      "epoch": 0.054951191726048204,
+      "grad_norm": 0.45024290680885315,
+      "learning_rate": 9.983708820643173e-05,
+      "loss": 1.116,
+      "step": 615
+    },
+    {
+      "epoch": 0.05504054325731007,
+      "grad_norm": 0.4610383212566376,
+      "learning_rate": 9.983591892736647e-05,
+      "loss": 1.0447,
+      "step": 616
+    },
+    {
+      "epoch": 0.05512989478857194,
+      "grad_norm": 0.5474917888641357,
+      "learning_rate": 9.983474547401182e-05,
+      "loss": 1.0719,
+      "step": 617
+    },
+    {
+      "epoch": 0.055219246319833805,
+      "grad_norm": 0.3890456259250641,
+      "learning_rate": 9.9833567846466e-05,
+      "loss": 1.0726,
+      "step": 618
+    },
+    {
+      "epoch": 0.055308597851095674,
+      "grad_norm": 0.495962917804718,
+      "learning_rate": 9.983238604482771e-05,
+      "loss": 1.1493,
+      "step": 619
+    },
+    {
+      "epoch": 0.05539794938235754,
+      "grad_norm": 0.46767839789390564,
+      "learning_rate": 9.983120006919591e-05,
+      "loss": 1.1153,
+      "step": 620
+    },
+    {
+      "epoch": 0.055487300913619406,
+      "grad_norm": 0.49606284499168396,
+      "learning_rate": 9.983000991966993e-05,
+      "loss": 1.1023,
+      "step": 621
+    },
+    {
+      "epoch": 0.055576652444881275,
+      "grad_norm": 0.45764997601509094,
+      "learning_rate": 9.982881559634947e-05,
+      "loss": 1.0941,
+      "step": 622
+    },
+    {
+      "epoch": 0.055666003976143144,
+      "grad_norm": 0.5058903694152832,
+      "learning_rate": 9.982761709933457e-05,
+      "loss": 1.0503,
+      "step": 623
+    },
+    {
+      "epoch": 0.055755355507405006,
+      "grad_norm": 0.4385363757610321,
+      "learning_rate": 9.982641442872562e-05,
+      "loss": 1.1275,
+      "step": 624
+    },
+    {
+      "epoch": 0.055844707038666876,
+      "grad_norm": 0.43508297204971313,
+      "learning_rate": 9.982520758462335e-05,
+      "loss": 1.1353,
+      "step": 625
+    },
+    {
+      "epoch": 0.055934058569928745,
+      "grad_norm": 0.49032166600227356,
+      "learning_rate": 9.982399656712884e-05,
+      "loss": 1.0699,
+      "step": 626
+    },
+    {
+      "epoch": 0.05602341010119061,
+      "grad_norm": 0.5139771699905396,
+      "learning_rate": 9.982278137634353e-05,
+      "loss": 1.0745,
+      "step": 627
+    },
+    {
+      "epoch": 0.056112761632452476,
+      "grad_norm": 0.4978711009025574,
+      "learning_rate": 9.982156201236921e-05,
+      "loss": 1.0771,
+      "step": 628
+    },
+    {
+      "epoch": 0.056202113163714346,
+      "grad_norm": 0.45232483744621277,
+      "learning_rate": 9.9820338475308e-05,
+      "loss": 1.1367,
+      "step": 629
+    },
+    {
+      "epoch": 0.05629146469497621,
+      "grad_norm": 0.5391058921813965,
+      "learning_rate": 9.981911076526243e-05,
+      "loss": 0.9913,
+      "step": 630
+    },
+    {
+      "epoch": 0.05638081622623808,
+      "grad_norm": 0.4652855396270752,
+      "learning_rate": 9.981787888233527e-05,
+      "loss": 1.0855,
+      "step": 631
+    },
+    {
+      "epoch": 0.056470167757499946,
+      "grad_norm": 0.4794527590274811,
+      "learning_rate": 9.981664282662974e-05,
+      "loss": 1.2006,
+      "step": 632
+    },
+    {
+      "epoch": 0.05655951928876181,
+      "grad_norm": 0.4287259578704834,
+      "learning_rate": 9.981540259824938e-05,
+      "loss": 1.0524,
+      "step": 633
+    },
+    {
+      "epoch": 0.05664887082002368,
+      "grad_norm": 0.4706929922103882,
+      "learning_rate": 9.981415819729804e-05,
+      "loss": 0.9986,
+      "step": 634
+    },
+    {
+      "epoch": 0.05673822235128555,
+      "grad_norm": 0.5352094769477844,
+      "learning_rate": 9.981290962387998e-05,
+      "loss": 1.0363,
+      "step": 635
+    },
+    {
+      "epoch": 0.05682757388254741,
+      "grad_norm": 0.4566076695919037,
+      "learning_rate": 9.981165687809976e-05,
+      "loss": 1.089,
+      "step": 636
+    },
+    {
+      "epoch": 0.05691692541380928,
+      "grad_norm": 0.5029745697975159,
+      "learning_rate": 9.981039996006234e-05,
+      "loss": 0.9928,
+      "step": 637
+    },
+    {
+      "epoch": 0.05700627694507115,
+      "grad_norm": 0.49635255336761475,
+      "learning_rate": 9.980913886987296e-05,
+      "loss": 1.1236,
+      "step": 638
+    },
+    {
+      "epoch": 0.05709562847633301,
+      "grad_norm": 0.5090769529342651,
+      "learning_rate": 9.98078736076373e-05,
+      "loss": 1.0525,
+      "step": 639
+    },
+    {
+      "epoch": 0.05718498000759488,
+      "grad_norm": 0.46991288661956787,
+      "learning_rate": 9.980660417346129e-05,
+      "loss": 1.1371,
+      "step": 640
+    },
+    {
+      "epoch": 0.05727433153885675,
+      "grad_norm": 0.5006728172302246,
+      "learning_rate": 9.980533056745128e-05,
+      "loss": 1.0505,
+      "step": 641
+    },
+    {
+      "epoch": 0.05736368307011861,
+      "grad_norm": 0.45537087321281433,
+      "learning_rate": 9.980405278971396e-05,
+      "loss": 1.1296,
+      "step": 642
+    },
+    {
+      "epoch": 0.05745303460138048,
+      "grad_norm": 0.4972824454307556,
+      "learning_rate": 9.980277084035634e-05,
+      "loss": 1.1512,
+      "step": 643
+    },
+    {
+      "epoch": 0.05754238613264235,
+      "grad_norm": 0.6011673212051392,
+      "learning_rate": 9.980148471948581e-05,
+      "loss": 1.0242,
+      "step": 644
+    },
+    {
+      "epoch": 0.05763173766390421,
+      "grad_norm": 0.4560789167881012,
+      "learning_rate": 9.980019442721008e-05,
+      "loss": 1.1034,
+      "step": 645
+    },
+    {
+      "epoch": 0.05772108919516608,
+      "grad_norm": 0.46284574270248413,
+      "learning_rate": 9.979889996363723e-05,
+      "loss": 1.1125,
+      "step": 646
+    },
+    {
+      "epoch": 0.05781044072642795,
+      "grad_norm": 0.5078014731407166,
+      "learning_rate": 9.97976013288757e-05,
+      "loss": 1.0878,
+      "step": 647
+    },
+    {
+      "epoch": 0.05789979225768982,
+      "grad_norm": 0.4990587830543518,
+      "learning_rate": 9.979629852303426e-05,
+      "loss": 1.0623,
+      "step": 648
+    },
+    {
+      "epoch": 0.05798914378895168,
+      "grad_norm": 0.4506543278694153,
+      "learning_rate": 9.979499154622201e-05,
+      "loss": 1.1416,
+      "step": 649
+    },
+    {
+      "epoch": 0.05807849532021355,
+      "grad_norm": 0.4424755573272705,
+      "learning_rate": 9.979368039854847e-05,
+      "loss": 1.094,
+      "step": 650
+    },
+    {
+      "epoch": 0.05816784685147542,
+      "grad_norm": 0.4033477008342743,
+      "learning_rate": 9.979236508012341e-05,
+      "loss": 1.13,
+      "step": 651
+    },
+    {
+      "epoch": 0.05825719838273728,
+      "grad_norm": 0.43958303332328796,
+      "learning_rate": 9.979104559105703e-05,
+      "loss": 1.1477,
+      "step": 652
+    },
+    {
+      "epoch": 0.05834654991399915,
+      "grad_norm": 0.44301077723503113,
+      "learning_rate": 9.978972193145986e-05,
+      "loss": 1.1097,
+      "step": 653
+    },
+    {
+      "epoch": 0.05843590144526102,
+      "grad_norm": 0.44206470251083374,
+      "learning_rate": 9.978839410144274e-05,
+      "loss": 1.0702,
+      "step": 654
+    },
+    {
+      "epoch": 0.05852525297652288,
+      "grad_norm": 0.47731223702430725,
+      "learning_rate": 9.978706210111692e-05,
+      "loss": 1.109,
+      "step": 655
+    },
+    {
+      "epoch": 0.05861460450778475,
+      "grad_norm": 0.47083383798599243,
+      "learning_rate": 9.978572593059394e-05,
+      "loss": 1.1559,
+      "step": 656
+    },
+    {
+      "epoch": 0.05870395603904662,
+      "grad_norm": 0.5734443068504333,
+      "learning_rate": 9.978438558998575e-05,
+      "loss": 1.0478,
+      "step": 657
+    },
+    {
+      "epoch": 0.058793307570308484,
+      "grad_norm": 0.4939190745353699,
+      "learning_rate": 9.978304107940461e-05,
+      "loss": 1.0989,
+      "step": 658
+    },
+    {
+      "epoch": 0.05888265910157035,
+      "grad_norm": 0.42895951867103577,
+      "learning_rate": 9.978169239896311e-05,
+      "loss": 1.1302,
+      "step": 659
+    },
+    {
+      "epoch": 0.05897201063283222,
+      "grad_norm": 0.513512372970581,
+      "learning_rate": 9.978033954877425e-05,
+      "loss": 1.1085,
+      "step": 660
+    },
+    {
+      "epoch": 0.059061362164094085,
+      "grad_norm": 0.4882372319698334,
+      "learning_rate": 9.977898252895134e-05,
+      "loss": 1.0566,
+      "step": 661
+    },
+    {
+      "epoch": 0.059150713695355954,
+      "grad_norm": 0.4673426151275635,
+      "learning_rate": 9.977762133960802e-05,
+      "loss": 1.028,
+      "step": 662
+    },
+    {
+      "epoch": 0.05924006522661782,
+      "grad_norm": 0.523703932762146,
+      "learning_rate": 9.977625598085834e-05,
+      "loss": 1.0267,
+      "step": 663
+    },
+    {
+      "epoch": 0.059329416757879685,
+      "grad_norm": 0.5089814066886902,
+      "learning_rate": 9.977488645281662e-05,
+      "loss": 1.1286,
+      "step": 664
+    },
+    {
+      "epoch": 0.059418768289141555,
+      "grad_norm": 0.48868119716644287,
+      "learning_rate": 9.977351275559763e-05,
+      "loss": 1.0537,
+      "step": 665
+    },
+    {
+      "epoch": 0.059508119820403424,
+      "grad_norm": 0.5465951561927795,
+      "learning_rate": 9.977213488931638e-05,
+      "loss": 1.1084,
+      "step": 666
+    },
+    {
+      "epoch": 0.059597471351665286,
+      "grad_norm": 0.49999743700027466,
+      "learning_rate": 9.97707528540883e-05,
+      "loss": 1.094,
+      "step": 667
+    },
+    {
+      "epoch": 0.059686822882927155,
+      "grad_norm": 0.5092505216598511,
+      "learning_rate": 9.976936665002916e-05,
+      "loss": 1.0616,
+      "step": 668
+    },
+    {
+      "epoch": 0.059776174414189025,
+      "grad_norm": 0.5057411193847656,
+      "learning_rate": 9.976797627725505e-05,
+      "loss": 1.0719,
+      "step": 669
+    },
+    {
+      "epoch": 0.05986552594545089,
+      "grad_norm": 0.5061649084091187,
+      "learning_rate": 9.976658173588244e-05,
+      "loss": 1.1306,
+      "step": 670
+    },
+    {
+      "epoch": 0.059954877476712756,
+      "grad_norm": 0.4602451026439667,
+      "learning_rate": 9.976518302602813e-05,
+      "loss": 1.074,
+      "step": 671
+    },
+    {
+      "epoch": 0.060044229007974625,
+      "grad_norm": 0.47477665543556213,
+      "learning_rate": 9.97637801478093e-05,
+      "loss": 1.0751,
+      "step": 672
+    },
+    {
+      "epoch": 0.06013358053923649,
+      "grad_norm": 0.4398420751094818,
+      "learning_rate": 9.976237310134342e-05,
+      "loss": 1.1357,
+      "step": 673
+    },
+    {
+      "epoch": 0.06022293207049836,
+      "grad_norm": 0.49362048506736755,
+      "learning_rate": 9.976096188674837e-05,
+      "loss": 1.1307,
+      "step": 674
+    },
+    {
+      "epoch": 0.060312283601760226,
+      "grad_norm": 0.4223507344722748,
+      "learning_rate": 9.975954650414236e-05,
+      "loss": 1.122,
+      "step": 675
+    },
+    {
+      "epoch": 0.060401635133022095,
+      "grad_norm": 0.4884486496448517,
+      "learning_rate": 9.975812695364392e-05,
+      "loss": 1.1486,
+      "step": 676
+    },
+    {
+      "epoch": 0.06049098666428396,
+      "grad_norm": 0.4124647080898285,
+      "learning_rate": 9.975670323537197e-05,
+      "loss": 1.0908,
+      "step": 677
+    },
+    {
+      "epoch": 0.06058033819554583,
+      "grad_norm": 0.5047022700309753,
+      "learning_rate": 9.975527534944574e-05,
+      "loss": 0.9936,
+      "step": 678
+    },
+    {
+      "epoch": 0.060669689726807696,
+      "grad_norm": 0.509061872959137,
+      "learning_rate": 9.975384329598486e-05,
+      "loss": 1.0482,
+      "step": 679
+    },
+    {
+      "epoch": 0.06075904125806956,
+      "grad_norm": 0.5170226693153381,
+      "learning_rate": 9.975240707510926e-05,
+      "loss": 1.0872,
+      "step": 680
+    },
+    {
+      "epoch": 0.06084839278933143,
+      "grad_norm": 0.43989554047584534,
+      "learning_rate": 9.975096668693926e-05,
+      "loss": 1.1313,
+      "step": 681
+    },
+    {
+      "epoch": 0.0609377443205933,
+      "grad_norm": 0.43840011954307556,
+      "learning_rate": 9.974952213159547e-05,
+      "loss": 1.1161,
+      "step": 682
+    },
+    {
+      "epoch": 0.06102709585185516,
+      "grad_norm": 0.5931734442710876,
+      "learning_rate": 9.974807340919893e-05,
+      "loss": 0.9472,
+      "step": 683
+    },
+    {
+      "epoch": 0.06111644738311703,
+      "grad_norm": 0.4824213683605194,
+      "learning_rate": 9.974662051987096e-05,
+      "loss": 1.0452,
+      "step": 684
+    },
+    {
+      "epoch": 0.0612057989143789,
+      "grad_norm": 0.5604041218757629,
+      "learning_rate": 9.974516346373326e-05,
+      "loss": 1.1357,
+      "step": 685
+    },
+    {
+      "epoch": 0.06129515044564076,
+      "grad_norm": 0.42983385920524597,
+      "learning_rate": 9.974370224090788e-05,
+      "loss": 1.166,
+      "step": 686
+    },
+    {
+      "epoch": 0.06138450197690263,
+      "grad_norm": 0.4323766827583313,
+      "learning_rate": 9.97422368515172e-05,
+      "loss": 1.0949,
+      "step": 687
+    },
+    {
+      "epoch": 0.0614738535081645,
+      "grad_norm": 0.4329332113265991,
+      "learning_rate": 9.974076729568396e-05,
+      "loss": 1.1318,
+      "step": 688
+    },
+    {
+      "epoch": 0.06156320503942636,
+      "grad_norm": 0.48779281973838806,
+      "learning_rate": 9.973929357353126e-05,
+      "loss": 1.1396,
+      "step": 689
+    },
+    {
+      "epoch": 0.06165255657068823,
+      "grad_norm": 0.485921710729599,
+      "learning_rate": 9.973781568518256e-05,
+      "loss": 1.0039,
+      "step": 690
+    },
+    {
+      "epoch": 0.0617419081019501,
+      "grad_norm": 0.45170992612838745,
+      "learning_rate": 9.973633363076163e-05,
+      "loss": 1.0345,
+      "step": 691
+    },
+    {
+      "epoch": 0.06183125963321196,
+      "grad_norm": 0.5609152317047119,
+      "learning_rate": 9.973484741039258e-05,
+      "loss": 1.0191,
+      "step": 692
+    },
+    {
+      "epoch": 0.06192061116447383,
+      "grad_norm": 0.4308573007583618,
+      "learning_rate": 9.973335702419995e-05,
+      "loss": 1.0891,
+      "step": 693
+    },
+    {
+      "epoch": 0.0620099626957357,
+      "grad_norm": 0.4530807137489319,
+      "learning_rate": 9.973186247230855e-05,
+      "loss": 1.1436,
+      "step": 694
+    },
+    {
+      "epoch": 0.06209931422699756,
+      "grad_norm": 0.4708486795425415,
+      "learning_rate": 9.973036375484354e-05,
+      "loss": 1.0609,
+      "step": 695
+    },
+    {
+      "epoch": 0.06218866575825943,
+      "grad_norm": 0.4028553068637848,
+      "learning_rate": 9.97288608719305e-05,
+      "loss": 1.0699,
+      "step": 696
+    },
+    {
+      "epoch": 0.0622780172895213,
+      "grad_norm": 0.4526921510696411,
+      "learning_rate": 9.97273538236953e-05,
+      "loss": 1.1414,
+      "step": 697
+    },
+    {
+      "epoch": 0.06236736882078316,
+      "grad_norm": 0.46698683500289917,
+      "learning_rate": 9.972584261026413e-05,
+      "loss": 1.1178,
+      "step": 698
+    },
+    {
+      "epoch": 0.06245672035204503,
+      "grad_norm": 0.5231309533119202,
+      "learning_rate": 9.972432723176361e-05,
+      "loss": 1.1514,
+      "step": 699
+    },
+    {
+      "epoch": 0.0625460718833069,
+      "grad_norm": 0.38658031821250916,
+      "learning_rate": 9.972280768832068e-05,
+      "loss": 1.1243,
+      "step": 700
+    },
+    {
+      "epoch": 0.06263542341456876,
+      "grad_norm": 0.4694380760192871,
+      "learning_rate": 9.972128398006259e-05,
+      "loss": 1.1258,
+      "step": 701
+    },
+    {
+      "epoch": 0.06272477494583063,
+      "grad_norm": 0.4181990623474121,
+      "learning_rate": 9.971975610711697e-05,
+      "loss": 1.1512,
+      "step": 702
+    },
+    {
+      "epoch": 0.0628141264770925,
+      "grad_norm": 0.41654473543167114,
+      "learning_rate": 9.971822406961179e-05,
+      "loss": 1.1766,
+      "step": 703
+    },
+    {
+      "epoch": 0.06290347800835437,
+      "grad_norm": 0.44843196868896484,
+      "learning_rate": 9.971668786767541e-05,
+      "loss": 1.0496,
+      "step": 704
+    },
+    {
+      "epoch": 0.06299282953961624,
+      "grad_norm": 0.5500750541687012,
+      "learning_rate": 9.971514750143647e-05,
+      "loss": 1.0574,
+      "step": 705
+    },
+    {
+      "epoch": 0.0630821810708781,
+      "grad_norm": 0.4476911127567291,
+      "learning_rate": 9.971360297102401e-05,
+      "loss": 1.0639,
+      "step": 706
+    },
+    {
+      "epoch": 0.06317153260213997,
+      "grad_norm": 0.46156466007232666,
+      "learning_rate": 9.971205427656738e-05,
+      "loss": 1.1028,
+      "step": 707
+    },
+    {
+      "epoch": 0.06326088413340183,
+      "grad_norm": 0.45186087489128113,
+      "learning_rate": 9.971050141819632e-05,
+      "loss": 1.0948,
+      "step": 708
+    },
+    {
+      "epoch": 0.0633502356646637,
+      "grad_norm": 0.4499582052230835,
+      "learning_rate": 9.970894439604088e-05,
+      "loss": 1.0978,
+      "step": 709
+    },
+    {
+      "epoch": 0.06343958719592557,
+      "grad_norm": 0.4321632385253906,
+      "learning_rate": 9.970738321023149e-05,
+      "loss": 1.1102,
+      "step": 710
+    },
+    {
+      "epoch": 0.06352893872718744,
+      "grad_norm": 0.4565223455429077,
+      "learning_rate": 9.970581786089891e-05,
+      "loss": 1.0825,
+      "step": 711
+    },
+    {
+      "epoch": 0.0636182902584493,
+      "grad_norm": 0.4617339074611664,
+      "learning_rate": 9.970424834817428e-05,
+      "loss": 1.0773,
+      "step": 712
+    },
+    {
+      "epoch": 0.06370764178971117,
+      "grad_norm": 0.4778772294521332,
+      "learning_rate": 9.970267467218904e-05,
+      "loss": 1.0918,
+      "step": 713
+    },
+    {
+      "epoch": 0.06379699332097304,
+      "grad_norm": 0.46952787041664124,
+      "learning_rate": 9.970109683307498e-05,
+      "loss": 1.0939,
+      "step": 714
+    },
+    {
+      "epoch": 0.0638863448522349,
+      "grad_norm": 0.4276919364929199,
+      "learning_rate": 9.969951483096429e-05,
+      "loss": 1.0416,
+      "step": 715
+    },
+    {
+      "epoch": 0.06397569638349677,
+      "grad_norm": 0.45656710863113403,
+      "learning_rate": 9.969792866598948e-05,
+      "loss": 1.0634,
+      "step": 716
+    },
+    {
+      "epoch": 0.06406504791475864,
+      "grad_norm": 0.4644415080547333,
+      "learning_rate": 9.96963383382834e-05,
+      "loss": 1.1223,
+      "step": 717
+    },
+    {
+      "epoch": 0.06415439944602051,
+      "grad_norm": 0.5773541331291199,
+      "learning_rate": 9.969474384797926e-05,
+      "loss": 1.0547,
+      "step": 718
+    },
+    {
+      "epoch": 0.06424375097728237,
+      "grad_norm": 0.5227424502372742,
+      "learning_rate": 9.969314519521063e-05,
+      "loss": 1.1057,
+      "step": 719
+    },
+    {
+      "epoch": 0.06433310250854424,
+      "grad_norm": 0.4427839517593384,
+      "learning_rate": 9.969154238011138e-05,
+      "loss": 1.0599,
+      "step": 720
+    },
+    {
+      "epoch": 0.0644224540398061,
+      "grad_norm": 0.472834050655365,
+      "learning_rate": 9.968993540281579e-05,
+      "loss": 1.0368,
+      "step": 721
+    },
+    {
+      "epoch": 0.06451180557106798,
+      "grad_norm": 0.4237770438194275,
+      "learning_rate": 9.968832426345845e-05,
+      "loss": 1.1564,
+      "step": 722
+    },
+    {
+      "epoch": 0.06460115710232985,
+      "grad_norm": 0.3942003548145294,
+      "learning_rate": 9.968670896217431e-05,
+      "loss": 1.1415,
+      "step": 723
+    },
+    {
+      "epoch": 0.06469050863359171,
+      "grad_norm": 0.43824538588523865,
+      "learning_rate": 9.968508949909868e-05,
+      "loss": 1.118,
+      "step": 724
+    },
+    {
+      "epoch": 0.06477986016485357,
+      "grad_norm": 0.4967290461063385,
+      "learning_rate": 9.968346587436719e-05,
+      "loss": 1.0834,
+      "step": 725
+    },
+    {
+      "epoch": 0.06486921169611544,
+      "grad_norm": 0.5201587080955505,
+      "learning_rate": 9.968183808811586e-05,
+      "loss": 1.1773,
+      "step": 726
+    },
+    {
+      "epoch": 0.06495856322737731,
+      "grad_norm": 0.42590588331222534,
+      "learning_rate": 9.968020614048101e-05,
+      "loss": 1.1176,
+      "step": 727
+    },
+    {
+      "epoch": 0.06504791475863918,
+      "grad_norm": 0.4532450735569,
+      "learning_rate": 9.967857003159933e-05,
+      "loss": 1.0961,
+      "step": 728
+    },
+    {
+      "epoch": 0.06513726628990105,
+      "grad_norm": 0.49376237392425537,
+      "learning_rate": 9.96769297616079e-05,
+      "loss": 1.1232,
+      "step": 729
+    },
+    {
+      "epoch": 0.06522661782116292,
+      "grad_norm": 0.4753684401512146,
+      "learning_rate": 9.967528533064408e-05,
+      "loss": 1.0848,
+      "step": 730
+    },
+    {
+      "epoch": 0.06531596935242477,
+      "grad_norm": 0.478083997964859,
+      "learning_rate": 9.96736367388456e-05,
+      "loss": 1.1025,
+      "step": 731
+    },
+    {
+      "epoch": 0.06540532088368664,
+      "grad_norm": 0.46657711267471313,
+      "learning_rate": 9.967198398635056e-05,
+      "loss": 1.1299,
+      "step": 732
+    },
+    {
+      "epoch": 0.06549467241494851,
+      "grad_norm": 0.4510549306869507,
+      "learning_rate": 9.96703270732974e-05,
+      "loss": 1.1,
+      "step": 733
+    },
+    {
+      "epoch": 0.06558402394621038,
+      "grad_norm": 0.4335935115814209,
+      "learning_rate": 9.96686659998249e-05,
+      "loss": 1.0771,
+      "step": 734
+    },
+    {
+      "epoch": 0.06567337547747225,
+      "grad_norm": 0.39717864990234375,
+      "learning_rate": 9.96670007660722e-05,
+      "loss": 1.1836,
+      "step": 735
+    },
+    {
+      "epoch": 0.06576272700873412,
+      "grad_norm": 0.5292769074440002,
+      "learning_rate": 9.966533137217878e-05,
+      "loss": 1.0035,
+      "step": 736
+    },
+    {
+      "epoch": 0.06585207853999597,
+      "grad_norm": 0.4478752911090851,
+      "learning_rate": 9.966365781828443e-05,
+      "loss": 1.0829,
+      "step": 737
+    },
+    {
+      "epoch": 0.06594143007125784,
+      "grad_norm": 0.4209703207015991,
+      "learning_rate": 9.966198010452939e-05,
+      "loss": 1.1075,
+      "step": 738
+    },
+    {
+      "epoch": 0.06603078160251971,
+      "grad_norm": 0.5448471903800964,
+      "learning_rate": 9.966029823105416e-05,
+      "loss": 1.0948,
+      "step": 739
+    },
+    {
+      "epoch": 0.06612013313378158,
+      "grad_norm": 0.40626657009124756,
+      "learning_rate": 9.965861219799958e-05,
+      "loss": 1.1325,
+      "step": 740
+    },
+    {
+      "epoch": 0.06620948466504345,
+      "grad_norm": 0.41559332609176636,
+      "learning_rate": 9.965692200550693e-05,
+      "loss": 1.1513,
+      "step": 741
+    },
+    {
+      "epoch": 0.06629883619630532,
+      "grad_norm": 0.4277205169200897,
+      "learning_rate": 9.965522765371777e-05,
+      "loss": 1.1243,
+      "step": 742
+    },
+    {
+      "epoch": 0.06638818772756717,
+      "grad_norm": 0.4078103303909302,
+      "learning_rate": 9.965352914277399e-05,
+      "loss": 1.1174,
+      "step": 743
+    },
+    {
+      "epoch": 0.06647753925882904,
+      "grad_norm": 0.4964846968650818,
+      "learning_rate": 9.965182647281788e-05,
+      "loss": 1.0881,
+      "step": 744
+    },
+    {
+      "epoch": 0.06656689079009091,
+      "grad_norm": 0.442452609539032,
+      "learning_rate": 9.965011964399204e-05,
+      "loss": 1.0964,
+      "step": 745
+    },
+    {
+      "epoch": 0.06665624232135278,
+      "grad_norm": 0.4539310038089752,
+      "learning_rate": 9.964840865643948e-05,
+      "loss": 1.0969,
+      "step": 746
+    },
+    {
+      "epoch": 0.06674559385261465,
+      "grad_norm": 0.4463444948196411,
+      "learning_rate": 9.964669351030345e-05,
+      "loss": 1.1043,
+      "step": 747
+    },
+    {
+      "epoch": 0.06683494538387652,
+      "grad_norm": 0.4720766246318817,
+      "learning_rate": 9.964497420572765e-05,
+      "loss": 1.0618,
+      "step": 748
+    },
+    {
+      "epoch": 0.06692429691513839,
+      "grad_norm": 0.42916223406791687,
+      "learning_rate": 9.964325074285609e-05,
+      "loss": 1.1004,
+      "step": 749
+    },
+    {
+      "epoch": 0.06701364844640025,
+      "grad_norm": 0.4502628743648529,
+      "learning_rate": 9.96415231218331e-05,
+      "loss": 1.0862,
+      "step": 750
+    },
+    {
+      "epoch": 0.06710299997766211,
+      "grad_norm": 0.4129611551761627,
+      "learning_rate": 9.963979134280343e-05,
+      "loss": 1.1337,
+      "step": 751
+    },
+    {
+      "epoch": 0.06719235150892398,
+      "grad_norm": 0.4612719714641571,
+      "learning_rate": 9.963805540591211e-05,
+      "loss": 1.1208,
+      "step": 752
+    },
+    {
+      "epoch": 0.06728170304018585,
+      "grad_norm": 0.5239313840866089,
+      "learning_rate": 9.963631531130455e-05,
+      "loss": 1.1208,
+      "step": 753
+    },
+    {
+      "epoch": 0.06737105457144772,
+      "grad_norm": 0.4685198664665222,
+      "learning_rate": 9.963457105912647e-05,
+      "loss": 1.0893,
+      "step": 754
+    },
+    {
+      "epoch": 0.06746040610270959,
+      "grad_norm": 0.4528690278530121,
+      "learning_rate": 9.963282264952403e-05,
+      "loss": 1.042,
+      "step": 755
+    },
+    {
+      "epoch": 0.06754975763397145,
+      "grad_norm": 0.5180427432060242,
+      "learning_rate": 9.963107008264364e-05,
+      "loss": 1.0768,
+      "step": 756
+    },
+    {
+      "epoch": 0.06763910916523332,
+      "grad_norm": 0.4094788432121277,
+      "learning_rate": 9.96293133586321e-05,
+      "loss": 1.0914,
+      "step": 757
+    },
+    {
+      "epoch": 0.06772846069649519,
+      "grad_norm": 0.4870153069496155,
+      "learning_rate": 9.962755247763654e-05,
+      "loss": 1.019,
+      "step": 758
+    },
+    {
+      "epoch": 0.06781781222775705,
+      "grad_norm": 0.4720255434513092,
+      "learning_rate": 9.962578743980449e-05,
+      "loss": 1.0146,
+      "step": 759
+    },
+    {
+      "epoch": 0.06790716375901892,
+      "grad_norm": 0.45057880878448486,
+      "learning_rate": 9.962401824528376e-05,
+      "loss": 1.0702,
+      "step": 760
+    },
+    {
+      "epoch": 0.06799651529028079,
+      "grad_norm": 0.44907015562057495,
+      "learning_rate": 9.962224489422254e-05,
+      "loss": 1.0297,
+      "step": 761
+    },
+    {
+      "epoch": 0.06808586682154265,
+      "grad_norm": 0.4270571172237396,
+      "learning_rate": 9.962046738676938e-05,
+      "loss": 1.0743,
+      "step": 762
+    },
+    {
+      "epoch": 0.06817521835280452,
+      "grad_norm": 0.47250601649284363,
+      "learning_rate": 9.961868572307315e-05,
+      "loss": 1.1362,
+      "step": 763
+    },
+    {
+      "epoch": 0.06826456988406639,
+      "grad_norm": 0.5045757293701172,
+      "learning_rate": 9.96168999032831e-05,
+      "loss": 1.04,
+      "step": 764
+    },
+    {
+      "epoch": 0.06835392141532826,
+      "grad_norm": 0.4600796699523926,
+      "learning_rate": 9.961510992754882e-05,
+      "loss": 1.0663,
+      "step": 765
+    },
+    {
+      "epoch": 0.06844327294659013,
+      "grad_norm": 0.4397517442703247,
+      "learning_rate": 9.961331579602022e-05,
+      "loss": 1.1174,
+      "step": 766
+    },
+    {
+      "epoch": 0.068532624477852,
+      "grad_norm": 0.4315062165260315,
+      "learning_rate": 9.961151750884758e-05,
+      "loss": 1.0979,
+      "step": 767
+    },
+    {
+      "epoch": 0.06862197600911385,
+      "grad_norm": 0.41421815752983093,
+      "learning_rate": 9.960971506618152e-05,
+      "loss": 1.0429,
+      "step": 768
+    },
+    {
+      "epoch": 0.06871132754037572,
+      "grad_norm": 0.49965426325798035,
+      "learning_rate": 9.960790846817303e-05,
+      "loss": 1.1674,
+      "step": 769
+    },
+    {
+      "epoch": 0.06880067907163759,
+      "grad_norm": 0.4487605392932892,
+      "learning_rate": 9.960609771497341e-05,
+      "loss": 1.1274,
+      "step": 770
+    },
+    {
+      "epoch": 0.06889003060289946,
+      "grad_norm": 0.5031136870384216,
+      "learning_rate": 9.960428280673435e-05,
+      "loss": 1.0788,
+      "step": 771
+    },
+    {
+      "epoch": 0.06897938213416133,
+      "grad_norm": 0.4574654996395111,
+      "learning_rate": 9.960246374360787e-05,
+      "loss": 1.1115,
+      "step": 772
+    },
+    {
+      "epoch": 0.0690687336654232,
+      "grad_norm": 0.3996691405773163,
+      "learning_rate": 9.960064052574632e-05,
+      "loss": 1.1178,
+      "step": 773
+    },
+    {
+      "epoch": 0.06915808519668505,
+      "grad_norm": 0.45163261890411377,
+      "learning_rate": 9.959881315330241e-05,
+      "loss": 1.1403,
+      "step": 774
+    },
+    {
+      "epoch": 0.06924743672794692,
+      "grad_norm": 0.48119327425956726,
+      "learning_rate": 9.959698162642923e-05,
+      "loss": 1.1177,
+      "step": 775
+    },
+    {
+      "epoch": 0.06933678825920879,
+      "grad_norm": 0.4419304132461548,
+      "learning_rate": 9.959514594528018e-05,
+      "loss": 1.0816,
+      "step": 776
+    },
+    {
+      "epoch": 0.06942613979047066,
+      "grad_norm": 0.48990410566329956,
+      "learning_rate": 9.959330611000898e-05,
+      "loss": 1.1581,
+      "step": 777
+    },
+    {
+      "epoch": 0.06951549132173253,
+      "grad_norm": 0.4900878071784973,
+      "learning_rate": 9.95914621207698e-05,
+      "loss": 1.0604,
+      "step": 778
+    },
+    {
+      "epoch": 0.0696048428529944,
+      "grad_norm": 0.473408967256546,
+      "learning_rate": 9.958961397771704e-05,
+      "loss": 1.0345,
+      "step": 779
+    },
+    {
+      "epoch": 0.06969419438425627,
+      "grad_norm": 0.4891606867313385,
+      "learning_rate": 9.958776168100555e-05,
+      "loss": 1.0432,
+      "step": 780
+    },
+    {
+      "epoch": 0.06978354591551812,
+      "grad_norm": 0.5358079075813293,
+      "learning_rate": 9.958590523079041e-05,
+      "loss": 1.1014,
+      "step": 781
+    },
+    {
+      "epoch": 0.06987289744677999,
+      "grad_norm": 0.455711305141449,
+      "learning_rate": 9.95840446272272e-05,
+      "loss": 1.0498,
+      "step": 782
+    },
+    {
+      "epoch": 0.06996224897804186,
+      "grad_norm": 0.42493581771850586,
+      "learning_rate": 9.958217987047169e-05,
+      "loss": 1.0354,
+      "step": 783
+    },
+    {
+      "epoch": 0.07005160050930373,
+      "grad_norm": 0.46666839718818665,
+      "learning_rate": 9.958031096068012e-05,
+      "loss": 1.0204,
+      "step": 784
+    },
+    {
+      "epoch": 0.0701409520405656,
+      "grad_norm": 0.40961867570877075,
+      "learning_rate": 9.957843789800902e-05,
+      "loss": 1.1588,
+      "step": 785
+    },
+    {
+      "epoch": 0.07023030357182747,
+      "grad_norm": 0.4467061758041382,
+      "learning_rate": 9.957656068261527e-05,
+      "loss": 1.0933,
+      "step": 786
+    },
+    {
+      "epoch": 0.07031965510308932,
+      "grad_norm": 0.4334403872489929,
+      "learning_rate": 9.957467931465613e-05,
+      "loss": 1.1161,
+      "step": 787
+    },
+    {
+      "epoch": 0.07040900663435119,
+      "grad_norm": 0.5627568364143372,
+      "learning_rate": 9.957279379428917e-05,
+      "loss": 1.0551,
+      "step": 788
+    },
+    {
+      "epoch": 0.07049835816561306,
+      "grad_norm": 0.422397643327713,
+      "learning_rate": 9.95709041216723e-05,
+      "loss": 1.1089,
+      "step": 789
+    },
+    {
+      "epoch": 0.07058770969687493,
+      "grad_norm": 0.5722343325614929,
+      "learning_rate": 9.956901029696384e-05,
+      "loss": 1.0943,
+      "step": 790
+    },
+    {
+      "epoch": 0.0706770612281368,
+      "grad_norm": 0.4556920826435089,
+      "learning_rate": 9.95671123203224e-05,
+      "loss": 1.0813,
+      "step": 791
+    },
+    {
+      "epoch": 0.07076641275939867,
+      "grad_norm": 0.43123292922973633,
+      "learning_rate": 9.956521019190694e-05,
+      "loss": 1.0318,
+      "step": 792
+    },
+    {
+      "epoch": 0.07085576429066053,
+      "grad_norm": 0.5033695101737976,
+      "learning_rate": 9.956330391187682e-05,
+      "loss": 1.0647,
+      "step": 793
+    },
+    {
+      "epoch": 0.0709451158219224,
+      "grad_norm": 0.45806631445884705,
+      "learning_rate": 9.956139348039168e-05,
+      "loss": 1.1287,
+      "step": 794
+    },
+    {
+      "epoch": 0.07103446735318426,
+      "grad_norm": 0.43176013231277466,
+      "learning_rate": 9.955947889761155e-05,
+      "loss": 1.061,
+      "step": 795
+    },
+    {
+      "epoch": 0.07112381888444613,
+      "grad_norm": 0.4724084138870239,
+      "learning_rate": 9.95575601636968e-05,
+      "loss": 1.0868,
+      "step": 796
+    },
+    {
+      "epoch": 0.071213170415708,
+      "grad_norm": 0.4638439416885376,
+      "learning_rate": 9.955563727880814e-05,
+      "loss": 1.0874,
+      "step": 797
+    },
+    {
+      "epoch": 0.07130252194696987,
+      "grad_norm": 0.4180169403553009,
+      "learning_rate": 9.955371024310662e-05,
+      "loss": 1.1168,
+      "step": 798
+    },
+    {
+      "epoch": 0.07139187347823173,
+      "grad_norm": 0.455720454454422,
+      "learning_rate": 9.955177905675367e-05,
+      "loss": 1.1141,
+      "step": 799
+    },
+    {
+      "epoch": 0.0714812250094936,
+      "grad_norm": 0.4505751132965088,
+      "learning_rate": 9.954984371991105e-05,
+      "loss": 1.0927,
+      "step": 800
+    },
+    {
+      "epoch": 0.07157057654075547,
+      "grad_norm": 0.4087337553501129,
+      "learning_rate": 9.954790423274085e-05,
+      "loss": 1.104,
+      "step": 801
+    },
+    {
+      "epoch": 0.07165992807201733,
+      "grad_norm": 0.46492061018943787,
+      "learning_rate": 9.954596059540553e-05,
+      "loss": 1.0794,
+      "step": 802
+    },
+    {
+      "epoch": 0.0717492796032792,
+      "grad_norm": 0.4779999554157257,
+      "learning_rate": 9.954401280806789e-05,
+      "loss": 1.0552,
+      "step": 803
+    },
+    {
+      "epoch": 0.07183863113454107,
+      "grad_norm": 0.480494886636734,
+      "learning_rate": 9.954206087089107e-05,
+      "loss": 1.0328,
+      "step": 804
+    },
+    {
+      "epoch": 0.07192798266580294,
+      "grad_norm": 0.40479815006256104,
+      "learning_rate": 9.954010478403857e-05,
+      "loss": 1.0973,
+      "step": 805
+    },
+    {
+      "epoch": 0.0720173341970648,
+      "grad_norm": 0.49492841958999634,
+      "learning_rate": 9.953814454767423e-05,
+      "loss": 1.0628,
+      "step": 806
+    },
+    {
+      "epoch": 0.07210668572832667,
+      "grad_norm": 0.49477580189704895,
+      "learning_rate": 9.953618016196224e-05,
+      "loss": 1.083,
+      "step": 807
+    },
+    {
+      "epoch": 0.07219603725958854,
+      "grad_norm": 0.4487774670124054,
+      "learning_rate": 9.953421162706717e-05,
+      "loss": 1.0459,
+      "step": 808
+    },
+    {
+      "epoch": 0.0722853887908504,
+      "grad_norm": 0.4162748157978058,
+      "learning_rate": 9.953223894315386e-05,
+      "loss": 1.1345,
+      "step": 809
+    },
+    {
+      "epoch": 0.07237474032211227,
+      "grad_norm": 0.4790276288986206,
+      "learning_rate": 9.953026211038757e-05,
+      "loss": 1.1452,
+      "step": 810
+    },
+    {
+      "epoch": 0.07246409185337414,
+      "grad_norm": 0.4182452857494354,
+      "learning_rate": 9.952828112893388e-05,
+      "loss": 1.06,
+      "step": 811
+    },
+    {
+      "epoch": 0.072553443384636,
+      "grad_norm": 0.4895060360431671,
+      "learning_rate": 9.95262959989587e-05,
+      "loss": 1.0784,
+      "step": 812
+    },
+    {
+      "epoch": 0.07264279491589787,
+      "grad_norm": 0.44416457414627075,
+      "learning_rate": 9.952430672062831e-05,
+      "loss": 1.1275,
+      "step": 813
+    },
+    {
+      "epoch": 0.07273214644715974,
+      "grad_norm": 0.3927420377731323,
+      "learning_rate": 9.952231329410936e-05,
+      "loss": 1.1797,
+      "step": 814
+    },
+    {
+      "epoch": 0.0728214979784216,
+      "grad_norm": 0.43822696805000305,
+      "learning_rate": 9.952031571956878e-05,
+      "loss": 1.074,
+      "step": 815
+    },
+    {
+      "epoch": 0.07291084950968348,
+      "grad_norm": 0.44089022278785706,
+      "learning_rate": 9.951831399717394e-05,
+      "loss": 1.0785,
+      "step": 816
+    },
+    {
+      "epoch": 0.07300020104094535,
+      "grad_norm": 0.4124377369880676,
+      "learning_rate": 9.951630812709245e-05,
+      "loss": 1.065,
+      "step": 817
+    },
+    {
+      "epoch": 0.0730895525722072,
+      "grad_norm": 0.43494531512260437,
+      "learning_rate": 9.951429810949237e-05,
+      "loss": 1.0736,
+      "step": 818
+    },
+    {
+      "epoch": 0.07317890410346907,
+      "grad_norm": 0.4180351793766022,
+      "learning_rate": 9.951228394454201e-05,
+      "loss": 1.0566,
+      "step": 819
+    },
+    {
+      "epoch": 0.07326825563473094,
+      "grad_norm": 0.4254007637500763,
+      "learning_rate": 9.951026563241014e-05,
+      "loss": 1.079,
+      "step": 820
+    },
+    {
+      "epoch": 0.07335760716599281,
+      "grad_norm": 0.42448723316192627,
+      "learning_rate": 9.950824317326577e-05,
+      "loss": 1.1101,
+      "step": 821
+    },
+    {
+      "epoch": 0.07344695869725468,
+      "grad_norm": 0.4987415075302124,
+      "learning_rate": 9.95062165672783e-05,
+      "loss": 1.0266,
+      "step": 822
+    },
+    {
+      "epoch": 0.07353631022851655,
+      "grad_norm": 0.4084853529930115,
+      "learning_rate": 9.95041858146175e-05,
+      "loss": 1.0966,
+      "step": 823
+    },
+    {
+      "epoch": 0.0736256617597784,
+      "grad_norm": 0.4334554374217987,
+      "learning_rate": 9.950215091545347e-05,
+      "loss": 1.11,
+      "step": 824
+    },
+    {
+      "epoch": 0.07371501329104027,
+      "grad_norm": 0.5075578093528748,
+      "learning_rate": 9.950011186995665e-05,
+      "loss": 1.014,
+      "step": 825
+    },
+    {
+      "epoch": 0.07380436482230214,
+      "grad_norm": 0.4870114028453827,
+      "learning_rate": 9.94980686782978e-05,
+      "loss": 1.0139,
+      "step": 826
+    },
+    {
+      "epoch": 0.07389371635356401,
+      "grad_norm": 0.44856953620910645,
+      "learning_rate": 9.949602134064812e-05,
+      "loss": 1.089,
+      "step": 827
+    },
+    {
+      "epoch": 0.07398306788482588,
+      "grad_norm": 0.48632118105888367,
+      "learning_rate": 9.949396985717904e-05,
+      "loss": 1.1153,
+      "step": 828
+    },
+    {
+      "epoch": 0.07407241941608775,
+      "grad_norm": 0.44190341234207153,
+      "learning_rate": 9.949191422806244e-05,
+      "loss": 1.2339,
+      "step": 829
+    },
+    {
+      "epoch": 0.0741617709473496,
+      "grad_norm": 0.4298502504825592,
+      "learning_rate": 9.948985445347046e-05,
+      "loss": 1.1638,
+      "step": 830
+    },
+    {
+      "epoch": 0.07425112247861147,
+      "grad_norm": 0.4637526869773865,
+      "learning_rate": 9.948779053357564e-05,
+      "loss": 1.0537,
+      "step": 831
+    },
+    {
+      "epoch": 0.07434047400987334,
+      "grad_norm": 0.4532431364059448,
+      "learning_rate": 9.948572246855086e-05,
+      "loss": 1.0724,
+      "step": 832
+    },
+    {
+      "epoch": 0.07442982554113521,
+      "grad_norm": 0.52618807554245,
+      "learning_rate": 9.948365025856936e-05,
+      "loss": 1.0945,
+      "step": 833
+    },
+    {
+      "epoch": 0.07451917707239708,
+      "grad_norm": 0.4329455494880676,
+      "learning_rate": 9.948157390380468e-05,
+      "loss": 1.0166,
+      "step": 834
+    },
+    {
+      "epoch": 0.07460852860365895,
+      "grad_norm": 0.4594184160232544,
+      "learning_rate": 9.947949340443076e-05,
+      "loss": 1.1553,
+      "step": 835
+    },
+    {
+      "epoch": 0.07469788013492082,
+      "grad_norm": 0.4881901443004608,
+      "learning_rate": 9.947740876062185e-05,
+      "loss": 1.0782,
+      "step": 836
+    },
+    {
+      "epoch": 0.07478723166618267,
+      "grad_norm": 0.5016196370124817,
+      "learning_rate": 9.947531997255256e-05,
+      "loss": 1.0892,
+      "step": 837
+    },
+    {
+      "epoch": 0.07487658319744454,
+      "grad_norm": 0.4345821738243103,
+      "learning_rate": 9.947322704039785e-05,
+      "loss": 1.1057,
+      "step": 838
+    },
+    {
+      "epoch": 0.07496593472870641,
+      "grad_norm": 0.4423231780529022,
+      "learning_rate": 9.947112996433305e-05,
+      "loss": 1.0884,
+      "step": 839
+    },
+    {
+      "epoch": 0.07505528625996828,
+      "grad_norm": 0.5169707536697388,
+      "learning_rate": 9.946902874453376e-05,
+      "loss": 1.015,
+      "step": 840
+    },
+    {
+      "epoch": 0.07514463779123015,
+      "grad_norm": 0.44377315044403076,
+      "learning_rate": 9.946692338117603e-05,
+      "loss": 1.0713,
+      "step": 841
+    },
+    {
+      "epoch": 0.07523398932249202,
+      "grad_norm": 0.4438510239124298,
+      "learning_rate": 9.94648138744362e-05,
+      "loss": 1.0522,
+      "step": 842
+    },
+    {
+      "epoch": 0.07532334085375388,
+      "grad_norm": 0.5507922768592834,
+      "learning_rate": 9.946270022449093e-05,
+      "loss": 1.0294,
+      "step": 843
+    },
+    {
+      "epoch": 0.07541269238501574,
+      "grad_norm": 0.48594483733177185,
+      "learning_rate": 9.946058243151728e-05,
+      "loss": 1.0541,
+      "step": 844
+    },
+    {
+      "epoch": 0.07550204391627761,
+      "grad_norm": 0.4724047780036926,
+      "learning_rate": 9.945846049569265e-05,
+      "loss": 1.0744,
+      "step": 845
+    },
+    {
+      "epoch": 0.07559139544753948,
+      "grad_norm": 0.5080140233039856,
+      "learning_rate": 9.945633441719476e-05,
+      "loss": 1.0812,
+      "step": 846
+    },
+    {
+      "epoch": 0.07568074697880135,
+      "grad_norm": 0.44215860962867737,
+      "learning_rate": 9.945420419620171e-05,
+      "loss": 1.1124,
+      "step": 847
+    },
+    {
+      "epoch": 0.07577009851006322,
+      "grad_norm": 0.43543803691864014,
+      "learning_rate": 9.94520698328919e-05,
+      "loss": 1.108,
+      "step": 848
+    },
+    {
+      "epoch": 0.07585945004132508,
+      "grad_norm": 0.4514320194721222,
+      "learning_rate": 9.944993132744411e-05,
+      "loss": 1.0689,
+      "step": 849
+    },
+    {
+      "epoch": 0.07594880157258695,
+      "grad_norm": 0.4318578243255615,
+      "learning_rate": 9.94477886800375e-05,
+      "loss": 1.0977,
+      "step": 850
+    },
+    {
+      "epoch": 0.07603815310384882,
+      "grad_norm": 0.4266031086444855,
+      "learning_rate": 9.944564189085149e-05,
+      "loss": 1.0782,
+      "step": 851
+    },
+    {
+      "epoch": 0.07612750463511068,
+      "grad_norm": 0.3994167149066925,
+      "learning_rate": 9.944349096006593e-05,
+      "loss": 1.1211,
+      "step": 852
+    },
+    {
+      "epoch": 0.07621685616637255,
+      "grad_norm": 0.4137265980243683,
+      "learning_rate": 9.944133588786097e-05,
+      "loss": 1.0609,
+      "step": 853
+    },
+    {
+      "epoch": 0.07630620769763442,
+      "grad_norm": 0.4087986648082733,
+      "learning_rate": 9.943917667441712e-05,
+      "loss": 1.0968,
+      "step": 854
+    },
+    {
+      "epoch": 0.07639555922889628,
+      "grad_norm": 0.4582999646663666,
+      "learning_rate": 9.943701331991524e-05,
+      "loss": 1.0594,
+      "step": 855
+    },
+    {
+      "epoch": 0.07648491076015815,
+      "grad_norm": 0.44180095195770264,
+      "learning_rate": 9.943484582453653e-05,
+      "loss": 1.0959,
+      "step": 856
+    },
+    {
+      "epoch": 0.07657426229142002,
+      "grad_norm": 0.5081307291984558,
+      "learning_rate": 9.943267418846256e-05,
+      "loss": 1.0916,
+      "step": 857
+    },
+    {
+      "epoch": 0.07666361382268189,
+      "grad_norm": 0.4701635241508484,
+      "learning_rate": 9.94304984118752e-05,
+      "loss": 1.0602,
+      "step": 858
+    },
+    {
+      "epoch": 0.07675296535394376,
+      "grad_norm": 0.42284098267555237,
+      "learning_rate": 9.942831849495671e-05,
+      "loss": 1.1025,
+      "step": 859
+    },
+    {
+      "epoch": 0.07684231688520562,
+      "grad_norm": 0.4433005154132843,
+      "learning_rate": 9.942613443788967e-05,
+      "loss": 1.1403,
+      "step": 860
+    },
+    {
+      "epoch": 0.0769316684164675,
+      "grad_norm": 0.41836610436439514,
+      "learning_rate": 9.942394624085703e-05,
+      "loss": 1.0193,
+      "step": 861
+    },
+    {
+      "epoch": 0.07702101994772935,
+      "grad_norm": 0.5095844268798828,
+      "learning_rate": 9.942175390404208e-05,
+      "loss": 1.0305,
+      "step": 862
+    },
+    {
+      "epoch": 0.07711037147899122,
+      "grad_norm": 0.4428594410419464,
+      "learning_rate": 9.941955742762843e-05,
+      "loss": 1.0768,
+      "step": 863
+    },
+    {
+      "epoch": 0.07719972301025309,
+      "grad_norm": 0.5194300413131714,
+      "learning_rate": 9.941735681180009e-05,
+      "loss": 1.0837,
+      "step": 864
+    },
+    {
+      "epoch": 0.07728907454151496,
+      "grad_norm": 0.5022509098052979,
+      "learning_rate": 9.941515205674134e-05,
+      "loss": 1.0751,
+      "step": 865
+    },
+    {
+      "epoch": 0.07737842607277683,
+      "grad_norm": 0.45329055190086365,
+      "learning_rate": 9.94129431626369e-05,
+      "loss": 1.1309,
+      "step": 866
+    },
+    {
+      "epoch": 0.0774677776040387,
+      "grad_norm": 0.44875457882881165,
+      "learning_rate": 9.941073012967174e-05,
+      "loss": 0.9794,
+      "step": 867
+    },
+    {
+      "epoch": 0.07755712913530055,
+      "grad_norm": 0.5031179189682007,
+      "learning_rate": 9.940851295803128e-05,
+      "loss": 1.086,
+      "step": 868
+    },
+    {
+      "epoch": 0.07764648066656242,
+      "grad_norm": 0.42640596628189087,
+      "learning_rate": 9.94062916479012e-05,
+      "loss": 1.068,
+      "step": 869
+    },
+    {
+      "epoch": 0.07773583219782429,
+      "grad_norm": 0.39963454008102417,
+      "learning_rate": 9.940406619946754e-05,
+      "loss": 1.1031,
+      "step": 870
+    },
+    {
+      "epoch": 0.07782518372908616,
+      "grad_norm": 0.42264753580093384,
+      "learning_rate": 9.940183661291674e-05,
+      "loss": 1.0831,
+      "step": 871
+    },
+    {
+      "epoch": 0.07791453526034803,
+      "grad_norm": 0.45042434334754944,
+      "learning_rate": 9.939960288843553e-05,
+      "loss": 1.0708,
+      "step": 872
+    },
+    {
+      "epoch": 0.0780038867916099,
+      "grad_norm": 0.41273242235183716,
+      "learning_rate": 9.939736502621104e-05,
+      "loss": 1.122,
+      "step": 873
+    },
+    {
+      "epoch": 0.07809323832287175,
+      "grad_norm": 0.43225687742233276,
+      "learning_rate": 9.939512302643066e-05,
+      "loss": 1.1367,
+      "step": 874
+    },
+    {
+      "epoch": 0.07818258985413362,
+      "grad_norm": 0.4413902759552002,
+      "learning_rate": 9.939287688928223e-05,
+      "loss": 1.0814,
+      "step": 875
+    },
+    {
+      "epoch": 0.07827194138539549,
+      "grad_norm": 0.4078238308429718,
+      "learning_rate": 9.939062661495386e-05,
+      "loss": 1.0772,
+      "step": 876
+    },
+    {
+      "epoch": 0.07836129291665736,
+      "grad_norm": 0.3733663260936737,
+      "learning_rate": 9.938837220363406e-05,
+      "loss": 1.1371,
+      "step": 877
+    },
+    {
+      "epoch": 0.07845064444791923,
+      "grad_norm": 0.39089709520339966,
+      "learning_rate": 9.938611365551164e-05,
+      "loss": 1.0742,
+      "step": 878
+    },
+    {
+      "epoch": 0.0785399959791811,
+      "grad_norm": 0.41924816370010376,
+      "learning_rate": 9.938385097077579e-05,
+      "loss": 1.1673,
+      "step": 879
+    },
+    {
+      "epoch": 0.07862934751044295,
+      "grad_norm": 0.5209245681762695,
+      "learning_rate": 9.9381584149616e-05,
+      "loss": 1.0325,
+      "step": 880
+    },
+    {
+      "epoch": 0.07871869904170482,
+      "grad_norm": 0.38127946853637695,
+      "learning_rate": 9.937931319222218e-05,
+      "loss": 1.1191,
+      "step": 881
+    },
+    {
+      "epoch": 0.07880805057296669,
+      "grad_norm": 0.4516022503376007,
+      "learning_rate": 9.937703809878455e-05,
+      "loss": 1.0864,
+      "step": 882
+    },
+    {
+      "epoch": 0.07889740210422856,
+      "grad_norm": 0.5131936073303223,
+      "learning_rate": 9.937475886949364e-05,
+      "loss": 1.0033,
+      "step": 883
+    },
+    {
+      "epoch": 0.07898675363549043,
+      "grad_norm": 0.3989183008670807,
+      "learning_rate": 9.937247550454039e-05,
+      "loss": 1.1386,
+      "step": 884
+    },
+    {
+      "epoch": 0.0790761051667523,
+      "grad_norm": 0.4165286123752594,
+      "learning_rate": 9.937018800411604e-05,
+      "loss": 1.131,
+      "step": 885
+    },
+    {
+      "epoch": 0.07916545669801416,
+      "grad_norm": 0.49382612109184265,
+      "learning_rate": 9.936789636841219e-05,
+      "loss": 1.0344,
+      "step": 886
+    },
+    {
+      "epoch": 0.07925480822927602,
+      "grad_norm": 0.4288516938686371,
+      "learning_rate": 9.93656005976208e-05,
+      "loss": 1.0715,
+      "step": 887
+    },
+    {
+      "epoch": 0.0793441597605379,
+      "grad_norm": 0.5689799785614014,
+      "learning_rate": 9.936330069193415e-05,
+      "loss": 1.0104,
+      "step": 888
+    },
+    {
+      "epoch": 0.07943351129179976,
+      "grad_norm": 0.40826788544654846,
+      "learning_rate": 9.936099665154491e-05,
+      "loss": 1.1411,
+      "step": 889
+    },
+    {
+      "epoch": 0.07952286282306163,
+      "grad_norm": 0.4648611843585968,
+      "learning_rate": 9.935868847664605e-05,
+      "loss": 1.1341,
+      "step": 890
+    },
+    {
+      "epoch": 0.0796122143543235,
+      "grad_norm": 0.45681998133659363,
+      "learning_rate": 9.935637616743089e-05,
+      "loss": 1.1331,
+      "step": 891
+    },
+    {
+      "epoch": 0.07970156588558537,
+      "grad_norm": 0.4700930416584015,
+      "learning_rate": 9.935405972409313e-05,
+      "loss": 1.0357,
+      "step": 892
+    },
+    {
+      "epoch": 0.07979091741684723,
+      "grad_norm": 0.42509207129478455,
+      "learning_rate": 9.93517391468268e-05,
+      "loss": 1.1024,
+      "step": 893
+    },
+    {
+      "epoch": 0.0798802689481091,
+      "grad_norm": 0.46535390615463257,
+      "learning_rate": 9.934941443582626e-05,
+      "loss": 1.1248,
+      "step": 894
+    },
+    {
+      "epoch": 0.07996962047937096,
+      "grad_norm": 0.5609625577926636,
+      "learning_rate": 9.934708559128622e-05,
+      "loss": 0.947,
+      "step": 895
+    },
+    {
+      "epoch": 0.08005897201063283,
+      "grad_norm": 0.5007188320159912,
+      "learning_rate": 9.934475261340177e-05,
+      "loss": 1.1,
+      "step": 896
+    },
+    {
+      "epoch": 0.0801483235418947,
+      "grad_norm": 0.5065818428993225,
+      "learning_rate": 9.934241550236831e-05,
+      "loss": 1.0299,
+      "step": 897
+    },
+    {
+      "epoch": 0.08023767507315657,
+      "grad_norm": 0.44581353664398193,
+      "learning_rate": 9.934007425838161e-05,
+      "loss": 1.1251,
+      "step": 898
+    },
+    {
+      "epoch": 0.08032702660441843,
+      "grad_norm": 0.4115321636199951,
+      "learning_rate": 9.933772888163776e-05,
+      "loss": 1.1807,
+      "step": 899
+    },
+    {
+      "epoch": 0.0804163781356803,
+      "grad_norm": 0.47303566336631775,
+      "learning_rate": 9.933537937233321e-05,
+      "loss": 1.1472,
+      "step": 900
+    },
+    {
+      "epoch": 0.08050572966694217,
+      "grad_norm": 0.4314580261707306,
+      "learning_rate": 9.933302573066477e-05,
+      "loss": 1.089,
+      "step": 901
+    },
+    {
+      "epoch": 0.08059508119820404,
+      "grad_norm": 0.47710415720939636,
+      "learning_rate": 9.933066795682955e-05,
+      "loss": 1.0226,
+      "step": 902
+    },
+    {
+      "epoch": 0.0806844327294659,
+      "grad_norm": 0.4161222577095032,
+      "learning_rate": 9.932830605102508e-05,
+      "loss": 1.0913,
+      "step": 903
+    },
+    {
+      "epoch": 0.08077378426072777,
+      "grad_norm": 0.485124409198761,
+      "learning_rate": 9.932594001344918e-05,
+      "loss": 1.0963,
+      "step": 904
+    },
+    {
+      "epoch": 0.08086313579198963,
+      "grad_norm": 0.47519686818122864,
+      "learning_rate": 9.93235698443e-05,
+      "loss": 1.0747,
+      "step": 905
+    },
+    {
+      "epoch": 0.0809524873232515,
+      "grad_norm": 0.4680500328540802,
+      "learning_rate": 9.932119554377611e-05,
+      "loss": 1.1009,
+      "step": 906
+    },
+    {
+      "epoch": 0.08104183885451337,
+      "grad_norm": 0.49420708417892456,
+      "learning_rate": 9.931881711207638e-05,
+      "loss": 1.0773,
+      "step": 907
+    },
+    {
+      "epoch": 0.08113119038577524,
+      "grad_norm": 0.3618301749229431,
+      "learning_rate": 9.93164345494e-05,
+      "loss": 1.0759,
+      "step": 908
+    },
+    {
+      "epoch": 0.0812205419170371,
+      "grad_norm": 0.4082967936992645,
+      "learning_rate": 9.931404785594656e-05,
+      "loss": 1.133,
+      "step": 909
+    },
+    {
+      "epoch": 0.08130989344829898,
+      "grad_norm": 0.5318341851234436,
+      "learning_rate": 9.931165703191595e-05,
+      "loss": 1.0725,
+      "step": 910
+    },
+    {
+      "epoch": 0.08139924497956083,
+      "grad_norm": 0.46658989787101746,
+      "learning_rate": 9.930926207750845e-05,
+      "loss": 1.0376,
+      "step": 911
+    },
+    {
+      "epoch": 0.0814885965108227,
+      "grad_norm": 0.5218292474746704,
+      "learning_rate": 9.930686299292464e-05,
+      "loss": 1.0949,
+      "step": 912
+    },
+    {
+      "epoch": 0.08157794804208457,
+      "grad_norm": 0.48732948303222656,
+      "learning_rate": 9.930445977836548e-05,
+      "loss": 1.0491,
+      "step": 913
+    },
+    {
+      "epoch": 0.08166729957334644,
+      "grad_norm": 0.40489867329597473,
+      "learning_rate": 9.930205243403229e-05,
+      "loss": 1.0856,
+      "step": 914
+    },
+    {
+      "epoch": 0.08175665110460831,
+      "grad_norm": 0.5249372720718384,
+      "learning_rate": 9.929964096012668e-05,
+      "loss": 1.0358,
+      "step": 915
+    },
+    {
+      "epoch": 0.08184600263587018,
+      "grad_norm": 0.4804689288139343,
+      "learning_rate": 9.929722535685062e-05,
+      "loss": 1.0793,
+      "step": 916
+    },
+    {
+      "epoch": 0.08193535416713203,
+      "grad_norm": 0.43911850452423096,
+      "learning_rate": 9.929480562440649e-05,
+      "loss": 1.1816,
+      "step": 917
+    },
+    {
+      "epoch": 0.0820247056983939,
+      "grad_norm": 0.4654462933540344,
+      "learning_rate": 9.929238176299693e-05,
+      "loss": 1.1595,
+      "step": 918
+    },
+    {
+      "epoch": 0.08211405722965577,
+      "grad_norm": 0.44041964411735535,
+      "learning_rate": 9.928995377282498e-05,
+      "loss": 1.0732,
+      "step": 919
+    },
+    {
+      "epoch": 0.08220340876091764,
+      "grad_norm": 0.4381254315376282,
+      "learning_rate": 9.928752165409401e-05,
+      "loss": 1.0728,
+      "step": 920
+    },
+    {
+      "epoch": 0.08229276029217951,
+      "grad_norm": 0.5617743730545044,
+      "learning_rate": 9.928508540700774e-05,
+      "loss": 1.0313,
+      "step": 921
+    },
+    {
+      "epoch": 0.08238211182344138,
+      "grad_norm": 0.4664022922515869,
+      "learning_rate": 9.928264503177023e-05,
+      "loss": 1.0456,
+      "step": 922
+    },
+    {
+      "epoch": 0.08247146335470325,
+      "grad_norm": 0.4458063542842865,
+      "learning_rate": 9.928020052858587e-05,
+      "loss": 1.0782,
+      "step": 923
+    },
+    {
+      "epoch": 0.0825608148859651,
+      "grad_norm": 0.44314441084861755,
+      "learning_rate": 9.927775189765943e-05,
+      "loss": 1.0905,
+      "step": 924
+    },
+    {
+      "epoch": 0.08265016641722697,
+      "grad_norm": 0.4562947154045105,
+      "learning_rate": 9.927529913919601e-05,
+      "loss": 1.0686,
+      "step": 925
+    },
+    {
+      "epoch": 0.08273951794848884,
+      "grad_norm": 0.4639774560928345,
+      "learning_rate": 9.927284225340105e-05,
+      "loss": 1.0761,
+      "step": 926
+    },
+    {
+      "epoch": 0.08282886947975071,
+      "grad_norm": 0.4482592046260834,
+      "learning_rate": 9.927038124048034e-05,
+      "loss": 1.0584,
+      "step": 927
+    },
+    {
+      "epoch": 0.08291822101101258,
+      "grad_norm": 0.4356086254119873,
+      "learning_rate": 9.926791610064002e-05,
+      "loss": 1.0996,
+      "step": 928
+    },
+    {
+      "epoch": 0.08300757254227445,
+      "grad_norm": 0.3952994644641876,
+      "learning_rate": 9.926544683408656e-05,
+      "loss": 1.1176,
+      "step": 929
+    },
+    {
+      "epoch": 0.0830969240735363,
+      "grad_norm": 0.4380575716495514,
+      "learning_rate": 9.92629734410268e-05,
+      "loss": 1.0432,
+      "step": 930
+    },
+    {
+      "epoch": 0.08318627560479817,
+      "grad_norm": 0.54874587059021,
+      "learning_rate": 9.92604959216679e-05,
+      "loss": 0.9305,
+      "step": 931
+    },
+    {
+      "epoch": 0.08327562713606004,
+      "grad_norm": 0.5379965901374817,
+      "learning_rate": 9.925801427621739e-05,
+      "loss": 1.1302,
+      "step": 932
+    },
+    {
+      "epoch": 0.08336497866732191,
+      "grad_norm": 0.4604237377643585,
+      "learning_rate": 9.925552850488314e-05,
+      "loss": 1.0884,
+      "step": 933
+    },
+    {
+      "epoch": 0.08345433019858378,
+      "grad_norm": 0.4671541750431061,
+      "learning_rate": 9.925303860787335e-05,
+      "loss": 0.9864,
+      "step": 934
+    },
+    {
+      "epoch": 0.08354368172984565,
+      "grad_norm": 0.48653197288513184,
+      "learning_rate": 9.925054458539658e-05,
+      "loss": 1.1002,
+      "step": 935
+    },
+    {
+      "epoch": 0.0836330332611075,
+      "grad_norm": 0.4400951564311981,
+      "learning_rate": 9.924804643766172e-05,
+      "loss": 1.1034,
+      "step": 936
+    },
+    {
+      "epoch": 0.08372238479236938,
+      "grad_norm": 0.4999461770057678,
+      "learning_rate": 9.924554416487802e-05,
+      "loss": 1.1624,
+      "step": 937
+    },
+    {
+      "epoch": 0.08381173632363124,
+      "grad_norm": 0.45174461603164673,
+      "learning_rate": 9.92430377672551e-05,
+      "loss": 1.0431,
+      "step": 938
+    },
+    {
+      "epoch": 0.08390108785489311,
+      "grad_norm": 0.4591819643974304,
+      "learning_rate": 9.924052724500284e-05,
+      "loss": 1.0554,
+      "step": 939
+    },
+    {
+      "epoch": 0.08399043938615498,
+      "grad_norm": 0.4611966609954834,
+      "learning_rate": 9.923801259833159e-05,
+      "loss": 1.0504,
+      "step": 940
+    },
+    {
+      "epoch": 0.08407979091741685,
+      "grad_norm": 0.3897397518157959,
+      "learning_rate": 9.923549382745192e-05,
+      "loss": 1.0838,
+      "step": 941
+    },
+    {
+      "epoch": 0.08416914244867871,
+      "grad_norm": 0.3815537095069885,
+      "learning_rate": 9.923297093257485e-05,
+      "loss": 1.095,
+      "step": 942
+    },
+    {
+      "epoch": 0.08425849397994058,
+      "grad_norm": 0.4189860224723816,
+      "learning_rate": 9.923044391391165e-05,
+      "loss": 1.0924,
+      "step": 943
+    },
+    {
+      "epoch": 0.08434784551120245,
+      "grad_norm": 0.45325958728790283,
+      "learning_rate": 9.922791277167404e-05,
+      "loss": 1.0615,
+      "step": 944
+    },
+    {
+      "epoch": 0.08443719704246432,
+      "grad_norm": 0.4996306300163269,
+      "learning_rate": 9.9225377506074e-05,
+      "loss": 1.0995,
+      "step": 945
+    },
+    {
+      "epoch": 0.08452654857372618,
+      "grad_norm": 0.4180799424648285,
+      "learning_rate": 9.922283811732388e-05,
+      "loss": 1.1514,
+      "step": 946
+    },
+    {
+      "epoch": 0.08461590010498805,
+      "grad_norm": 0.433929443359375,
+      "learning_rate": 9.92202946056364e-05,
+      "loss": 1.056,
+      "step": 947
+    },
+    {
+      "epoch": 0.08470525163624992,
+      "grad_norm": 0.479571133852005,
+      "learning_rate": 9.921774697122459e-05,
+      "loss": 1.0351,
+      "step": 948
+    },
+    {
+      "epoch": 0.08479460316751178,
+      "grad_norm": 0.46040210127830505,
+      "learning_rate": 9.921519521430185e-05,
+      "loss": 0.9939,
+      "step": 949
+    },
+    {
+      "epoch": 0.08488395469877365,
+      "grad_norm": 0.43157997727394104,
+      "learning_rate": 9.92126393350819e-05,
+      "loss": 1.0588,
+      "step": 950
+    },
+    {
+      "epoch": 0.08497330623003552,
+      "grad_norm": 0.4109559655189514,
+      "learning_rate": 9.921007933377887e-05,
+      "loss": 1.1222,
+      "step": 951
+    },
+    {
+      "epoch": 0.08506265776129739,
+      "grad_norm": 0.4423241913318634,
+      "learning_rate": 9.920751521060712e-05,
+      "loss": 1.0276,
+      "step": 952
+    },
+    {
+      "epoch": 0.08515200929255926,
+      "grad_norm": 0.454683393239975,
+      "learning_rate": 9.920494696578146e-05,
+      "loss": 1.0437,
+      "step": 953
+    },
+    {
+      "epoch": 0.08524136082382112,
+      "grad_norm": 0.47767403721809387,
+      "learning_rate": 9.920237459951702e-05,
+      "loss": 1.0884,
+      "step": 954
+    },
+    {
+      "epoch": 0.08533071235508298,
+      "grad_norm": 0.47979000210762024,
+      "learning_rate": 9.919979811202923e-05,
+      "loss": 1.0598,
+      "step": 955
+    },
+    {
+      "epoch": 0.08542006388634485,
+      "grad_norm": 0.5115445852279663,
+      "learning_rate": 9.919721750353395e-05,
+      "loss": 1.024,
+      "step": 956
+    },
+    {
+      "epoch": 0.08550941541760672,
+      "grad_norm": 0.4317951798439026,
+      "learning_rate": 9.919463277424727e-05,
+      "loss": 1.1506,
+      "step": 957
+    },
+    {
+      "epoch": 0.08559876694886859,
+      "grad_norm": 0.4962863624095917,
+      "learning_rate": 9.919204392438573e-05,
+      "loss": 1.0262,
+      "step": 958
+    },
+    {
+      "epoch": 0.08568811848013046,
+      "grad_norm": 0.44821789860725403,
+      "learning_rate": 9.918945095416616e-05,
+      "loss": 1.0722,
+      "step": 959
+    },
+    {
+      "epoch": 0.08577747001139233,
+      "grad_norm": 0.44719305634498596,
+      "learning_rate": 9.918685386380573e-05,
+      "loss": 1.0362,
+      "step": 960
+    },
+    {
+      "epoch": 0.08586682154265418,
+      "grad_norm": 0.4721403121948242,
+      "learning_rate": 9.918425265352202e-05,
+      "loss": 0.972,
+      "step": 961
+    },
+    {
+      "epoch": 0.08595617307391605,
+      "grad_norm": 0.43076246976852417,
+      "learning_rate": 9.918164732353288e-05,
+      "loss": 1.0746,
+      "step": 962
+    },
+    {
+      "epoch": 0.08604552460517792,
+      "grad_norm": 0.4410285949707031,
+      "learning_rate": 9.917903787405653e-05,
+      "loss": 1.0802,
+      "step": 963
+    },
+    {
+      "epoch": 0.08613487613643979,
+      "grad_norm": 0.4716387689113617,
+      "learning_rate": 9.917642430531155e-05,
+      "loss": 1.165,
+      "step": 964
+    },
+    {
+      "epoch": 0.08622422766770166,
+      "grad_norm": 0.38331499695777893,
+      "learning_rate": 9.917380661751685e-05,
+      "loss": 1.0947,
+      "step": 965
+    },
+    {
+      "epoch": 0.08631357919896353,
+      "grad_norm": 0.47889336943626404,
+      "learning_rate": 9.917118481089169e-05,
+      "loss": 1.0482,
+      "step": 966
+    },
+    {
+      "epoch": 0.08640293073022538,
+      "grad_norm": 0.4541834890842438,
+      "learning_rate": 9.916855888565569e-05,
+      "loss": 1.0942,
+      "step": 967
+    },
+    {
+      "epoch": 0.08649228226148725,
+      "grad_norm": 0.4245148003101349,
+      "learning_rate": 9.916592884202878e-05,
+      "loss": 1.0635,
+      "step": 968
+    },
+    {
+      "epoch": 0.08658163379274912,
+      "grad_norm": 0.40217721462249756,
+      "learning_rate": 9.916329468023124e-05,
+      "loss": 1.0662,
+      "step": 969
+    },
+    {
+      "epoch": 0.08667098532401099,
+      "grad_norm": 0.4782769978046417,
+      "learning_rate": 9.916065640048374e-05,
+      "loss": 1.0232,
+      "step": 970
+    },
+    {
+      "epoch": 0.08676033685527286,
+      "grad_norm": 0.4472961127758026,
+      "learning_rate": 9.915801400300727e-05,
+      "loss": 1.0723,
+      "step": 971
+    },
+    {
+      "epoch": 0.08684968838653473,
+      "grad_norm": 0.4397634267807007,
+      "learning_rate": 9.91553674880231e-05,
+      "loss": 1.1551,
+      "step": 972
+    },
+    {
+      "epoch": 0.08693903991779658,
+      "grad_norm": 0.481342077255249,
+      "learning_rate": 9.915271685575297e-05,
+      "loss": 1.1133,
+      "step": 973
+    },
+    {
+      "epoch": 0.08702839144905845,
+      "grad_norm": 0.3950006365776062,
+      "learning_rate": 9.915006210641886e-05,
+      "loss": 1.0542,
+      "step": 974
+    },
+    {
+      "epoch": 0.08711774298032032,
+      "grad_norm": 0.46636098623275757,
+      "learning_rate": 9.914740324024316e-05,
+      "loss": 1.0893,
+      "step": 975
+    },
+    {
+      "epoch": 0.08720709451158219,
+      "grad_norm": 0.48313936591148376,
+      "learning_rate": 9.914474025744856e-05,
+      "loss": 1.0318,
+      "step": 976
+    },
+    {
+      "epoch": 0.08729644604284406,
+      "grad_norm": 0.480648934841156,
+      "learning_rate": 9.914207315825812e-05,
+      "loss": 1.1249,
+      "step": 977
+    },
+    {
+      "epoch": 0.08738579757410593,
+      "grad_norm": 0.4614575207233429,
+      "learning_rate": 9.913940194289524e-05,
+      "loss": 1.133,
+      "step": 978
+    },
+    {
+      "epoch": 0.0874751491053678,
+      "grad_norm": 0.4390462040901184,
+      "learning_rate": 9.913672661158364e-05,
+      "loss": 1.1004,
+      "step": 979
+    },
+    {
+      "epoch": 0.08756450063662966,
+      "grad_norm": 0.5013588070869446,
+      "learning_rate": 9.913404716454744e-05,
+      "loss": 1.1423,
+      "step": 980
+    },
+    {
+      "epoch": 0.08765385216789152,
+      "grad_norm": 0.40994441509246826,
+      "learning_rate": 9.913136360201106e-05,
+      "loss": 1.058,
+      "step": 981
+    },
+    {
+      "epoch": 0.0877432036991534,
+      "grad_norm": 0.4297054409980774,
+      "learning_rate": 9.912867592419928e-05,
+      "loss": 1.0135,
+      "step": 982
+    },
+    {
+      "epoch": 0.08783255523041526,
+      "grad_norm": 0.4328972399234772,
+      "learning_rate": 9.91259841313372e-05,
+      "loss": 1.1261,
+      "step": 983
+    },
+    {
+      "epoch": 0.08792190676167713,
+      "grad_norm": 0.42971837520599365,
+      "learning_rate": 9.912328822365033e-05,
+      "loss": 1.1578,
+      "step": 984
+    },
+    {
+      "epoch": 0.088011258292939,
+      "grad_norm": 0.4185878336429596,
+      "learning_rate": 9.912058820136443e-05,
+      "loss": 1.0351,
+      "step": 985
+    },
+    {
+      "epoch": 0.08810060982420086,
+      "grad_norm": 0.4461815655231476,
+      "learning_rate": 9.911788406470569e-05,
+      "loss": 1.0864,
+      "step": 986
+    },
+    {
+      "epoch": 0.08818996135546273,
+      "grad_norm": 0.402701199054718,
+      "learning_rate": 9.911517581390059e-05,
+      "loss": 1.1004,
+      "step": 987
+    },
+    {
+      "epoch": 0.0882793128867246,
+      "grad_norm": 0.4615456163883209,
+      "learning_rate": 9.9112463449176e-05,
+      "loss": 1.0427,
+      "step": 988
+    },
+    {
+      "epoch": 0.08836866441798646,
+      "grad_norm": 0.49826958775520325,
+      "learning_rate": 9.91097469707591e-05,
+      "loss": 1.0445,
+      "step": 989
+    },
+    {
+      "epoch": 0.08845801594924833,
+      "grad_norm": 0.5109001994132996,
+      "learning_rate": 9.91070263788774e-05,
+      "loss": 1.0151,
+      "step": 990
+    },
+    {
+      "epoch": 0.0885473674805102,
+      "grad_norm": 0.4601687788963318,
+      "learning_rate": 9.910430167375881e-05,
+      "loss": 0.9897,
+      "step": 991
+    },
+    {
+      "epoch": 0.08863671901177206,
+      "grad_norm": 0.48032835125923157,
+      "learning_rate": 9.910157285563154e-05,
+      "loss": 1.1166,
+      "step": 992
+    },
+    {
+      "epoch": 0.08872607054303393,
+      "grad_norm": 0.3616820275783539,
+      "learning_rate": 9.909883992472415e-05,
+      "loss": 1.1109,
+      "step": 993
+    },
+    {
+      "epoch": 0.0888154220742958,
+      "grad_norm": 0.5358753204345703,
+      "learning_rate": 9.909610288126557e-05,
+      "loss": 1.0476,
+      "step": 994
+    },
+    {
+      "epoch": 0.08890477360555767,
+      "grad_norm": 0.4149110019207001,
+      "learning_rate": 9.909336172548505e-05,
+      "loss": 1.068,
+      "step": 995
+    },
+    {
+      "epoch": 0.08899412513681954,
+      "grad_norm": 0.43261730670928955,
+      "learning_rate": 9.909061645761217e-05,
+      "loss": 1.0922,
+      "step": 996
+    },
+    {
+      "epoch": 0.0890834766680814,
+      "grad_norm": 0.504966139793396,
+      "learning_rate": 9.90878670778769e-05,
+      "loss": 1.0179,
+      "step": 997
+    },
+    {
+      "epoch": 0.08917282819934326,
+      "grad_norm": 0.42445361614227295,
+      "learning_rate": 9.908511358650953e-05,
+      "loss": 1.0927,
+      "step": 998
+    },
+    {
+      "epoch": 0.08926217973060513,
+      "grad_norm": 0.4488779306411743,
+      "learning_rate": 9.908235598374069e-05,
+      "loss": 1.0652,
+      "step": 999
+    },
+    {
+      "epoch": 0.089351531261867,
+      "grad_norm": 0.4371225833892822,
+      "learning_rate": 9.907959426980136e-05,
+      "loss": 1.0874,
+      "step": 1000
+    },
+    {
+      "epoch": 0.08944088279312887,
+      "grad_norm": 0.3913322389125824,
+      "learning_rate": 9.907682844492284e-05,
+      "loss": 1.0903,
+      "step": 1001
+    },
+    {
+      "epoch": 0.08953023432439074,
+      "grad_norm": 0.38679686188697815,
+      "learning_rate": 9.907405850933681e-05,
+      "loss": 1.0884,
+      "step": 1002
+    },
+    {
+      "epoch": 0.0896195858556526,
+      "grad_norm": 0.4290361702442169,
+      "learning_rate": 9.907128446327531e-05,
+      "loss": 1.087,
+      "step": 1003
+    },
+    {
+      "epoch": 0.08970893738691448,
+      "grad_norm": 0.42051953077316284,
+      "learning_rate": 9.906850630697068e-05,
+      "loss": 1.1921,
+      "step": 1004
+    },
+    {
+      "epoch": 0.08979828891817633,
+      "grad_norm": 0.4305054247379303,
+      "learning_rate": 9.90657240406556e-05,
+      "loss": 1.1087,
+      "step": 1005
+    },
+    {
+      "epoch": 0.0898876404494382,
+      "grad_norm": 0.4126817286014557,
+      "learning_rate": 9.906293766456312e-05,
+      "loss": 1.0889,
+      "step": 1006
+    },
+    {
+      "epoch": 0.08997699198070007,
+      "grad_norm": 0.4437747895717621,
+      "learning_rate": 9.906014717892666e-05,
+      "loss": 1.0852,
+      "step": 1007
+    },
+    {
+      "epoch": 0.09006634351196194,
+      "grad_norm": 0.40008100867271423,
+      "learning_rate": 9.905735258397993e-05,
+      "loss": 1.0639,
+      "step": 1008
+    },
+    {
+      "epoch": 0.09015569504322381,
+      "grad_norm": 0.4763084650039673,
+      "learning_rate": 9.905455387995699e-05,
+      "loss": 0.9696,
+      "step": 1009
+    },
+    {
+      "epoch": 0.09024504657448568,
+      "grad_norm": 0.5551483631134033,
+      "learning_rate": 9.905175106709228e-05,
+      "loss": 1.0921,
+      "step": 1010
+    },
+    {
+      "epoch": 0.09033439810574753,
+      "grad_norm": 0.398017555475235,
+      "learning_rate": 9.904894414562056e-05,
+      "loss": 1.0868,
+      "step": 1011
+    },
+    {
+      "epoch": 0.0904237496370094,
+      "grad_norm": 0.41361182928085327,
+      "learning_rate": 9.904613311577695e-05,
+      "loss": 1.0827,
+      "step": 1012
+    },
+    {
+      "epoch": 0.09051310116827127,
+      "grad_norm": 0.43735066056251526,
+      "learning_rate": 9.90433179777969e-05,
+      "loss": 1.0752,
+      "step": 1013
+    },
+    {
+      "epoch": 0.09060245269953314,
+      "grad_norm": 0.43464162945747375,
+      "learning_rate": 9.904049873191621e-05,
+      "loss": 1.1453,
+      "step": 1014
+    },
+    {
+      "epoch": 0.09069180423079501,
+      "grad_norm": 0.42488962411880493,
+      "learning_rate": 9.903767537837101e-05,
+      "loss": 1.0359,
+      "step": 1015
+    },
+    {
+      "epoch": 0.09078115576205688,
+      "grad_norm": 0.4419018626213074,
+      "learning_rate": 9.90348479173978e-05,
+      "loss": 1.0873,
+      "step": 1016
+    },
+    {
+      "epoch": 0.09087050729331873,
+      "grad_norm": 0.45977696776390076,
+      "learning_rate": 9.903201634923338e-05,
+      "loss": 1.0001,
+      "step": 1017
+    },
+    {
+      "epoch": 0.0909598588245806,
+      "grad_norm": 0.4522443115711212,
+      "learning_rate": 9.902918067411497e-05,
+      "loss": 1.0555,
+      "step": 1018
+    },
+    {
+      "epoch": 0.09104921035584247,
+      "grad_norm": 0.43287041783332825,
+      "learning_rate": 9.902634089228007e-05,
+      "loss": 1.0233,
+      "step": 1019
+    },
+    {
+      "epoch": 0.09113856188710434,
+      "grad_norm": 0.5111227035522461,
+      "learning_rate": 9.902349700396651e-05,
+      "loss": 0.9717,
+      "step": 1020
+    },
+    {
+      "epoch": 0.09122791341836621,
+      "grad_norm": 0.44105076789855957,
+      "learning_rate": 9.902064900941255e-05,
+      "loss": 0.9953,
+      "step": 1021
+    },
+    {
+      "epoch": 0.09131726494962808,
+      "grad_norm": 0.46010032296180725,
+      "learning_rate": 9.90177969088567e-05,
+      "loss": 1.0938,
+      "step": 1022
+    },
+    {
+      "epoch": 0.09140661648088994,
+      "grad_norm": 0.4434414803981781,
+      "learning_rate": 9.901494070253788e-05,
+      "loss": 1.1415,
+      "step": 1023
+    },
+    {
+      "epoch": 0.0914959680121518,
+      "grad_norm": 0.4490472972393036,
+      "learning_rate": 9.90120803906953e-05,
+      "loss": 1.0611,
+      "step": 1024
+    },
+    {
+      "epoch": 0.09158531954341367,
+      "grad_norm": 0.3900289833545685,
+      "learning_rate": 9.900921597356856e-05,
+      "loss": 1.1096,
+      "step": 1025
+    },
+    {
+      "epoch": 0.09167467107467554,
+      "grad_norm": 0.478982150554657,
+      "learning_rate": 9.900634745139758e-05,
+      "loss": 1.0677,
+      "step": 1026
+    },
+    {
+      "epoch": 0.09176402260593741,
+      "grad_norm": 0.37474942207336426,
+      "learning_rate": 9.900347482442262e-05,
+      "loss": 1.0386,
+      "step": 1027
+    },
+    {
+      "epoch": 0.09185337413719928,
+      "grad_norm": 0.5137802958488464,
+      "learning_rate": 9.900059809288431e-05,
+      "loss": 1.0332,
+      "step": 1028
+    },
+    {
+      "epoch": 0.09194272566846114,
+      "grad_norm": 0.4314192831516266,
+      "learning_rate": 9.899771725702362e-05,
+      "loss": 1.0801,
+      "step": 1029
+    },
+    {
+      "epoch": 0.092032077199723,
+      "grad_norm": 0.48988327383995056,
+      "learning_rate": 9.899483231708181e-05,
+      "loss": 1.0857,
+      "step": 1030
+    },
+    {
+      "epoch": 0.09212142873098488,
+      "grad_norm": 0.4243643283843994,
+      "learning_rate": 9.899194327330056e-05,
+      "loss": 1.096,
+      "step": 1031
+    },
+    {
+      "epoch": 0.09221078026224674,
+      "grad_norm": 0.4380730986595154,
+      "learning_rate": 9.898905012592183e-05,
+      "loss": 1.0709,
+      "step": 1032
+    },
+    {
+      "epoch": 0.09230013179350861,
+      "grad_norm": 0.4659455716609955,
+      "learning_rate": 9.898615287518798e-05,
+      "loss": 1.0621,
+      "step": 1033
+    },
+    {
+      "epoch": 0.09238948332477048,
+      "grad_norm": 0.46971428394317627,
+      "learning_rate": 9.898325152134167e-05,
+      "loss": 1.1165,
+      "step": 1034
+    },
+    {
+      "epoch": 0.09247883485603235,
+      "grad_norm": 0.47089582681655884,
+      "learning_rate": 9.898034606462592e-05,
+      "loss": 1.1352,
+      "step": 1035
+    },
+    {
+      "epoch": 0.09256818638729421,
+      "grad_norm": 0.44664955139160156,
+      "learning_rate": 9.897743650528408e-05,
+      "loss": 1.0728,
+      "step": 1036
+    },
+    {
+      "epoch": 0.09265753791855608,
+      "grad_norm": 0.4320451617240906,
+      "learning_rate": 9.897452284355989e-05,
+      "loss": 1.1007,
+      "step": 1037
+    },
+    {
+      "epoch": 0.09274688944981795,
+      "grad_norm": 0.4943344295024872,
+      "learning_rate": 9.897160507969738e-05,
+      "loss": 1.0915,
+      "step": 1038
+    },
+    {
+      "epoch": 0.09283624098107982,
+      "grad_norm": 0.43696850538253784,
+      "learning_rate": 9.896868321394093e-05,
+      "loss": 1.1057,
+      "step": 1039
+    },
+    {
+      "epoch": 0.09292559251234168,
+      "grad_norm": 0.39653998613357544,
+      "learning_rate": 9.896575724653529e-05,
+      "loss": 1.0422,
+      "step": 1040
+    },
+    {
+      "epoch": 0.09301494404360355,
+      "grad_norm": 0.4009700417518616,
+      "learning_rate": 9.896282717772556e-05,
+      "loss": 1.0917,
+      "step": 1041
+    },
+    {
+      "epoch": 0.09310429557486541,
+      "grad_norm": 0.4337768852710724,
+      "learning_rate": 9.895989300775714e-05,
+      "loss": 1.089,
+      "step": 1042
+    },
+    {
+      "epoch": 0.09319364710612728,
+      "grad_norm": 0.4776664078235626,
+      "learning_rate": 9.895695473687581e-05,
+      "loss": 1.0316,
+      "step": 1043
+    },
+    {
+      "epoch": 0.09328299863738915,
+      "grad_norm": 0.4160580337047577,
+      "learning_rate": 9.895401236532769e-05,
+      "loss": 1.0437,
+      "step": 1044
+    },
+    {
+      "epoch": 0.09337235016865102,
+      "grad_norm": 0.4131687879562378,
+      "learning_rate": 9.895106589335919e-05,
+      "loss": 1.0765,
+      "step": 1045
+    },
+    {
+      "epoch": 0.09346170169991289,
+      "grad_norm": 0.39853063225746155,
+      "learning_rate": 9.894811532121716e-05,
+      "loss": 1.0726,
+      "step": 1046
+    },
+    {
+      "epoch": 0.09355105323117476,
+      "grad_norm": 0.46733424067497253,
+      "learning_rate": 9.894516064914871e-05,
+      "loss": 1.0561,
+      "step": 1047
+    },
+    {
+      "epoch": 0.09364040476243661,
+      "grad_norm": 0.3866161108016968,
+      "learning_rate": 9.894220187740135e-05,
+      "loss": 1.0816,
+      "step": 1048
+    },
+    {
+      "epoch": 0.09372975629369848,
+      "grad_norm": 0.42076459527015686,
+      "learning_rate": 9.89392390062229e-05,
+      "loss": 0.9866,
+      "step": 1049
+    },
+    {
+      "epoch": 0.09381910782496035,
+      "grad_norm": 0.4648950695991516,
+      "learning_rate": 9.893627203586152e-05,
+      "loss": 1.0732,
+      "step": 1050
+    },
+    {
+      "epoch": 0.09390845935622222,
+      "grad_norm": 0.4551086723804474,
+      "learning_rate": 9.893330096656574e-05,
+      "loss": 1.1067,
+      "step": 1051
+    },
+    {
+      "epoch": 0.09399781088748409,
+      "grad_norm": 0.43443188071250916,
+      "learning_rate": 9.893032579858442e-05,
+      "loss": 1.0777,
+      "step": 1052
+    },
+    {
+      "epoch": 0.09408716241874596,
+      "grad_norm": 0.4357977509498596,
+      "learning_rate": 9.892734653216673e-05,
+      "loss": 0.995,
+      "step": 1053
+    },
+    {
+      "epoch": 0.09417651395000781,
+      "grad_norm": 0.4856785833835602,
+      "learning_rate": 9.892436316756226e-05,
+      "loss": 1.0041,
+      "step": 1054
+    },
+    {
+      "epoch": 0.09426586548126968,
+      "grad_norm": 0.4476493299007416,
+      "learning_rate": 9.892137570502087e-05,
+      "loss": 1.0459,
+      "step": 1055
+    },
+    {
+      "epoch": 0.09435521701253155,
+      "grad_norm": 0.4645354449748993,
+      "learning_rate": 9.89183841447928e-05,
+      "loss": 1.1058,
+      "step": 1056
+    },
+    {
+      "epoch": 0.09444456854379342,
+      "grad_norm": 0.45503050088882446,
+      "learning_rate": 9.891538848712863e-05,
+      "loss": 1.0721,
+      "step": 1057
+    },
+    {
+      "epoch": 0.09453392007505529,
+      "grad_norm": 0.4932888448238373,
+      "learning_rate": 9.891238873227925e-05,
+      "loss": 1.0064,
+      "step": 1058
+    },
+    {
+      "epoch": 0.09462327160631716,
+      "grad_norm": 0.43276292085647583,
+      "learning_rate": 9.890938488049597e-05,
+      "loss": 1.1481,
+      "step": 1059
+    },
+    {
+      "epoch": 0.09471262313757901,
+      "grad_norm": 0.4331585764884949,
+      "learning_rate": 9.890637693203038e-05,
+      "loss": 1.0468,
+      "step": 1060
+    },
+    {
+      "epoch": 0.09480197466884088,
+      "grad_norm": 0.5495041608810425,
+      "learning_rate": 9.89033648871344e-05,
+      "loss": 1.0436,
+      "step": 1061
+    },
+    {
+      "epoch": 0.09489132620010275,
+      "grad_norm": 0.47634175419807434,
+      "learning_rate": 9.890034874606033e-05,
+      "loss": 1.0175,
+      "step": 1062
+    },
+    {
+      "epoch": 0.09498067773136462,
+      "grad_norm": 0.400700181722641,
+      "learning_rate": 9.889732850906083e-05,
+      "loss": 1.1379,
+      "step": 1063
+    },
+    {
+      "epoch": 0.09507002926262649,
+      "grad_norm": 0.49066829681396484,
+      "learning_rate": 9.889430417638885e-05,
+      "loss": 1.115,
+      "step": 1064
+    },
+    {
+      "epoch": 0.09515938079388836,
+      "grad_norm": 0.3815631568431854,
+      "learning_rate": 9.889127574829773e-05,
+      "loss": 1.1274,
+      "step": 1065
+    },
+    {
+      "epoch": 0.09524873232515023,
+      "grad_norm": 0.5222175717353821,
+      "learning_rate": 9.88882432250411e-05,
+      "loss": 1.0232,
+      "step": 1066
+    },
+    {
+      "epoch": 0.09533808385641208,
+      "grad_norm": 0.4751056134700775,
+      "learning_rate": 9.888520660687302e-05,
+      "loss": 1.1066,
+      "step": 1067
+    },
+    {
+      "epoch": 0.09542743538767395,
+      "grad_norm": 0.5163382887840271,
+      "learning_rate": 9.888216589404779e-05,
+      "loss": 1.0504,
+      "step": 1068
+    },
+    {
+      "epoch": 0.09551678691893582,
+      "grad_norm": 0.5940535068511963,
+      "learning_rate": 9.887912108682011e-05,
+      "loss": 1.0346,
+      "step": 1069
+    },
+    {
+      "epoch": 0.09560613845019769,
+      "grad_norm": 0.38179200887680054,
+      "learning_rate": 9.887607218544503e-05,
+      "loss": 1.1381,
+      "step": 1070
+    },
+    {
+      "epoch": 0.09569548998145956,
+      "grad_norm": 0.3916855454444885,
+      "learning_rate": 9.887301919017794e-05,
+      "loss": 1.1062,
+      "step": 1071
+    },
+    {
+      "epoch": 0.09578484151272143,
+      "grad_norm": 0.48447951674461365,
+      "learning_rate": 9.886996210127452e-05,
+      "loss": 1.0219,
+      "step": 1072
+    },
+    {
+      "epoch": 0.09587419304398329,
+      "grad_norm": 0.4082842171192169,
+      "learning_rate": 9.886690091899088e-05,
+      "loss": 1.0546,
+      "step": 1073
+    },
+    {
+      "epoch": 0.09596354457524515,
+      "grad_norm": 0.4271855354309082,
+      "learning_rate": 9.886383564358339e-05,
+      "loss": 1.05,
+      "step": 1074
+    },
+    {
+      "epoch": 0.09605289610650702,
+      "grad_norm": 0.37129080295562744,
+      "learning_rate": 9.886076627530883e-05,
+      "loss": 1.0979,
+      "step": 1075
+    },
+    {
+      "epoch": 0.0961422476377689,
+      "grad_norm": 0.4285633862018585,
+      "learning_rate": 9.885769281442426e-05,
+      "loss": 1.0915,
+      "step": 1076
+    },
+    {
+      "epoch": 0.09623159916903076,
+      "grad_norm": 0.5488273501396179,
+      "learning_rate": 9.885461526118713e-05,
+      "loss": 0.9937,
+      "step": 1077
+    },
+    {
+      "epoch": 0.09632095070029263,
+      "grad_norm": 0.42456749081611633,
+      "learning_rate": 9.885153361585523e-05,
+      "loss": 1.1105,
+      "step": 1078
+    },
+    {
+      "epoch": 0.09641030223155449,
+      "grad_norm": 0.38298559188842773,
+      "learning_rate": 9.884844787868667e-05,
+      "loss": 1.1525,
+      "step": 1079
+    },
+    {
+      "epoch": 0.09649965376281636,
+      "grad_norm": 0.46398019790649414,
+      "learning_rate": 9.884535804993991e-05,
+      "loss": 1.0575,
+      "step": 1080
+    },
+    {
+      "epoch": 0.09658900529407823,
+      "grad_norm": 0.47431719303131104,
+      "learning_rate": 9.884226412987375e-05,
+      "loss": 1.0777,
+      "step": 1081
+    },
+    {
+      "epoch": 0.0966783568253401,
+      "grad_norm": 0.4811946451663971,
+      "learning_rate": 9.883916611874735e-05,
+      "loss": 1.0186,
+      "step": 1082
+    },
+    {
+      "epoch": 0.09676770835660196,
+      "grad_norm": 0.4855489730834961,
+      "learning_rate": 9.883606401682022e-05,
+      "loss": 1.1432,
+      "step": 1083
+    },
+    {
+      "epoch": 0.09685705988786383,
+      "grad_norm": 0.423043817281723,
+      "learning_rate": 9.883295782435216e-05,
+      "loss": 1.0769,
+      "step": 1084
+    },
+    {
+      "epoch": 0.09694641141912569,
+      "grad_norm": 0.5006414651870728,
+      "learning_rate": 9.882984754160334e-05,
+      "loss": 1.0417,
+      "step": 1085
+    },
+    {
+      "epoch": 0.09703576295038756,
+      "grad_norm": 0.41438305377960205,
+      "learning_rate": 9.882673316883432e-05,
+      "loss": 1.091,
+      "step": 1086
+    },
+    {
+      "epoch": 0.09712511448164943,
+      "grad_norm": 0.4482746124267578,
+      "learning_rate": 9.882361470630594e-05,
+      "loss": 1.0692,
+      "step": 1087
+    },
+    {
+      "epoch": 0.0972144660129113,
+      "grad_norm": 0.43151143193244934,
+      "learning_rate": 9.882049215427941e-05,
+      "loss": 1.0767,
+      "step": 1088
+    },
+    {
+      "epoch": 0.09730381754417317,
+      "grad_norm": 0.47808754444122314,
+      "learning_rate": 9.881736551301627e-05,
+      "loss": 1.1291,
+      "step": 1089
+    },
+    {
+      "epoch": 0.09739316907543503,
+      "grad_norm": 0.3970921039581299,
+      "learning_rate": 9.88142347827784e-05,
+      "loss": 1.0847,
+      "step": 1090
+    },
+    {
+      "epoch": 0.0974825206066969,
+      "grad_norm": 0.4331444501876831,
+      "learning_rate": 9.881109996382807e-05,
+      "loss": 1.0468,
+      "step": 1091
+    },
+    {
+      "epoch": 0.09757187213795876,
+      "grad_norm": 0.48726001381874084,
+      "learning_rate": 9.880796105642782e-05,
+      "loss": 1.0272,
+      "step": 1092
+    },
+    {
+      "epoch": 0.09766122366922063,
+      "grad_norm": 0.39687567949295044,
+      "learning_rate": 9.880481806084057e-05,
+      "loss": 1.0803,
+      "step": 1093
+    },
+    {
+      "epoch": 0.0977505752004825,
+      "grad_norm": 0.510676920413971,
+      "learning_rate": 9.880167097732957e-05,
+      "loss": 0.9923,
+      "step": 1094
+    },
+    {
+      "epoch": 0.09783992673174437,
+      "grad_norm": 0.4363247752189636,
+      "learning_rate": 9.879851980615847e-05,
+      "loss": 1.0961,
+      "step": 1095
+    },
+    {
+      "epoch": 0.09792927826300624,
+      "grad_norm": 0.4566675126552582,
+      "learning_rate": 9.879536454759115e-05,
+      "loss": 0.9882,
+      "step": 1096
+    },
+    {
+      "epoch": 0.0980186297942681,
+      "grad_norm": 0.3779030442237854,
+      "learning_rate": 9.879220520189195e-05,
+      "loss": 1.1436,
+      "step": 1097
+    },
+    {
+      "epoch": 0.09810798132552996,
+      "grad_norm": 0.45519158244132996,
+      "learning_rate": 9.878904176932546e-05,
+      "loss": 1.0497,
+      "step": 1098
+    },
+    {
+      "epoch": 0.09819733285679183,
+      "grad_norm": 0.3910900056362152,
+      "learning_rate": 9.878587425015668e-05,
+      "loss": 1.0654,
+      "step": 1099
+    },
+    {
+      "epoch": 0.0982866843880537,
+      "grad_norm": 0.45868462324142456,
+      "learning_rate": 9.878270264465091e-05,
+      "loss": 1.0361,
+      "step": 1100
+    },
+    {
+      "epoch": 0.09837603591931557,
+      "grad_norm": 0.435067743062973,
+      "learning_rate": 9.877952695307382e-05,
+      "loss": 1.0632,
+      "step": 1101
+    },
+    {
+      "epoch": 0.09846538745057744,
+      "grad_norm": 0.5049869418144226,
+      "learning_rate": 9.877634717569137e-05,
+      "loss": 0.9519,
+      "step": 1102
+    },
+    {
+      "epoch": 0.09855473898183931,
+      "grad_norm": 0.48381614685058594,
+      "learning_rate": 9.877316331276995e-05,
+      "loss": 1.0368,
+      "step": 1103
+    },
+    {
+      "epoch": 0.09864409051310116,
+      "grad_norm": 0.4501636326313019,
+      "learning_rate": 9.876997536457619e-05,
+      "loss": 0.9595,
+      "step": 1104
+    },
+    {
+      "epoch": 0.09873344204436303,
+      "grad_norm": 0.4924948811531067,
+      "learning_rate": 9.876678333137716e-05,
+      "loss": 1.0592,
+      "step": 1105
+    },
+    {
+      "epoch": 0.0988227935756249,
+      "grad_norm": 0.387658029794693,
+      "learning_rate": 9.876358721344022e-05,
+      "loss": 1.0821,
+      "step": 1106
+    },
+    {
+      "epoch": 0.09891214510688677,
+      "grad_norm": 0.5033579468727112,
+      "learning_rate": 9.876038701103305e-05,
+      "loss": 1.0504,
+      "step": 1107
+    },
+    {
+      "epoch": 0.09900149663814864,
+      "grad_norm": 0.47466421127319336,
+      "learning_rate": 9.875718272442373e-05,
+      "loss": 1.009,
+      "step": 1108
+    },
+    {
+      "epoch": 0.09909084816941051,
+      "grad_norm": 0.43417561054229736,
+      "learning_rate": 9.875397435388063e-05,
+      "loss": 1.0584,
+      "step": 1109
+    },
+    {
+      "epoch": 0.09918019970067236,
+      "grad_norm": 0.4432922899723053,
+      "learning_rate": 9.875076189967252e-05,
+      "loss": 1.0067,
+      "step": 1110
+    },
+    {
+      "epoch": 0.09926955123193423,
+      "grad_norm": 0.45428282022476196,
+      "learning_rate": 9.874754536206843e-05,
+      "loss": 1.1137,
+      "step": 1111
+    },
+    {
+      "epoch": 0.0993589027631961,
+      "grad_norm": 0.4388342797756195,
+      "learning_rate": 9.874432474133781e-05,
+      "loss": 1.0513,
+      "step": 1112
+    },
+    {
+      "epoch": 0.09944825429445797,
+      "grad_norm": 0.42276880145072937,
+      "learning_rate": 9.874110003775043e-05,
+      "loss": 1.0552,
+      "step": 1113
+    },
+    {
+      "epoch": 0.09953760582571984,
+      "grad_norm": 0.4311312735080719,
+      "learning_rate": 9.873787125157636e-05,
+      "loss": 1.0679,
+      "step": 1114
+    },
+    {
+      "epoch": 0.09962695735698171,
+      "grad_norm": 0.38914674520492554,
+      "learning_rate": 9.873463838308608e-05,
+      "loss": 1.0837,
+      "step": 1115
+    },
+    {
+      "epoch": 0.09971630888824357,
+      "grad_norm": 0.46259498596191406,
+      "learning_rate": 9.873140143255036e-05,
+      "loss": 1.0772,
+      "step": 1116
+    },
+    {
+      "epoch": 0.09980566041950543,
+      "grad_norm": 0.4010103642940521,
+      "learning_rate": 9.872816040024034e-05,
+      "loss": 1.0801,
+      "step": 1117
+    },
+    {
+      "epoch": 0.0998950119507673,
+      "grad_norm": 0.3827916383743286,
+      "learning_rate": 9.872491528642746e-05,
+      "loss": 1.1128,
+      "step": 1118
+    },
+    {
+      "epoch": 0.09998436348202917,
+      "grad_norm": 0.41605910658836365,
+      "learning_rate": 9.872166609138355e-05,
+      "loss": 1.0274,
+      "step": 1119
+    },
+    {
+      "epoch": 0.10007371501329104,
+      "grad_norm": 0.44363489747047424,
+      "learning_rate": 9.87184128153808e-05,
+      "loss": 1.0553,
+      "step": 1120
+    },
+    {
+      "epoch": 0.10016306654455291,
+      "grad_norm": 0.4547342360019684,
+      "learning_rate": 9.871515545869166e-05,
+      "loss": 1.0827,
+      "step": 1121
+    },
+    {
+      "epoch": 0.10025241807581478,
+      "grad_norm": 0.4331105649471283,
+      "learning_rate": 9.871189402158898e-05,
+      "loss": 1.0772,
+      "step": 1122
+    },
+    {
+      "epoch": 0.10034176960707664,
+      "grad_norm": 0.47358986735343933,
+      "learning_rate": 9.870862850434593e-05,
+      "loss": 1.124,
+      "step": 1123
+    },
+    {
+      "epoch": 0.1004311211383385,
+      "grad_norm": 0.4455796480178833,
+      "learning_rate": 9.870535890723607e-05,
+      "loss": 1.0435,
+      "step": 1124
+    },
+    {
+      "epoch": 0.10052047266960037,
+      "grad_norm": 0.42114776372909546,
+      "learning_rate": 9.870208523053323e-05,
+      "loss": 1.1176,
+      "step": 1125
+    },
+    {
+      "epoch": 0.10060982420086224,
+      "grad_norm": 0.4320701062679291,
+      "learning_rate": 9.869880747451164e-05,
+      "loss": 1.0214,
+      "step": 1126
+    },
+    {
+      "epoch": 0.10069917573212411,
+      "grad_norm": 0.41799092292785645,
+      "learning_rate": 9.86955256394458e-05,
+      "loss": 1.0928,
+      "step": 1127
+    },
+    {
+      "epoch": 0.10078852726338598,
+      "grad_norm": 0.4450794458389282,
+      "learning_rate": 9.869223972561066e-05,
+      "loss": 1.1019,
+      "step": 1128
+    },
+    {
+      "epoch": 0.10087787879464784,
+      "grad_norm": 0.42303231358528137,
+      "learning_rate": 9.86889497332814e-05,
+      "loss": 1.0618,
+      "step": 1129
+    },
+    {
+      "epoch": 0.10096723032590971,
+      "grad_norm": 0.5273473858833313,
+      "learning_rate": 9.868565566273363e-05,
+      "loss": 1.0161,
+      "step": 1130
+    },
+    {
+      "epoch": 0.10105658185717158,
+      "grad_norm": 0.4437713921070099,
+      "learning_rate": 9.868235751424324e-05,
+      "loss": 1.0227,
+      "step": 1131
+    },
+    {
+      "epoch": 0.10114593338843345,
+      "grad_norm": 0.41443753242492676,
+      "learning_rate": 9.86790552880865e-05,
+      "loss": 1.036,
+      "step": 1132
+    },
+    {
+      "epoch": 0.10123528491969531,
+      "grad_norm": 0.5034554600715637,
+      "learning_rate": 9.867574898453998e-05,
+      "loss": 0.9698,
+      "step": 1133
+    },
+    {
+      "epoch": 0.10132463645095718,
+      "grad_norm": 0.4698033630847931,
+      "learning_rate": 9.867243860388065e-05,
+      "loss": 1.0825,
+      "step": 1134
+    },
+    {
+      "epoch": 0.10141398798221904,
+      "grad_norm": 0.48983025550842285,
+      "learning_rate": 9.866912414638577e-05,
+      "loss": 1.0424,
+      "step": 1135
+    },
+    {
+      "epoch": 0.10150333951348091,
+      "grad_norm": 0.5005711317062378,
+      "learning_rate": 9.866580561233297e-05,
+      "loss": 1.0549,
+      "step": 1136
+    },
+    {
+      "epoch": 0.10159269104474278,
+      "grad_norm": 0.44725582003593445,
+      "learning_rate": 9.866248300200021e-05,
+      "loss": 1.0877,
+      "step": 1137
+    },
+    {
+      "epoch": 0.10168204257600465,
+      "grad_norm": 0.4655127227306366,
+      "learning_rate": 9.865915631566579e-05,
+      "loss": 1.1502,
+      "step": 1138
+    },
+    {
+      "epoch": 0.10177139410726652,
+      "grad_norm": 0.4426875114440918,
+      "learning_rate": 9.865582555360837e-05,
+      "loss": 1.1273,
+      "step": 1139
+    },
+    {
+      "epoch": 0.10186074563852839,
+      "grad_norm": 0.4345450699329376,
+      "learning_rate": 9.865249071610691e-05,
+      "loss": 1.1256,
+      "step": 1140
+    },
+    {
+      "epoch": 0.10195009716979024,
+      "grad_norm": 0.4371299743652344,
+      "learning_rate": 9.864915180344076e-05,
+      "loss": 1.0296,
+      "step": 1141
+    },
+    {
+      "epoch": 0.10203944870105211,
+      "grad_norm": 0.44082698225975037,
+      "learning_rate": 9.864580881588959e-05,
+      "loss": 1.0628,
+      "step": 1142
+    },
+    {
+      "epoch": 0.10212880023231398,
+      "grad_norm": 0.5174263715744019,
+      "learning_rate": 9.864246175373339e-05,
+      "loss": 0.9576,
+      "step": 1143
+    },
+    {
+      "epoch": 0.10221815176357585,
+      "grad_norm": 0.42104923725128174,
+      "learning_rate": 9.863911061725256e-05,
+      "loss": 1.017,
+      "step": 1144
+    },
+    {
+      "epoch": 0.10230750329483772,
+      "grad_norm": 0.39812588691711426,
+      "learning_rate": 9.863575540672772e-05,
+      "loss": 1.0241,
+      "step": 1145
+    },
+    {
+      "epoch": 0.10239685482609959,
+      "grad_norm": 0.5662054419517517,
+      "learning_rate": 9.863239612243997e-05,
+      "loss": 1.0541,
+      "step": 1146
+    },
+    {
+      "epoch": 0.10248620635736144,
+      "grad_norm": 0.4463253319263458,
+      "learning_rate": 9.862903276467066e-05,
+      "loss": 1.0583,
+      "step": 1147
+    },
+    {
+      "epoch": 0.10257555788862331,
+      "grad_norm": 0.3903212249279022,
+      "learning_rate": 9.86256653337015e-05,
+      "loss": 1.1279,
+      "step": 1148
+    },
+    {
+      "epoch": 0.10266490941988518,
+      "grad_norm": 0.518621563911438,
+      "learning_rate": 9.862229382981454e-05,
+      "loss": 1.0526,
+      "step": 1149
+    },
+    {
+      "epoch": 0.10275426095114705,
+      "grad_norm": 0.5142053365707397,
+      "learning_rate": 9.861891825329222e-05,
+      "loss": 1.0619,
+      "step": 1150
+    },
+    {
+      "epoch": 0.10284361248240892,
+      "grad_norm": 0.45500150322914124,
+      "learning_rate": 9.861553860441725e-05,
+      "loss": 1.0992,
+      "step": 1151
+    },
+    {
+      "epoch": 0.10293296401367079,
+      "grad_norm": 0.4826822578907013,
+      "learning_rate": 9.861215488347272e-05,
+      "loss": 1.0001,
+      "step": 1152
+    },
+    {
+      "epoch": 0.10302231554493266,
+      "grad_norm": 0.44567134976387024,
+      "learning_rate": 9.860876709074204e-05,
+      "loss": 1.1272,
+      "step": 1153
+    },
+    {
+      "epoch": 0.10311166707619451,
+      "grad_norm": 0.4271107614040375,
+      "learning_rate": 9.860537522650898e-05,
+      "loss": 1.0903,
+      "step": 1154
+    },
+    {
+      "epoch": 0.10320101860745638,
+      "grad_norm": 0.4095713198184967,
+      "learning_rate": 9.860197929105768e-05,
+      "loss": 1.0451,
+      "step": 1155
+    },
+    {
+      "epoch": 0.10329037013871825,
+      "grad_norm": 0.42766252160072327,
+      "learning_rate": 9.859857928467254e-05,
+      "loss": 1.0661,
+      "step": 1156
+    },
+    {
+      "epoch": 0.10337972166998012,
+      "grad_norm": 0.4731602072715759,
+      "learning_rate": 9.859517520763834e-05,
+      "loss": 1.0923,
+      "step": 1157
+    },
+    {
+      "epoch": 0.10346907320124199,
+      "grad_norm": 0.5062028169631958,
+      "learning_rate": 9.859176706024024e-05,
+      "loss": 0.9769,
+      "step": 1158
+    },
+    {
+      "epoch": 0.10355842473250386,
+      "grad_norm": 0.433315634727478,
+      "learning_rate": 9.858835484276369e-05,
+      "loss": 1.0987,
+      "step": 1159
+    },
+    {
+      "epoch": 0.10364777626376571,
+      "grad_norm": 0.4237388074398041,
+      "learning_rate": 9.858493855549453e-05,
+      "loss": 1.0466,
+      "step": 1160
+    },
+    {
+      "epoch": 0.10373712779502758,
+      "grad_norm": 0.45744040608406067,
+      "learning_rate": 9.858151819871888e-05,
+      "loss": 1.039,
+      "step": 1161
+    },
+    {
+      "epoch": 0.10382647932628945,
+      "grad_norm": 0.5162191390991211,
+      "learning_rate": 9.857809377272323e-05,
+      "loss": 1.0173,
+      "step": 1162
+    },
+    {
+      "epoch": 0.10391583085755132,
+      "grad_norm": 0.4239135682582855,
+      "learning_rate": 9.857466527779444e-05,
+      "loss": 1.044,
+      "step": 1163
+    },
+    {
+      "epoch": 0.10400518238881319,
+      "grad_norm": 0.5142119526863098,
+      "learning_rate": 9.857123271421964e-05,
+      "loss": 1.0666,
+      "step": 1164
+    },
+    {
+      "epoch": 0.10409453392007506,
+      "grad_norm": 0.49715521931648254,
+      "learning_rate": 9.856779608228638e-05,
+      "loss": 1.0999,
+      "step": 1165
+    },
+    {
+      "epoch": 0.10418388545133692,
+      "grad_norm": 0.4236762523651123,
+      "learning_rate": 9.856435538228251e-05,
+      "loss": 1.0383,
+      "step": 1166
+    },
+    {
+      "epoch": 0.10427323698259879,
+      "grad_norm": 0.4198687970638275,
+      "learning_rate": 9.85609106144962e-05,
+      "loss": 1.0291,
+      "step": 1167
+    },
+    {
+      "epoch": 0.10436258851386065,
+      "grad_norm": 0.4170401990413666,
+      "learning_rate": 9.8557461779216e-05,
+      "loss": 1.056,
+      "step": 1168
+    },
+    {
+      "epoch": 0.10445194004512252,
+      "grad_norm": 0.4177343547344208,
+      "learning_rate": 9.855400887673082e-05,
+      "loss": 1.0697,
+      "step": 1169
+    },
+    {
+      "epoch": 0.1045412915763844,
+      "grad_norm": 0.4959930181503296,
+      "learning_rate": 9.855055190732983e-05,
+      "loss": 1.1104,
+      "step": 1170
+    },
+    {
+      "epoch": 0.10463064310764626,
+      "grad_norm": 0.4635033905506134,
+      "learning_rate": 9.85470908713026e-05,
+      "loss": 0.9893,
+      "step": 1171
+    },
+    {
+      "epoch": 0.10471999463890812,
+      "grad_norm": 0.42118409276008606,
+      "learning_rate": 9.854362576893905e-05,
+      "loss": 1.0985,
+      "step": 1172
+    },
+    {
+      "epoch": 0.10480934617016999,
+      "grad_norm": 0.41519343852996826,
+      "learning_rate": 9.85401566005294e-05,
+      "loss": 1.0862,
+      "step": 1173
+    },
+    {
+      "epoch": 0.10489869770143186,
+      "grad_norm": 0.42314785718917847,
+      "learning_rate": 9.853668336636422e-05,
+      "loss": 1.1004,
+      "step": 1174
+    },
+    {
+      "epoch": 0.10498804923269373,
+      "grad_norm": 0.4355357885360718,
+      "learning_rate": 9.853320606673446e-05,
+      "loss": 1.0732,
+      "step": 1175
+    },
+    {
+      "epoch": 0.1050774007639556,
+      "grad_norm": 0.3647025525569916,
+      "learning_rate": 9.852972470193136e-05,
+      "loss": 1.0505,
+      "step": 1176
+    },
+    {
+      "epoch": 0.10516675229521746,
+      "grad_norm": 0.4869689643383026,
+      "learning_rate": 9.852623927224653e-05,
+      "loss": 1.0354,
+      "step": 1177
+    },
+    {
+      "epoch": 0.10525610382647933,
+      "grad_norm": 0.38133373856544495,
+      "learning_rate": 9.85227497779719e-05,
+      "loss": 1.0863,
+      "step": 1178
+    },
+    {
+      "epoch": 0.10534545535774119,
+      "grad_norm": 0.5018585920333862,
+      "learning_rate": 9.851925621939976e-05,
+      "loss": 1.093,
+      "step": 1179
+    },
+    {
+      "epoch": 0.10543480688900306,
+      "grad_norm": 0.44408929347991943,
+      "learning_rate": 9.851575859682274e-05,
+      "loss": 1.0619,
+      "step": 1180
+    },
+    {
+      "epoch": 0.10552415842026493,
+      "grad_norm": 0.4483299255371094,
+      "learning_rate": 9.85122569105338e-05,
+      "loss": 1.0438,
+      "step": 1181
+    },
+    {
+      "epoch": 0.1056135099515268,
+      "grad_norm": 0.3859767019748688,
+      "learning_rate": 9.850875116082623e-05,
+      "loss": 1.1045,
+      "step": 1182
+    },
+    {
+      "epoch": 0.10570286148278867,
+      "grad_norm": 0.4310990571975708,
+      "learning_rate": 9.850524134799371e-05,
+      "loss": 1.0604,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10579221301405053,
+      "grad_norm": 0.5007196664810181,
+      "learning_rate": 9.850172747233018e-05,
+      "loss": 0.9943,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10588156454531239,
+      "grad_norm": 0.4328349530696869,
+      "learning_rate": 9.849820953412997e-05,
+      "loss": 1.0766,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10597091607657426,
+      "grad_norm": 0.4742787480354309,
+      "learning_rate": 9.849468753368777e-05,
+      "loss": 1.0821,
+      "step": 1186
+    },
+    {
+      "epoch": 0.10606026760783613,
+      "grad_norm": 0.43033096194267273,
+      "learning_rate": 9.849116147129857e-05,
+      "loss": 1.0192,
+      "step": 1187
+    },
+    {
+      "epoch": 0.106149619139098,
+      "grad_norm": 0.416620671749115,
+      "learning_rate": 9.848763134725771e-05,
+      "loss": 1.1052,
+      "step": 1188
+    },
+    {
+      "epoch": 0.10623897067035987,
+      "grad_norm": 0.44439324736595154,
+      "learning_rate": 9.848409716186091e-05,
+      "loss": 1.0822,
+      "step": 1189
+    },
+    {
+      "epoch": 0.10632832220162174,
+      "grad_norm": 0.4544062316417694,
+      "learning_rate": 9.848055891540416e-05,
+      "loss": 1.0642,
+      "step": 1190
+    },
+    {
+      "epoch": 0.10641767373288359,
+      "grad_norm": 0.5025431513786316,
+      "learning_rate": 9.847701660818381e-05,
+      "loss": 0.9832,
+      "step": 1191
+    },
+    {
+      "epoch": 0.10650702526414546,
+      "grad_norm": 0.4252472519874573,
+      "learning_rate": 9.847347024049662e-05,
+      "loss": 1.0422,
+      "step": 1192
+    },
+    {
+      "epoch": 0.10659637679540733,
+      "grad_norm": 0.43434327840805054,
+      "learning_rate": 9.84699198126396e-05,
+      "loss": 1.1426,
+      "step": 1193
+    },
+    {
+      "epoch": 0.1066857283266692,
+      "grad_norm": 0.40715593099594116,
+      "learning_rate": 9.846636532491014e-05,
+      "loss": 1.0338,
+      "step": 1194
+    },
+    {
+      "epoch": 0.10677507985793107,
+      "grad_norm": 0.47930577397346497,
+      "learning_rate": 9.846280677760596e-05,
+      "loss": 1.1011,
+      "step": 1195
+    },
+    {
+      "epoch": 0.10686443138919294,
+      "grad_norm": 0.45403239130973816,
+      "learning_rate": 9.845924417102514e-05,
+      "loss": 1.0613,
+      "step": 1196
+    },
+    {
+      "epoch": 0.10695378292045479,
+      "grad_norm": 0.3767177164554596,
+      "learning_rate": 9.84556775054661e-05,
+      "loss": 1.1216,
+      "step": 1197
+    },
+    {
+      "epoch": 0.10704313445171666,
+      "grad_norm": 0.41710513830184937,
+      "learning_rate": 9.845210678122756e-05,
+      "loss": 1.0588,
+      "step": 1198
+    },
+    {
+      "epoch": 0.10713248598297853,
+      "grad_norm": 0.49159350991249084,
+      "learning_rate": 9.844853199860861e-05,
+      "loss": 1.0458,
+      "step": 1199
+    },
+    {
+      "epoch": 0.1072218375142404,
+      "grad_norm": 0.4163805842399597,
+      "learning_rate": 9.844495315790869e-05,
+      "loss": 1.0478,
+      "step": 1200
+    },
+    {
+      "epoch": 0.10731118904550227,
+      "grad_norm": 0.3860858678817749,
+      "learning_rate": 9.844137025942754e-05,
+      "loss": 1.0728,
+      "step": 1201
+    },
+    {
+      "epoch": 0.10740054057676414,
+      "grad_norm": 0.44660601019859314,
+      "learning_rate": 9.84377833034653e-05,
+      "loss": 1.1211,
+      "step": 1202
+    },
+    {
+      "epoch": 0.107489892108026,
+      "grad_norm": 0.4287334382534027,
+      "learning_rate": 9.843419229032238e-05,
+      "loss": 1.0144,
+      "step": 1203
+    },
+    {
+      "epoch": 0.10757924363928786,
+      "grad_norm": 0.4286378026008606,
+      "learning_rate": 9.843059722029959e-05,
+      "loss": 1.1338,
+      "step": 1204
+    },
+    {
+      "epoch": 0.10766859517054973,
+      "grad_norm": 0.3777160048484802,
+      "learning_rate": 9.842699809369806e-05,
+      "loss": 1.0877,
+      "step": 1205
+    },
+    {
+      "epoch": 0.1077579467018116,
+      "grad_norm": 0.40929606556892395,
+      "learning_rate": 9.842339491081924e-05,
+      "loss": 1.0229,
+      "step": 1206
+    },
+    {
+      "epoch": 0.10784729823307347,
+      "grad_norm": 0.40165361762046814,
+      "learning_rate": 9.841978767196495e-05,
+      "loss": 1.042,
+      "step": 1207
+    },
+    {
+      "epoch": 0.10793664976433534,
+      "grad_norm": 0.4323919713497162,
+      "learning_rate": 9.841617637743731e-05,
+      "loss": 1.0781,
+      "step": 1208
+    },
+    {
+      "epoch": 0.10802600129559721,
+      "grad_norm": 0.5180752873420715,
+      "learning_rate": 9.841256102753882e-05,
+      "loss": 1.1395,
+      "step": 1209
+    },
+    {
+      "epoch": 0.10811535282685907,
+      "grad_norm": 0.40795522928237915,
+      "learning_rate": 9.84089416225723e-05,
+      "loss": 1.0676,
+      "step": 1210
+    },
+    {
+      "epoch": 0.10820470435812093,
+      "grad_norm": 0.4201282560825348,
+      "learning_rate": 9.840531816284093e-05,
+      "loss": 1.0493,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1082940558893828,
+      "grad_norm": 0.4624415636062622,
+      "learning_rate": 9.84016906486482e-05,
+      "loss": 1.0656,
+      "step": 1212
+    },
+    {
+      "epoch": 0.10838340742064467,
+      "grad_norm": 0.415123850107193,
+      "learning_rate": 9.839805908029795e-05,
+      "loss": 1.0556,
+      "step": 1213
+    },
+    {
+      "epoch": 0.10847275895190654,
+      "grad_norm": 0.41990217566490173,
+      "learning_rate": 9.839442345809435e-05,
+      "loss": 1.0851,
+      "step": 1214
+    },
+    {
+      "epoch": 0.10856211048316841,
+      "grad_norm": 0.43788111209869385,
+      "learning_rate": 9.839078378234196e-05,
+      "loss": 1.106,
+      "step": 1215
+    },
+    {
+      "epoch": 0.10865146201443027,
+      "grad_norm": 0.4749810993671417,
+      "learning_rate": 9.838714005334562e-05,
+      "loss": 1.0649,
+      "step": 1216
+    },
+    {
+      "epoch": 0.10874081354569214,
+      "grad_norm": 0.4938737750053406,
+      "learning_rate": 9.838349227141051e-05,
+      "loss": 1.0409,
+      "step": 1217
+    },
+    {
+      "epoch": 0.108830165076954,
+      "grad_norm": 0.48466798663139343,
+      "learning_rate": 9.83798404368422e-05,
+      "loss": 1.1215,
+      "step": 1218
+    },
+    {
+      "epoch": 0.10891951660821587,
+      "grad_norm": 0.3890851140022278,
+      "learning_rate": 9.837618454994657e-05,
+      "loss": 1.0944,
+      "step": 1219
+    },
+    {
+      "epoch": 0.10900886813947774,
+      "grad_norm": 0.4145745635032654,
+      "learning_rate": 9.837252461102981e-05,
+      "loss": 1.1748,
+      "step": 1220
+    },
+    {
+      "epoch": 0.10909821967073961,
+      "grad_norm": 0.4121316969394684,
+      "learning_rate": 9.836886062039853e-05,
+      "loss": 1.0474,
+      "step": 1221
+    },
+    {
+      "epoch": 0.10918757120200147,
+      "grad_norm": 0.3703002333641052,
+      "learning_rate": 9.836519257835957e-05,
+      "loss": 1.1139,
+      "step": 1222
+    },
+    {
+      "epoch": 0.10927692273326334,
+      "grad_norm": 0.44502654671669006,
+      "learning_rate": 9.836152048522022e-05,
+      "loss": 1.0527,
+      "step": 1223
+    },
+    {
+      "epoch": 0.1093662742645252,
+      "grad_norm": 0.41378626227378845,
+      "learning_rate": 9.835784434128802e-05,
+      "loss": 1.1426,
+      "step": 1224
+    },
+    {
+      "epoch": 0.10945562579578708,
+      "grad_norm": 0.5465456247329712,
+      "learning_rate": 9.83541641468709e-05,
+      "loss": 1.057,
+      "step": 1225
+    },
+    {
+      "epoch": 0.10954497732704895,
+      "grad_norm": 0.49641984701156616,
+      "learning_rate": 9.835047990227712e-05,
+      "loss": 1.0913,
+      "step": 1226
+    },
+    {
+      "epoch": 0.10963432885831081,
+      "grad_norm": 0.5275992155075073,
+      "learning_rate": 9.834679160781526e-05,
+      "loss": 0.987,
+      "step": 1227
+    },
+    {
+      "epoch": 0.10972368038957267,
+      "grad_norm": 0.43388912081718445,
+      "learning_rate": 9.834309926379426e-05,
+      "loss": 1.0185,
+      "step": 1228
+    },
+    {
+      "epoch": 0.10981303192083454,
+      "grad_norm": 0.5272001028060913,
+      "learning_rate": 9.833940287052341e-05,
+      "loss": 1.0355,
+      "step": 1229
+    },
+    {
+      "epoch": 0.10990238345209641,
+      "grad_norm": 0.39785364270210266,
+      "learning_rate": 9.833570242831229e-05,
+      "loss": 1.0603,
+      "step": 1230
+    },
+    {
+      "epoch": 0.10999173498335828,
+      "grad_norm": 0.5039787292480469,
+      "learning_rate": 9.833199793747089e-05,
+      "loss": 1.0754,
+      "step": 1231
+    },
+    {
+      "epoch": 0.11008108651462015,
+      "grad_norm": 0.5028790831565857,
+      "learning_rate": 9.832828939830947e-05,
+      "loss": 1.0872,
+      "step": 1232
+    },
+    {
+      "epoch": 0.11017043804588202,
+      "grad_norm": 0.4891465902328491,
+      "learning_rate": 9.832457681113866e-05,
+      "loss": 0.9673,
+      "step": 1233
+    },
+    {
+      "epoch": 0.11025978957714389,
+      "grad_norm": 0.5017561316490173,
+      "learning_rate": 9.832086017626947e-05,
+      "loss": 1.0323,
+      "step": 1234
+    },
+    {
+      "epoch": 0.11034914110840574,
+      "grad_norm": 0.41430050134658813,
+      "learning_rate": 9.831713949401316e-05,
+      "loss": 1.0685,
+      "step": 1235
+    },
+    {
+      "epoch": 0.11043849263966761,
+      "grad_norm": 0.4018838703632355,
+      "learning_rate": 9.831341476468139e-05,
+      "loss": 1.0245,
+      "step": 1236
+    },
+    {
+      "epoch": 0.11052784417092948,
+      "grad_norm": 0.48818472027778625,
+      "learning_rate": 9.830968598858614e-05,
+      "loss": 1.0407,
+      "step": 1237
+    },
+    {
+      "epoch": 0.11061719570219135,
+      "grad_norm": 0.45621582865715027,
+      "learning_rate": 9.830595316603976e-05,
+      "loss": 1.0304,
+      "step": 1238
+    },
+    {
+      "epoch": 0.11070654723345322,
+      "grad_norm": 0.5320197343826294,
+      "learning_rate": 9.83022162973549e-05,
+      "loss": 1.0693,
+      "step": 1239
+    },
+    {
+      "epoch": 0.11079589876471509,
+      "grad_norm": 0.4279773533344269,
+      "learning_rate": 9.829847538284455e-05,
+      "loss": 1.1345,
+      "step": 1240
+    },
+    {
+      "epoch": 0.11088525029597694,
+      "grad_norm": 0.4108751118183136,
+      "learning_rate": 9.829473042282207e-05,
+      "loss": 1.1014,
+      "step": 1241
+    },
+    {
+      "epoch": 0.11097460182723881,
+      "grad_norm": 0.5319353938102722,
+      "learning_rate": 9.829098141760111e-05,
+      "loss": 0.9735,
+      "step": 1242
+    },
+    {
+      "epoch": 0.11106395335850068,
+      "grad_norm": 0.42923703789711,
+      "learning_rate": 9.828722836749575e-05,
+      "loss": 1.0158,
+      "step": 1243
+    },
+    {
+      "epoch": 0.11115330488976255,
+      "grad_norm": 0.4357326924800873,
+      "learning_rate": 9.828347127282027e-05,
+      "loss": 1.0393,
+      "step": 1244
+    },
+    {
+      "epoch": 0.11124265642102442,
+      "grad_norm": 0.3578936755657196,
+      "learning_rate": 9.827971013388944e-05,
+      "loss": 1.0764,
+      "step": 1245
+    },
+    {
+      "epoch": 0.11133200795228629,
+      "grad_norm": 0.35362938046455383,
+      "learning_rate": 9.827594495101823e-05,
+      "loss": 1.1115,
+      "step": 1246
+    },
+    {
+      "epoch": 0.11142135948354814,
+      "grad_norm": 0.41567638516426086,
+      "learning_rate": 9.827217572452208e-05,
+      "loss": 1.1021,
+      "step": 1247
+    },
+    {
+      "epoch": 0.11151071101481001,
+      "grad_norm": 0.40736740827560425,
+      "learning_rate": 9.826840245471665e-05,
+      "loss": 1.03,
+      "step": 1248
+    },
+    {
+      "epoch": 0.11160006254607188,
+      "grad_norm": 0.4363911747932434,
+      "learning_rate": 9.826462514191801e-05,
+      "loss": 1.0462,
+      "step": 1249
+    },
+    {
+      "epoch": 0.11168941407733375,
+      "grad_norm": 0.4702802896499634,
+      "learning_rate": 9.826084378644254e-05,
+      "loss": 1.029,
+      "step": 1250
+    },
+    {
+      "epoch": 0.11177876560859562,
+      "grad_norm": 0.45278483629226685,
+      "learning_rate": 9.8257058388607e-05,
+      "loss": 1.1249,
+      "step": 1251
+    },
+    {
+      "epoch": 0.11186811713985749,
+      "grad_norm": 0.47644123435020447,
+      "learning_rate": 9.825326894872842e-05,
+      "loss": 1.0096,
+      "step": 1252
+    },
+    {
+      "epoch": 0.11195746867111935,
+      "grad_norm": 0.48887526988983154,
+      "learning_rate": 9.824947546712424e-05,
+      "loss": 1.0465,
+      "step": 1253
+    },
+    {
+      "epoch": 0.11204682020238121,
+      "grad_norm": 0.4377509355545044,
+      "learning_rate": 9.824567794411216e-05,
+      "loss": 1.049,
+      "step": 1254
+    },
+    {
+      "epoch": 0.11213617173364308,
+      "grad_norm": 0.5497344732284546,
+      "learning_rate": 9.824187638001032e-05,
+      "loss": 0.9813,
+      "step": 1255
+    },
+    {
+      "epoch": 0.11222552326490495,
+      "grad_norm": 0.39412158727645874,
+      "learning_rate": 9.82380707751371e-05,
+      "loss": 1.0328,
+      "step": 1256
+    },
+    {
+      "epoch": 0.11231487479616682,
+      "grad_norm": 0.5611281394958496,
+      "learning_rate": 9.823426112981126e-05,
+      "loss": 0.9994,
+      "step": 1257
+    },
+    {
+      "epoch": 0.11240422632742869,
+      "grad_norm": 0.4828655421733856,
+      "learning_rate": 9.823044744435193e-05,
+      "loss": 1.062,
+      "step": 1258
+    },
+    {
+      "epoch": 0.11249357785869055,
+      "grad_norm": 0.3836156129837036,
+      "learning_rate": 9.822662971907852e-05,
+      "loss": 1.1001,
+      "step": 1259
+    },
+    {
+      "epoch": 0.11258292938995242,
+      "grad_norm": 0.42271852493286133,
+      "learning_rate": 9.822280795431082e-05,
+      "loss": 1.0411,
+      "step": 1260
+    },
+    {
+      "epoch": 0.11267228092121429,
+      "grad_norm": 0.4254275858402252,
+      "learning_rate": 9.821898215036891e-05,
+      "loss": 1.0963,
+      "step": 1261
+    },
+    {
+      "epoch": 0.11276163245247615,
+      "grad_norm": 0.46899744868278503,
+      "learning_rate": 9.821515230757329e-05,
+      "loss": 1.054,
+      "step": 1262
+    },
+    {
+      "epoch": 0.11285098398373802,
+      "grad_norm": 0.4628446400165558,
+      "learning_rate": 9.821131842624471e-05,
+      "loss": 1.0433,
+      "step": 1263
+    },
+    {
+      "epoch": 0.11294033551499989,
+      "grad_norm": 0.4287131726741791,
+      "learning_rate": 9.820748050670433e-05,
+      "loss": 1.0086,
+      "step": 1264
+    },
+    {
+      "epoch": 0.11302968704626176,
+      "grad_norm": 0.4528293013572693,
+      "learning_rate": 9.820363854927362e-05,
+      "loss": 1.0296,
+      "step": 1265
+    },
+    {
+      "epoch": 0.11311903857752362,
+      "grad_norm": 0.5807567834854126,
+      "learning_rate": 9.819979255427434e-05,
+      "loss": 1.0242,
+      "step": 1266
+    },
+    {
+      "epoch": 0.11320839010878549,
+      "grad_norm": 0.4787788391113281,
+      "learning_rate": 9.819594252202866e-05,
+      "loss": 1.0764,
+      "step": 1267
+    },
+    {
+      "epoch": 0.11329774164004736,
+      "grad_norm": 0.4397558271884918,
+      "learning_rate": 9.819208845285908e-05,
+      "loss": 1.0668,
+      "step": 1268
+    },
+    {
+      "epoch": 0.11338709317130923,
+      "grad_norm": 0.5197393894195557,
+      "learning_rate": 9.81882303470884e-05,
+      "loss": 1.0614,
+      "step": 1269
+    },
+    {
+      "epoch": 0.1134764447025711,
+      "grad_norm": 0.4312942326068878,
+      "learning_rate": 9.818436820503976e-05,
+      "loss": 1.1124,
+      "step": 1270
+    },
+    {
+      "epoch": 0.11356579623383296,
+      "grad_norm": 0.36695683002471924,
+      "learning_rate": 9.818050202703668e-05,
+      "loss": 1.0542,
+      "step": 1271
+    },
+    {
+      "epoch": 0.11365514776509482,
+      "grad_norm": 0.47956258058547974,
+      "learning_rate": 9.817663181340299e-05,
+      "loss": 1.0063,
+      "step": 1272
+    },
+    {
+      "epoch": 0.11374449929635669,
+      "grad_norm": 0.40785545110702515,
+      "learning_rate": 9.817275756446287e-05,
+      "loss": 1.0145,
+      "step": 1273
+    },
+    {
+      "epoch": 0.11383385082761856,
+      "grad_norm": 0.5532562732696533,
+      "learning_rate": 9.816887928054082e-05,
+      "loss": 1.0415,
+      "step": 1274
+    },
+    {
+      "epoch": 0.11392320235888043,
+      "grad_norm": 0.4029194414615631,
+      "learning_rate": 9.816499696196167e-05,
+      "loss": 1.0276,
+      "step": 1275
+    },
+    {
+      "epoch": 0.1140125538901423,
+      "grad_norm": 0.38297927379608154,
+      "learning_rate": 9.816111060905062e-05,
+      "loss": 1.0914,
+      "step": 1276
+    },
+    {
+      "epoch": 0.11410190542140417,
+      "grad_norm": 0.44156894087791443,
+      "learning_rate": 9.815722022213322e-05,
+      "loss": 1.0469,
+      "step": 1277
+    },
+    {
+      "epoch": 0.11419125695266602,
+      "grad_norm": 0.44452789425849915,
+      "learning_rate": 9.81533258015353e-05,
+      "loss": 0.9441,
+      "step": 1278
+    },
+    {
+      "epoch": 0.11428060848392789,
+      "grad_norm": 0.4572380483150482,
+      "learning_rate": 9.814942734758306e-05,
+      "loss": 1.0353,
+      "step": 1279
+    },
+    {
+      "epoch": 0.11436996001518976,
+      "grad_norm": 0.39455848932266235,
+      "learning_rate": 9.814552486060305e-05,
+      "loss": 1.0873,
+      "step": 1280
+    },
+    {
+      "epoch": 0.11445931154645163,
+      "grad_norm": 0.3743246793746948,
+      "learning_rate": 9.814161834092212e-05,
+      "loss": 1.0684,
+      "step": 1281
+    },
+    {
+      "epoch": 0.1145486630777135,
+      "grad_norm": 0.4070795774459839,
+      "learning_rate": 9.81377077888675e-05,
+      "loss": 1.143,
+      "step": 1282
+    },
+    {
+      "epoch": 0.11463801460897537,
+      "grad_norm": 0.4575822353363037,
+      "learning_rate": 9.813379320476677e-05,
+      "loss": 1.0032,
+      "step": 1283
+    },
+    {
+      "epoch": 0.11472736614023722,
+      "grad_norm": 0.4624904692173004,
+      "learning_rate": 9.812987458894778e-05,
+      "loss": 1.1421,
+      "step": 1284
+    },
+    {
+      "epoch": 0.11481671767149909,
+      "grad_norm": 0.45701712369918823,
+      "learning_rate": 9.812595194173875e-05,
+      "loss": 1.0939,
+      "step": 1285
+    },
+    {
+      "epoch": 0.11490606920276096,
+      "grad_norm": 0.4664194583892822,
+      "learning_rate": 9.812202526346827e-05,
+      "loss": 1.0381,
+      "step": 1286
+    },
+    {
+      "epoch": 0.11499542073402283,
+      "grad_norm": 0.42624375224113464,
+      "learning_rate": 9.811809455446523e-05,
+      "loss": 1.0903,
+      "step": 1287
+    },
+    {
+      "epoch": 0.1150847722652847,
+      "grad_norm": 0.40351542830467224,
+      "learning_rate": 9.811415981505887e-05,
+      "loss": 1.0591,
+      "step": 1288
+    },
+    {
+      "epoch": 0.11517412379654657,
+      "grad_norm": 0.46651527285575867,
+      "learning_rate": 9.811022104557877e-05,
+      "loss": 1.0647,
+      "step": 1289
+    },
+    {
+      "epoch": 0.11526347532780842,
+      "grad_norm": 0.4483588635921478,
+      "learning_rate": 9.810627824635483e-05,
+      "loss": 1.0734,
+      "step": 1290
+    },
+    {
+      "epoch": 0.11535282685907029,
+      "grad_norm": 0.45187729597091675,
+      "learning_rate": 9.810233141771732e-05,
+      "loss": 1.0024,
+      "step": 1291
+    },
+    {
+      "epoch": 0.11544217839033216,
+      "grad_norm": 0.43645814061164856,
+      "learning_rate": 9.809838055999681e-05,
+      "loss": 1.1,
+      "step": 1292
+    },
+    {
+      "epoch": 0.11553152992159403,
+      "grad_norm": 0.42564859986305237,
+      "learning_rate": 9.809442567352425e-05,
+      "loss": 1.0762,
+      "step": 1293
+    },
+    {
+      "epoch": 0.1156208814528559,
+      "grad_norm": 0.419881135225296,
+      "learning_rate": 9.809046675863087e-05,
+      "loss": 1.1093,
+      "step": 1294
+    },
+    {
+      "epoch": 0.11571023298411777,
+      "grad_norm": 0.46777603030204773,
+      "learning_rate": 9.808650381564831e-05,
+      "loss": 1.0415,
+      "step": 1295
+    },
+    {
+      "epoch": 0.11579958451537964,
+      "grad_norm": 0.4529683589935303,
+      "learning_rate": 9.80825368449085e-05,
+      "loss": 1.0256,
+      "step": 1296
+    },
+    {
+      "epoch": 0.1158889360466415,
+      "grad_norm": 0.3957824110984802,
+      "learning_rate": 9.807856584674368e-05,
+      "loss": 1.1027,
+      "step": 1297
+    },
+    {
+      "epoch": 0.11597828757790336,
+      "grad_norm": 0.3844730854034424,
+      "learning_rate": 9.807459082148648e-05,
+      "loss": 1.0713,
+      "step": 1298
+    },
+    {
+      "epoch": 0.11606763910916523,
+      "grad_norm": 0.3445946276187897,
+      "learning_rate": 9.80706117694699e-05,
+      "loss": 1.1058,
+      "step": 1299
+    },
+    {
+      "epoch": 0.1161569906404271,
+      "grad_norm": 0.39715394377708435,
+      "learning_rate": 9.806662869102717e-05,
+      "loss": 1.0711,
+      "step": 1300
+    },
+    {
+      "epoch": 0.11624634217168897,
+      "grad_norm": 0.5070072412490845,
+      "learning_rate": 9.806264158649193e-05,
+      "loss": 1.0142,
+      "step": 1301
+    },
+    {
+      "epoch": 0.11633569370295084,
+      "grad_norm": 0.4780667722225189,
+      "learning_rate": 9.805865045619813e-05,
+      "loss": 1.0362,
+      "step": 1302
+    },
+    {
+      "epoch": 0.1164250452342127,
+      "grad_norm": 0.45890098810195923,
+      "learning_rate": 9.80546553004801e-05,
+      "loss": 0.9997,
+      "step": 1303
+    },
+    {
+      "epoch": 0.11651439676547456,
+      "grad_norm": 0.46433109045028687,
+      "learning_rate": 9.805065611967248e-05,
+      "loss": 0.9268,
+      "step": 1304
+    },
+    {
+      "epoch": 0.11660374829673643,
+      "grad_norm": 0.43207046389579773,
+      "learning_rate": 9.804665291411022e-05,
+      "loss": 1.0814,
+      "step": 1305
+    },
+    {
+      "epoch": 0.1166930998279983,
+      "grad_norm": 0.3997035622596741,
+      "learning_rate": 9.804264568412862e-05,
+      "loss": 1.0454,
+      "step": 1306
+    },
+    {
+      "epoch": 0.11678245135926017,
+      "grad_norm": 0.4062361717224121,
+      "learning_rate": 9.803863443006336e-05,
+      "loss": 1.1057,
+      "step": 1307
+    },
+    {
+      "epoch": 0.11687180289052204,
+      "grad_norm": 0.3779062032699585,
+      "learning_rate": 9.80346191522504e-05,
+      "loss": 1.1132,
+      "step": 1308
+    },
+    {
+      "epoch": 0.1169611544217839,
+      "grad_norm": 0.4400861859321594,
+      "learning_rate": 9.803059985102609e-05,
+      "loss": 1.0412,
+      "step": 1309
+    },
+    {
+      "epoch": 0.11705050595304577,
+      "grad_norm": 0.48002249002456665,
+      "learning_rate": 9.802657652672706e-05,
+      "loss": 0.9414,
+      "step": 1310
+    },
+    {
+      "epoch": 0.11713985748430764,
+      "grad_norm": 0.5008269548416138,
+      "learning_rate": 9.802254917969032e-05,
+      "loss": 1.0162,
+      "step": 1311
+    },
+    {
+      "epoch": 0.1172292090155695,
+      "grad_norm": 0.4323752224445343,
+      "learning_rate": 9.801851781025322e-05,
+      "loss": 1.1209,
+      "step": 1312
+    },
+    {
+      "epoch": 0.11731856054683137,
+      "grad_norm": 0.5104453563690186,
+      "learning_rate": 9.80144824187534e-05,
+      "loss": 0.9974,
+      "step": 1313
+    },
+    {
+      "epoch": 0.11740791207809324,
+      "grad_norm": 0.4443637430667877,
+      "learning_rate": 9.801044300552887e-05,
+      "loss": 1.0159,
+      "step": 1314
+    },
+    {
+      "epoch": 0.1174972636093551,
+      "grad_norm": 0.44141674041748047,
+      "learning_rate": 9.800639957091799e-05,
+      "loss": 1.0203,
+      "step": 1315
+    },
+    {
+      "epoch": 0.11758661514061697,
+      "grad_norm": 0.44127362966537476,
+      "learning_rate": 9.800235211525945e-05,
+      "loss": 1.1218,
+      "step": 1316
+    },
+    {
+      "epoch": 0.11767596667187884,
+      "grad_norm": 0.401253342628479,
+      "learning_rate": 9.799830063889223e-05,
+      "loss": 1.1097,
+      "step": 1317
+    },
+    {
+      "epoch": 0.1177653182031407,
+      "grad_norm": 0.5445295572280884,
+      "learning_rate": 9.799424514215572e-05,
+      "loss": 1.0315,
+      "step": 1318
+    },
+    {
+      "epoch": 0.11785466973440258,
+      "grad_norm": 0.3961075246334076,
+      "learning_rate": 9.79901856253896e-05,
+      "loss": 1.0794,
+      "step": 1319
+    },
+    {
+      "epoch": 0.11794402126566444,
+      "grad_norm": 0.4312506318092346,
+      "learning_rate": 9.798612208893389e-05,
+      "loss": 1.0732,
+      "step": 1320
+    },
+    {
+      "epoch": 0.11803337279692631,
+      "grad_norm": 0.5154565572738647,
+      "learning_rate": 9.798205453312895e-05,
+      "loss": 1.1045,
+      "step": 1321
+    },
+    {
+      "epoch": 0.11812272432818817,
+      "grad_norm": 0.4929812550544739,
+      "learning_rate": 9.79779829583155e-05,
+      "loss": 1.009,
+      "step": 1322
+    },
+    {
+      "epoch": 0.11821207585945004,
+      "grad_norm": 0.39952999353408813,
+      "learning_rate": 9.797390736483459e-05,
+      "loss": 1.0558,
+      "step": 1323
+    },
+    {
+      "epoch": 0.11830142739071191,
+      "grad_norm": 0.47065678238868713,
+      "learning_rate": 9.796982775302755e-05,
+      "loss": 1.0119,
+      "step": 1324
+    },
+    {
+      "epoch": 0.11839077892197378,
+      "grad_norm": 0.39249664545059204,
+      "learning_rate": 9.796574412323611e-05,
+      "loss": 1.0805,
+      "step": 1325
+    },
+    {
+      "epoch": 0.11848013045323565,
+      "grad_norm": 0.38709941506385803,
+      "learning_rate": 9.796165647580233e-05,
+      "loss": 1.0919,
+      "step": 1326
+    },
+    {
+      "epoch": 0.11856948198449752,
+      "grad_norm": 0.4576520323753357,
+      "learning_rate": 9.795756481106857e-05,
+      "loss": 1.082,
+      "step": 1327
+    },
+    {
+      "epoch": 0.11865883351575937,
+      "grad_norm": 0.42923304438591003,
+      "learning_rate": 9.795346912937757e-05,
+      "loss": 0.9918,
+      "step": 1328
+    },
+    {
+      "epoch": 0.11874818504702124,
+      "grad_norm": 0.38031134009361267,
+      "learning_rate": 9.79493694310724e-05,
+      "loss": 1.0198,
+      "step": 1329
+    },
+    {
+      "epoch": 0.11883753657828311,
+      "grad_norm": 0.39930686354637146,
+      "learning_rate": 9.794526571649643e-05,
+      "loss": 1.0082,
+      "step": 1330
+    },
+    {
+      "epoch": 0.11892688810954498,
+      "grad_norm": 0.463544636964798,
+      "learning_rate": 9.794115798599339e-05,
+      "loss": 1.1003,
+      "step": 1331
+    },
+    {
+      "epoch": 0.11901623964080685,
+      "grad_norm": 0.42002353072166443,
+      "learning_rate": 9.793704623990736e-05,
+      "loss": 1.0518,
+      "step": 1332
+    },
+    {
+      "epoch": 0.11910559117206872,
+      "grad_norm": 0.39078399538993835,
+      "learning_rate": 9.793293047858274e-05,
+      "loss": 1.0775,
+      "step": 1333
+    },
+    {
+      "epoch": 0.11919494270333057,
+      "grad_norm": 0.44137701392173767,
+      "learning_rate": 9.792881070236426e-05,
+      "loss": 1.0588,
+      "step": 1334
+    },
+    {
+      "epoch": 0.11928429423459244,
+      "grad_norm": 0.40128040313720703,
+      "learning_rate": 9.7924686911597e-05,
+      "loss": 1.0676,
+      "step": 1335
+    },
+    {
+      "epoch": 0.11937364576585431,
+      "grad_norm": 0.46440789103507996,
+      "learning_rate": 9.792055910662636e-05,
+      "loss": 1.1613,
+      "step": 1336
+    },
+    {
+      "epoch": 0.11946299729711618,
+      "grad_norm": 0.36990535259246826,
+      "learning_rate": 9.791642728779811e-05,
+      "loss": 1.0511,
+      "step": 1337
+    },
+    {
+      "epoch": 0.11955234882837805,
+      "grad_norm": 0.47823598980903625,
+      "learning_rate": 9.791229145545831e-05,
+      "loss": 0.9977,
+      "step": 1338
+    },
+    {
+      "epoch": 0.11964170035963992,
+      "grad_norm": 0.3946368992328644,
+      "learning_rate": 9.790815160995342e-05,
+      "loss": 1.0871,
+      "step": 1339
+    },
+    {
+      "epoch": 0.11973105189090177,
+      "grad_norm": 0.5002843141555786,
+      "learning_rate": 9.790400775163014e-05,
+      "loss": 0.9361,
+      "step": 1340
+    },
+    {
+      "epoch": 0.11982040342216364,
+      "grad_norm": 0.4414635896682739,
+      "learning_rate": 9.789985988083558e-05,
+      "loss": 1.0857,
+      "step": 1341
+    },
+    {
+      "epoch": 0.11990975495342551,
+      "grad_norm": 0.4567340314388275,
+      "learning_rate": 9.789570799791721e-05,
+      "loss": 0.9954,
+      "step": 1342
+    },
+    {
+      "epoch": 0.11999910648468738,
+      "grad_norm": 0.46271204948425293,
+      "learning_rate": 9.789155210322276e-05,
+      "loss": 0.9978,
+      "step": 1343
+    },
+    {
+      "epoch": 0.12008845801594925,
+      "grad_norm": 0.45726093649864197,
+      "learning_rate": 9.788739219710032e-05,
+      "loss": 1.0101,
+      "step": 1344
+    },
+    {
+      "epoch": 0.12017780954721112,
+      "grad_norm": 0.4213901162147522,
+      "learning_rate": 9.788322827989836e-05,
+      "loss": 1.061,
+      "step": 1345
+    },
+    {
+      "epoch": 0.12026716107847298,
+      "grad_norm": 0.5426849126815796,
+      "learning_rate": 9.787906035196562e-05,
+      "loss": 1.0625,
+      "step": 1346
+    },
+    {
+      "epoch": 0.12035651260973484,
+      "grad_norm": 0.4493926167488098,
+      "learning_rate": 9.787488841365122e-05,
+      "loss": 1.021,
+      "step": 1347
+    },
+    {
+      "epoch": 0.12044586414099671,
+      "grad_norm": 0.47702398896217346,
+      "learning_rate": 9.787071246530459e-05,
+      "loss": 1.0161,
+      "step": 1348
+    },
+    {
+      "epoch": 0.12053521567225858,
+      "grad_norm": 0.5239232778549194,
+      "learning_rate": 9.786653250727555e-05,
+      "loss": 1.0321,
+      "step": 1349
+    },
+    {
+      "epoch": 0.12062456720352045,
+      "grad_norm": 0.4591730237007141,
+      "learning_rate": 9.786234853991418e-05,
+      "loss": 0.9884,
+      "step": 1350
+    },
+    {
+      "epoch": 0.12071391873478232,
+      "grad_norm": 0.39837637543678284,
+      "learning_rate": 9.785816056357095e-05,
+      "loss": 1.0764,
+      "step": 1351
+    },
+    {
+      "epoch": 0.12080327026604419,
+      "grad_norm": 0.42193183302879333,
+      "learning_rate": 9.785396857859664e-05,
+      "loss": 1.0636,
+      "step": 1352
+    },
+    {
+      "epoch": 0.12089262179730605,
+      "grad_norm": 0.48460525274276733,
+      "learning_rate": 9.784977258534239e-05,
+      "loss": 1.0683,
+      "step": 1353
+    },
+    {
+      "epoch": 0.12098197332856792,
+      "grad_norm": 0.45954951643943787,
+      "learning_rate": 9.784557258415963e-05,
+      "loss": 1.0613,
+      "step": 1354
+    },
+    {
+      "epoch": 0.12107132485982978,
+      "grad_norm": 0.4340897500514984,
+      "learning_rate": 9.784136857540015e-05,
+      "loss": 0.9937,
+      "step": 1355
+    },
+    {
+      "epoch": 0.12116067639109165,
+      "grad_norm": 0.39842215180397034,
+      "learning_rate": 9.783716055941612e-05,
+      "loss": 1.0619,
+      "step": 1356
+    },
+    {
+      "epoch": 0.12125002792235352,
+      "grad_norm": 0.38093817234039307,
+      "learning_rate": 9.783294853655999e-05,
+      "loss": 1.0357,
+      "step": 1357
+    },
+    {
+      "epoch": 0.12133937945361539,
+      "grad_norm": 0.5032150149345398,
+      "learning_rate": 9.782873250718455e-05,
+      "loss": 0.914,
+      "step": 1358
+    },
+    {
+      "epoch": 0.12142873098487725,
+      "grad_norm": 0.4284474551677704,
+      "learning_rate": 9.782451247164295e-05,
+      "loss": 1.0058,
+      "step": 1359
+    },
+    {
+      "epoch": 0.12151808251613912,
+      "grad_norm": 0.40385428071022034,
+      "learning_rate": 9.782028843028865e-05,
+      "loss": 1.0313,
+      "step": 1360
+    },
+    {
+      "epoch": 0.12160743404740099,
+      "grad_norm": 0.48033663630485535,
+      "learning_rate": 9.781606038347547e-05,
+      "loss": 1.005,
+      "step": 1361
+    },
+    {
+      "epoch": 0.12169678557866286,
+      "grad_norm": 0.3936779201030731,
+      "learning_rate": 9.781182833155755e-05,
+      "loss": 1.0181,
+      "step": 1362
+    },
+    {
+      "epoch": 0.12178613710992472,
+      "grad_norm": 0.46711423993110657,
+      "learning_rate": 9.780759227488936e-05,
+      "loss": 1.0816,
+      "step": 1363
+    },
+    {
+      "epoch": 0.1218754886411866,
+      "grad_norm": 0.43602418899536133,
+      "learning_rate": 9.780335221382574e-05,
+      "loss": 1.1094,
+      "step": 1364
+    },
+    {
+      "epoch": 0.12196484017244845,
+      "grad_norm": 0.3990572690963745,
+      "learning_rate": 9.779910814872182e-05,
+      "loss": 0.9995,
+      "step": 1365
+    },
+    {
+      "epoch": 0.12205419170371032,
+      "grad_norm": 0.5048510432243347,
+      "learning_rate": 9.77948600799331e-05,
+      "loss": 1.0082,
+      "step": 1366
+    },
+    {
+      "epoch": 0.12214354323497219,
+      "grad_norm": 0.3975992202758789,
+      "learning_rate": 9.779060800781537e-05,
+      "loss": 1.0578,
+      "step": 1367
+    },
+    {
+      "epoch": 0.12223289476623406,
+      "grad_norm": 0.4395469129085541,
+      "learning_rate": 9.778635193272483e-05,
+      "loss": 1.0318,
+      "step": 1368
+    },
+    {
+      "epoch": 0.12232224629749593,
+      "grad_norm": 0.4118629992008209,
+      "learning_rate": 9.778209185501794e-05,
+      "loss": 1.0436,
+      "step": 1369
+    },
+    {
+      "epoch": 0.1224115978287578,
+      "grad_norm": 0.4724205732345581,
+      "learning_rate": 9.777782777505153e-05,
+      "loss": 0.9656,
+      "step": 1370
+    },
+    {
+      "epoch": 0.12250094936001965,
+      "grad_norm": 0.49588900804519653,
+      "learning_rate": 9.777355969318278e-05,
+      "loss": 0.9988,
+      "step": 1371
+    },
+    {
+      "epoch": 0.12259030089128152,
+      "grad_norm": 0.4462149441242218,
+      "learning_rate": 9.776928760976918e-05,
+      "loss": 1.0914,
+      "step": 1372
+    },
+    {
+      "epoch": 0.12267965242254339,
+      "grad_norm": 0.3685830533504486,
+      "learning_rate": 9.776501152516855e-05,
+      "loss": 1.0973,
+      "step": 1373
+    },
+    {
+      "epoch": 0.12276900395380526,
+      "grad_norm": 0.5166415572166443,
+      "learning_rate": 9.776073143973904e-05,
+      "loss": 0.9822,
+      "step": 1374
+    },
+    {
+      "epoch": 0.12285835548506713,
+      "grad_norm": 0.4535563588142395,
+      "learning_rate": 9.775644735383922e-05,
+      "loss": 1.0124,
+      "step": 1375
+    },
+    {
+      "epoch": 0.122947707016329,
+      "grad_norm": 0.40610021352767944,
+      "learning_rate": 9.775215926782788e-05,
+      "loss": 1.0909,
+      "step": 1376
+    },
+    {
+      "epoch": 0.12303705854759087,
+      "grad_norm": 0.4109049141407013,
+      "learning_rate": 9.774786718206419e-05,
+      "loss": 1.0908,
+      "step": 1377
+    },
+    {
+      "epoch": 0.12312641007885272,
+      "grad_norm": 0.4295797348022461,
+      "learning_rate": 9.774357109690767e-05,
+      "loss": 1.0351,
+      "step": 1378
+    },
+    {
+      "epoch": 0.12321576161011459,
+      "grad_norm": 0.49584078788757324,
+      "learning_rate": 9.773927101271816e-05,
+      "loss": 1.0216,
+      "step": 1379
+    },
+    {
+      "epoch": 0.12330511314137646,
+      "grad_norm": 0.44206714630126953,
+      "learning_rate": 9.773496692985584e-05,
+      "loss": 0.9854,
+      "step": 1380
+    },
+    {
+      "epoch": 0.12339446467263833,
+      "grad_norm": 0.39388856291770935,
+      "learning_rate": 9.773065884868122e-05,
+      "loss": 1.0904,
+      "step": 1381
+    },
+    {
+      "epoch": 0.1234838162039002,
+      "grad_norm": 0.4499436914920807,
+      "learning_rate": 9.772634676955515e-05,
+      "loss": 1.0563,
+      "step": 1382
+    },
+    {
+      "epoch": 0.12357316773516207,
+      "grad_norm": 0.48682132363319397,
+      "learning_rate": 9.772203069283881e-05,
+      "loss": 1.0727,
+      "step": 1383
+    },
+    {
+      "epoch": 0.12366251926642392,
+      "grad_norm": 0.38047704100608826,
+      "learning_rate": 9.771771061889373e-05,
+      "loss": 1.0871,
+      "step": 1384
+    },
+    {
+      "epoch": 0.12375187079768579,
+      "grad_norm": 0.46172595024108887,
+      "learning_rate": 9.771338654808173e-05,
+      "loss": 1.0829,
+      "step": 1385
+    },
+    {
+      "epoch": 0.12384122232894766,
+      "grad_norm": 0.44145411252975464,
+      "learning_rate": 9.770905848076504e-05,
+      "loss": 1.0631,
+      "step": 1386
+    },
+    {
+      "epoch": 0.12393057386020953,
+      "grad_norm": 0.4803364872932434,
+      "learning_rate": 9.770472641730615e-05,
+      "loss": 0.9805,
+      "step": 1387
+    },
+    {
+      "epoch": 0.1240199253914714,
+      "grad_norm": 0.4117061197757721,
+      "learning_rate": 9.770039035806792e-05,
+      "loss": 1.023,
+      "step": 1388
+    },
+    {
+      "epoch": 0.12410927692273327,
+      "grad_norm": 0.4862026572227478,
+      "learning_rate": 9.769605030341357e-05,
+      "loss": 1.0725,
+      "step": 1389
+    },
+    {
+      "epoch": 0.12419862845399512,
+      "grad_norm": 0.4829283654689789,
+      "learning_rate": 9.769170625370658e-05,
+      "loss": 0.993,
+      "step": 1390
+    },
+    {
+      "epoch": 0.124287979985257,
+      "grad_norm": 0.4735516309738159,
+      "learning_rate": 9.768735820931085e-05,
+      "loss": 1.059,
+      "step": 1391
+    },
+    {
+      "epoch": 0.12437733151651886,
+      "grad_norm": 0.4585529565811157,
+      "learning_rate": 9.768300617059055e-05,
+      "loss": 1.0455,
+      "step": 1392
+    },
+    {
+      "epoch": 0.12446668304778073,
+      "grad_norm": 0.40741658210754395,
+      "learning_rate": 9.767865013791022e-05,
+      "loss": 0.9718,
+      "step": 1393
+    },
+    {
+      "epoch": 0.1245560345790426,
+      "grad_norm": 0.4379764497280121,
+      "learning_rate": 9.767429011163473e-05,
+      "loss": 1.0147,
+      "step": 1394
+    },
+    {
+      "epoch": 0.12464538611030447,
+      "grad_norm": 0.46133920550346375,
+      "learning_rate": 9.766992609212926e-05,
+      "loss": 1.037,
+      "step": 1395
+    },
+    {
+      "epoch": 0.12473473764156633,
+      "grad_norm": 0.4180471897125244,
+      "learning_rate": 9.766555807975936e-05,
+      "loss": 1.0846,
+      "step": 1396
+    },
+    {
+      "epoch": 0.1248240891728282,
+      "grad_norm": 0.5262375473976135,
+      "learning_rate": 9.76611860748909e-05,
+      "loss": 1.0615,
+      "step": 1397
+    },
+    {
+      "epoch": 0.12491344070409006,
+      "grad_norm": 0.4973163604736328,
+      "learning_rate": 9.765681007789008e-05,
+      "loss": 1.0329,
+      "step": 1398
+    },
+    {
+      "epoch": 0.12500279223535193,
+      "grad_norm": 0.48663395643234253,
+      "learning_rate": 9.765243008912342e-05,
+      "loss": 1.011,
+      "step": 1399
+    },
+    {
+      "epoch": 0.1250921437666138,
+      "grad_norm": 0.52272629737854,
+      "learning_rate": 9.76480461089578e-05,
+      "loss": 1.0271,
+      "step": 1400
+    },
+    {
+      "epoch": 0.12518149529787567,
+      "grad_norm": 0.5239830613136292,
+      "learning_rate": 9.764365813776042e-05,
+      "loss": 1.061,
+      "step": 1401
+    },
+    {
+      "epoch": 0.12527084682913753,
+      "grad_norm": 0.44980114698410034,
+      "learning_rate": 9.763926617589883e-05,
+      "loss": 1.0671,
+      "step": 1402
+    },
+    {
+      "epoch": 0.1253601983603994,
+      "grad_norm": 0.4098733961582184,
+      "learning_rate": 9.763487022374092e-05,
+      "loss": 0.9897,
+      "step": 1403
+    },
+    {
+      "epoch": 0.12544954989166127,
+      "grad_norm": 0.360730916261673,
+      "learning_rate": 9.763047028165484e-05,
+      "loss": 1.0978,
+      "step": 1404
+    },
+    {
+      "epoch": 0.12553890142292312,
+      "grad_norm": 0.518429160118103,
+      "learning_rate": 9.762606635000919e-05,
+      "loss": 0.9659,
+      "step": 1405
+    },
+    {
+      "epoch": 0.125628252954185,
+      "grad_norm": 0.4658845365047455,
+      "learning_rate": 9.762165842917283e-05,
+      "loss": 1.0005,
+      "step": 1406
+    },
+    {
+      "epoch": 0.12571760448544686,
+      "grad_norm": 0.3965921700000763,
+      "learning_rate": 9.761724651951498e-05,
+      "loss": 1.029,
+      "step": 1407
+    },
+    {
+      "epoch": 0.12580695601670874,
+      "grad_norm": 0.4184513986110687,
+      "learning_rate": 9.761283062140514e-05,
+      "loss": 0.9864,
+      "step": 1408
+    },
+    {
+      "epoch": 0.1258963075479706,
+      "grad_norm": 0.37843790650367737,
+      "learning_rate": 9.760841073521323e-05,
+      "loss": 1.1095,
+      "step": 1409
+    },
+    {
+      "epoch": 0.12598565907923248,
+      "grad_norm": 0.4006551504135132,
+      "learning_rate": 9.760398686130946e-05,
+      "loss": 1.0732,
+      "step": 1410
+    },
+    {
+      "epoch": 0.12607501061049434,
+      "grad_norm": 0.407956600189209,
+      "learning_rate": 9.759955900006436e-05,
+      "loss": 1.064,
+      "step": 1411
+    },
+    {
+      "epoch": 0.1261643621417562,
+      "grad_norm": 0.42004308104515076,
+      "learning_rate": 9.759512715184881e-05,
+      "loss": 1.0558,
+      "step": 1412
+    },
+    {
+      "epoch": 0.12625371367301808,
+      "grad_norm": 0.411745548248291,
+      "learning_rate": 9.759069131703406e-05,
+      "loss": 1.0318,
+      "step": 1413
+    },
+    {
+      "epoch": 0.12634306520427993,
+      "grad_norm": 0.40602388978004456,
+      "learning_rate": 9.75862514959916e-05,
+      "loss": 1.0533,
+      "step": 1414
+    },
+    {
+      "epoch": 0.1264324167355418,
+      "grad_norm": 0.4616459906101227,
+      "learning_rate": 9.758180768909337e-05,
+      "loss": 1.0691,
+      "step": 1415
+    },
+    {
+      "epoch": 0.12652176826680367,
+      "grad_norm": 0.5043428540229797,
+      "learning_rate": 9.757735989671156e-05,
+      "loss": 1.0047,
+      "step": 1416
+    },
+    {
+      "epoch": 0.12661111979806555,
+      "grad_norm": 0.4009447693824768,
+      "learning_rate": 9.75729081192187e-05,
+      "loss": 1.0422,
+      "step": 1417
+    },
+    {
+      "epoch": 0.1267004713293274,
+      "grad_norm": 0.3939031660556793,
+      "learning_rate": 9.756845235698772e-05,
+      "loss": 1.0726,
+      "step": 1418
+    },
+    {
+      "epoch": 0.12678982286058926,
+      "grad_norm": 0.42009732127189636,
+      "learning_rate": 9.756399261039179e-05,
+      "loss": 1.0992,
+      "step": 1419
+    },
+    {
+      "epoch": 0.12687917439185115,
+      "grad_norm": 0.475175142288208,
+      "learning_rate": 9.75595288798045e-05,
+      "loss": 1.079,
+      "step": 1420
+    },
+    {
+      "epoch": 0.126968525923113,
+      "grad_norm": 0.42921051383018494,
+      "learning_rate": 9.755506116559971e-05,
+      "loss": 0.999,
+      "step": 1421
+    },
+    {
+      "epoch": 0.12705787745437488,
+      "grad_norm": 0.4029146730899811,
+      "learning_rate": 9.755058946815164e-05,
+      "loss": 1.0903,
+      "step": 1422
+    },
+    {
+      "epoch": 0.12714722898563674,
+      "grad_norm": 0.4076593518257141,
+      "learning_rate": 9.754611378783486e-05,
+      "loss": 1.1189,
+      "step": 1423
+    },
+    {
+      "epoch": 0.1272365805168986,
+      "grad_norm": 0.41739514470100403,
+      "learning_rate": 9.754163412502424e-05,
+      "loss": 1.0114,
+      "step": 1424
+    },
+    {
+      "epoch": 0.12732593204816048,
+      "grad_norm": 0.4175979197025299,
+      "learning_rate": 9.7537150480095e-05,
+      "loss": 1.0852,
+      "step": 1425
+    },
+    {
+      "epoch": 0.12741528357942233,
+      "grad_norm": 0.5235275030136108,
+      "learning_rate": 9.75326628534227e-05,
+      "loss": 1.0744,
+      "step": 1426
+    },
+    {
+      "epoch": 0.12750463511068422,
+      "grad_norm": 0.41696855425834656,
+      "learning_rate": 9.752817124538324e-05,
+      "loss": 1.0345,
+      "step": 1427
+    },
+    {
+      "epoch": 0.12759398664194607,
+      "grad_norm": 0.4305160641670227,
+      "learning_rate": 9.752367565635281e-05,
+      "loss": 1.0633,
+      "step": 1428
+    },
+    {
+      "epoch": 0.12768333817320796,
+      "grad_norm": 0.4274474084377289,
+      "learning_rate": 9.751917608670797e-05,
+      "loss": 1.0398,
+      "step": 1429
+    },
+    {
+      "epoch": 0.1277726897044698,
+      "grad_norm": 0.41223591566085815,
+      "learning_rate": 9.751467253682563e-05,
+      "loss": 1.0496,
+      "step": 1430
+    },
+    {
+      "epoch": 0.12786204123573167,
+      "grad_norm": 0.4205789566040039,
+      "learning_rate": 9.751016500708298e-05,
+      "loss": 0.9926,
+      "step": 1431
+    },
+    {
+      "epoch": 0.12795139276699355,
+      "grad_norm": 0.3847178816795349,
+      "learning_rate": 9.75056534978576e-05,
+      "loss": 1.0912,
+      "step": 1432
+    },
+    {
+      "epoch": 0.1280407442982554,
+      "grad_norm": 0.3665478527545929,
+      "learning_rate": 9.750113800952738e-05,
+      "loss": 1.0538,
+      "step": 1433
+    },
+    {
+      "epoch": 0.1281300958295173,
+      "grad_norm": 0.4411013126373291,
+      "learning_rate": 9.74966185424705e-05,
+      "loss": 1.0351,
+      "step": 1434
+    },
+    {
+      "epoch": 0.12821944736077914,
+      "grad_norm": 0.5092450380325317,
+      "learning_rate": 9.749209509706555e-05,
+      "loss": 1.0013,
+      "step": 1435
+    },
+    {
+      "epoch": 0.12830879889204103,
+      "grad_norm": 0.39857053756713867,
+      "learning_rate": 9.74875676736914e-05,
+      "loss": 1.0148,
+      "step": 1436
+    },
+    {
+      "epoch": 0.12839815042330288,
+      "grad_norm": 0.4596467912197113,
+      "learning_rate": 9.74830362727273e-05,
+      "loss": 1.0708,
+      "step": 1437
+    },
+    {
+      "epoch": 0.12848750195456474,
+      "grad_norm": 0.4768296182155609,
+      "learning_rate": 9.747850089455275e-05,
+      "loss": 1.0696,
+      "step": 1438
+    },
+    {
+      "epoch": 0.12857685348582662,
+      "grad_norm": 0.39800745248794556,
+      "learning_rate": 9.747396153954767e-05,
+      "loss": 1.0498,
+      "step": 1439
+    },
+    {
+      "epoch": 0.12866620501708848,
+      "grad_norm": 0.36220765113830566,
+      "learning_rate": 9.746941820809229e-05,
+      "loss": 1.0571,
+      "step": 1440
+    },
+    {
+      "epoch": 0.12875555654835036,
+      "grad_norm": 0.4107576012611389,
+      "learning_rate": 9.746487090056713e-05,
+      "loss": 1.0934,
+      "step": 1441
+    },
+    {
+      "epoch": 0.1288449080796122,
+      "grad_norm": 0.5009557604789734,
+      "learning_rate": 9.746031961735311e-05,
+      "loss": 1.021,
+      "step": 1442
+    },
+    {
+      "epoch": 0.12893425961087407,
+      "grad_norm": 0.43069615960121155,
+      "learning_rate": 9.745576435883142e-05,
+      "loss": 1.0339,
+      "step": 1443
+    },
+    {
+      "epoch": 0.12902361114213595,
+      "grad_norm": 0.515500009059906,
+      "learning_rate": 9.745120512538362e-05,
+      "loss": 0.9844,
+      "step": 1444
+    },
+    {
+      "epoch": 0.1291129626733978,
+      "grad_norm": 0.49051618576049805,
+      "learning_rate": 9.744664191739161e-05,
+      "loss": 1.0346,
+      "step": 1445
+    },
+    {
+      "epoch": 0.1292023142046597,
+      "grad_norm": 0.4764910042285919,
+      "learning_rate": 9.744207473523759e-05,
+      "loss": 1.0409,
+      "step": 1446
+    },
+    {
+      "epoch": 0.12929166573592155,
+      "grad_norm": 0.37674903869628906,
+      "learning_rate": 9.74375035793041e-05,
+      "loss": 1.0546,
+      "step": 1447
+    },
+    {
+      "epoch": 0.12938101726718343,
+      "grad_norm": 0.524577796459198,
+      "learning_rate": 9.743292844997407e-05,
+      "loss": 1.0141,
+      "step": 1448
+    },
+    {
+      "epoch": 0.12947036879844528,
+      "grad_norm": 0.43543311953544617,
+      "learning_rate": 9.742834934763066e-05,
+      "loss": 0.9927,
+      "step": 1449
+    },
+    {
+      "epoch": 0.12955972032970714,
+      "grad_norm": 0.4244259297847748,
+      "learning_rate": 9.742376627265745e-05,
+      "loss": 1.0312,
+      "step": 1450
+    },
+    {
+      "epoch": 0.12964907186096902,
+      "grad_norm": 0.41922134160995483,
+      "learning_rate": 9.74191792254383e-05,
+      "loss": 1.0319,
+      "step": 1451
+    },
+    {
+      "epoch": 0.12973842339223088,
+      "grad_norm": 0.38788026571273804,
+      "learning_rate": 9.741458820635745e-05,
+      "loss": 1.1334,
+      "step": 1452
+    },
+    {
+      "epoch": 0.12982777492349276,
+      "grad_norm": 0.42001378536224365,
+      "learning_rate": 9.740999321579943e-05,
+      "loss": 1.0905,
+      "step": 1453
+    },
+    {
+      "epoch": 0.12991712645475462,
+      "grad_norm": 0.41198793053627014,
+      "learning_rate": 9.740539425414912e-05,
+      "loss": 1.0529,
+      "step": 1454
+    },
+    {
+      "epoch": 0.13000647798601647,
+      "grad_norm": 0.36228737235069275,
+      "learning_rate": 9.740079132179175e-05,
+      "loss": 1.0103,
+      "step": 1455
+    },
+    {
+      "epoch": 0.13009582951727836,
+      "grad_norm": 0.48605382442474365,
+      "learning_rate": 9.739618441911285e-05,
+      "loss": 1.0234,
+      "step": 1456
+    },
+    {
+      "epoch": 0.1301851810485402,
+      "grad_norm": 0.49182093143463135,
+      "learning_rate": 9.739157354649829e-05,
+      "loss": 0.9932,
+      "step": 1457
+    },
+    {
+      "epoch": 0.1302745325798021,
+      "grad_norm": 0.3926076591014862,
+      "learning_rate": 9.738695870433428e-05,
+      "loss": 1.0766,
+      "step": 1458
+    },
+    {
+      "epoch": 0.13036388411106395,
+      "grad_norm": 0.5217881202697754,
+      "learning_rate": 9.738233989300739e-05,
+      "loss": 1.0239,
+      "step": 1459
+    },
+    {
+      "epoch": 0.13045323564232583,
+      "grad_norm": 0.4107765555381775,
+      "learning_rate": 9.737771711290447e-05,
+      "loss": 1.0517,
+      "step": 1460
+    },
+    {
+      "epoch": 0.1305425871735877,
+      "grad_norm": 0.48638463020324707,
+      "learning_rate": 9.737309036441271e-05,
+      "loss": 0.9408,
+      "step": 1461
+    },
+    {
+      "epoch": 0.13063193870484954,
+      "grad_norm": 0.4553143084049225,
+      "learning_rate": 9.736845964791968e-05,
+      "loss": 1.0842,
+      "step": 1462
+    },
+    {
+      "epoch": 0.13072129023611143,
+      "grad_norm": 0.42393583059310913,
+      "learning_rate": 9.736382496381325e-05,
+      "loss": 1.0693,
+      "step": 1463
+    },
+    {
+      "epoch": 0.13081064176737328,
+      "grad_norm": 0.43372994661331177,
+      "learning_rate": 9.735918631248162e-05,
+      "loss": 1.0923,
+      "step": 1464
+    },
+    {
+      "epoch": 0.13089999329863516,
+      "grad_norm": 0.42476969957351685,
+      "learning_rate": 9.735454369431332e-05,
+      "loss": 1.1099,
+      "step": 1465
+    },
+    {
+      "epoch": 0.13098934482989702,
+      "grad_norm": 0.4615803360939026,
+      "learning_rate": 9.734989710969722e-05,
+      "loss": 1.0546,
+      "step": 1466
+    },
+    {
+      "epoch": 0.1310786963611589,
+      "grad_norm": 0.3829108774662018,
+      "learning_rate": 9.734524655902253e-05,
+      "loss": 1.0674,
+      "step": 1467
+    },
+    {
+      "epoch": 0.13116804789242076,
+      "grad_norm": 0.4632830321788788,
+      "learning_rate": 9.734059204267878e-05,
+      "loss": 1.0149,
+      "step": 1468
+    },
+    {
+      "epoch": 0.1312573994236826,
+      "grad_norm": 0.4512673020362854,
+      "learning_rate": 9.733593356105581e-05,
+      "loss": 1.018,
+      "step": 1469
+    },
+    {
+      "epoch": 0.1313467509549445,
+      "grad_norm": 0.5254853963851929,
+      "learning_rate": 9.733127111454385e-05,
+      "loss": 0.9973,
+      "step": 1470
+    },
+    {
+      "epoch": 0.13143610248620635,
+      "grad_norm": 0.4638168215751648,
+      "learning_rate": 9.732660470353343e-05,
+      "loss": 1.0137,
+      "step": 1471
+    },
+    {
+      "epoch": 0.13152545401746824,
+      "grad_norm": 0.40837833285331726,
+      "learning_rate": 9.732193432841539e-05,
+      "loss": 1.1323,
+      "step": 1472
+    },
+    {
+      "epoch": 0.1316148055487301,
+      "grad_norm": 0.39538437128067017,
+      "learning_rate": 9.731725998958095e-05,
+      "loss": 1.0323,
+      "step": 1473
+    },
+    {
+      "epoch": 0.13170415707999195,
+      "grad_norm": 0.429143488407135,
+      "learning_rate": 9.73125816874216e-05,
+      "loss": 1.0897,
+      "step": 1474
+    },
+    {
+      "epoch": 0.13179350861125383,
+      "grad_norm": 0.48212477564811707,
+      "learning_rate": 9.730789942232923e-05,
+      "loss": 0.9875,
+      "step": 1475
+    },
+    {
+      "epoch": 0.13188286014251568,
+      "grad_norm": 0.37595105171203613,
+      "learning_rate": 9.730321319469601e-05,
+      "loss": 1.0909,
+      "step": 1476
+    },
+    {
+      "epoch": 0.13197221167377757,
+      "grad_norm": 0.4069534242153168,
+      "learning_rate": 9.729852300491447e-05,
+      "loss": 1.0292,
+      "step": 1477
+    },
+    {
+      "epoch": 0.13206156320503942,
+      "grad_norm": 0.4816598892211914,
+      "learning_rate": 9.729382885337747e-05,
+      "loss": 1.0329,
+      "step": 1478
+    },
+    {
+      "epoch": 0.1321509147363013,
+      "grad_norm": 0.4182417392730713,
+      "learning_rate": 9.728913074047819e-05,
+      "loss": 1.0831,
+      "step": 1479
+    },
+    {
+      "epoch": 0.13224026626756316,
+      "grad_norm": 0.4294276535511017,
+      "learning_rate": 9.728442866661013e-05,
+      "loss": 1.0979,
+      "step": 1480
+    },
+    {
+      "epoch": 0.13232961779882502,
+      "grad_norm": 0.4855419099330902,
+      "learning_rate": 9.727972263216716e-05,
+      "loss": 0.994,
+      "step": 1481
+    },
+    {
+      "epoch": 0.1324189693300869,
+      "grad_norm": 0.3743133544921875,
+      "learning_rate": 9.727501263754346e-05,
+      "loss": 1.1104,
+      "step": 1482
+    },
+    {
+      "epoch": 0.13250832086134876,
+      "grad_norm": 0.47284814715385437,
+      "learning_rate": 9.727029868313352e-05,
+      "loss": 1.0346,
+      "step": 1483
+    },
+    {
+      "epoch": 0.13259767239261064,
+      "grad_norm": 0.4814388155937195,
+      "learning_rate": 9.726558076933221e-05,
+      "loss": 1.0394,
+      "step": 1484
+    },
+    {
+      "epoch": 0.1326870239238725,
+      "grad_norm": 0.48564615845680237,
+      "learning_rate": 9.726085889653469e-05,
+      "loss": 1.0232,
+      "step": 1485
+    },
+    {
+      "epoch": 0.13277637545513435,
+      "grad_norm": 0.4509584605693817,
+      "learning_rate": 9.725613306513648e-05,
+      "loss": 1.0218,
+      "step": 1486
+    },
+    {
+      "epoch": 0.13286572698639623,
+      "grad_norm": 0.45788347721099854,
+      "learning_rate": 9.725140327553342e-05,
+      "loss": 1.0633,
+      "step": 1487
+    },
+    {
+      "epoch": 0.1329550785176581,
+      "grad_norm": 0.5591961741447449,
+      "learning_rate": 9.724666952812166e-05,
+      "loss": 1.0504,
+      "step": 1488
+    },
+    {
+      "epoch": 0.13304443004891997,
+      "grad_norm": 0.49788787961006165,
+      "learning_rate": 9.724193182329772e-05,
+      "loss": 0.9796,
+      "step": 1489
+    },
+    {
+      "epoch": 0.13313378158018183,
+      "grad_norm": 0.5562415719032288,
+      "learning_rate": 9.723719016145843e-05,
+      "loss": 0.9633,
+      "step": 1490
+    },
+    {
+      "epoch": 0.1332231331114437,
+      "grad_norm": 0.49074721336364746,
+      "learning_rate": 9.723244454300093e-05,
+      "loss": 1.0881,
+      "step": 1491
+    },
+    {
+      "epoch": 0.13331248464270556,
+      "grad_norm": 0.42609742283821106,
+      "learning_rate": 9.722769496832275e-05,
+      "loss": 1.0958,
+      "step": 1492
+    },
+    {
+      "epoch": 0.13340183617396742,
+      "grad_norm": 0.5812770128250122,
+      "learning_rate": 9.722294143782171e-05,
+      "loss": 1.1263,
+      "step": 1493
+    },
+    {
+      "epoch": 0.1334911877052293,
+      "grad_norm": 0.4071921408176422,
+      "learning_rate": 9.721818395189597e-05,
+      "loss": 1.0965,
+      "step": 1494
+    },
+    {
+      "epoch": 0.13358053923649116,
+      "grad_norm": 0.4325934648513794,
+      "learning_rate": 9.7213422510944e-05,
+      "loss": 1.0716,
+      "step": 1495
+    },
+    {
+      "epoch": 0.13366989076775304,
+      "grad_norm": 0.4646300673484802,
+      "learning_rate": 9.720865711536464e-05,
+      "loss": 1.0785,
+      "step": 1496
+    },
+    {
+      "epoch": 0.1337592422990149,
+      "grad_norm": 0.4847221076488495,
+      "learning_rate": 9.720388776555704e-05,
+      "loss": 1.0767,
+      "step": 1497
+    },
+    {
+      "epoch": 0.13384859383027678,
+      "grad_norm": 0.3830418884754181,
+      "learning_rate": 9.71991144619207e-05,
+      "loss": 1.0585,
+      "step": 1498
+    },
+    {
+      "epoch": 0.13393794536153864,
+      "grad_norm": 0.4215502142906189,
+      "learning_rate": 9.719433720485539e-05,
+      "loss": 1.0535,
+      "step": 1499
+    },
+    {
+      "epoch": 0.1340272968928005,
+      "grad_norm": 0.4030102789402008,
+      "learning_rate": 9.718955599476129e-05,
+      "loss": 1.0467,
+      "step": 1500
+    },
+    {
+      "epoch": 0.13411664842406237,
+      "grad_norm": 0.4001905024051666,
+      "learning_rate": 9.718477083203887e-05,
+      "loss": 1.0354,
+      "step": 1501
+    },
+    {
+      "epoch": 0.13420599995532423,
+      "grad_norm": 0.6067075133323669,
+      "learning_rate": 9.717998171708895e-05,
+      "loss": 0.9813,
+      "step": 1502
+    },
+    {
+      "epoch": 0.1342953514865861,
+      "grad_norm": 0.4722452163696289,
+      "learning_rate": 9.717518865031266e-05,
+      "loss": 1.0531,
+      "step": 1503
+    },
+    {
+      "epoch": 0.13438470301784797,
+      "grad_norm": 0.45017173886299133,
+      "learning_rate": 9.717039163211146e-05,
+      "loss": 0.9219,
+      "step": 1504
+    },
+    {
+      "epoch": 0.13447405454910982,
+      "grad_norm": 0.5280361771583557,
+      "learning_rate": 9.716559066288715e-05,
+      "loss": 0.9697,
+      "step": 1505
+    },
+    {
+      "epoch": 0.1345634060803717,
+      "grad_norm": 0.4430614709854126,
+      "learning_rate": 9.716078574304189e-05,
+      "loss": 1.0229,
+      "step": 1506
+    },
+    {
+      "epoch": 0.13465275761163356,
+      "grad_norm": 0.4364091157913208,
+      "learning_rate": 9.715597687297813e-05,
+      "loss": 1.0449,
+      "step": 1507
+    },
+    {
+      "epoch": 0.13474210914289544,
+      "grad_norm": 0.43157270550727844,
+      "learning_rate": 9.715116405309865e-05,
+      "loss": 1.0936,
+      "step": 1508
+    },
+    {
+      "epoch": 0.1348314606741573,
+      "grad_norm": 0.439143568277359,
+      "learning_rate": 9.714634728380658e-05,
+      "loss": 1.0145,
+      "step": 1509
+    },
+    {
+      "epoch": 0.13492081220541918,
+      "grad_norm": 0.3858429789543152,
+      "learning_rate": 9.714152656550539e-05,
+      "loss": 1.1219,
+      "step": 1510
+    },
+    {
+      "epoch": 0.13501016373668104,
+      "grad_norm": 0.4329967200756073,
+      "learning_rate": 9.713670189859887e-05,
+      "loss": 0.9739,
+      "step": 1511
+    },
+    {
+      "epoch": 0.1350995152679429,
+      "grad_norm": 0.4508189260959625,
+      "learning_rate": 9.713187328349111e-05,
+      "loss": 1.0605,
+      "step": 1512
+    },
+    {
+      "epoch": 0.13518886679920478,
+      "grad_norm": 0.4645385444164276,
+      "learning_rate": 9.712704072058656e-05,
+      "loss": 0.9873,
+      "step": 1513
+    },
+    {
+      "epoch": 0.13527821833046663,
+      "grad_norm": 0.5293644666671753,
+      "learning_rate": 9.712220421029003e-05,
+      "loss": 1.0814,
+      "step": 1514
+    },
+    {
+      "epoch": 0.13536756986172852,
+      "grad_norm": 0.4389038681983948,
+      "learning_rate": 9.711736375300661e-05,
+      "loss": 1.0325,
+      "step": 1515
+    },
+    {
+      "epoch": 0.13545692139299037,
+      "grad_norm": 0.349090039730072,
+      "learning_rate": 9.711251934914174e-05,
+      "loss": 1.1654,
+      "step": 1516
+    },
+    {
+      "epoch": 0.13554627292425223,
+      "grad_norm": 0.5453407168388367,
+      "learning_rate": 9.710767099910119e-05,
+      "loss": 1.0776,
+      "step": 1517
+    },
+    {
+      "epoch": 0.1356356244555141,
+      "grad_norm": 0.4334562420845032,
+      "learning_rate": 9.710281870329105e-05,
+      "loss": 1.0226,
+      "step": 1518
+    },
+    {
+      "epoch": 0.13572497598677596,
+      "grad_norm": 0.3693782687187195,
+      "learning_rate": 9.709796246211777e-05,
+      "loss": 1.0437,
+      "step": 1519
+    },
+    {
+      "epoch": 0.13581432751803785,
+      "grad_norm": 0.49413368105888367,
+      "learning_rate": 9.709310227598811e-05,
+      "loss": 1.0567,
+      "step": 1520
+    },
+    {
+      "epoch": 0.1359036790492997,
+      "grad_norm": 0.4487200081348419,
+      "learning_rate": 9.708823814530917e-05,
+      "loss": 0.981,
+      "step": 1521
+    },
+    {
+      "epoch": 0.13599303058056159,
+      "grad_norm": 0.38888999819755554,
+      "learning_rate": 9.708337007048834e-05,
+      "loss": 1.0563,
+      "step": 1522
+    },
+    {
+      "epoch": 0.13608238211182344,
+      "grad_norm": 0.3906208276748657,
+      "learning_rate": 9.70784980519334e-05,
+      "loss": 1.0467,
+      "step": 1523
+    },
+    {
+      "epoch": 0.1361717336430853,
+      "grad_norm": 0.3793841302394867,
+      "learning_rate": 9.707362209005244e-05,
+      "loss": 1.0433,
+      "step": 1524
+    },
+    {
+      "epoch": 0.13626108517434718,
+      "grad_norm": 0.38040420413017273,
+      "learning_rate": 9.706874218525385e-05,
+      "loss": 1.0493,
+      "step": 1525
+    },
+    {
+      "epoch": 0.13635043670560903,
+      "grad_norm": 0.5357340574264526,
+      "learning_rate": 9.706385833794638e-05,
+      "loss": 0.9978,
+      "step": 1526
+    },
+    {
+      "epoch": 0.13643978823687092,
+      "grad_norm": 0.45937198400497437,
+      "learning_rate": 9.705897054853912e-05,
+      "loss": 1.0814,
+      "step": 1527
+    },
+    {
+      "epoch": 0.13652913976813277,
+      "grad_norm": 0.41274699568748474,
+      "learning_rate": 9.705407881744146e-05,
+      "loss": 1.0937,
+      "step": 1528
+    },
+    {
+      "epoch": 0.13661849129939466,
+      "grad_norm": 0.4878060519695282,
+      "learning_rate": 9.704918314506313e-05,
+      "loss": 1.0107,
+      "step": 1529
+    },
+    {
+      "epoch": 0.1367078428306565,
+      "grad_norm": 0.4097362756729126,
+      "learning_rate": 9.704428353181421e-05,
+      "loss": 1.1267,
+      "step": 1530
+    },
+    {
+      "epoch": 0.13679719436191837,
+      "grad_norm": 0.40500837564468384,
+      "learning_rate": 9.703937997810511e-05,
+      "loss": 1.0243,
+      "step": 1531
+    },
+    {
+      "epoch": 0.13688654589318025,
+      "grad_norm": 0.4004482626914978,
+      "learning_rate": 9.70344724843465e-05,
+      "loss": 1.0469,
+      "step": 1532
+    },
+    {
+      "epoch": 0.1369758974244421,
+      "grad_norm": 0.5236450433731079,
+      "learning_rate": 9.702956105094948e-05,
+      "loss": 0.9977,
+      "step": 1533
+    },
+    {
+      "epoch": 0.137065248955704,
+      "grad_norm": 0.4368230998516083,
+      "learning_rate": 9.702464567832543e-05,
+      "loss": 1.0354,
+      "step": 1534
+    },
+    {
+      "epoch": 0.13715460048696584,
+      "grad_norm": 0.4011850357055664,
+      "learning_rate": 9.701972636688606e-05,
+      "loss": 1.0555,
+      "step": 1535
+    },
+    {
+      "epoch": 0.1372439520182277,
+      "grad_norm": 0.4206826984882355,
+      "learning_rate": 9.701480311704339e-05,
+      "loss": 1.007,
+      "step": 1536
+    },
+    {
+      "epoch": 0.13733330354948958,
+      "grad_norm": 0.45997175574302673,
+      "learning_rate": 9.700987592920983e-05,
+      "loss": 1.0538,
+      "step": 1537
+    },
+    {
+      "epoch": 0.13742265508075144,
+      "grad_norm": 0.4319978356361389,
+      "learning_rate": 9.700494480379807e-05,
+      "loss": 0.9962,
+      "step": 1538
+    },
+    {
+      "epoch": 0.13751200661201332,
+      "grad_norm": 0.4265231490135193,
+      "learning_rate": 9.700000974122115e-05,
+      "loss": 0.9786,
+      "step": 1539
+    },
+    {
+      "epoch": 0.13760135814327518,
+      "grad_norm": 0.4962487518787384,
+      "learning_rate": 9.699507074189242e-05,
+      "loss": 1.0389,
+      "step": 1540
+    },
+    {
+      "epoch": 0.13769070967453706,
+      "grad_norm": 0.40575921535491943,
+      "learning_rate": 9.699012780622561e-05,
+      "loss": 1.0018,
+      "step": 1541
+    },
+    {
+      "epoch": 0.13778006120579891,
+      "grad_norm": 0.4042010009288788,
+      "learning_rate": 9.698518093463469e-05,
+      "loss": 1.0876,
+      "step": 1542
+    },
+    {
+      "epoch": 0.13786941273706077,
+      "grad_norm": 0.37715786695480347,
+      "learning_rate": 9.698023012753405e-05,
+      "loss": 1.0111,
+      "step": 1543
+    },
+    {
+      "epoch": 0.13795876426832265,
+      "grad_norm": 0.4954730272293091,
+      "learning_rate": 9.697527538533837e-05,
+      "loss": 0.9983,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1380481157995845,
+      "grad_norm": 0.41885560750961304,
+      "learning_rate": 9.697031670846265e-05,
+      "loss": 1.042,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1381374673308464,
+      "grad_norm": 0.3857840895652771,
+      "learning_rate": 9.696535409732224e-05,
+      "loss": 1.0516,
+      "step": 1546
+    },
+    {
+      "epoch": 0.13822681886210825,
+      "grad_norm": 0.3953034281730652,
+      "learning_rate": 9.696038755233282e-05,
+      "loss": 1.1183,
+      "step": 1547
+    },
+    {
+      "epoch": 0.1383161703933701,
+      "grad_norm": 0.4239901006221771,
+      "learning_rate": 9.695541707391036e-05,
+      "loss": 1.0304,
+      "step": 1548
+    },
+    {
+      "epoch": 0.13840552192463199,
+      "grad_norm": 0.4670591652393341,
+      "learning_rate": 9.695044266247122e-05,
+      "loss": 1.0292,
+      "step": 1549
+    },
+    {
+      "epoch": 0.13849487345589384,
+      "grad_norm": 0.40943849086761475,
+      "learning_rate": 9.694546431843205e-05,
+      "loss": 1.0806,
+      "step": 1550
+    },
+    {
+      "epoch": 0.13858422498715572,
+      "grad_norm": 0.4126596748828888,
+      "learning_rate": 9.694048204220985e-05,
+      "loss": 1.0379,
+      "step": 1551
+    },
+    {
+      "epoch": 0.13867357651841758,
+      "grad_norm": 0.4415280222892761,
+      "learning_rate": 9.693549583422191e-05,
+      "loss": 1.0358,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13876292804967946,
+      "grad_norm": 0.4215603768825531,
+      "learning_rate": 9.69305056948859e-05,
+      "loss": 1.0874,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13885227958094132,
+      "grad_norm": 0.42005017399787903,
+      "learning_rate": 9.692551162461981e-05,
+      "loss": 1.0552,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13894163111220317,
+      "grad_norm": 0.4029698669910431,
+      "learning_rate": 9.692051362384193e-05,
+      "loss": 1.0081,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13903098264346506,
+      "grad_norm": 0.5039706826210022,
+      "learning_rate": 9.691551169297089e-05,
+      "loss": 1.0135,
+      "step": 1556
+    },
+    {
+      "epoch": 0.1391203341747269,
+      "grad_norm": 0.48726481199264526,
+      "learning_rate": 9.691050583242567e-05,
+      "loss": 1.011,
+      "step": 1557
+    },
+    {
+      "epoch": 0.1392096857059888,
+      "grad_norm": 0.4722976088523865,
+      "learning_rate": 9.690549604262555e-05,
+      "loss": 0.9555,
+      "step": 1558
+    },
+    {
+      "epoch": 0.13929903723725065,
+      "grad_norm": 0.44578781723976135,
+      "learning_rate": 9.690048232399017e-05,
+      "loss": 1.0367,
+      "step": 1559
+    },
+    {
+      "epoch": 0.13938838876851253,
+      "grad_norm": 0.47405532002449036,
+      "learning_rate": 9.689546467693946e-05,
+      "loss": 1.0977,
+      "step": 1560
+    },
+    {
+      "epoch": 0.1394777402997744,
+      "grad_norm": 0.4186593294143677,
+      "learning_rate": 9.689044310189371e-05,
+      "loss": 1.0139,
+      "step": 1561
+    },
+    {
+      "epoch": 0.13956709183103624,
+      "grad_norm": 0.3761426508426666,
+      "learning_rate": 9.688541759927354e-05,
+      "loss": 1.0328,
+      "step": 1562
+    },
+    {
+      "epoch": 0.13965644336229813,
+      "grad_norm": 0.4144037365913391,
+      "learning_rate": 9.688038816949989e-05,
+      "loss": 1.0543,
+      "step": 1563
+    },
+    {
+      "epoch": 0.13974579489355998,
+      "grad_norm": 0.5622879862785339,
+      "learning_rate": 9.687535481299402e-05,
+      "loss": 0.9816,
+      "step": 1564
+    },
+    {
+      "epoch": 0.13983514642482187,
+      "grad_norm": 0.39679154753685,
+      "learning_rate": 9.687031753017753e-05,
+      "loss": 1.075,
+      "step": 1565
+    },
+    {
+      "epoch": 0.13992449795608372,
+      "grad_norm": 0.48569121956825256,
+      "learning_rate": 9.686527632147234e-05,
+      "loss": 0.9407,
+      "step": 1566
+    },
+    {
+      "epoch": 0.14001384948734558,
+      "grad_norm": 0.43438124656677246,
+      "learning_rate": 9.68602311873007e-05,
+      "loss": 1.0061,
+      "step": 1567
+    },
+    {
+      "epoch": 0.14010320101860746,
+      "grad_norm": 0.4626932144165039,
+      "learning_rate": 9.685518212808522e-05,
+      "loss": 0.9583,
+      "step": 1568
+    },
+    {
+      "epoch": 0.14019255254986931,
+      "grad_norm": 0.5100006461143494,
+      "learning_rate": 9.685012914424878e-05,
+      "loss": 0.9871,
+      "step": 1569
+    },
+    {
+      "epoch": 0.1402819040811312,
+      "grad_norm": 0.4893263280391693,
+      "learning_rate": 9.684507223621465e-05,
+      "loss": 0.9678,
+      "step": 1570
+    },
+    {
+      "epoch": 0.14037125561239305,
+      "grad_norm": 0.40761256217956543,
+      "learning_rate": 9.684001140440639e-05,
+      "loss": 1.0929,
+      "step": 1571
+    },
+    {
+      "epoch": 0.14046060714365494,
+      "grad_norm": 0.4098682403564453,
+      "learning_rate": 9.68349466492479e-05,
+      "loss": 1.1117,
+      "step": 1572
+    },
+    {
+      "epoch": 0.1405499586749168,
+      "grad_norm": 0.36809906363487244,
+      "learning_rate": 9.682987797116339e-05,
+      "loss": 1.0983,
+      "step": 1573
+    },
+    {
+      "epoch": 0.14063931020617865,
+      "grad_norm": 0.4545081853866577,
+      "learning_rate": 9.682480537057743e-05,
+      "loss": 1.0238,
+      "step": 1574
+    },
+    {
+      "epoch": 0.14072866173744053,
+      "grad_norm": 0.46472567319869995,
+      "learning_rate": 9.681972884791492e-05,
+      "loss": 1.0105,
+      "step": 1575
+    },
+    {
+      "epoch": 0.14081801326870239,
+      "grad_norm": 0.4712905287742615,
+      "learning_rate": 9.681464840360103e-05,
+      "loss": 0.9822,
+      "step": 1576
+    },
+    {
+      "epoch": 0.14090736479996427,
+      "grad_norm": 0.4782291352748871,
+      "learning_rate": 9.680956403806135e-05,
+      "loss": 1.0426,
+      "step": 1577
+    },
+    {
+      "epoch": 0.14099671633122612,
+      "grad_norm": 0.4774461090564728,
+      "learning_rate": 9.680447575172173e-05,
+      "loss": 1.0958,
+      "step": 1578
+    },
+    {
+      "epoch": 0.141086067862488,
+      "grad_norm": 0.41189044713974,
+      "learning_rate": 9.679938354500835e-05,
+      "loss": 1.0193,
+      "step": 1579
+    },
+    {
+      "epoch": 0.14117541939374986,
+      "grad_norm": 0.45908036828041077,
+      "learning_rate": 9.679428741834776e-05,
+      "loss": 1.0789,
+      "step": 1580
+    },
+    {
+      "epoch": 0.14126477092501172,
+      "grad_norm": 0.439585417509079,
+      "learning_rate": 9.67891873721668e-05,
+      "loss": 1.047,
+      "step": 1581
+    },
+    {
+      "epoch": 0.1413541224562736,
+      "grad_norm": 0.3884674906730652,
+      "learning_rate": 9.678408340689267e-05,
+      "loss": 1.0571,
+      "step": 1582
+    },
+    {
+      "epoch": 0.14144347398753546,
+      "grad_norm": 0.46724075078964233,
+      "learning_rate": 9.677897552295288e-05,
+      "loss": 1.0396,
+      "step": 1583
+    },
+    {
+      "epoch": 0.14153282551879734,
+      "grad_norm": 0.4285435378551483,
+      "learning_rate": 9.677386372077524e-05,
+      "loss": 0.9729,
+      "step": 1584
+    },
+    {
+      "epoch": 0.1416221770500592,
+      "grad_norm": 0.41435936093330383,
+      "learning_rate": 9.676874800078796e-05,
+      "loss": 1.0059,
+      "step": 1585
+    },
+    {
+      "epoch": 0.14171152858132105,
+      "grad_norm": 0.4586467146873474,
+      "learning_rate": 9.67636283634195e-05,
+      "loss": 1.0339,
+      "step": 1586
+    },
+    {
+      "epoch": 0.14180088011258293,
+      "grad_norm": 0.5122067928314209,
+      "learning_rate": 9.675850480909872e-05,
+      "loss": 1.001,
+      "step": 1587
+    },
+    {
+      "epoch": 0.1418902316438448,
+      "grad_norm": 0.42933905124664307,
+      "learning_rate": 9.675337733825473e-05,
+      "loss": 1.0386,
+      "step": 1588
+    },
+    {
+      "epoch": 0.14197958317510667,
+      "grad_norm": 0.4264965355396271,
+      "learning_rate": 9.674824595131704e-05,
+      "loss": 1.0794,
+      "step": 1589
+    },
+    {
+      "epoch": 0.14206893470636853,
+      "grad_norm": 0.4125423729419708,
+      "learning_rate": 9.674311064871547e-05,
+      "loss": 1.1203,
+      "step": 1590
+    },
+    {
+      "epoch": 0.1421582862376304,
+      "grad_norm": 0.3605044484138489,
+      "learning_rate": 9.67379714308801e-05,
+      "loss": 1.0368,
+      "step": 1591
+    },
+    {
+      "epoch": 0.14224763776889227,
+      "grad_norm": 0.45730555057525635,
+      "learning_rate": 9.673282829824146e-05,
+      "loss": 1.0428,
+      "step": 1592
+    },
+    {
+      "epoch": 0.14233698930015412,
+      "grad_norm": 0.45879489183425903,
+      "learning_rate": 9.672768125123031e-05,
+      "loss": 1.055,
+      "step": 1593
+    },
+    {
+      "epoch": 0.142426340831416,
+      "grad_norm": 0.4261813461780548,
+      "learning_rate": 9.672253029027777e-05,
+      "loss": 1.0486,
+      "step": 1594
+    },
+    {
+      "epoch": 0.14251569236267786,
+      "grad_norm": 0.3854468762874603,
+      "learning_rate": 9.671737541581529e-05,
+      "loss": 1.0092,
+      "step": 1595
+    },
+    {
+      "epoch": 0.14260504389393974,
+      "grad_norm": 0.4169517457485199,
+      "learning_rate": 9.671221662827465e-05,
+      "loss": 1.063,
+      "step": 1596
+    },
+    {
+      "epoch": 0.1426943954252016,
+      "grad_norm": 0.41241368651390076,
+      "learning_rate": 9.670705392808796e-05,
+      "loss": 1.0649,
+      "step": 1597
+    },
+    {
+      "epoch": 0.14278374695646345,
+      "grad_norm": 0.3985016644001007,
+      "learning_rate": 9.670188731568764e-05,
+      "loss": 1.0126,
+      "step": 1598
+    },
+    {
+      "epoch": 0.14287309848772534,
+      "grad_norm": 0.395868182182312,
+      "learning_rate": 9.669671679150642e-05,
+      "loss": 1.0801,
+      "step": 1599
+    },
+    {
+      "epoch": 0.1429624500189872,
+      "grad_norm": 0.3894529938697815,
+      "learning_rate": 9.669154235597746e-05,
+      "loss": 1.0315,
+      "step": 1600
+    },
+    {
+      "epoch": 0.14305180155024907,
+      "grad_norm": 0.4293975234031677,
+      "learning_rate": 9.66863640095341e-05,
+      "loss": 1.0163,
+      "step": 1601
+    },
+    {
+      "epoch": 0.14314115308151093,
+      "grad_norm": 0.3886623978614807,
+      "learning_rate": 9.668118175261015e-05,
+      "loss": 1.0487,
+      "step": 1602
+    },
+    {
+      "epoch": 0.1432305046127728,
+      "grad_norm": 0.46701672673225403,
+      "learning_rate": 9.66759955856396e-05,
+      "loss": 1.0463,
+      "step": 1603
+    },
+    {
+      "epoch": 0.14331985614403467,
+      "grad_norm": 0.4126945436000824,
+      "learning_rate": 9.667080550905691e-05,
+      "loss": 1.0268,
+      "step": 1604
+    },
+    {
+      "epoch": 0.14340920767529652,
+      "grad_norm": 0.4913039207458496,
+      "learning_rate": 9.666561152329679e-05,
+      "loss": 0.9639,
+      "step": 1605
+    },
+    {
+      "epoch": 0.1434985592065584,
+      "grad_norm": 0.3920893967151642,
+      "learning_rate": 9.666041362879427e-05,
+      "loss": 1.0301,
+      "step": 1606
+    },
+    {
+      "epoch": 0.14358791073782026,
+      "grad_norm": 0.38557082414627075,
+      "learning_rate": 9.665521182598476e-05,
+      "loss": 1.0137,
+      "step": 1607
+    },
+    {
+      "epoch": 0.14367726226908215,
+      "grad_norm": 0.532264769077301,
+      "learning_rate": 9.665000611530392e-05,
+      "loss": 0.9705,
+      "step": 1608
+    },
+    {
+      "epoch": 0.143766613800344,
+      "grad_norm": 0.42515647411346436,
+      "learning_rate": 9.664479649718784e-05,
+      "loss": 1.0133,
+      "step": 1609
+    },
+    {
+      "epoch": 0.14385596533160588,
+      "grad_norm": 0.42825859785079956,
+      "learning_rate": 9.663958297207286e-05,
+      "loss": 1.0272,
+      "step": 1610
+    },
+    {
+      "epoch": 0.14394531686286774,
+      "grad_norm": 0.3822021782398224,
+      "learning_rate": 9.663436554039567e-05,
+      "loss": 1.0168,
+      "step": 1611
+    },
+    {
+      "epoch": 0.1440346683941296,
+      "grad_norm": 0.43384209275245667,
+      "learning_rate": 9.662914420259325e-05,
+      "loss": 1.031,
+      "step": 1612
+    },
+    {
+      "epoch": 0.14412401992539148,
+      "grad_norm": 0.4819833040237427,
+      "learning_rate": 9.662391895910299e-05,
+      "loss": 1.011,
+      "step": 1613
+    },
+    {
+      "epoch": 0.14421337145665333,
+      "grad_norm": 0.38827648758888245,
+      "learning_rate": 9.661868981036255e-05,
+      "loss": 1.0561,
+      "step": 1614
+    },
+    {
+      "epoch": 0.14430272298791522,
+      "grad_norm": 0.5991657376289368,
+      "learning_rate": 9.661345675680991e-05,
+      "loss": 0.9748,
+      "step": 1615
+    },
+    {
+      "epoch": 0.14439207451917707,
+      "grad_norm": 0.568602442741394,
+      "learning_rate": 9.660821979888339e-05,
+      "loss": 1.0853,
+      "step": 1616
+    },
+    {
+      "epoch": 0.14448142605043893,
+      "grad_norm": 0.4651549458503723,
+      "learning_rate": 9.660297893702168e-05,
+      "loss": 1.0078,
+      "step": 1617
+    },
+    {
+      "epoch": 0.1445707775817008,
+      "grad_norm": 0.426548570394516,
+      "learning_rate": 9.659773417166371e-05,
+      "loss": 0.9844,
+      "step": 1618
+    },
+    {
+      "epoch": 0.14466012911296267,
+      "grad_norm": 0.4951413571834564,
+      "learning_rate": 9.659248550324882e-05,
+      "loss": 0.8695,
+      "step": 1619
+    },
+    {
+      "epoch": 0.14474948064422455,
+      "grad_norm": 0.4711915850639343,
+      "learning_rate": 9.658723293221661e-05,
+      "loss": 1.03,
+      "step": 1620
+    },
+    {
+      "epoch": 0.1448388321754864,
+      "grad_norm": 0.38477903604507446,
+      "learning_rate": 9.658197645900708e-05,
+      "loss": 1.0262,
+      "step": 1621
+    },
+    {
+      "epoch": 0.1449281837067483,
+      "grad_norm": 0.5240241885185242,
+      "learning_rate": 9.657671608406047e-05,
+      "loss": 0.9587,
+      "step": 1622
+    },
+    {
+      "epoch": 0.14501753523801014,
+      "grad_norm": 0.4152171015739441,
+      "learning_rate": 9.657145180781743e-05,
+      "loss": 1.0534,
+      "step": 1623
+    },
+    {
+      "epoch": 0.145106886769272,
+      "grad_norm": 0.43236833810806274,
+      "learning_rate": 9.65661836307189e-05,
+      "loss": 1.0896,
+      "step": 1624
+    },
+    {
+      "epoch": 0.14519623830053388,
+      "grad_norm": 0.5443985462188721,
+      "learning_rate": 9.656091155320611e-05,
+      "loss": 1.0644,
+      "step": 1625
+    },
+    {
+      "epoch": 0.14528558983179574,
+      "grad_norm": 0.4063294529914856,
+      "learning_rate": 9.655563557572069e-05,
+      "loss": 1.0385,
+      "step": 1626
+    },
+    {
+      "epoch": 0.14537494136305762,
+      "grad_norm": 0.5186159610748291,
+      "learning_rate": 9.655035569870452e-05,
+      "loss": 1.0842,
+      "step": 1627
+    },
+    {
+      "epoch": 0.14546429289431947,
+      "grad_norm": 0.4194406569004059,
+      "learning_rate": 9.654507192259989e-05,
+      "loss": 1.0013,
+      "step": 1628
+    },
+    {
+      "epoch": 0.14555364442558133,
+      "grad_norm": 0.4991320073604584,
+      "learning_rate": 9.653978424784934e-05,
+      "loss": 1.096,
+      "step": 1629
+    },
+    {
+      "epoch": 0.1456429959568432,
+      "grad_norm": 0.4811127185821533,
+      "learning_rate": 9.653449267489579e-05,
+      "loss": 0.9865,
+      "step": 1630
+    },
+    {
+      "epoch": 0.14573234748810507,
+      "grad_norm": 0.4587719440460205,
+      "learning_rate": 9.652919720418245e-05,
+      "loss": 1.0453,
+      "step": 1631
+    },
+    {
+      "epoch": 0.14582169901936695,
+      "grad_norm": 0.40233114361763,
+      "learning_rate": 9.652389783615287e-05,
+      "loss": 1.1234,
+      "step": 1632
+    },
+    {
+      "epoch": 0.1459110505506288,
+      "grad_norm": 0.4691876471042633,
+      "learning_rate": 9.651859457125096e-05,
+      "loss": 0.9824,
+      "step": 1633
+    },
+    {
+      "epoch": 0.1460004020818907,
+      "grad_norm": 0.41825005412101746,
+      "learning_rate": 9.651328740992088e-05,
+      "loss": 1.0688,
+      "step": 1634
+    },
+    {
+      "epoch": 0.14608975361315255,
+      "grad_norm": 0.4713682532310486,
+      "learning_rate": 9.65079763526072e-05,
+      "loss": 1.0345,
+      "step": 1635
+    },
+    {
+      "epoch": 0.1461791051444144,
+      "grad_norm": 0.41411858797073364,
+      "learning_rate": 9.650266139975474e-05,
+      "loss": 1.0163,
+      "step": 1636
+    },
+    {
+      "epoch": 0.14626845667567628,
+      "grad_norm": 0.3815430700778961,
+      "learning_rate": 9.649734255180873e-05,
+      "loss": 1.0626,
+      "step": 1637
+    },
+    {
+      "epoch": 0.14635780820693814,
+      "grad_norm": 0.41789570450782776,
+      "learning_rate": 9.649201980921465e-05,
+      "loss": 0.9784,
+      "step": 1638
+    },
+    {
+      "epoch": 0.14644715973820002,
+      "grad_norm": 0.43766146898269653,
+      "learning_rate": 9.648669317241833e-05,
+      "loss": 1.0253,
+      "step": 1639
+    },
+    {
+      "epoch": 0.14653651126946188,
+      "grad_norm": 0.43024829030036926,
+      "learning_rate": 9.648136264186595e-05,
+      "loss": 1.0291,
+      "step": 1640
+    },
+    {
+      "epoch": 0.14662586280072376,
+      "grad_norm": 0.39453184604644775,
+      "learning_rate": 9.647602821800399e-05,
+      "loss": 1.0444,
+      "step": 1641
+    },
+    {
+      "epoch": 0.14671521433198562,
+      "grad_norm": 0.44370511174201965,
+      "learning_rate": 9.647068990127928e-05,
+      "loss": 1.0723,
+      "step": 1642
+    },
+    {
+      "epoch": 0.14680456586324747,
+      "grad_norm": 0.46351152658462524,
+      "learning_rate": 9.646534769213893e-05,
+      "loss": 1.0513,
+      "step": 1643
+    },
+    {
+      "epoch": 0.14689391739450935,
+      "grad_norm": 0.4535273313522339,
+      "learning_rate": 9.646000159103043e-05,
+      "loss": 1.0497,
+      "step": 1644
+    },
+    {
+      "epoch": 0.1469832689257712,
+      "grad_norm": 0.43159547448158264,
+      "learning_rate": 9.645465159840157e-05,
+      "loss": 1.0305,
+      "step": 1645
+    },
+    {
+      "epoch": 0.1470726204570331,
+      "grad_norm": 0.4147747755050659,
+      "learning_rate": 9.644929771470048e-05,
+      "loss": 1.0223,
+      "step": 1646
+    },
+    {
+      "epoch": 0.14716197198829495,
+      "grad_norm": 0.45313653349876404,
+      "learning_rate": 9.64439399403756e-05,
+      "loss": 0.9908,
+      "step": 1647
+    },
+    {
+      "epoch": 0.1472513235195568,
+      "grad_norm": 0.3770694434642792,
+      "learning_rate": 9.643857827587566e-05,
+      "loss": 1.0579,
+      "step": 1648
+    },
+    {
+      "epoch": 0.1473406750508187,
+      "grad_norm": 0.3648872971534729,
+      "learning_rate": 9.643321272164981e-05,
+      "loss": 1.0841,
+      "step": 1649
+    },
+    {
+      "epoch": 0.14743002658208054,
+      "grad_norm": 0.48402151465415955,
+      "learning_rate": 9.642784327814744e-05,
+      "loss": 1.0109,
+      "step": 1650
+    },
+    {
+      "epoch": 0.14751937811334243,
+      "grad_norm": 0.42030149698257446,
+      "learning_rate": 9.642246994581832e-05,
+      "loss": 1.0677,
+      "step": 1651
+    },
+    {
+      "epoch": 0.14760872964460428,
+      "grad_norm": 0.46909406781196594,
+      "learning_rate": 9.641709272511252e-05,
+      "loss": 0.9945,
+      "step": 1652
+    },
+    {
+      "epoch": 0.14769808117586616,
+      "grad_norm": 0.4196245074272156,
+      "learning_rate": 9.641171161648043e-05,
+      "loss": 1.1265,
+      "step": 1653
+    },
+    {
+      "epoch": 0.14778743270712802,
+      "grad_norm": 0.3893241286277771,
+      "learning_rate": 9.640632662037278e-05,
+      "loss": 1.0552,
+      "step": 1654
+    },
+    {
+      "epoch": 0.14787678423838987,
+      "grad_norm": 0.44918498396873474,
+      "learning_rate": 9.640093773724061e-05,
+      "loss": 1.0326,
+      "step": 1655
+    },
+    {
+      "epoch": 0.14796613576965176,
+      "grad_norm": 0.5244982242584229,
+      "learning_rate": 9.63955449675353e-05,
+      "loss": 1.0296,
+      "step": 1656
+    },
+    {
+      "epoch": 0.1480554873009136,
+      "grad_norm": 0.4323536157608032,
+      "learning_rate": 9.639014831170858e-05,
+      "loss": 1.0664,
+      "step": 1657
+    },
+    {
+      "epoch": 0.1481448388321755,
+      "grad_norm": 0.437716007232666,
+      "learning_rate": 9.638474777021244e-05,
+      "loss": 1.0745,
+      "step": 1658
+    },
+    {
+      "epoch": 0.14823419036343735,
+      "grad_norm": 0.39507436752319336,
+      "learning_rate": 9.637934334349927e-05,
+      "loss": 1.0435,
+      "step": 1659
+    },
+    {
+      "epoch": 0.1483235418946992,
+      "grad_norm": 0.422969251871109,
+      "learning_rate": 9.63739350320217e-05,
+      "loss": 1.0611,
+      "step": 1660
+    },
+    {
+      "epoch": 0.1484128934259611,
+      "grad_norm": 0.3984132409095764,
+      "learning_rate": 9.636852283623276e-05,
+      "loss": 1.0717,
+      "step": 1661
+    },
+    {
+      "epoch": 0.14850224495722295,
+      "grad_norm": 0.4262081980705261,
+      "learning_rate": 9.63631067565858e-05,
+      "loss": 1.0278,
+      "step": 1662
+    },
+    {
+      "epoch": 0.14859159648848483,
+      "grad_norm": 0.43201524019241333,
+      "learning_rate": 9.635768679353445e-05,
+      "loss": 0.9729,
+      "step": 1663
+    },
+    {
+      "epoch": 0.14868094801974668,
+      "grad_norm": 0.4402812719345093,
+      "learning_rate": 9.635226294753267e-05,
+      "loss": 1.0292,
+      "step": 1664
+    },
+    {
+      "epoch": 0.14877029955100857,
+      "grad_norm": 0.41984882950782776,
+      "learning_rate": 9.634683521903483e-05,
+      "loss": 1.036,
+      "step": 1665
+    },
+    {
+      "epoch": 0.14885965108227042,
+      "grad_norm": 0.38695448637008667,
+      "learning_rate": 9.634140360849548e-05,
+      "loss": 1.0454,
+      "step": 1666
+    },
+    {
+      "epoch": 0.14894900261353228,
+      "grad_norm": 0.44035449624061584,
+      "learning_rate": 9.633596811636964e-05,
+      "loss": 1.0825,
+      "step": 1667
+    },
+    {
+      "epoch": 0.14903835414479416,
+      "grad_norm": 0.4655371904373169,
+      "learning_rate": 9.633052874311255e-05,
+      "loss": 1.0182,
+      "step": 1668
+    },
+    {
+      "epoch": 0.14912770567605602,
+      "grad_norm": 0.42064300179481506,
+      "learning_rate": 9.632508548917983e-05,
+      "loss": 1.0113,
+      "step": 1669
+    },
+    {
+      "epoch": 0.1492170572073179,
+      "grad_norm": 0.4009134769439697,
+      "learning_rate": 9.631963835502742e-05,
+      "loss": 1.0587,
+      "step": 1670
+    },
+    {
+      "epoch": 0.14930640873857975,
+      "grad_norm": 0.3674651086330414,
+      "learning_rate": 9.631418734111155e-05,
+      "loss": 1.0356,
+      "step": 1671
+    },
+    {
+      "epoch": 0.14939576026984164,
+      "grad_norm": 0.416415810585022,
+      "learning_rate": 9.630873244788883e-05,
+      "loss": 1.0074,
+      "step": 1672
+    },
+    {
+      "epoch": 0.1494851118011035,
+      "grad_norm": 0.5505238771438599,
+      "learning_rate": 9.630327367581616e-05,
+      "loss": 1.0533,
+      "step": 1673
+    },
+    {
+      "epoch": 0.14957446333236535,
+      "grad_norm": 0.42817139625549316,
+      "learning_rate": 9.629781102535074e-05,
+      "loss": 0.9811,
+      "step": 1674
+    },
+    {
+      "epoch": 0.14966381486362723,
+      "grad_norm": 0.4695313274860382,
+      "learning_rate": 9.629234449695015e-05,
+      "loss": 1.1321,
+      "step": 1675
+    },
+    {
+      "epoch": 0.1497531663948891,
+      "grad_norm": 0.5143523216247559,
+      "learning_rate": 9.628687409107229e-05,
+      "loss": 0.9853,
+      "step": 1676
+    },
+    {
+      "epoch": 0.14984251792615097,
+      "grad_norm": 0.4100930392742157,
+      "learning_rate": 9.628139980817532e-05,
+      "loss": 1.1377,
+      "step": 1677
+    },
+    {
+      "epoch": 0.14993186945741283,
+      "grad_norm": 0.41417646408081055,
+      "learning_rate": 9.627592164871779e-05,
+      "loss": 0.9765,
+      "step": 1678
+    },
+    {
+      "epoch": 0.15002122098867468,
+      "grad_norm": 0.3846787214279175,
+      "learning_rate": 9.627043961315856e-05,
+      "loss": 1.0585,
+      "step": 1679
+    },
+    {
+      "epoch": 0.15011057251993656,
+      "grad_norm": 0.4256773889064789,
+      "learning_rate": 9.626495370195683e-05,
+      "loss": 1.0301,
+      "step": 1680
+    },
+    {
+      "epoch": 0.15019992405119842,
+      "grad_norm": 0.5811467170715332,
+      "learning_rate": 9.625946391557204e-05,
+      "loss": 1.0538,
+      "step": 1681
+    },
+    {
+      "epoch": 0.1502892755824603,
+      "grad_norm": 0.4346529543399811,
+      "learning_rate": 9.625397025446408e-05,
+      "loss": 1.0657,
+      "step": 1682
+    },
+    {
+      "epoch": 0.15037862711372216,
+      "grad_norm": 0.3570214509963989,
+      "learning_rate": 9.624847271909308e-05,
+      "loss": 1.0752,
+      "step": 1683
+    },
+    {
+      "epoch": 0.15046797864498404,
+      "grad_norm": 0.4259663224220276,
+      "learning_rate": 9.62429713099195e-05,
+      "loss": 1.1056,
+      "step": 1684
+    },
+    {
+      "epoch": 0.1505573301762459,
+      "grad_norm": 0.3912825882434845,
+      "learning_rate": 9.623746602740417e-05,
+      "loss": 1.0072,
+      "step": 1685
+    },
+    {
+      "epoch": 0.15064668170750775,
+      "grad_norm": 0.41152167320251465,
+      "learning_rate": 9.623195687200822e-05,
+      "loss": 1.0224,
+      "step": 1686
+    },
+    {
+      "epoch": 0.15073603323876963,
+      "grad_norm": 0.397203266620636,
+      "learning_rate": 9.622644384419306e-05,
+      "loss": 1.0702,
+      "step": 1687
+    },
+    {
+      "epoch": 0.1508253847700315,
+      "grad_norm": 0.4435591399669647,
+      "learning_rate": 9.62209269444205e-05,
+      "loss": 0.9521,
+      "step": 1688
+    },
+    {
+      "epoch": 0.15091473630129337,
+      "grad_norm": 0.36750173568725586,
+      "learning_rate": 9.621540617315262e-05,
+      "loss": 0.9843,
+      "step": 1689
+    },
+    {
+      "epoch": 0.15100408783255523,
+      "grad_norm": 0.4787936210632324,
+      "learning_rate": 9.620988153085187e-05,
+      "loss": 1.0932,
+      "step": 1690
+    },
+    {
+      "epoch": 0.15109343936381708,
+      "grad_norm": 0.44318270683288574,
+      "learning_rate": 9.620435301798097e-05,
+      "loss": 1.0223,
+      "step": 1691
+    },
+    {
+      "epoch": 0.15118279089507897,
+      "grad_norm": 0.47096431255340576,
+      "learning_rate": 9.619882063500299e-05,
+      "loss": 1.0623,
+      "step": 1692
+    },
+    {
+      "epoch": 0.15127214242634082,
+      "grad_norm": 0.44428104162216187,
+      "learning_rate": 9.619328438238135e-05,
+      "loss": 1.0413,
+      "step": 1693
+    },
+    {
+      "epoch": 0.1513614939576027,
+      "grad_norm": 0.4424484670162201,
+      "learning_rate": 9.618774426057975e-05,
+      "loss": 1.0759,
+      "step": 1694
+    },
+    {
+      "epoch": 0.15145084548886456,
+      "grad_norm": 0.3779531717300415,
+      "learning_rate": 9.618220027006225e-05,
+      "loss": 1.0476,
+      "step": 1695
+    },
+    {
+      "epoch": 0.15154019702012644,
+      "grad_norm": 0.40432384610176086,
+      "learning_rate": 9.617665241129321e-05,
+      "loss": 1.1215,
+      "step": 1696
+    },
+    {
+      "epoch": 0.1516295485513883,
+      "grad_norm": 0.48599207401275635,
+      "learning_rate": 9.617110068473732e-05,
+      "loss": 1.0024,
+      "step": 1697
+    },
+    {
+      "epoch": 0.15171890008265015,
+      "grad_norm": 0.49614617228507996,
+      "learning_rate": 9.61655450908596e-05,
+      "loss": 0.9352,
+      "step": 1698
+    },
+    {
+      "epoch": 0.15180825161391204,
+      "grad_norm": 0.5036540627479553,
+      "learning_rate": 9.615998563012538e-05,
+      "loss": 0.9844,
+      "step": 1699
+    },
+    {
+      "epoch": 0.1518976031451739,
+      "grad_norm": 0.3981391191482544,
+      "learning_rate": 9.615442230300036e-05,
+      "loss": 1.0222,
+      "step": 1700
+    },
+    {
+      "epoch": 0.15198695467643578,
+      "grad_norm": 0.4211319386959076,
+      "learning_rate": 9.614885510995047e-05,
+      "loss": 1.0769,
+      "step": 1701
+    },
+    {
+      "epoch": 0.15207630620769763,
+      "grad_norm": 0.4759863317012787,
+      "learning_rate": 9.614328405144207e-05,
+      "loss": 0.999,
+      "step": 1702
+    },
+    {
+      "epoch": 0.15216565773895951,
+      "grad_norm": 0.48188042640686035,
+      "learning_rate": 9.613770912794178e-05,
+      "loss": 0.9254,
+      "step": 1703
+    },
+    {
+      "epoch": 0.15225500927022137,
+      "grad_norm": 0.38263580203056335,
+      "learning_rate": 9.613213033991655e-05,
+      "loss": 1.0517,
+      "step": 1704
+    },
+    {
+      "epoch": 0.15234436080148323,
+      "grad_norm": 0.44071927666664124,
+      "learning_rate": 9.612654768783368e-05,
+      "loss": 1.0019,
+      "step": 1705
+    },
+    {
+      "epoch": 0.1524337123327451,
+      "grad_norm": 0.44804859161376953,
+      "learning_rate": 9.612096117216076e-05,
+      "loss": 1.062,
+      "step": 1706
+    },
+    {
+      "epoch": 0.15252306386400696,
+      "grad_norm": 0.4194650948047638,
+      "learning_rate": 9.611537079336574e-05,
+      "loss": 1.0101,
+      "step": 1707
+    },
+    {
+      "epoch": 0.15261241539526885,
+      "grad_norm": 0.4241783618927002,
+      "learning_rate": 9.610977655191684e-05,
+      "loss": 1.0197,
+      "step": 1708
+    },
+    {
+      "epoch": 0.1527017669265307,
+      "grad_norm": 0.48272955417633057,
+      "learning_rate": 9.610417844828268e-05,
+      "loss": 1.014,
+      "step": 1709
+    },
+    {
+      "epoch": 0.15279111845779256,
+      "grad_norm": 0.4096718728542328,
+      "learning_rate": 9.609857648293212e-05,
+      "loss": 1.0106,
+      "step": 1710
+    },
+    {
+      "epoch": 0.15288046998905444,
+      "grad_norm": 0.3911011815071106,
+      "learning_rate": 9.609297065633443e-05,
+      "loss": 1.0542,
+      "step": 1711
+    },
+    {
+      "epoch": 0.1529698215203163,
+      "grad_norm": 0.43495795130729675,
+      "learning_rate": 9.60873609689591e-05,
+      "loss": 1.0039,
+      "step": 1712
+    },
+    {
+      "epoch": 0.15305917305157818,
+      "grad_norm": 0.4525046646595001,
+      "learning_rate": 9.608174742127605e-05,
+      "loss": 0.9883,
+      "step": 1713
+    },
+    {
+      "epoch": 0.15314852458284003,
+      "grad_norm": 0.3769172430038452,
+      "learning_rate": 9.607613001375546e-05,
+      "loss": 1.1064,
+      "step": 1714
+    },
+    {
+      "epoch": 0.15323787611410192,
+      "grad_norm": 0.435320645570755,
+      "learning_rate": 9.607050874686784e-05,
+      "loss": 1.0457,
+      "step": 1715
+    },
+    {
+      "epoch": 0.15332722764536377,
+      "grad_norm": 0.38679543137550354,
+      "learning_rate": 9.606488362108404e-05,
+      "loss": 1.0426,
+      "step": 1716
+    },
+    {
+      "epoch": 0.15341657917662563,
+      "grad_norm": 0.446005254983902,
+      "learning_rate": 9.605925463687522e-05,
+      "loss": 1.041,
+      "step": 1717
+    },
+    {
+      "epoch": 0.1535059307078875,
+      "grad_norm": 0.4369790554046631,
+      "learning_rate": 9.605362179471287e-05,
+      "loss": 0.9566,
+      "step": 1718
+    },
+    {
+      "epoch": 0.15359528223914937,
+      "grad_norm": 0.3627423942089081,
+      "learning_rate": 9.604798509506879e-05,
+      "loss": 1.0247,
+      "step": 1719
+    },
+    {
+      "epoch": 0.15368463377041125,
+      "grad_norm": 0.40787839889526367,
+      "learning_rate": 9.604234453841512e-05,
+      "loss": 1.1035,
+      "step": 1720
+    },
+    {
+      "epoch": 0.1537739853016731,
+      "grad_norm": 0.4434973895549774,
+      "learning_rate": 9.603670012522432e-05,
+      "loss": 1.0767,
+      "step": 1721
+    },
+    {
+      "epoch": 0.153863336832935,
+      "grad_norm": 0.4459492564201355,
+      "learning_rate": 9.603105185596917e-05,
+      "loss": 0.9853,
+      "step": 1722
+    },
+    {
+      "epoch": 0.15395268836419684,
+      "grad_norm": 0.45601245760917664,
+      "learning_rate": 9.602539973112278e-05,
+      "loss": 1.0228,
+      "step": 1723
+    },
+    {
+      "epoch": 0.1540420398954587,
+      "grad_norm": 0.44149455428123474,
+      "learning_rate": 9.601974375115856e-05,
+      "loss": 1.0443,
+      "step": 1724
+    },
+    {
+      "epoch": 0.15413139142672058,
+      "grad_norm": 0.5166652798652649,
+      "learning_rate": 9.601408391655027e-05,
+      "loss": 0.9864,
+      "step": 1725
+    },
+    {
+      "epoch": 0.15422074295798244,
+      "grad_norm": 0.5009913444519043,
+      "learning_rate": 9.600842022777198e-05,
+      "loss": 1.0516,
+      "step": 1726
+    },
+    {
+      "epoch": 0.15431009448924432,
+      "grad_norm": 0.4483872652053833,
+      "learning_rate": 9.600275268529807e-05,
+      "loss": 0.9622,
+      "step": 1727
+    },
+    {
+      "epoch": 0.15439944602050618,
+      "grad_norm": 0.4618902802467346,
+      "learning_rate": 9.599708128960327e-05,
+      "loss": 0.9304,
+      "step": 1728
+    },
+    {
+      "epoch": 0.15448879755176803,
+      "grad_norm": 0.39204496145248413,
+      "learning_rate": 9.599140604116263e-05,
+      "loss": 1.0674,
+      "step": 1729
+    },
+    {
+      "epoch": 0.15457814908302991,
+      "grad_norm": 0.44558286666870117,
+      "learning_rate": 9.598572694045149e-05,
+      "loss": 1.0133,
+      "step": 1730
+    },
+    {
+      "epoch": 0.15466750061429177,
+      "grad_norm": 0.39136767387390137,
+      "learning_rate": 9.598004398794556e-05,
+      "loss": 0.9855,
+      "step": 1731
+    },
+    {
+      "epoch": 0.15475685214555365,
+      "grad_norm": 0.47076520323753357,
+      "learning_rate": 9.597435718412082e-05,
+      "loss": 1.0118,
+      "step": 1732
+    },
+    {
+      "epoch": 0.1548462036768155,
+      "grad_norm": 0.373670369386673,
+      "learning_rate": 9.596866652945362e-05,
+      "loss": 1.0864,
+      "step": 1733
+    },
+    {
+      "epoch": 0.1549355552080774,
+      "grad_norm": 0.3965423107147217,
+      "learning_rate": 9.59629720244206e-05,
+      "loss": 0.9902,
+      "step": 1734
+    },
+    {
+      "epoch": 0.15502490673933925,
+      "grad_norm": 0.4660866856575012,
+      "learning_rate": 9.595727366949875e-05,
+      "loss": 1.0477,
+      "step": 1735
+    },
+    {
+      "epoch": 0.1551142582706011,
+      "grad_norm": 0.4512030780315399,
+      "learning_rate": 9.595157146516535e-05,
+      "loss": 1.058,
+      "step": 1736
+    },
+    {
+      "epoch": 0.15520360980186299,
+      "grad_norm": 0.4368734359741211,
+      "learning_rate": 9.594586541189804e-05,
+      "loss": 1.0307,
+      "step": 1737
+    },
+    {
+      "epoch": 0.15529296133312484,
+      "grad_norm": 0.5563750863075256,
+      "learning_rate": 9.594015551017477e-05,
+      "loss": 0.9361,
+      "step": 1738
+    },
+    {
+      "epoch": 0.15538231286438672,
+      "grad_norm": 0.43998295068740845,
+      "learning_rate": 9.593444176047378e-05,
+      "loss": 1.041,
+      "step": 1739
+    },
+    {
+      "epoch": 0.15547166439564858,
+      "grad_norm": 0.6449422836303711,
+      "learning_rate": 9.592872416327365e-05,
+      "loss": 0.9818,
+      "step": 1740
+    },
+    {
+      "epoch": 0.15556101592691043,
+      "grad_norm": 0.40925419330596924,
+      "learning_rate": 9.592300271905332e-05,
+      "loss": 1.0854,
+      "step": 1741
+    },
+    {
+      "epoch": 0.15565036745817232,
+      "grad_norm": 0.3952295780181885,
+      "learning_rate": 9.591727742829199e-05,
+      "loss": 1.1511,
+      "step": 1742
+    },
+    {
+      "epoch": 0.15573971898943417,
+      "grad_norm": 0.49510762095451355,
+      "learning_rate": 9.591154829146927e-05,
+      "loss": 0.9822,
+      "step": 1743
+    },
+    {
+      "epoch": 0.15582907052069606,
+      "grad_norm": 0.43776780366897583,
+      "learning_rate": 9.590581530906497e-05,
+      "loss": 1.0781,
+      "step": 1744
+    },
+    {
+      "epoch": 0.1559184220519579,
+      "grad_norm": 0.40279847383499146,
+      "learning_rate": 9.590007848155932e-05,
+      "loss": 1.0741,
+      "step": 1745
+    },
+    {
+      "epoch": 0.1560077735832198,
+      "grad_norm": 0.36434587836265564,
+      "learning_rate": 9.589433780943284e-05,
+      "loss": 1.0873,
+      "step": 1746
+    },
+    {
+      "epoch": 0.15609712511448165,
+      "grad_norm": 0.3859378695487976,
+      "learning_rate": 9.588859329316637e-05,
+      "loss": 1.0719,
+      "step": 1747
+    },
+    {
+      "epoch": 0.1561864766457435,
+      "grad_norm": 0.34931498765945435,
+      "learning_rate": 9.588284493324106e-05,
+      "loss": 1.0645,
+      "step": 1748
+    },
+    {
+      "epoch": 0.1562758281770054,
+      "grad_norm": 0.3696591556072235,
+      "learning_rate": 9.587709273013845e-05,
+      "loss": 1.0927,
+      "step": 1749
+    },
+    {
+      "epoch": 0.15636517970826724,
+      "grad_norm": 0.4064505100250244,
+      "learning_rate": 9.587133668434027e-05,
+      "loss": 1.0693,
+      "step": 1750
+    },
+    {
+      "epoch": 0.15645453123952913,
+      "grad_norm": 0.5394856333732605,
+      "learning_rate": 9.586557679632871e-05,
+      "loss": 1.0407,
+      "step": 1751
+    },
+    {
+      "epoch": 0.15654388277079098,
+      "grad_norm": 0.47162312269210815,
+      "learning_rate": 9.58598130665862e-05,
+      "loss": 1.0118,
+      "step": 1752
+    },
+    {
+      "epoch": 0.15663323430205287,
+      "grad_norm": 0.46519139409065247,
+      "learning_rate": 9.585404549559551e-05,
+      "loss": 1.0815,
+      "step": 1753
+    },
+    {
+      "epoch": 0.15672258583331472,
+      "grad_norm": 0.4289345443248749,
+      "learning_rate": 9.584827408383974e-05,
+      "loss": 1.039,
+      "step": 1754
+    },
+    {
+      "epoch": 0.15681193736457658,
+      "grad_norm": 0.4443482756614685,
+      "learning_rate": 9.58424988318023e-05,
+      "loss": 1.0424,
+      "step": 1755
+    },
+    {
+      "epoch": 0.15690128889583846,
+      "grad_norm": 0.4442220628261566,
+      "learning_rate": 9.583671973996695e-05,
+      "loss": 1.0615,
+      "step": 1756
+    },
+    {
+      "epoch": 0.15699064042710031,
+      "grad_norm": 0.4155905544757843,
+      "learning_rate": 9.583093680881774e-05,
+      "loss": 1.0328,
+      "step": 1757
+    },
+    {
+      "epoch": 0.1570799919583622,
+      "grad_norm": 0.40333032608032227,
+      "learning_rate": 9.582515003883904e-05,
+      "loss": 1.0237,
+      "step": 1758
+    },
+    {
+      "epoch": 0.15716934348962405,
+      "grad_norm": 0.4274482727050781,
+      "learning_rate": 9.581935943051557e-05,
+      "loss": 1.0246,
+      "step": 1759
+    },
+    {
+      "epoch": 0.1572586950208859,
+      "grad_norm": 0.47601258754730225,
+      "learning_rate": 9.581356498433233e-05,
+      "loss": 1.0917,
+      "step": 1760
+    },
+    {
+      "epoch": 0.1573480465521478,
+      "grad_norm": 0.37168294191360474,
+      "learning_rate": 9.580776670077471e-05,
+      "loss": 1.0442,
+      "step": 1761
+    },
+    {
+      "epoch": 0.15743739808340965,
+      "grad_norm": 0.48139792680740356,
+      "learning_rate": 9.580196458032836e-05,
+      "loss": 1.1097,
+      "step": 1762
+    },
+    {
+      "epoch": 0.15752674961467153,
+      "grad_norm": 0.49613094329833984,
+      "learning_rate": 9.579615862347924e-05,
+      "loss": 1.0116,
+      "step": 1763
+    },
+    {
+      "epoch": 0.15761610114593338,
+      "grad_norm": 0.3607613742351532,
+      "learning_rate": 9.579034883071368e-05,
+      "loss": 1.0563,
+      "step": 1764
+    },
+    {
+      "epoch": 0.15770545267719527,
+      "grad_norm": 0.5845707654953003,
+      "learning_rate": 9.578453520251833e-05,
+      "loss": 1.0129,
+      "step": 1765
+    },
+    {
+      "epoch": 0.15779480420845712,
+      "grad_norm": 0.4461520314216614,
+      "learning_rate": 9.577871773938011e-05,
+      "loss": 0.99,
+      "step": 1766
+    },
+    {
+      "epoch": 0.15788415573971898,
+      "grad_norm": 0.4064408242702484,
+      "learning_rate": 9.577289644178634e-05,
+      "loss": 1.0397,
+      "step": 1767
+    },
+    {
+      "epoch": 0.15797350727098086,
+      "grad_norm": 0.475175678730011,
+      "learning_rate": 9.576707131022458e-05,
+      "loss": 0.9823,
+      "step": 1768
+    },
+    {
+      "epoch": 0.15806285880224272,
+      "grad_norm": 0.5856608152389526,
+      "learning_rate": 9.576124234518275e-05,
+      "loss": 0.9467,
+      "step": 1769
+    },
+    {
+      "epoch": 0.1581522103335046,
+      "grad_norm": 0.44236186146736145,
+      "learning_rate": 9.57554095471491e-05,
+      "loss": 0.9745,
+      "step": 1770
+    },
+    {
+      "epoch": 0.15824156186476646,
+      "grad_norm": 0.38273531198501587,
+      "learning_rate": 9.574957291661218e-05,
+      "loss": 1.0545,
+      "step": 1771
+    },
+    {
+      "epoch": 0.1583309133960283,
+      "grad_norm": 0.37142929434776306,
+      "learning_rate": 9.574373245406086e-05,
+      "loss": 0.9557,
+      "step": 1772
+    },
+    {
+      "epoch": 0.1584202649272902,
+      "grad_norm": 0.4939393997192383,
+      "learning_rate": 9.573788815998437e-05,
+      "loss": 1.027,
+      "step": 1773
+    },
+    {
+      "epoch": 0.15850961645855205,
+      "grad_norm": 0.5472790598869324,
+      "learning_rate": 9.573204003487221e-05,
+      "loss": 0.987,
+      "step": 1774
+    },
+    {
+      "epoch": 0.15859896798981393,
+      "grad_norm": 0.44142383337020874,
+      "learning_rate": 9.572618807921423e-05,
+      "loss": 1.0675,
+      "step": 1775
+    },
+    {
+      "epoch": 0.1586883195210758,
+      "grad_norm": 0.4367615878582001,
+      "learning_rate": 9.572033229350059e-05,
+      "loss": 1.0437,
+      "step": 1776
+    },
+    {
+      "epoch": 0.15877767105233767,
+      "grad_norm": 0.38631755113601685,
+      "learning_rate": 9.571447267822178e-05,
+      "loss": 1.0157,
+      "step": 1777
+    },
+    {
+      "epoch": 0.15886702258359953,
+      "grad_norm": 0.5570868849754333,
+      "learning_rate": 9.57086092338686e-05,
+      "loss": 0.9977,
+      "step": 1778
+    },
+    {
+      "epoch": 0.15895637411486138,
+      "grad_norm": 0.4448052942752838,
+      "learning_rate": 9.570274196093217e-05,
+      "loss": 1.0383,
+      "step": 1779
+    },
+    {
+      "epoch": 0.15904572564612326,
+      "grad_norm": 0.46690064668655396,
+      "learning_rate": 9.569687085990395e-05,
+      "loss": 1.0864,
+      "step": 1780
+    },
+    {
+      "epoch": 0.15913507717738512,
+      "grad_norm": 0.411737322807312,
+      "learning_rate": 9.569099593127571e-05,
+      "loss": 0.9839,
+      "step": 1781
+    },
+    {
+      "epoch": 0.159224428708647,
+      "grad_norm": 0.4243115484714508,
+      "learning_rate": 9.568511717553952e-05,
+      "loss": 0.9728,
+      "step": 1782
+    },
+    {
+      "epoch": 0.15931378023990886,
+      "grad_norm": 0.36506015062332153,
+      "learning_rate": 9.567923459318782e-05,
+      "loss": 1.0646,
+      "step": 1783
+    },
+    {
+      "epoch": 0.15940313177117074,
+      "grad_norm": 0.47241315245628357,
+      "learning_rate": 9.56733481847133e-05,
+      "loss": 1.0306,
+      "step": 1784
+    },
+    {
+      "epoch": 0.1594924833024326,
+      "grad_norm": 0.4867679178714752,
+      "learning_rate": 9.566745795060901e-05,
+      "loss": 0.9727,
+      "step": 1785
+    },
+    {
+      "epoch": 0.15958183483369445,
+      "grad_norm": 0.47150614857673645,
+      "learning_rate": 9.566156389136835e-05,
+      "loss": 1.0212,
+      "step": 1786
+    },
+    {
+      "epoch": 0.15967118636495634,
+      "grad_norm": 0.4018579423427582,
+      "learning_rate": 9.5655666007485e-05,
+      "loss": 1.0142,
+      "step": 1787
+    },
+    {
+      "epoch": 0.1597605378962182,
+      "grad_norm": 0.4174135625362396,
+      "learning_rate": 9.564976429945298e-05,
+      "loss": 1.0768,
+      "step": 1788
+    },
+    {
+      "epoch": 0.15984988942748007,
+      "grad_norm": 0.5170420408248901,
+      "learning_rate": 9.56438587677666e-05,
+      "loss": 0.9997,
+      "step": 1789
+    },
+    {
+      "epoch": 0.15993924095874193,
+      "grad_norm": 0.4216405153274536,
+      "learning_rate": 9.563794941292051e-05,
+      "loss": 1.0355,
+      "step": 1790
+    },
+    {
+      "epoch": 0.16002859249000378,
+      "grad_norm": 0.431983083486557,
+      "learning_rate": 9.563203623540969e-05,
+      "loss": 1.0086,
+      "step": 1791
+    },
+    {
+      "epoch": 0.16011794402126567,
+      "grad_norm": 0.45916739106178284,
+      "learning_rate": 9.562611923572944e-05,
+      "loss": 1.0688,
+      "step": 1792
+    },
+    {
+      "epoch": 0.16020729555252752,
+      "grad_norm": 0.45291900634765625,
+      "learning_rate": 9.562019841437537e-05,
+      "loss": 1.057,
+      "step": 1793
+    },
+    {
+      "epoch": 0.1602966470837894,
+      "grad_norm": 0.4525638520717621,
+      "learning_rate": 9.56142737718434e-05,
+      "loss": 1.1029,
+      "step": 1794
+    },
+    {
+      "epoch": 0.16038599861505126,
+      "grad_norm": 0.4891865849494934,
+      "learning_rate": 9.560834530862979e-05,
+      "loss": 1.0798,
+      "step": 1795
+    },
+    {
+      "epoch": 0.16047535014631314,
+      "grad_norm": 0.3669840097427368,
+      "learning_rate": 9.560241302523109e-05,
+      "loss": 1.0062,
+      "step": 1796
+    },
+    {
+      "epoch": 0.160564701677575,
+      "grad_norm": 0.4095466732978821,
+      "learning_rate": 9.559647692214424e-05,
+      "loss": 0.9993,
+      "step": 1797
+    },
+    {
+      "epoch": 0.16065405320883686,
+      "grad_norm": 0.5019549131393433,
+      "learning_rate": 9.559053699986642e-05,
+      "loss": 1.0768,
+      "step": 1798
+    },
+    {
+      "epoch": 0.16074340474009874,
+      "grad_norm": 0.40123990178108215,
+      "learning_rate": 9.558459325889515e-05,
+      "loss": 1.0714,
+      "step": 1799
+    },
+    {
+      "epoch": 0.1608327562713606,
+      "grad_norm": 0.4176787734031677,
+      "learning_rate": 9.557864569972832e-05,
+      "loss": 1.0862,
+      "step": 1800
+    },
+    {
+      "epoch": 0.16092210780262248,
+      "grad_norm": 0.4730657935142517,
+      "learning_rate": 9.557269432286407e-05,
+      "loss": 1.0686,
+      "step": 1801
+    },
+    {
+      "epoch": 0.16101145933388433,
+      "grad_norm": 0.4306131899356842,
+      "learning_rate": 9.55667391288009e-05,
+      "loss": 0.9852,
+      "step": 1802
+    },
+    {
+      "epoch": 0.1611008108651462,
+      "grad_norm": 0.40398168563842773,
+      "learning_rate": 9.556078011803762e-05,
+      "loss": 1.003,
+      "step": 1803
+    },
+    {
+      "epoch": 0.16119016239640807,
+      "grad_norm": 0.3656451106071472,
+      "learning_rate": 9.555481729107336e-05,
+      "loss": 1.0516,
+      "step": 1804
+    },
+    {
+      "epoch": 0.16127951392766993,
+      "grad_norm": 0.5433897376060486,
+      "learning_rate": 9.554885064840758e-05,
+      "loss": 0.9463,
+      "step": 1805
+    },
+    {
+      "epoch": 0.1613688654589318,
+      "grad_norm": 0.3712410628795624,
+      "learning_rate": 9.554288019054003e-05,
+      "loss": 1.0896,
+      "step": 1806
+    },
+    {
+      "epoch": 0.16145821699019366,
+      "grad_norm": 0.4026222825050354,
+      "learning_rate": 9.553690591797082e-05,
+      "loss": 1.1181,
+      "step": 1807
+    },
+    {
+      "epoch": 0.16154756852145555,
+      "grad_norm": 0.45911315083503723,
+      "learning_rate": 9.553092783120034e-05,
+      "loss": 0.9981,
+      "step": 1808
+    },
+    {
+      "epoch": 0.1616369200527174,
+      "grad_norm": 0.44033879041671753,
+      "learning_rate": 9.552494593072935e-05,
+      "loss": 1.0534,
+      "step": 1809
+    },
+    {
+      "epoch": 0.16172627158397926,
+      "grad_norm": 0.42678460478782654,
+      "learning_rate": 9.551896021705886e-05,
+      "loss": 1.0123,
+      "step": 1810
+    },
+    {
+      "epoch": 0.16181562311524114,
+      "grad_norm": 0.48469212651252747,
+      "learning_rate": 9.551297069069027e-05,
+      "loss": 0.9846,
+      "step": 1811
+    },
+    {
+      "epoch": 0.161904974646503,
+      "grad_norm": 0.4134121537208557,
+      "learning_rate": 9.550697735212523e-05,
+      "loss": 1.1542,
+      "step": 1812
+    },
+    {
+      "epoch": 0.16199432617776488,
+      "grad_norm": 0.44429683685302734,
+      "learning_rate": 9.550098020186579e-05,
+      "loss": 1.0126,
+      "step": 1813
+    },
+    {
+      "epoch": 0.16208367770902674,
+      "grad_norm": 0.4219866096973419,
+      "learning_rate": 9.549497924041424e-05,
+      "loss": 1.0309,
+      "step": 1814
+    },
+    {
+      "epoch": 0.16217302924028862,
+      "grad_norm": 0.38047802448272705,
+      "learning_rate": 9.548897446827322e-05,
+      "loss": 1.0875,
+      "step": 1815
+    },
+    {
+      "epoch": 0.16226238077155047,
+      "grad_norm": 0.37720173597335815,
+      "learning_rate": 9.548296588594575e-05,
+      "loss": 1.0008,
+      "step": 1816
+    },
+    {
+      "epoch": 0.16235173230281233,
+      "grad_norm": 0.4290127754211426,
+      "learning_rate": 9.547695349393504e-05,
+      "loss": 1.0353,
+      "step": 1817
+    },
+    {
+      "epoch": 0.1624410838340742,
+      "grad_norm": 0.4377609193325043,
+      "learning_rate": 9.547093729274474e-05,
+      "loss": 1.0352,
+      "step": 1818
+    },
+    {
+      "epoch": 0.16253043536533607,
+      "grad_norm": 0.3721230924129486,
+      "learning_rate": 9.546491728287876e-05,
+      "loss": 1.0833,
+      "step": 1819
+    },
+    {
+      "epoch": 0.16261978689659795,
+      "grad_norm": 0.43184563517570496,
+      "learning_rate": 9.545889346484134e-05,
+      "loss": 0.9966,
+      "step": 1820
+    },
+    {
+      "epoch": 0.1627091384278598,
+      "grad_norm": 0.41010618209838867,
+      "learning_rate": 9.545286583913702e-05,
+      "loss": 0.9989,
+      "step": 1821
+    },
+    {
+      "epoch": 0.16279848995912166,
+      "grad_norm": 0.5316981077194214,
+      "learning_rate": 9.54468344062707e-05,
+      "loss": 0.9844,
+      "step": 1822
+    },
+    {
+      "epoch": 0.16288784149038354,
+      "grad_norm": 0.35896286368370056,
+      "learning_rate": 9.544079916674757e-05,
+      "loss": 1.0765,
+      "step": 1823
+    },
+    {
+      "epoch": 0.1629771930216454,
+      "grad_norm": 0.45125991106033325,
+      "learning_rate": 9.543476012107313e-05,
+      "loss": 1.1044,
+      "step": 1824
+    },
+    {
+      "epoch": 0.16306654455290728,
+      "grad_norm": 0.4255354404449463,
+      "learning_rate": 9.542871726975327e-05,
+      "loss": 1.0465,
+      "step": 1825
+    },
+    {
+      "epoch": 0.16315589608416914,
+      "grad_norm": 0.4363675117492676,
+      "learning_rate": 9.542267061329406e-05,
+      "loss": 1.0555,
+      "step": 1826
+    },
+    {
+      "epoch": 0.16324524761543102,
+      "grad_norm": 0.3892308175563812,
+      "learning_rate": 9.541662015220205e-05,
+      "loss": 1.055,
+      "step": 1827
+    },
+    {
+      "epoch": 0.16333459914669288,
+      "grad_norm": 0.43629783391952515,
+      "learning_rate": 9.541056588698397e-05,
+      "loss": 1.0117,
+      "step": 1828
+    },
+    {
+      "epoch": 0.16342395067795473,
+      "grad_norm": 0.454475998878479,
+      "learning_rate": 9.540450781814696e-05,
+      "loss": 1.0849,
+      "step": 1829
+    },
+    {
+      "epoch": 0.16351330220921662,
+      "grad_norm": 0.41698095202445984,
+      "learning_rate": 9.539844594619845e-05,
+      "loss": 1.0137,
+      "step": 1830
+    },
+    {
+      "epoch": 0.16360265374047847,
+      "grad_norm": 0.41076821088790894,
+      "learning_rate": 9.539238027164619e-05,
+      "loss": 0.9891,
+      "step": 1831
+    },
+    {
+      "epoch": 0.16369200527174035,
+      "grad_norm": 0.4605240821838379,
+      "learning_rate": 9.538631079499823e-05,
+      "loss": 0.9447,
+      "step": 1832
+    },
+    {
+      "epoch": 0.1637813568030022,
+      "grad_norm": 0.36705470085144043,
+      "learning_rate": 9.538023751676294e-05,
+      "loss": 1.023,
+      "step": 1833
+    },
+    {
+      "epoch": 0.16387070833426406,
+      "grad_norm": 0.44232022762298584,
+      "learning_rate": 9.537416043744905e-05,
+      "loss": 0.9757,
+      "step": 1834
+    },
+    {
+      "epoch": 0.16396005986552595,
+      "grad_norm": 0.41922441124916077,
+      "learning_rate": 9.536807955756557e-05,
+      "loss": 1.0834,
+      "step": 1835
+    },
+    {
+      "epoch": 0.1640494113967878,
+      "grad_norm": 0.44340333342552185,
+      "learning_rate": 9.536199487762182e-05,
+      "loss": 0.9799,
+      "step": 1836
+    },
+    {
+      "epoch": 0.1641387629280497,
+      "grad_norm": 0.5596421360969543,
+      "learning_rate": 9.53559063981275e-05,
+      "loss": 0.921,
+      "step": 1837
+    },
+    {
+      "epoch": 0.16422811445931154,
+      "grad_norm": 0.3959425091743469,
+      "learning_rate": 9.534981411959255e-05,
+      "loss": 1.0426,
+      "step": 1838
+    },
+    {
+      "epoch": 0.16431746599057342,
+      "grad_norm": 0.431082546710968,
+      "learning_rate": 9.534371804252728e-05,
+      "loss": 1.1178,
+      "step": 1839
+    },
+    {
+      "epoch": 0.16440681752183528,
+      "grad_norm": 0.42235735058784485,
+      "learning_rate": 9.533761816744228e-05,
+      "loss": 1.0315,
+      "step": 1840
+    },
+    {
+      "epoch": 0.16449616905309714,
+      "grad_norm": 0.4385235905647278,
+      "learning_rate": 9.533151449484851e-05,
+      "loss": 1.0135,
+      "step": 1841
+    },
+    {
+      "epoch": 0.16458552058435902,
+      "grad_norm": 0.41497084498405457,
+      "learning_rate": 9.53254070252572e-05,
+      "loss": 1.0006,
+      "step": 1842
+    },
+    {
+      "epoch": 0.16467487211562087,
+      "grad_norm": 0.4509193003177643,
+      "learning_rate": 9.531929575917991e-05,
+      "loss": 0.9943,
+      "step": 1843
+    },
+    {
+      "epoch": 0.16476422364688276,
+      "grad_norm": 0.45146405696868896,
+      "learning_rate": 9.531318069712854e-05,
+      "loss": 0.9926,
+      "step": 1844
+    },
+    {
+      "epoch": 0.1648535751781446,
+      "grad_norm": 0.43658965826034546,
+      "learning_rate": 9.530706183961526e-05,
+      "loss": 0.9999,
+      "step": 1845
+    },
+    {
+      "epoch": 0.1649429267094065,
+      "grad_norm": 0.4209139347076416,
+      "learning_rate": 9.530093918715264e-05,
+      "loss": 1.0795,
+      "step": 1846
+    },
+    {
+      "epoch": 0.16503227824066835,
+      "grad_norm": 0.39283883571624756,
+      "learning_rate": 9.529481274025347e-05,
+      "loss": 1.101,
+      "step": 1847
+    },
+    {
+      "epoch": 0.1651216297719302,
+      "grad_norm": 0.5160419344902039,
+      "learning_rate": 9.528868249943095e-05,
+      "loss": 0.9985,
+      "step": 1848
+    },
+    {
+      "epoch": 0.1652109813031921,
+      "grad_norm": 0.5293583869934082,
+      "learning_rate": 9.528254846519851e-05,
+      "loss": 0.9466,
+      "step": 1849
+    },
+    {
+      "epoch": 0.16530033283445394,
+      "grad_norm": 0.4382696747779846,
+      "learning_rate": 9.527641063806996e-05,
+      "loss": 1.0173,
+      "step": 1850
+    },
+    {
+      "epoch": 0.16538968436571583,
+      "grad_norm": 0.4551234841346741,
+      "learning_rate": 9.52702690185594e-05,
+      "loss": 0.9932,
+      "step": 1851
+    },
+    {
+      "epoch": 0.16547903589697768,
+      "grad_norm": 0.42946553230285645,
+      "learning_rate": 9.526412360718127e-05,
+      "loss": 1.0688,
+      "step": 1852
+    },
+    {
+      "epoch": 0.16556838742823954,
+      "grad_norm": 0.5565560460090637,
+      "learning_rate": 9.525797440445031e-05,
+      "loss": 1.024,
+      "step": 1853
+    },
+    {
+      "epoch": 0.16565773895950142,
+      "grad_norm": 0.483884334564209,
+      "learning_rate": 9.525182141088159e-05,
+      "loss": 0.9546,
+      "step": 1854
+    },
+    {
+      "epoch": 0.16574709049076328,
+      "grad_norm": 0.4862859547138214,
+      "learning_rate": 9.524566462699045e-05,
+      "loss": 0.9695,
+      "step": 1855
+    },
+    {
+      "epoch": 0.16583644202202516,
+      "grad_norm": 0.5029091238975525,
+      "learning_rate": 9.523950405329262e-05,
+      "loss": 0.9478,
+      "step": 1856
+    },
+    {
+      "epoch": 0.16592579355328702,
+      "grad_norm": 0.6389296650886536,
+      "learning_rate": 9.523333969030413e-05,
+      "loss": 1.0239,
+      "step": 1857
+    },
+    {
+      "epoch": 0.1660151450845489,
+      "grad_norm": 0.39252007007598877,
+      "learning_rate": 9.522717153854125e-05,
+      "loss": 1.1161,
+      "step": 1858
+    },
+    {
+      "epoch": 0.16610449661581075,
+      "grad_norm": 0.45961377024650574,
+      "learning_rate": 9.522099959852071e-05,
+      "loss": 1.0234,
+      "step": 1859
+    },
+    {
+      "epoch": 0.1661938481470726,
+      "grad_norm": 0.3859967291355133,
+      "learning_rate": 9.52148238707594e-05,
+      "loss": 1.0622,
+      "step": 1860
+    },
+    {
+      "epoch": 0.1662831996783345,
+      "grad_norm": 0.48423507809638977,
+      "learning_rate": 9.520864435577466e-05,
+      "loss": 1.0117,
+      "step": 1861
+    },
+    {
+      "epoch": 0.16637255120959635,
+      "grad_norm": 0.4309901297092438,
+      "learning_rate": 9.520246105408403e-05,
+      "loss": 1.0523,
+      "step": 1862
+    },
+    {
+      "epoch": 0.16646190274085823,
+      "grad_norm": 0.487351655960083,
+      "learning_rate": 9.519627396620549e-05,
+      "loss": 0.9941,
+      "step": 1863
+    },
+    {
+      "epoch": 0.1665512542721201,
+      "grad_norm": 0.44148099422454834,
+      "learning_rate": 9.519008309265724e-05,
+      "loss": 1.0301,
+      "step": 1864
+    },
+    {
+      "epoch": 0.16664060580338197,
+      "grad_norm": 0.4186917245388031,
+      "learning_rate": 9.518388843395786e-05,
+      "loss": 0.996,
+      "step": 1865
+    },
+    {
+      "epoch": 0.16672995733464382,
+      "grad_norm": 0.4065192937850952,
+      "learning_rate": 9.517768999062617e-05,
+      "loss": 1.0351,
+      "step": 1866
+    },
+    {
+      "epoch": 0.16681930886590568,
+      "grad_norm": 0.402235209941864,
+      "learning_rate": 9.51714877631814e-05,
+      "loss": 1.0227,
+      "step": 1867
+    },
+    {
+      "epoch": 0.16690866039716756,
+      "grad_norm": 0.4134058952331543,
+      "learning_rate": 9.516528175214303e-05,
+      "loss": 1.0518,
+      "step": 1868
+    },
+    {
+      "epoch": 0.16699801192842942,
+      "grad_norm": 0.39137229323387146,
+      "learning_rate": 9.515907195803088e-05,
+      "loss": 1.0648,
+      "step": 1869
+    },
+    {
+      "epoch": 0.1670873634596913,
+      "grad_norm": 0.43055975437164307,
+      "learning_rate": 9.51528583813651e-05,
+      "loss": 0.9951,
+      "step": 1870
+    },
+    {
+      "epoch": 0.16717671499095316,
+      "grad_norm": 0.45151713490486145,
+      "learning_rate": 9.514664102266615e-05,
+      "loss": 1.0613,
+      "step": 1871
+    },
+    {
+      "epoch": 0.167266066522215,
+      "grad_norm": 0.45818963646888733,
+      "learning_rate": 9.514041988245477e-05,
+      "loss": 1.0874,
+      "step": 1872
+    },
+    {
+      "epoch": 0.1673554180534769,
+      "grad_norm": 0.38174670934677124,
+      "learning_rate": 9.513419496125206e-05,
+      "loss": 1.0888,
+      "step": 1873
+    },
+    {
+      "epoch": 0.16744476958473875,
+      "grad_norm": 0.4613979160785675,
+      "learning_rate": 9.512796625957943e-05,
+      "loss": 0.9875,
+      "step": 1874
+    },
+    {
+      "epoch": 0.16753412111600063,
+      "grad_norm": 0.49428126215934753,
+      "learning_rate": 9.512173377795859e-05,
+      "loss": 0.9716,
+      "step": 1875
+    },
+    {
+      "epoch": 0.1676234726472625,
+      "grad_norm": 0.5154160261154175,
+      "learning_rate": 9.511549751691159e-05,
+      "loss": 0.9448,
+      "step": 1876
+    },
+    {
+      "epoch": 0.16771282417852437,
+      "grad_norm": 0.4421161711215973,
+      "learning_rate": 9.510925747696077e-05,
+      "loss": 1.009,
+      "step": 1877
+    },
+    {
+      "epoch": 0.16780217570978623,
+      "grad_norm": 0.5333375930786133,
+      "learning_rate": 9.510301365862882e-05,
+      "loss": 1.0682,
+      "step": 1878
+    },
+    {
+      "epoch": 0.16789152724104808,
+      "grad_norm": 0.35133177042007446,
+      "learning_rate": 9.509676606243869e-05,
+      "loss": 1.0646,
+      "step": 1879
+    },
+    {
+      "epoch": 0.16798087877230997,
+      "grad_norm": 0.39709219336509705,
+      "learning_rate": 9.509051468891372e-05,
+      "loss": 1.0559,
+      "step": 1880
+    },
+    {
+      "epoch": 0.16807023030357182,
+      "grad_norm": 0.45971524715423584,
+      "learning_rate": 9.508425953857752e-05,
+      "loss": 1.0145,
+      "step": 1881
+    },
+    {
+      "epoch": 0.1681595818348337,
+      "grad_norm": 0.44881150126457214,
+      "learning_rate": 9.507800061195401e-05,
+      "loss": 1.0404,
+      "step": 1882
+    },
+    {
+      "epoch": 0.16824893336609556,
+      "grad_norm": 0.5401568412780762,
+      "learning_rate": 9.507173790956746e-05,
+      "loss": 0.9762,
+      "step": 1883
+    },
+    {
+      "epoch": 0.16833828489735742,
+      "grad_norm": 0.48548710346221924,
+      "learning_rate": 9.506547143194242e-05,
+      "loss": 1.0197,
+      "step": 1884
+    },
+    {
+      "epoch": 0.1684276364286193,
+      "grad_norm": 0.4435642659664154,
+      "learning_rate": 9.505920117960379e-05,
+      "loss": 1.0289,
+      "step": 1885
+    },
+    {
+      "epoch": 0.16851698795988115,
+      "grad_norm": 0.4363146722316742,
+      "learning_rate": 9.505292715307676e-05,
+      "loss": 1.0346,
+      "step": 1886
+    },
+    {
+      "epoch": 0.16860633949114304,
+      "grad_norm": 0.45172473788261414,
+      "learning_rate": 9.504664935288685e-05,
+      "loss": 1.0222,
+      "step": 1887
+    },
+    {
+      "epoch": 0.1686956910224049,
+      "grad_norm": 0.4839716851711273,
+      "learning_rate": 9.504036777955991e-05,
+      "loss": 1.0304,
+      "step": 1888
+    },
+    {
+      "epoch": 0.16878504255366678,
+      "grad_norm": 0.4805682599544525,
+      "learning_rate": 9.503408243362206e-05,
+      "loss": 1.0433,
+      "step": 1889
+    },
+    {
+      "epoch": 0.16887439408492863,
+      "grad_norm": 0.4286407232284546,
+      "learning_rate": 9.502779331559977e-05,
+      "loss": 1.059,
+      "step": 1890
+    },
+    {
+      "epoch": 0.16896374561619049,
+      "grad_norm": 0.39787065982818604,
+      "learning_rate": 9.502150042601985e-05,
+      "loss": 1.1139,
+      "step": 1891
+    },
+    {
+      "epoch": 0.16905309714745237,
+      "grad_norm": 0.42047420144081116,
+      "learning_rate": 9.501520376540936e-05,
+      "loss": 1.0147,
+      "step": 1892
+    },
+    {
+      "epoch": 0.16914244867871422,
+      "grad_norm": 0.4411332309246063,
+      "learning_rate": 9.500890333429573e-05,
+      "loss": 1.0576,
+      "step": 1893
+    },
+    {
+      "epoch": 0.1692318002099761,
+      "grad_norm": 0.43769532442092896,
+      "learning_rate": 9.500259913320668e-05,
+      "loss": 1.0455,
+      "step": 1894
+    },
+    {
+      "epoch": 0.16932115174123796,
+      "grad_norm": 0.4141598343849182,
+      "learning_rate": 9.499629116267026e-05,
+      "loss": 1.0782,
+      "step": 1895
+    },
+    {
+      "epoch": 0.16941050327249985,
+      "grad_norm": 0.4736109972000122,
+      "learning_rate": 9.498997942321483e-05,
+      "loss": 0.9884,
+      "step": 1896
+    },
+    {
+      "epoch": 0.1694998548037617,
+      "grad_norm": 0.38666075468063354,
+      "learning_rate": 9.498366391536907e-05,
+      "loss": 1.0701,
+      "step": 1897
+    },
+    {
+      "epoch": 0.16958920633502356,
+      "grad_norm": 0.4399631917476654,
+      "learning_rate": 9.497734463966196e-05,
+      "loss": 0.996,
+      "step": 1898
+    },
+    {
+      "epoch": 0.16967855786628544,
+      "grad_norm": 0.3586561977863312,
+      "learning_rate": 9.497102159662281e-05,
+      "loss": 1.0366,
+      "step": 1899
+    },
+    {
+      "epoch": 0.1697679093975473,
+      "grad_norm": 0.3949701189994812,
+      "learning_rate": 9.496469478678126e-05,
+      "loss": 1.0616,
+      "step": 1900
+    },
+    {
+      "epoch": 0.16985726092880918,
+      "grad_norm": 0.38532134890556335,
+      "learning_rate": 9.495836421066722e-05,
+      "loss": 1.0576,
+      "step": 1901
+    },
+    {
+      "epoch": 0.16994661246007103,
+      "grad_norm": 0.4000793993473053,
+      "learning_rate": 9.495202986881095e-05,
+      "loss": 1.0262,
+      "step": 1902
+    },
+    {
+      "epoch": 0.1700359639913329,
+      "grad_norm": 0.3810192048549652,
+      "learning_rate": 9.494569176174304e-05,
+      "loss": 1.0097,
+      "step": 1903
+    },
+    {
+      "epoch": 0.17012531552259477,
+      "grad_norm": 0.4256387948989868,
+      "learning_rate": 9.493934988999436e-05,
+      "loss": 1.0186,
+      "step": 1904
+    },
+    {
+      "epoch": 0.17021466705385663,
+      "grad_norm": 0.4718528389930725,
+      "learning_rate": 9.493300425409609e-05,
+      "loss": 0.9866,
+      "step": 1905
+    },
+    {
+      "epoch": 0.1703040185851185,
+      "grad_norm": 0.3986504077911377,
+      "learning_rate": 9.492665485457976e-05,
+      "loss": 1.0735,
+      "step": 1906
+    },
+    {
+      "epoch": 0.17039337011638037,
+      "grad_norm": 0.4253017008304596,
+      "learning_rate": 9.492030169197722e-05,
+      "loss": 1.0168,
+      "step": 1907
+    },
+    {
+      "epoch": 0.17048272164764225,
+      "grad_norm": 0.34923726320266724,
+      "learning_rate": 9.49139447668206e-05,
+      "loss": 1.1098,
+      "step": 1908
+    },
+    {
+      "epoch": 0.1705720731789041,
+      "grad_norm": 0.46308934688568115,
+      "learning_rate": 9.490758407964234e-05,
+      "loss": 0.9807,
+      "step": 1909
+    },
+    {
+      "epoch": 0.17066142471016596,
+      "grad_norm": 0.5132005214691162,
+      "learning_rate": 9.490121963097525e-05,
+      "loss": 1.0316,
+      "step": 1910
+    },
+    {
+      "epoch": 0.17075077624142784,
+      "grad_norm": 0.4535066485404968,
+      "learning_rate": 9.489485142135238e-05,
+      "loss": 0.9653,
+      "step": 1911
+    },
+    {
+      "epoch": 0.1708401277726897,
+      "grad_norm": 0.41993486881256104,
+      "learning_rate": 9.488847945130718e-05,
+      "loss": 1.0554,
+      "step": 1912
+    },
+    {
+      "epoch": 0.17092947930395158,
+      "grad_norm": 0.46821948885917664,
+      "learning_rate": 9.488210372137335e-05,
+      "loss": 0.9511,
+      "step": 1913
+    },
+    {
+      "epoch": 0.17101883083521344,
+      "grad_norm": 0.5106322765350342,
+      "learning_rate": 9.487572423208491e-05,
+      "loss": 0.9633,
+      "step": 1914
+    },
+    {
+      "epoch": 0.1711081823664753,
+      "grad_norm": 0.4303719997406006,
+      "learning_rate": 9.486934098397622e-05,
+      "loss": 1.0664,
+      "step": 1915
+    },
+    {
+      "epoch": 0.17119753389773718,
+      "grad_norm": 0.3766723573207855,
+      "learning_rate": 9.486295397758196e-05,
+      "loss": 1.0226,
+      "step": 1916
+    },
+    {
+      "epoch": 0.17128688542899903,
+      "grad_norm": 0.4431716799736023,
+      "learning_rate": 9.48565632134371e-05,
+      "loss": 1.0226,
+      "step": 1917
+    },
+    {
+      "epoch": 0.1713762369602609,
+      "grad_norm": 0.41526105999946594,
+      "learning_rate": 9.485016869207695e-05,
+      "loss": 1.0072,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17146558849152277,
+      "grad_norm": 0.4168999195098877,
+      "learning_rate": 9.484377041403706e-05,
+      "loss": 1.0846,
+      "step": 1919
+    },
+    {
+      "epoch": 0.17155494002278465,
+      "grad_norm": 0.4972972571849823,
+      "learning_rate": 9.483736837985344e-05,
+      "loss": 1.0172,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1716442915540465,
+      "grad_norm": 0.4009465277194977,
+      "learning_rate": 9.483096259006228e-05,
+      "loss": 1.04,
+      "step": 1921
+    },
+    {
+      "epoch": 0.17173364308530836,
+      "grad_norm": 0.37553495168685913,
+      "learning_rate": 9.482455304520013e-05,
+      "loss": 1.0406,
+      "step": 1922
+    },
+    {
+      "epoch": 0.17182299461657025,
+      "grad_norm": 0.37648624181747437,
+      "learning_rate": 9.481813974580386e-05,
+      "loss": 1.0371,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1719123461478321,
+      "grad_norm": 0.3838346302509308,
+      "learning_rate": 9.481172269241067e-05,
+      "loss": 0.9912,
+      "step": 1924
+    },
+    {
+      "epoch": 0.17200169767909398,
+      "grad_norm": 0.4350638687610626,
+      "learning_rate": 9.480530188555805e-05,
+      "loss": 1.0651,
+      "step": 1925
+    },
+    {
+      "epoch": 0.17209104921035584,
+      "grad_norm": 0.46230390667915344,
+      "learning_rate": 9.47988773257838e-05,
+      "loss": 0.9876,
+      "step": 1926
+    },
+    {
+      "epoch": 0.17218040074161772,
+      "grad_norm": 0.41337013244628906,
+      "learning_rate": 9.479244901362605e-05,
+      "loss": 1.09,
+      "step": 1927
+    },
+    {
+      "epoch": 0.17226975227287958,
+      "grad_norm": 0.4298297166824341,
+      "learning_rate": 9.478601694962323e-05,
+      "loss": 1.1069,
+      "step": 1928
+    },
+    {
+      "epoch": 0.17235910380414143,
+      "grad_norm": 0.4469965398311615,
+      "learning_rate": 9.477958113431413e-05,
+      "loss": 1.0354,
+      "step": 1929
+    },
+    {
+      "epoch": 0.17244845533540332,
+      "grad_norm": 0.4125996530056,
+      "learning_rate": 9.477314156823779e-05,
+      "loss": 1.0402,
+      "step": 1930
+    },
+    {
+      "epoch": 0.17253780686666517,
+      "grad_norm": 0.44828668236732483,
+      "learning_rate": 9.476669825193359e-05,
+      "loss": 1.0471,
+      "step": 1931
+    },
+    {
+      "epoch": 0.17262715839792706,
+      "grad_norm": 0.4745740592479706,
+      "learning_rate": 9.476025118594124e-05,
+      "loss": 1.01,
+      "step": 1932
+    },
+    {
+      "epoch": 0.1727165099291889,
+      "grad_norm": 0.46582111716270447,
+      "learning_rate": 9.475380037080073e-05,
+      "loss": 1.0378,
+      "step": 1933
+    },
+    {
+      "epoch": 0.17280586146045077,
+      "grad_norm": 0.45007333159446716,
+      "learning_rate": 9.47473458070524e-05,
+      "loss": 1.0021,
+      "step": 1934
+    },
+    {
+      "epoch": 0.17289521299171265,
+      "grad_norm": 0.3929324448108673,
+      "learning_rate": 9.474088749523689e-05,
+      "loss": 0.9972,
+      "step": 1935
+    },
+    {
+      "epoch": 0.1729845645229745,
+      "grad_norm": 0.42300376296043396,
+      "learning_rate": 9.473442543589515e-05,
+      "loss": 1.0603,
+      "step": 1936
+    },
+    {
+      "epoch": 0.1730739160542364,
+      "grad_norm": 0.5826932191848755,
+      "learning_rate": 9.472795962956844e-05,
+      "loss": 0.9595,
+      "step": 1937
+    },
+    {
+      "epoch": 0.17316326758549824,
+      "grad_norm": 0.5283807516098022,
+      "learning_rate": 9.472149007679836e-05,
+      "loss": 0.9726,
+      "step": 1938
+    },
+    {
+      "epoch": 0.17325261911676013,
+      "grad_norm": 0.4322817921638489,
+      "learning_rate": 9.471501677812677e-05,
+      "loss": 0.9882,
+      "step": 1939
+    },
+    {
+      "epoch": 0.17334197064802198,
+      "grad_norm": 0.42819416522979736,
+      "learning_rate": 9.47085397340959e-05,
+      "loss": 1.0776,
+      "step": 1940
+    },
+    {
+      "epoch": 0.17343132217928384,
+      "grad_norm": 0.379272997379303,
+      "learning_rate": 9.470205894524829e-05,
+      "loss": 1.0924,
+      "step": 1941
+    },
+    {
+      "epoch": 0.17352067371054572,
+      "grad_norm": 0.4515071213245392,
+      "learning_rate": 9.469557441212674e-05,
+      "loss": 1.0024,
+      "step": 1942
+    },
+    {
+      "epoch": 0.17361002524180758,
+      "grad_norm": 0.4782482981681824,
+      "learning_rate": 9.468908613527441e-05,
+      "loss": 0.9794,
+      "step": 1943
+    },
+    {
+      "epoch": 0.17369937677306946,
+      "grad_norm": 0.3846750855445862,
+      "learning_rate": 9.468259411523476e-05,
+      "loss": 0.9786,
+      "step": 1944
+    },
+    {
+      "epoch": 0.1737887283043313,
+      "grad_norm": 0.4089057743549347,
+      "learning_rate": 9.467609835255158e-05,
+      "loss": 1.1039,
+      "step": 1945
+    },
+    {
+      "epoch": 0.17387807983559317,
+      "grad_norm": 0.5269849896430969,
+      "learning_rate": 9.466959884776894e-05,
+      "loss": 1.0726,
+      "step": 1946
+    },
+    {
+      "epoch": 0.17396743136685505,
+      "grad_norm": 0.43416985869407654,
+      "learning_rate": 9.466309560143126e-05,
+      "loss": 1.0341,
+      "step": 1947
+    },
+    {
+      "epoch": 0.1740567828981169,
+      "grad_norm": 0.41887369751930237,
+      "learning_rate": 9.465658861408324e-05,
+      "loss": 1.0519,
+      "step": 1948
+    },
+    {
+      "epoch": 0.1741461344293788,
+      "grad_norm": 0.47824710607528687,
+      "learning_rate": 9.465007788626993e-05,
+      "loss": 1.0775,
+      "step": 1949
+    },
+    {
+      "epoch": 0.17423548596064065,
+      "grad_norm": 0.3974583148956299,
+      "learning_rate": 9.464356341853666e-05,
+      "loss": 1.026,
+      "step": 1950
+    },
+    {
+      "epoch": 0.17432483749190253,
+      "grad_norm": 0.4442320764064789,
+      "learning_rate": 9.463704521142909e-05,
+      "loss": 1.0781,
+      "step": 1951
+    },
+    {
+      "epoch": 0.17441418902316438,
+      "grad_norm": 0.4253045618534088,
+      "learning_rate": 9.463052326549317e-05,
+      "loss": 1.0449,
+      "step": 1952
+    },
+    {
+      "epoch": 0.17450354055442624,
+      "grad_norm": 0.40720415115356445,
+      "learning_rate": 9.462399758127521e-05,
+      "loss": 1.0281,
+      "step": 1953
+    },
+    {
+      "epoch": 0.17459289208568812,
+      "grad_norm": 0.46845048666000366,
+      "learning_rate": 9.46174681593218e-05,
+      "loss": 1.0313,
+      "step": 1954
+    },
+    {
+      "epoch": 0.17468224361694998,
+      "grad_norm": 0.4253866374492645,
+      "learning_rate": 9.461093500017984e-05,
+      "loss": 1.0191,
+      "step": 1955
+    },
+    {
+      "epoch": 0.17477159514821186,
+      "grad_norm": 0.41982749104499817,
+      "learning_rate": 9.460439810439655e-05,
+      "loss": 1.0332,
+      "step": 1956
+    },
+    {
+      "epoch": 0.17486094667947372,
+      "grad_norm": 0.45301470160484314,
+      "learning_rate": 9.459785747251948e-05,
+      "loss": 0.9582,
+      "step": 1957
+    },
+    {
+      "epoch": 0.1749502982107356,
+      "grad_norm": 0.44690030813217163,
+      "learning_rate": 9.459131310509646e-05,
+      "loss": 1.0285,
+      "step": 1958
+    },
+    {
+      "epoch": 0.17503964974199746,
+      "grad_norm": 0.5572661757469177,
+      "learning_rate": 9.458476500267566e-05,
+      "loss": 0.9837,
+      "step": 1959
+    },
+    {
+      "epoch": 0.1751290012732593,
+      "grad_norm": 0.3876509666442871,
+      "learning_rate": 9.457821316580555e-05,
+      "loss": 1.058,
+      "step": 1960
+    },
+    {
+      "epoch": 0.1752183528045212,
+      "grad_norm": 0.4823840856552124,
+      "learning_rate": 9.457165759503493e-05,
+      "loss": 0.9689,
+      "step": 1961
+    },
+    {
+      "epoch": 0.17530770433578305,
+      "grad_norm": 0.4866883456707001,
+      "learning_rate": 9.456509829091287e-05,
+      "loss": 1.0991,
+      "step": 1962
+    },
+    {
+      "epoch": 0.17539705586704493,
+      "grad_norm": 0.47182294726371765,
+      "learning_rate": 9.455853525398881e-05,
+      "loss": 0.898,
+      "step": 1963
+    },
+    {
+      "epoch": 0.1754864073983068,
+      "grad_norm": 0.4766868054866791,
+      "learning_rate": 9.455196848481244e-05,
+      "loss": 1.017,
+      "step": 1964
+    },
+    {
+      "epoch": 0.17557575892956864,
+      "grad_norm": 0.47620826959609985,
+      "learning_rate": 9.454539798393385e-05,
+      "loss": 1.117,
+      "step": 1965
+    },
+    {
+      "epoch": 0.17566511046083053,
+      "grad_norm": 0.391517698764801,
+      "learning_rate": 9.453882375190335e-05,
+      "loss": 1.0444,
+      "step": 1966
+    },
+    {
+      "epoch": 0.17575446199209238,
+      "grad_norm": 0.4250124394893646,
+      "learning_rate": 9.45322457892716e-05,
+      "loss": 1.1342,
+      "step": 1967
+    },
+    {
+      "epoch": 0.17584381352335426,
+      "grad_norm": 0.4004350006580353,
+      "learning_rate": 9.45256640965896e-05,
+      "loss": 1.0022,
+      "step": 1968
+    },
+    {
+      "epoch": 0.17593316505461612,
+      "grad_norm": 0.4620151221752167,
+      "learning_rate": 9.451907867440862e-05,
+      "loss": 1.0139,
+      "step": 1969
+    },
+    {
+      "epoch": 0.176022516585878,
+      "grad_norm": 0.43265095353126526,
+      "learning_rate": 9.451248952328025e-05,
+      "loss": 1.1093,
+      "step": 1970
+    },
+    {
+      "epoch": 0.17611186811713986,
+      "grad_norm": 0.5694372653961182,
+      "learning_rate": 9.450589664375643e-05,
+      "loss": 0.9312,
+      "step": 1971
+    },
+    {
+      "epoch": 0.1762012196484017,
+      "grad_norm": 0.4211016297340393,
+      "learning_rate": 9.449930003638935e-05,
+      "loss": 1.0137,
+      "step": 1972
+    },
+    {
+      "epoch": 0.1762905711796636,
+      "grad_norm": 0.44613519310951233,
+      "learning_rate": 9.449269970173158e-05,
+      "loss": 0.969,
+      "step": 1973
+    },
+    {
+      "epoch": 0.17637992271092545,
+      "grad_norm": 0.3986136019229889,
+      "learning_rate": 9.448609564033593e-05,
+      "loss": 1.1211,
+      "step": 1974
+    },
+    {
+      "epoch": 0.17646927424218734,
+      "grad_norm": 0.5278311967849731,
+      "learning_rate": 9.44794878527556e-05,
+      "loss": 0.9798,
+      "step": 1975
+    },
+    {
+      "epoch": 0.1765586257734492,
+      "grad_norm": 0.5130774974822998,
+      "learning_rate": 9.447287633954405e-05,
+      "loss": 1.0396,
+      "step": 1976
+    },
+    {
+      "epoch": 0.17664797730471105,
+      "grad_norm": 0.483941912651062,
+      "learning_rate": 9.446626110125505e-05,
+      "loss": 0.9771,
+      "step": 1977
+    },
+    {
+      "epoch": 0.17673732883597293,
+      "grad_norm": 0.46051788330078125,
+      "learning_rate": 9.445964213844269e-05,
+      "loss": 1.0256,
+      "step": 1978
+    },
+    {
+      "epoch": 0.17682668036723478,
+      "grad_norm": 0.5272265076637268,
+      "learning_rate": 9.445301945166143e-05,
+      "loss": 0.9884,
+      "step": 1979
+    },
+    {
+      "epoch": 0.17691603189849667,
+      "grad_norm": 0.4559291899204254,
+      "learning_rate": 9.444639304146593e-05,
+      "loss": 0.9899,
+      "step": 1980
+    },
+    {
+      "epoch": 0.17700538342975852,
+      "grad_norm": 0.42420053482055664,
+      "learning_rate": 9.443976290841126e-05,
+      "loss": 0.9969,
+      "step": 1981
+    },
+    {
+      "epoch": 0.1770947349610204,
+      "grad_norm": 0.5223949551582336,
+      "learning_rate": 9.443312905305274e-05,
+      "loss": 1.0163,
+      "step": 1982
+    },
+    {
+      "epoch": 0.17718408649228226,
+      "grad_norm": 0.4252959191799164,
+      "learning_rate": 9.442649147594606e-05,
+      "loss": 1.0653,
+      "step": 1983
+    },
+    {
+      "epoch": 0.17727343802354412,
+      "grad_norm": 0.40141063928604126,
+      "learning_rate": 9.441985017764715e-05,
+      "loss": 1.0184,
+      "step": 1984
+    },
+    {
+      "epoch": 0.177362789554806,
+      "grad_norm": 0.4690384566783905,
+      "learning_rate": 9.44132051587123e-05,
+      "loss": 0.984,
+      "step": 1985
+    },
+    {
+      "epoch": 0.17745214108606785,
+      "grad_norm": 0.412387490272522,
+      "learning_rate": 9.440655641969814e-05,
+      "loss": 1.0184,
+      "step": 1986
+    },
+    {
+      "epoch": 0.17754149261732974,
+      "grad_norm": 0.37352073192596436,
+      "learning_rate": 9.43999039611615e-05,
+      "loss": 1.0471,
+      "step": 1987
+    },
+    {
+      "epoch": 0.1776308441485916,
+      "grad_norm": 0.4083310663700104,
+      "learning_rate": 9.439324778365965e-05,
+      "loss": 1.043,
+      "step": 1988
+    },
+    {
+      "epoch": 0.17772019567985348,
+      "grad_norm": 0.41002601385116577,
+      "learning_rate": 9.43865878877501e-05,
+      "loss": 0.9932,
+      "step": 1989
+    },
+    {
+      "epoch": 0.17780954721111533,
+      "grad_norm": 0.5643458962440491,
+      "learning_rate": 9.437992427399069e-05,
+      "loss": 0.9207,
+      "step": 1990
+    },
+    {
+      "epoch": 0.1778988987423772,
+      "grad_norm": 0.43547579646110535,
+      "learning_rate": 9.437325694293957e-05,
+      "loss": 1.0288,
+      "step": 1991
+    },
+    {
+      "epoch": 0.17798825027363907,
+      "grad_norm": 0.41671237349510193,
+      "learning_rate": 9.43665858951552e-05,
+      "loss": 1.0792,
+      "step": 1992
+    },
+    {
+      "epoch": 0.17807760180490093,
+      "grad_norm": 0.3886220455169678,
+      "learning_rate": 9.435991113119634e-05,
+      "loss": 1.0739,
+      "step": 1993
+    },
+    {
+      "epoch": 0.1781669533361628,
+      "grad_norm": 0.42051705718040466,
+      "learning_rate": 9.435323265162207e-05,
+      "loss": 0.9779,
+      "step": 1994
+    },
+    {
+      "epoch": 0.17825630486742466,
+      "grad_norm": 0.4543735086917877,
+      "learning_rate": 9.43465504569918e-05,
+      "loss": 0.9808,
+      "step": 1995
+    },
+    {
+      "epoch": 0.17834565639868652,
+      "grad_norm": 0.43816715478897095,
+      "learning_rate": 9.433986454786523e-05,
+      "loss": 1.028,
+      "step": 1996
+    },
+    {
+      "epoch": 0.1784350079299484,
+      "grad_norm": 0.3944636881351471,
+      "learning_rate": 9.433317492480238e-05,
+      "loss": 1.0616,
+      "step": 1997
+    },
+    {
+      "epoch": 0.17852435946121026,
+      "grad_norm": 0.39649829268455505,
+      "learning_rate": 9.432648158836357e-05,
+      "loss": 0.9368,
+      "step": 1998
+    },
+    {
+      "epoch": 0.17861371099247214,
+      "grad_norm": 0.422755628824234,
+      "learning_rate": 9.431978453910943e-05,
+      "loss": 1.065,
+      "step": 1999
+    },
+    {
+      "epoch": 0.178703062523734,
+      "grad_norm": 0.40051138401031494,
+      "learning_rate": 9.431308377760094e-05,
+      "loss": 1.0022,
+      "step": 2000
+    },
+    {
+      "epoch": 0.17879241405499588,
+      "grad_norm": 0.514773428440094,
+      "learning_rate": 9.430637930439933e-05,
+      "loss": 0.9755,
+      "step": 2001
+    },
+    {
+      "epoch": 0.17888176558625773,
+      "grad_norm": 0.36524417996406555,
+      "learning_rate": 9.429967112006619e-05,
+      "loss": 1.0829,
+      "step": 2002
+    },
+    {
+      "epoch": 0.1789711171175196,
+      "grad_norm": 0.3744485378265381,
+      "learning_rate": 9.429295922516337e-05,
+      "loss": 1.0275,
+      "step": 2003
+    },
+    {
+      "epoch": 0.17906046864878147,
+      "grad_norm": 0.4923553168773651,
+      "learning_rate": 9.42862436202531e-05,
+      "loss": 0.9636,
+      "step": 2004
+    },
+    {
+      "epoch": 0.17914982018004333,
+      "grad_norm": 0.3743354082107544,
+      "learning_rate": 9.427952430589789e-05,
+      "loss": 1.0459,
+      "step": 2005
+    },
+    {
+      "epoch": 0.1792391717113052,
+      "grad_norm": 0.4594390392303467,
+      "learning_rate": 9.42728012826605e-05,
+      "loss": 1.0414,
+      "step": 2006
+    },
+    {
+      "epoch": 0.17932852324256707,
+      "grad_norm": 0.4210093915462494,
+      "learning_rate": 9.42660745511041e-05,
+      "loss": 1.088,
+      "step": 2007
+    },
+    {
+      "epoch": 0.17941787477382895,
+      "grad_norm": 0.4122314751148224,
+      "learning_rate": 9.425934411179211e-05,
+      "loss": 0.9287,
+      "step": 2008
+    },
+    {
+      "epoch": 0.1795072263050908,
+      "grad_norm": 0.41803067922592163,
+      "learning_rate": 9.425260996528829e-05,
+      "loss": 1.0501,
+      "step": 2009
+    },
+    {
+      "epoch": 0.17959657783635266,
+      "grad_norm": 0.4542330503463745,
+      "learning_rate": 9.424587211215669e-05,
+      "loss": 1.0908,
+      "step": 2010
+    },
+    {
+      "epoch": 0.17968592936761454,
+      "grad_norm": 0.4137341380119324,
+      "learning_rate": 9.423913055296165e-05,
+      "loss": 0.9796,
+      "step": 2011
+    },
+    {
+      "epoch": 0.1797752808988764,
+      "grad_norm": 0.3948180079460144,
+      "learning_rate": 9.42323852882679e-05,
+      "loss": 1.0888,
+      "step": 2012
+    },
+    {
+      "epoch": 0.17986463243013828,
+      "grad_norm": 0.398710161447525,
+      "learning_rate": 9.422563631864038e-05,
+      "loss": 1.0466,
+      "step": 2013
+    },
+    {
+      "epoch": 0.17995398396140014,
+      "grad_norm": 0.42401716113090515,
+      "learning_rate": 9.421888364464442e-05,
+      "loss": 1.0015,
+      "step": 2014
+    },
+    {
+      "epoch": 0.180043335492662,
+      "grad_norm": 0.45904120802879333,
+      "learning_rate": 9.42121272668456e-05,
+      "loss": 0.9718,
+      "step": 2015
+    },
+    {
+      "epoch": 0.18013268702392388,
+      "grad_norm": 0.45379403233528137,
+      "learning_rate": 9.420536718580986e-05,
+      "loss": 1.0008,
+      "step": 2016
+    },
+    {
+      "epoch": 0.18022203855518573,
+      "grad_norm": 0.4098423421382904,
+      "learning_rate": 9.419860340210342e-05,
+      "loss": 1.1183,
+      "step": 2017
+    },
+    {
+      "epoch": 0.18031139008644761,
+      "grad_norm": 0.3899945616722107,
+      "learning_rate": 9.419183591629284e-05,
+      "loss": 1.0653,
+      "step": 2018
+    },
+    {
+      "epoch": 0.18040074161770947,
+      "grad_norm": 0.44282397627830505,
+      "learning_rate": 9.418506472894492e-05,
+      "loss": 1.0232,
+      "step": 2019
+    },
+    {
+      "epoch": 0.18049009314897135,
+      "grad_norm": 0.3919029235839844,
+      "learning_rate": 9.417828984062687e-05,
+      "loss": 1.031,
+      "step": 2020
+    },
+    {
+      "epoch": 0.1805794446802332,
+      "grad_norm": 0.3580346703529358,
+      "learning_rate": 9.417151125190614e-05,
+      "loss": 1.1478,
+      "step": 2021
+    },
+    {
+      "epoch": 0.18066879621149506,
+      "grad_norm": 0.4946131706237793,
+      "learning_rate": 9.416472896335051e-05,
+      "loss": 0.9887,
+      "step": 2022
+    },
+    {
+      "epoch": 0.18075814774275695,
+      "grad_norm": 0.4798043966293335,
+      "learning_rate": 9.415794297552805e-05,
+      "loss": 0.9341,
+      "step": 2023
+    },
+    {
+      "epoch": 0.1808474992740188,
+      "grad_norm": 0.42261967062950134,
+      "learning_rate": 9.415115328900719e-05,
+      "loss": 1.063,
+      "step": 2024
+    },
+    {
+      "epoch": 0.18093685080528069,
+      "grad_norm": 0.4271705448627472,
+      "learning_rate": 9.414435990435663e-05,
+      "loss": 1.0327,
+      "step": 2025
+    },
+    {
+      "epoch": 0.18102620233654254,
+      "grad_norm": 0.394653856754303,
+      "learning_rate": 9.413756282214537e-05,
+      "loss": 1.1644,
+      "step": 2026
+    },
+    {
+      "epoch": 0.1811155538678044,
+      "grad_norm": 0.43346062302589417,
+      "learning_rate": 9.413076204294275e-05,
+      "loss": 1.0556,
+      "step": 2027
+    },
+    {
+      "epoch": 0.18120490539906628,
+      "grad_norm": 0.390903115272522,
+      "learning_rate": 9.412395756731843e-05,
+      "loss": 1.0185,
+      "step": 2028
+    },
+    {
+      "epoch": 0.18129425693032813,
+      "grad_norm": 0.3963910937309265,
+      "learning_rate": 9.41171493958423e-05,
+      "loss": 1.0377,
+      "step": 2029
+    },
+    {
+      "epoch": 0.18138360846159002,
+      "grad_norm": 0.4594999849796295,
+      "learning_rate": 9.41103375290847e-05,
+      "loss": 1.0494,
+      "step": 2030
+    },
+    {
+      "epoch": 0.18147295999285187,
+      "grad_norm": 0.42716097831726074,
+      "learning_rate": 9.410352196761612e-05,
+      "loss": 0.9836,
+      "step": 2031
+    },
+    {
+      "epoch": 0.18156231152411376,
+      "grad_norm": 0.43300801515579224,
+      "learning_rate": 9.409670271200745e-05,
+      "loss": 1.0358,
+      "step": 2032
+    },
+    {
+      "epoch": 0.1816516630553756,
+      "grad_norm": 0.3748253583908081,
+      "learning_rate": 9.408987976282993e-05,
+      "loss": 1.0185,
+      "step": 2033
+    },
+    {
+      "epoch": 0.18174101458663747,
+      "grad_norm": 0.4405686855316162,
+      "learning_rate": 9.408305312065498e-05,
+      "loss": 1.0303,
+      "step": 2034
+    },
+    {
+      "epoch": 0.18183036611789935,
+      "grad_norm": 0.4379626512527466,
+      "learning_rate": 9.407622278605445e-05,
+      "loss": 0.984,
+      "step": 2035
+    },
+    {
+      "epoch": 0.1819197176491612,
+      "grad_norm": 0.39666110277175903,
+      "learning_rate": 9.406938875960045e-05,
+      "loss": 1.1574,
+      "step": 2036
+    },
+    {
+      "epoch": 0.1820090691804231,
+      "grad_norm": 0.39308610558509827,
+      "learning_rate": 9.406255104186541e-05,
+      "loss": 1.0372,
+      "step": 2037
+    },
+    {
+      "epoch": 0.18209842071168494,
+      "grad_norm": 0.4402790069580078,
+      "learning_rate": 9.405570963342203e-05,
+      "loss": 1.102,
+      "step": 2038
+    },
+    {
+      "epoch": 0.18218777224294683,
+      "grad_norm": 0.3761218190193176,
+      "learning_rate": 9.404886453484336e-05,
+      "loss": 1.0392,
+      "step": 2039
+    },
+    {
+      "epoch": 0.18227712377420868,
+      "grad_norm": 0.62852942943573,
+      "learning_rate": 9.404201574670278e-05,
+      "loss": 0.9496,
+      "step": 2040
+    },
+    {
+      "epoch": 0.18236647530547054,
+      "grad_norm": 0.4357655644416809,
+      "learning_rate": 9.40351632695739e-05,
+      "loss": 1.0507,
+      "step": 2041
+    },
+    {
+      "epoch": 0.18245582683673242,
+      "grad_norm": 0.4296508729457855,
+      "learning_rate": 9.402830710403074e-05,
+      "loss": 1.0769,
+      "step": 2042
+    },
+    {
+      "epoch": 0.18254517836799428,
+      "grad_norm": 0.4572366774082184,
+      "learning_rate": 9.402144725064753e-05,
+      "loss": 0.9844,
+      "step": 2043
+    },
+    {
+      "epoch": 0.18263452989925616,
+      "grad_norm": 0.48715341091156006,
+      "learning_rate": 9.40145837099989e-05,
+      "loss": 0.9948,
+      "step": 2044
+    },
+    {
+      "epoch": 0.18272388143051801,
+      "grad_norm": 0.4071628749370575,
+      "learning_rate": 9.40077164826597e-05,
+      "loss": 1.0108,
+      "step": 2045
+    },
+    {
+      "epoch": 0.18281323296177987,
+      "grad_norm": 0.4973810017108917,
+      "learning_rate": 9.400084556920517e-05,
+      "loss": 1.0148,
+      "step": 2046
+    },
+    {
+      "epoch": 0.18290258449304175,
+      "grad_norm": 0.4013235867023468,
+      "learning_rate": 9.399397097021082e-05,
+      "loss": 0.9818,
+      "step": 2047
+    },
+    {
+      "epoch": 0.1829919360243036,
+      "grad_norm": 0.4264853596687317,
+      "learning_rate": 9.398709268625244e-05,
+      "loss": 1.0554,
+      "step": 2048
+    },
+    {
+      "epoch": 0.1830812875555655,
+      "grad_norm": 0.5234047174453735,
+      "learning_rate": 9.398021071790617e-05,
+      "loss": 1.0184,
+      "step": 2049
+    },
+    {
+      "epoch": 0.18317063908682735,
+      "grad_norm": 0.4011201560497284,
+      "learning_rate": 9.397332506574848e-05,
+      "loss": 1.0643,
+      "step": 2050
+    },
+    {
+      "epoch": 0.18325999061808923,
+      "grad_norm": 0.4270339906215668,
+      "learning_rate": 9.396643573035608e-05,
+      "loss": 1.0095,
+      "step": 2051
+    },
+    {
+      "epoch": 0.18334934214935109,
+      "grad_norm": 0.3894960880279541,
+      "learning_rate": 9.395954271230604e-05,
+      "loss": 1.0769,
+      "step": 2052
+    },
+    {
+      "epoch": 0.18343869368061294,
+      "grad_norm": 0.3954775035381317,
+      "learning_rate": 9.395264601217573e-05,
+      "loss": 1.0095,
+      "step": 2053
+    },
+    {
+      "epoch": 0.18352804521187482,
+      "grad_norm": 0.4338105618953705,
+      "learning_rate": 9.394574563054282e-05,
+      "loss": 0.9931,
+      "step": 2054
+    },
+    {
+      "epoch": 0.18361739674313668,
+      "grad_norm": 0.4074452519416809,
+      "learning_rate": 9.393884156798527e-05,
+      "loss": 1.0317,
+      "step": 2055
+    },
+    {
+      "epoch": 0.18370674827439856,
+      "grad_norm": 0.39676398038864136,
+      "learning_rate": 9.393193382508138e-05,
+      "loss": 0.9515,
+      "step": 2056
+    },
+    {
+      "epoch": 0.18379609980566042,
+      "grad_norm": 0.42293867468833923,
+      "learning_rate": 9.392502240240977e-05,
+      "loss": 1.1357,
+      "step": 2057
+    },
+    {
+      "epoch": 0.18388545133692227,
+      "grad_norm": 0.44433289766311646,
+      "learning_rate": 9.391810730054932e-05,
+      "loss": 1.0308,
+      "step": 2058
+    },
+    {
+      "epoch": 0.18397480286818416,
+      "grad_norm": 0.4930141270160675,
+      "learning_rate": 9.391118852007926e-05,
+      "loss": 1.0061,
+      "step": 2059
+    },
+    {
+      "epoch": 0.184064154399446,
+      "grad_norm": 0.3936423361301422,
+      "learning_rate": 9.390426606157907e-05,
+      "loss": 1.0441,
+      "step": 2060
+    },
+    {
+      "epoch": 0.1841535059307079,
+      "grad_norm": 0.5013602375984192,
+      "learning_rate": 9.389733992562863e-05,
+      "loss": 1.0648,
+      "step": 2061
+    },
+    {
+      "epoch": 0.18424285746196975,
+      "grad_norm": 0.4181208610534668,
+      "learning_rate": 9.389041011280808e-05,
+      "loss": 1.0172,
+      "step": 2062
+    },
+    {
+      "epoch": 0.18433220899323163,
+      "grad_norm": 0.3877794146537781,
+      "learning_rate": 9.388347662369782e-05,
+      "loss": 1.0403,
+      "step": 2063
+    },
+    {
+      "epoch": 0.1844215605244935,
+      "grad_norm": 0.4322667419910431,
+      "learning_rate": 9.387653945887864e-05,
+      "loss": 1.0691,
+      "step": 2064
+    },
+    {
+      "epoch": 0.18451091205575534,
+      "grad_norm": 0.4271637201309204,
+      "learning_rate": 9.386959861893158e-05,
+      "loss": 1.0122,
+      "step": 2065
+    },
+    {
+      "epoch": 0.18460026358701723,
+      "grad_norm": 0.3978729248046875,
+      "learning_rate": 9.386265410443802e-05,
+      "loss": 1.0317,
+      "step": 2066
+    },
+    {
+      "epoch": 0.18468961511827908,
+      "grad_norm": 0.40776076912879944,
+      "learning_rate": 9.385570591597966e-05,
+      "loss": 0.9875,
+      "step": 2067
+    },
+    {
+      "epoch": 0.18477896664954097,
+      "grad_norm": 0.3995235860347748,
+      "learning_rate": 9.384875405413843e-05,
+      "loss": 0.9726,
+      "step": 2068
+    },
+    {
+      "epoch": 0.18486831818080282,
+      "grad_norm": 0.3666049540042877,
+      "learning_rate": 9.384179851949666e-05,
+      "loss": 1.0619,
+      "step": 2069
+    },
+    {
+      "epoch": 0.1849576697120647,
+      "grad_norm": 0.37846672534942627,
+      "learning_rate": 9.383483931263695e-05,
+      "loss": 1.0445,
+      "step": 2070
+    },
+    {
+      "epoch": 0.18504702124332656,
+      "grad_norm": 0.44983261823654175,
+      "learning_rate": 9.382787643414221e-05,
+      "loss": 0.9651,
+      "step": 2071
+    },
+    {
+      "epoch": 0.18513637277458841,
+      "grad_norm": 0.46223461627960205,
+      "learning_rate": 9.382090988459564e-05,
+      "loss": 0.9688,
+      "step": 2072
+    },
+    {
+      "epoch": 0.1852257243058503,
+      "grad_norm": 0.3715665638446808,
+      "learning_rate": 9.381393966458077e-05,
+      "loss": 1.0355,
+      "step": 2073
+    },
+    {
+      "epoch": 0.18531507583711215,
+      "grad_norm": 0.45313575863838196,
+      "learning_rate": 9.380696577468142e-05,
+      "loss": 0.9863,
+      "step": 2074
+    },
+    {
+      "epoch": 0.18540442736837404,
+      "grad_norm": 0.45069244503974915,
+      "learning_rate": 9.379998821548175e-05,
+      "loss": 1.0141,
+      "step": 2075
+    },
+    {
+      "epoch": 0.1854937788996359,
+      "grad_norm": 0.3656700551509857,
+      "learning_rate": 9.37930069875662e-05,
+      "loss": 1.0549,
+      "step": 2076
+    },
+    {
+      "epoch": 0.18558313043089775,
+      "grad_norm": 0.44431185722351074,
+      "learning_rate": 9.37860220915195e-05,
+      "loss": 1.0339,
+      "step": 2077
+    },
+    {
+      "epoch": 0.18567248196215963,
+      "grad_norm": 0.5284201502799988,
+      "learning_rate": 9.377903352792672e-05,
+      "loss": 0.9624,
+      "step": 2078
+    },
+    {
+      "epoch": 0.18576183349342149,
+      "grad_norm": 0.3846912682056427,
+      "learning_rate": 9.377204129737325e-05,
+      "loss": 1.0481,
+      "step": 2079
+    },
+    {
+      "epoch": 0.18585118502468337,
+      "grad_norm": 0.41815292835235596,
+      "learning_rate": 9.376504540044472e-05,
+      "loss": 1.0371,
+      "step": 2080
+    },
+    {
+      "epoch": 0.18594053655594522,
+      "grad_norm": 0.38547441363334656,
+      "learning_rate": 9.375804583772716e-05,
+      "loss": 1.0483,
+      "step": 2081
+    },
+    {
+      "epoch": 0.1860298880872071,
+      "grad_norm": 0.4458550810813904,
+      "learning_rate": 9.375104260980683e-05,
+      "loss": 1.0046,
+      "step": 2082
+    },
+    {
+      "epoch": 0.18611923961846896,
+      "grad_norm": 0.5194621086120605,
+      "learning_rate": 9.374403571727032e-05,
+      "loss": 1.0876,
+      "step": 2083
+    },
+    {
+      "epoch": 0.18620859114973082,
+      "grad_norm": 0.41328129172325134,
+      "learning_rate": 9.373702516070453e-05,
+      "loss": 1.0493,
+      "step": 2084
+    },
+    {
+      "epoch": 0.1862979426809927,
+      "grad_norm": 0.4268978536128998,
+      "learning_rate": 9.373001094069671e-05,
+      "loss": 0.998,
+      "step": 2085
+    },
+    {
+      "epoch": 0.18638729421225456,
+      "grad_norm": 0.4924493134021759,
+      "learning_rate": 9.372299305783432e-05,
+      "loss": 1.0129,
+      "step": 2086
+    },
+    {
+      "epoch": 0.18647664574351644,
+      "grad_norm": 0.49052175879478455,
+      "learning_rate": 9.371597151270521e-05,
+      "loss": 0.9506,
+      "step": 2087
+    },
+    {
+      "epoch": 0.1865659972747783,
+      "grad_norm": 0.44871586561203003,
+      "learning_rate": 9.370894630589753e-05,
+      "loss": 0.9917,
+      "step": 2088
+    },
+    {
+      "epoch": 0.18665534880604015,
+      "grad_norm": 0.4465029239654541,
+      "learning_rate": 9.370191743799968e-05,
+      "loss": 1.0708,
+      "step": 2089
+    },
+    {
+      "epoch": 0.18674470033730203,
+      "grad_norm": 0.42947274446487427,
+      "learning_rate": 9.369488490960042e-05,
+      "loss": 1.0145,
+      "step": 2090
+    },
+    {
+      "epoch": 0.1868340518685639,
+      "grad_norm": 0.43883195519447327,
+      "learning_rate": 9.368784872128878e-05,
+      "loss": 1.0238,
+      "step": 2091
+    },
+    {
+      "epoch": 0.18692340339982577,
+      "grad_norm": 0.4147821068763733,
+      "learning_rate": 9.368080887365413e-05,
+      "loss": 1.019,
+      "step": 2092
+    },
+    {
+      "epoch": 0.18701275493108763,
+      "grad_norm": 0.4450010061264038,
+      "learning_rate": 9.367376536728613e-05,
+      "loss": 0.9557,
+      "step": 2093
+    },
+    {
+      "epoch": 0.1871021064623495,
+      "grad_norm": 0.5454333424568176,
+      "learning_rate": 9.366671820277477e-05,
+      "loss": 1.013,
+      "step": 2094
+    },
+    {
+      "epoch": 0.18719145799361137,
+      "grad_norm": 0.41965803503990173,
+      "learning_rate": 9.36596673807103e-05,
+      "loss": 0.9919,
+      "step": 2095
+    },
+    {
+      "epoch": 0.18728080952487322,
+      "grad_norm": 0.4064487814903259,
+      "learning_rate": 9.365261290168331e-05,
+      "loss": 0.9643,
+      "step": 2096
+    },
+    {
+      "epoch": 0.1873701610561351,
+      "grad_norm": 0.46642303466796875,
+      "learning_rate": 9.364555476628467e-05,
+      "loss": 0.9944,
+      "step": 2097
+    },
+    {
+      "epoch": 0.18745951258739696,
+      "grad_norm": 0.41478562355041504,
+      "learning_rate": 9.36384929751056e-05,
+      "loss": 1.0993,
+      "step": 2098
+    },
+    {
+      "epoch": 0.18754886411865884,
+      "grad_norm": 0.3898516893386841,
+      "learning_rate": 9.36314275287376e-05,
+      "loss": 1.0821,
+      "step": 2099
+    },
+    {
+      "epoch": 0.1876382156499207,
+      "grad_norm": 0.5555103421211243,
+      "learning_rate": 9.362435842777246e-05,
+      "loss": 0.8719,
+      "step": 2100
+    },
+    {
+      "epoch": 0.18772756718118258,
+      "grad_norm": 0.38932761549949646,
+      "learning_rate": 9.36172856728023e-05,
+      "loss": 1.0611,
+      "step": 2101
+    },
+    {
+      "epoch": 0.18781691871244444,
+      "grad_norm": 0.411034494638443,
+      "learning_rate": 9.361020926441955e-05,
+      "loss": 0.9853,
+      "step": 2102
+    },
+    {
+      "epoch": 0.1879062702437063,
+      "grad_norm": 0.47794967889785767,
+      "learning_rate": 9.36031292032169e-05,
+      "loss": 1.0755,
+      "step": 2103
+    },
+    {
+      "epoch": 0.18799562177496817,
+      "grad_norm": 0.40055814385414124,
+      "learning_rate": 9.359604548978742e-05,
+      "loss": 1.072,
+      "step": 2104
+    },
+    {
+      "epoch": 0.18808497330623003,
+      "grad_norm": 0.4187031686306,
+      "learning_rate": 9.358895812472442e-05,
+      "loss": 1.0448,
+      "step": 2105
+    },
+    {
+      "epoch": 0.1881743248374919,
+      "grad_norm": 0.39350834488868713,
+      "learning_rate": 9.358186710862156e-05,
+      "loss": 1.0356,
+      "step": 2106
+    },
+    {
+      "epoch": 0.18826367636875377,
+      "grad_norm": 0.36614513397216797,
+      "learning_rate": 9.357477244207278e-05,
+      "loss": 1.0865,
+      "step": 2107
+    },
+    {
+      "epoch": 0.18835302790001562,
+      "grad_norm": 0.4072873890399933,
+      "learning_rate": 9.356767412567234e-05,
+      "loss": 1.043,
+      "step": 2108
+    },
+    {
+      "epoch": 0.1884423794312775,
+      "grad_norm": 0.3540877103805542,
+      "learning_rate": 9.356057216001477e-05,
+      "loss": 1.0828,
+      "step": 2109
+    },
+    {
+      "epoch": 0.18853173096253936,
+      "grad_norm": 0.41696932911872864,
+      "learning_rate": 9.355346654569497e-05,
+      "loss": 1.0219,
+      "step": 2110
+    },
+    {
+      "epoch": 0.18862108249380125,
+      "grad_norm": 0.5428374409675598,
+      "learning_rate": 9.354635728330811e-05,
+      "loss": 1.0378,
+      "step": 2111
+    },
+    {
+      "epoch": 0.1887104340250631,
+      "grad_norm": 0.3925424814224243,
+      "learning_rate": 9.353924437344966e-05,
+      "loss": 1.0572,
+      "step": 2112
+    },
+    {
+      "epoch": 0.18879978555632498,
+      "grad_norm": 0.3727796971797943,
+      "learning_rate": 9.353212781671538e-05,
+      "loss": 1.0736,
+      "step": 2113
+    },
+    {
+      "epoch": 0.18888913708758684,
+      "grad_norm": 0.4229578971862793,
+      "learning_rate": 9.352500761370139e-05,
+      "loss": 1.0785,
+      "step": 2114
+    },
+    {
+      "epoch": 0.1889784886188487,
+      "grad_norm": 0.4206151068210602,
+      "learning_rate": 9.351788376500406e-05,
+      "loss": 1.0187,
+      "step": 2115
+    },
+    {
+      "epoch": 0.18906784015011058,
+      "grad_norm": 0.4051399230957031,
+      "learning_rate": 9.35107562712201e-05,
+      "loss": 1.0228,
+      "step": 2116
+    },
+    {
+      "epoch": 0.18915719168137243,
+      "grad_norm": 0.3675979971885681,
+      "learning_rate": 9.350362513294651e-05,
+      "loss": 1.024,
+      "step": 2117
+    },
+    {
+      "epoch": 0.18924654321263432,
+      "grad_norm": 0.43674102425575256,
+      "learning_rate": 9.349649035078061e-05,
+      "loss": 1.0003,
+      "step": 2118
+    },
+    {
+      "epoch": 0.18933589474389617,
+      "grad_norm": 0.4083952009677887,
+      "learning_rate": 9.348935192532001e-05,
+      "loss": 1.0628,
+      "step": 2119
+    },
+    {
+      "epoch": 0.18942524627515803,
+      "grad_norm": 0.4974057972431183,
+      "learning_rate": 9.34822098571626e-05,
+      "loss": 1.0283,
+      "step": 2120
+    },
+    {
+      "epoch": 0.1895145978064199,
+      "grad_norm": 0.39554762840270996,
+      "learning_rate": 9.347506414690663e-05,
+      "loss": 1.0382,
+      "step": 2121
+    },
+    {
+      "epoch": 0.18960394933768177,
+      "grad_norm": 0.40948179364204407,
+      "learning_rate": 9.346791479515063e-05,
+      "loss": 1.0227,
+      "step": 2122
+    },
+    {
+      "epoch": 0.18969330086894365,
+      "grad_norm": 0.42662203311920166,
+      "learning_rate": 9.346076180249345e-05,
+      "loss": 1.0179,
+      "step": 2123
+    },
+    {
+      "epoch": 0.1897826524002055,
+      "grad_norm": 0.40149882435798645,
+      "learning_rate": 9.345360516953418e-05,
+      "loss": 1.0281,
+      "step": 2124
+    },
+    {
+      "epoch": 0.1898720039314674,
+      "grad_norm": 0.49369674921035767,
+      "learning_rate": 9.344644489687233e-05,
+      "loss": 0.9582,
+      "step": 2125
+    },
+    {
+      "epoch": 0.18996135546272924,
+      "grad_norm": 0.43468207120895386,
+      "learning_rate": 9.343928098510759e-05,
+      "loss": 1.0257,
+      "step": 2126
+    },
+    {
+      "epoch": 0.1900507069939911,
+      "grad_norm": 0.5939388275146484,
+      "learning_rate": 9.343211343484004e-05,
+      "loss": 1.0418,
+      "step": 2127
+    },
+    {
+      "epoch": 0.19014005852525298,
+      "grad_norm": 0.42721638083457947,
+      "learning_rate": 9.342494224667004e-05,
+      "loss": 0.9871,
+      "step": 2128
+    },
+    {
+      "epoch": 0.19022941005651484,
+      "grad_norm": 0.4589712917804718,
+      "learning_rate": 9.341776742119825e-05,
+      "loss": 0.9894,
+      "step": 2129
+    },
+    {
+      "epoch": 0.19031876158777672,
+      "grad_norm": 0.3878852128982544,
+      "learning_rate": 9.341058895902563e-05,
+      "loss": 1.0514,
+      "step": 2130
+    },
+    {
+      "epoch": 0.19040811311903857,
+      "grad_norm": 0.47722190618515015,
+      "learning_rate": 9.340340686075349e-05,
+      "loss": 0.9354,
+      "step": 2131
+    },
+    {
+      "epoch": 0.19049746465030046,
+      "grad_norm": 0.4715014398097992,
+      "learning_rate": 9.339622112698334e-05,
+      "loss": 1.0487,
+      "step": 2132
+    },
+    {
+      "epoch": 0.1905868161815623,
+      "grad_norm": 0.44013863801956177,
+      "learning_rate": 9.338903175831712e-05,
+      "loss": 1.0149,
+      "step": 2133
+    },
+    {
+      "epoch": 0.19067616771282417,
+      "grad_norm": 0.4588448107242584,
+      "learning_rate": 9.3381838755357e-05,
+      "loss": 1.0304,
+      "step": 2134
+    },
+    {
+      "epoch": 0.19076551924408605,
+      "grad_norm": 0.47280123829841614,
+      "learning_rate": 9.337464211870546e-05,
+      "loss": 1.0329,
+      "step": 2135
+    },
+    {
+      "epoch": 0.1908548707753479,
+      "grad_norm": 0.4202229380607605,
+      "learning_rate": 9.336744184896531e-05,
+      "loss": 1.1011,
+      "step": 2136
+    },
+    {
+      "epoch": 0.1909442223066098,
+      "grad_norm": 0.3864685297012329,
+      "learning_rate": 9.336023794673962e-05,
+      "loss": 1.0398,
+      "step": 2137
+    },
+    {
+      "epoch": 0.19103357383787165,
+      "grad_norm": 0.3656110465526581,
+      "learning_rate": 9.335303041263183e-05,
+      "loss": 0.9896,
+      "step": 2138
+    },
+    {
+      "epoch": 0.1911229253691335,
+      "grad_norm": 0.3890793025493622,
+      "learning_rate": 9.334581924724564e-05,
+      "loss": 1.0371,
+      "step": 2139
+    },
+    {
+      "epoch": 0.19121227690039538,
+      "grad_norm": 0.4095704257488251,
+      "learning_rate": 9.333860445118505e-05,
+      "loss": 1.0538,
+      "step": 2140
+    },
+    {
+      "epoch": 0.19130162843165724,
+      "grad_norm": 0.3988470733165741,
+      "learning_rate": 9.333138602505437e-05,
+      "loss": 0.9512,
+      "step": 2141
+    },
+    {
+      "epoch": 0.19139097996291912,
+      "grad_norm": 0.42787396907806396,
+      "learning_rate": 9.332416396945824e-05,
+      "loss": 1.0253,
+      "step": 2142
+    },
+    {
+      "epoch": 0.19148033149418098,
+      "grad_norm": 0.44595029950141907,
+      "learning_rate": 9.331693828500159e-05,
+      "loss": 1.005,
+      "step": 2143
+    },
+    {
+      "epoch": 0.19156968302544286,
+      "grad_norm": 0.44697070121765137,
+      "learning_rate": 9.33097089722896e-05,
+      "loss": 0.9655,
+      "step": 2144
+    },
+    {
+      "epoch": 0.19165903455670472,
+      "grad_norm": 0.45841777324676514,
+      "learning_rate": 9.330247603192786e-05,
+      "loss": 1.0242,
+      "step": 2145
+    },
+    {
+      "epoch": 0.19174838608796657,
+      "grad_norm": 0.4379693865776062,
+      "learning_rate": 9.32952394645222e-05,
+      "loss": 1.0129,
+      "step": 2146
+    },
+    {
+      "epoch": 0.19183773761922845,
+      "grad_norm": 0.3796353042125702,
+      "learning_rate": 9.32879992706787e-05,
+      "loss": 1.0516,
+      "step": 2147
+    },
+    {
+      "epoch": 0.1919270891504903,
+      "grad_norm": 0.5354682207107544,
+      "learning_rate": 9.328075545100385e-05,
+      "loss": 0.85,
+      "step": 2148
+    },
+    {
+      "epoch": 0.1920164406817522,
+      "grad_norm": 0.5834673643112183,
+      "learning_rate": 9.32735080061044e-05,
+      "loss": 1.0113,
+      "step": 2149
+    },
+    {
+      "epoch": 0.19210579221301405,
+      "grad_norm": 0.46997979283332825,
+      "learning_rate": 9.326625693658739e-05,
+      "loss": 1.0355,
+      "step": 2150
+    },
+    {
+      "epoch": 0.19219514374427593,
+      "grad_norm": 0.47971826791763306,
+      "learning_rate": 9.325900224306019e-05,
+      "loss": 0.9207,
+      "step": 2151
+    },
+    {
+      "epoch": 0.1922844952755378,
+      "grad_norm": 0.49349215626716614,
+      "learning_rate": 9.325174392613043e-05,
+      "loss": 0.9519,
+      "step": 2152
+    },
+    {
+      "epoch": 0.19237384680679964,
+      "grad_norm": 0.603679895401001,
+      "learning_rate": 9.32444819864061e-05,
+      "loss": 0.9713,
+      "step": 2153
+    },
+    {
+      "epoch": 0.19246319833806153,
+      "grad_norm": 0.44981256127357483,
+      "learning_rate": 9.323721642449543e-05,
+      "loss": 0.9577,
+      "step": 2154
+    },
+    {
+      "epoch": 0.19255254986932338,
+      "grad_norm": 0.3963879644870758,
+      "learning_rate": 9.322994724100702e-05,
+      "loss": 1.0324,
+      "step": 2155
+    },
+    {
+      "epoch": 0.19264190140058526,
+      "grad_norm": 0.4557402729988098,
+      "learning_rate": 9.322267443654972e-05,
+      "loss": 1.042,
+      "step": 2156
+    },
+    {
+      "epoch": 0.19273125293184712,
+      "grad_norm": 0.4886689782142639,
+      "learning_rate": 9.321539801173274e-05,
+      "loss": 1.01,
+      "step": 2157
+    },
+    {
+      "epoch": 0.19282060446310897,
+      "grad_norm": 0.416227787733078,
+      "learning_rate": 9.320811796716552e-05,
+      "loss": 0.9619,
+      "step": 2158
+    },
+    {
+      "epoch": 0.19290995599437086,
+      "grad_norm": 0.44444751739501953,
+      "learning_rate": 9.320083430345785e-05,
+      "loss": 1.0472,
+      "step": 2159
+    },
+    {
+      "epoch": 0.1929993075256327,
+      "grad_norm": 0.4092031717300415,
+      "learning_rate": 9.319354702121981e-05,
+      "loss": 1.0692,
+      "step": 2160
+    },
+    {
+      "epoch": 0.1930886590568946,
+      "grad_norm": 0.3944936692714691,
+      "learning_rate": 9.318625612106182e-05,
+      "loss": 1.0244,
+      "step": 2161
+    },
+    {
+      "epoch": 0.19317801058815645,
+      "grad_norm": 0.457409530878067,
+      "learning_rate": 9.317896160359454e-05,
+      "loss": 1.1001,
+      "step": 2162
+    },
+    {
+      "epoch": 0.19326736211941833,
+      "grad_norm": 0.4606925845146179,
+      "learning_rate": 9.317166346942897e-05,
+      "loss": 1.0092,
+      "step": 2163
+    },
+    {
+      "epoch": 0.1933567136506802,
+      "grad_norm": 0.458124041557312,
+      "learning_rate": 9.31643617191764e-05,
+      "loss": 1.0144,
+      "step": 2164
+    },
+    {
+      "epoch": 0.19344606518194205,
+      "grad_norm": 0.3879139721393585,
+      "learning_rate": 9.315705635344844e-05,
+      "loss": 0.9889,
+      "step": 2165
+    },
+    {
+      "epoch": 0.19353541671320393,
+      "grad_norm": 0.3774491548538208,
+      "learning_rate": 9.314974737285699e-05,
+      "loss": 1.0031,
+      "step": 2166
+    },
+    {
+      "epoch": 0.19362476824446578,
+      "grad_norm": 0.38061612844467163,
+      "learning_rate": 9.314243477801425e-05,
+      "loss": 0.9962,
+      "step": 2167
+    },
+    {
+      "epoch": 0.19371411977572767,
+      "grad_norm": 0.4952597916126251,
+      "learning_rate": 9.313511856953274e-05,
+      "loss": 1.0836,
+      "step": 2168
+    },
+    {
+      "epoch": 0.19380347130698952,
+      "grad_norm": 0.41984468698501587,
+      "learning_rate": 9.312779874802526e-05,
+      "loss": 1.0027,
+      "step": 2169
+    },
+    {
+      "epoch": 0.19389282283825138,
+      "grad_norm": 0.47165659070014954,
+      "learning_rate": 9.312047531410493e-05,
+      "loss": 1.0573,
+      "step": 2170
+    },
+    {
+      "epoch": 0.19398217436951326,
+      "grad_norm": 0.39716780185699463,
+      "learning_rate": 9.311314826838515e-05,
+      "loss": 1.0426,
+      "step": 2171
+    },
+    {
+      "epoch": 0.19407152590077512,
+      "grad_norm": 0.4184114336967468,
+      "learning_rate": 9.310581761147966e-05,
+      "loss": 1.0309,
+      "step": 2172
+    },
+    {
+      "epoch": 0.194160877432037,
+      "grad_norm": 0.4498063027858734,
+      "learning_rate": 9.309848334400246e-05,
+      "loss": 0.9592,
+      "step": 2173
+    },
+    {
+      "epoch": 0.19425022896329885,
+      "grad_norm": 0.4691358208656311,
+      "learning_rate": 9.30911454665679e-05,
+      "loss": 1.0557,
+      "step": 2174
+    },
+    {
+      "epoch": 0.19433958049456074,
+      "grad_norm": 0.4362775981426239,
+      "learning_rate": 9.308380397979057e-05,
+      "loss": 1.1005,
+      "step": 2175
+    },
+    {
+      "epoch": 0.1944289320258226,
+      "grad_norm": 0.39508792757987976,
+      "learning_rate": 9.307645888428542e-05,
+      "loss": 1.0745,
+      "step": 2176
+    },
+    {
+      "epoch": 0.19451828355708445,
+      "grad_norm": 0.4104553461074829,
+      "learning_rate": 9.306911018066769e-05,
+      "loss": 1.0801,
+      "step": 2177
+    },
+    {
+      "epoch": 0.19460763508834633,
+      "grad_norm": 0.4288639426231384,
+      "learning_rate": 9.306175786955289e-05,
+      "loss": 1.1468,
+      "step": 2178
+    },
+    {
+      "epoch": 0.1946969866196082,
+      "grad_norm": 0.4734762907028198,
+      "learning_rate": 9.305440195155686e-05,
+      "loss": 1.0742,
+      "step": 2179
+    },
+    {
+      "epoch": 0.19478633815087007,
+      "grad_norm": 0.45509761571884155,
+      "learning_rate": 9.304704242729575e-05,
+      "loss": 0.9943,
+      "step": 2180
+    },
+    {
+      "epoch": 0.19487568968213193,
+      "grad_norm": 0.412614643573761,
+      "learning_rate": 9.303967929738598e-05,
+      "loss": 1.0259,
+      "step": 2181
+    },
+    {
+      "epoch": 0.1949650412133938,
+      "grad_norm": 0.38879382610321045,
+      "learning_rate": 9.30323125624443e-05,
+      "loss": 1.0245,
+      "step": 2182
+    },
+    {
+      "epoch": 0.19505439274465566,
+      "grad_norm": 0.459376722574234,
+      "learning_rate": 9.302494222308774e-05,
+      "loss": 1.0003,
+      "step": 2183
+    },
+    {
+      "epoch": 0.19514374427591752,
+      "grad_norm": 0.41094890236854553,
+      "learning_rate": 9.301756827993367e-05,
+      "loss": 0.9735,
+      "step": 2184
+    },
+    {
+      "epoch": 0.1952330958071794,
+      "grad_norm": 0.4334050118923187,
+      "learning_rate": 9.301019073359972e-05,
+      "loss": 0.9722,
+      "step": 2185
+    },
+    {
+      "epoch": 0.19532244733844126,
+      "grad_norm": 0.43469616770744324,
+      "learning_rate": 9.300280958470384e-05,
+      "loss": 1.0839,
+      "step": 2186
+    },
+    {
+      "epoch": 0.19541179886970314,
+      "grad_norm": 0.45029255747795105,
+      "learning_rate": 9.299542483386428e-05,
+      "loss": 0.9987,
+      "step": 2187
+    },
+    {
+      "epoch": 0.195501150400965,
+      "grad_norm": 0.42933982610702515,
+      "learning_rate": 9.298803648169958e-05,
+      "loss": 1.0202,
+      "step": 2188
+    },
+    {
+      "epoch": 0.19559050193222685,
+      "grad_norm": 0.37860357761383057,
+      "learning_rate": 9.298064452882862e-05,
+      "loss": 1.053,
+      "step": 2189
+    },
+    {
+      "epoch": 0.19567985346348873,
+      "grad_norm": 0.43233370780944824,
+      "learning_rate": 9.297324897587054e-05,
+      "loss": 1.07,
+      "step": 2190
+    },
+    {
+      "epoch": 0.1957692049947506,
+      "grad_norm": 0.4745257794857025,
+      "learning_rate": 9.296584982344478e-05,
+      "loss": 0.951,
+      "step": 2191
+    },
+    {
+      "epoch": 0.19585855652601247,
+      "grad_norm": 0.453050434589386,
+      "learning_rate": 9.295844707217114e-05,
+      "loss": 0.9839,
+      "step": 2192
+    },
+    {
+      "epoch": 0.19594790805727433,
+      "grad_norm": 0.4505593776702881,
+      "learning_rate": 9.295104072266965e-05,
+      "loss": 1.0556,
+      "step": 2193
+    },
+    {
+      "epoch": 0.1960372595885362,
+      "grad_norm": 0.41230955719947815,
+      "learning_rate": 9.294363077556066e-05,
+      "loss": 1.0569,
+      "step": 2194
+    },
+    {
+      "epoch": 0.19612661111979807,
+      "grad_norm": 0.38011908531188965,
+      "learning_rate": 9.293621723146485e-05,
+      "loss": 1.0311,
+      "step": 2195
+    },
+    {
+      "epoch": 0.19621596265105992,
+      "grad_norm": 0.37792640924453735,
+      "learning_rate": 9.292880009100318e-05,
+      "loss": 1.0998,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1963053141823218,
+      "grad_norm": 0.42640724778175354,
+      "learning_rate": 9.292137935479692e-05,
+      "loss": 1.0311,
+      "step": 2197
+    },
+    {
+      "epoch": 0.19639466571358366,
+      "grad_norm": 0.39782777428627014,
+      "learning_rate": 9.291395502346763e-05,
+      "loss": 1.035,
+      "step": 2198
+    },
+    {
+      "epoch": 0.19648401724484554,
+      "grad_norm": 0.39150485396385193,
+      "learning_rate": 9.290652709763717e-05,
+      "loss": 1.0524,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1965733687761074,
+      "grad_norm": 0.41855573654174805,
+      "learning_rate": 9.289909557792771e-05,
+      "loss": 1.05,
+      "step": 2200
+    },
+    {
+      "epoch": 0.19666272030736925,
+      "grad_norm": 0.4688778817653656,
+      "learning_rate": 9.289166046496172e-05,
+      "loss": 1.0219,
+      "step": 2201
+    },
+    {
+      "epoch": 0.19675207183863114,
+      "grad_norm": 0.4490063488483429,
+      "learning_rate": 9.2884221759362e-05,
+      "loss": 0.9859,
+      "step": 2202
+    },
+    {
+      "epoch": 0.196841423369893,
+      "grad_norm": 0.37221378087997437,
+      "learning_rate": 9.287677946175157e-05,
+      "loss": 1.0359,
+      "step": 2203
+    },
+    {
+      "epoch": 0.19693077490115488,
+      "grad_norm": 0.5122733116149902,
+      "learning_rate": 9.286933357275385e-05,
+      "loss": 0.9099,
+      "step": 2204
+    },
+    {
+      "epoch": 0.19702012643241673,
+      "grad_norm": 0.48560839891433716,
+      "learning_rate": 9.286188409299246e-05,
+      "loss": 0.9573,
+      "step": 2205
+    },
+    {
+      "epoch": 0.19710947796367861,
+      "grad_norm": 0.42828473448753357,
+      "learning_rate": 9.285443102309142e-05,
+      "loss": 1.0508,
+      "step": 2206
+    },
+    {
+      "epoch": 0.19719882949494047,
+      "grad_norm": 0.417341947555542,
+      "learning_rate": 9.284697436367497e-05,
+      "loss": 1.085,
+      "step": 2207
+    },
+    {
+      "epoch": 0.19728818102620232,
+      "grad_norm": 0.4165228307247162,
+      "learning_rate": 9.283951411536773e-05,
+      "loss": 1.1126,
+      "step": 2208
+    },
+    {
+      "epoch": 0.1973775325574642,
+      "grad_norm": 0.4660586714744568,
+      "learning_rate": 9.283205027879454e-05,
+      "loss": 0.9542,
+      "step": 2209
+    },
+    {
+      "epoch": 0.19746688408872606,
+      "grad_norm": 0.41941720247268677,
+      "learning_rate": 9.282458285458055e-05,
+      "loss": 1.0903,
+      "step": 2210
+    },
+    {
+      "epoch": 0.19755623561998795,
+      "grad_norm": 0.4800897538661957,
+      "learning_rate": 9.281711184335131e-05,
+      "loss": 0.9575,
+      "step": 2211
+    },
+    {
+      "epoch": 0.1976455871512498,
+      "grad_norm": 0.39509686827659607,
+      "learning_rate": 9.280963724573253e-05,
+      "loss": 1.0522,
+      "step": 2212
+    },
+    {
+      "epoch": 0.19773493868251169,
+      "grad_norm": 0.4314636290073395,
+      "learning_rate": 9.280215906235032e-05,
+      "loss": 1.0846,
+      "step": 2213
+    },
+    {
+      "epoch": 0.19782429021377354,
+      "grad_norm": 0.39220911264419556,
+      "learning_rate": 9.279467729383105e-05,
+      "loss": 0.9884,
+      "step": 2214
+    },
+    {
+      "epoch": 0.1979136417450354,
+      "grad_norm": 0.45912548899650574,
+      "learning_rate": 9.27871919408014e-05,
+      "loss": 0.984,
+      "step": 2215
+    },
+    {
+      "epoch": 0.19800299327629728,
+      "grad_norm": 0.48036208748817444,
+      "learning_rate": 9.277970300388834e-05,
+      "loss": 0.9564,
+      "step": 2216
+    },
+    {
+      "epoch": 0.19809234480755913,
+      "grad_norm": 0.5072821974754333,
+      "learning_rate": 9.277221048371917e-05,
+      "loss": 0.911,
+      "step": 2217
+    },
+    {
+      "epoch": 0.19818169633882102,
+      "grad_norm": 0.48071086406707764,
+      "learning_rate": 9.276471438092145e-05,
+      "loss": 1.0484,
+      "step": 2218
+    },
+    {
+      "epoch": 0.19827104787008287,
+      "grad_norm": 0.5344045758247375,
+      "learning_rate": 9.275721469612304e-05,
+      "loss": 0.8887,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19836039940134473,
+      "grad_norm": 0.43411949276924133,
+      "learning_rate": 9.274971142995216e-05,
+      "loss": 1.0001,
+      "step": 2220
+    },
+    {
+      "epoch": 0.1984497509326066,
+      "grad_norm": 0.4209814667701721,
+      "learning_rate": 9.274220458303727e-05,
+      "loss": 1.0129,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19853910246386847,
+      "grad_norm": 0.33118653297424316,
+      "learning_rate": 9.273469415600713e-05,
+      "loss": 1.0303,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19862845399513035,
+      "grad_norm": 0.4255508780479431,
+      "learning_rate": 9.272718014949085e-05,
+      "loss": 0.9976,
+      "step": 2223
+    },
+    {
+      "epoch": 0.1987178055263922,
+      "grad_norm": 0.4046938121318817,
+      "learning_rate": 9.271966256411779e-05,
+      "loss": 1.0221,
+      "step": 2224
+    },
+    {
+      "epoch": 0.1988071570576541,
+      "grad_norm": 0.4717459976673126,
+      "learning_rate": 9.271214140051763e-05,
+      "loss": 0.9862,
+      "step": 2225
+    },
+    {
+      "epoch": 0.19889650858891594,
+      "grad_norm": 0.4104228615760803,
+      "learning_rate": 9.270461665932034e-05,
+      "loss": 1.024,
+      "step": 2226
+    },
+    {
+      "epoch": 0.1989858601201778,
+      "grad_norm": 0.41744041442871094,
+      "learning_rate": 9.269708834115622e-05,
+      "loss": 1.0642,
+      "step": 2227
+    },
+    {
+      "epoch": 0.19907521165143968,
+      "grad_norm": 0.4957825541496277,
+      "learning_rate": 9.268955644665582e-05,
+      "loss": 0.978,
+      "step": 2228
+    },
+    {
+      "epoch": 0.19916456318270154,
+      "grad_norm": 0.4245834946632385,
+      "learning_rate": 9.268202097645005e-05,
+      "loss": 0.9519,
+      "step": 2229
+    },
+    {
+      "epoch": 0.19925391471396342,
+      "grad_norm": 0.4112485647201538,
+      "learning_rate": 9.267448193117005e-05,
+      "loss": 1.0033,
+      "step": 2230
+    },
+    {
+      "epoch": 0.19934326624522528,
+      "grad_norm": 0.45706307888031006,
+      "learning_rate": 9.266693931144732e-05,
+      "loss": 1.0904,
+      "step": 2231
+    },
+    {
+      "epoch": 0.19943261777648713,
+      "grad_norm": 0.44358548521995544,
+      "learning_rate": 9.265939311791362e-05,
+      "loss": 0.9685,
+      "step": 2232
+    },
+    {
+      "epoch": 0.19952196930774901,
+      "grad_norm": 0.4372340440750122,
+      "learning_rate": 9.265184335120103e-05,
+      "loss": 1.0331,
+      "step": 2233
+    },
+    {
+      "epoch": 0.19961132083901087,
+      "grad_norm": 0.5012243986129761,
+      "learning_rate": 9.264429001194193e-05,
+      "loss": 0.9479,
+      "step": 2234
+    },
+    {
+      "epoch": 0.19970067237027275,
+      "grad_norm": 0.45009467005729675,
+      "learning_rate": 9.263673310076897e-05,
+      "loss": 0.9737,
+      "step": 2235
+    },
+    {
+      "epoch": 0.1997900239015346,
+      "grad_norm": 0.3963136076927185,
+      "learning_rate": 9.262917261831515e-05,
+      "loss": 1.0775,
+      "step": 2236
+    },
+    {
+      "epoch": 0.1998793754327965,
+      "grad_norm": 0.4253182113170624,
+      "learning_rate": 9.262160856521372e-05,
+      "loss": 1.013,
+      "step": 2237
+    },
+    {
+      "epoch": 0.19996872696405835,
+      "grad_norm": 0.40562012791633606,
+      "learning_rate": 9.261404094209827e-05,
+      "loss": 1.0651,
+      "step": 2238
+    },
+    {
+      "epoch": 0.2000580784953202,
+      "grad_norm": 0.42175614833831787,
+      "learning_rate": 9.260646974960265e-05,
+      "loss": 1.052,
+      "step": 2239
+    },
+    {
+      "epoch": 0.20014743002658208,
+      "grad_norm": 0.3817167580127716,
+      "learning_rate": 9.259889498836105e-05,
+      "loss": 1.0804,
+      "step": 2240
+    },
+    {
+      "epoch": 0.20023678155784394,
+      "grad_norm": 0.42921021580696106,
+      "learning_rate": 9.259131665900792e-05,
+      "loss": 0.9756,
+      "step": 2241
+    },
+    {
+      "epoch": 0.20032613308910582,
+      "grad_norm": 0.4635266363620758,
+      "learning_rate": 9.258373476217801e-05,
+      "loss": 0.9926,
+      "step": 2242
+    },
+    {
+      "epoch": 0.20041548462036768,
+      "grad_norm": 0.45946842432022095,
+      "learning_rate": 9.257614929850642e-05,
+      "loss": 0.9873,
+      "step": 2243
+    },
+    {
+      "epoch": 0.20050483615162956,
+      "grad_norm": 0.4113311767578125,
+      "learning_rate": 9.256856026862847e-05,
+      "loss": 1.0364,
+      "step": 2244
+    },
+    {
+      "epoch": 0.20059418768289142,
+      "grad_norm": 0.38544759154319763,
+      "learning_rate": 9.256096767317989e-05,
+      "loss": 1.0698,
+      "step": 2245
+    },
+    {
+      "epoch": 0.20068353921415327,
+      "grad_norm": 0.483548104763031,
+      "learning_rate": 9.255337151279658e-05,
+      "loss": 0.9635,
+      "step": 2246
+    },
+    {
+      "epoch": 0.20077289074541516,
+      "grad_norm": 0.41710737347602844,
+      "learning_rate": 9.254577178811482e-05,
+      "loss": 0.9602,
+      "step": 2247
+    },
+    {
+      "epoch": 0.200862242276677,
+      "grad_norm": 0.41707658767700195,
+      "learning_rate": 9.253816849977117e-05,
+      "loss": 1.0529,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2009515938079389,
+      "grad_norm": 0.49464061856269836,
+      "learning_rate": 9.253056164840248e-05,
+      "loss": 1.0603,
+      "step": 2249
+    },
+    {
+      "epoch": 0.20104094533920075,
+      "grad_norm": 0.4921278655529022,
+      "learning_rate": 9.252295123464592e-05,
+      "loss": 1.0333,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2011302968704626,
+      "grad_norm": 0.41099363565444946,
+      "learning_rate": 9.251533725913893e-05,
+      "loss": 0.9884,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2012196484017245,
+      "grad_norm": 0.40399447083473206,
+      "learning_rate": 9.250771972251925e-05,
+      "loss": 0.9832,
+      "step": 2252
+    },
+    {
+      "epoch": 0.20130899993298634,
+      "grad_norm": 0.39591044187545776,
+      "learning_rate": 9.250009862542495e-05,
+      "loss": 1.0229,
+      "step": 2253
+    },
+    {
+      "epoch": 0.20139835146424823,
+      "grad_norm": 0.4617488980293274,
+      "learning_rate": 9.249247396849437e-05,
+      "loss": 0.9989,
+      "step": 2254
+    },
+    {
+      "epoch": 0.20148770299551008,
+      "grad_norm": 0.3816708028316498,
+      "learning_rate": 9.248484575236616e-05,
+      "loss": 1.1098,
+      "step": 2255
+    },
+    {
+      "epoch": 0.20157705452677196,
+      "grad_norm": 0.43088236451148987,
+      "learning_rate": 9.247721397767926e-05,
+      "loss": 0.9852,
+      "step": 2256
+    },
+    {
+      "epoch": 0.20166640605803382,
+      "grad_norm": 0.39949992299079895,
+      "learning_rate": 9.246957864507292e-05,
+      "loss": 1.0444,
+      "step": 2257
+    },
+    {
+      "epoch": 0.20175575758929568,
+      "grad_norm": 0.5272142291069031,
+      "learning_rate": 9.246193975518667e-05,
+      "loss": 0.9864,
+      "step": 2258
+    },
+    {
+      "epoch": 0.20184510912055756,
+      "grad_norm": 0.44182446599006653,
+      "learning_rate": 9.245429730866035e-05,
+      "loss": 0.9412,
+      "step": 2259
+    },
+    {
+      "epoch": 0.20193446065181941,
+      "grad_norm": 0.4208958148956299,
+      "learning_rate": 9.244665130613411e-05,
+      "loss": 1.0687,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2020238121830813,
+      "grad_norm": 0.5523428916931152,
+      "learning_rate": 9.243900174824838e-05,
+      "loss": 0.9329,
+      "step": 2261
+    },
+    {
+      "epoch": 0.20211316371434315,
+      "grad_norm": 0.3945762813091278,
+      "learning_rate": 9.243134863564387e-05,
+      "loss": 1.0095,
+      "step": 2262
+    },
+    {
+      "epoch": 0.202202515245605,
+      "grad_norm": 0.5793361067771912,
+      "learning_rate": 9.242369196896163e-05,
+      "loss": 0.9279,
+      "step": 2263
+    },
+    {
+      "epoch": 0.2022918667768669,
+      "grad_norm": 0.4429938495159149,
+      "learning_rate": 9.241603174884299e-05,
+      "loss": 1.0672,
+      "step": 2264
+    },
+    {
+      "epoch": 0.20238121830812875,
+      "grad_norm": 0.4614643454551697,
+      "learning_rate": 9.240836797592958e-05,
+      "loss": 1.0062,
+      "step": 2265
+    },
+    {
+      "epoch": 0.20247056983939063,
+      "grad_norm": 0.4438598155975342,
+      "learning_rate": 9.240070065086328e-05,
+      "loss": 0.9787,
+      "step": 2266
+    },
+    {
+      "epoch": 0.20255992137065248,
+      "grad_norm": 0.3669438362121582,
+      "learning_rate": 9.239302977428637e-05,
+      "loss": 1.0367,
+      "step": 2267
+    },
+    {
+      "epoch": 0.20264927290191437,
+      "grad_norm": 0.43813052773475647,
+      "learning_rate": 9.238535534684131e-05,
+      "loss": 0.9767,
+      "step": 2268
+    },
+    {
+      "epoch": 0.20273862443317622,
+      "grad_norm": 0.4687936007976532,
+      "learning_rate": 9.237767736917098e-05,
+      "loss": 1.0335,
+      "step": 2269
+    },
+    {
+      "epoch": 0.20282797596443808,
+      "grad_norm": 0.3983701467514038,
+      "learning_rate": 9.236999584191843e-05,
+      "loss": 1.0086,
+      "step": 2270
+    },
+    {
+      "epoch": 0.20291732749569996,
+      "grad_norm": 0.4753643870353699,
+      "learning_rate": 9.23623107657271e-05,
+      "loss": 0.9726,
+      "step": 2271
+    },
+    {
+      "epoch": 0.20300667902696182,
+      "grad_norm": 0.4003112316131592,
+      "learning_rate": 9.235462214124071e-05,
+      "loss": 1.0838,
+      "step": 2272
+    },
+    {
+      "epoch": 0.2030960305582237,
+      "grad_norm": 0.4177161157131195,
+      "learning_rate": 9.234692996910323e-05,
+      "loss": 1.0637,
+      "step": 2273
+    },
+    {
+      "epoch": 0.20318538208948556,
+      "grad_norm": 0.48143237829208374,
+      "learning_rate": 9.233923424995899e-05,
+      "loss": 0.9929,
+      "step": 2274
+    },
+    {
+      "epoch": 0.20327473362074744,
+      "grad_norm": 0.4232881963253021,
+      "learning_rate": 9.233153498445258e-05,
+      "loss": 1.0583,
+      "step": 2275
+    },
+    {
+      "epoch": 0.2033640851520093,
+      "grad_norm": 0.47992992401123047,
+      "learning_rate": 9.232383217322889e-05,
+      "loss": 0.9638,
+      "step": 2276
+    },
+    {
+      "epoch": 0.20345343668327115,
+      "grad_norm": 0.4345097839832306,
+      "learning_rate": 9.23161258169331e-05,
+      "loss": 1.0855,
+      "step": 2277
+    },
+    {
+      "epoch": 0.20354278821453303,
+      "grad_norm": 0.45901912450790405,
+      "learning_rate": 9.230841591621073e-05,
+      "loss": 1.0315,
+      "step": 2278
+    },
+    {
+      "epoch": 0.2036321397457949,
+      "grad_norm": 0.45470988750457764,
+      "learning_rate": 9.230070247170755e-05,
+      "loss": 1.0359,
+      "step": 2279
+    },
+    {
+      "epoch": 0.20372149127705677,
+      "grad_norm": 0.4291987419128418,
+      "learning_rate": 9.229298548406964e-05,
+      "loss": 1.0296,
+      "step": 2280
+    },
+    {
+      "epoch": 0.20381084280831863,
+      "grad_norm": 0.3951917588710785,
+      "learning_rate": 9.228526495394339e-05,
+      "loss": 0.9465,
+      "step": 2281
+    },
+    {
+      "epoch": 0.20390019433958048,
+      "grad_norm": 0.3695099353790283,
+      "learning_rate": 9.227754088197548e-05,
+      "loss": 1.03,
+      "step": 2282
+    },
+    {
+      "epoch": 0.20398954587084236,
+      "grad_norm": 0.40235844254493713,
+      "learning_rate": 9.226981326881286e-05,
+      "loss": 1.0398,
+      "step": 2283
+    },
+    {
+      "epoch": 0.20407889740210422,
+      "grad_norm": 0.42109382152557373,
+      "learning_rate": 9.226208211510282e-05,
+      "loss": 1.0524,
+      "step": 2284
+    },
+    {
+      "epoch": 0.2041682489333661,
+      "grad_norm": 0.4391646981239319,
+      "learning_rate": 9.225434742149293e-05,
+      "loss": 1.0501,
+      "step": 2285
+    },
+    {
+      "epoch": 0.20425760046462796,
+      "grad_norm": 0.45787525177001953,
+      "learning_rate": 9.224660918863104e-05,
+      "loss": 0.9669,
+      "step": 2286
+    },
+    {
+      "epoch": 0.20434695199588984,
+      "grad_norm": 0.3951761722564697,
+      "learning_rate": 9.22388674171653e-05,
+      "loss": 0.9985,
+      "step": 2287
+    },
+    {
+      "epoch": 0.2044363035271517,
+      "grad_norm": 0.5199114680290222,
+      "learning_rate": 9.22311221077442e-05,
+      "loss": 0.946,
+      "step": 2288
+    },
+    {
+      "epoch": 0.20452565505841355,
+      "grad_norm": 0.41517964005470276,
+      "learning_rate": 9.222337326101647e-05,
+      "loss": 1.0267,
+      "step": 2289
+    },
+    {
+      "epoch": 0.20461500658967544,
+      "grad_norm": 0.45859935879707336,
+      "learning_rate": 9.221562087763114e-05,
+      "loss": 0.9967,
+      "step": 2290
+    },
+    {
+      "epoch": 0.2047043581209373,
+      "grad_norm": 0.4338325560092926,
+      "learning_rate": 9.220786495823758e-05,
+      "loss": 1.023,
+      "step": 2291
+    },
+    {
+      "epoch": 0.20479370965219917,
+      "grad_norm": 0.41822728514671326,
+      "learning_rate": 9.220010550348544e-05,
+      "loss": 1.0318,
+      "step": 2292
+    },
+    {
+      "epoch": 0.20488306118346103,
+      "grad_norm": 0.4380994141101837,
+      "learning_rate": 9.219234251402464e-05,
+      "loss": 1.0465,
+      "step": 2293
+    },
+    {
+      "epoch": 0.20497241271472288,
+      "grad_norm": 0.4628269672393799,
+      "learning_rate": 9.218457599050542e-05,
+      "loss": 0.9637,
+      "step": 2294
+    },
+    {
+      "epoch": 0.20506176424598477,
+      "grad_norm": 0.41425222158432007,
+      "learning_rate": 9.217680593357829e-05,
+      "loss": 1.0134,
+      "step": 2295
+    },
+    {
+      "epoch": 0.20515111577724662,
+      "grad_norm": 0.33953461050987244,
+      "learning_rate": 9.216903234389412e-05,
+      "loss": 1.0589,
+      "step": 2296
+    },
+    {
+      "epoch": 0.2052404673085085,
+      "grad_norm": 0.42335033416748047,
+      "learning_rate": 9.216125522210398e-05,
+      "loss": 0.9652,
+      "step": 2297
+    },
+    {
+      "epoch": 0.20532981883977036,
+      "grad_norm": 0.39172297716140747,
+      "learning_rate": 9.21534745688593e-05,
+      "loss": 1.0338,
+      "step": 2298
+    },
+    {
+      "epoch": 0.20541917037103224,
+      "grad_norm": 0.4049052298069,
+      "learning_rate": 9.214569038481183e-05,
+      "loss": 0.9537,
+      "step": 2299
+    },
+    {
+      "epoch": 0.2055085219022941,
+      "grad_norm": 0.4806058406829834,
+      "learning_rate": 9.213790267061352e-05,
+      "loss": 0.9846,
+      "step": 2300
+    },
+    {
+      "epoch": 0.20559787343355596,
+      "grad_norm": 0.437541663646698,
+      "learning_rate": 9.213011142691671e-05,
+      "loss": 0.9776,
+      "step": 2301
+    },
+    {
+      "epoch": 0.20568722496481784,
+      "grad_norm": 0.42535480856895447,
+      "learning_rate": 9.2122316654374e-05,
+      "loss": 0.9601,
+      "step": 2302
+    },
+    {
+      "epoch": 0.2057765764960797,
+      "grad_norm": 0.3970091640949249,
+      "learning_rate": 9.211451835363828e-05,
+      "loss": 0.9903,
+      "step": 2303
+    },
+    {
+      "epoch": 0.20586592802734158,
+      "grad_norm": 0.4786219894886017,
+      "learning_rate": 9.210671652536274e-05,
+      "loss": 0.9416,
+      "step": 2304
+    },
+    {
+      "epoch": 0.20595527955860343,
+      "grad_norm": 0.4173687696456909,
+      "learning_rate": 9.209891117020087e-05,
+      "loss": 1.0231,
+      "step": 2305
+    },
+    {
+      "epoch": 0.20604463108986532,
+      "grad_norm": 0.49292880296707153,
+      "learning_rate": 9.209110228880642e-05,
+      "loss": 0.8597,
+      "step": 2306
+    },
+    {
+      "epoch": 0.20613398262112717,
+      "grad_norm": 0.4839416742324829,
+      "learning_rate": 9.208328988183352e-05,
+      "loss": 1.0004,
+      "step": 2307
+    },
+    {
+      "epoch": 0.20622333415238903,
+      "grad_norm": 0.5510335564613342,
+      "learning_rate": 9.207547394993651e-05,
+      "loss": 0.9456,
+      "step": 2308
+    },
+    {
+      "epoch": 0.2063126856836509,
+      "grad_norm": 0.3770674169063568,
+      "learning_rate": 9.206765449377006e-05,
+      "loss": 1.0004,
+      "step": 2309
+    },
+    {
+      "epoch": 0.20640203721491276,
+      "grad_norm": 0.43098023533821106,
+      "learning_rate": 9.205983151398915e-05,
+      "loss": 1.0377,
+      "step": 2310
+    },
+    {
+      "epoch": 0.20649138874617465,
+      "grad_norm": 0.4793153405189514,
+      "learning_rate": 9.205200501124902e-05,
+      "loss": 0.9746,
+      "step": 2311
+    },
+    {
+      "epoch": 0.2065807402774365,
+      "grad_norm": 0.5100858807563782,
+      "learning_rate": 9.204417498620522e-05,
+      "loss": 1.021,
+      "step": 2312
+    },
+    {
+      "epoch": 0.20667009180869836,
+      "grad_norm": 0.42289164662361145,
+      "learning_rate": 9.203634143951361e-05,
+      "loss": 0.9899,
+      "step": 2313
+    },
+    {
+      "epoch": 0.20675944333996024,
+      "grad_norm": 0.4890194237232208,
+      "learning_rate": 9.202850437183033e-05,
+      "loss": 0.94,
+      "step": 2314
+    },
+    {
+      "epoch": 0.2068487948712221,
+      "grad_norm": 0.3784908652305603,
+      "learning_rate": 9.202066378381183e-05,
+      "loss": 1.0157,
+      "step": 2315
+    },
+    {
+      "epoch": 0.20693814640248398,
+      "grad_norm": 0.4029442369937897,
+      "learning_rate": 9.201281967611481e-05,
+      "loss": 1.0325,
+      "step": 2316
+    },
+    {
+      "epoch": 0.20702749793374584,
+      "grad_norm": 0.413496732711792,
+      "learning_rate": 9.200497204939633e-05,
+      "loss": 1.0768,
+      "step": 2317
+    },
+    {
+      "epoch": 0.20711684946500772,
+      "grad_norm": 0.4051424562931061,
+      "learning_rate": 9.19971209043137e-05,
+      "loss": 1.0192,
+      "step": 2318
+    },
+    {
+      "epoch": 0.20720620099626957,
+      "grad_norm": 0.399827241897583,
+      "learning_rate": 9.198926624152453e-05,
+      "loss": 1.0137,
+      "step": 2319
+    },
+    {
+      "epoch": 0.20729555252753143,
+      "grad_norm": 0.41111016273498535,
+      "learning_rate": 9.198140806168673e-05,
+      "loss": 1.0135,
+      "step": 2320
+    },
+    {
+      "epoch": 0.2073849040587933,
+      "grad_norm": 0.49021974205970764,
+      "learning_rate": 9.197354636545853e-05,
+      "loss": 1.0126,
+      "step": 2321
+    },
+    {
+      "epoch": 0.20747425559005517,
+      "grad_norm": 0.37917208671569824,
+      "learning_rate": 9.196568115349842e-05,
+      "loss": 1.0455,
+      "step": 2322
+    },
+    {
+      "epoch": 0.20756360712131705,
+      "grad_norm": 0.4736994802951813,
+      "learning_rate": 9.195781242646517e-05,
+      "loss": 1.0129,
+      "step": 2323
+    },
+    {
+      "epoch": 0.2076529586525789,
+      "grad_norm": 0.444503515958786,
+      "learning_rate": 9.194994018501793e-05,
+      "loss": 0.998,
+      "step": 2324
+    },
+    {
+      "epoch": 0.2077423101838408,
+      "grad_norm": 0.4442484378814697,
+      "learning_rate": 9.194206442981601e-05,
+      "loss": 0.9962,
+      "step": 2325
+    },
+    {
+      "epoch": 0.20783166171510264,
+      "grad_norm": 0.48136886954307556,
+      "learning_rate": 9.193418516151912e-05,
+      "loss": 1.0792,
+      "step": 2326
+    },
+    {
+      "epoch": 0.2079210132463645,
+      "grad_norm": 0.4007897675037384,
+      "learning_rate": 9.192630238078725e-05,
+      "loss": 0.9787,
+      "step": 2327
+    },
+    {
+      "epoch": 0.20801036477762638,
+      "grad_norm": 0.42562204599380493,
+      "learning_rate": 9.191841608828066e-05,
+      "loss": 1.0511,
+      "step": 2328
+    },
+    {
+      "epoch": 0.20809971630888824,
+      "grad_norm": 0.4017636179924011,
+      "learning_rate": 9.19105262846599e-05,
+      "loss": 1.0076,
+      "step": 2329
+    },
+    {
+      "epoch": 0.20818906784015012,
+      "grad_norm": 0.4797520637512207,
+      "learning_rate": 9.190263297058583e-05,
+      "loss": 1.0116,
+      "step": 2330
+    },
+    {
+      "epoch": 0.20827841937141198,
+      "grad_norm": 0.37784191966056824,
+      "learning_rate": 9.189473614671959e-05,
+      "loss": 1.0355,
+      "step": 2331
+    },
+    {
+      "epoch": 0.20836777090267383,
+      "grad_norm": 0.43021056056022644,
+      "learning_rate": 9.188683581372264e-05,
+      "loss": 1.0424,
+      "step": 2332
+    },
+    {
+      "epoch": 0.20845712243393572,
+      "grad_norm": 0.3950195610523224,
+      "learning_rate": 9.187893197225672e-05,
+      "loss": 1.0493,
+      "step": 2333
+    },
+    {
+      "epoch": 0.20854647396519757,
+      "grad_norm": 0.40289580821990967,
+      "learning_rate": 9.187102462298384e-05,
+      "loss": 1.0145,
+      "step": 2334
+    },
+    {
+      "epoch": 0.20863582549645945,
+      "grad_norm": 0.3758835196495056,
+      "learning_rate": 9.186311376656633e-05,
+      "loss": 1.0352,
+      "step": 2335
+    },
+    {
+      "epoch": 0.2087251770277213,
+      "grad_norm": 0.44723740220069885,
+      "learning_rate": 9.185519940366682e-05,
+      "loss": 1.0553,
+      "step": 2336
+    },
+    {
+      "epoch": 0.2088145285589832,
+      "grad_norm": 0.4580692946910858,
+      "learning_rate": 9.18472815349482e-05,
+      "loss": 1.0657,
+      "step": 2337
+    },
+    {
+      "epoch": 0.20890388009024505,
+      "grad_norm": 0.5509561896324158,
+      "learning_rate": 9.183936016107371e-05,
+      "loss": 0.9759,
+      "step": 2338
+    },
+    {
+      "epoch": 0.2089932316215069,
+      "grad_norm": 0.5175040364265442,
+      "learning_rate": 9.183143528270682e-05,
+      "loss": 1.0432,
+      "step": 2339
+    },
+    {
+      "epoch": 0.2090825831527688,
+      "grad_norm": 0.40656107664108276,
+      "learning_rate": 9.182350690051133e-05,
+      "loss": 1.0077,
+      "step": 2340
+    },
+    {
+      "epoch": 0.20917193468403064,
+      "grad_norm": 0.44654014706611633,
+      "learning_rate": 9.181557501515134e-05,
+      "loss": 1.0218,
+      "step": 2341
+    },
+    {
+      "epoch": 0.20926128621529252,
+      "grad_norm": 0.4651091396808624,
+      "learning_rate": 9.180763962729123e-05,
+      "loss": 1.0464,
+      "step": 2342
+    },
+    {
+      "epoch": 0.20935063774655438,
+      "grad_norm": 0.4404200613498688,
+      "learning_rate": 9.179970073759565e-05,
+      "loss": 1.0241,
+      "step": 2343
+    },
+    {
+      "epoch": 0.20943998927781624,
+      "grad_norm": 0.5655425190925598,
+      "learning_rate": 9.17917583467296e-05,
+      "loss": 0.9452,
+      "step": 2344
+    },
+    {
+      "epoch": 0.20952934080907812,
+      "grad_norm": 0.442924827337265,
+      "learning_rate": 9.178381245535829e-05,
+      "loss": 1.0046,
+      "step": 2345
+    },
+    {
+      "epoch": 0.20961869234033997,
+      "grad_norm": 0.40375635027885437,
+      "learning_rate": 9.177586306414731e-05,
+      "loss": 1.0673,
+      "step": 2346
+    },
+    {
+      "epoch": 0.20970804387160186,
+      "grad_norm": 0.46348804235458374,
+      "learning_rate": 9.176791017376252e-05,
+      "loss": 1.0156,
+      "step": 2347
+    },
+    {
+      "epoch": 0.2097973954028637,
+      "grad_norm": 0.41273656487464905,
+      "learning_rate": 9.175995378487003e-05,
+      "loss": 1.0553,
+      "step": 2348
+    },
+    {
+      "epoch": 0.2098867469341256,
+      "grad_norm": 0.37597787380218506,
+      "learning_rate": 9.175199389813627e-05,
+      "loss": 0.9665,
+      "step": 2349
+    },
+    {
+      "epoch": 0.20997609846538745,
+      "grad_norm": 0.4291061758995056,
+      "learning_rate": 9.174403051422798e-05,
+      "loss": 0.9993,
+      "step": 2350
+    },
+    {
+      "epoch": 0.2100654499966493,
+      "grad_norm": 0.363323837518692,
+      "learning_rate": 9.173606363381219e-05,
+      "loss": 1.0601,
+      "step": 2351
+    },
+    {
+      "epoch": 0.2101548015279112,
+      "grad_norm": 0.4082275629043579,
+      "learning_rate": 9.172809325755618e-05,
+      "loss": 1.0879,
+      "step": 2352
+    },
+    {
+      "epoch": 0.21024415305917304,
+      "grad_norm": 0.41467830538749695,
+      "learning_rate": 9.172011938612757e-05,
+      "loss": 1.0513,
+      "step": 2353
+    },
+    {
+      "epoch": 0.21033350459043493,
+      "grad_norm": 0.40125423669815063,
+      "learning_rate": 9.171214202019428e-05,
+      "loss": 1.0362,
+      "step": 2354
+    },
+    {
+      "epoch": 0.21042285612169678,
+      "grad_norm": 0.38793444633483887,
+      "learning_rate": 9.170416116042444e-05,
+      "loss": 0.9959,
+      "step": 2355
+    },
+    {
+      "epoch": 0.21051220765295867,
+      "grad_norm": 0.4493177533149719,
+      "learning_rate": 9.169617680748659e-05,
+      "loss": 1.004,
+      "step": 2356
+    },
+    {
+      "epoch": 0.21060155918422052,
+      "grad_norm": 0.39111337065696716,
+      "learning_rate": 9.168818896204948e-05,
+      "loss": 0.9994,
+      "step": 2357
+    },
+    {
+      "epoch": 0.21069091071548238,
+      "grad_norm": 0.37774431705474854,
+      "learning_rate": 9.168019762478218e-05,
+      "loss": 1.0901,
+      "step": 2358
+    },
+    {
+      "epoch": 0.21078026224674426,
+      "grad_norm": 0.3941672444343567,
+      "learning_rate": 9.167220279635406e-05,
+      "loss": 1.0383,
+      "step": 2359
+    },
+    {
+      "epoch": 0.21086961377800612,
+      "grad_norm": 0.39567404985427856,
+      "learning_rate": 9.166420447743475e-05,
+      "loss": 0.986,
+      "step": 2360
+    },
+    {
+      "epoch": 0.210958965309268,
+      "grad_norm": 0.5099696516990662,
+      "learning_rate": 9.165620266869421e-05,
+      "loss": 0.9966,
+      "step": 2361
+    },
+    {
+      "epoch": 0.21104831684052985,
+      "grad_norm": 0.4867641031742096,
+      "learning_rate": 9.164819737080267e-05,
+      "loss": 1.037,
+      "step": 2362
+    },
+    {
+      "epoch": 0.2111376683717917,
+      "grad_norm": 0.41066691279411316,
+      "learning_rate": 9.164018858443066e-05,
+      "loss": 1.0574,
+      "step": 2363
+    },
+    {
+      "epoch": 0.2112270199030536,
+      "grad_norm": 0.4427512586116791,
+      "learning_rate": 9.163217631024901e-05,
+      "loss": 0.989,
+      "step": 2364
+    },
+    {
+      "epoch": 0.21131637143431545,
+      "grad_norm": 0.4066883325576782,
+      "learning_rate": 9.162416054892882e-05,
+      "loss": 0.979,
+      "step": 2365
+    },
+    {
+      "epoch": 0.21140572296557733,
+      "grad_norm": 0.4055411219596863,
+      "learning_rate": 9.161614130114151e-05,
+      "loss": 1.0365,
+      "step": 2366
+    },
+    {
+      "epoch": 0.21149507449683919,
+      "grad_norm": 0.4873175621032715,
+      "learning_rate": 9.160811856755877e-05,
+      "loss": 0.9807,
+      "step": 2367
+    },
+    {
+      "epoch": 0.21158442602810107,
+      "grad_norm": 0.4377719461917877,
+      "learning_rate": 9.160009234885258e-05,
+      "loss": 1.0499,
+      "step": 2368
+    },
+    {
+      "epoch": 0.21167377755936292,
+      "grad_norm": 0.4379160702228546,
+      "learning_rate": 9.159206264569524e-05,
+      "loss": 1.0161,
+      "step": 2369
+    },
+    {
+      "epoch": 0.21176312909062478,
+      "grad_norm": 0.41256406903266907,
+      "learning_rate": 9.158402945875932e-05,
+      "loss": 0.9943,
+      "step": 2370
+    },
+    {
+      "epoch": 0.21185248062188666,
+      "grad_norm": 0.3979935646057129,
+      "learning_rate": 9.157599278871767e-05,
+      "loss": 1.0566,
+      "step": 2371
+    },
+    {
+      "epoch": 0.21194183215314852,
+      "grad_norm": 0.4434359073638916,
+      "learning_rate": 9.156795263624345e-05,
+      "loss": 0.9642,
+      "step": 2372
+    },
+    {
+      "epoch": 0.2120311836844104,
+      "grad_norm": 0.42993849515914917,
+      "learning_rate": 9.155990900201012e-05,
+      "loss": 0.9881,
+      "step": 2373
+    },
+    {
+      "epoch": 0.21212053521567226,
+      "grad_norm": 0.4635986387729645,
+      "learning_rate": 9.155186188669143e-05,
+      "loss": 0.988,
+      "step": 2374
+    },
+    {
+      "epoch": 0.2122098867469341,
+      "grad_norm": 0.3796444535255432,
+      "learning_rate": 9.154381129096137e-05,
+      "loss": 1.0141,
+      "step": 2375
+    },
+    {
+      "epoch": 0.212299238278196,
+      "grad_norm": 0.468389093875885,
+      "learning_rate": 9.15357572154943e-05,
+      "loss": 0.9412,
+      "step": 2376
+    },
+    {
+      "epoch": 0.21238858980945785,
+      "grad_norm": 0.410773366689682,
+      "learning_rate": 9.152769966096482e-05,
+      "loss": 1.0115,
+      "step": 2377
+    },
+    {
+      "epoch": 0.21247794134071973,
+      "grad_norm": 0.41390907764434814,
+      "learning_rate": 9.151963862804784e-05,
+      "loss": 1.0499,
+      "step": 2378
+    },
+    {
+      "epoch": 0.2125672928719816,
+      "grad_norm": 0.4567076563835144,
+      "learning_rate": 9.151157411741858e-05,
+      "loss": 0.9748,
+      "step": 2379
+    },
+    {
+      "epoch": 0.21265664440324347,
+      "grad_norm": 0.4838281273841858,
+      "learning_rate": 9.150350612975247e-05,
+      "loss": 1.0076,
+      "step": 2380
+    },
+    {
+      "epoch": 0.21274599593450533,
+      "grad_norm": 0.4495866000652313,
+      "learning_rate": 9.149543466572535e-05,
+      "loss": 0.959,
+      "step": 2381
+    },
+    {
+      "epoch": 0.21283534746576718,
+      "grad_norm": 0.4260130226612091,
+      "learning_rate": 9.148735972601326e-05,
+      "loss": 1.0522,
+      "step": 2382
+    },
+    {
+      "epoch": 0.21292469899702907,
+      "grad_norm": 0.3835292160511017,
+      "learning_rate": 9.147928131129256e-05,
+      "loss": 1.0342,
+      "step": 2383
+    },
+    {
+      "epoch": 0.21301405052829092,
+      "grad_norm": 0.4930969774723053,
+      "learning_rate": 9.147119942223993e-05,
+      "loss": 0.9595,
+      "step": 2384
+    },
+    {
+      "epoch": 0.2131034020595528,
+      "grad_norm": 0.6031138896942139,
+      "learning_rate": 9.14631140595323e-05,
+      "loss": 1.0145,
+      "step": 2385
+    },
+    {
+      "epoch": 0.21319275359081466,
+      "grad_norm": 0.5279069542884827,
+      "learning_rate": 9.145502522384688e-05,
+      "loss": 0.9652,
+      "step": 2386
+    },
+    {
+      "epoch": 0.21328210512207654,
+      "grad_norm": 0.4844546616077423,
+      "learning_rate": 9.144693291586124e-05,
+      "loss": 1.0302,
+      "step": 2387
+    },
+    {
+      "epoch": 0.2133714566533384,
+      "grad_norm": 0.4257798194885254,
+      "learning_rate": 9.143883713625317e-05,
+      "loss": 1.0513,
+      "step": 2388
+    },
+    {
+      "epoch": 0.21346080818460025,
+      "grad_norm": 0.41385725140571594,
+      "learning_rate": 9.143073788570077e-05,
+      "loss": 0.9939,
+      "step": 2389
+    },
+    {
+      "epoch": 0.21355015971586214,
+      "grad_norm": 0.3999519944190979,
+      "learning_rate": 9.142263516488245e-05,
+      "loss": 0.9706,
+      "step": 2390
+    },
+    {
+      "epoch": 0.213639511247124,
+      "grad_norm": 0.3895846903324127,
+      "learning_rate": 9.141452897447692e-05,
+      "loss": 1.074,
+      "step": 2391
+    },
+    {
+      "epoch": 0.21372886277838588,
+      "grad_norm": 0.44434553384780884,
+      "learning_rate": 9.140641931516314e-05,
+      "loss": 1.0137,
+      "step": 2392
+    },
+    {
+      "epoch": 0.21381821430964773,
+      "grad_norm": 0.3749370574951172,
+      "learning_rate": 9.139830618762038e-05,
+      "loss": 1.0124,
+      "step": 2393
+    },
+    {
+      "epoch": 0.21390756584090959,
+      "grad_norm": 0.37694430351257324,
+      "learning_rate": 9.139018959252819e-05,
+      "loss": 1.0097,
+      "step": 2394
+    },
+    {
+      "epoch": 0.21399691737217147,
+      "grad_norm": 0.3813773989677429,
+      "learning_rate": 9.138206953056644e-05,
+      "loss": 1.0833,
+      "step": 2395
+    },
+    {
+      "epoch": 0.21408626890343332,
+      "grad_norm": 0.4502156972885132,
+      "learning_rate": 9.137394600241527e-05,
+      "loss": 0.9692,
+      "step": 2396
+    },
+    {
+      "epoch": 0.2141756204346952,
+      "grad_norm": 0.41509172320365906,
+      "learning_rate": 9.136581900875512e-05,
+      "loss": 0.9825,
+      "step": 2397
+    },
+    {
+      "epoch": 0.21426497196595706,
+      "grad_norm": 0.4003329277038574,
+      "learning_rate": 9.135768855026668e-05,
+      "loss": 1.0834,
+      "step": 2398
+    },
+    {
+      "epoch": 0.21435432349721895,
+      "grad_norm": 0.4339703917503357,
+      "learning_rate": 9.1349554627631e-05,
+      "loss": 1.0946,
+      "step": 2399
+    },
+    {
+      "epoch": 0.2144436750284808,
+      "grad_norm": 0.43198466300964355,
+      "learning_rate": 9.134141724152934e-05,
+      "loss": 0.9449,
+      "step": 2400
+    },
+    {
+      "epoch": 0.21453302655974266,
+      "grad_norm": 0.44103294610977173,
+      "learning_rate": 9.133327639264332e-05,
+      "loss": 0.9653,
+      "step": 2401
+    },
+    {
+      "epoch": 0.21462237809100454,
+      "grad_norm": 0.4824722707271576,
+      "learning_rate": 9.132513208165486e-05,
+      "loss": 1.0428,
+      "step": 2402
+    },
+    {
+      "epoch": 0.2147117296222664,
+      "grad_norm": 0.49322792887687683,
+      "learning_rate": 9.131698430924605e-05,
+      "loss": 1.0646,
+      "step": 2403
+    },
+    {
+      "epoch": 0.21480108115352828,
+      "grad_norm": 0.4989064931869507,
+      "learning_rate": 9.130883307609942e-05,
+      "loss": 1.0013,
+      "step": 2404
+    },
+    {
+      "epoch": 0.21489043268479013,
+      "grad_norm": 0.459974467754364,
+      "learning_rate": 9.130067838289769e-05,
+      "loss": 1.0519,
+      "step": 2405
+    },
+    {
+      "epoch": 0.214979784216052,
+      "grad_norm": 0.44831836223602295,
+      "learning_rate": 9.129252023032391e-05,
+      "loss": 1.055,
+      "step": 2406
+    },
+    {
+      "epoch": 0.21506913574731387,
+      "grad_norm": 0.4193100929260254,
+      "learning_rate": 9.128435861906142e-05,
+      "loss": 0.9937,
+      "step": 2407
+    },
+    {
+      "epoch": 0.21515848727857573,
+      "grad_norm": 0.4522268772125244,
+      "learning_rate": 9.127619354979384e-05,
+      "loss": 1.076,
+      "step": 2408
+    },
+    {
+      "epoch": 0.2152478388098376,
+      "grad_norm": 0.4380955100059509,
+      "learning_rate": 9.126802502320509e-05,
+      "loss": 0.9917,
+      "step": 2409
+    },
+    {
+      "epoch": 0.21533719034109947,
+      "grad_norm": 0.45167332887649536,
+      "learning_rate": 9.125985303997933e-05,
+      "loss": 1.0884,
+      "step": 2410
+    },
+    {
+      "epoch": 0.21542654187236135,
+      "grad_norm": 0.4360027611255646,
+      "learning_rate": 9.125167760080108e-05,
+      "loss": 0.9882,
+      "step": 2411
+    },
+    {
+      "epoch": 0.2155158934036232,
+      "grad_norm": 0.390898197889328,
+      "learning_rate": 9.124349870635515e-05,
+      "loss": 1.0076,
+      "step": 2412
+    },
+    {
+      "epoch": 0.21560524493488506,
+      "grad_norm": 0.4726807475090027,
+      "learning_rate": 9.123531635732656e-05,
+      "loss": 0.9496,
+      "step": 2413
+    },
+    {
+      "epoch": 0.21569459646614694,
+      "grad_norm": 0.4268726706504822,
+      "learning_rate": 9.122713055440069e-05,
+      "loss": 1.0008,
+      "step": 2414
+    },
+    {
+      "epoch": 0.2157839479974088,
+      "grad_norm": 0.43627238273620605,
+      "learning_rate": 9.121894129826318e-05,
+      "loss": 0.9642,
+      "step": 2415
+    },
+    {
+      "epoch": 0.21587329952867068,
+      "grad_norm": 0.4274177849292755,
+      "learning_rate": 9.121074858959997e-05,
+      "loss": 1.0273,
+      "step": 2416
+    },
+    {
+      "epoch": 0.21596265105993254,
+      "grad_norm": 0.4479106068611145,
+      "learning_rate": 9.12025524290973e-05,
+      "loss": 0.9998,
+      "step": 2417
+    },
+    {
+      "epoch": 0.21605200259119442,
+      "grad_norm": 0.45606303215026855,
+      "learning_rate": 9.119435281744169e-05,
+      "loss": 0.9753,
+      "step": 2418
+    },
+    {
+      "epoch": 0.21614135412245628,
+      "grad_norm": 0.40159735083580017,
+      "learning_rate": 9.118614975531991e-05,
+      "loss": 1.011,
+      "step": 2419
+    },
+    {
+      "epoch": 0.21623070565371813,
+      "grad_norm": 0.45181921124458313,
+      "learning_rate": 9.117794324341908e-05,
+      "loss": 1.0253,
+      "step": 2420
+    },
+    {
+      "epoch": 0.21632005718498,
+      "grad_norm": 0.41297322511672974,
+      "learning_rate": 9.116973328242658e-05,
+      "loss": 1.001,
+      "step": 2421
+    },
+    {
+      "epoch": 0.21640940871624187,
+      "grad_norm": 0.39003559947013855,
+      "learning_rate": 9.116151987303007e-05,
+      "loss": 1.0044,
+      "step": 2422
+    },
+    {
+      "epoch": 0.21649876024750375,
+      "grad_norm": 0.45136478543281555,
+      "learning_rate": 9.115330301591753e-05,
+      "loss": 0.9569,
+      "step": 2423
+    },
+    {
+      "epoch": 0.2165881117787656,
+      "grad_norm": 0.4354073107242584,
+      "learning_rate": 9.114508271177722e-05,
+      "loss": 0.9862,
+      "step": 2424
+    },
+    {
+      "epoch": 0.21667746331002746,
+      "grad_norm": 0.3921179175376892,
+      "learning_rate": 9.113685896129763e-05,
+      "loss": 0.9894,
+      "step": 2425
+    },
+    {
+      "epoch": 0.21676681484128935,
+      "grad_norm": 0.42139732837677,
+      "learning_rate": 9.112863176516762e-05,
+      "loss": 0.9777,
+      "step": 2426
+    },
+    {
+      "epoch": 0.2168561663725512,
+      "grad_norm": 0.5038270950317383,
+      "learning_rate": 9.11204011240763e-05,
+      "loss": 1.1157,
+      "step": 2427
+    },
+    {
+      "epoch": 0.21694551790381308,
+      "grad_norm": 0.5017008781433105,
+      "learning_rate": 9.111216703871308e-05,
+      "loss": 0.9634,
+      "step": 2428
+    },
+    {
+      "epoch": 0.21703486943507494,
+      "grad_norm": 0.5621154308319092,
+      "learning_rate": 9.110392950976764e-05,
+      "loss": 1.0222,
+      "step": 2429
+    },
+    {
+      "epoch": 0.21712422096633682,
+      "grad_norm": 0.43633437156677246,
+      "learning_rate": 9.109568853792998e-05,
+      "loss": 1.1109,
+      "step": 2430
+    },
+    {
+      "epoch": 0.21721357249759868,
+      "grad_norm": 0.48714277148246765,
+      "learning_rate": 9.108744412389034e-05,
+      "loss": 0.9751,
+      "step": 2431
+    },
+    {
+      "epoch": 0.21730292402886053,
+      "grad_norm": 0.4233207702636719,
+      "learning_rate": 9.107919626833931e-05,
+      "loss": 1.0543,
+      "step": 2432
+    },
+    {
+      "epoch": 0.21739227556012242,
+      "grad_norm": 0.3775826096534729,
+      "learning_rate": 9.107094497196771e-05,
+      "loss": 1.0672,
+      "step": 2433
+    },
+    {
+      "epoch": 0.21748162709138427,
+      "grad_norm": 0.47611549496650696,
+      "learning_rate": 9.106269023546667e-05,
+      "loss": 1.0062,
+      "step": 2434
+    },
+    {
+      "epoch": 0.21757097862264616,
+      "grad_norm": 0.40070483088493347,
+      "learning_rate": 9.105443205952765e-05,
+      "loss": 0.9908,
+      "step": 2435
+    },
+    {
+      "epoch": 0.217660330153908,
+      "grad_norm": 0.4242134392261505,
+      "learning_rate": 9.104617044484233e-05,
+      "loss": 1.0447,
+      "step": 2436
+    },
+    {
+      "epoch": 0.21774968168516987,
+      "grad_norm": 0.4634241461753845,
+      "learning_rate": 9.103790539210271e-05,
+      "loss": 1.0203,
+      "step": 2437
+    },
+    {
+      "epoch": 0.21783903321643175,
+      "grad_norm": 0.46753108501434326,
+      "learning_rate": 9.102963690200108e-05,
+      "loss": 0.892,
+      "step": 2438
+    },
+    {
+      "epoch": 0.2179283847476936,
+      "grad_norm": 0.47275984287261963,
+      "learning_rate": 9.102136497523002e-05,
+      "loss": 1.0528,
+      "step": 2439
+    },
+    {
+      "epoch": 0.2180177362789555,
+      "grad_norm": 0.4334014058113098,
+      "learning_rate": 9.101308961248238e-05,
+      "loss": 0.9923,
+      "step": 2440
+    },
+    {
+      "epoch": 0.21810708781021734,
+      "grad_norm": 0.41837531328201294,
+      "learning_rate": 9.100481081445132e-05,
+      "loss": 0.9946,
+      "step": 2441
+    },
+    {
+      "epoch": 0.21819643934147923,
+      "grad_norm": 0.4113021194934845,
+      "learning_rate": 9.099652858183028e-05,
+      "loss": 1.0441,
+      "step": 2442
+    },
+    {
+      "epoch": 0.21828579087274108,
+      "grad_norm": 0.4260026514530182,
+      "learning_rate": 9.098824291531296e-05,
+      "loss": 1.0988,
+      "step": 2443
+    },
+    {
+      "epoch": 0.21837514240400294,
+      "grad_norm": 0.4235369563102722,
+      "learning_rate": 9.097995381559341e-05,
+      "loss": 1.0867,
+      "step": 2444
+    },
+    {
+      "epoch": 0.21846449393526482,
+      "grad_norm": 0.4195026755332947,
+      "learning_rate": 9.097166128336592e-05,
+      "loss": 1.0309,
+      "step": 2445
+    },
+    {
+      "epoch": 0.21855384546652667,
+      "grad_norm": 0.4182673990726471,
+      "learning_rate": 9.096336531932506e-05,
+      "loss": 1.0566,
+      "step": 2446
+    },
+    {
+      "epoch": 0.21864319699778856,
+      "grad_norm": 0.40884897112846375,
+      "learning_rate": 9.095506592416572e-05,
+      "loss": 1.0218,
+      "step": 2447
+    },
+    {
+      "epoch": 0.2187325485290504,
+      "grad_norm": 0.4604688584804535,
+      "learning_rate": 9.094676309858305e-05,
+      "loss": 1.0578,
+      "step": 2448
+    },
+    {
+      "epoch": 0.2188219000603123,
+      "grad_norm": 0.4072805941104889,
+      "learning_rate": 9.093845684327251e-05,
+      "loss": 1.0117,
+      "step": 2449
+    },
+    {
+      "epoch": 0.21891125159157415,
+      "grad_norm": 0.3600502014160156,
+      "learning_rate": 9.093014715892984e-05,
+      "loss": 1.0826,
+      "step": 2450
+    },
+    {
+      "epoch": 0.219000603122836,
+      "grad_norm": 0.421089231967926,
+      "learning_rate": 9.092183404625107e-05,
+      "loss": 1.0072,
+      "step": 2451
+    },
+    {
+      "epoch": 0.2190899546540979,
+      "grad_norm": 0.4491601586341858,
+      "learning_rate": 9.09135175059325e-05,
+      "loss": 1.0409,
+      "step": 2452
+    },
+    {
+      "epoch": 0.21917930618535975,
+      "grad_norm": 0.4364774823188782,
+      "learning_rate": 9.090519753867072e-05,
+      "loss": 1.0142,
+      "step": 2453
+    },
+    {
+      "epoch": 0.21926865771662163,
+      "grad_norm": 0.4364645779132843,
+      "learning_rate": 9.089687414516265e-05,
+      "loss": 1.0297,
+      "step": 2454
+    },
+    {
+      "epoch": 0.21935800924788348,
+      "grad_norm": 0.4864216148853302,
+      "learning_rate": 9.088854732610543e-05,
+      "loss": 1.0635,
+      "step": 2455
+    },
+    {
+      "epoch": 0.21944736077914534,
+      "grad_norm": 0.4514232873916626,
+      "learning_rate": 9.088021708219652e-05,
+      "loss": 1.0224,
+      "step": 2456
+    },
+    {
+      "epoch": 0.21953671231040722,
+      "grad_norm": 0.3978760540485382,
+      "learning_rate": 9.087188341413369e-05,
+      "loss": 1.0442,
+      "step": 2457
+    },
+    {
+      "epoch": 0.21962606384166908,
+      "grad_norm": 0.36813050508499146,
+      "learning_rate": 9.086354632261496e-05,
+      "loss": 1.0114,
+      "step": 2458
+    },
+    {
+      "epoch": 0.21971541537293096,
+      "grad_norm": 0.4238460063934326,
+      "learning_rate": 9.085520580833866e-05,
+      "loss": 1.0065,
+      "step": 2459
+    },
+    {
+      "epoch": 0.21980476690419282,
+      "grad_norm": 0.57405024766922,
+      "learning_rate": 9.084686187200338e-05,
+      "loss": 0.9361,
+      "step": 2460
+    },
+    {
+      "epoch": 0.2198941184354547,
+      "grad_norm": 0.5002081990242004,
+      "learning_rate": 9.0838514514308e-05,
+      "loss": 1.0109,
+      "step": 2461
+    },
+    {
+      "epoch": 0.21998346996671655,
+      "grad_norm": 0.397339403629303,
+      "learning_rate": 9.083016373595174e-05,
+      "loss": 1.0188,
+      "step": 2462
+    },
+    {
+      "epoch": 0.2200728214979784,
+      "grad_norm": 0.4922274947166443,
+      "learning_rate": 9.082180953763406e-05,
+      "loss": 0.9552,
+      "step": 2463
+    },
+    {
+      "epoch": 0.2201621730292403,
+      "grad_norm": 0.49494630098342896,
+      "learning_rate": 9.08134519200547e-05,
+      "loss": 1.0058,
+      "step": 2464
+    },
+    {
+      "epoch": 0.22025152456050215,
+      "grad_norm": 0.44430068135261536,
+      "learning_rate": 9.080509088391369e-05,
+      "loss": 1.0486,
+      "step": 2465
+    },
+    {
+      "epoch": 0.22034087609176403,
+      "grad_norm": 0.4087202548980713,
+      "learning_rate": 9.079672642991137e-05,
+      "loss": 1.0006,
+      "step": 2466
+    },
+    {
+      "epoch": 0.2204302276230259,
+      "grad_norm": 0.49828004837036133,
+      "learning_rate": 9.078835855874835e-05,
+      "loss": 1.0229,
+      "step": 2467
+    },
+    {
+      "epoch": 0.22051957915428777,
+      "grad_norm": 0.42710915207862854,
+      "learning_rate": 9.077998727112554e-05,
+      "loss": 1.0254,
+      "step": 2468
+    },
+    {
+      "epoch": 0.22060893068554963,
+      "grad_norm": 0.40693971514701843,
+      "learning_rate": 9.077161256774409e-05,
+      "loss": 1.0551,
+      "step": 2469
+    },
+    {
+      "epoch": 0.22069828221681148,
+      "grad_norm": 0.4143005609512329,
+      "learning_rate": 9.076323444930551e-05,
+      "loss": 0.9634,
+      "step": 2470
+    },
+    {
+      "epoch": 0.22078763374807336,
+      "grad_norm": 0.4316956102848053,
+      "learning_rate": 9.075485291651154e-05,
+      "loss": 0.9414,
+      "step": 2471
+    },
+    {
+      "epoch": 0.22087698527933522,
+      "grad_norm": 0.42413267493247986,
+      "learning_rate": 9.07464679700642e-05,
+      "loss": 0.9748,
+      "step": 2472
+    },
+    {
+      "epoch": 0.2209663368105971,
+      "grad_norm": 0.4333771765232086,
+      "learning_rate": 9.073807961066588e-05,
+      "loss": 0.9875,
+      "step": 2473
+    },
+    {
+      "epoch": 0.22105568834185896,
+      "grad_norm": 0.4280893802642822,
+      "learning_rate": 9.072968783901913e-05,
+      "loss": 0.972,
+      "step": 2474
+    },
+    {
+      "epoch": 0.2211450398731208,
+      "grad_norm": 0.4792003333568573,
+      "learning_rate": 9.072129265582689e-05,
+      "loss": 1.0064,
+      "step": 2475
+    },
+    {
+      "epoch": 0.2212343914043827,
+      "grad_norm": 0.44787389039993286,
+      "learning_rate": 9.071289406179231e-05,
+      "loss": 0.9806,
+      "step": 2476
+    },
+    {
+      "epoch": 0.22132374293564455,
+      "grad_norm": 0.43630531430244446,
+      "learning_rate": 9.070449205761891e-05,
+      "loss": 0.9906,
+      "step": 2477
+    },
+    {
+      "epoch": 0.22141309446690643,
+      "grad_norm": 0.46949827671051025,
+      "learning_rate": 9.069608664401041e-05,
+      "loss": 1.0052,
+      "step": 2478
+    },
+    {
+      "epoch": 0.2215024459981683,
+      "grad_norm": 0.37613534927368164,
+      "learning_rate": 9.068767782167086e-05,
+      "loss": 0.9737,
+      "step": 2479
+    },
+    {
+      "epoch": 0.22159179752943017,
+      "grad_norm": 0.4791412353515625,
+      "learning_rate": 9.06792655913046e-05,
+      "loss": 0.9505,
+      "step": 2480
+    },
+    {
+      "epoch": 0.22168114906069203,
+      "grad_norm": 0.4607903063297272,
+      "learning_rate": 9.067084995361623e-05,
+      "loss": 1.1162,
+      "step": 2481
+    },
+    {
+      "epoch": 0.22177050059195388,
+      "grad_norm": 0.40373310446739197,
+      "learning_rate": 9.066243090931066e-05,
+      "loss": 0.9888,
+      "step": 2482
+    },
+    {
+      "epoch": 0.22185985212321577,
+      "grad_norm": 0.40429624915122986,
+      "learning_rate": 9.065400845909308e-05,
+      "loss": 1.0024,
+      "step": 2483
+    },
+    {
+      "epoch": 0.22194920365447762,
+      "grad_norm": 0.40631914138793945,
+      "learning_rate": 9.064558260366893e-05,
+      "loss": 0.9719,
+      "step": 2484
+    },
+    {
+      "epoch": 0.2220385551857395,
+      "grad_norm": 0.43395867943763733,
+      "learning_rate": 9.063715334374401e-05,
+      "loss": 0.9634,
+      "step": 2485
+    },
+    {
+      "epoch": 0.22212790671700136,
+      "grad_norm": 0.4469437003135681,
+      "learning_rate": 9.062872068002432e-05,
+      "loss": 0.9666,
+      "step": 2486
+    },
+    {
+      "epoch": 0.22221725824826322,
+      "grad_norm": 0.5009754300117493,
+      "learning_rate": 9.062028461321621e-05,
+      "loss": 0.9931,
+      "step": 2487
+    },
+    {
+      "epoch": 0.2223066097795251,
+      "grad_norm": 0.4312589764595032,
+      "learning_rate": 9.061184514402627e-05,
+      "loss": 1.0201,
+      "step": 2488
+    },
+    {
+      "epoch": 0.22239596131078695,
+      "grad_norm": 0.45867425203323364,
+      "learning_rate": 9.060340227316142e-05,
+      "loss": 1.0177,
+      "step": 2489
+    },
+    {
+      "epoch": 0.22248531284204884,
+      "grad_norm": 0.39511245489120483,
+      "learning_rate": 9.059495600132883e-05,
+      "loss": 1.0383,
+      "step": 2490
+    },
+    {
+      "epoch": 0.2225746643733107,
+      "grad_norm": 0.4442897439002991,
+      "learning_rate": 9.058650632923595e-05,
+      "loss": 0.9576,
+      "step": 2491
+    },
+    {
+      "epoch": 0.22266401590457258,
+      "grad_norm": 0.556317925453186,
+      "learning_rate": 9.057805325759057e-05,
+      "loss": 0.9748,
+      "step": 2492
+    },
+    {
+      "epoch": 0.22275336743583443,
+      "grad_norm": 0.49973076581954956,
+      "learning_rate": 9.056959678710067e-05,
+      "loss": 0.8914,
+      "step": 2493
+    },
+    {
+      "epoch": 0.2228427189670963,
+      "grad_norm": 0.5200818181037903,
+      "learning_rate": 9.056113691847461e-05,
+      "loss": 0.9384,
+      "step": 2494
+    },
+    {
+      "epoch": 0.22293207049835817,
+      "grad_norm": 0.4469231963157654,
+      "learning_rate": 9.055267365242099e-05,
+      "loss": 0.9844,
+      "step": 2495
+    },
+    {
+      "epoch": 0.22302142202962003,
+      "grad_norm": 0.4275433123111725,
+      "learning_rate": 9.054420698964868e-05,
+      "loss": 1.0414,
+      "step": 2496
+    },
+    {
+      "epoch": 0.2231107735608819,
+      "grad_norm": 0.36767879128456116,
+      "learning_rate": 9.053573693086687e-05,
+      "loss": 1.026,
+      "step": 2497
+    },
+    {
+      "epoch": 0.22320012509214376,
+      "grad_norm": 0.3746047019958496,
+      "learning_rate": 9.052726347678502e-05,
+      "loss": 1.0667,
+      "step": 2498
+    },
+    {
+      "epoch": 0.22328947662340565,
+      "grad_norm": 0.44350770115852356,
+      "learning_rate": 9.051878662811286e-05,
+      "loss": 1.0337,
+      "step": 2499
+    },
+    {
+      "epoch": 0.2233788281546675,
+      "grad_norm": 0.42601439356803894,
+      "learning_rate": 9.051030638556041e-05,
+      "loss": 1.0673,
+      "step": 2500
+    },
+    {
+      "epoch": 0.22346817968592936,
+      "grad_norm": 0.4272880554199219,
+      "learning_rate": 9.050182274983798e-05,
+      "loss": 1.0239,
+      "step": 2501
+    },
+    {
+      "epoch": 0.22355753121719124,
+      "grad_norm": 0.4652038812637329,
+      "learning_rate": 9.04933357216562e-05,
+      "loss": 0.959,
+      "step": 2502
+    },
+    {
+      "epoch": 0.2236468827484531,
+      "grad_norm": 0.4481859505176544,
+      "learning_rate": 9.048484530172592e-05,
+      "loss": 1.0601,
+      "step": 2503
+    },
+    {
+      "epoch": 0.22373623427971498,
+      "grad_norm": 0.4447591304779053,
+      "learning_rate": 9.04763514907583e-05,
+      "loss": 1.0246,
+      "step": 2504
+    },
+    {
+      "epoch": 0.22382558581097683,
+      "grad_norm": 0.37949299812316895,
+      "learning_rate": 9.046785428946481e-05,
+      "loss": 1.0761,
+      "step": 2505
+    },
+    {
+      "epoch": 0.2239149373422387,
+      "grad_norm": 0.44117259979248047,
+      "learning_rate": 9.045935369855716e-05,
+      "loss": 1.0082,
+      "step": 2506
+    },
+    {
+      "epoch": 0.22400428887350057,
+      "grad_norm": 0.4014883041381836,
+      "learning_rate": 9.045084971874738e-05,
+      "loss": 1.0867,
+      "step": 2507
+    },
+    {
+      "epoch": 0.22409364040476243,
+      "grad_norm": 0.5892843008041382,
+      "learning_rate": 9.044234235074775e-05,
+      "loss": 0.9446,
+      "step": 2508
+    },
+    {
+      "epoch": 0.2241829919360243,
+      "grad_norm": 0.4554404318332672,
+      "learning_rate": 9.043383159527087e-05,
+      "loss": 0.939,
+      "step": 2509
+    },
+    {
+      "epoch": 0.22427234346728617,
+      "grad_norm": 0.3739197552204132,
+      "learning_rate": 9.04253174530296e-05,
+      "loss": 1.0603,
+      "step": 2510
+    },
+    {
+      "epoch": 0.22436169499854805,
+      "grad_norm": 0.48913314938545227,
+      "learning_rate": 9.041679992473708e-05,
+      "loss": 0.9852,
+      "step": 2511
+    },
+    {
+      "epoch": 0.2244510465298099,
+      "grad_norm": 0.3724241256713867,
+      "learning_rate": 9.040827901110676e-05,
+      "loss": 1.0422,
+      "step": 2512
+    },
+    {
+      "epoch": 0.22454039806107176,
+      "grad_norm": 0.4462158679962158,
+      "learning_rate": 9.039975471285235e-05,
+      "loss": 1.0337,
+      "step": 2513
+    },
+    {
+      "epoch": 0.22462974959233364,
+      "grad_norm": 0.4481222629547119,
+      "learning_rate": 9.039122703068785e-05,
+      "loss": 1.0114,
+      "step": 2514
+    },
+    {
+      "epoch": 0.2247191011235955,
+      "grad_norm": 0.4646155536174774,
+      "learning_rate": 9.038269596532755e-05,
+      "loss": 1.0595,
+      "step": 2515
+    },
+    {
+      "epoch": 0.22480845265485738,
+      "grad_norm": 0.4418215751647949,
+      "learning_rate": 9.037416151748603e-05,
+      "loss": 0.9829,
+      "step": 2516
+    },
+    {
+      "epoch": 0.22489780418611924,
+      "grad_norm": 0.4447876811027527,
+      "learning_rate": 9.036562368787811e-05,
+      "loss": 0.9914,
+      "step": 2517
+    },
+    {
+      "epoch": 0.2249871557173811,
+      "grad_norm": 0.47525134682655334,
+      "learning_rate": 9.035708247721895e-05,
+      "loss": 1.0124,
+      "step": 2518
+    },
+    {
+      "epoch": 0.22507650724864298,
+      "grad_norm": 0.5248557925224304,
+      "learning_rate": 9.034853788622393e-05,
+      "loss": 0.9379,
+      "step": 2519
+    },
+    {
+      "epoch": 0.22516585877990483,
+      "grad_norm": 0.4031184911727905,
+      "learning_rate": 9.033998991560881e-05,
+      "loss": 0.9834,
+      "step": 2520
+    },
+    {
+      "epoch": 0.22525521031116671,
+      "grad_norm": 0.38345515727996826,
+      "learning_rate": 9.033143856608952e-05,
+      "loss": 1.043,
+      "step": 2521
+    },
+    {
+      "epoch": 0.22534456184242857,
+      "grad_norm": 0.40574783086776733,
+      "learning_rate": 9.032288383838236e-05,
+      "loss": 1.0062,
+      "step": 2522
+    },
+    {
+      "epoch": 0.22543391337369045,
+      "grad_norm": 0.5696341395378113,
+      "learning_rate": 9.031432573320387e-05,
+      "loss": 0.9409,
+      "step": 2523
+    },
+    {
+      "epoch": 0.2255232649049523,
+      "grad_norm": 0.39741194248199463,
+      "learning_rate": 9.030576425127087e-05,
+      "loss": 1.0018,
+      "step": 2524
+    },
+    {
+      "epoch": 0.22561261643621416,
+      "grad_norm": 0.41686055064201355,
+      "learning_rate": 9.029719939330047e-05,
+      "loss": 0.9734,
+      "step": 2525
+    },
+    {
+      "epoch": 0.22570196796747605,
+      "grad_norm": 0.39542368054389954,
+      "learning_rate": 9.028863116001012e-05,
+      "loss": 0.9796,
+      "step": 2526
+    },
+    {
+      "epoch": 0.2257913194987379,
+      "grad_norm": 0.4378524720668793,
+      "learning_rate": 9.028005955211744e-05,
+      "loss": 0.9701,
+      "step": 2527
+    },
+    {
+      "epoch": 0.22588067102999979,
+      "grad_norm": 0.5109492540359497,
+      "learning_rate": 9.027148457034043e-05,
+      "loss": 1.0008,
+      "step": 2528
+    },
+    {
+      "epoch": 0.22597002256126164,
+      "grad_norm": 0.35746294260025024,
+      "learning_rate": 9.02629062153973e-05,
+      "loss": 1.037,
+      "step": 2529
+    },
+    {
+      "epoch": 0.22605937409252352,
+      "grad_norm": 0.44290691614151,
+      "learning_rate": 9.025432448800662e-05,
+      "loss": 1.0122,
+      "step": 2530
+    },
+    {
+      "epoch": 0.22614872562378538,
+      "grad_norm": 0.4638045132160187,
+      "learning_rate": 9.02457393888872e-05,
+      "loss": 1.0214,
+      "step": 2531
+    },
+    {
+      "epoch": 0.22623807715504723,
+      "grad_norm": 0.41130268573760986,
+      "learning_rate": 9.023715091875809e-05,
+      "loss": 1.0102,
+      "step": 2532
+    },
+    {
+      "epoch": 0.22632742868630912,
+      "grad_norm": 0.4205174446105957,
+      "learning_rate": 9.022855907833871e-05,
+      "loss": 0.9937,
+      "step": 2533
+    },
+    {
+      "epoch": 0.22641678021757097,
+      "grad_norm": 0.4878970980644226,
+      "learning_rate": 9.02199638683487e-05,
+      "loss": 0.9084,
+      "step": 2534
+    },
+    {
+      "epoch": 0.22650613174883286,
+      "grad_norm": 0.44246265292167664,
+      "learning_rate": 9.0211365289508e-05,
+      "loss": 0.946,
+      "step": 2535
+    },
+    {
+      "epoch": 0.2265954832800947,
+      "grad_norm": 0.4307198226451874,
+      "learning_rate": 9.020276334253683e-05,
+      "loss": 1.0387,
+      "step": 2536
+    },
+    {
+      "epoch": 0.22668483481135657,
+      "grad_norm": 0.45680615305900574,
+      "learning_rate": 9.019415802815569e-05,
+      "loss": 1.0461,
+      "step": 2537
+    },
+    {
+      "epoch": 0.22677418634261845,
+      "grad_norm": 0.3967481851577759,
+      "learning_rate": 9.01855493470854e-05,
+      "loss": 1.0014,
+      "step": 2538
+    },
+    {
+      "epoch": 0.2268635378738803,
+      "grad_norm": 0.49700799584388733,
+      "learning_rate": 9.0176937300047e-05,
+      "loss": 1.0862,
+      "step": 2539
+    },
+    {
+      "epoch": 0.2269528894051422,
+      "grad_norm": 0.4415007531642914,
+      "learning_rate": 9.016832188776183e-05,
+      "loss": 1.049,
+      "step": 2540
+    },
+    {
+      "epoch": 0.22704224093640404,
+      "grad_norm": 0.39791885018348694,
+      "learning_rate": 9.015970311095156e-05,
+      "loss": 0.9441,
+      "step": 2541
+    },
+    {
+      "epoch": 0.22713159246766593,
+      "grad_norm": 0.41884180903434753,
+      "learning_rate": 9.015108097033806e-05,
+      "loss": 1.0447,
+      "step": 2542
+    },
+    {
+      "epoch": 0.22722094399892778,
+      "grad_norm": 0.4121531546115875,
+      "learning_rate": 9.014245546664357e-05,
+      "loss": 1.0104,
+      "step": 2543
+    },
+    {
+      "epoch": 0.22731029553018964,
+      "grad_norm": 0.43119367957115173,
+      "learning_rate": 9.013382660059053e-05,
+      "loss": 0.9879,
+      "step": 2544
+    },
+    {
+      "epoch": 0.22739964706145152,
+      "grad_norm": 0.4344024360179901,
+      "learning_rate": 9.012519437290172e-05,
+      "loss": 1.0233,
+      "step": 2545
+    },
+    {
+      "epoch": 0.22748899859271338,
+      "grad_norm": 0.49549728631973267,
+      "learning_rate": 9.011655878430019e-05,
+      "loss": 0.9671,
+      "step": 2546
+    },
+    {
+      "epoch": 0.22757835012397526,
+      "grad_norm": 0.4547901153564453,
+      "learning_rate": 9.010791983550923e-05,
+      "loss": 0.9462,
+      "step": 2547
+    },
+    {
+      "epoch": 0.22766770165523711,
+      "grad_norm": 0.40302136540412903,
+      "learning_rate": 9.009927752725247e-05,
+      "loss": 1.0536,
+      "step": 2548
+    },
+    {
+      "epoch": 0.22775705318649897,
+      "grad_norm": 0.4038864076137543,
+      "learning_rate": 9.009063186025379e-05,
+      "loss": 1.0665,
+      "step": 2549
+    },
+    {
+      "epoch": 0.22784640471776085,
+      "grad_norm": 0.4060629606246948,
+      "learning_rate": 9.008198283523737e-05,
+      "loss": 0.9799,
+      "step": 2550
+    },
+    {
+      "epoch": 0.2279357562490227,
+      "grad_norm": 0.49202004075050354,
+      "learning_rate": 9.007333045292764e-05,
+      "loss": 1.0176,
+      "step": 2551
+    },
+    {
+      "epoch": 0.2280251077802846,
+      "grad_norm": 0.45856913924217224,
+      "learning_rate": 9.006467471404932e-05,
+      "loss": 0.9464,
+      "step": 2552
+    },
+    {
+      "epoch": 0.22811445931154645,
+      "grad_norm": 0.44552847743034363,
+      "learning_rate": 9.005601561932745e-05,
+      "loss": 1.0197,
+      "step": 2553
+    },
+    {
+      "epoch": 0.22820381084280833,
+      "grad_norm": 0.40756505727767944,
+      "learning_rate": 9.00473531694873e-05,
+      "loss": 0.9694,
+      "step": 2554
+    },
+    {
+      "epoch": 0.22829316237407019,
+      "grad_norm": 0.4505196511745453,
+      "learning_rate": 9.003868736525444e-05,
+      "loss": 1.0085,
+      "step": 2555
+    },
+    {
+      "epoch": 0.22838251390533204,
+      "grad_norm": 0.40980064868927,
+      "learning_rate": 9.003001820735474e-05,
+      "loss": 1.0383,
+      "step": 2556
+    },
+    {
+      "epoch": 0.22847186543659392,
+      "grad_norm": 0.37999671697616577,
+      "learning_rate": 9.002134569651433e-05,
+      "loss": 1.026,
+      "step": 2557
+    },
+    {
+      "epoch": 0.22856121696785578,
+      "grad_norm": 0.4312988817691803,
+      "learning_rate": 9.00126698334596e-05,
+      "loss": 1.0029,
+      "step": 2558
+    },
+    {
+      "epoch": 0.22865056849911766,
+      "grad_norm": 0.45107796788215637,
+      "learning_rate": 9.000399061891728e-05,
+      "loss": 1.0555,
+      "step": 2559
+    },
+    {
+      "epoch": 0.22873992003037952,
+      "grad_norm": 0.39357370138168335,
+      "learning_rate": 8.999530805361434e-05,
+      "loss": 1.0421,
+      "step": 2560
+    },
+    {
+      "epoch": 0.2288292715616414,
+      "grad_norm": 0.4320610761642456,
+      "learning_rate": 8.998662213827802e-05,
+      "loss": 1.0491,
+      "step": 2561
+    },
+    {
+      "epoch": 0.22891862309290326,
+      "grad_norm": 0.4132126271724701,
+      "learning_rate": 8.997793287363588e-05,
+      "loss": 0.9682,
+      "step": 2562
+    },
+    {
+      "epoch": 0.2290079746241651,
+      "grad_norm": 0.38975366950035095,
+      "learning_rate": 8.996924026041573e-05,
+      "loss": 1.0131,
+      "step": 2563
+    },
+    {
+      "epoch": 0.229097326155427,
+      "grad_norm": 0.42048487067222595,
+      "learning_rate": 8.996054429934567e-05,
+      "loss": 1.0186,
+      "step": 2564
+    },
+    {
+      "epoch": 0.22918667768668885,
+      "grad_norm": 0.40763533115386963,
+      "learning_rate": 8.995184499115405e-05,
+      "loss": 0.9722,
+      "step": 2565
+    },
+    {
+      "epoch": 0.22927602921795073,
+      "grad_norm": 0.47947773337364197,
+      "learning_rate": 8.994314233656958e-05,
+      "loss": 0.9427,
+      "step": 2566
+    },
+    {
+      "epoch": 0.2293653807492126,
+      "grad_norm": 0.4561311602592468,
+      "learning_rate": 8.993443633632116e-05,
+      "loss": 0.9793,
+      "step": 2567
+    },
+    {
+      "epoch": 0.22945473228047444,
+      "grad_norm": 0.3957562744617462,
+      "learning_rate": 8.992572699113804e-05,
+      "loss": 1.0227,
+      "step": 2568
+    },
+    {
+      "epoch": 0.22954408381173633,
+      "grad_norm": 0.3969632685184479,
+      "learning_rate": 8.99170143017497e-05,
+      "loss": 1.001,
+      "step": 2569
+    },
+    {
+      "epoch": 0.22963343534299818,
+      "grad_norm": 0.4251578748226166,
+      "learning_rate": 8.990829826888592e-05,
+      "loss": 1.0243,
+      "step": 2570
+    },
+    {
+      "epoch": 0.22972278687426007,
+      "grad_norm": 0.48977455496788025,
+      "learning_rate": 8.989957889327678e-05,
+      "loss": 0.9637,
+      "step": 2571
+    },
+    {
+      "epoch": 0.22981213840552192,
+      "grad_norm": 0.42375892400741577,
+      "learning_rate": 8.98908561756526e-05,
+      "loss": 0.9662,
+      "step": 2572
+    },
+    {
+      "epoch": 0.2299014899367838,
+      "grad_norm": 0.3831101655960083,
+      "learning_rate": 8.988213011674402e-05,
+      "loss": 1.0247,
+      "step": 2573
+    },
+    {
+      "epoch": 0.22999084146804566,
+      "grad_norm": 0.4255763590335846,
+      "learning_rate": 8.987340071728192e-05,
+      "loss": 1.0308,
+      "step": 2574
+    },
+    {
+      "epoch": 0.23008019299930751,
+      "grad_norm": 0.4933546185493469,
+      "learning_rate": 8.986466797799749e-05,
+      "loss": 0.9334,
+      "step": 2575
+    },
+    {
+      "epoch": 0.2301695445305694,
+      "grad_norm": 0.4253925085067749,
+      "learning_rate": 8.985593189962221e-05,
+      "loss": 1.0554,
+      "step": 2576
+    },
+    {
+      "epoch": 0.23025889606183125,
+      "grad_norm": 0.4756745398044586,
+      "learning_rate": 8.984719248288778e-05,
+      "loss": 0.9267,
+      "step": 2577
+    },
+    {
+      "epoch": 0.23034824759309314,
+      "grad_norm": 0.41270750761032104,
+      "learning_rate": 8.983844972852625e-05,
+      "loss": 0.9618,
+      "step": 2578
+    },
+    {
+      "epoch": 0.230437599124355,
+      "grad_norm": 0.4245428144931793,
+      "learning_rate": 8.982970363726989e-05,
+      "loss": 1.0846,
+      "step": 2579
+    },
+    {
+      "epoch": 0.23052695065561685,
+      "grad_norm": 0.595262348651886,
+      "learning_rate": 8.98209542098513e-05,
+      "loss": 0.9651,
+      "step": 2580
+    },
+    {
+      "epoch": 0.23061630218687873,
+      "grad_norm": 0.4939884543418884,
+      "learning_rate": 8.981220144700335e-05,
+      "loss": 1.0355,
+      "step": 2581
+    },
+    {
+      "epoch": 0.23070565371814059,
+      "grad_norm": 0.430448979139328,
+      "learning_rate": 8.980344534945915e-05,
+      "loss": 1.0304,
+      "step": 2582
+    },
+    {
+      "epoch": 0.23079500524940247,
+      "grad_norm": 0.4117549657821655,
+      "learning_rate": 8.979468591795213e-05,
+      "loss": 1.024,
+      "step": 2583
+    },
+    {
+      "epoch": 0.23088435678066432,
+      "grad_norm": 0.3863102197647095,
+      "learning_rate": 8.978592315321597e-05,
+      "loss": 1.0221,
+      "step": 2584
+    },
+    {
+      "epoch": 0.2309737083119262,
+      "grad_norm": 0.37732282280921936,
+      "learning_rate": 8.977715705598469e-05,
+      "loss": 1.0347,
+      "step": 2585
+    },
+    {
+      "epoch": 0.23106305984318806,
+      "grad_norm": 0.3825008273124695,
+      "learning_rate": 8.976838762699249e-05,
+      "loss": 1.1106,
+      "step": 2586
+    },
+    {
+      "epoch": 0.23115241137444992,
+      "grad_norm": 0.38581812381744385,
+      "learning_rate": 8.975961486697392e-05,
+      "loss": 0.9931,
+      "step": 2587
+    },
+    {
+      "epoch": 0.2312417629057118,
+      "grad_norm": 0.48402634263038635,
+      "learning_rate": 8.975083877666382e-05,
+      "loss": 0.9774,
+      "step": 2588
+    },
+    {
+      "epoch": 0.23133111443697366,
+      "grad_norm": 0.4185434877872467,
+      "learning_rate": 8.974205935679725e-05,
+      "loss": 1.0566,
+      "step": 2589
+    },
+    {
+      "epoch": 0.23142046596823554,
+      "grad_norm": 0.37079721689224243,
+      "learning_rate": 8.973327660810958e-05,
+      "loss": 1.0591,
+      "step": 2590
+    },
+    {
+      "epoch": 0.2315098174994974,
+      "grad_norm": 0.44048118591308594,
+      "learning_rate": 8.972449053133647e-05,
+      "loss": 1.0344,
+      "step": 2591
+    },
+    {
+      "epoch": 0.23159916903075928,
+      "grad_norm": 0.3690267503261566,
+      "learning_rate": 8.971570112721385e-05,
+      "loss": 1.0284,
+      "step": 2592
+    },
+    {
+      "epoch": 0.23168852056202113,
+      "grad_norm": 0.38595226407051086,
+      "learning_rate": 8.970690839647792e-05,
+      "loss": 1.0338,
+      "step": 2593
+    },
+    {
+      "epoch": 0.231777872093283,
+      "grad_norm": 0.427207350730896,
+      "learning_rate": 8.969811233986519e-05,
+      "loss": 0.9762,
+      "step": 2594
+    },
+    {
+      "epoch": 0.23186722362454487,
+      "grad_norm": 0.4339447617530823,
+      "learning_rate": 8.968931295811236e-05,
+      "loss": 0.9732,
+      "step": 2595
+    },
+    {
+      "epoch": 0.23195657515580673,
+      "grad_norm": 0.5293307304382324,
+      "learning_rate": 8.968051025195653e-05,
+      "loss": 0.9787,
+      "step": 2596
+    },
+    {
+      "epoch": 0.2320459266870686,
+      "grad_norm": 0.44257649779319763,
+      "learning_rate": 8.9671704222135e-05,
+      "loss": 0.9879,
+      "step": 2597
+    },
+    {
+      "epoch": 0.23213527821833047,
+      "grad_norm": 0.4917474091053009,
+      "learning_rate": 8.966289486938538e-05,
+      "loss": 1.0853,
+      "step": 2598
+    },
+    {
+      "epoch": 0.23222462974959232,
+      "grad_norm": 0.41678449511528015,
+      "learning_rate": 8.965408219444554e-05,
+      "loss": 1.0143,
+      "step": 2599
+    },
+    {
+      "epoch": 0.2323139812808542,
+      "grad_norm": 0.49743688106536865,
+      "learning_rate": 8.964526619805362e-05,
+      "loss": 0.9425,
+      "step": 2600
+    },
+    {
+      "epoch": 0.23240333281211606,
+      "grad_norm": 0.4517749845981598,
+      "learning_rate": 8.963644688094807e-05,
+      "loss": 1.0329,
+      "step": 2601
+    },
+    {
+      "epoch": 0.23249268434337794,
+      "grad_norm": 0.48534515500068665,
+      "learning_rate": 8.96276242438676e-05,
+      "loss": 0.9492,
+      "step": 2602
+    },
+    {
+      "epoch": 0.2325820358746398,
+      "grad_norm": 0.44116154313087463,
+      "learning_rate": 8.96187982875512e-05,
+      "loss": 0.9327,
+      "step": 2603
+    },
+    {
+      "epoch": 0.23267138740590168,
+      "grad_norm": 0.377189576625824,
+      "learning_rate": 8.960996901273815e-05,
+      "loss": 1.0739,
+      "step": 2604
+    },
+    {
+      "epoch": 0.23276073893716354,
+      "grad_norm": 0.40555354952812195,
+      "learning_rate": 8.960113642016797e-05,
+      "loss": 1.0098,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2328500904684254,
+      "grad_norm": 0.64093416929245,
+      "learning_rate": 8.95923005105805e-05,
+      "loss": 0.8996,
+      "step": 2606
+    },
+    {
+      "epoch": 0.23293944199968727,
+      "grad_norm": 0.4546898305416107,
+      "learning_rate": 8.958346128471584e-05,
+      "loss": 0.979,
+      "step": 2607
+    },
+    {
+      "epoch": 0.23302879353094913,
+      "grad_norm": 0.4061562120914459,
+      "learning_rate": 8.957461874331436e-05,
+      "loss": 1.0129,
+      "step": 2608
+    },
+    {
+      "epoch": 0.233118145062211,
+      "grad_norm": 0.45059117674827576,
+      "learning_rate": 8.956577288711673e-05,
+      "loss": 1.0107,
+      "step": 2609
+    },
+    {
+      "epoch": 0.23320749659347287,
+      "grad_norm": 0.4028307795524597,
+      "learning_rate": 8.955692371686388e-05,
+      "loss": 1.0118,
+      "step": 2610
+    },
+    {
+      "epoch": 0.23329684812473475,
+      "grad_norm": 0.5479282736778259,
+      "learning_rate": 8.954807123329704e-05,
+      "loss": 0.9751,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2333861996559966,
+      "grad_norm": 0.4075985550880432,
+      "learning_rate": 8.953921543715767e-05,
+      "loss": 1.0192,
+      "step": 2612
+    },
+    {
+      "epoch": 0.23347555118725846,
+      "grad_norm": 0.41962021589279175,
+      "learning_rate": 8.953035632918754e-05,
+      "loss": 0.9887,
+      "step": 2613
+    },
+    {
+      "epoch": 0.23356490271852035,
+      "grad_norm": 0.46383902430534363,
+      "learning_rate": 8.952149391012872e-05,
+      "loss": 1.0025,
+      "step": 2614
+    },
+    {
+      "epoch": 0.2336542542497822,
+      "grad_norm": 0.51430743932724,
+      "learning_rate": 8.95126281807235e-05,
+      "loss": 1.0262,
+      "step": 2615
+    },
+    {
+      "epoch": 0.23374360578104408,
+      "grad_norm": 0.45170411467552185,
+      "learning_rate": 8.95037591417145e-05,
+      "loss": 0.9814,
+      "step": 2616
+    },
+    {
+      "epoch": 0.23383295731230594,
+      "grad_norm": 0.4477570652961731,
+      "learning_rate": 8.94948867938446e-05,
+      "loss": 1.0245,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2339223088435678,
+      "grad_norm": 0.3824307918548584,
+      "learning_rate": 8.948601113785693e-05,
+      "loss": 1.0084,
+      "step": 2618
+    },
+    {
+      "epoch": 0.23401166037482968,
+      "grad_norm": 0.5118735432624817,
+      "learning_rate": 8.947713217449495e-05,
+      "loss": 0.9164,
+      "step": 2619
+    },
+    {
+      "epoch": 0.23410101190609153,
+      "grad_norm": 0.4788876175880432,
+      "learning_rate": 8.946824990450236e-05,
+      "loss": 1.0392,
+      "step": 2620
+    },
+    {
+      "epoch": 0.23419036343735342,
+      "grad_norm": 0.43572187423706055,
+      "learning_rate": 8.945936432862312e-05,
+      "loss": 0.9694,
+      "step": 2621
+    },
+    {
+      "epoch": 0.23427971496861527,
+      "grad_norm": 0.437338650226593,
+      "learning_rate": 8.945047544760153e-05,
+      "loss": 0.9954,
+      "step": 2622
+    },
+    {
+      "epoch": 0.23436906649987715,
+      "grad_norm": 0.40690669417381287,
+      "learning_rate": 8.944158326218208e-05,
+      "loss": 1.0038,
+      "step": 2623
+    },
+    {
+      "epoch": 0.234458418031139,
+      "grad_norm": 0.40458807349205017,
+      "learning_rate": 8.943268777310964e-05,
+      "loss": 0.9944,
+      "step": 2624
+    },
+    {
+      "epoch": 0.23454776956240087,
+      "grad_norm": 0.396119624376297,
+      "learning_rate": 8.942378898112928e-05,
+      "loss": 0.9914,
+      "step": 2625
+    },
+    {
+      "epoch": 0.23463712109366275,
+      "grad_norm": 0.4142768085002899,
+      "learning_rate": 8.941488688698634e-05,
+      "loss": 1.1052,
+      "step": 2626
+    },
+    {
+      "epoch": 0.2347264726249246,
+      "grad_norm": 0.516160786151886,
+      "learning_rate": 8.940598149142652e-05,
+      "loss": 1.0511,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2348158241561865,
+      "grad_norm": 0.41457611322402954,
+      "learning_rate": 8.93970727951957e-05,
+      "loss": 0.9767,
+      "step": 2628
+    },
+    {
+      "epoch": 0.23490517568744834,
+      "grad_norm": 0.41176801919937134,
+      "learning_rate": 8.938816079904009e-05,
+      "loss": 1.0566,
+      "step": 2629
+    },
+    {
+      "epoch": 0.2349945272187102,
+      "grad_norm": 0.42432960867881775,
+      "learning_rate": 8.937924550370618e-05,
+      "loss": 1.0717,
+      "step": 2630
+    },
+    {
+      "epoch": 0.23508387874997208,
+      "grad_norm": 0.5278024077415466,
+      "learning_rate": 8.937032690994068e-05,
+      "loss": 0.9483,
+      "step": 2631
+    },
+    {
+      "epoch": 0.23517323028123394,
+      "grad_norm": 0.39076733589172363,
+      "learning_rate": 8.936140501849066e-05,
+      "loss": 1.0294,
+      "step": 2632
+    },
+    {
+      "epoch": 0.23526258181249582,
+      "grad_norm": 0.41586795449256897,
+      "learning_rate": 8.935247983010339e-05,
+      "loss": 0.9949,
+      "step": 2633
+    },
+    {
+      "epoch": 0.23535193334375767,
+      "grad_norm": 0.4252259135246277,
+      "learning_rate": 8.93435513455265e-05,
+      "loss": 0.9886,
+      "step": 2634
+    },
+    {
+      "epoch": 0.23544128487501956,
+      "grad_norm": 0.41890689730644226,
+      "learning_rate": 8.93346195655078e-05,
+      "loss": 1.0233,
+      "step": 2635
+    },
+    {
+      "epoch": 0.2355306364062814,
+      "grad_norm": 0.38035479187965393,
+      "learning_rate": 8.932568449079541e-05,
+      "loss": 1.0627,
+      "step": 2636
+    },
+    {
+      "epoch": 0.23561998793754327,
+      "grad_norm": 0.414369136095047,
+      "learning_rate": 8.931674612213778e-05,
+      "loss": 0.974,
+      "step": 2637
+    },
+    {
+      "epoch": 0.23570933946880515,
+      "grad_norm": 0.3929893374443054,
+      "learning_rate": 8.930780446028359e-05,
+      "loss": 1.031,
+      "step": 2638
+    },
+    {
+      "epoch": 0.235798691000067,
+      "grad_norm": 0.45517128705978394,
+      "learning_rate": 8.929885950598177e-05,
+      "loss": 0.996,
+      "step": 2639
+    },
+    {
+      "epoch": 0.2358880425313289,
+      "grad_norm": 0.39736053347587585,
+      "learning_rate": 8.928991125998157e-05,
+      "loss": 1.001,
+      "step": 2640
+    },
+    {
+      "epoch": 0.23597739406259075,
+      "grad_norm": 0.38671961426734924,
+      "learning_rate": 8.92809597230325e-05,
+      "loss": 1.0809,
+      "step": 2641
+    },
+    {
+      "epoch": 0.23606674559385263,
+      "grad_norm": 0.4453861117362976,
+      "learning_rate": 8.927200489588435e-05,
+      "loss": 1.0721,
+      "step": 2642
+    },
+    {
+      "epoch": 0.23615609712511448,
+      "grad_norm": 0.4946734309196472,
+      "learning_rate": 8.926304677928718e-05,
+      "loss": 0.9619,
+      "step": 2643
+    },
+    {
+      "epoch": 0.23624544865637634,
+      "grad_norm": 0.34986773133277893,
+      "learning_rate": 8.925408537399133e-05,
+      "loss": 1.0727,
+      "step": 2644
+    },
+    {
+      "epoch": 0.23633480018763822,
+      "grad_norm": 0.5045349597930908,
+      "learning_rate": 8.924512068074742e-05,
+      "loss": 0.9678,
+      "step": 2645
+    },
+    {
+      "epoch": 0.23642415171890008,
+      "grad_norm": 0.4377862215042114,
+      "learning_rate": 8.923615270030632e-05,
+      "loss": 0.9723,
+      "step": 2646
+    },
+    {
+      "epoch": 0.23651350325016196,
+      "grad_norm": 0.5501147508621216,
+      "learning_rate": 8.922718143341921e-05,
+      "loss": 0.9724,
+      "step": 2647
+    },
+    {
+      "epoch": 0.23660285478142382,
+      "grad_norm": 0.42605921626091003,
+      "learning_rate": 8.921820688083753e-05,
+      "loss": 1.0548,
+      "step": 2648
+    },
+    {
+      "epoch": 0.23669220631268567,
+      "grad_norm": 0.36013084650039673,
+      "learning_rate": 8.920922904331297e-05,
+      "loss": 1.0721,
+      "step": 2649
+    },
+    {
+      "epoch": 0.23678155784394755,
+      "grad_norm": 0.4354332685470581,
+      "learning_rate": 8.920024792159754e-05,
+      "loss": 0.9691,
+      "step": 2650
+    },
+    {
+      "epoch": 0.2368709093752094,
+      "grad_norm": 0.45081332325935364,
+      "learning_rate": 8.919126351644351e-05,
+      "loss": 0.8911,
+      "step": 2651
+    },
+    {
+      "epoch": 0.2369602609064713,
+      "grad_norm": 0.41693955659866333,
+      "learning_rate": 8.918227582860341e-05,
+      "loss": 1.0437,
+      "step": 2652
+    },
+    {
+      "epoch": 0.23704961243773315,
+      "grad_norm": 0.4205388128757477,
+      "learning_rate": 8.917328485883005e-05,
+      "loss": 1.0024,
+      "step": 2653
+    },
+    {
+      "epoch": 0.23713896396899503,
+      "grad_norm": 0.3820997476577759,
+      "learning_rate": 8.916429060787654e-05,
+      "loss": 1.0151,
+      "step": 2654
+    },
+    {
+      "epoch": 0.2372283155002569,
+      "grad_norm": 0.5409601330757141,
+      "learning_rate": 8.915529307649621e-05,
+      "loss": 0.931,
+      "step": 2655
+    },
+    {
+      "epoch": 0.23731766703151874,
+      "grad_norm": 0.40477868914604187,
+      "learning_rate": 8.914629226544273e-05,
+      "loss": 1.0052,
+      "step": 2656
+    },
+    {
+      "epoch": 0.23740701856278063,
+      "grad_norm": 0.5156925320625305,
+      "learning_rate": 8.913728817547002e-05,
+      "loss": 0.9712,
+      "step": 2657
+    },
+    {
+      "epoch": 0.23749637009404248,
+      "grad_norm": 0.4694739282131195,
+      "learning_rate": 8.912828080733223e-05,
+      "loss": 0.9844,
+      "step": 2658
+    },
+    {
+      "epoch": 0.23758572162530436,
+      "grad_norm": 0.520553708076477,
+      "learning_rate": 8.911927016178385e-05,
+      "loss": 0.9508,
+      "step": 2659
+    },
+    {
+      "epoch": 0.23767507315656622,
+      "grad_norm": 0.4804941415786743,
+      "learning_rate": 8.911025623957961e-05,
+      "loss": 0.9502,
+      "step": 2660
+    },
+    {
+      "epoch": 0.23776442468782807,
+      "grad_norm": 0.49052369594573975,
+      "learning_rate": 8.910123904147452e-05,
+      "loss": 0.9866,
+      "step": 2661
+    },
+    {
+      "epoch": 0.23785377621908996,
+      "grad_norm": 0.45931488275527954,
+      "learning_rate": 8.909221856822388e-05,
+      "loss": 0.9628,
+      "step": 2662
+    },
+    {
+      "epoch": 0.2379431277503518,
+      "grad_norm": 0.43489325046539307,
+      "learning_rate": 8.908319482058325e-05,
+      "loss": 1.0695,
+      "step": 2663
+    },
+    {
+      "epoch": 0.2380324792816137,
+      "grad_norm": 0.47782567143440247,
+      "learning_rate": 8.907416779930843e-05,
+      "loss": 0.9529,
+      "step": 2664
+    },
+    {
+      "epoch": 0.23812183081287555,
+      "grad_norm": 0.4324779808521271,
+      "learning_rate": 8.906513750515559e-05,
+      "loss": 0.9556,
+      "step": 2665
+    },
+    {
+      "epoch": 0.23821118234413743,
+      "grad_norm": 0.4681794047355652,
+      "learning_rate": 8.905610393888106e-05,
+      "loss": 0.9987,
+      "step": 2666
+    },
+    {
+      "epoch": 0.2383005338753993,
+      "grad_norm": 0.443891704082489,
+      "learning_rate": 8.904706710124152e-05,
+      "loss": 1.0109,
+      "step": 2667
+    },
+    {
+      "epoch": 0.23838988540666115,
+      "grad_norm": 0.46368691325187683,
+      "learning_rate": 8.90380269929939e-05,
+      "loss": 0.9511,
+      "step": 2668
+    },
+    {
+      "epoch": 0.23847923693792303,
+      "grad_norm": 0.5165659189224243,
+      "learning_rate": 8.90289836148954e-05,
+      "loss": 0.9759,
+      "step": 2669
+    },
+    {
+      "epoch": 0.23856858846918488,
+      "grad_norm": 0.42744940519332886,
+      "learning_rate": 8.90199369677035e-05,
+      "loss": 0.9876,
+      "step": 2670
+    },
+    {
+      "epoch": 0.23865794000044677,
+      "grad_norm": 0.41597381234169006,
+      "learning_rate": 8.901088705217598e-05,
+      "loss": 1.0021,
+      "step": 2671
+    },
+    {
+      "epoch": 0.23874729153170862,
+      "grad_norm": 0.3914920687675476,
+      "learning_rate": 8.900183386907082e-05,
+      "loss": 1.1311,
+      "step": 2672
+    },
+    {
+      "epoch": 0.2388366430629705,
+      "grad_norm": 0.49854883551597595,
+      "learning_rate": 8.899277741914633e-05,
+      "loss": 0.9727,
+      "step": 2673
+    },
+    {
+      "epoch": 0.23892599459423236,
+      "grad_norm": 0.4740012586116791,
+      "learning_rate": 8.898371770316111e-05,
+      "loss": 0.9915,
+      "step": 2674
+    },
+    {
+      "epoch": 0.23901534612549422,
+      "grad_norm": 0.4568859040737152,
+      "learning_rate": 8.897465472187401e-05,
+      "loss": 0.981,
+      "step": 2675
+    },
+    {
+      "epoch": 0.2391046976567561,
+      "grad_norm": 0.4102030098438263,
+      "learning_rate": 8.896558847604414e-05,
+      "loss": 0.9928,
+      "step": 2676
+    },
+    {
+      "epoch": 0.23919404918801795,
+      "grad_norm": 0.50264573097229,
+      "learning_rate": 8.895651896643088e-05,
+      "loss": 0.9527,
+      "step": 2677
+    },
+    {
+      "epoch": 0.23928340071927984,
+      "grad_norm": 0.4080445170402527,
+      "learning_rate": 8.894744619379391e-05,
+      "loss": 1.1131,
+      "step": 2678
+    },
+    {
+      "epoch": 0.2393727522505417,
+      "grad_norm": 0.45948582887649536,
+      "learning_rate": 8.893837015889317e-05,
+      "loss": 0.9385,
+      "step": 2679
+    },
+    {
+      "epoch": 0.23946210378180355,
+      "grad_norm": 0.4374150335788727,
+      "learning_rate": 8.892929086248888e-05,
+      "loss": 0.9693,
+      "step": 2680
+    },
+    {
+      "epoch": 0.23955145531306543,
+      "grad_norm": 0.4575091302394867,
+      "learning_rate": 8.892020830534152e-05,
+      "loss": 0.9821,
+      "step": 2681
+    },
+    {
+      "epoch": 0.2396408068443273,
+      "grad_norm": 0.39326122403144836,
+      "learning_rate": 8.891112248821186e-05,
+      "loss": 1.0787,
+      "step": 2682
+    },
+    {
+      "epoch": 0.23973015837558917,
+      "grad_norm": 0.4379619061946869,
+      "learning_rate": 8.890203341186092e-05,
+      "loss": 1.0188,
+      "step": 2683
+    },
+    {
+      "epoch": 0.23981950990685102,
+      "grad_norm": 0.47353631258010864,
+      "learning_rate": 8.889294107705002e-05,
+      "loss": 1.0197,
+      "step": 2684
+    },
+    {
+      "epoch": 0.2399088614381129,
+      "grad_norm": 0.4937237799167633,
+      "learning_rate": 8.888384548454075e-05,
+      "loss": 0.9871,
+      "step": 2685
+    },
+    {
+      "epoch": 0.23999821296937476,
+      "grad_norm": 0.39890146255493164,
+      "learning_rate": 8.887474663509493e-05,
+      "loss": 1.0274,
+      "step": 2686
+    },
+    {
+      "epoch": 0.24008756450063662,
+      "grad_norm": 0.3992265462875366,
+      "learning_rate": 8.886564452947471e-05,
+      "loss": 1.0258,
+      "step": 2687
+    },
+    {
+      "epoch": 0.2401769160318985,
+      "grad_norm": 0.4458499550819397,
+      "learning_rate": 8.885653916844248e-05,
+      "loss": 0.9608,
+      "step": 2688
+    },
+    {
+      "epoch": 0.24026626756316036,
+      "grad_norm": 0.39621952176094055,
+      "learning_rate": 8.884743055276092e-05,
+      "loss": 0.9997,
+      "step": 2689
+    },
+    {
+      "epoch": 0.24035561909442224,
+      "grad_norm": 0.44387656450271606,
+      "learning_rate": 8.883831868319297e-05,
+      "loss": 1.0045,
+      "step": 2690
+    },
+    {
+      "epoch": 0.2404449706256841,
+      "grad_norm": 0.38763174414634705,
+      "learning_rate": 8.882920356050184e-05,
+      "loss": 1.0502,
+      "step": 2691
+    },
+    {
+      "epoch": 0.24053432215694595,
+      "grad_norm": 0.4442698657512665,
+      "learning_rate": 8.882008518545101e-05,
+      "loss": 0.9964,
+      "step": 2692
+    },
+    {
+      "epoch": 0.24062367368820783,
+      "grad_norm": 0.39915651082992554,
+      "learning_rate": 8.881096355880428e-05,
+      "loss": 1.0732,
+      "step": 2693
+    },
+    {
+      "epoch": 0.2407130252194697,
+      "grad_norm": 0.458362877368927,
+      "learning_rate": 8.880183868132563e-05,
+      "loss": 1.0264,
+      "step": 2694
+    },
+    {
+      "epoch": 0.24080237675073157,
+      "grad_norm": 0.4793623685836792,
+      "learning_rate": 8.87927105537794e-05,
+      "loss": 0.931,
+      "step": 2695
+    },
+    {
+      "epoch": 0.24089172828199343,
+      "grad_norm": 0.4859834909439087,
+      "learning_rate": 8.878357917693016e-05,
+      "loss": 0.9891,
+      "step": 2696
+    },
+    {
+      "epoch": 0.2409810798132553,
+      "grad_norm": 0.4228796362876892,
+      "learning_rate": 8.877444455154278e-05,
+      "loss": 1.0693,
+      "step": 2697
+    },
+    {
+      "epoch": 0.24107043134451717,
+      "grad_norm": 0.37542724609375,
+      "learning_rate": 8.876530667838234e-05,
+      "loss": 1.0717,
+      "step": 2698
+    },
+    {
+      "epoch": 0.24115978287577902,
+      "grad_norm": 0.44685491919517517,
+      "learning_rate": 8.875616555821426e-05,
+      "loss": 1.0261,
+      "step": 2699
+    },
+    {
+      "epoch": 0.2412491344070409,
+      "grad_norm": 0.37416496872901917,
+      "learning_rate": 8.874702119180421e-05,
+      "loss": 1.1008,
+      "step": 2700
+    },
+    {
+      "epoch": 0.24133848593830276,
+      "grad_norm": 0.3862450420856476,
+      "learning_rate": 8.873787357991812e-05,
+      "loss": 1.0433,
+      "step": 2701
+    },
+    {
+      "epoch": 0.24142783746956464,
+      "grad_norm": 0.4019640386104584,
+      "learning_rate": 8.87287227233222e-05,
+      "loss": 1.0231,
+      "step": 2702
+    },
+    {
+      "epoch": 0.2415171890008265,
+      "grad_norm": 0.4376417398452759,
+      "learning_rate": 8.871956862278294e-05,
+      "loss": 1.0422,
+      "step": 2703
+    },
+    {
+      "epoch": 0.24160654053208838,
+      "grad_norm": 0.39731988310813904,
+      "learning_rate": 8.871041127906707e-05,
+      "loss": 1.0429,
+      "step": 2704
+    },
+    {
+      "epoch": 0.24169589206335024,
+      "grad_norm": 0.44788575172424316,
+      "learning_rate": 8.870125069294166e-05,
+      "loss": 1.0384,
+      "step": 2705
+    },
+    {
+      "epoch": 0.2417852435946121,
+      "grad_norm": 0.43335989117622375,
+      "learning_rate": 8.869208686517395e-05,
+      "loss": 1.0378,
+      "step": 2706
+    },
+    {
+      "epoch": 0.24187459512587398,
+      "grad_norm": 0.4897671639919281,
+      "learning_rate": 8.868291979653154e-05,
+      "loss": 0.9621,
+      "step": 2707
+    },
+    {
+      "epoch": 0.24196394665713583,
+      "grad_norm": 0.4613479971885681,
+      "learning_rate": 8.867374948778228e-05,
+      "loss": 0.9983,
+      "step": 2708
+    },
+    {
+      "epoch": 0.24205329818839771,
+      "grad_norm": 0.5041788220405579,
+      "learning_rate": 8.866457593969427e-05,
+      "loss": 1.0287,
+      "step": 2709
+    },
+    {
+      "epoch": 0.24214264971965957,
+      "grad_norm": 0.45073777437210083,
+      "learning_rate": 8.865539915303588e-05,
+      "loss": 0.9705,
+      "step": 2710
+    },
+    {
+      "epoch": 0.24223200125092142,
+      "grad_norm": 0.39561158418655396,
+      "learning_rate": 8.864621912857578e-05,
+      "loss": 1.0513,
+      "step": 2711
+    },
+    {
+      "epoch": 0.2423213527821833,
+      "grad_norm": 0.40363892912864685,
+      "learning_rate": 8.86370358670829e-05,
+      "loss": 0.9696,
+      "step": 2712
+    },
+    {
+      "epoch": 0.24241070431344516,
+      "grad_norm": 0.4564259648323059,
+      "learning_rate": 8.86278493693264e-05,
+      "loss": 1.0501,
+      "step": 2713
+    },
+    {
+      "epoch": 0.24250005584470705,
+      "grad_norm": 0.42725691199302673,
+      "learning_rate": 8.861865963607578e-05,
+      "loss": 0.9827,
+      "step": 2714
+    },
+    {
+      "epoch": 0.2425894073759689,
+      "grad_norm": 0.4821705222129822,
+      "learning_rate": 8.860946666810079e-05,
+      "loss": 0.952,
+      "step": 2715
+    },
+    {
+      "epoch": 0.24267875890723078,
+      "grad_norm": 0.42694029211997986,
+      "learning_rate": 8.86002704661714e-05,
+      "loss": 0.9642,
+      "step": 2716
+    },
+    {
+      "epoch": 0.24276811043849264,
+      "grad_norm": 0.45525282621383667,
+      "learning_rate": 8.85910710310579e-05,
+      "loss": 0.9894,
+      "step": 2717
+    },
+    {
+      "epoch": 0.2428574619697545,
+      "grad_norm": 0.4108658730983734,
+      "learning_rate": 8.858186836353087e-05,
+      "loss": 0.966,
+      "step": 2718
+    },
+    {
+      "epoch": 0.24294681350101638,
+      "grad_norm": 0.40456750988960266,
+      "learning_rate": 8.857266246436111e-05,
+      "loss": 1.0883,
+      "step": 2719
+    },
+    {
+      "epoch": 0.24303616503227823,
+      "grad_norm": 0.4102852940559387,
+      "learning_rate": 8.856345333431971e-05,
+      "loss": 1.0069,
+      "step": 2720
+    },
+    {
+      "epoch": 0.24312551656354012,
+      "grad_norm": 0.40076547861099243,
+      "learning_rate": 8.855424097417802e-05,
+      "loss": 1.0423,
+      "step": 2721
+    },
+    {
+      "epoch": 0.24321486809480197,
+      "grad_norm": 0.3830002248287201,
+      "learning_rate": 8.854502538470771e-05,
+      "loss": 1.003,
+      "step": 2722
+    },
+    {
+      "epoch": 0.24330421962606383,
+      "grad_norm": 0.4300539493560791,
+      "learning_rate": 8.853580656668065e-05,
+      "loss": 1.0192,
+      "step": 2723
+    },
+    {
+      "epoch": 0.2433935711573257,
+      "grad_norm": 0.3924980163574219,
+      "learning_rate": 8.852658452086904e-05,
+      "loss": 1.021,
+      "step": 2724
+    },
+    {
+      "epoch": 0.24348292268858757,
+      "grad_norm": 0.3826366364955902,
+      "learning_rate": 8.851735924804531e-05,
+      "loss": 1.0014,
+      "step": 2725
+    },
+    {
+      "epoch": 0.24357227421984945,
+      "grad_norm": 0.48502808809280396,
+      "learning_rate": 8.850813074898217e-05,
+      "loss": 1.0423,
+      "step": 2726
+    },
+    {
+      "epoch": 0.2436616257511113,
+      "grad_norm": 0.40080320835113525,
+      "learning_rate": 8.849889902445263e-05,
+      "loss": 1.0273,
+      "step": 2727
+    },
+    {
+      "epoch": 0.2437509772823732,
+      "grad_norm": 0.4349658489227295,
+      "learning_rate": 8.848966407522992e-05,
+      "loss": 1.0189,
+      "step": 2728
+    },
+    {
+      "epoch": 0.24384032881363504,
+      "grad_norm": 0.4866269826889038,
+      "learning_rate": 8.848042590208756e-05,
+      "loss": 0.9429,
+      "step": 2729
+    },
+    {
+      "epoch": 0.2439296803448969,
+      "grad_norm": 0.38999444246292114,
+      "learning_rate": 8.847118450579937e-05,
+      "loss": 0.964,
+      "step": 2730
+    },
+    {
+      "epoch": 0.24401903187615878,
+      "grad_norm": 0.39741405844688416,
+      "learning_rate": 8.84619398871394e-05,
+      "loss": 1.038,
+      "step": 2731
+    },
+    {
+      "epoch": 0.24410838340742064,
+      "grad_norm": 0.42767640948295593,
+      "learning_rate": 8.845269204688199e-05,
+      "loss": 1.041,
+      "step": 2732
+    },
+    {
+      "epoch": 0.24419773493868252,
+      "grad_norm": 0.42031916975975037,
+      "learning_rate": 8.844344098580176e-05,
+      "loss": 0.9304,
+      "step": 2733
+    },
+    {
+      "epoch": 0.24428708646994438,
+      "grad_norm": 0.34102049469947815,
+      "learning_rate": 8.843418670467353e-05,
+      "loss": 1.021,
+      "step": 2734
+    },
+    {
+      "epoch": 0.24437643800120626,
+      "grad_norm": 0.387407511472702,
+      "learning_rate": 8.842492920427252e-05,
+      "loss": 1.0539,
+      "step": 2735
+    },
+    {
+      "epoch": 0.24446578953246811,
+      "grad_norm": 0.373221218585968,
+      "learning_rate": 8.84156684853741e-05,
+      "loss": 1.0037,
+      "step": 2736
+    },
+    {
+      "epoch": 0.24455514106372997,
+      "grad_norm": 0.3756154477596283,
+      "learning_rate": 8.840640454875396e-05,
+      "loss": 1.0619,
+      "step": 2737
+    },
+    {
+      "epoch": 0.24464449259499185,
+      "grad_norm": 0.47706297039985657,
+      "learning_rate": 8.839713739518807e-05,
+      "loss": 0.8904,
+      "step": 2738
+    },
+    {
+      "epoch": 0.2447338441262537,
+      "grad_norm": 0.49002113938331604,
+      "learning_rate": 8.838786702545262e-05,
+      "loss": 1.0568,
+      "step": 2739
+    },
+    {
+      "epoch": 0.2448231956575156,
+      "grad_norm": 0.4314875304698944,
+      "learning_rate": 8.837859344032413e-05,
+      "loss": 0.9789,
+      "step": 2740
+    },
+    {
+      "epoch": 0.24491254718877745,
+      "grad_norm": 0.4264450967311859,
+      "learning_rate": 8.836931664057935e-05,
+      "loss": 0.9799,
+      "step": 2741
+    },
+    {
+      "epoch": 0.2450018987200393,
+      "grad_norm": 0.4409767985343933,
+      "learning_rate": 8.836003662699533e-05,
+      "loss": 0.9571,
+      "step": 2742
+    },
+    {
+      "epoch": 0.24509125025130118,
+      "grad_norm": 0.47590428590774536,
+      "learning_rate": 8.835075340034933e-05,
+      "loss": 0.9076,
+      "step": 2743
+    },
+    {
+      "epoch": 0.24518060178256304,
+      "grad_norm": 0.4312697947025299,
+      "learning_rate": 8.834146696141895e-05,
+      "loss": 0.9393,
+      "step": 2744
+    },
+    {
+      "epoch": 0.24526995331382492,
+      "grad_norm": 0.42560774087905884,
+      "learning_rate": 8.833217731098203e-05,
+      "loss": 1.0549,
+      "step": 2745
+    },
+    {
+      "epoch": 0.24535930484508678,
+      "grad_norm": 0.44607481360435486,
+      "learning_rate": 8.832288444981666e-05,
+      "loss": 0.9967,
+      "step": 2746
+    },
+    {
+      "epoch": 0.24544865637634866,
+      "grad_norm": 0.49672216176986694,
+      "learning_rate": 8.831358837870122e-05,
+      "loss": 1.0662,
+      "step": 2747
+    },
+    {
+      "epoch": 0.24553800790761052,
+      "grad_norm": 0.44877761602401733,
+      "learning_rate": 8.830428909841437e-05,
+      "loss": 1.0782,
+      "step": 2748
+    },
+    {
+      "epoch": 0.24562735943887237,
+      "grad_norm": 0.5102913975715637,
+      "learning_rate": 8.829498660973501e-05,
+      "loss": 0.9089,
+      "step": 2749
+    },
+    {
+      "epoch": 0.24571671097013426,
+      "grad_norm": 0.4264692962169647,
+      "learning_rate": 8.828568091344234e-05,
+      "loss": 1.003,
+      "step": 2750
+    },
+    {
+      "epoch": 0.2458060625013961,
+      "grad_norm": 0.45728904008865356,
+      "learning_rate": 8.827637201031577e-05,
+      "loss": 0.9578,
+      "step": 2751
+    },
+    {
+      "epoch": 0.245895414032658,
+      "grad_norm": 0.4567210078239441,
+      "learning_rate": 8.826705990113506e-05,
+      "loss": 1.0127,
+      "step": 2752
+    },
+    {
+      "epoch": 0.24598476556391985,
+      "grad_norm": 0.4181562662124634,
+      "learning_rate": 8.825774458668019e-05,
+      "loss": 1.0222,
+      "step": 2753
+    },
+    {
+      "epoch": 0.24607411709518173,
+      "grad_norm": 0.415363609790802,
+      "learning_rate": 8.824842606773142e-05,
+      "loss": 0.9997,
+      "step": 2754
+    },
+    {
+      "epoch": 0.2461634686264436,
+      "grad_norm": 0.43965113162994385,
+      "learning_rate": 8.823910434506925e-05,
+      "loss": 1.0537,
+      "step": 2755
+    },
+    {
+      "epoch": 0.24625282015770544,
+      "grad_norm": 0.36847561597824097,
+      "learning_rate": 8.82297794194745e-05,
+      "loss": 1.0743,
+      "step": 2756
+    },
+    {
+      "epoch": 0.24634217168896733,
+      "grad_norm": 0.5313237309455872,
+      "learning_rate": 8.822045129172822e-05,
+      "loss": 1.0059,
+      "step": 2757
+    },
+    {
+      "epoch": 0.24643152322022918,
+      "grad_norm": 0.41933882236480713,
+      "learning_rate": 8.821111996261176e-05,
+      "loss": 0.9882,
+      "step": 2758
+    },
+    {
+      "epoch": 0.24652087475149106,
+      "grad_norm": 0.43498140573501587,
+      "learning_rate": 8.820178543290668e-05,
+      "loss": 1.0393,
+      "step": 2759
+    },
+    {
+      "epoch": 0.24661022628275292,
+      "grad_norm": 0.5251235365867615,
+      "learning_rate": 8.819244770339488e-05,
+      "loss": 1.0354,
+      "step": 2760
+    },
+    {
+      "epoch": 0.24669957781401478,
+      "grad_norm": 0.4627592861652374,
+      "learning_rate": 8.818310677485848e-05,
+      "loss": 0.9657,
+      "step": 2761
+    },
+    {
+      "epoch": 0.24678892934527666,
+      "grad_norm": 0.4479570984840393,
+      "learning_rate": 8.817376264807989e-05,
+      "loss": 0.9823,
+      "step": 2762
+    },
+    {
+      "epoch": 0.24687828087653851,
+      "grad_norm": 0.40790751576423645,
+      "learning_rate": 8.816441532384177e-05,
+      "loss": 1.0679,
+      "step": 2763
+    },
+    {
+      "epoch": 0.2469676324078004,
+      "grad_norm": 0.5192294120788574,
+      "learning_rate": 8.815506480292706e-05,
+      "loss": 1.0415,
+      "step": 2764
+    },
+    {
+      "epoch": 0.24705698393906225,
+      "grad_norm": 0.5589134097099304,
+      "learning_rate": 8.814571108611896e-05,
+      "loss": 0.9845,
+      "step": 2765
+    },
+    {
+      "epoch": 0.24714633547032414,
+      "grad_norm": 0.38889801502227783,
+      "learning_rate": 8.813635417420096e-05,
+      "loss": 1.066,
+      "step": 2766
+    },
+    {
+      "epoch": 0.247235687001586,
+      "grad_norm": 0.41053298115730286,
+      "learning_rate": 8.812699406795682e-05,
+      "loss": 0.99,
+      "step": 2767
+    },
+    {
+      "epoch": 0.24732503853284785,
+      "grad_norm": 0.4181362986564636,
+      "learning_rate": 8.81176307681705e-05,
+      "loss": 0.9906,
+      "step": 2768
+    },
+    {
+      "epoch": 0.24741439006410973,
+      "grad_norm": 0.42989280819892883,
+      "learning_rate": 8.810826427562629e-05,
+      "loss": 0.9261,
+      "step": 2769
+    },
+    {
+      "epoch": 0.24750374159537158,
+      "grad_norm": 0.3895682096481323,
+      "learning_rate": 8.809889459110875e-05,
+      "loss": 0.9863,
+      "step": 2770
+    },
+    {
+      "epoch": 0.24759309312663347,
+      "grad_norm": 0.4124111235141754,
+      "learning_rate": 8.808952171540268e-05,
+      "loss": 0.9897,
+      "step": 2771
+    },
+    {
+      "epoch": 0.24768244465789532,
+      "grad_norm": 0.4256143271923065,
+      "learning_rate": 8.808014564929316e-05,
+      "loss": 0.9599,
+      "step": 2772
+    },
+    {
+      "epoch": 0.24777179618915718,
+      "grad_norm": 0.4850587248802185,
+      "learning_rate": 8.807076639356556e-05,
+      "loss": 0.9145,
+      "step": 2773
+    },
+    {
+      "epoch": 0.24786114772041906,
+      "grad_norm": 0.46698129177093506,
+      "learning_rate": 8.806138394900544e-05,
+      "loss": 1.0141,
+      "step": 2774
+    },
+    {
+      "epoch": 0.24795049925168092,
+      "grad_norm": 0.4754186272621155,
+      "learning_rate": 8.805199831639872e-05,
+      "loss": 1.0272,
+      "step": 2775
+    },
+    {
+      "epoch": 0.2480398507829428,
+      "grad_norm": 0.4320179224014282,
+      "learning_rate": 8.804260949653154e-05,
+      "loss": 0.9598,
+      "step": 2776
+    },
+    {
+      "epoch": 0.24812920231420466,
+      "grad_norm": 0.5230510830879211,
+      "learning_rate": 8.80332174901903e-05,
+      "loss": 0.9854,
+      "step": 2777
+    },
+    {
+      "epoch": 0.24821855384546654,
+      "grad_norm": 0.4366409182548523,
+      "learning_rate": 8.80238222981617e-05,
+      "loss": 1.0322,
+      "step": 2778
+    },
+    {
+      "epoch": 0.2483079053767284,
+      "grad_norm": 0.46669307351112366,
+      "learning_rate": 8.801442392123267e-05,
+      "loss": 1.0078,
+      "step": 2779
+    },
+    {
+      "epoch": 0.24839725690799025,
+      "grad_norm": 0.570253849029541,
+      "learning_rate": 8.800502236019044e-05,
+      "loss": 1.0106,
+      "step": 2780
+    },
+    {
+      "epoch": 0.24848660843925213,
+      "grad_norm": 0.4147552251815796,
+      "learning_rate": 8.799561761582247e-05,
+      "loss": 1.0206,
+      "step": 2781
+    },
+    {
+      "epoch": 0.248575959970514,
+      "grad_norm": 0.39246666431427,
+      "learning_rate": 8.798620968891653e-05,
+      "loss": 1.0486,
+      "step": 2782
+    },
+    {
+      "epoch": 0.24866531150177587,
+      "grad_norm": 0.4565700590610504,
+      "learning_rate": 8.797679858026062e-05,
+      "loss": 0.9924,
+      "step": 2783
+    },
+    {
+      "epoch": 0.24875466303303773,
+      "grad_norm": 0.38775768876075745,
+      "learning_rate": 8.796738429064303e-05,
+      "loss": 1.0482,
+      "step": 2784
+    },
+    {
+      "epoch": 0.2488440145642996,
+      "grad_norm": 0.4036676287651062,
+      "learning_rate": 8.795796682085231e-05,
+      "loss": 1.0342,
+      "step": 2785
+    },
+    {
+      "epoch": 0.24893336609556146,
+      "grad_norm": 0.5216470956802368,
+      "learning_rate": 8.794854617167725e-05,
+      "loss": 1.0855,
+      "step": 2786
+    },
+    {
+      "epoch": 0.24902271762682332,
+      "grad_norm": 0.4413949251174927,
+      "learning_rate": 8.793912234390695e-05,
+      "loss": 0.9762,
+      "step": 2787
+    },
+    {
+      "epoch": 0.2491120691580852,
+      "grad_norm": 0.4308202862739563,
+      "learning_rate": 8.792969533833076e-05,
+      "loss": 0.9775,
+      "step": 2788
+    },
+    {
+      "epoch": 0.24920142068934706,
+      "grad_norm": 0.5357450246810913,
+      "learning_rate": 8.792026515573828e-05,
+      "loss": 1.0041,
+      "step": 2789
+    },
+    {
+      "epoch": 0.24929077222060894,
+      "grad_norm": 0.35964709520339966,
+      "learning_rate": 8.791083179691939e-05,
+      "loss": 1.012,
+      "step": 2790
+    },
+    {
+      "epoch": 0.2493801237518708,
+      "grad_norm": 0.535650908946991,
+      "learning_rate": 8.790139526266423e-05,
+      "loss": 1.0273,
+      "step": 2791
+    },
+    {
+      "epoch": 0.24946947528313265,
+      "grad_norm": 0.4403007924556732,
+      "learning_rate": 8.789195555376323e-05,
+      "loss": 0.9549,
+      "step": 2792
+    },
+    {
+      "epoch": 0.24955882681439454,
+      "grad_norm": 0.3707777261734009,
+      "learning_rate": 8.788251267100704e-05,
+      "loss": 1.0049,
+      "step": 2793
+    },
+    {
+      "epoch": 0.2496481783456564,
+      "grad_norm": 0.4524572491645813,
+      "learning_rate": 8.787306661518662e-05,
+      "loss": 1.0156,
+      "step": 2794
+    },
+    {
+      "epoch": 0.24973752987691827,
+      "grad_norm": 0.4172338545322418,
+      "learning_rate": 8.786361738709319e-05,
+      "loss": 0.9847,
+      "step": 2795
+    },
+    {
+      "epoch": 0.24982688140818013,
+      "grad_norm": 0.40239232778549194,
+      "learning_rate": 8.78541649875182e-05,
+      "loss": 1.0954,
+      "step": 2796
+    },
+    {
+      "epoch": 0.249916232939442,
+      "grad_norm": 0.44025593996047974,
+      "learning_rate": 8.784470941725338e-05,
+      "loss": 1.0388,
+      "step": 2797
+    },
+    {
+      "epoch": 0.25000558447070387,
+      "grad_norm": 0.3682102859020233,
+      "learning_rate": 8.783525067709075e-05,
+      "loss": 1.0017,
+      "step": 2798
+    },
+    {
+      "epoch": 0.25009493600196575,
+      "grad_norm": 0.5301668643951416,
+      "learning_rate": 8.782578876782259e-05,
+      "loss": 0.8846,
+      "step": 2799
+    },
+    {
+      "epoch": 0.2501842875332276,
+      "grad_norm": 0.5186197757720947,
+      "learning_rate": 8.781632369024141e-05,
+      "loss": 0.9255,
+      "step": 2800
+    },
+    {
+      "epoch": 0.25027363906448946,
+      "grad_norm": 0.4258555769920349,
+      "learning_rate": 8.780685544514006e-05,
+      "loss": 0.9937,
+      "step": 2801
+    },
+    {
+      "epoch": 0.25036299059575134,
+      "grad_norm": 0.44377055764198303,
+      "learning_rate": 8.779738403331157e-05,
+      "loss": 1.0151,
+      "step": 2802
+    },
+    {
+      "epoch": 0.2504523421270132,
+      "grad_norm": 0.3976726830005646,
+      "learning_rate": 8.778790945554926e-05,
+      "loss": 1.0148,
+      "step": 2803
+    },
+    {
+      "epoch": 0.25054169365827506,
+      "grad_norm": 0.34984302520751953,
+      "learning_rate": 8.777843171264675e-05,
+      "loss": 1.0027,
+      "step": 2804
+    },
+    {
+      "epoch": 0.25063104518953694,
+      "grad_norm": 0.36528366804122925,
+      "learning_rate": 8.776895080539789e-05,
+      "loss": 1.0559,
+      "step": 2805
+    },
+    {
+      "epoch": 0.2507203967207988,
+      "grad_norm": 0.48383477330207825,
+      "learning_rate": 8.775946673459681e-05,
+      "loss": 1.0472,
+      "step": 2806
+    },
+    {
+      "epoch": 0.25080974825206065,
+      "grad_norm": 0.3872027099132538,
+      "learning_rate": 8.774997950103791e-05,
+      "loss": 1.0441,
+      "step": 2807
+    },
+    {
+      "epoch": 0.25089909978332253,
+      "grad_norm": 0.45994165539741516,
+      "learning_rate": 8.774048910551584e-05,
+      "loss": 0.9451,
+      "step": 2808
+    },
+    {
+      "epoch": 0.2509884513145844,
+      "grad_norm": 0.39448827505111694,
+      "learning_rate": 8.773099554882552e-05,
+      "loss": 1.0241,
+      "step": 2809
+    },
+    {
+      "epoch": 0.25107780284584624,
+      "grad_norm": 0.39023977518081665,
+      "learning_rate": 8.772149883176215e-05,
+      "loss": 0.9925,
+      "step": 2810
+    },
+    {
+      "epoch": 0.2511671543771081,
+      "grad_norm": 0.49682340025901794,
+      "learning_rate": 8.771199895512115e-05,
+      "loss": 0.9678,
+      "step": 2811
+    },
+    {
+      "epoch": 0.25125650590837,
+      "grad_norm": 0.5279167890548706,
+      "learning_rate": 8.770249591969829e-05,
+      "loss": 0.9971,
+      "step": 2812
+    },
+    {
+      "epoch": 0.2513458574396319,
+      "grad_norm": 0.46056196093559265,
+      "learning_rate": 8.769298972628948e-05,
+      "loss": 1.0181,
+      "step": 2813
+    },
+    {
+      "epoch": 0.2514352089708937,
+      "grad_norm": 0.40414974093437195,
+      "learning_rate": 8.768348037569102e-05,
+      "loss": 0.9936,
+      "step": 2814
+    },
+    {
+      "epoch": 0.2515245605021556,
+      "grad_norm": 0.45471900701522827,
+      "learning_rate": 8.76739678686994e-05,
+      "loss": 1.0091,
+      "step": 2815
+    },
+    {
+      "epoch": 0.2516139120334175,
+      "grad_norm": 0.4152718186378479,
+      "learning_rate": 8.766445220611139e-05,
+      "loss": 1.0165,
+      "step": 2816
+    },
+    {
+      "epoch": 0.2517032635646793,
+      "grad_norm": 0.3915504515171051,
+      "learning_rate": 8.765493338872403e-05,
+      "loss": 1.028,
+      "step": 2817
+    },
+    {
+      "epoch": 0.2517926150959412,
+      "grad_norm": 0.4392106533050537,
+      "learning_rate": 8.764541141733464e-05,
+      "loss": 0.9371,
+      "step": 2818
+    },
+    {
+      "epoch": 0.2518819666272031,
+      "grad_norm": 0.39547622203826904,
+      "learning_rate": 8.763588629274077e-05,
+      "loss": 1.018,
+      "step": 2819
+    },
+    {
+      "epoch": 0.25197131815846496,
+      "grad_norm": 0.35008326172828674,
+      "learning_rate": 8.762635801574025e-05,
+      "loss": 0.9658,
+      "step": 2820
+    },
+    {
+      "epoch": 0.2520606696897268,
+      "grad_norm": 0.4563378393650055,
+      "learning_rate": 8.761682658713119e-05,
+      "loss": 0.966,
+      "step": 2821
+    },
+    {
+      "epoch": 0.2521500212209887,
+      "grad_norm": 0.3921094536781311,
+      "learning_rate": 8.760729200771192e-05,
+      "loss": 1.0282,
+      "step": 2822
+    },
+    {
+      "epoch": 0.25223937275225056,
+      "grad_norm": 0.4152204990386963,
+      "learning_rate": 8.759775427828108e-05,
+      "loss": 1.0107,
+      "step": 2823
+    },
+    {
+      "epoch": 0.2523287242835124,
+      "grad_norm": 0.47061243653297424,
+      "learning_rate": 8.758821339963756e-05,
+      "loss": 1.0095,
+      "step": 2824
+    },
+    {
+      "epoch": 0.25241807581477427,
+      "grad_norm": 0.49046018719673157,
+      "learning_rate": 8.75786693725805e-05,
+      "loss": 1.0251,
+      "step": 2825
+    },
+    {
+      "epoch": 0.25250742734603615,
+      "grad_norm": 0.43877169489860535,
+      "learning_rate": 8.756912219790933e-05,
+      "loss": 0.9794,
+      "step": 2826
+    },
+    {
+      "epoch": 0.25259677887729803,
+      "grad_norm": 0.4174705445766449,
+      "learning_rate": 8.755957187642372e-05,
+      "loss": 0.9845,
+      "step": 2827
+    },
+    {
+      "epoch": 0.25268613040855986,
+      "grad_norm": 0.3626020550727844,
+      "learning_rate": 8.755001840892361e-05,
+      "loss": 1.0018,
+      "step": 2828
+    },
+    {
+      "epoch": 0.25277548193982174,
+      "grad_norm": 0.48767632246017456,
+      "learning_rate": 8.754046179620919e-05,
+      "loss": 1.0055,
+      "step": 2829
+    },
+    {
+      "epoch": 0.2528648334710836,
+      "grad_norm": 0.4239792823791504,
+      "learning_rate": 8.753090203908095e-05,
+      "loss": 0.9721,
+      "step": 2830
+    },
+    {
+      "epoch": 0.25295418500234546,
+      "grad_norm": 0.4804183542728424,
+      "learning_rate": 8.752133913833962e-05,
+      "loss": 0.9585,
+      "step": 2831
+    },
+    {
+      "epoch": 0.25304353653360734,
+      "grad_norm": 0.45902127027511597,
+      "learning_rate": 8.751177309478618e-05,
+      "loss": 1.0059,
+      "step": 2832
+    },
+    {
+      "epoch": 0.2531328880648692,
+      "grad_norm": 0.473166823387146,
+      "learning_rate": 8.750220390922188e-05,
+      "loss": 0.9744,
+      "step": 2833
+    },
+    {
+      "epoch": 0.2532222395961311,
+      "grad_norm": 0.475212037563324,
+      "learning_rate": 8.74926315824483e-05,
+      "loss": 1.0587,
+      "step": 2834
+    },
+    {
+      "epoch": 0.25331159112739293,
+      "grad_norm": 0.4180943965911865,
+      "learning_rate": 8.748305611526715e-05,
+      "loss": 0.9937,
+      "step": 2835
+    },
+    {
+      "epoch": 0.2534009426586548,
+      "grad_norm": 0.4551735520362854,
+      "learning_rate": 8.747347750848052e-05,
+      "loss": 0.968,
+      "step": 2836
+    },
+    {
+      "epoch": 0.2534902941899167,
+      "grad_norm": 0.3868304491043091,
+      "learning_rate": 8.74638957628907e-05,
+      "loss": 0.9912,
+      "step": 2837
+    },
+    {
+      "epoch": 0.2535796457211785,
+      "grad_norm": 0.43097737431526184,
+      "learning_rate": 8.745431087930028e-05,
+      "loss": 1.0214,
+      "step": 2838
+    },
+    {
+      "epoch": 0.2536689972524404,
+      "grad_norm": 0.37458547949790955,
+      "learning_rate": 8.74447228585121e-05,
+      "loss": 1.017,
+      "step": 2839
+    },
+    {
+      "epoch": 0.2537583487837023,
+      "grad_norm": 0.5047030448913574,
+      "learning_rate": 8.743513170132924e-05,
+      "loss": 0.9561,
+      "step": 2840
+    },
+    {
+      "epoch": 0.2538477003149642,
+      "grad_norm": 0.3879890441894531,
+      "learning_rate": 8.742553740855506e-05,
+      "loss": 1.0405,
+      "step": 2841
+    },
+    {
+      "epoch": 0.253937051846226,
+      "grad_norm": 0.44273656606674194,
+      "learning_rate": 8.74159399809932e-05,
+      "loss": 1.0599,
+      "step": 2842
+    },
+    {
+      "epoch": 0.2540264033774879,
+      "grad_norm": 0.4389507472515106,
+      "learning_rate": 8.740633941944754e-05,
+      "loss": 1.0179,
+      "step": 2843
+    },
+    {
+      "epoch": 0.25411575490874977,
+      "grad_norm": 0.44529905915260315,
+      "learning_rate": 8.739673572472225e-05,
+      "loss": 1.0689,
+      "step": 2844
+    },
+    {
+      "epoch": 0.2542051064400116,
+      "grad_norm": 0.4119149148464203,
+      "learning_rate": 8.738712889762171e-05,
+      "loss": 0.9636,
+      "step": 2845
+    },
+    {
+      "epoch": 0.2542944579712735,
+      "grad_norm": 0.4760536551475525,
+      "learning_rate": 8.73775189389506e-05,
+      "loss": 0.963,
+      "step": 2846
+    },
+    {
+      "epoch": 0.25438380950253536,
+      "grad_norm": 0.4643342196941376,
+      "learning_rate": 8.736790584951387e-05,
+      "loss": 0.9906,
+      "step": 2847
+    },
+    {
+      "epoch": 0.2544731610337972,
+      "grad_norm": 0.402329683303833,
+      "learning_rate": 8.735828963011671e-05,
+      "loss": 1.0063,
+      "step": 2848
+    },
+    {
+      "epoch": 0.2545625125650591,
+      "grad_norm": 0.41884222626686096,
+      "learning_rate": 8.734867028156458e-05,
+      "loss": 1.0401,
+      "step": 2849
+    },
+    {
+      "epoch": 0.25465186409632096,
+      "grad_norm": 0.3897888660430908,
+      "learning_rate": 8.733904780466321e-05,
+      "loss": 1.057,
+      "step": 2850
+    },
+    {
+      "epoch": 0.25474121562758284,
+      "grad_norm": 0.4296242296695709,
+      "learning_rate": 8.732942220021858e-05,
+      "loss": 0.9684,
+      "step": 2851
+    },
+    {
+      "epoch": 0.25483056715884467,
+      "grad_norm": 0.47800928354263306,
+      "learning_rate": 8.731979346903693e-05,
+      "loss": 1.0239,
+      "step": 2852
+    },
+    {
+      "epoch": 0.25491991869010655,
+      "grad_norm": 0.4393937587738037,
+      "learning_rate": 8.731016161192479e-05,
+      "loss": 1.0716,
+      "step": 2853
+    },
+    {
+      "epoch": 0.25500927022136843,
+      "grad_norm": 0.46051809191703796,
+      "learning_rate": 8.730052662968891e-05,
+      "loss": 1.0035,
+      "step": 2854
+    },
+    {
+      "epoch": 0.25509862175263026,
+      "grad_norm": 0.3697504997253418,
+      "learning_rate": 8.729088852313633e-05,
+      "loss": 0.9667,
+      "step": 2855
+    },
+    {
+      "epoch": 0.25518797328389214,
+      "grad_norm": 0.4968481659889221,
+      "learning_rate": 8.728124729307434e-05,
+      "loss": 0.9228,
+      "step": 2856
+    },
+    {
+      "epoch": 0.255277324815154,
+      "grad_norm": 0.46425577998161316,
+      "learning_rate": 8.727160294031051e-05,
+      "loss": 1.0125,
+      "step": 2857
+    },
+    {
+      "epoch": 0.2553666763464159,
+      "grad_norm": 0.40039297938346863,
+      "learning_rate": 8.726195546565263e-05,
+      "loss": 1.0081,
+      "step": 2858
+    },
+    {
+      "epoch": 0.25545602787767774,
+      "grad_norm": 0.48892778158187866,
+      "learning_rate": 8.725230486990882e-05,
+      "loss": 0.9115,
+      "step": 2859
+    },
+    {
+      "epoch": 0.2555453794089396,
+      "grad_norm": 0.3827580511569977,
+      "learning_rate": 8.724265115388739e-05,
+      "loss": 1.0488,
+      "step": 2860
+    },
+    {
+      "epoch": 0.2556347309402015,
+      "grad_norm": 0.45204082131385803,
+      "learning_rate": 8.723299431839693e-05,
+      "loss": 1.0264,
+      "step": 2861
+    },
+    {
+      "epoch": 0.25572408247146333,
+      "grad_norm": 0.46111056208610535,
+      "learning_rate": 8.722333436424633e-05,
+      "loss": 0.9979,
+      "step": 2862
+    },
+    {
+      "epoch": 0.2558134340027252,
+      "grad_norm": 0.40219980478286743,
+      "learning_rate": 8.721367129224471e-05,
+      "loss": 1.0384,
+      "step": 2863
+    },
+    {
+      "epoch": 0.2559027855339871,
+      "grad_norm": 0.37858057022094727,
+      "learning_rate": 8.720400510320146e-05,
+      "loss": 1.0113,
+      "step": 2864
+    },
+    {
+      "epoch": 0.255992137065249,
+      "grad_norm": 0.4286115765571594,
+      "learning_rate": 8.71943357979262e-05,
+      "loss": 0.9691,
+      "step": 2865
+    },
+    {
+      "epoch": 0.2560814885965108,
+      "grad_norm": 0.4207289218902588,
+      "learning_rate": 8.718466337722885e-05,
+      "loss": 0.9316,
+      "step": 2866
+    },
+    {
+      "epoch": 0.2561708401277727,
+      "grad_norm": 0.4365525245666504,
+      "learning_rate": 8.717498784191958e-05,
+      "loss": 0.99,
+      "step": 2867
+    },
+    {
+      "epoch": 0.2562601916590346,
+      "grad_norm": 0.4580113887786865,
+      "learning_rate": 8.716530919280883e-05,
+      "loss": 0.9109,
+      "step": 2868
+    },
+    {
+      "epoch": 0.2563495431902964,
+      "grad_norm": 0.4263405203819275,
+      "learning_rate": 8.715562743070729e-05,
+      "loss": 0.9391,
+      "step": 2869
+    },
+    {
+      "epoch": 0.2564388947215583,
+      "grad_norm": 0.4391098916530609,
+      "learning_rate": 8.71459425564259e-05,
+      "loss": 1.0104,
+      "step": 2870
+    },
+    {
+      "epoch": 0.25652824625282017,
+      "grad_norm": 0.45276468992233276,
+      "learning_rate": 8.713625457077585e-05,
+      "loss": 1.1036,
+      "step": 2871
+    },
+    {
+      "epoch": 0.25661759778408205,
+      "grad_norm": 0.45625758171081543,
+      "learning_rate": 8.712656347456867e-05,
+      "loss": 0.9503,
+      "step": 2872
+    },
+    {
+      "epoch": 0.2567069493153439,
+      "grad_norm": 0.40368926525115967,
+      "learning_rate": 8.711686926861604e-05,
+      "loss": 1.0154,
+      "step": 2873
+    },
+    {
+      "epoch": 0.25679630084660576,
+      "grad_norm": 0.43693703413009644,
+      "learning_rate": 8.710717195372997e-05,
+      "loss": 1.0454,
+      "step": 2874
+    },
+    {
+      "epoch": 0.25688565237786765,
+      "grad_norm": 0.38676273822784424,
+      "learning_rate": 8.709747153072272e-05,
+      "loss": 1.0724,
+      "step": 2875
+    },
+    {
+      "epoch": 0.2569750039091295,
+      "grad_norm": 0.441709041595459,
+      "learning_rate": 8.708776800040679e-05,
+      "loss": 0.9994,
+      "step": 2876
+    },
+    {
+      "epoch": 0.25706435544039136,
+      "grad_norm": 0.40171146392822266,
+      "learning_rate": 8.707806136359497e-05,
+      "loss": 1.0599,
+      "step": 2877
+    },
+    {
+      "epoch": 0.25715370697165324,
+      "grad_norm": 0.40110501646995544,
+      "learning_rate": 8.706835162110028e-05,
+      "loss": 1.015,
+      "step": 2878
+    },
+    {
+      "epoch": 0.25724305850291507,
+      "grad_norm": 0.4818125069141388,
+      "learning_rate": 8.705863877373603e-05,
+      "loss": 1.0023,
+      "step": 2879
+    },
+    {
+      "epoch": 0.25733241003417695,
+      "grad_norm": 0.3713032901287079,
+      "learning_rate": 8.704892282231575e-05,
+      "loss": 1.0398,
+      "step": 2880
+    },
+    {
+      "epoch": 0.25742176156543883,
+      "grad_norm": 0.509074330329895,
+      "learning_rate": 8.70392037676533e-05,
+      "loss": 0.9623,
+      "step": 2881
+    },
+    {
+      "epoch": 0.2575111130967007,
+      "grad_norm": 0.4063872694969177,
+      "learning_rate": 8.70294816105627e-05,
+      "loss": 0.9864,
+      "step": 2882
+    },
+    {
+      "epoch": 0.25760046462796254,
+      "grad_norm": 0.4200826585292816,
+      "learning_rate": 8.701975635185833e-05,
+      "loss": 0.9905,
+      "step": 2883
+    },
+    {
+      "epoch": 0.2576898161592244,
+      "grad_norm": 0.44294077157974243,
+      "learning_rate": 8.701002799235475e-05,
+      "loss": 0.9429,
+      "step": 2884
+    },
+    {
+      "epoch": 0.2577791676904863,
+      "grad_norm": 0.3688490092754364,
+      "learning_rate": 8.700029653286684e-05,
+      "loss": 1.0168,
+      "step": 2885
+    },
+    {
+      "epoch": 0.25786851922174814,
+      "grad_norm": 0.381213903427124,
+      "learning_rate": 8.699056197420967e-05,
+      "loss": 1.0477,
+      "step": 2886
+    },
+    {
+      "epoch": 0.25795787075301,
+      "grad_norm": 0.4321887195110321,
+      "learning_rate": 8.698082431719867e-05,
+      "loss": 0.9478,
+      "step": 2887
+    },
+    {
+      "epoch": 0.2580472222842719,
+      "grad_norm": 0.39009973406791687,
+      "learning_rate": 8.697108356264944e-05,
+      "loss": 0.9456,
+      "step": 2888
+    },
+    {
+      "epoch": 0.2581365738155338,
+      "grad_norm": 0.39922964572906494,
+      "learning_rate": 8.696133971137788e-05,
+      "loss": 1.0529,
+      "step": 2889
+    },
+    {
+      "epoch": 0.2582259253467956,
+      "grad_norm": 0.45770078897476196,
+      "learning_rate": 8.695159276420013e-05,
+      "loss": 0.9618,
+      "step": 2890
+    },
+    {
+      "epoch": 0.2583152768780575,
+      "grad_norm": 0.4154913127422333,
+      "learning_rate": 8.694184272193262e-05,
+      "loss": 1.016,
+      "step": 2891
+    },
+    {
+      "epoch": 0.2584046284093194,
+      "grad_norm": 0.3861916661262512,
+      "learning_rate": 8.6932089585392e-05,
+      "loss": 1.0038,
+      "step": 2892
+    },
+    {
+      "epoch": 0.2584939799405812,
+      "grad_norm": 0.39484497904777527,
+      "learning_rate": 8.692233335539521e-05,
+      "loss": 0.9882,
+      "step": 2893
+    },
+    {
+      "epoch": 0.2585833314718431,
+      "grad_norm": 0.36382368206977844,
+      "learning_rate": 8.691257403275945e-05,
+      "loss": 1.087,
+      "step": 2894
+    },
+    {
+      "epoch": 0.258672683003105,
+      "grad_norm": 0.42892515659332275,
+      "learning_rate": 8.690281161830216e-05,
+      "loss": 1.1341,
+      "step": 2895
+    },
+    {
+      "epoch": 0.25876203453436686,
+      "grad_norm": 0.42802634835243225,
+      "learning_rate": 8.689304611284103e-05,
+      "loss": 0.993,
+      "step": 2896
+    },
+    {
+      "epoch": 0.2588513860656287,
+      "grad_norm": 0.6133880019187927,
+      "learning_rate": 8.688327751719403e-05,
+      "loss": 0.9809,
+      "step": 2897
+    },
+    {
+      "epoch": 0.25894073759689057,
+      "grad_norm": 0.4578530788421631,
+      "learning_rate": 8.68735058321794e-05,
+      "loss": 0.9413,
+      "step": 2898
+    },
+    {
+      "epoch": 0.25903008912815245,
+      "grad_norm": 0.4118945896625519,
+      "learning_rate": 8.68637310586156e-05,
+      "loss": 0.9512,
+      "step": 2899
+    },
+    {
+      "epoch": 0.2591194406594143,
+      "grad_norm": 0.4931046962738037,
+      "learning_rate": 8.685395319732141e-05,
+      "loss": 1.0285,
+      "step": 2900
+    },
+    {
+      "epoch": 0.25920879219067616,
+      "grad_norm": 0.4491409361362457,
+      "learning_rate": 8.684417224911578e-05,
+      "loss": 1.0044,
+      "step": 2901
+    },
+    {
+      "epoch": 0.25929814372193805,
+      "grad_norm": 0.37248972058296204,
+      "learning_rate": 8.683438821481802e-05,
+      "loss": 1.0201,
+      "step": 2902
+    },
+    {
+      "epoch": 0.25938749525319993,
+      "grad_norm": 0.46489718556404114,
+      "learning_rate": 8.68246010952476e-05,
+      "loss": 0.9569,
+      "step": 2903
+    },
+    {
+      "epoch": 0.25947684678446176,
+      "grad_norm": 0.4184110164642334,
+      "learning_rate": 8.681481089122432e-05,
+      "loss": 0.9632,
+      "step": 2904
+    },
+    {
+      "epoch": 0.25956619831572364,
+      "grad_norm": 0.38718414306640625,
+      "learning_rate": 8.68050176035682e-05,
+      "loss": 1.0251,
+      "step": 2905
+    },
+    {
+      "epoch": 0.2596555498469855,
+      "grad_norm": 0.3973325490951538,
+      "learning_rate": 8.679522123309956e-05,
+      "loss": 1.0239,
+      "step": 2906
+    },
+    {
+      "epoch": 0.25974490137824735,
+      "grad_norm": 0.37302878499031067,
+      "learning_rate": 8.678542178063893e-05,
+      "loss": 1.0898,
+      "step": 2907
+    },
+    {
+      "epoch": 0.25983425290950923,
+      "grad_norm": 0.46653488278388977,
+      "learning_rate": 8.677561924700713e-05,
+      "loss": 1.0144,
+      "step": 2908
+    },
+    {
+      "epoch": 0.2599236044407711,
+      "grad_norm": 0.43336376547813416,
+      "learning_rate": 8.676581363302518e-05,
+      "loss": 0.9507,
+      "step": 2909
+    },
+    {
+      "epoch": 0.26001295597203294,
+      "grad_norm": 0.5737956762313843,
+      "learning_rate": 8.675600493951448e-05,
+      "loss": 0.9631,
+      "step": 2910
+    },
+    {
+      "epoch": 0.2601023075032948,
+      "grad_norm": 0.460018128156662,
+      "learning_rate": 8.674619316729657e-05,
+      "loss": 0.9619,
+      "step": 2911
+    },
+    {
+      "epoch": 0.2601916590345567,
+      "grad_norm": 0.420095831155777,
+      "learning_rate": 8.673637831719328e-05,
+      "loss": 0.9587,
+      "step": 2912
+    },
+    {
+      "epoch": 0.2602810105658186,
+      "grad_norm": 0.46309584379196167,
+      "learning_rate": 8.672656039002674e-05,
+      "loss": 0.9664,
+      "step": 2913
+    },
+    {
+      "epoch": 0.2603703620970804,
+      "grad_norm": 0.5068766474723816,
+      "learning_rate": 8.671673938661929e-05,
+      "loss": 1.0086,
+      "step": 2914
+    },
+    {
+      "epoch": 0.2604597136283423,
+      "grad_norm": 0.46311527490615845,
+      "learning_rate": 8.670691530779354e-05,
+      "loss": 0.9858,
+      "step": 2915
+    },
+    {
+      "epoch": 0.2605490651596042,
+      "grad_norm": 0.4441559612751007,
+      "learning_rate": 8.669708815437237e-05,
+      "loss": 0.9832,
+      "step": 2916
+    },
+    {
+      "epoch": 0.260638416690866,
+      "grad_norm": 0.4474044442176819,
+      "learning_rate": 8.668725792717889e-05,
+      "loss": 1.0313,
+      "step": 2917
+    },
+    {
+      "epoch": 0.2607277682221279,
+      "grad_norm": 0.5184552669525146,
+      "learning_rate": 8.667742462703649e-05,
+      "loss": 1.0102,
+      "step": 2918
+    },
+    {
+      "epoch": 0.2608171197533898,
+      "grad_norm": 0.4744894206523895,
+      "learning_rate": 8.666758825476886e-05,
+      "loss": 0.9695,
+      "step": 2919
+    },
+    {
+      "epoch": 0.26090647128465166,
+      "grad_norm": 0.40858200192451477,
+      "learning_rate": 8.665774881119985e-05,
+      "loss": 0.9797,
+      "step": 2920
+    },
+    {
+      "epoch": 0.2609958228159135,
+      "grad_norm": 0.394593745470047,
+      "learning_rate": 8.664790629715363e-05,
+      "loss": 1.0465,
+      "step": 2921
+    },
+    {
+      "epoch": 0.2610851743471754,
+      "grad_norm": 0.4352075457572937,
+      "learning_rate": 8.663806071345462e-05,
+      "loss": 0.9832,
+      "step": 2922
+    },
+    {
+      "epoch": 0.26117452587843726,
+      "grad_norm": 0.4699738323688507,
+      "learning_rate": 8.662821206092748e-05,
+      "loss": 1.064,
+      "step": 2923
+    },
+    {
+      "epoch": 0.2612638774096991,
+      "grad_norm": 0.4367484450340271,
+      "learning_rate": 8.661836034039717e-05,
+      "loss": 0.9967,
+      "step": 2924
+    },
+    {
+      "epoch": 0.26135322894096097,
+      "grad_norm": 0.3603065609931946,
+      "learning_rate": 8.660850555268886e-05,
+      "loss": 1.0077,
+      "step": 2925
+    },
+    {
+      "epoch": 0.26144258047222285,
+      "grad_norm": 0.3636876344680786,
+      "learning_rate": 8.659864769862798e-05,
+      "loss": 1.0128,
+      "step": 2926
+    },
+    {
+      "epoch": 0.26153193200348474,
+      "grad_norm": 0.37722858786582947,
+      "learning_rate": 8.658878677904024e-05,
+      "loss": 0.9895,
+      "step": 2927
+    },
+    {
+      "epoch": 0.26162128353474656,
+      "grad_norm": 0.4058395028114319,
+      "learning_rate": 8.65789227947516e-05,
+      "loss": 0.965,
+      "step": 2928
+    },
+    {
+      "epoch": 0.26171063506600845,
+      "grad_norm": 0.3944026529788971,
+      "learning_rate": 8.656905574658829e-05,
+      "loss": 1.0081,
+      "step": 2929
+    },
+    {
+      "epoch": 0.26179998659727033,
+      "grad_norm": 0.5204667448997498,
+      "learning_rate": 8.655918563537675e-05,
+      "loss": 1.0273,
+      "step": 2930
+    },
+    {
+      "epoch": 0.26188933812853216,
+      "grad_norm": 0.466569721698761,
+      "learning_rate": 8.654931246194372e-05,
+      "loss": 1.0126,
+      "step": 2931
+    },
+    {
+      "epoch": 0.26197868965979404,
+      "grad_norm": 0.49643585085868835,
+      "learning_rate": 8.653943622711618e-05,
+      "loss": 0.9925,
+      "step": 2932
+    },
+    {
+      "epoch": 0.2620680411910559,
+      "grad_norm": 0.42305490374565125,
+      "learning_rate": 8.652955693172137e-05,
+      "loss": 0.9461,
+      "step": 2933
+    },
+    {
+      "epoch": 0.2621573927223178,
+      "grad_norm": 0.5041871666908264,
+      "learning_rate": 8.65196745765868e-05,
+      "loss": 0.9714,
+      "step": 2934
+    },
+    {
+      "epoch": 0.26224674425357963,
+      "grad_norm": 0.4066472351551056,
+      "learning_rate": 8.65097891625402e-05,
+      "loss": 1.0175,
+      "step": 2935
+    },
+    {
+      "epoch": 0.2623360957848415,
+      "grad_norm": 0.41952818632125854,
+      "learning_rate": 8.649990069040961e-05,
+      "loss": 0.998,
+      "step": 2936
+    },
+    {
+      "epoch": 0.2624254473161034,
+      "grad_norm": 0.44271522760391235,
+      "learning_rate": 8.649000916102325e-05,
+      "loss": 0.9336,
+      "step": 2937
+    },
+    {
+      "epoch": 0.2625147988473652,
+      "grad_norm": 0.47163864970207214,
+      "learning_rate": 8.64801145752097e-05,
+      "loss": 1.0735,
+      "step": 2938
+    },
+    {
+      "epoch": 0.2626041503786271,
+      "grad_norm": 0.40220436453819275,
+      "learning_rate": 8.647021693379768e-05,
+      "loss": 0.99,
+      "step": 2939
+    },
+    {
+      "epoch": 0.262693501909889,
+      "grad_norm": 0.37812304496765137,
+      "learning_rate": 8.646031623761626e-05,
+      "loss": 1.0802,
+      "step": 2940
+    },
+    {
+      "epoch": 0.2627828534411508,
+      "grad_norm": 0.4696109890937805,
+      "learning_rate": 8.645041248749471e-05,
+      "loss": 0.9638,
+      "step": 2941
+    },
+    {
+      "epoch": 0.2628722049724127,
+      "grad_norm": 0.4705543518066406,
+      "learning_rate": 8.644050568426259e-05,
+      "loss": 0.9096,
+      "step": 2942
+    },
+    {
+      "epoch": 0.2629615565036746,
+      "grad_norm": 0.4492562413215637,
+      "learning_rate": 8.643059582874969e-05,
+      "loss": 0.9718,
+      "step": 2943
+    },
+    {
+      "epoch": 0.26305090803493647,
+      "grad_norm": 0.44293731451034546,
+      "learning_rate": 8.642068292178605e-05,
+      "loss": 0.9978,
+      "step": 2944
+    },
+    {
+      "epoch": 0.2631402595661983,
+      "grad_norm": 0.4119532108306885,
+      "learning_rate": 8.641076696420201e-05,
+      "loss": 1.0805,
+      "step": 2945
+    },
+    {
+      "epoch": 0.2632296110974602,
+      "grad_norm": 0.5227283239364624,
+      "learning_rate": 8.640084795682813e-05,
+      "loss": 0.9369,
+      "step": 2946
+    },
+    {
+      "epoch": 0.26331896262872206,
+      "grad_norm": 0.3896709382534027,
+      "learning_rate": 8.639092590049521e-05,
+      "loss": 0.9635,
+      "step": 2947
+    },
+    {
+      "epoch": 0.2634083141599839,
+      "grad_norm": 0.3982143998146057,
+      "learning_rate": 8.638100079603437e-05,
+      "loss": 1.0144,
+      "step": 2948
+    },
+    {
+      "epoch": 0.2634976656912458,
+      "grad_norm": 0.40874722599983215,
+      "learning_rate": 8.63710726442769e-05,
+      "loss": 1.0396,
+      "step": 2949
+    },
+    {
+      "epoch": 0.26358701722250766,
+      "grad_norm": 0.4392448365688324,
+      "learning_rate": 8.636114144605442e-05,
+      "loss": 1.0296,
+      "step": 2950
+    },
+    {
+      "epoch": 0.26367636875376954,
+      "grad_norm": 0.44011494517326355,
+      "learning_rate": 8.635120720219876e-05,
+      "loss": 1.0233,
+      "step": 2951
+    },
+    {
+      "epoch": 0.26376572028503137,
+      "grad_norm": 0.4407230615615845,
+      "learning_rate": 8.634126991354202e-05,
+      "loss": 1.0963,
+      "step": 2952
+    },
+    {
+      "epoch": 0.26385507181629325,
+      "grad_norm": 0.562527596950531,
+      "learning_rate": 8.633132958091655e-05,
+      "loss": 0.9788,
+      "step": 2953
+    },
+    {
+      "epoch": 0.26394442334755513,
+      "grad_norm": 0.44428539276123047,
+      "learning_rate": 8.632138620515498e-05,
+      "loss": 1.0182,
+      "step": 2954
+    },
+    {
+      "epoch": 0.26403377487881696,
+      "grad_norm": 0.4478207230567932,
+      "learning_rate": 8.631143978709013e-05,
+      "loss": 0.981,
+      "step": 2955
+    },
+    {
+      "epoch": 0.26412312641007885,
+      "grad_norm": 0.48832935094833374,
+      "learning_rate": 8.630149032755517e-05,
+      "loss": 1.0158,
+      "step": 2956
+    },
+    {
+      "epoch": 0.26421247794134073,
+      "grad_norm": 0.47887471318244934,
+      "learning_rate": 8.629153782738344e-05,
+      "loss": 0.9785,
+      "step": 2957
+    },
+    {
+      "epoch": 0.2643018294726026,
+      "grad_norm": 0.481168657541275,
+      "learning_rate": 8.628158228740857e-05,
+      "loss": 0.949,
+      "step": 2958
+    },
+    {
+      "epoch": 0.26439118100386444,
+      "grad_norm": 0.42248255014419556,
+      "learning_rate": 8.627162370846446e-05,
+      "loss": 1.0021,
+      "step": 2959
+    },
+    {
+      "epoch": 0.2644805325351263,
+      "grad_norm": 0.3783293664455414,
+      "learning_rate": 8.626166209138524e-05,
+      "loss": 1.0025,
+      "step": 2960
+    },
+    {
+      "epoch": 0.2645698840663882,
+      "grad_norm": 0.40361276268959045,
+      "learning_rate": 8.62516974370053e-05,
+      "loss": 0.9737,
+      "step": 2961
+    },
+    {
+      "epoch": 0.26465923559765003,
+      "grad_norm": 0.4194972813129425,
+      "learning_rate": 8.624172974615926e-05,
+      "loss": 1.0377,
+      "step": 2962
+    },
+    {
+      "epoch": 0.2647485871289119,
+      "grad_norm": 0.4278354346752167,
+      "learning_rate": 8.623175901968206e-05,
+      "loss": 1.0373,
+      "step": 2963
+    },
+    {
+      "epoch": 0.2648379386601738,
+      "grad_norm": 0.430337518453598,
+      "learning_rate": 8.622178525840885e-05,
+      "loss": 0.9437,
+      "step": 2964
+    },
+    {
+      "epoch": 0.2649272901914357,
+      "grad_norm": 0.39128750562667847,
+      "learning_rate": 8.6211808463175e-05,
+      "loss": 1.03,
+      "step": 2965
+    },
+    {
+      "epoch": 0.2650166417226975,
+      "grad_norm": 0.4273848235607147,
+      "learning_rate": 8.620182863481622e-05,
+      "loss": 0.967,
+      "step": 2966
+    },
+    {
+      "epoch": 0.2651059932539594,
+      "grad_norm": 0.43095484375953674,
+      "learning_rate": 8.619184577416842e-05,
+      "loss": 1.0008,
+      "step": 2967
+    },
+    {
+      "epoch": 0.2651953447852213,
+      "grad_norm": 0.4028542935848236,
+      "learning_rate": 8.618185988206775e-05,
+      "loss": 1.0687,
+      "step": 2968
+    },
+    {
+      "epoch": 0.2652846963164831,
+      "grad_norm": 0.45364266633987427,
+      "learning_rate": 8.617187095935065e-05,
+      "loss": 0.9845,
+      "step": 2969
+    },
+    {
+      "epoch": 0.265374047847745,
+      "grad_norm": 0.4723089039325714,
+      "learning_rate": 8.616187900685377e-05,
+      "loss": 1.0094,
+      "step": 2970
+    },
+    {
+      "epoch": 0.26546339937900687,
+      "grad_norm": 0.45047450065612793,
+      "learning_rate": 8.615188402541408e-05,
+      "loss": 1.025,
+      "step": 2971
+    },
+    {
+      "epoch": 0.2655527509102687,
+      "grad_norm": 0.44591468572616577,
+      "learning_rate": 8.614188601586875e-05,
+      "loss": 0.9725,
+      "step": 2972
+    },
+    {
+      "epoch": 0.2656421024415306,
+      "grad_norm": 0.5437367558479309,
+      "learning_rate": 8.613188497905523e-05,
+      "loss": 0.9072,
+      "step": 2973
+    },
+    {
+      "epoch": 0.26573145397279246,
+      "grad_norm": 0.36033257842063904,
+      "learning_rate": 8.612188091581119e-05,
+      "loss": 0.9875,
+      "step": 2974
+    },
+    {
+      "epoch": 0.26582080550405435,
+      "grad_norm": 0.4065442979335785,
+      "learning_rate": 8.611187382697458e-05,
+      "loss": 1.0593,
+      "step": 2975
+    },
+    {
+      "epoch": 0.2659101570353162,
+      "grad_norm": 0.42632168531417847,
+      "learning_rate": 8.610186371338365e-05,
+      "loss": 1.0284,
+      "step": 2976
+    },
+    {
+      "epoch": 0.26599950856657806,
+      "grad_norm": 0.45668208599090576,
+      "learning_rate": 8.609185057587678e-05,
+      "loss": 1.0186,
+      "step": 2977
+    },
+    {
+      "epoch": 0.26608886009783994,
+      "grad_norm": 0.4407256245613098,
+      "learning_rate": 8.608183441529274e-05,
+      "loss": 0.9508,
+      "step": 2978
+    },
+    {
+      "epoch": 0.26617821162910177,
+      "grad_norm": 0.4113801419734955,
+      "learning_rate": 8.607181523247045e-05,
+      "loss": 1.0019,
+      "step": 2979
+    },
+    {
+      "epoch": 0.26626756316036365,
+      "grad_norm": 0.5497756004333496,
+      "learning_rate": 8.606179302824914e-05,
+      "loss": 1.0021,
+      "step": 2980
+    },
+    {
+      "epoch": 0.26635691469162553,
+      "grad_norm": 0.3906475901603699,
+      "learning_rate": 8.605176780346826e-05,
+      "loss": 1.003,
+      "step": 2981
+    },
+    {
+      "epoch": 0.2664462662228874,
+      "grad_norm": 0.4320390820503235,
+      "learning_rate": 8.604173955896756e-05,
+      "loss": 0.9869,
+      "step": 2982
+    },
+    {
+      "epoch": 0.26653561775414925,
+      "grad_norm": 0.47781896591186523,
+      "learning_rate": 8.6031708295587e-05,
+      "loss": 0.9723,
+      "step": 2983
+    },
+    {
+      "epoch": 0.26662496928541113,
+      "grad_norm": 0.4359282851219177,
+      "learning_rate": 8.602167401416678e-05,
+      "loss": 1.0909,
+      "step": 2984
+    },
+    {
+      "epoch": 0.266714320816673,
+      "grad_norm": 0.3747998774051666,
+      "learning_rate": 8.601163671554739e-05,
+      "loss": 0.9894,
+      "step": 2985
+    },
+    {
+      "epoch": 0.26680367234793484,
+      "grad_norm": 0.4070686101913452,
+      "learning_rate": 8.60015964005696e-05,
+      "loss": 0.9766,
+      "step": 2986
+    },
+    {
+      "epoch": 0.2668930238791967,
+      "grad_norm": 0.43301576375961304,
+      "learning_rate": 8.599155307007434e-05,
+      "loss": 0.9774,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2669823754104586,
+      "grad_norm": 0.436482697725296,
+      "learning_rate": 8.598150672490288e-05,
+      "loss": 1.0462,
+      "step": 2988
+    },
+    {
+      "epoch": 0.2670717269417205,
+      "grad_norm": 0.5423337817192078,
+      "learning_rate": 8.59714573658967e-05,
+      "loss": 0.9865,
+      "step": 2989
+    },
+    {
+      "epoch": 0.2671610784729823,
+      "grad_norm": 0.42116260528564453,
+      "learning_rate": 8.596140499389751e-05,
+      "loss": 1.0272,
+      "step": 2990
+    },
+    {
+      "epoch": 0.2672504300042442,
+      "grad_norm": 0.4866194725036621,
+      "learning_rate": 8.595134960974735e-05,
+      "loss": 1.0235,
+      "step": 2991
+    },
+    {
+      "epoch": 0.2673397815355061,
+      "grad_norm": 0.39662399888038635,
+      "learning_rate": 8.594129121428844e-05,
+      "loss": 0.9877,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2674291330667679,
+      "grad_norm": 0.48044952750205994,
+      "learning_rate": 8.59312298083633e-05,
+      "loss": 0.9672,
+      "step": 2993
+    },
+    {
+      "epoch": 0.2675184845980298,
+      "grad_norm": 0.40628960728645325,
+      "learning_rate": 8.592116539281466e-05,
+      "loss": 1.0034,
+      "step": 2994
+    },
+    {
+      "epoch": 0.2676078361292917,
+      "grad_norm": 0.3931500017642975,
+      "learning_rate": 8.591109796848551e-05,
+      "loss": 1.0643,
+      "step": 2995
+    },
+    {
+      "epoch": 0.26769718766055356,
+      "grad_norm": 0.4383373260498047,
+      "learning_rate": 8.590102753621913e-05,
+      "loss": 0.943,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2677865391918154,
+      "grad_norm": 0.41193974018096924,
+      "learning_rate": 8.5890954096859e-05,
+      "loss": 1.0634,
+      "step": 2997
+    },
+    {
+      "epoch": 0.26787589072307727,
+      "grad_norm": 0.480523020029068,
+      "learning_rate": 8.588087765124893e-05,
+      "loss": 0.9083,
+      "step": 2998
+    },
+    {
+      "epoch": 0.26796524225433915,
+      "grad_norm": 0.592286229133606,
+      "learning_rate": 8.587079820023287e-05,
+      "loss": 0.9664,
+      "step": 2999
+    },
+    {
+      "epoch": 0.268054593785601,
+      "grad_norm": 0.361767053604126,
+      "learning_rate": 8.586071574465511e-05,
+      "loss": 1.0011,
+      "step": 3000
+    },
+    {
+      "epoch": 0.26814394531686286,
+      "grad_norm": 0.38865599036216736,
+      "learning_rate": 8.585063028536016e-05,
+      "loss": 1.0153,
+      "step": 3001
+    },
+    {
+      "epoch": 0.26823329684812475,
+      "grad_norm": 0.4458223879337311,
+      "learning_rate": 8.584054182319279e-05,
+      "loss": 1.1077,
+      "step": 3002
+    },
+    {
+      "epoch": 0.2683226483793866,
+      "grad_norm": 0.4879177212715149,
+      "learning_rate": 8.583045035899799e-05,
+      "loss": 0.9969,
+      "step": 3003
+    },
+    {
+      "epoch": 0.26841199991064846,
+      "grad_norm": 0.47598427534103394,
+      "learning_rate": 8.582035589362107e-05,
+      "loss": 0.9383,
+      "step": 3004
+    },
+    {
+      "epoch": 0.26850135144191034,
+      "grad_norm": 0.5034103393554688,
+      "learning_rate": 8.581025842790751e-05,
+      "loss": 1.068,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2685907029731722,
+      "grad_norm": 0.3905804753303528,
+      "learning_rate": 8.58001579627031e-05,
+      "loss": 1.0858,
+      "step": 3006
+    },
+    {
+      "epoch": 0.26868005450443405,
+      "grad_norm": 0.3951498866081238,
+      "learning_rate": 8.579005449885385e-05,
+      "loss": 0.9513,
+      "step": 3007
+    },
+    {
+      "epoch": 0.26876940603569593,
+      "grad_norm": 0.44809505343437195,
+      "learning_rate": 8.577994803720606e-05,
+      "loss": 0.9417,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2688587575669578,
+      "grad_norm": 0.39570125937461853,
+      "learning_rate": 8.576983857860624e-05,
+      "loss": 1.0068,
+      "step": 3009
+    },
+    {
+      "epoch": 0.26894810909821965,
+      "grad_norm": 0.4790659546852112,
+      "learning_rate": 8.575972612390114e-05,
+      "loss": 0.9794,
+      "step": 3010
+    },
+    {
+      "epoch": 0.26903746062948153,
+      "grad_norm": 0.4047996699810028,
+      "learning_rate": 8.57496106739378e-05,
+      "loss": 0.9618,
+      "step": 3011
+    },
+    {
+      "epoch": 0.2691268121607434,
+      "grad_norm": 0.4825955629348755,
+      "learning_rate": 8.573949222956352e-05,
+      "loss": 0.9828,
+      "step": 3012
+    },
+    {
+      "epoch": 0.2692161636920053,
+      "grad_norm": 0.4488469362258911,
+      "learning_rate": 8.57293707916258e-05,
+      "loss": 0.9346,
+      "step": 3013
+    },
+    {
+      "epoch": 0.2693055152232671,
+      "grad_norm": 0.41295936703681946,
+      "learning_rate": 8.571924636097243e-05,
+      "loss": 1.007,
+      "step": 3014
+    },
+    {
+      "epoch": 0.269394866754529,
+      "grad_norm": 0.43881046772003174,
+      "learning_rate": 8.570911893845144e-05,
+      "loss": 1.061,
+      "step": 3015
+    },
+    {
+      "epoch": 0.2694842182857909,
+      "grad_norm": 0.386459618806839,
+      "learning_rate": 8.569898852491112e-05,
+      "loss": 1.0555,
+      "step": 3016
+    },
+    {
+      "epoch": 0.2695735698170527,
+      "grad_norm": 0.3662417531013489,
+      "learning_rate": 8.568885512119997e-05,
+      "loss": 1.056,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2696629213483146,
+      "grad_norm": 0.38702377676963806,
+      "learning_rate": 8.56787187281668e-05,
+      "loss": 1.0434,
+      "step": 3018
+    },
+    {
+      "epoch": 0.2697522728795765,
+      "grad_norm": 0.40131479501724243,
+      "learning_rate": 8.566857934666062e-05,
+      "loss": 0.9943,
+      "step": 3019
+    },
+    {
+      "epoch": 0.26984162441083837,
+      "grad_norm": 0.47031551599502563,
+      "learning_rate": 8.565843697753072e-05,
+      "loss": 0.973,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2699309759421002,
+      "grad_norm": 0.47709816694259644,
+      "learning_rate": 8.564829162162664e-05,
+      "loss": 0.9718,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2700203274733621,
+      "grad_norm": 0.4619588553905487,
+      "learning_rate": 8.563814327979814e-05,
+      "loss": 0.979,
+      "step": 3022
+    },
+    {
+      "epoch": 0.27010967900462396,
+      "grad_norm": 0.4841311573982239,
+      "learning_rate": 8.562799195289527e-05,
+      "loss": 0.9617,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2701990305358858,
+      "grad_norm": 0.385013222694397,
+      "learning_rate": 8.561783764176832e-05,
+      "loss": 1.0135,
+      "step": 3024
+    },
+    {
+      "epoch": 0.27028838206714767,
+      "grad_norm": 0.48148104548454285,
+      "learning_rate": 8.560768034726778e-05,
+      "loss": 0.9748,
+      "step": 3025
+    },
+    {
+      "epoch": 0.27037773359840955,
+      "grad_norm": 0.42841196060180664,
+      "learning_rate": 8.559752007024448e-05,
+      "loss": 0.9614,
+      "step": 3026
+    },
+    {
+      "epoch": 0.27046708512967144,
+      "grad_norm": 0.42978012561798096,
+      "learning_rate": 8.558735681154943e-05,
+      "loss": 1.0068,
+      "step": 3027
+    },
+    {
+      "epoch": 0.27055643666093326,
+      "grad_norm": 0.4296395778656006,
+      "learning_rate": 8.557719057203391e-05,
+      "loss": 1.0504,
+      "step": 3028
+    },
+    {
+      "epoch": 0.27064578819219515,
+      "grad_norm": 0.41547369956970215,
+      "learning_rate": 8.556702135254946e-05,
+      "loss": 1.0337,
+      "step": 3029
+    },
+    {
+      "epoch": 0.27073513972345703,
+      "grad_norm": 0.445972204208374,
+      "learning_rate": 8.555684915394786e-05,
+      "loss": 0.9933,
+      "step": 3030
+    },
+    {
+      "epoch": 0.27082449125471886,
+      "grad_norm": 0.5036845207214355,
+      "learning_rate": 8.554667397708112e-05,
+      "loss": 0.9972,
+      "step": 3031
+    },
+    {
+      "epoch": 0.27091384278598074,
+      "grad_norm": 0.4610268175601959,
+      "learning_rate": 8.553649582280155e-05,
+      "loss": 0.9926,
+      "step": 3032
+    },
+    {
+      "epoch": 0.2710031943172426,
+      "grad_norm": 0.5067682266235352,
+      "learning_rate": 8.552631469196164e-05,
+      "loss": 1.0285,
+      "step": 3033
+    },
+    {
+      "epoch": 0.27109254584850445,
+      "grad_norm": 0.3850218951702118,
+      "learning_rate": 8.551613058541421e-05,
+      "loss": 1.001,
+      "step": 3034
+    },
+    {
+      "epoch": 0.27118189737976633,
+      "grad_norm": 0.39576455950737,
+      "learning_rate": 8.550594350401225e-05,
+      "loss": 0.9882,
+      "step": 3035
+    },
+    {
+      "epoch": 0.2712712489110282,
+      "grad_norm": 0.4152897894382477,
+      "learning_rate": 8.549575344860907e-05,
+      "loss": 0.9561,
+      "step": 3036
+    },
+    {
+      "epoch": 0.2713606004422901,
+      "grad_norm": 0.6863245964050293,
+      "learning_rate": 8.548556042005819e-05,
+      "loss": 0.9862,
+      "step": 3037
+    },
+    {
+      "epoch": 0.27144995197355193,
+      "grad_norm": 0.3862869441509247,
+      "learning_rate": 8.547536441921336e-05,
+      "loss": 1.0157,
+      "step": 3038
+    },
+    {
+      "epoch": 0.2715393035048138,
+      "grad_norm": 0.37958061695098877,
+      "learning_rate": 8.546516544692861e-05,
+      "loss": 1.0548,
+      "step": 3039
+    },
+    {
+      "epoch": 0.2716286550360757,
+      "grad_norm": 0.4158911108970642,
+      "learning_rate": 8.545496350405825e-05,
+      "loss": 1.0377,
+      "step": 3040
+    },
+    {
+      "epoch": 0.2717180065673375,
+      "grad_norm": 0.43958452343940735,
+      "learning_rate": 8.544475859145676e-05,
+      "loss": 1.0113,
+      "step": 3041
+    },
+    {
+      "epoch": 0.2718073580985994,
+      "grad_norm": 0.4888957440853119,
+      "learning_rate": 8.543455070997892e-05,
+      "loss": 1.0082,
+      "step": 3042
+    },
+    {
+      "epoch": 0.2718967096298613,
+      "grad_norm": 0.4653918147087097,
+      "learning_rate": 8.542433986047977e-05,
+      "loss": 0.9827,
+      "step": 3043
+    },
+    {
+      "epoch": 0.27198606116112317,
+      "grad_norm": 0.4065740704536438,
+      "learning_rate": 8.541412604381454e-05,
+      "loss": 1.0425,
+      "step": 3044
+    },
+    {
+      "epoch": 0.272075412692385,
+      "grad_norm": 0.3832918703556061,
+      "learning_rate": 8.540390926083876e-05,
+      "loss": 0.9736,
+      "step": 3045
+    },
+    {
+      "epoch": 0.2721647642236469,
+      "grad_norm": 0.42073532938957214,
+      "learning_rate": 8.539368951240823e-05,
+      "loss": 1.0916,
+      "step": 3046
+    },
+    {
+      "epoch": 0.27225411575490877,
+      "grad_norm": 0.4821849763393402,
+      "learning_rate": 8.538346679937891e-05,
+      "loss": 1.0015,
+      "step": 3047
+    },
+    {
+      "epoch": 0.2723434672861706,
+      "grad_norm": 0.3975672125816345,
+      "learning_rate": 8.53732411226071e-05,
+      "loss": 0.9954,
+      "step": 3048
+    },
+    {
+      "epoch": 0.2724328188174325,
+      "grad_norm": 0.4945456385612488,
+      "learning_rate": 8.536301248294929e-05,
+      "loss": 1.0665,
+      "step": 3049
+    },
+    {
+      "epoch": 0.27252217034869436,
+      "grad_norm": 0.451886922121048,
+      "learning_rate": 8.535278088126225e-05,
+      "loss": 0.9445,
+      "step": 3050
+    },
+    {
+      "epoch": 0.27261152187995624,
+      "grad_norm": 0.4293072819709778,
+      "learning_rate": 8.534254631840296e-05,
+      "loss": 0.9888,
+      "step": 3051
+    },
+    {
+      "epoch": 0.27270087341121807,
+      "grad_norm": 0.4669989347457886,
+      "learning_rate": 8.53323087952287e-05,
+      "loss": 0.9923,
+      "step": 3052
+    },
+    {
+      "epoch": 0.27279022494247995,
+      "grad_norm": 0.4209103286266327,
+      "learning_rate": 8.532206831259696e-05,
+      "loss": 0.9666,
+      "step": 3053
+    },
+    {
+      "epoch": 0.27287957647374184,
+      "grad_norm": 0.41052526235580444,
+      "learning_rate": 8.531182487136549e-05,
+      "loss": 1.0202,
+      "step": 3054
+    },
+    {
+      "epoch": 0.27296892800500366,
+      "grad_norm": 0.46915799379348755,
+      "learning_rate": 8.53015784723923e-05,
+      "loss": 0.8682,
+      "step": 3055
+    },
+    {
+      "epoch": 0.27305827953626555,
+      "grad_norm": 0.3582940697669983,
+      "learning_rate": 8.529132911653563e-05,
+      "loss": 0.9678,
+      "step": 3056
+    },
+    {
+      "epoch": 0.27314763106752743,
+      "grad_norm": 0.4747898280620575,
+      "learning_rate": 8.528107680465394e-05,
+      "loss": 0.9526,
+      "step": 3057
+    },
+    {
+      "epoch": 0.2732369825987893,
+      "grad_norm": 0.43291470408439636,
+      "learning_rate": 8.527082153760601e-05,
+      "loss": 0.9883,
+      "step": 3058
+    },
+    {
+      "epoch": 0.27332633413005114,
+      "grad_norm": 0.4209512770175934,
+      "learning_rate": 8.526056331625083e-05,
+      "loss": 0.9941,
+      "step": 3059
+    },
+    {
+      "epoch": 0.273415685661313,
+      "grad_norm": 0.4599602520465851,
+      "learning_rate": 8.525030214144763e-05,
+      "loss": 1.0283,
+      "step": 3060
+    },
+    {
+      "epoch": 0.2735050371925749,
+      "grad_norm": 0.3656606078147888,
+      "learning_rate": 8.524003801405587e-05,
+      "loss": 1.017,
+      "step": 3061
+    },
+    {
+      "epoch": 0.27359438872383673,
+      "grad_norm": 0.43189749121665955,
+      "learning_rate": 8.522977093493528e-05,
+      "loss": 0.9632,
+      "step": 3062
+    },
+    {
+      "epoch": 0.2736837402550986,
+      "grad_norm": 0.5547308325767517,
+      "learning_rate": 8.521950090494587e-05,
+      "loss": 1.0423,
+      "step": 3063
+    },
+    {
+      "epoch": 0.2737730917863605,
+      "grad_norm": 0.48824217915534973,
+      "learning_rate": 8.520922792494783e-05,
+      "loss": 0.919,
+      "step": 3064
+    },
+    {
+      "epoch": 0.27386244331762233,
+      "grad_norm": 0.4252563416957855,
+      "learning_rate": 8.519895199580168e-05,
+      "loss": 0.9322,
+      "step": 3065
+    },
+    {
+      "epoch": 0.2739517948488842,
+      "grad_norm": 0.40307125449180603,
+      "learning_rate": 8.518867311836808e-05,
+      "loss": 1.0316,
+      "step": 3066
+    },
+    {
+      "epoch": 0.2740411463801461,
+      "grad_norm": 0.38896623253822327,
+      "learning_rate": 8.517839129350802e-05,
+      "loss": 1.0407,
+      "step": 3067
+    },
+    {
+      "epoch": 0.274130497911408,
+      "grad_norm": 0.40794113278388977,
+      "learning_rate": 8.516810652208272e-05,
+      "loss": 0.9766,
+      "step": 3068
+    },
+    {
+      "epoch": 0.2742198494426698,
+      "grad_norm": 0.4464515745639801,
+      "learning_rate": 8.515781880495363e-05,
+      "loss": 0.9997,
+      "step": 3069
+    },
+    {
+      "epoch": 0.2743092009739317,
+      "grad_norm": 0.3915344476699829,
+      "learning_rate": 8.514752814298248e-05,
+      "loss": 0.9563,
+      "step": 3070
+    },
+    {
+      "epoch": 0.27439855250519357,
+      "grad_norm": 0.3987552225589752,
+      "learning_rate": 8.513723453703119e-05,
+      "loss": 1.0145,
+      "step": 3071
+    },
+    {
+      "epoch": 0.2744879040364554,
+      "grad_norm": 0.3828818202018738,
+      "learning_rate": 8.512693798796196e-05,
+      "loss": 0.993,
+      "step": 3072
+    },
+    {
+      "epoch": 0.2745772555677173,
+      "grad_norm": 0.47670143842697144,
+      "learning_rate": 8.511663849663727e-05,
+      "loss": 0.9994,
+      "step": 3073
+    },
+    {
+      "epoch": 0.27466660709897917,
+      "grad_norm": 0.5165907144546509,
+      "learning_rate": 8.510633606391977e-05,
+      "loss": 0.974,
+      "step": 3074
+    },
+    {
+      "epoch": 0.27475595863024105,
+      "grad_norm": 0.49067389965057373,
+      "learning_rate": 8.509603069067243e-05,
+      "loss": 0.9133,
+      "step": 3075
+    },
+    {
+      "epoch": 0.2748453101615029,
+      "grad_norm": 0.3810344636440277,
+      "learning_rate": 8.50857223777584e-05,
+      "loss": 0.988,
+      "step": 3076
+    },
+    {
+      "epoch": 0.27493466169276476,
+      "grad_norm": 0.4502268135547638,
+      "learning_rate": 8.507541112604116e-05,
+      "loss": 0.9268,
+      "step": 3077
+    },
+    {
+      "epoch": 0.27502401322402664,
+      "grad_norm": 0.42537322640419006,
+      "learning_rate": 8.506509693638435e-05,
+      "loss": 1.0575,
+      "step": 3078
+    },
+    {
+      "epoch": 0.27511336475528847,
+      "grad_norm": 0.37440937757492065,
+      "learning_rate": 8.505477980965191e-05,
+      "loss": 1.0067,
+      "step": 3079
+    },
+    {
+      "epoch": 0.27520271628655035,
+      "grad_norm": 0.4221092164516449,
+      "learning_rate": 8.5044459746708e-05,
+      "loss": 0.9788,
+      "step": 3080
+    },
+    {
+      "epoch": 0.27529206781781224,
+      "grad_norm": 0.5111864805221558,
+      "learning_rate": 8.503413674841703e-05,
+      "loss": 0.9639,
+      "step": 3081
+    },
+    {
+      "epoch": 0.2753814193490741,
+      "grad_norm": 0.4051772654056549,
+      "learning_rate": 8.502381081564369e-05,
+      "loss": 1.0491,
+      "step": 3082
+    },
+    {
+      "epoch": 0.27547077088033595,
+      "grad_norm": 0.4286561608314514,
+      "learning_rate": 8.501348194925285e-05,
+      "loss": 0.9884,
+      "step": 3083
+    },
+    {
+      "epoch": 0.27556012241159783,
+      "grad_norm": 0.4711834490299225,
+      "learning_rate": 8.500315015010968e-05,
+      "loss": 1.025,
+      "step": 3084
+    },
+    {
+      "epoch": 0.2756494739428597,
+      "grad_norm": 0.43955665826797485,
+      "learning_rate": 8.499281541907959e-05,
+      "loss": 0.9961,
+      "step": 3085
+    },
+    {
+      "epoch": 0.27573882547412154,
+      "grad_norm": 0.38096296787261963,
+      "learning_rate": 8.498247775702821e-05,
+      "loss": 1.0081,
+      "step": 3086
+    },
+    {
+      "epoch": 0.2758281770053834,
+      "grad_norm": 0.40169036388397217,
+      "learning_rate": 8.497213716482142e-05,
+      "loss": 1.0185,
+      "step": 3087
+    },
+    {
+      "epoch": 0.2759175285366453,
+      "grad_norm": 0.4468224346637726,
+      "learning_rate": 8.496179364332539e-05,
+      "loss": 1.0679,
+      "step": 3088
+    },
+    {
+      "epoch": 0.2760068800679072,
+      "grad_norm": 0.5090979337692261,
+      "learning_rate": 8.495144719340646e-05,
+      "loss": 1.0041,
+      "step": 3089
+    },
+    {
+      "epoch": 0.276096231599169,
+      "grad_norm": 0.4413253366947174,
+      "learning_rate": 8.494109781593127e-05,
+      "loss": 0.998,
+      "step": 3090
+    },
+    {
+      "epoch": 0.2761855831304309,
+      "grad_norm": 0.5128616094589233,
+      "learning_rate": 8.49307455117667e-05,
+      "loss": 0.9282,
+      "step": 3091
+    },
+    {
+      "epoch": 0.2762749346616928,
+      "grad_norm": 0.4610481858253479,
+      "learning_rate": 8.492039028177986e-05,
+      "loss": 0.9687,
+      "step": 3092
+    },
+    {
+      "epoch": 0.2763642861929546,
+      "grad_norm": 0.42069047689437866,
+      "learning_rate": 8.491003212683811e-05,
+      "loss": 0.9726,
+      "step": 3093
+    },
+    {
+      "epoch": 0.2764536377242165,
+      "grad_norm": 0.42152658104896545,
+      "learning_rate": 8.489967104780902e-05,
+      "loss": 0.968,
+      "step": 3094
+    },
+    {
+      "epoch": 0.2765429892554784,
+      "grad_norm": 0.40881800651550293,
+      "learning_rate": 8.48893070455605e-05,
+      "loss": 1.0844,
+      "step": 3095
+    },
+    {
+      "epoch": 0.2766323407867402,
+      "grad_norm": 0.4400191009044647,
+      "learning_rate": 8.487894012096065e-05,
+      "loss": 0.9969,
+      "step": 3096
+    },
+    {
+      "epoch": 0.2767216923180021,
+      "grad_norm": 0.4783930480480194,
+      "learning_rate": 8.486857027487775e-05,
+      "loss": 0.9458,
+      "step": 3097
+    },
+    {
+      "epoch": 0.27681104384926397,
+      "grad_norm": 0.35160624980926514,
+      "learning_rate": 8.485819750818043e-05,
+      "loss": 1.0798,
+      "step": 3098
+    },
+    {
+      "epoch": 0.27690039538052585,
+      "grad_norm": 0.4336118996143341,
+      "learning_rate": 8.484782182173749e-05,
+      "loss": 1.0089,
+      "step": 3099
+    },
+    {
+      "epoch": 0.2769897469117877,
+      "grad_norm": 0.3933138847351074,
+      "learning_rate": 8.483744321641805e-05,
+      "loss": 1.0987,
+      "step": 3100
+    },
+    {
+      "epoch": 0.27707909844304957,
+      "grad_norm": 0.44518598914146423,
+      "learning_rate": 8.482706169309138e-05,
+      "loss": 0.9875,
+      "step": 3101
+    },
+    {
+      "epoch": 0.27716844997431145,
+      "grad_norm": 0.4272666275501251,
+      "learning_rate": 8.481667725262708e-05,
+      "loss": 0.986,
+      "step": 3102
+    },
+    {
+      "epoch": 0.2772578015055733,
+      "grad_norm": 0.40855729579925537,
+      "learning_rate": 8.480628989589491e-05,
+      "loss": 0.9967,
+      "step": 3103
+    },
+    {
+      "epoch": 0.27734715303683516,
+      "grad_norm": 0.42171216011047363,
+      "learning_rate": 8.479589962376497e-05,
+      "loss": 1.0869,
+      "step": 3104
+    },
+    {
+      "epoch": 0.27743650456809704,
+      "grad_norm": 0.39265212416648865,
+      "learning_rate": 8.478550643710754e-05,
+      "loss": 1.0718,
+      "step": 3105
+    },
+    {
+      "epoch": 0.2775258560993589,
+      "grad_norm": 0.4548582136631012,
+      "learning_rate": 8.477511033679317e-05,
+      "loss": 0.9926,
+      "step": 3106
+    },
+    {
+      "epoch": 0.27761520763062075,
+      "grad_norm": 0.37569287419319153,
+      "learning_rate": 8.476471132369262e-05,
+      "loss": 1.022,
+      "step": 3107
+    },
+    {
+      "epoch": 0.27770455916188264,
+      "grad_norm": 0.3934466242790222,
+      "learning_rate": 8.475430939867693e-05,
+      "loss": 0.9714,
+      "step": 3108
+    },
+    {
+      "epoch": 0.2777939106931445,
+      "grad_norm": 0.4344901144504547,
+      "learning_rate": 8.474390456261738e-05,
+      "loss": 0.9954,
+      "step": 3109
+    },
+    {
+      "epoch": 0.27788326222440635,
+      "grad_norm": 0.4985685646533966,
+      "learning_rate": 8.473349681638547e-05,
+      "loss": 0.949,
+      "step": 3110
+    },
+    {
+      "epoch": 0.27797261375566823,
+      "grad_norm": 0.40631502866744995,
+      "learning_rate": 8.472308616085298e-05,
+      "loss": 1.0339,
+      "step": 3111
+    },
+    {
+      "epoch": 0.2780619652869301,
+      "grad_norm": 0.4710947871208191,
+      "learning_rate": 8.47126725968919e-05,
+      "loss": 0.9509,
+      "step": 3112
+    },
+    {
+      "epoch": 0.278151316818192,
+      "grad_norm": 0.4414302110671997,
+      "learning_rate": 8.470225612537448e-05,
+      "loss": 0.9992,
+      "step": 3113
+    },
+    {
+      "epoch": 0.2782406683494538,
+      "grad_norm": 0.3889710605144501,
+      "learning_rate": 8.46918367471732e-05,
+      "loss": 0.976,
+      "step": 3114
+    },
+    {
+      "epoch": 0.2783300198807157,
+      "grad_norm": 0.41074416041374207,
+      "learning_rate": 8.468141446316082e-05,
+      "loss": 1.0087,
+      "step": 3115
+    },
+    {
+      "epoch": 0.2784193714119776,
+      "grad_norm": 0.47910332679748535,
+      "learning_rate": 8.46709892742103e-05,
+      "loss": 0.9827,
+      "step": 3116
+    },
+    {
+      "epoch": 0.2785087229432394,
+      "grad_norm": 0.3920195400714874,
+      "learning_rate": 8.466056118119485e-05,
+      "loss": 1.0017,
+      "step": 3117
+    },
+    {
+      "epoch": 0.2785980744745013,
+      "grad_norm": 0.4296383261680603,
+      "learning_rate": 8.465013018498795e-05,
+      "loss": 0.9948,
+      "step": 3118
+    },
+    {
+      "epoch": 0.2786874260057632,
+      "grad_norm": 0.408699631690979,
+      "learning_rate": 8.463969628646332e-05,
+      "loss": 1.0166,
+      "step": 3119
+    },
+    {
+      "epoch": 0.27877677753702507,
+      "grad_norm": 0.46718454360961914,
+      "learning_rate": 8.462925948649488e-05,
+      "loss": 0.9799,
+      "step": 3120
+    },
+    {
+      "epoch": 0.2788661290682869,
+      "grad_norm": 0.40248745679855347,
+      "learning_rate": 8.461881978595683e-05,
+      "loss": 1.0439,
+      "step": 3121
+    },
+    {
+      "epoch": 0.2789554805995488,
+      "grad_norm": 0.4227806329727173,
+      "learning_rate": 8.460837718572361e-05,
+      "loss": 1.0268,
+      "step": 3122
+    },
+    {
+      "epoch": 0.27904483213081066,
+      "grad_norm": 0.44259488582611084,
+      "learning_rate": 8.45979316866699e-05,
+      "loss": 1.0697,
+      "step": 3123
+    },
+    {
+      "epoch": 0.2791341836620725,
+      "grad_norm": 0.43050849437713623,
+      "learning_rate": 8.458748328967065e-05,
+      "loss": 1.0158,
+      "step": 3124
+    },
+    {
+      "epoch": 0.27922353519333437,
+      "grad_norm": 0.4049340486526489,
+      "learning_rate": 8.457703199560098e-05,
+      "loss": 0.9342,
+      "step": 3125
+    },
+    {
+      "epoch": 0.27931288672459625,
+      "grad_norm": 0.43839579820632935,
+      "learning_rate": 8.456657780533632e-05,
+      "loss": 0.951,
+      "step": 3126
+    },
+    {
+      "epoch": 0.27940223825585814,
+      "grad_norm": 0.44916754961013794,
+      "learning_rate": 8.45561207197523e-05,
+      "loss": 0.9717,
+      "step": 3127
+    },
+    {
+      "epoch": 0.27949158978711997,
+      "grad_norm": 0.4199538230895996,
+      "learning_rate": 8.454566073972485e-05,
+      "loss": 0.9371,
+      "step": 3128
+    },
+    {
+      "epoch": 0.27958094131838185,
+      "grad_norm": 0.4070294201374054,
+      "learning_rate": 8.453519786613007e-05,
+      "loss": 0.9972,
+      "step": 3129
+    },
+    {
+      "epoch": 0.27967029284964373,
+      "grad_norm": 0.4184653162956238,
+      "learning_rate": 8.452473209984435e-05,
+      "loss": 0.9807,
+      "step": 3130
+    },
+    {
+      "epoch": 0.27975964438090556,
+      "grad_norm": 0.4994652271270752,
+      "learning_rate": 8.451426344174433e-05,
+      "loss": 1.0829,
+      "step": 3131
+    },
+    {
+      "epoch": 0.27984899591216744,
+      "grad_norm": 0.4521782696247101,
+      "learning_rate": 8.450379189270683e-05,
+      "loss": 1.0126,
+      "step": 3132
+    },
+    {
+      "epoch": 0.2799383474434293,
+      "grad_norm": 0.43609222769737244,
+      "learning_rate": 8.449331745360898e-05,
+      "loss": 1.0078,
+      "step": 3133
+    },
+    {
+      "epoch": 0.28002769897469115,
+      "grad_norm": 0.4149281680583954,
+      "learning_rate": 8.448284012532812e-05,
+      "loss": 0.9348,
+      "step": 3134
+    },
+    {
+      "epoch": 0.28011705050595304,
+      "grad_norm": 0.4595349431037903,
+      "learning_rate": 8.447235990874182e-05,
+      "loss": 1.0162,
+      "step": 3135
+    },
+    {
+      "epoch": 0.2802064020372149,
+      "grad_norm": 0.48324936628341675,
+      "learning_rate": 8.446187680472797e-05,
+      "loss": 0.9531,
+      "step": 3136
+    },
+    {
+      "epoch": 0.2802957535684768,
+      "grad_norm": 0.40329238772392273,
+      "learning_rate": 8.445139081416458e-05,
+      "loss": 0.977,
+      "step": 3137
+    },
+    {
+      "epoch": 0.28038510509973863,
+      "grad_norm": 0.5319374203681946,
+      "learning_rate": 8.444090193792997e-05,
+      "loss": 0.9266,
+      "step": 3138
+    },
+    {
+      "epoch": 0.2804744566310005,
+      "grad_norm": 0.43720632791519165,
+      "learning_rate": 8.443041017690271e-05,
+      "loss": 1.049,
+      "step": 3139
+    },
+    {
+      "epoch": 0.2805638081622624,
+      "grad_norm": 0.4096292555332184,
+      "learning_rate": 8.441991553196162e-05,
+      "loss": 1.027,
+      "step": 3140
+    },
+    {
+      "epoch": 0.2806531596935242,
+      "grad_norm": 0.4749426543712616,
+      "learning_rate": 8.440941800398571e-05,
+      "loss": 0.9608,
+      "step": 3141
+    },
+    {
+      "epoch": 0.2807425112247861,
+      "grad_norm": 0.40563464164733887,
+      "learning_rate": 8.439891759385425e-05,
+      "loss": 1.025,
+      "step": 3142
+    },
+    {
+      "epoch": 0.280831862756048,
+      "grad_norm": 0.4449627101421356,
+      "learning_rate": 8.438841430244678e-05,
+      "loss": 0.9698,
+      "step": 3143
+    },
+    {
+      "epoch": 0.2809212142873099,
+      "grad_norm": 0.4511934816837311,
+      "learning_rate": 8.437790813064305e-05,
+      "loss": 0.9556,
+      "step": 3144
+    },
+    {
+      "epoch": 0.2810105658185717,
+      "grad_norm": 0.4390813112258911,
+      "learning_rate": 8.43673990793231e-05,
+      "loss": 1.0457,
+      "step": 3145
+    },
+    {
+      "epoch": 0.2810999173498336,
+      "grad_norm": 0.4567374885082245,
+      "learning_rate": 8.435688714936711e-05,
+      "loss": 0.9987,
+      "step": 3146
+    },
+    {
+      "epoch": 0.28118926888109547,
+      "grad_norm": 0.39277079701423645,
+      "learning_rate": 8.434637234165563e-05,
+      "loss": 1.0383,
+      "step": 3147
+    },
+    {
+      "epoch": 0.2812786204123573,
+      "grad_norm": 0.36896324157714844,
+      "learning_rate": 8.433585465706934e-05,
+      "loss": 0.9997,
+      "step": 3148
+    },
+    {
+      "epoch": 0.2813679719436192,
+      "grad_norm": 0.40107041597366333,
+      "learning_rate": 8.432533409648925e-05,
+      "loss": 1.0192,
+      "step": 3149
+    },
+    {
+      "epoch": 0.28145732347488106,
+      "grad_norm": 0.5373698472976685,
+      "learning_rate": 8.431481066079654e-05,
+      "loss": 0.9916,
+      "step": 3150
+    },
+    {
+      "epoch": 0.28154667500614294,
+      "grad_norm": 0.4341030418872833,
+      "learning_rate": 8.430428435087267e-05,
+      "loss": 1.0095,
+      "step": 3151
+    },
+    {
+      "epoch": 0.28163602653740477,
+      "grad_norm": 0.4856875240802765,
+      "learning_rate": 8.429375516759933e-05,
+      "loss": 0.9375,
+      "step": 3152
+    },
+    {
+      "epoch": 0.28172537806866665,
+      "grad_norm": 0.4683944880962372,
+      "learning_rate": 8.428322311185844e-05,
+      "loss": 1.0203,
+      "step": 3153
+    },
+    {
+      "epoch": 0.28181472959992854,
+      "grad_norm": 0.39955028891563416,
+      "learning_rate": 8.427268818453218e-05,
+      "loss": 1.0851,
+      "step": 3154
+    },
+    {
+      "epoch": 0.28190408113119036,
+      "grad_norm": 0.4985107481479645,
+      "learning_rate": 8.426215038650296e-05,
+      "loss": 0.9497,
+      "step": 3155
+    },
+    {
+      "epoch": 0.28199343266245225,
+      "grad_norm": 0.40328237414360046,
+      "learning_rate": 8.425160971865345e-05,
+      "loss": 1.019,
+      "step": 3156
+    },
+    {
+      "epoch": 0.28208278419371413,
+      "grad_norm": 0.4598018527030945,
+      "learning_rate": 8.424106618186653e-05,
+      "loss": 0.8932,
+      "step": 3157
+    },
+    {
+      "epoch": 0.282172135724976,
+      "grad_norm": 0.46386003494262695,
+      "learning_rate": 8.423051977702532e-05,
+      "loss": 1.0043,
+      "step": 3158
+    },
+    {
+      "epoch": 0.28226148725623784,
+      "grad_norm": 0.4086954593658447,
+      "learning_rate": 8.421997050501323e-05,
+      "loss": 1.0223,
+      "step": 3159
+    },
+    {
+      "epoch": 0.2823508387874997,
+      "grad_norm": 0.4212207794189453,
+      "learning_rate": 8.420941836671384e-05,
+      "loss": 0.9869,
+      "step": 3160
+    },
+    {
+      "epoch": 0.2824401903187616,
+      "grad_norm": 0.3890816271305084,
+      "learning_rate": 8.419886336301099e-05,
+      "loss": 1.0268,
+      "step": 3161
+    },
+    {
+      "epoch": 0.28252954185002344,
+      "grad_norm": 0.39444929361343384,
+      "learning_rate": 8.418830549478885e-05,
+      "loss": 1.0987,
+      "step": 3162
+    },
+    {
+      "epoch": 0.2826188933812853,
+      "grad_norm": 0.43861931562423706,
+      "learning_rate": 8.417774476293167e-05,
+      "loss": 0.922,
+      "step": 3163
+    },
+    {
+      "epoch": 0.2827082449125472,
+      "grad_norm": 0.42500317096710205,
+      "learning_rate": 8.416718116832406e-05,
+      "loss": 0.9877,
+      "step": 3164
+    },
+    {
+      "epoch": 0.28279759644380903,
+      "grad_norm": 0.4124399721622467,
+      "learning_rate": 8.415661471185084e-05,
+      "loss": 1.0055,
+      "step": 3165
+    },
+    {
+      "epoch": 0.2828869479750709,
+      "grad_norm": 0.41332152485847473,
+      "learning_rate": 8.414604539439704e-05,
+      "loss": 1.0422,
+      "step": 3166
+    },
+    {
+      "epoch": 0.2829762995063328,
+      "grad_norm": 0.42248642444610596,
+      "learning_rate": 8.413547321684798e-05,
+      "loss": 1.0358,
+      "step": 3167
+    },
+    {
+      "epoch": 0.2830656510375947,
+      "grad_norm": 0.42159736156463623,
+      "learning_rate": 8.412489818008918e-05,
+      "loss": 1.0113,
+      "step": 3168
+    },
+    {
+      "epoch": 0.2831550025688565,
+      "grad_norm": 0.41881147027015686,
+      "learning_rate": 8.41143202850064e-05,
+      "loss": 1.0388,
+      "step": 3169
+    },
+    {
+      "epoch": 0.2832443541001184,
+      "grad_norm": 0.4449400007724762,
+      "learning_rate": 8.41037395324857e-05,
+      "loss": 1.0147,
+      "step": 3170
+    },
+    {
+      "epoch": 0.2833337056313803,
+      "grad_norm": 0.4245634377002716,
+      "learning_rate": 8.409315592341324e-05,
+      "loss": 1.0661,
+      "step": 3171
+    },
+    {
+      "epoch": 0.2834230571626421,
+      "grad_norm": 0.4041271209716797,
+      "learning_rate": 8.408256945867561e-05,
+      "loss": 1.0927,
+      "step": 3172
+    },
+    {
+      "epoch": 0.283512408693904,
+      "grad_norm": 0.5533398389816284,
+      "learning_rate": 8.407198013915947e-05,
+      "loss": 0.9166,
+      "step": 3173
+    },
+    {
+      "epoch": 0.28360176022516587,
+      "grad_norm": 0.39345189929008484,
+      "learning_rate": 8.406138796575183e-05,
+      "loss": 0.9596,
+      "step": 3174
+    },
+    {
+      "epoch": 0.28369111175642775,
+      "grad_norm": 0.39271482825279236,
+      "learning_rate": 8.405079293933986e-05,
+      "loss": 1.0173,
+      "step": 3175
+    },
+    {
+      "epoch": 0.2837804632876896,
+      "grad_norm": 0.38316190242767334,
+      "learning_rate": 8.404019506081104e-05,
+      "loss": 0.9547,
+      "step": 3176
+    },
+    {
+      "epoch": 0.28386981481895146,
+      "grad_norm": 0.3933804929256439,
+      "learning_rate": 8.402959433105303e-05,
+      "loss": 0.9311,
+      "step": 3177
+    },
+    {
+      "epoch": 0.28395916635021334,
+      "grad_norm": 0.39881816506385803,
+      "learning_rate": 8.401899075095378e-05,
+      "loss": 0.9956,
+      "step": 3178
+    },
+    {
+      "epoch": 0.28404851788147517,
+      "grad_norm": 0.4697539806365967,
+      "learning_rate": 8.400838432140142e-05,
+      "loss": 0.935,
+      "step": 3179
+    },
+    {
+      "epoch": 0.28413786941273705,
+      "grad_norm": 0.4180028438568115,
+      "learning_rate": 8.399777504328438e-05,
+      "loss": 1.0263,
+      "step": 3180
+    },
+    {
+      "epoch": 0.28422722094399894,
+      "grad_norm": 0.4734386205673218,
+      "learning_rate": 8.398716291749129e-05,
+      "loss": 0.911,
+      "step": 3181
+    },
+    {
+      "epoch": 0.2843165724752608,
+      "grad_norm": 0.41403621435165405,
+      "learning_rate": 8.397654794491103e-05,
+      "loss": 1.0165,
+      "step": 3182
+    },
+    {
+      "epoch": 0.28440592400652265,
+      "grad_norm": 0.5192162990570068,
+      "learning_rate": 8.396593012643272e-05,
+      "loss": 0.9693,
+      "step": 3183
+    },
+    {
+      "epoch": 0.28449527553778453,
+      "grad_norm": 0.40429311990737915,
+      "learning_rate": 8.39553094629457e-05,
+      "loss": 1.0161,
+      "step": 3184
+    },
+    {
+      "epoch": 0.2845846270690464,
+      "grad_norm": 0.43155255913734436,
+      "learning_rate": 8.394468595533958e-05,
+      "loss": 1.0094,
+      "step": 3185
+    },
+    {
+      "epoch": 0.28467397860030824,
+      "grad_norm": 0.4038214683532715,
+      "learning_rate": 8.393405960450418e-05,
+      "loss": 1.0635,
+      "step": 3186
+    },
+    {
+      "epoch": 0.2847633301315701,
+      "grad_norm": 0.5239593982696533,
+      "learning_rate": 8.392343041132959e-05,
+      "loss": 0.9088,
+      "step": 3187
+    },
+    {
+      "epoch": 0.284852681662832,
+      "grad_norm": 0.4169578552246094,
+      "learning_rate": 8.391279837670609e-05,
+      "loss": 0.9349,
+      "step": 3188
+    },
+    {
+      "epoch": 0.2849420331940939,
+      "grad_norm": 0.5473026633262634,
+      "learning_rate": 8.390216350152425e-05,
+      "loss": 1.0211,
+      "step": 3189
+    },
+    {
+      "epoch": 0.2850313847253557,
+      "grad_norm": 0.45344579219818115,
+      "learning_rate": 8.389152578667485e-05,
+      "loss": 1.0067,
+      "step": 3190
+    },
+    {
+      "epoch": 0.2851207362566176,
+      "grad_norm": 0.4618963599205017,
+      "learning_rate": 8.38808852330489e-05,
+      "loss": 0.996,
+      "step": 3191
+    },
+    {
+      "epoch": 0.2852100877878795,
+      "grad_norm": 0.5052065253257751,
+      "learning_rate": 8.387024184153766e-05,
+      "loss": 0.9071,
+      "step": 3192
+    },
+    {
+      "epoch": 0.2852994393191413,
+      "grad_norm": 0.413125604391098,
+      "learning_rate": 8.385959561303265e-05,
+      "loss": 1.022,
+      "step": 3193
+    },
+    {
+      "epoch": 0.2853887908504032,
+      "grad_norm": 0.3971850574016571,
+      "learning_rate": 8.384894654842559e-05,
+      "loss": 1.0205,
+      "step": 3194
+    },
+    {
+      "epoch": 0.2854781423816651,
+      "grad_norm": 0.3980804681777954,
+      "learning_rate": 8.383829464860844e-05,
+      "loss": 1.0146,
+      "step": 3195
+    },
+    {
+      "epoch": 0.2855674939129269,
+      "grad_norm": 0.46941179037094116,
+      "learning_rate": 8.382763991447344e-05,
+      "loss": 1.0008,
+      "step": 3196
+    },
+    {
+      "epoch": 0.2856568454441888,
+      "grad_norm": 0.38151049613952637,
+      "learning_rate": 8.381698234691301e-05,
+      "loss": 1.0283,
+      "step": 3197
+    },
+    {
+      "epoch": 0.2857461969754507,
+      "grad_norm": 0.5456905364990234,
+      "learning_rate": 8.380632194681986e-05,
+      "loss": 0.92,
+      "step": 3198
+    },
+    {
+      "epoch": 0.28583554850671256,
+      "grad_norm": 0.3786112368106842,
+      "learning_rate": 8.379565871508688e-05,
+      "loss": 0.9825,
+      "step": 3199
+    },
+    {
+      "epoch": 0.2859249000379744,
+      "grad_norm": 0.48761385679244995,
+      "learning_rate": 8.378499265260724e-05,
+      "loss": 0.9437,
+      "step": 3200
+    },
+    {
+      "epoch": 0.28601425156923627,
+      "grad_norm": 0.4335952401161194,
+      "learning_rate": 8.377432376027437e-05,
+      "loss": 0.9783,
+      "step": 3201
+    },
+    {
+      "epoch": 0.28610360310049815,
+      "grad_norm": 0.44660720229148865,
+      "learning_rate": 8.376365203898184e-05,
+      "loss": 1.0364,
+      "step": 3202
+    },
+    {
+      "epoch": 0.28619295463176,
+      "grad_norm": 0.4328174889087677,
+      "learning_rate": 8.375297748962358e-05,
+      "loss": 1.002,
+      "step": 3203
+    },
+    {
+      "epoch": 0.28628230616302186,
+      "grad_norm": 0.4029334783554077,
+      "learning_rate": 8.374230011309368e-05,
+      "loss": 0.9958,
+      "step": 3204
+    },
+    {
+      "epoch": 0.28637165769428374,
+      "grad_norm": 0.49282363057136536,
+      "learning_rate": 8.373161991028646e-05,
+      "loss": 0.8783,
+      "step": 3205
+    },
+    {
+      "epoch": 0.2864610092255456,
+      "grad_norm": 0.4617169499397278,
+      "learning_rate": 8.372093688209655e-05,
+      "loss": 1.0666,
+      "step": 3206
+    },
+    {
+      "epoch": 0.28655036075680745,
+      "grad_norm": 0.44560667872428894,
+      "learning_rate": 8.371025102941872e-05,
+      "loss": 0.9826,
+      "step": 3207
+    },
+    {
+      "epoch": 0.28663971228806934,
+      "grad_norm": 0.4756043255329132,
+      "learning_rate": 8.369956235314802e-05,
+      "loss": 0.9956,
+      "step": 3208
+    },
+    {
+      "epoch": 0.2867290638193312,
+      "grad_norm": 0.4700027108192444,
+      "learning_rate": 8.368887085417977e-05,
+      "loss": 1.0006,
+      "step": 3209
+    },
+    {
+      "epoch": 0.28681841535059305,
+      "grad_norm": 0.47527703642845154,
+      "learning_rate": 8.367817653340951e-05,
+      "loss": 0.9828,
+      "step": 3210
+    },
+    {
+      "epoch": 0.28690776688185493,
+      "grad_norm": 0.5730937719345093,
+      "learning_rate": 8.366747939173297e-05,
+      "loss": 0.9436,
+      "step": 3211
+    },
+    {
+      "epoch": 0.2869971184131168,
+      "grad_norm": 0.47630396485328674,
+      "learning_rate": 8.365677943004617e-05,
+      "loss": 0.9992,
+      "step": 3212
+    },
+    {
+      "epoch": 0.2870864699443787,
+      "grad_norm": 0.40669214725494385,
+      "learning_rate": 8.364607664924533e-05,
+      "loss": 1.015,
+      "step": 3213
+    },
+    {
+      "epoch": 0.2871758214756405,
+      "grad_norm": 0.3980100750923157,
+      "learning_rate": 8.363537105022692e-05,
+      "loss": 0.968,
+      "step": 3214
+    },
+    {
+      "epoch": 0.2872651730069024,
+      "grad_norm": 0.4197221100330353,
+      "learning_rate": 8.362466263388766e-05,
+      "loss": 1.0064,
+      "step": 3215
+    },
+    {
+      "epoch": 0.2873545245381643,
+      "grad_norm": 0.42029672861099243,
+      "learning_rate": 8.36139514011245e-05,
+      "loss": 1.0118,
+      "step": 3216
+    },
+    {
+      "epoch": 0.2874438760694261,
+      "grad_norm": 0.3882405757904053,
+      "learning_rate": 8.36032373528346e-05,
+      "loss": 1.0255,
+      "step": 3217
+    },
+    {
+      "epoch": 0.287533227600688,
+      "grad_norm": 0.4027383625507355,
+      "learning_rate": 8.359252048991539e-05,
+      "loss": 1.013,
+      "step": 3218
+    },
+    {
+      "epoch": 0.2876225791319499,
+      "grad_norm": 0.4605361819267273,
+      "learning_rate": 8.358180081326454e-05,
+      "loss": 0.9371,
+      "step": 3219
+    },
+    {
+      "epoch": 0.28771193066321177,
+      "grad_norm": 0.4364441931247711,
+      "learning_rate": 8.357107832377988e-05,
+      "loss": 0.9737,
+      "step": 3220
+    },
+    {
+      "epoch": 0.2878012821944736,
+      "grad_norm": 0.4793910086154938,
+      "learning_rate": 8.356035302235959e-05,
+      "loss": 0.9819,
+      "step": 3221
+    },
+    {
+      "epoch": 0.2878906337257355,
+      "grad_norm": 0.39052850008010864,
+      "learning_rate": 8.354962490990202e-05,
+      "loss": 1.0593,
+      "step": 3222
+    },
+    {
+      "epoch": 0.28797998525699736,
+      "grad_norm": 0.4997557997703552,
+      "learning_rate": 8.353889398730572e-05,
+      "loss": 1.0613,
+      "step": 3223
+    },
+    {
+      "epoch": 0.2880693367882592,
+      "grad_norm": 0.398324191570282,
+      "learning_rate": 8.352816025546957e-05,
+      "loss": 0.9801,
+      "step": 3224
+    },
+    {
+      "epoch": 0.2881586883195211,
+      "grad_norm": 0.5084810256958008,
+      "learning_rate": 8.351742371529261e-05,
+      "loss": 1.0013,
+      "step": 3225
+    },
+    {
+      "epoch": 0.28824803985078296,
+      "grad_norm": 0.4650806486606598,
+      "learning_rate": 8.350668436767413e-05,
+      "loss": 0.9618,
+      "step": 3226
+    },
+    {
+      "epoch": 0.2883373913820448,
+      "grad_norm": 0.43210747838020325,
+      "learning_rate": 8.349594221351368e-05,
+      "loss": 0.9195,
+      "step": 3227
+    },
+    {
+      "epoch": 0.28842674291330667,
+      "grad_norm": 0.4068807363510132,
+      "learning_rate": 8.348519725371105e-05,
+      "loss": 1.1076,
+      "step": 3228
+    },
+    {
+      "epoch": 0.28851609444456855,
+      "grad_norm": 0.4721238911151886,
+      "learning_rate": 8.347444948916622e-05,
+      "loss": 1.0048,
+      "step": 3229
+    },
+    {
+      "epoch": 0.28860544597583043,
+      "grad_norm": 0.4541776776313782,
+      "learning_rate": 8.346369892077942e-05,
+      "loss": 0.9736,
+      "step": 3230
+    },
+    {
+      "epoch": 0.28869479750709226,
+      "grad_norm": 0.5041652917861938,
+      "learning_rate": 8.345294554945113e-05,
+      "loss": 0.9833,
+      "step": 3231
+    },
+    {
+      "epoch": 0.28878414903835414,
+      "grad_norm": 0.48984166979789734,
+      "learning_rate": 8.344218937608211e-05,
+      "loss": 0.9144,
+      "step": 3232
+    },
+    {
+      "epoch": 0.288873500569616,
+      "grad_norm": 0.3938770294189453,
+      "learning_rate": 8.343143040157322e-05,
+      "loss": 0.9965,
+      "step": 3233
+    },
+    {
+      "epoch": 0.28896285210087785,
+      "grad_norm": 0.36259424686431885,
+      "learning_rate": 8.34206686268257e-05,
+      "loss": 1.0121,
+      "step": 3234
+    },
+    {
+      "epoch": 0.28905220363213974,
+      "grad_norm": 0.46610236167907715,
+      "learning_rate": 8.340990405274091e-05,
+      "loss": 0.9109,
+      "step": 3235
+    },
+    {
+      "epoch": 0.2891415551634016,
+      "grad_norm": 0.39891326427459717,
+      "learning_rate": 8.339913668022057e-05,
+      "loss": 0.9904,
+      "step": 3236
+    },
+    {
+      "epoch": 0.2892309066946635,
+      "grad_norm": 0.4756152629852295,
+      "learning_rate": 8.338836651016652e-05,
+      "loss": 1.0002,
+      "step": 3237
+    },
+    {
+      "epoch": 0.28932025822592533,
+      "grad_norm": 0.42526543140411377,
+      "learning_rate": 8.337759354348087e-05,
+      "loss": 1.0233,
+      "step": 3238
+    },
+    {
+      "epoch": 0.2894096097571872,
+      "grad_norm": 0.4663248360157013,
+      "learning_rate": 8.336681778106598e-05,
+      "loss": 1.0187,
+      "step": 3239
+    },
+    {
+      "epoch": 0.2894989612884491,
+      "grad_norm": 0.39710041880607605,
+      "learning_rate": 8.335603922382444e-05,
+      "loss": 1.0259,
+      "step": 3240
+    },
+    {
+      "epoch": 0.2895883128197109,
+      "grad_norm": 0.4082708954811096,
+      "learning_rate": 8.334525787265905e-05,
+      "loss": 0.9772,
+      "step": 3241
+    },
+    {
+      "epoch": 0.2896776643509728,
+      "grad_norm": 0.41001391410827637,
+      "learning_rate": 8.333447372847291e-05,
+      "loss": 0.987,
+      "step": 3242
+    },
+    {
+      "epoch": 0.2897670158822347,
+      "grad_norm": 0.39033588767051697,
+      "learning_rate": 8.332368679216927e-05,
+      "loss": 0.9732,
+      "step": 3243
+    },
+    {
+      "epoch": 0.2898563674134966,
+      "grad_norm": 0.4427189230918884,
+      "learning_rate": 8.331289706465162e-05,
+      "loss": 1.0081,
+      "step": 3244
+    },
+    {
+      "epoch": 0.2899457189447584,
+      "grad_norm": 0.3782840669155121,
+      "learning_rate": 8.330210454682379e-05,
+      "loss": 1.0221,
+      "step": 3245
+    },
+    {
+      "epoch": 0.2900350704760203,
+      "grad_norm": 0.460595041513443,
+      "learning_rate": 8.32913092395897e-05,
+      "loss": 1.0137,
+      "step": 3246
+    },
+    {
+      "epoch": 0.29012442200728217,
+      "grad_norm": 0.45908665657043457,
+      "learning_rate": 8.328051114385362e-05,
+      "loss": 0.9745,
+      "step": 3247
+    },
+    {
+      "epoch": 0.290213773538544,
+      "grad_norm": 0.5565299987792969,
+      "learning_rate": 8.326971026052e-05,
+      "loss": 0.87,
+      "step": 3248
+    },
+    {
+      "epoch": 0.2903031250698059,
+      "grad_norm": 0.37024036049842834,
+      "learning_rate": 8.32589065904935e-05,
+      "loss": 1.031,
+      "step": 3249
+    },
+    {
+      "epoch": 0.29039247660106776,
+      "grad_norm": 0.42501765489578247,
+      "learning_rate": 8.324810013467905e-05,
+      "loss": 0.9865,
+      "step": 3250
+    },
+    {
+      "epoch": 0.29048182813232964,
+      "grad_norm": 0.38723427057266235,
+      "learning_rate": 8.323729089398181e-05,
+      "loss": 1.0355,
+      "step": 3251
+    },
+    {
+      "epoch": 0.2905711796635915,
+      "grad_norm": 0.37960848212242126,
+      "learning_rate": 8.322647886930718e-05,
+      "loss": 0.955,
+      "step": 3252
+    },
+    {
+      "epoch": 0.29066053119485336,
+      "grad_norm": 0.3721432089805603,
+      "learning_rate": 8.32156640615608e-05,
+      "loss": 0.982,
+      "step": 3253
+    },
+    {
+      "epoch": 0.29074988272611524,
+      "grad_norm": 0.43282589316368103,
+      "learning_rate": 8.320484647164848e-05,
+      "loss": 1.0314,
+      "step": 3254
+    },
+    {
+      "epoch": 0.29083923425737707,
+      "grad_norm": 0.4642455577850342,
+      "learning_rate": 8.319402610047633e-05,
+      "loss": 0.9311,
+      "step": 3255
+    },
+    {
+      "epoch": 0.29092858578863895,
+      "grad_norm": 0.3716782331466675,
+      "learning_rate": 8.318320294895067e-05,
+      "loss": 1.0325,
+      "step": 3256
+    },
+    {
+      "epoch": 0.29101793731990083,
+      "grad_norm": 0.4215312600135803,
+      "learning_rate": 8.317237701797807e-05,
+      "loss": 0.9463,
+      "step": 3257
+    },
+    {
+      "epoch": 0.29110728885116266,
+      "grad_norm": 0.49139776825904846,
+      "learning_rate": 8.316154830846528e-05,
+      "loss": 0.9384,
+      "step": 3258
+    },
+    {
+      "epoch": 0.29119664038242454,
+      "grad_norm": 0.4280232787132263,
+      "learning_rate": 8.315071682131936e-05,
+      "loss": 1.0093,
+      "step": 3259
+    },
+    {
+      "epoch": 0.2912859919136864,
+      "grad_norm": 0.42184942960739136,
+      "learning_rate": 8.313988255744754e-05,
+      "loss": 0.9151,
+      "step": 3260
+    },
+    {
+      "epoch": 0.2913753434449483,
+      "grad_norm": 0.44842901825904846,
+      "learning_rate": 8.312904551775731e-05,
+      "loss": 1.0307,
+      "step": 3261
+    },
+    {
+      "epoch": 0.29146469497621014,
+      "grad_norm": 0.39945071935653687,
+      "learning_rate": 8.311820570315639e-05,
+      "loss": 1.0681,
+      "step": 3262
+    },
+    {
+      "epoch": 0.291554046507472,
+      "grad_norm": 0.4484541416168213,
+      "learning_rate": 8.310736311455271e-05,
+      "loss": 0.9827,
+      "step": 3263
+    },
+    {
+      "epoch": 0.2916433980387339,
+      "grad_norm": 0.39152559638023376,
+      "learning_rate": 8.30965177528545e-05,
+      "loss": 1.0095,
+      "step": 3264
+    },
+    {
+      "epoch": 0.29173274956999573,
+      "grad_norm": 0.4719206392765045,
+      "learning_rate": 8.308566961897012e-05,
+      "loss": 0.9007,
+      "step": 3265
+    },
+    {
+      "epoch": 0.2918221011012576,
+      "grad_norm": 0.46025243401527405,
+      "learning_rate": 8.307481871380824e-05,
+      "loss": 1.0282,
+      "step": 3266
+    },
+    {
+      "epoch": 0.2919114526325195,
+      "grad_norm": 0.41656240820884705,
+      "learning_rate": 8.306396503827775e-05,
+      "loss": 0.9827,
+      "step": 3267
+    },
+    {
+      "epoch": 0.2920008041637814,
+      "grad_norm": 0.440974622964859,
+      "learning_rate": 8.305310859328777e-05,
+      "loss": 0.9863,
+      "step": 3268
+    },
+    {
+      "epoch": 0.2920901556950432,
+      "grad_norm": 0.3946766257286072,
+      "learning_rate": 8.30422493797476e-05,
+      "loss": 0.9565,
+      "step": 3269
+    },
+    {
+      "epoch": 0.2921795072263051,
+      "grad_norm": 0.47271499037742615,
+      "learning_rate": 8.303138739856684e-05,
+      "loss": 0.9492,
+      "step": 3270
+    },
+    {
+      "epoch": 0.292268858757567,
+      "grad_norm": 0.4621809124946594,
+      "learning_rate": 8.302052265065531e-05,
+      "loss": 0.926,
+      "step": 3271
+    },
+    {
+      "epoch": 0.2923582102888288,
+      "grad_norm": 0.4402875006198883,
+      "learning_rate": 8.300965513692303e-05,
+      "loss": 1.0155,
+      "step": 3272
+    },
+    {
+      "epoch": 0.2924475618200907,
+      "grad_norm": 0.4282824695110321,
+      "learning_rate": 8.299878485828028e-05,
+      "loss": 1.0171,
+      "step": 3273
+    },
+    {
+      "epoch": 0.29253691335135257,
+      "grad_norm": 0.47499823570251465,
+      "learning_rate": 8.298791181563754e-05,
+      "loss": 0.9616,
+      "step": 3274
+    },
+    {
+      "epoch": 0.29262626488261445,
+      "grad_norm": 0.4334074556827545,
+      "learning_rate": 8.297703600990556e-05,
+      "loss": 0.9689,
+      "step": 3275
+    },
+    {
+      "epoch": 0.2927156164138763,
+      "grad_norm": 0.42571699619293213,
+      "learning_rate": 8.296615744199532e-05,
+      "loss": 1.152,
+      "step": 3276
+    },
+    {
+      "epoch": 0.29280496794513816,
+      "grad_norm": 0.473748117685318,
+      "learning_rate": 8.295527611281799e-05,
+      "loss": 0.9805,
+      "step": 3277
+    },
+    {
+      "epoch": 0.29289431947640004,
+      "grad_norm": 0.4350266456604004,
+      "learning_rate": 8.2944392023285e-05,
+      "loss": 1.0552,
+      "step": 3278
+    },
+    {
+      "epoch": 0.29298367100766187,
+      "grad_norm": 0.43737518787384033,
+      "learning_rate": 8.293350517430805e-05,
+      "loss": 0.9802,
+      "step": 3279
+    },
+    {
+      "epoch": 0.29307302253892376,
+      "grad_norm": 0.45549145340919495,
+      "learning_rate": 8.292261556679897e-05,
+      "loss": 0.9499,
+      "step": 3280
+    },
+    {
+      "epoch": 0.29316237407018564,
+      "grad_norm": 0.4989168345928192,
+      "learning_rate": 8.29117232016699e-05,
+      "loss": 0.9162,
+      "step": 3281
+    },
+    {
+      "epoch": 0.2932517256014475,
+      "grad_norm": 0.4217113256454468,
+      "learning_rate": 8.29008280798332e-05,
+      "loss": 0.9838,
+      "step": 3282
+    },
+    {
+      "epoch": 0.29334107713270935,
+      "grad_norm": 0.382791668176651,
+      "learning_rate": 8.288993020220147e-05,
+      "loss": 1.0894,
+      "step": 3283
+    },
+    {
+      "epoch": 0.29343042866397123,
+      "grad_norm": 0.4287773072719574,
+      "learning_rate": 8.287902956968748e-05,
+      "loss": 0.9894,
+      "step": 3284
+    },
+    {
+      "epoch": 0.2935197801952331,
+      "grad_norm": 0.4374977648258209,
+      "learning_rate": 8.286812618320431e-05,
+      "loss": 0.9899,
+      "step": 3285
+    },
+    {
+      "epoch": 0.29360913172649494,
+      "grad_norm": 0.37822476029396057,
+      "learning_rate": 8.285722004366522e-05,
+      "loss": 1.0091,
+      "step": 3286
+    },
+    {
+      "epoch": 0.2936984832577568,
+      "grad_norm": 0.4243577718734741,
+      "learning_rate": 8.284631115198371e-05,
+      "loss": 0.9796,
+      "step": 3287
+    },
+    {
+      "epoch": 0.2937878347890187,
+      "grad_norm": 0.556516706943512,
+      "learning_rate": 8.283539950907356e-05,
+      "loss": 0.9766,
+      "step": 3288
+    },
+    {
+      "epoch": 0.29387718632028054,
+      "grad_norm": 0.4003793001174927,
+      "learning_rate": 8.282448511584866e-05,
+      "loss": 0.9797,
+      "step": 3289
+    },
+    {
+      "epoch": 0.2939665378515424,
+      "grad_norm": 0.39269542694091797,
+      "learning_rate": 8.281356797322327e-05,
+      "loss": 0.966,
+      "step": 3290
+    },
+    {
+      "epoch": 0.2940558893828043,
+      "grad_norm": 0.4365537464618683,
+      "learning_rate": 8.280264808211179e-05,
+      "loss": 1.0566,
+      "step": 3291
+    },
+    {
+      "epoch": 0.2941452409140662,
+      "grad_norm": 0.42626434564590454,
+      "learning_rate": 8.279172544342888e-05,
+      "loss": 0.9358,
+      "step": 3292
+    },
+    {
+      "epoch": 0.294234592445328,
+      "grad_norm": 0.5638636946678162,
+      "learning_rate": 8.278080005808943e-05,
+      "loss": 0.9842,
+      "step": 3293
+    },
+    {
+      "epoch": 0.2943239439765899,
+      "grad_norm": 0.3928448259830475,
+      "learning_rate": 8.276987192700856e-05,
+      "loss": 1.0545,
+      "step": 3294
+    },
+    {
+      "epoch": 0.2944132955078518,
+      "grad_norm": 0.4263659119606018,
+      "learning_rate": 8.275894105110161e-05,
+      "loss": 0.9396,
+      "step": 3295
+    },
+    {
+      "epoch": 0.2945026470391136,
+      "grad_norm": 0.4779483377933502,
+      "learning_rate": 8.274800743128417e-05,
+      "loss": 1.0356,
+      "step": 3296
+    },
+    {
+      "epoch": 0.2945919985703755,
+      "grad_norm": 0.3762441575527191,
+      "learning_rate": 8.273707106847202e-05,
+      "loss": 1.0131,
+      "step": 3297
+    },
+    {
+      "epoch": 0.2946813501016374,
+      "grad_norm": 0.4361181855201721,
+      "learning_rate": 8.272613196358124e-05,
+      "loss": 1.0383,
+      "step": 3298
+    },
+    {
+      "epoch": 0.29477070163289926,
+      "grad_norm": 0.3914470076560974,
+      "learning_rate": 8.271519011752807e-05,
+      "loss": 0.9641,
+      "step": 3299
+    },
+    {
+      "epoch": 0.2948600531641611,
+      "grad_norm": 0.4699605703353882,
+      "learning_rate": 8.270424553122899e-05,
+      "loss": 0.9384,
+      "step": 3300
+    },
+    {
+      "epoch": 0.29494940469542297,
+      "grad_norm": 0.4258545935153961,
+      "learning_rate": 8.269329820560075e-05,
+      "loss": 1.0225,
+      "step": 3301
+    },
+    {
+      "epoch": 0.29503875622668485,
+      "grad_norm": 0.4467248320579529,
+      "learning_rate": 8.26823481415603e-05,
+      "loss": 0.9794,
+      "step": 3302
+    },
+    {
+      "epoch": 0.2951281077579467,
+      "grad_norm": 0.49532341957092285,
+      "learning_rate": 8.267139534002483e-05,
+      "loss": 0.9664,
+      "step": 3303
+    },
+    {
+      "epoch": 0.29521745928920856,
+      "grad_norm": 0.5230326056480408,
+      "learning_rate": 8.266043980191175e-05,
+      "loss": 1.001,
+      "step": 3304
+    },
+    {
+      "epoch": 0.29530681082047044,
+      "grad_norm": 0.46619758009910583,
+      "learning_rate": 8.26494815281387e-05,
+      "loss": 0.8984,
+      "step": 3305
+    },
+    {
+      "epoch": 0.2953961623517323,
+      "grad_norm": 0.514583170413971,
+      "learning_rate": 8.263852051962356e-05,
+      "loss": 0.9236,
+      "step": 3306
+    },
+    {
+      "epoch": 0.29548551388299416,
+      "grad_norm": 0.48377975821495056,
+      "learning_rate": 8.26275567772844e-05,
+      "loss": 1.0234,
+      "step": 3307
+    },
+    {
+      "epoch": 0.29557486541425604,
+      "grad_norm": 0.3904837667942047,
+      "learning_rate": 8.261659030203961e-05,
+      "loss": 1.072,
+      "step": 3308
+    },
+    {
+      "epoch": 0.2956642169455179,
+      "grad_norm": 0.3960415720939636,
+      "learning_rate": 8.260562109480768e-05,
+      "loss": 0.9875,
+      "step": 3309
+    },
+    {
+      "epoch": 0.29575356847677975,
+      "grad_norm": 0.4395782947540283,
+      "learning_rate": 8.259464915650745e-05,
+      "loss": 0.9658,
+      "step": 3310
+    },
+    {
+      "epoch": 0.29584292000804163,
+      "grad_norm": 0.548772394657135,
+      "learning_rate": 8.258367448805791e-05,
+      "loss": 0.9851,
+      "step": 3311
+    },
+    {
+      "epoch": 0.2959322715393035,
+      "grad_norm": 0.38434407114982605,
+      "learning_rate": 8.257269709037832e-05,
+      "loss": 1.0152,
+      "step": 3312
+    },
+    {
+      "epoch": 0.2960216230705654,
+      "grad_norm": 0.44232288002967834,
+      "learning_rate": 8.256171696438817e-05,
+      "loss": 0.9907,
+      "step": 3313
+    },
+    {
+      "epoch": 0.2961109746018272,
+      "grad_norm": 0.4280986487865448,
+      "learning_rate": 8.25507341110071e-05,
+      "loss": 0.9695,
+      "step": 3314
+    },
+    {
+      "epoch": 0.2962003261330891,
+      "grad_norm": 0.5346389412879944,
+      "learning_rate": 8.253974853115511e-05,
+      "loss": 0.9704,
+      "step": 3315
+    },
+    {
+      "epoch": 0.296289677664351,
+      "grad_norm": 0.43099406361579895,
+      "learning_rate": 8.252876022575234e-05,
+      "loss": 1.0432,
+      "step": 3316
+    },
+    {
+      "epoch": 0.2963790291956128,
+      "grad_norm": 0.412399560213089,
+      "learning_rate": 8.251776919571915e-05,
+      "loss": 1.0178,
+      "step": 3317
+    },
+    {
+      "epoch": 0.2964683807268747,
+      "grad_norm": 0.40774041414260864,
+      "learning_rate": 8.250677544197619e-05,
+      "loss": 1.0056,
+      "step": 3318
+    },
+    {
+      "epoch": 0.2965577322581366,
+      "grad_norm": 0.4034601151943207,
+      "learning_rate": 8.249577896544427e-05,
+      "loss": 1.0198,
+      "step": 3319
+    },
+    {
+      "epoch": 0.2966470837893984,
+      "grad_norm": 0.46452754735946655,
+      "learning_rate": 8.248477976704449e-05,
+      "loss": 1.0138,
+      "step": 3320
+    },
+    {
+      "epoch": 0.2967364353206603,
+      "grad_norm": 0.512606143951416,
+      "learning_rate": 8.247377784769814e-05,
+      "loss": 1.0394,
+      "step": 3321
+    },
+    {
+      "epoch": 0.2968257868519222,
+      "grad_norm": 0.557276725769043,
+      "learning_rate": 8.246277320832676e-05,
+      "loss": 1.0412,
+      "step": 3322
+    },
+    {
+      "epoch": 0.29691513838318406,
+      "grad_norm": 0.4517113268375397,
+      "learning_rate": 8.245176584985208e-05,
+      "loss": 0.9373,
+      "step": 3323
+    },
+    {
+      "epoch": 0.2970044899144459,
+      "grad_norm": 0.3471103608608246,
+      "learning_rate": 8.24407557731961e-05,
+      "loss": 1.0308,
+      "step": 3324
+    },
+    {
+      "epoch": 0.2970938414457078,
+      "grad_norm": 0.48082980513572693,
+      "learning_rate": 8.242974297928105e-05,
+      "loss": 0.9768,
+      "step": 3325
+    },
+    {
+      "epoch": 0.29718319297696966,
+      "grad_norm": 0.47311216592788696,
+      "learning_rate": 8.241872746902935e-05,
+      "loss": 0.9338,
+      "step": 3326
+    },
+    {
+      "epoch": 0.2972725445082315,
+      "grad_norm": 0.5065577626228333,
+      "learning_rate": 8.240770924336364e-05,
+      "loss": 0.9229,
+      "step": 3327
+    },
+    {
+      "epoch": 0.29736189603949337,
+      "grad_norm": 0.45983678102493286,
+      "learning_rate": 8.239668830320686e-05,
+      "loss": 1.0353,
+      "step": 3328
+    },
+    {
+      "epoch": 0.29745124757075525,
+      "grad_norm": 0.5233861804008484,
+      "learning_rate": 8.23856646494821e-05,
+      "loss": 0.9412,
+      "step": 3329
+    },
+    {
+      "epoch": 0.29754059910201713,
+      "grad_norm": 0.4303194284439087,
+      "learning_rate": 8.237463828311272e-05,
+      "loss": 1.0683,
+      "step": 3330
+    },
+    {
+      "epoch": 0.29762995063327896,
+      "grad_norm": 0.4418349862098694,
+      "learning_rate": 8.23636092050223e-05,
+      "loss": 0.9342,
+      "step": 3331
+    },
+    {
+      "epoch": 0.29771930216454084,
+      "grad_norm": 0.41626089811325073,
+      "learning_rate": 8.235257741613463e-05,
+      "loss": 0.9939,
+      "step": 3332
+    },
+    {
+      "epoch": 0.2978086536958027,
+      "grad_norm": 0.41136887669563293,
+      "learning_rate": 8.234154291737375e-05,
+      "loss": 1.0507,
+      "step": 3333
+    },
+    {
+      "epoch": 0.29789800522706456,
+      "grad_norm": 0.4649127125740051,
+      "learning_rate": 8.233050570966392e-05,
+      "loss": 1.0003,
+      "step": 3334
+    },
+    {
+      "epoch": 0.29798735675832644,
+      "grad_norm": 0.43239861726760864,
+      "learning_rate": 8.231946579392961e-05,
+      "loss": 0.9848,
+      "step": 3335
+    },
+    {
+      "epoch": 0.2980767082895883,
+      "grad_norm": 0.395386278629303,
+      "learning_rate": 8.230842317109555e-05,
+      "loss": 1.0269,
+      "step": 3336
+    },
+    {
+      "epoch": 0.2981660598208502,
+      "grad_norm": 0.5006843209266663,
+      "learning_rate": 8.229737784208667e-05,
+      "loss": 1.0536,
+      "step": 3337
+    },
+    {
+      "epoch": 0.29825541135211203,
+      "grad_norm": 0.39409273862838745,
+      "learning_rate": 8.22863298078281e-05,
+      "loss": 0.9827,
+      "step": 3338
+    },
+    {
+      "epoch": 0.2983447628833739,
+      "grad_norm": 0.4018579125404358,
+      "learning_rate": 8.22752790692453e-05,
+      "loss": 1.0009,
+      "step": 3339
+    },
+    {
+      "epoch": 0.2984341144146358,
+      "grad_norm": 0.4127073585987091,
+      "learning_rate": 8.226422562726383e-05,
+      "loss": 1.0012,
+      "step": 3340
+    },
+    {
+      "epoch": 0.2985234659458976,
+      "grad_norm": 0.4360758662223816,
+      "learning_rate": 8.225316948280955e-05,
+      "loss": 1.0098,
+      "step": 3341
+    },
+    {
+      "epoch": 0.2986128174771595,
+      "grad_norm": 0.4666980803012848,
+      "learning_rate": 8.224211063680853e-05,
+      "loss": 1.0034,
+      "step": 3342
+    },
+    {
+      "epoch": 0.2987021690084214,
+      "grad_norm": 0.4008931517601013,
+      "learning_rate": 8.223104909018707e-05,
+      "loss": 1.0339,
+      "step": 3343
+    },
+    {
+      "epoch": 0.2987915205396833,
+      "grad_norm": 0.38966014981269836,
+      "learning_rate": 8.22199848438717e-05,
+      "loss": 1.0023,
+      "step": 3344
+    },
+    {
+      "epoch": 0.2988808720709451,
+      "grad_norm": 0.48258697986602783,
+      "learning_rate": 8.220891789878915e-05,
+      "loss": 0.9615,
+      "step": 3345
+    },
+    {
+      "epoch": 0.298970223602207,
+      "grad_norm": 0.4819931387901306,
+      "learning_rate": 8.21978482558664e-05,
+      "loss": 0.9779,
+      "step": 3346
+    },
+    {
+      "epoch": 0.29905957513346887,
+      "grad_norm": 0.48947110772132874,
+      "learning_rate": 8.218677591603066e-05,
+      "loss": 0.9124,
+      "step": 3347
+    },
+    {
+      "epoch": 0.2991489266647307,
+      "grad_norm": 0.39361852407455444,
+      "learning_rate": 8.217570088020936e-05,
+      "loss": 1.079,
+      "step": 3348
+    },
+    {
+      "epoch": 0.2992382781959926,
+      "grad_norm": 0.5385176539421082,
+      "learning_rate": 8.216462314933013e-05,
+      "loss": 0.9939,
+      "step": 3349
+    },
+    {
+      "epoch": 0.29932762972725446,
+      "grad_norm": 0.4513779878616333,
+      "learning_rate": 8.215354272432086e-05,
+      "loss": 0.9908,
+      "step": 3350
+    },
+    {
+      "epoch": 0.2994169812585163,
+      "grad_norm": 0.4059743285179138,
+      "learning_rate": 8.214245960610967e-05,
+      "loss": 1.02,
+      "step": 3351
+    },
+    {
+      "epoch": 0.2995063327897782,
+      "grad_norm": 0.43754130601882935,
+      "learning_rate": 8.213137379562485e-05,
+      "loss": 0.9945,
+      "step": 3352
+    },
+    {
+      "epoch": 0.29959568432104006,
+      "grad_norm": 0.414631724357605,
+      "learning_rate": 8.2120285293795e-05,
+      "loss": 1.0089,
+      "step": 3353
+    },
+    {
+      "epoch": 0.29968503585230194,
+      "grad_norm": 0.518393337726593,
+      "learning_rate": 8.210919410154888e-05,
+      "loss": 1.0141,
+      "step": 3354
+    },
+    {
+      "epoch": 0.29977438738356377,
+      "grad_norm": 0.3928556442260742,
+      "learning_rate": 8.209810021981548e-05,
+      "loss": 1.0099,
+      "step": 3355
+    },
+    {
+      "epoch": 0.29986373891482565,
+      "grad_norm": 0.4404885470867157,
+      "learning_rate": 8.208700364952404e-05,
+      "loss": 1.0929,
+      "step": 3356
+    },
+    {
+      "epoch": 0.29995309044608753,
+      "grad_norm": 0.5233496427536011,
+      "learning_rate": 8.207590439160404e-05,
+      "loss": 0.9228,
+      "step": 3357
+    },
+    {
+      "epoch": 0.30004244197734936,
+      "grad_norm": 0.35965949296951294,
+      "learning_rate": 8.206480244698514e-05,
+      "loss": 1.0296,
+      "step": 3358
+    },
+    {
+      "epoch": 0.30013179350861124,
+      "grad_norm": 0.45257434248924255,
+      "learning_rate": 8.205369781659724e-05,
+      "loss": 0.9396,
+      "step": 3359
+    },
+    {
+      "epoch": 0.3002211450398731,
+      "grad_norm": 0.4101848304271698,
+      "learning_rate": 8.204259050137048e-05,
+      "loss": 1.0811,
+      "step": 3360
+    },
+    {
+      "epoch": 0.300310496571135,
+      "grad_norm": 0.48168009519577026,
+      "learning_rate": 8.203148050223521e-05,
+      "loss": 0.9708,
+      "step": 3361
+    },
+    {
+      "epoch": 0.30039984810239684,
+      "grad_norm": 0.434770405292511,
+      "learning_rate": 8.202036782012203e-05,
+      "loss": 0.9487,
+      "step": 3362
+    },
+    {
+      "epoch": 0.3004891996336587,
+      "grad_norm": 0.3673984706401825,
+      "learning_rate": 8.200925245596174e-05,
+      "loss": 1.0221,
+      "step": 3363
+    },
+    {
+      "epoch": 0.3005785511649206,
+      "grad_norm": 0.4212033152580261,
+      "learning_rate": 8.199813441068535e-05,
+      "loss": 1.0128,
+      "step": 3364
+    },
+    {
+      "epoch": 0.30066790269618243,
+      "grad_norm": 0.42803311347961426,
+      "learning_rate": 8.198701368522412e-05,
+      "loss": 1.0614,
+      "step": 3365
+    },
+    {
+      "epoch": 0.3007572542274443,
+      "grad_norm": 0.48348864912986755,
+      "learning_rate": 8.197589028050956e-05,
+      "loss": 0.9918,
+      "step": 3366
+    },
+    {
+      "epoch": 0.3008466057587062,
+      "grad_norm": 0.455049067735672,
+      "learning_rate": 8.196476419747333e-05,
+      "loss": 1.0028,
+      "step": 3367
+    },
+    {
+      "epoch": 0.3009359572899681,
+      "grad_norm": 0.397243857383728,
+      "learning_rate": 8.195363543704739e-05,
+      "loss": 0.9751,
+      "step": 3368
+    },
+    {
+      "epoch": 0.3010253088212299,
+      "grad_norm": 0.3902703523635864,
+      "learning_rate": 8.194250400016388e-05,
+      "loss": 1.0078,
+      "step": 3369
+    },
+    {
+      "epoch": 0.3011146603524918,
+      "grad_norm": 0.41560593247413635,
+      "learning_rate": 8.193136988775516e-05,
+      "loss": 0.9526,
+      "step": 3370
+    },
+    {
+      "epoch": 0.3012040118837537,
+      "grad_norm": 0.44474300742149353,
+      "learning_rate": 8.192023310075387e-05,
+      "loss": 0.8866,
+      "step": 3371
+    },
+    {
+      "epoch": 0.3012933634150155,
+      "grad_norm": 0.34557124972343445,
+      "learning_rate": 8.190909364009281e-05,
+      "loss": 1.0232,
+      "step": 3372
+    },
+    {
+      "epoch": 0.3013827149462774,
+      "grad_norm": 0.4132451117038727,
+      "learning_rate": 8.1897951506705e-05,
+      "loss": 1.0186,
+      "step": 3373
+    },
+    {
+      "epoch": 0.30147206647753927,
+      "grad_norm": 0.4821149706840515,
+      "learning_rate": 8.188680670152379e-05,
+      "loss": 0.9414,
+      "step": 3374
+    },
+    {
+      "epoch": 0.30156141800880115,
+      "grad_norm": 0.43094414472579956,
+      "learning_rate": 8.18756592254826e-05,
+      "loss": 0.9497,
+      "step": 3375
+    },
+    {
+      "epoch": 0.301650769540063,
+      "grad_norm": 0.3759251832962036,
+      "learning_rate": 8.18645090795152e-05,
+      "loss": 1.0405,
+      "step": 3376
+    },
+    {
+      "epoch": 0.30174012107132486,
+      "grad_norm": 0.5242640376091003,
+      "learning_rate": 8.185335626455549e-05,
+      "loss": 0.9464,
+      "step": 3377
+    },
+    {
+      "epoch": 0.30182947260258675,
+      "grad_norm": 0.4868447780609131,
+      "learning_rate": 8.184220078153768e-05,
+      "loss": 0.9646,
+      "step": 3378
+    },
+    {
+      "epoch": 0.3019188241338486,
+      "grad_norm": 0.441510945558548,
+      "learning_rate": 8.183104263139613e-05,
+      "loss": 0.9802,
+      "step": 3379
+    },
+    {
+      "epoch": 0.30200817566511046,
+      "grad_norm": 0.39666980504989624,
+      "learning_rate": 8.181988181506546e-05,
+      "loss": 1.0346,
+      "step": 3380
+    },
+    {
+      "epoch": 0.30209752719637234,
+      "grad_norm": 0.3387872874736786,
+      "learning_rate": 8.180871833348052e-05,
+      "loss": 0.9721,
+      "step": 3381
+    },
+    {
+      "epoch": 0.30218687872763417,
+      "grad_norm": 0.4515991508960724,
+      "learning_rate": 8.179755218757636e-05,
+      "loss": 0.9902,
+      "step": 3382
+    },
+    {
+      "epoch": 0.30227623025889605,
+      "grad_norm": 0.4831897020339966,
+      "learning_rate": 8.178638337828828e-05,
+      "loss": 0.9039,
+      "step": 3383
+    },
+    {
+      "epoch": 0.30236558179015793,
+      "grad_norm": 0.44388824701309204,
+      "learning_rate": 8.177521190655178e-05,
+      "loss": 0.9844,
+      "step": 3384
+    },
+    {
+      "epoch": 0.3024549333214198,
+      "grad_norm": 0.495235800743103,
+      "learning_rate": 8.176403777330255e-05,
+      "loss": 0.9928,
+      "step": 3385
+    },
+    {
+      "epoch": 0.30254428485268164,
+      "grad_norm": 0.5274266600608826,
+      "learning_rate": 8.17528609794766e-05,
+      "loss": 1.0191,
+      "step": 3386
+    },
+    {
+      "epoch": 0.3026336363839435,
+      "grad_norm": 0.3730257749557495,
+      "learning_rate": 8.174168152601007e-05,
+      "loss": 0.9684,
+      "step": 3387
+    },
+    {
+      "epoch": 0.3027229879152054,
+      "grad_norm": 0.45568758249282837,
+      "learning_rate": 8.173049941383938e-05,
+      "loss": 0.9999,
+      "step": 3388
+    },
+    {
+      "epoch": 0.30281233944646724,
+      "grad_norm": 0.4109109342098236,
+      "learning_rate": 8.171931464390115e-05,
+      "loss": 1.0108,
+      "step": 3389
+    },
+    {
+      "epoch": 0.3029016909777291,
+      "grad_norm": 0.4820617735385895,
+      "learning_rate": 8.170812721713218e-05,
+      "loss": 1.0419,
+      "step": 3390
+    },
+    {
+      "epoch": 0.302991042508991,
+      "grad_norm": 0.4437454044818878,
+      "learning_rate": 8.169693713446959e-05,
+      "loss": 0.9256,
+      "step": 3391
+    },
+    {
+      "epoch": 0.3030803940402529,
+      "grad_norm": 0.4973864257335663,
+      "learning_rate": 8.168574439685067e-05,
+      "loss": 0.9822,
+      "step": 3392
+    },
+    {
+      "epoch": 0.3031697455715147,
+      "grad_norm": 0.41152751445770264,
+      "learning_rate": 8.167454900521289e-05,
+      "loss": 1.02,
+      "step": 3393
+    },
+    {
+      "epoch": 0.3032590971027766,
+      "grad_norm": 0.42877259850502014,
+      "learning_rate": 8.1663350960494e-05,
+      "loss": 1.0445,
+      "step": 3394
+    },
+    {
+      "epoch": 0.3033484486340385,
+      "grad_norm": 0.5001733899116516,
+      "learning_rate": 8.165215026363198e-05,
+      "loss": 0.996,
+      "step": 3395
+    },
+    {
+      "epoch": 0.3034378001653003,
+      "grad_norm": 0.43617990612983704,
+      "learning_rate": 8.164094691556496e-05,
+      "loss": 0.9811,
+      "step": 3396
+    },
+    {
+      "epoch": 0.3035271516965622,
+      "grad_norm": 0.4765497148036957,
+      "learning_rate": 8.162974091723139e-05,
+      "loss": 0.976,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3036165032278241,
+      "grad_norm": 0.45245134830474854,
+      "learning_rate": 8.161853226956985e-05,
+      "loss": 1.0029,
+      "step": 3398
+    },
+    {
+      "epoch": 0.30370585475908596,
+      "grad_norm": 0.4027976989746094,
+      "learning_rate": 8.160732097351924e-05,
+      "loss": 1.0238,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3037952062903478,
+      "grad_norm": 0.4003666341304779,
+      "learning_rate": 8.159610703001857e-05,
+      "loss": 1.0662,
+      "step": 3400
+    },
+    {
+      "epoch": 0.30388455782160967,
+      "grad_norm": 0.4242064952850342,
+      "learning_rate": 8.158489044000713e-05,
+      "loss": 1.0077,
+      "step": 3401
+    },
+    {
+      "epoch": 0.30397390935287155,
+      "grad_norm": 0.3776360750198364,
+      "learning_rate": 8.157367120442445e-05,
+      "loss": 1.0848,
+      "step": 3402
+    },
+    {
+      "epoch": 0.3040632608841334,
+      "grad_norm": 0.40885689854621887,
+      "learning_rate": 8.156244932421026e-05,
+      "loss": 0.9957,
+      "step": 3403
+    },
+    {
+      "epoch": 0.30415261241539526,
+      "grad_norm": 0.44386911392211914,
+      "learning_rate": 8.155122480030453e-05,
+      "loss": 1.0783,
+      "step": 3404
+    },
+    {
+      "epoch": 0.30424196394665715,
+      "grad_norm": 0.38162335753440857,
+      "learning_rate": 8.153999763364738e-05,
+      "loss": 1.0232,
+      "step": 3405
+    },
+    {
+      "epoch": 0.30433131547791903,
+      "grad_norm": 0.3886288106441498,
+      "learning_rate": 8.152876782517925e-05,
+      "loss": 1.0595,
+      "step": 3406
+    },
+    {
+      "epoch": 0.30442066700918086,
+      "grad_norm": 0.43489503860473633,
+      "learning_rate": 8.151753537584073e-05,
+      "loss": 1.0397,
+      "step": 3407
+    },
+    {
+      "epoch": 0.30451001854044274,
+      "grad_norm": 0.4794921278953552,
+      "learning_rate": 8.150630028657267e-05,
+      "loss": 0.9464,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3045993700717046,
+      "grad_norm": 0.5062465071678162,
+      "learning_rate": 8.149506255831614e-05,
+      "loss": 0.9363,
+      "step": 3409
+    },
+    {
+      "epoch": 0.30468872160296645,
+      "grad_norm": 0.4547816514968872,
+      "learning_rate": 8.148382219201241e-05,
+      "loss": 0.981,
+      "step": 3410
+    },
+    {
+      "epoch": 0.30477807313422833,
+      "grad_norm": 0.43893638253211975,
+      "learning_rate": 8.147257918860296e-05,
+      "loss": 1.0097,
+      "step": 3411
+    },
+    {
+      "epoch": 0.3048674246654902,
+      "grad_norm": 0.4645942151546478,
+      "learning_rate": 8.146133354902954e-05,
+      "loss": 1.0032,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3049567761967521,
+      "grad_norm": 0.39374154806137085,
+      "learning_rate": 8.145008527423409e-05,
+      "loss": 1.0049,
+      "step": 3413
+    },
+    {
+      "epoch": 0.3050461277280139,
+      "grad_norm": 0.4022867977619171,
+      "learning_rate": 8.143883436515875e-05,
+      "loss": 0.9876,
+      "step": 3414
+    },
+    {
+      "epoch": 0.3051354792592758,
+      "grad_norm": 0.4411511719226837,
+      "learning_rate": 8.142758082274593e-05,
+      "loss": 0.9858,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3052248307905377,
+      "grad_norm": 0.4204576909542084,
+      "learning_rate": 8.141632464793822e-05,
+      "loss": 0.9826,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3053141823217995,
+      "grad_norm": 0.40963998436927795,
+      "learning_rate": 8.140506584167845e-05,
+      "loss": 0.9313,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3054035338530614,
+      "grad_norm": 0.40934035181999207,
+      "learning_rate": 8.139380440490965e-05,
+      "loss": 1.037,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3054928853843233,
+      "grad_norm": 0.3806384801864624,
+      "learning_rate": 8.13825403385751e-05,
+      "loss": 0.9968,
+      "step": 3419
+    },
+    {
+      "epoch": 0.3055822369155851,
+      "grad_norm": 0.3986847400665283,
+      "learning_rate": 8.137127364361828e-05,
+      "loss": 0.9992,
+      "step": 3420
+    },
+    {
+      "epoch": 0.305671588446847,
+      "grad_norm": 0.44412630796432495,
+      "learning_rate": 8.136000432098291e-05,
+      "loss": 1.0355,
+      "step": 3421
+    },
+    {
+      "epoch": 0.3057609399781089,
+      "grad_norm": 0.43642017245292664,
+      "learning_rate": 8.13487323716129e-05,
+      "loss": 1.0421,
+      "step": 3422
+    },
+    {
+      "epoch": 0.30585029150937076,
+      "grad_norm": 0.39217737317085266,
+      "learning_rate": 8.133745779645238e-05,
+      "loss": 1.0206,
+      "step": 3423
+    },
+    {
+      "epoch": 0.3059396430406326,
+      "grad_norm": 0.456626832485199,
+      "learning_rate": 8.132618059644577e-05,
+      "loss": 1.0341,
+      "step": 3424
+    },
+    {
+      "epoch": 0.3060289945718945,
+      "grad_norm": 0.5178848505020142,
+      "learning_rate": 8.13149007725376e-05,
+      "loss": 0.9424,
+      "step": 3425
+    },
+    {
+      "epoch": 0.30611834610315636,
+      "grad_norm": 0.49753326177597046,
+      "learning_rate": 8.13036183256727e-05,
+      "loss": 1.0006,
+      "step": 3426
+    },
+    {
+      "epoch": 0.3062076976344182,
+      "grad_norm": 0.48190826177597046,
+      "learning_rate": 8.129233325679611e-05,
+      "loss": 0.9446,
+      "step": 3427
+    },
+    {
+      "epoch": 0.30629704916568007,
+      "grad_norm": 0.431306928396225,
+      "learning_rate": 8.128104556685305e-05,
+      "loss": 0.9653,
+      "step": 3428
+    },
+    {
+      "epoch": 0.30638640069694195,
+      "grad_norm": 0.44903165102005005,
+      "learning_rate": 8.126975525678898e-05,
+      "loss": 1.0018,
+      "step": 3429
+    },
+    {
+      "epoch": 0.30647575222820383,
+      "grad_norm": 0.4540453553199768,
+      "learning_rate": 8.12584623275496e-05,
+      "loss": 0.9977,
+      "step": 3430
+    },
+    {
+      "epoch": 0.30656510375946566,
+      "grad_norm": 0.40107932686805725,
+      "learning_rate": 8.124716678008082e-05,
+      "loss": 1.058,
+      "step": 3431
+    },
+    {
+      "epoch": 0.30665445529072755,
+      "grad_norm": 0.3945641815662384,
+      "learning_rate": 8.123586861532873e-05,
+      "loss": 1.0387,
+      "step": 3432
+    },
+    {
+      "epoch": 0.30674380682198943,
+      "grad_norm": 0.45806485414505005,
+      "learning_rate": 8.12245678342397e-05,
+      "loss": 1.0011,
+      "step": 3433
+    },
+    {
+      "epoch": 0.30683315835325126,
+      "grad_norm": 0.49401962757110596,
+      "learning_rate": 8.12132644377603e-05,
+      "loss": 1.005,
+      "step": 3434
+    },
+    {
+      "epoch": 0.30692250988451314,
+      "grad_norm": 0.44378378987312317,
+      "learning_rate": 8.120195842683728e-05,
+      "loss": 0.9761,
+      "step": 3435
+    },
+    {
+      "epoch": 0.307011861415775,
+      "grad_norm": 0.4618094563484192,
+      "learning_rate": 8.119064980241766e-05,
+      "loss": 0.9972,
+      "step": 3436
+    },
+    {
+      "epoch": 0.3071012129470369,
+      "grad_norm": 0.4815685749053955,
+      "learning_rate": 8.117933856544864e-05,
+      "loss": 0.9633,
+      "step": 3437
+    },
+    {
+      "epoch": 0.30719056447829873,
+      "grad_norm": 0.43568116426467896,
+      "learning_rate": 8.116802471687768e-05,
+      "loss": 0.9158,
+      "step": 3438
+    },
+    {
+      "epoch": 0.3072799160095606,
+      "grad_norm": 0.4017315208911896,
+      "learning_rate": 8.11567082576524e-05,
+      "loss": 1.0004,
+      "step": 3439
+    },
+    {
+      "epoch": 0.3073692675408225,
+      "grad_norm": 0.43449345231056213,
+      "learning_rate": 8.114538918872072e-05,
+      "loss": 1.0021,
+      "step": 3440
+    },
+    {
+      "epoch": 0.3074586190720843,
+      "grad_norm": 0.37431442737579346,
+      "learning_rate": 8.11340675110307e-05,
+      "loss": 1.0407,
+      "step": 3441
+    },
+    {
+      "epoch": 0.3075479706033462,
+      "grad_norm": 0.42223483324050903,
+      "learning_rate": 8.112274322553067e-05,
+      "loss": 0.9816,
+      "step": 3442
+    },
+    {
+      "epoch": 0.3076373221346081,
+      "grad_norm": 0.5260857939720154,
+      "learning_rate": 8.111141633316914e-05,
+      "loss": 0.9194,
+      "step": 3443
+    },
+    {
+      "epoch": 0.30772667366587,
+      "grad_norm": 0.4374569356441498,
+      "learning_rate": 8.110008683489487e-05,
+      "loss": 1.0595,
+      "step": 3444
+    },
+    {
+      "epoch": 0.3078160251971318,
+      "grad_norm": 0.45745593309402466,
+      "learning_rate": 8.108875473165683e-05,
+      "loss": 0.988,
+      "step": 3445
+    },
+    {
+      "epoch": 0.3079053767283937,
+      "grad_norm": 0.42707595229148865,
+      "learning_rate": 8.10774200244042e-05,
+      "loss": 1.0022,
+      "step": 3446
+    },
+    {
+      "epoch": 0.30799472825965557,
+      "grad_norm": 0.4703771471977234,
+      "learning_rate": 8.106608271408638e-05,
+      "loss": 1.0088,
+      "step": 3447
+    },
+    {
+      "epoch": 0.3080840797909174,
+      "grad_norm": 0.5333291292190552,
+      "learning_rate": 8.105474280165298e-05,
+      "loss": 0.994,
+      "step": 3448
+    },
+    {
+      "epoch": 0.3081734313221793,
+      "grad_norm": 0.3782300055027008,
+      "learning_rate": 8.104340028805386e-05,
+      "loss": 1.0284,
+      "step": 3449
+    },
+    {
+      "epoch": 0.30826278285344116,
+      "grad_norm": 0.5105333924293518,
+      "learning_rate": 8.103205517423906e-05,
+      "loss": 0.9954,
+      "step": 3450
+    },
+    {
+      "epoch": 0.308352134384703,
+      "grad_norm": 0.37891459465026855,
+      "learning_rate": 8.102070746115888e-05,
+      "loss": 0.9678,
+      "step": 3451
+    },
+    {
+      "epoch": 0.3084414859159649,
+      "grad_norm": 0.47059276700019836,
+      "learning_rate": 8.100935714976377e-05,
+      "loss": 0.9004,
+      "step": 3452
+    },
+    {
+      "epoch": 0.30853083744722676,
+      "grad_norm": 0.38149774074554443,
+      "learning_rate": 8.099800424100448e-05,
+      "loss": 0.9746,
+      "step": 3453
+    },
+    {
+      "epoch": 0.30862018897848864,
+      "grad_norm": 0.4379188120365143,
+      "learning_rate": 8.09866487358319e-05,
+      "loss": 1.02,
+      "step": 3454
+    },
+    {
+      "epoch": 0.30870954050975047,
+      "grad_norm": 0.4322841465473175,
+      "learning_rate": 8.097529063519723e-05,
+      "loss": 0.9619,
+      "step": 3455
+    },
+    {
+      "epoch": 0.30879889204101235,
+      "grad_norm": 0.4443778097629547,
+      "learning_rate": 8.096392994005177e-05,
+      "loss": 0.9386,
+      "step": 3456
+    },
+    {
+      "epoch": 0.30888824357227423,
+      "grad_norm": 0.4804965853691101,
+      "learning_rate": 8.095256665134712e-05,
+      "loss": 0.9817,
+      "step": 3457
+    },
+    {
+      "epoch": 0.30897759510353606,
+      "grad_norm": 0.5008002519607544,
+      "learning_rate": 8.094120077003509e-05,
+      "loss": 0.9661,
+      "step": 3458
+    },
+    {
+      "epoch": 0.30906694663479795,
+      "grad_norm": 0.4188991189002991,
+      "learning_rate": 8.092983229706767e-05,
+      "loss": 0.9905,
+      "step": 3459
+    },
+    {
+      "epoch": 0.30915629816605983,
+      "grad_norm": 0.3990739583969116,
+      "learning_rate": 8.091846123339715e-05,
+      "loss": 1.0172,
+      "step": 3460
+    },
+    {
+      "epoch": 0.3092456496973217,
+      "grad_norm": 0.43267446756362915,
+      "learning_rate": 8.090708757997591e-05,
+      "loss": 1.0371,
+      "step": 3461
+    },
+    {
+      "epoch": 0.30933500122858354,
+      "grad_norm": 0.42393234372138977,
+      "learning_rate": 8.089571133775663e-05,
+      "loss": 1.0194,
+      "step": 3462
+    },
+    {
+      "epoch": 0.3094243527598454,
+      "grad_norm": 0.4767492711544037,
+      "learning_rate": 8.08843325076922e-05,
+      "loss": 1.0495,
+      "step": 3463
+    },
+    {
+      "epoch": 0.3095137042911073,
+      "grad_norm": 0.42636582255363464,
+      "learning_rate": 8.087295109073574e-05,
+      "loss": 1.0434,
+      "step": 3464
+    },
+    {
+      "epoch": 0.30960305582236913,
+      "grad_norm": 0.49471211433410645,
+      "learning_rate": 8.086156708784054e-05,
+      "loss": 0.9261,
+      "step": 3465
+    },
+    {
+      "epoch": 0.309692407353631,
+      "grad_norm": 0.4437471628189087,
+      "learning_rate": 8.085018049996011e-05,
+      "loss": 1.0086,
+      "step": 3466
+    },
+    {
+      "epoch": 0.3097817588848929,
+      "grad_norm": 0.44580212235450745,
+      "learning_rate": 8.083879132804826e-05,
+      "loss": 0.9378,
+      "step": 3467
+    },
+    {
+      "epoch": 0.3098711104161548,
+      "grad_norm": 0.4966757595539093,
+      "learning_rate": 8.082739957305889e-05,
+      "loss": 0.9868,
+      "step": 3468
+    },
+    {
+      "epoch": 0.3099604619474166,
+      "grad_norm": 0.5008419752120972,
+      "learning_rate": 8.081600523594622e-05,
+      "loss": 0.9713,
+      "step": 3469
+    },
+    {
+      "epoch": 0.3100498134786785,
+      "grad_norm": 0.44220054149627686,
+      "learning_rate": 8.080460831766464e-05,
+      "loss": 0.9581,
+      "step": 3470
+    },
+    {
+      "epoch": 0.3101391650099404,
+      "grad_norm": 0.47498950362205505,
+      "learning_rate": 8.079320881916877e-05,
+      "loss": 1.0412,
+      "step": 3471
+    },
+    {
+      "epoch": 0.3102285165412022,
+      "grad_norm": 0.48593392968177795,
+      "learning_rate": 8.07818067414134e-05,
+      "loss": 0.9469,
+      "step": 3472
+    },
+    {
+      "epoch": 0.3103178680724641,
+      "grad_norm": 0.3928394019603729,
+      "learning_rate": 8.077040208535362e-05,
+      "loss": 0.9524,
+      "step": 3473
+    },
+    {
+      "epoch": 0.31040721960372597,
+      "grad_norm": 0.38281822204589844,
+      "learning_rate": 8.075899485194467e-05,
+      "loss": 1.0437,
+      "step": 3474
+    },
+    {
+      "epoch": 0.31049657113498785,
+      "grad_norm": 0.408634752035141,
+      "learning_rate": 8.074758504214206e-05,
+      "loss": 0.9995,
+      "step": 3475
+    },
+    {
+      "epoch": 0.3105859226662497,
+      "grad_norm": 0.4135921001434326,
+      "learning_rate": 8.073617265690144e-05,
+      "loss": 1.0512,
+      "step": 3476
+    },
+    {
+      "epoch": 0.31067527419751156,
+      "grad_norm": 0.45093652606010437,
+      "learning_rate": 8.072475769717872e-05,
+      "loss": 0.9089,
+      "step": 3477
+    },
+    {
+      "epoch": 0.31076462572877345,
+      "grad_norm": 0.38925284147262573,
+      "learning_rate": 8.071334016393006e-05,
+      "loss": 0.958,
+      "step": 3478
+    },
+    {
+      "epoch": 0.3108539772600353,
+      "grad_norm": 0.4077800512313843,
+      "learning_rate": 8.070192005811177e-05,
+      "loss": 1.0573,
+      "step": 3479
+    },
+    {
+      "epoch": 0.31094332879129716,
+      "grad_norm": 0.48132577538490295,
+      "learning_rate": 8.069049738068041e-05,
+      "loss": 0.899,
+      "step": 3480
+    },
+    {
+      "epoch": 0.31103268032255904,
+      "grad_norm": 0.4954333007335663,
+      "learning_rate": 8.067907213259278e-05,
+      "loss": 0.9735,
+      "step": 3481
+    },
+    {
+      "epoch": 0.31112203185382087,
+      "grad_norm": 0.3865114152431488,
+      "learning_rate": 8.066764431480583e-05,
+      "loss": 0.9941,
+      "step": 3482
+    },
+    {
+      "epoch": 0.31121138338508275,
+      "grad_norm": 0.3530360460281372,
+      "learning_rate": 8.065621392827678e-05,
+      "loss": 1.0592,
+      "step": 3483
+    },
+    {
+      "epoch": 0.31130073491634463,
+      "grad_norm": 0.4171468913555145,
+      "learning_rate": 8.064478097396304e-05,
+      "loss": 0.9654,
+      "step": 3484
+    },
+    {
+      "epoch": 0.3113900864476065,
+      "grad_norm": 0.508972704410553,
+      "learning_rate": 8.063334545282224e-05,
+      "loss": 0.8936,
+      "step": 3485
+    },
+    {
+      "epoch": 0.31147943797886835,
+      "grad_norm": 0.3663281500339508,
+      "learning_rate": 8.062190736581223e-05,
+      "loss": 1.0223,
+      "step": 3486
+    },
+    {
+      "epoch": 0.31156878951013023,
+      "grad_norm": 0.4794585704803467,
+      "learning_rate": 8.061046671389107e-05,
+      "loss": 0.9199,
+      "step": 3487
+    },
+    {
+      "epoch": 0.3116581410413921,
+      "grad_norm": 0.4471113383769989,
+      "learning_rate": 8.059902349801704e-05,
+      "loss": 0.9633,
+      "step": 3488
+    },
+    {
+      "epoch": 0.31174749257265394,
+      "grad_norm": 0.3974318504333496,
+      "learning_rate": 8.058757771914865e-05,
+      "loss": 0.9769,
+      "step": 3489
+    },
+    {
+      "epoch": 0.3118368441039158,
+      "grad_norm": 0.4705379605293274,
+      "learning_rate": 8.057612937824456e-05,
+      "loss": 0.9843,
+      "step": 3490
+    },
+    {
+      "epoch": 0.3119261956351777,
+      "grad_norm": 0.3952665328979492,
+      "learning_rate": 8.056467847626373e-05,
+      "loss": 1.0787,
+      "step": 3491
+    },
+    {
+      "epoch": 0.3120155471664396,
+      "grad_norm": 0.451646089553833,
+      "learning_rate": 8.055322501416527e-05,
+      "loss": 1.0125,
+      "step": 3492
+    },
+    {
+      "epoch": 0.3121048986977014,
+      "grad_norm": 0.4705079197883606,
+      "learning_rate": 8.054176899290855e-05,
+      "loss": 0.9875,
+      "step": 3493
+    },
+    {
+      "epoch": 0.3121942502289633,
+      "grad_norm": 0.4389439523220062,
+      "learning_rate": 8.053031041345312e-05,
+      "loss": 1.0343,
+      "step": 3494
+    },
+    {
+      "epoch": 0.3122836017602252,
+      "grad_norm": 0.513843834400177,
+      "learning_rate": 8.051884927675877e-05,
+      "loss": 0.9618,
+      "step": 3495
+    },
+    {
+      "epoch": 0.312372953291487,
+      "grad_norm": 0.4346451163291931,
+      "learning_rate": 8.050738558378549e-05,
+      "loss": 1.0277,
+      "step": 3496
+    },
+    {
+      "epoch": 0.3124623048227489,
+      "grad_norm": 0.5713686347007751,
+      "learning_rate": 8.049591933549347e-05,
+      "loss": 0.9842,
+      "step": 3497
+    },
+    {
+      "epoch": 0.3125516563540108,
+      "grad_norm": 0.4521034359931946,
+      "learning_rate": 8.048445053284315e-05,
+      "loss": 0.9809,
+      "step": 3498
+    },
+    {
+      "epoch": 0.31264100788527266,
+      "grad_norm": 0.4172121584415436,
+      "learning_rate": 8.047297917679515e-05,
+      "loss": 1.0247,
+      "step": 3499
+    },
+    {
+      "epoch": 0.3127303594165345,
+      "grad_norm": 0.4868249297142029,
+      "learning_rate": 8.046150526831033e-05,
+      "loss": 0.9791,
+      "step": 3500
+    },
+    {
+      "epoch": 0.31281971094779637,
+      "grad_norm": 0.34857162833213806,
+      "learning_rate": 8.045002880834975e-05,
+      "loss": 1.0506,
+      "step": 3501
+    },
+    {
+      "epoch": 0.31290906247905825,
+      "grad_norm": 0.4393348693847656,
+      "learning_rate": 8.043854979787467e-05,
+      "loss": 1.0026,
+      "step": 3502
+    },
+    {
+      "epoch": 0.3129984140103201,
+      "grad_norm": 0.38047170639038086,
+      "learning_rate": 8.04270682378466e-05,
+      "loss": 1.0297,
+      "step": 3503
+    },
+    {
+      "epoch": 0.31308776554158196,
+      "grad_norm": 0.39295217394828796,
+      "learning_rate": 8.041558412922724e-05,
+      "loss": 0.997,
+      "step": 3504
+    },
+    {
+      "epoch": 0.31317711707284385,
+      "grad_norm": 0.4318108856678009,
+      "learning_rate": 8.04040974729785e-05,
+      "loss": 0.9394,
+      "step": 3505
+    },
+    {
+      "epoch": 0.31326646860410573,
+      "grad_norm": 0.41878437995910645,
+      "learning_rate": 8.039260827006252e-05,
+      "loss": 1.012,
+      "step": 3506
+    },
+    {
+      "epoch": 0.31335582013536756,
+      "grad_norm": 0.3880411982536316,
+      "learning_rate": 8.038111652144163e-05,
+      "loss": 0.962,
+      "step": 3507
+    },
+    {
+      "epoch": 0.31344517166662944,
+      "grad_norm": 0.39992988109588623,
+      "learning_rate": 8.036962222807838e-05,
+      "loss": 0.9838,
+      "step": 3508
+    },
+    {
+      "epoch": 0.3135345231978913,
+      "grad_norm": 0.3804689347743988,
+      "learning_rate": 8.035812539093557e-05,
+      "loss": 1.0372,
+      "step": 3509
+    },
+    {
+      "epoch": 0.31362387472915315,
+      "grad_norm": 0.4325745403766632,
+      "learning_rate": 8.034662601097615e-05,
+      "loss": 1.032,
+      "step": 3510
+    },
+    {
+      "epoch": 0.31371322626041503,
+      "grad_norm": 0.46068939566612244,
+      "learning_rate": 8.033512408916334e-05,
+      "loss": 0.9907,
+      "step": 3511
+    },
+    {
+      "epoch": 0.3138025777916769,
+      "grad_norm": 0.5449123382568359,
+      "learning_rate": 8.032361962646053e-05,
+      "loss": 0.9458,
+      "step": 3512
+    },
+    {
+      "epoch": 0.31389192932293875,
+      "grad_norm": 0.4366675913333893,
+      "learning_rate": 8.031211262383136e-05,
+      "loss": 1.0291,
+      "step": 3513
+    },
+    {
+      "epoch": 0.31398128085420063,
+      "grad_norm": 0.4750640392303467,
+      "learning_rate": 8.030060308223964e-05,
+      "loss": 0.9661,
+      "step": 3514
+    },
+    {
+      "epoch": 0.3140706323854625,
+      "grad_norm": 0.365343302488327,
+      "learning_rate": 8.028909100264943e-05,
+      "loss": 1.0437,
+      "step": 3515
+    },
+    {
+      "epoch": 0.3141599839167244,
+      "grad_norm": 0.580810546875,
+      "learning_rate": 8.0277576386025e-05,
+      "loss": 0.9558,
+      "step": 3516
+    },
+    {
+      "epoch": 0.3142493354479862,
+      "grad_norm": 0.5106691718101501,
+      "learning_rate": 8.026605923333081e-05,
+      "loss": 0.9663,
+      "step": 3517
+    },
+    {
+      "epoch": 0.3143386869792481,
+      "grad_norm": 0.48174557089805603,
+      "learning_rate": 8.025453954553155e-05,
+      "loss": 0.9992,
+      "step": 3518
+    },
+    {
+      "epoch": 0.31442803851051,
+      "grad_norm": 0.41393786668777466,
+      "learning_rate": 8.024301732359212e-05,
+      "loss": 1.0187,
+      "step": 3519
+    },
+    {
+      "epoch": 0.3145173900417718,
+      "grad_norm": 0.4682253897190094,
+      "learning_rate": 8.02314925684776e-05,
+      "loss": 0.9622,
+      "step": 3520
+    },
+    {
+      "epoch": 0.3146067415730337,
+      "grad_norm": 0.4221019446849823,
+      "learning_rate": 8.021996528115335e-05,
+      "loss": 1.0434,
+      "step": 3521
+    },
+    {
+      "epoch": 0.3146960931042956,
+      "grad_norm": 0.3874817490577698,
+      "learning_rate": 8.020843546258487e-05,
+      "loss": 0.9968,
+      "step": 3522
+    },
+    {
+      "epoch": 0.31478544463555747,
+      "grad_norm": 0.49449774622917175,
+      "learning_rate": 8.019690311373793e-05,
+      "loss": 0.9772,
+      "step": 3523
+    },
+    {
+      "epoch": 0.3148747961668193,
+      "grad_norm": 0.4564272463321686,
+      "learning_rate": 8.018536823557848e-05,
+      "loss": 0.9736,
+      "step": 3524
+    },
+    {
+      "epoch": 0.3149641476980812,
+      "grad_norm": 0.3975536525249481,
+      "learning_rate": 8.017383082907269e-05,
+      "loss": 0.988,
+      "step": 3525
+    },
+    {
+      "epoch": 0.31505349922934306,
+      "grad_norm": 0.42857542634010315,
+      "learning_rate": 8.016229089518694e-05,
+      "loss": 0.9846,
+      "step": 3526
+    },
+    {
+      "epoch": 0.3151428507606049,
+      "grad_norm": 0.34974923729896545,
+      "learning_rate": 8.01507484348878e-05,
+      "loss": 0.9828,
+      "step": 3527
+    },
+    {
+      "epoch": 0.31523220229186677,
+      "grad_norm": 0.39053013920783997,
+      "learning_rate": 8.013920344914212e-05,
+      "loss": 1.0341,
+      "step": 3528
+    },
+    {
+      "epoch": 0.31532155382312865,
+      "grad_norm": 0.44937631487846375,
+      "learning_rate": 8.012765593891688e-05,
+      "loss": 0.9921,
+      "step": 3529
+    },
+    {
+      "epoch": 0.31541090535439054,
+      "grad_norm": 0.5252249836921692,
+      "learning_rate": 8.011610590517932e-05,
+      "loss": 0.9491,
+      "step": 3530
+    },
+    {
+      "epoch": 0.31550025688565236,
+      "grad_norm": 0.3755653500556946,
+      "learning_rate": 8.010455334889689e-05,
+      "loss": 1.0336,
+      "step": 3531
+    },
+    {
+      "epoch": 0.31558960841691425,
+      "grad_norm": 0.47862106561660767,
+      "learning_rate": 8.00929982710372e-05,
+      "loss": 1.0282,
+      "step": 3532
+    },
+    {
+      "epoch": 0.31567895994817613,
+      "grad_norm": 0.47411367297172546,
+      "learning_rate": 8.008144067256815e-05,
+      "loss": 0.8744,
+      "step": 3533
+    },
+    {
+      "epoch": 0.31576831147943796,
+      "grad_norm": 0.41446375846862793,
+      "learning_rate": 8.006988055445778e-05,
+      "loss": 0.9799,
+      "step": 3534
+    },
+    {
+      "epoch": 0.31585766301069984,
+      "grad_norm": 0.45246684551239014,
+      "learning_rate": 8.00583179176744e-05,
+      "loss": 0.9723,
+      "step": 3535
+    },
+    {
+      "epoch": 0.3159470145419617,
+      "grad_norm": 0.46888962388038635,
+      "learning_rate": 8.004675276318651e-05,
+      "loss": 0.9581,
+      "step": 3536
+    },
+    {
+      "epoch": 0.3160363660732236,
+      "grad_norm": 0.4019709527492523,
+      "learning_rate": 8.00351850919628e-05,
+      "loss": 0.982,
+      "step": 3537
+    },
+    {
+      "epoch": 0.31612571760448543,
+      "grad_norm": 0.4450971186161041,
+      "learning_rate": 8.002361490497217e-05,
+      "loss": 0.9854,
+      "step": 3538
+    },
+    {
+      "epoch": 0.3162150691357473,
+      "grad_norm": 0.5295263528823853,
+      "learning_rate": 8.001204220318377e-05,
+      "loss": 0.9173,
+      "step": 3539
+    },
+    {
+      "epoch": 0.3163044206670092,
+      "grad_norm": 0.47927385568618774,
+      "learning_rate": 8.000046698756694e-05,
+      "loss": 0.8995,
+      "step": 3540
+    },
+    {
+      "epoch": 0.31639377219827103,
+      "grad_norm": 0.6096744537353516,
+      "learning_rate": 7.99888892590912e-05,
+      "loss": 0.9354,
+      "step": 3541
+    },
+    {
+      "epoch": 0.3164831237295329,
+      "grad_norm": 0.40292373299598694,
+      "learning_rate": 7.997730901872635e-05,
+      "loss": 1.0007,
+      "step": 3542
+    },
+    {
+      "epoch": 0.3165724752607948,
+      "grad_norm": 0.532189667224884,
+      "learning_rate": 7.996572626744232e-05,
+      "loss": 1.0307,
+      "step": 3543
+    },
+    {
+      "epoch": 0.3166618267920566,
+      "grad_norm": 0.5631407499313354,
+      "learning_rate": 7.99541410062093e-05,
+      "loss": 0.9184,
+      "step": 3544
+    },
+    {
+      "epoch": 0.3167511783233185,
+      "grad_norm": 0.4307250678539276,
+      "learning_rate": 7.994255323599769e-05,
+      "loss": 0.9451,
+      "step": 3545
+    },
+    {
+      "epoch": 0.3168405298545804,
+      "grad_norm": 0.4229245185852051,
+      "learning_rate": 7.993096295777807e-05,
+      "loss": 1.0172,
+      "step": 3546
+    },
+    {
+      "epoch": 0.31692988138584227,
+      "grad_norm": 0.4741484522819519,
+      "learning_rate": 7.991937017252126e-05,
+      "loss": 0.9643,
+      "step": 3547
+    },
+    {
+      "epoch": 0.3170192329171041,
+      "grad_norm": 0.46456822752952576,
+      "learning_rate": 7.990777488119829e-05,
+      "loss": 1.0751,
+      "step": 3548
+    },
+    {
+      "epoch": 0.317108584448366,
+      "grad_norm": 0.4328577518463135,
+      "learning_rate": 7.989617708478039e-05,
+      "loss": 0.9881,
+      "step": 3549
+    },
+    {
+      "epoch": 0.31719793597962787,
+      "grad_norm": 0.48607033491134644,
+      "learning_rate": 7.988457678423898e-05,
+      "loss": 1.0788,
+      "step": 3550
+    },
+    {
+      "epoch": 0.3172872875108897,
+      "grad_norm": 0.4115293622016907,
+      "learning_rate": 7.987297398054572e-05,
+      "loss": 1.005,
+      "step": 3551
+    },
+    {
+      "epoch": 0.3173766390421516,
+      "grad_norm": 0.4423523247241974,
+      "learning_rate": 7.986136867467247e-05,
+      "loss": 1.0001,
+      "step": 3552
+    },
+    {
+      "epoch": 0.31746599057341346,
+      "grad_norm": 0.4115248918533325,
+      "learning_rate": 7.984976086759128e-05,
+      "loss": 1.009,
+      "step": 3553
+    },
+    {
+      "epoch": 0.31755534210467534,
+      "grad_norm": 0.4350251853466034,
+      "learning_rate": 7.983815056027444e-05,
+      "loss": 0.9741,
+      "step": 3554
+    },
+    {
+      "epoch": 0.31764469363593717,
+      "grad_norm": 0.42540842294692993,
+      "learning_rate": 7.982653775369444e-05,
+      "loss": 0.9493,
+      "step": 3555
+    },
+    {
+      "epoch": 0.31773404516719905,
+      "grad_norm": 0.4265209436416626,
+      "learning_rate": 7.981492244882398e-05,
+      "loss": 0.9707,
+      "step": 3556
+    },
+    {
+      "epoch": 0.31782339669846094,
+      "grad_norm": 0.4104164242744446,
+      "learning_rate": 7.980330464663597e-05,
+      "loss": 1.0317,
+      "step": 3557
+    },
+    {
+      "epoch": 0.31791274822972276,
+      "grad_norm": 0.39542466402053833,
+      "learning_rate": 7.979168434810352e-05,
+      "loss": 0.9764,
+      "step": 3558
+    },
+    {
+      "epoch": 0.31800209976098465,
+      "grad_norm": 0.3717448115348816,
+      "learning_rate": 7.978006155419993e-05,
+      "loss": 0.9869,
+      "step": 3559
+    },
+    {
+      "epoch": 0.31809145129224653,
+      "grad_norm": 0.42824071645736694,
+      "learning_rate": 7.976843626589875e-05,
+      "loss": 0.9522,
+      "step": 3560
+    },
+    {
+      "epoch": 0.3181808028235084,
+      "grad_norm": 0.5165953636169434,
+      "learning_rate": 7.975680848417373e-05,
+      "loss": 1.0895,
+      "step": 3561
+    },
+    {
+      "epoch": 0.31827015435477024,
+      "grad_norm": 0.4496411681175232,
+      "learning_rate": 7.974517820999883e-05,
+      "loss": 0.9996,
+      "step": 3562
+    },
+    {
+      "epoch": 0.3183595058860321,
+      "grad_norm": 0.41469454765319824,
+      "learning_rate": 7.973354544434818e-05,
+      "loss": 0.9418,
+      "step": 3563
+    },
+    {
+      "epoch": 0.318448857417294,
+      "grad_norm": 0.4479805529117584,
+      "learning_rate": 7.972191018819615e-05,
+      "loss": 0.986,
+      "step": 3564
+    },
+    {
+      "epoch": 0.31853820894855583,
+      "grad_norm": 0.5563523173332214,
+      "learning_rate": 7.971027244251734e-05,
+      "loss": 0.9081,
+      "step": 3565
+    },
+    {
+      "epoch": 0.3186275604798177,
+      "grad_norm": 0.44090694189071655,
+      "learning_rate": 7.969863220828654e-05,
+      "loss": 0.968,
+      "step": 3566
+    },
+    {
+      "epoch": 0.3187169120110796,
+      "grad_norm": 0.4855956435203552,
+      "learning_rate": 7.968698948647872e-05,
+      "loss": 0.8986,
+      "step": 3567
+    },
+    {
+      "epoch": 0.3188062635423415,
+      "grad_norm": 0.5776668787002563,
+      "learning_rate": 7.967534427806909e-05,
+      "loss": 1.0122,
+      "step": 3568
+    },
+    {
+      "epoch": 0.3188956150736033,
+      "grad_norm": 0.487651526927948,
+      "learning_rate": 7.966369658403305e-05,
+      "loss": 0.9403,
+      "step": 3569
+    },
+    {
+      "epoch": 0.3189849666048652,
+      "grad_norm": 0.5066558122634888,
+      "learning_rate": 7.965204640534623e-05,
+      "loss": 0.9553,
+      "step": 3570
+    },
+    {
+      "epoch": 0.3190743181361271,
+      "grad_norm": 0.47500649094581604,
+      "learning_rate": 7.964039374298447e-05,
+      "loss": 0.9324,
+      "step": 3571
+    },
+    {
+      "epoch": 0.3191636696673889,
+      "grad_norm": 0.3772512376308441,
+      "learning_rate": 7.962873859792377e-05,
+      "loss": 0.9869,
+      "step": 3572
+    },
+    {
+      "epoch": 0.3192530211986508,
+      "grad_norm": 0.44486117362976074,
+      "learning_rate": 7.96170809711404e-05,
+      "loss": 1.0367,
+      "step": 3573
+    },
+    {
+      "epoch": 0.31934237272991267,
+      "grad_norm": 0.6170856356620789,
+      "learning_rate": 7.96054208636108e-05,
+      "loss": 1.091,
+      "step": 3574
+    },
+    {
+      "epoch": 0.3194317242611745,
+      "grad_norm": 0.46623945236206055,
+      "learning_rate": 7.959375827631165e-05,
+      "loss": 0.9985,
+      "step": 3575
+    },
+    {
+      "epoch": 0.3195210757924364,
+      "grad_norm": 0.38250869512557983,
+      "learning_rate": 7.958209321021979e-05,
+      "loss": 1.027,
+      "step": 3576
+    },
+    {
+      "epoch": 0.31961042732369827,
+      "grad_norm": 0.36755305528640747,
+      "learning_rate": 7.95704256663123e-05,
+      "loss": 1.0075,
+      "step": 3577
+    },
+    {
+      "epoch": 0.31969977885496015,
+      "grad_norm": 0.47335994243621826,
+      "learning_rate": 7.955875564556645e-05,
+      "loss": 0.9888,
+      "step": 3578
+    },
+    {
+      "epoch": 0.319789130386222,
+      "grad_norm": 0.4479726254940033,
+      "learning_rate": 7.954708314895975e-05,
+      "loss": 0.9571,
+      "step": 3579
+    },
+    {
+      "epoch": 0.31987848191748386,
+      "grad_norm": 0.4270458519458771,
+      "learning_rate": 7.953540817746988e-05,
+      "loss": 0.9431,
+      "step": 3580
+    },
+    {
+      "epoch": 0.31996783344874574,
+      "grad_norm": 0.42351189255714417,
+      "learning_rate": 7.952373073207478e-05,
+      "loss": 0.9591,
+      "step": 3581
+    },
+    {
+      "epoch": 0.32005718498000757,
+      "grad_norm": 0.44711586833000183,
+      "learning_rate": 7.951205081375249e-05,
+      "loss": 0.9685,
+      "step": 3582
+    },
+    {
+      "epoch": 0.32014653651126945,
+      "grad_norm": 0.44724342226982117,
+      "learning_rate": 7.950036842348139e-05,
+      "loss": 0.9795,
+      "step": 3583
+    },
+    {
+      "epoch": 0.32023588804253134,
+      "grad_norm": 0.4775746762752533,
+      "learning_rate": 7.948868356223997e-05,
+      "loss": 0.9054,
+      "step": 3584
+    },
+    {
+      "epoch": 0.3203252395737932,
+      "grad_norm": 0.4386691451072693,
+      "learning_rate": 7.9476996231007e-05,
+      "loss": 0.9949,
+      "step": 3585
+    },
+    {
+      "epoch": 0.32041459110505505,
+      "grad_norm": 0.40129563212394714,
+      "learning_rate": 7.946530643076138e-05,
+      "loss": 1.0254,
+      "step": 3586
+    },
+    {
+      "epoch": 0.32050394263631693,
+      "grad_norm": 0.4904455542564392,
+      "learning_rate": 7.945361416248226e-05,
+      "loss": 0.8934,
+      "step": 3587
+    },
+    {
+      "epoch": 0.3205932941675788,
+      "grad_norm": 0.4553501009941101,
+      "learning_rate": 7.9441919427149e-05,
+      "loss": 0.9707,
+      "step": 3588
+    },
+    {
+      "epoch": 0.32068264569884064,
+      "grad_norm": 0.38946229219436646,
+      "learning_rate": 7.943022222574116e-05,
+      "loss": 1.0248,
+      "step": 3589
+    },
+    {
+      "epoch": 0.3207719972301025,
+      "grad_norm": 0.39546647667884827,
+      "learning_rate": 7.941852255923852e-05,
+      "loss": 1.0132,
+      "step": 3590
+    },
+    {
+      "epoch": 0.3208613487613644,
+      "grad_norm": 0.4963313937187195,
+      "learning_rate": 7.940682042862104e-05,
+      "loss": 0.9254,
+      "step": 3591
+    },
+    {
+      "epoch": 0.3209507002926263,
+      "grad_norm": 0.42194679379463196,
+      "learning_rate": 7.939511583486887e-05,
+      "loss": 0.9982,
+      "step": 3592
+    },
+    {
+      "epoch": 0.3210400518238881,
+      "grad_norm": 0.4145607054233551,
+      "learning_rate": 7.938340877896244e-05,
+      "loss": 1.0461,
+      "step": 3593
+    },
+    {
+      "epoch": 0.32112940335515,
+      "grad_norm": 0.5402034521102905,
+      "learning_rate": 7.937169926188232e-05,
+      "loss": 1.0905,
+      "step": 3594
+    },
+    {
+      "epoch": 0.3212187548864119,
+      "grad_norm": 0.394927978515625,
+      "learning_rate": 7.935998728460929e-05,
+      "loss": 0.9712,
+      "step": 3595
+    },
+    {
+      "epoch": 0.3213081064176737,
+      "grad_norm": 0.393349289894104,
+      "learning_rate": 7.934827284812438e-05,
+      "loss": 0.9541,
+      "step": 3596
+    },
+    {
+      "epoch": 0.3213974579489356,
+      "grad_norm": 0.4421713352203369,
+      "learning_rate": 7.93365559534088e-05,
+      "loss": 1.0013,
+      "step": 3597
+    },
+    {
+      "epoch": 0.3214868094801975,
+      "grad_norm": 0.46195080876350403,
+      "learning_rate": 7.932483660144394e-05,
+      "loss": 0.9468,
+      "step": 3598
+    },
+    {
+      "epoch": 0.32157616101145936,
+      "grad_norm": 0.419318825006485,
+      "learning_rate": 7.931311479321144e-05,
+      "loss": 0.9858,
+      "step": 3599
+    },
+    {
+      "epoch": 0.3216655125427212,
+      "grad_norm": 0.4154645800590515,
+      "learning_rate": 7.93013905296931e-05,
+      "loss": 0.9703,
+      "step": 3600
+    },
+    {
+      "epoch": 0.32175486407398307,
+      "grad_norm": 0.5366727709770203,
+      "learning_rate": 7.9289663811871e-05,
+      "loss": 0.9062,
+      "step": 3601
+    },
+    {
+      "epoch": 0.32184421560524495,
+      "grad_norm": 0.48142287135124207,
+      "learning_rate": 7.927793464072734e-05,
+      "loss": 1.0538,
+      "step": 3602
+    },
+    {
+      "epoch": 0.3219335671365068,
+      "grad_norm": 0.4503115713596344,
+      "learning_rate": 7.926620301724459e-05,
+      "loss": 1.0272,
+      "step": 3603
+    },
+    {
+      "epoch": 0.32202291866776867,
+      "grad_norm": 0.536191999912262,
+      "learning_rate": 7.925446894240536e-05,
+      "loss": 0.9567,
+      "step": 3604
+    },
+    {
+      "epoch": 0.32211227019903055,
+      "grad_norm": 0.38099348545074463,
+      "learning_rate": 7.924273241719254e-05,
+      "loss": 0.9871,
+      "step": 3605
+    },
+    {
+      "epoch": 0.3222016217302924,
+      "grad_norm": 0.5245529413223267,
+      "learning_rate": 7.923099344258915e-05,
+      "loss": 0.9687,
+      "step": 3606
+    },
+    {
+      "epoch": 0.32229097326155426,
+      "grad_norm": 0.424612432718277,
+      "learning_rate": 7.921925201957851e-05,
+      "loss": 1.0479,
+      "step": 3607
+    },
+    {
+      "epoch": 0.32238032479281614,
+      "grad_norm": 0.4393298625946045,
+      "learning_rate": 7.920750814914404e-05,
+      "loss": 1.0369,
+      "step": 3608
+    },
+    {
+      "epoch": 0.322469676324078,
+      "grad_norm": 0.4350713789463043,
+      "learning_rate": 7.919576183226945e-05,
+      "loss": 1.0051,
+      "step": 3609
+    },
+    {
+      "epoch": 0.32255902785533985,
+      "grad_norm": 0.40692925453186035,
+      "learning_rate": 7.918401306993858e-05,
+      "loss": 1.0296,
+      "step": 3610
+    },
+    {
+      "epoch": 0.32264837938660174,
+      "grad_norm": 0.4172888696193695,
+      "learning_rate": 7.917226186313554e-05,
+      "loss": 0.9977,
+      "step": 3611
+    },
+    {
+      "epoch": 0.3227377309178636,
+      "grad_norm": 0.46099647879600525,
+      "learning_rate": 7.916050821284462e-05,
+      "loss": 0.9395,
+      "step": 3612
+    },
+    {
+      "epoch": 0.32282708244912545,
+      "grad_norm": 0.44035103917121887,
+      "learning_rate": 7.914875212005032e-05,
+      "loss": 0.9543,
+      "step": 3613
+    },
+    {
+      "epoch": 0.32291643398038733,
+      "grad_norm": 0.3710711896419525,
+      "learning_rate": 7.913699358573732e-05,
+      "loss": 0.9894,
+      "step": 3614
+    },
+    {
+      "epoch": 0.3230057855116492,
+      "grad_norm": 0.5159255862236023,
+      "learning_rate": 7.912523261089051e-05,
+      "loss": 0.9674,
+      "step": 3615
+    },
+    {
+      "epoch": 0.3230951370429111,
+      "grad_norm": 0.5290977954864502,
+      "learning_rate": 7.911346919649504e-05,
+      "loss": 1.0115,
+      "step": 3616
+    },
+    {
+      "epoch": 0.3231844885741729,
+      "grad_norm": 0.4615987241268158,
+      "learning_rate": 7.910170334353619e-05,
+      "loss": 0.9218,
+      "step": 3617
+    },
+    {
+      "epoch": 0.3232738401054348,
+      "grad_norm": 0.41924357414245605,
+      "learning_rate": 7.908993505299948e-05,
+      "loss": 1.0448,
+      "step": 3618
+    },
+    {
+      "epoch": 0.3233631916366967,
+      "grad_norm": 0.45864051580429077,
+      "learning_rate": 7.907816432587062e-05,
+      "loss": 1.0082,
+      "step": 3619
+    },
+    {
+      "epoch": 0.3234525431679585,
+      "grad_norm": 0.3755020797252655,
+      "learning_rate": 7.906639116313558e-05,
+      "loss": 0.9703,
+      "step": 3620
+    },
+    {
+      "epoch": 0.3235418946992204,
+      "grad_norm": 0.40216493606567383,
+      "learning_rate": 7.905461556578043e-05,
+      "loss": 1.0177,
+      "step": 3621
+    },
+    {
+      "epoch": 0.3236312462304823,
+      "grad_norm": 0.42985835671424866,
+      "learning_rate": 7.904283753479154e-05,
+      "loss": 0.9477,
+      "step": 3622
+    },
+    {
+      "epoch": 0.32372059776174417,
+      "grad_norm": 0.4799264371395111,
+      "learning_rate": 7.903105707115543e-05,
+      "loss": 1.0422,
+      "step": 3623
+    },
+    {
+      "epoch": 0.323809949293006,
+      "grad_norm": 0.3927200734615326,
+      "learning_rate": 7.901927417585884e-05,
+      "loss": 0.9533,
+      "step": 3624
+    },
+    {
+      "epoch": 0.3238993008242679,
+      "grad_norm": 0.4355437159538269,
+      "learning_rate": 7.90074888498887e-05,
+      "loss": 0.9825,
+      "step": 3625
+    },
+    {
+      "epoch": 0.32398865235552976,
+      "grad_norm": 0.40045395493507385,
+      "learning_rate": 7.899570109423217e-05,
+      "loss": 1.0001,
+      "step": 3626
+    },
+    {
+      "epoch": 0.3240780038867916,
+      "grad_norm": 0.4917450249195099,
+      "learning_rate": 7.898391090987662e-05,
+      "loss": 0.9796,
+      "step": 3627
+    },
+    {
+      "epoch": 0.32416735541805347,
+      "grad_norm": 0.46011456847190857,
+      "learning_rate": 7.897211829780959e-05,
+      "loss": 0.9242,
+      "step": 3628
+    },
+    {
+      "epoch": 0.32425670694931535,
+      "grad_norm": 0.4157116711139679,
+      "learning_rate": 7.896032325901883e-05,
+      "loss": 1.0034,
+      "step": 3629
+    },
+    {
+      "epoch": 0.32434605848057724,
+      "grad_norm": 0.3974458575248718,
+      "learning_rate": 7.894852579449227e-05,
+      "loss": 1.0186,
+      "step": 3630
+    },
+    {
+      "epoch": 0.32443541001183906,
+      "grad_norm": 0.38465315103530884,
+      "learning_rate": 7.893672590521814e-05,
+      "loss": 0.9477,
+      "step": 3631
+    },
+    {
+      "epoch": 0.32452476154310095,
+      "grad_norm": 0.4720938801765442,
+      "learning_rate": 7.892492359218477e-05,
+      "loss": 0.9311,
+      "step": 3632
+    },
+    {
+      "epoch": 0.32461411307436283,
+      "grad_norm": 0.4968278408050537,
+      "learning_rate": 7.89131188563807e-05,
+      "loss": 0.9531,
+      "step": 3633
+    },
+    {
+      "epoch": 0.32470346460562466,
+      "grad_norm": 0.43185052275657654,
+      "learning_rate": 7.890131169879477e-05,
+      "loss": 1.0749,
+      "step": 3634
+    },
+    {
+      "epoch": 0.32479281613688654,
+      "grad_norm": 0.4034889042377472,
+      "learning_rate": 7.888950212041591e-05,
+      "loss": 1.0049,
+      "step": 3635
+    },
+    {
+      "epoch": 0.3248821676681484,
+      "grad_norm": 0.4653837978839874,
+      "learning_rate": 7.88776901222333e-05,
+      "loss": 0.9604,
+      "step": 3636
+    },
+    {
+      "epoch": 0.32497151919941025,
+      "grad_norm": 0.4228241741657257,
+      "learning_rate": 7.886587570523634e-05,
+      "loss": 0.9343,
+      "step": 3637
+    },
+    {
+      "epoch": 0.32506087073067214,
+      "grad_norm": 0.3874700665473938,
+      "learning_rate": 7.88540588704146e-05,
+      "loss": 1.0345,
+      "step": 3638
+    },
+    {
+      "epoch": 0.325150222261934,
+      "grad_norm": 0.501092791557312,
+      "learning_rate": 7.884223961875785e-05,
+      "loss": 0.9159,
+      "step": 3639
+    },
+    {
+      "epoch": 0.3252395737931959,
+      "grad_norm": 0.6081518530845642,
+      "learning_rate": 7.88304179512561e-05,
+      "loss": 0.8758,
+      "step": 3640
+    },
+    {
+      "epoch": 0.32532892532445773,
+      "grad_norm": 0.4688172936439514,
+      "learning_rate": 7.881859386889954e-05,
+      "loss": 0.9757,
+      "step": 3641
+    },
+    {
+      "epoch": 0.3254182768557196,
+      "grad_norm": 0.4610939025878906,
+      "learning_rate": 7.880676737267857e-05,
+      "loss": 1.0032,
+      "step": 3642
+    },
+    {
+      "epoch": 0.3255076283869815,
+      "grad_norm": 0.49634119868278503,
+      "learning_rate": 7.879493846358377e-05,
+      "loss": 0.9689,
+      "step": 3643
+    },
+    {
+      "epoch": 0.3255969799182433,
+      "grad_norm": 0.4354616105556488,
+      "learning_rate": 7.878310714260593e-05,
+      "loss": 0.9444,
+      "step": 3644
+    },
+    {
+      "epoch": 0.3256863314495052,
+      "grad_norm": 0.4469475746154785,
+      "learning_rate": 7.877127341073606e-05,
+      "loss": 1.0258,
+      "step": 3645
+    },
+    {
+      "epoch": 0.3257756829807671,
+      "grad_norm": 0.4689541459083557,
+      "learning_rate": 7.875943726896538e-05,
+      "loss": 0.8939,
+      "step": 3646
+    },
+    {
+      "epoch": 0.325865034512029,
+      "grad_norm": 0.39675360918045044,
+      "learning_rate": 7.874759871828527e-05,
+      "loss": 0.9616,
+      "step": 3647
+    },
+    {
+      "epoch": 0.3259543860432908,
+      "grad_norm": 0.4091286361217499,
+      "learning_rate": 7.873575775968734e-05,
+      "loss": 0.8918,
+      "step": 3648
+    },
+    {
+      "epoch": 0.3260437375745527,
+      "grad_norm": 0.3711673319339752,
+      "learning_rate": 7.872391439416339e-05,
+      "loss": 1.0246,
+      "step": 3649
+    },
+    {
+      "epoch": 0.32613308910581457,
+      "grad_norm": 0.4348282217979431,
+      "learning_rate": 7.871206862270543e-05,
+      "loss": 0.9172,
+      "step": 3650
+    },
+    {
+      "epoch": 0.3262224406370764,
+      "grad_norm": 0.46988198161125183,
+      "learning_rate": 7.870022044630569e-05,
+      "loss": 0.9724,
+      "step": 3651
+    },
+    {
+      "epoch": 0.3263117921683383,
+      "grad_norm": 0.42179208993911743,
+      "learning_rate": 7.868836986595656e-05,
+      "loss": 0.9622,
+      "step": 3652
+    },
+    {
+      "epoch": 0.32640114369960016,
+      "grad_norm": 0.507832407951355,
+      "learning_rate": 7.867651688265066e-05,
+      "loss": 0.9428,
+      "step": 3653
+    },
+    {
+      "epoch": 0.32649049523086204,
+      "grad_norm": 0.458981990814209,
+      "learning_rate": 7.866466149738079e-05,
+      "loss": 1.0245,
+      "step": 3654
+    },
+    {
+      "epoch": 0.32657984676212387,
+      "grad_norm": 0.47192487120628357,
+      "learning_rate": 7.865280371113998e-05,
+      "loss": 0.9683,
+      "step": 3655
+    },
+    {
+      "epoch": 0.32666919829338575,
+      "grad_norm": 0.4830261170864105,
+      "learning_rate": 7.864094352492143e-05,
+      "loss": 1.0502,
+      "step": 3656
+    },
+    {
+      "epoch": 0.32675854982464764,
+      "grad_norm": 0.43681490421295166,
+      "learning_rate": 7.862908093971859e-05,
+      "loss": 0.9709,
+      "step": 3657
+    },
+    {
+      "epoch": 0.32684790135590946,
+      "grad_norm": 0.4974532127380371,
+      "learning_rate": 7.861721595652507e-05,
+      "loss": 1.012,
+      "step": 3658
+    },
+    {
+      "epoch": 0.32693725288717135,
+      "grad_norm": 0.43522706627845764,
+      "learning_rate": 7.860534857633464e-05,
+      "loss": 0.9935,
+      "step": 3659
+    },
+    {
+      "epoch": 0.32702660441843323,
+      "grad_norm": 0.3843841254711151,
+      "learning_rate": 7.859347880014138e-05,
+      "loss": 1.01,
+      "step": 3660
+    },
+    {
+      "epoch": 0.3271159559496951,
+      "grad_norm": 0.38631847500801086,
+      "learning_rate": 7.858160662893948e-05,
+      "loss": 1.0152,
+      "step": 3661
+    },
+    {
+      "epoch": 0.32720530748095694,
+      "grad_norm": 0.4131559431552887,
+      "learning_rate": 7.856973206372336e-05,
+      "loss": 0.9994,
+      "step": 3662
+    },
+    {
+      "epoch": 0.3272946590122188,
+      "grad_norm": 0.6184050440788269,
+      "learning_rate": 7.855785510548765e-05,
+      "loss": 0.97,
+      "step": 3663
+    },
+    {
+      "epoch": 0.3273840105434807,
+      "grad_norm": 0.4109181761741638,
+      "learning_rate": 7.854597575522717e-05,
+      "loss": 1.0286,
+      "step": 3664
+    },
+    {
+      "epoch": 0.32747336207474254,
+      "grad_norm": 0.46452227234840393,
+      "learning_rate": 7.853409401393694e-05,
+      "loss": 1.0194,
+      "step": 3665
+    },
+    {
+      "epoch": 0.3275627136060044,
+      "grad_norm": 0.44786307215690613,
+      "learning_rate": 7.85222098826122e-05,
+      "loss": 0.9468,
+      "step": 3666
+    },
+    {
+      "epoch": 0.3276520651372663,
+      "grad_norm": 0.4409159719944,
+      "learning_rate": 7.851032336224835e-05,
+      "loss": 0.9174,
+      "step": 3667
+    },
+    {
+      "epoch": 0.32774141666852813,
+      "grad_norm": 0.35179105401039124,
+      "learning_rate": 7.849843445384102e-05,
+      "loss": 1.0162,
+      "step": 3668
+    },
+    {
+      "epoch": 0.32783076819979,
+      "grad_norm": 0.45857298374176025,
+      "learning_rate": 7.848654315838603e-05,
+      "loss": 0.9139,
+      "step": 3669
+    },
+    {
+      "epoch": 0.3279201197310519,
+      "grad_norm": 0.3982926309108734,
+      "learning_rate": 7.84746494768794e-05,
+      "loss": 1.0394,
+      "step": 3670
+    },
+    {
+      "epoch": 0.3280094712623138,
+      "grad_norm": 0.48336413502693176,
+      "learning_rate": 7.846275341031736e-05,
+      "loss": 0.9504,
+      "step": 3671
+    },
+    {
+      "epoch": 0.3280988227935756,
+      "grad_norm": 0.46158286929130554,
+      "learning_rate": 7.845085495969635e-05,
+      "loss": 0.9069,
+      "step": 3672
+    },
+    {
+      "epoch": 0.3281881743248375,
+      "grad_norm": 0.5317728519439697,
+      "learning_rate": 7.843895412601296e-05,
+      "loss": 0.9455,
+      "step": 3673
+    },
+    {
+      "epoch": 0.3282775258560994,
+      "grad_norm": 0.4177683889865875,
+      "learning_rate": 7.842705091026403e-05,
+      "loss": 0.9303,
+      "step": 3674
+    },
+    {
+      "epoch": 0.3283668773873612,
+      "grad_norm": 0.4196977913379669,
+      "learning_rate": 7.841514531344655e-05,
+      "loss": 0.9551,
+      "step": 3675
+    },
+    {
+      "epoch": 0.3284562289186231,
+      "grad_norm": 0.48437947034835815,
+      "learning_rate": 7.840323733655778e-05,
+      "loss": 0.8823,
+      "step": 3676
+    },
+    {
+      "epoch": 0.32854558044988497,
+      "grad_norm": 0.38968583941459656,
+      "learning_rate": 7.839132698059515e-05,
+      "loss": 0.9749,
+      "step": 3677
+    },
+    {
+      "epoch": 0.32863493198114685,
+      "grad_norm": 0.41708287596702576,
+      "learning_rate": 7.837941424655624e-05,
+      "loss": 0.9639,
+      "step": 3678
+    },
+    {
+      "epoch": 0.3287242835124087,
+      "grad_norm": 0.43757346272468567,
+      "learning_rate": 7.836749913543888e-05,
+      "loss": 0.8771,
+      "step": 3679
+    },
+    {
+      "epoch": 0.32881363504367056,
+      "grad_norm": 0.513954758644104,
+      "learning_rate": 7.835558164824108e-05,
+      "loss": 0.9956,
+      "step": 3680
+    },
+    {
+      "epoch": 0.32890298657493244,
+      "grad_norm": 0.3900381624698639,
+      "learning_rate": 7.834366178596109e-05,
+      "loss": 1.0153,
+      "step": 3681
+    },
+    {
+      "epoch": 0.32899233810619427,
+      "grad_norm": 0.45596617460250854,
+      "learning_rate": 7.83317395495973e-05,
+      "loss": 1.1101,
+      "step": 3682
+    },
+    {
+      "epoch": 0.32908168963745615,
+      "grad_norm": 0.4868778586387634,
+      "learning_rate": 7.831981494014833e-05,
+      "loss": 1.0163,
+      "step": 3683
+    },
+    {
+      "epoch": 0.32917104116871804,
+      "grad_norm": 0.40496212244033813,
+      "learning_rate": 7.830788795861296e-05,
+      "loss": 0.9321,
+      "step": 3684
+    },
+    {
+      "epoch": 0.3292603926999799,
+      "grad_norm": 0.42820194363594055,
+      "learning_rate": 7.829595860599026e-05,
+      "loss": 1.0163,
+      "step": 3685
+    },
+    {
+      "epoch": 0.32934974423124175,
+      "grad_norm": 0.42636728286743164,
+      "learning_rate": 7.828402688327941e-05,
+      "loss": 0.9583,
+      "step": 3686
+    },
+    {
+      "epoch": 0.32943909576250363,
+      "grad_norm": 0.39956799149513245,
+      "learning_rate": 7.827209279147982e-05,
+      "loss": 1.0322,
+      "step": 3687
+    },
+    {
+      "epoch": 0.3295284472937655,
+      "grad_norm": 0.4089909791946411,
+      "learning_rate": 7.826015633159112e-05,
+      "loss": 1.0177,
+      "step": 3688
+    },
+    {
+      "epoch": 0.32961779882502734,
+      "grad_norm": 0.4677819013595581,
+      "learning_rate": 7.824821750461308e-05,
+      "loss": 0.9698,
+      "step": 3689
+    },
+    {
+      "epoch": 0.3297071503562892,
+      "grad_norm": 0.4185117781162262,
+      "learning_rate": 7.823627631154571e-05,
+      "loss": 1.001,
+      "step": 3690
+    },
+    {
+      "epoch": 0.3297965018875511,
+      "grad_norm": 0.3975558280944824,
+      "learning_rate": 7.822433275338923e-05,
+      "loss": 0.9953,
+      "step": 3691
+    },
+    {
+      "epoch": 0.329885853418813,
+      "grad_norm": 0.3640524744987488,
+      "learning_rate": 7.821238683114404e-05,
+      "loss": 1.0064,
+      "step": 3692
+    },
+    {
+      "epoch": 0.3299752049500748,
+      "grad_norm": 0.4032001197338104,
+      "learning_rate": 7.820043854581071e-05,
+      "loss": 1.01,
+      "step": 3693
+    },
+    {
+      "epoch": 0.3300645564813367,
+      "grad_norm": 0.45397523045539856,
+      "learning_rate": 7.818848789839008e-05,
+      "loss": 0.9609,
+      "step": 3694
+    },
+    {
+      "epoch": 0.3301539080125986,
+      "grad_norm": 0.3662114143371582,
+      "learning_rate": 7.81765348898831e-05,
+      "loss": 1.0157,
+      "step": 3695
+    },
+    {
+      "epoch": 0.3302432595438604,
+      "grad_norm": 0.38710111379623413,
+      "learning_rate": 7.816457952129099e-05,
+      "loss": 0.9811,
+      "step": 3696
+    },
+    {
+      "epoch": 0.3303326110751223,
+      "grad_norm": 0.40280595421791077,
+      "learning_rate": 7.815262179361514e-05,
+      "loss": 0.9899,
+      "step": 3697
+    },
+    {
+      "epoch": 0.3304219626063842,
+      "grad_norm": 0.4115069806575775,
+      "learning_rate": 7.814066170785714e-05,
+      "loss": 0.9829,
+      "step": 3698
+    },
+    {
+      "epoch": 0.330511314137646,
+      "grad_norm": 0.39041948318481445,
+      "learning_rate": 7.812869926501874e-05,
+      "loss": 0.9653,
+      "step": 3699
+    },
+    {
+      "epoch": 0.3306006656689079,
+      "grad_norm": 0.40543365478515625,
+      "learning_rate": 7.811673446610195e-05,
+      "loss": 1.001,
+      "step": 3700
+    },
+    {
+      "epoch": 0.3306900172001698,
+      "grad_norm": 0.36765506863594055,
+      "learning_rate": 7.810476731210896e-05,
+      "loss": 1.0908,
+      "step": 3701
+    },
+    {
+      "epoch": 0.33077936873143166,
+      "grad_norm": 0.45462504029273987,
+      "learning_rate": 7.809279780404212e-05,
+      "loss": 1.0355,
+      "step": 3702
+    },
+    {
+      "epoch": 0.3308687202626935,
+      "grad_norm": 0.5238141417503357,
+      "learning_rate": 7.808082594290402e-05,
+      "loss": 0.9617,
+      "step": 3703
+    },
+    {
+      "epoch": 0.33095807179395537,
+      "grad_norm": 0.4108685851097107,
+      "learning_rate": 7.806885172969742e-05,
+      "loss": 0.9806,
+      "step": 3704
+    },
+    {
+      "epoch": 0.33104742332521725,
+      "grad_norm": 0.320278525352478,
+      "learning_rate": 7.80568751654253e-05,
+      "loss": 1.0341,
+      "step": 3705
+    },
+    {
+      "epoch": 0.3311367748564791,
+      "grad_norm": 0.3879075050354004,
+      "learning_rate": 7.804489625109083e-05,
+      "loss": 0.9905,
+      "step": 3706
+    },
+    {
+      "epoch": 0.33122612638774096,
+      "grad_norm": 0.4919257164001465,
+      "learning_rate": 7.803291498769735e-05,
+      "loss": 0.9408,
+      "step": 3707
+    },
+    {
+      "epoch": 0.33131547791900284,
+      "grad_norm": 0.4680798053741455,
+      "learning_rate": 7.802093137624844e-05,
+      "loss": 0.8917,
+      "step": 3708
+    },
+    {
+      "epoch": 0.3314048294502647,
+      "grad_norm": 0.35378363728523254,
+      "learning_rate": 7.800894541774783e-05,
+      "loss": 1.0099,
+      "step": 3709
+    },
+    {
+      "epoch": 0.33149418098152655,
+      "grad_norm": 0.3704555928707123,
+      "learning_rate": 7.79969571131995e-05,
+      "loss": 0.9949,
+      "step": 3710
+    },
+    {
+      "epoch": 0.33158353251278844,
+      "grad_norm": 0.39882001280784607,
+      "learning_rate": 7.798496646360758e-05,
+      "loss": 0.9924,
+      "step": 3711
+    },
+    {
+      "epoch": 0.3316728840440503,
+      "grad_norm": 0.4005809724330902,
+      "learning_rate": 7.797297346997643e-05,
+      "loss": 0.9349,
+      "step": 3712
+    },
+    {
+      "epoch": 0.33176223557531215,
+      "grad_norm": 0.4036948084831238,
+      "learning_rate": 7.796097813331059e-05,
+      "loss": 0.9319,
+      "step": 3713
+    },
+    {
+      "epoch": 0.33185158710657403,
+      "grad_norm": 0.46617838740348816,
+      "learning_rate": 7.794898045461476e-05,
+      "loss": 0.904,
+      "step": 3714
+    },
+    {
+      "epoch": 0.3319409386378359,
+      "grad_norm": 0.38140738010406494,
+      "learning_rate": 7.79369804348939e-05,
+      "loss": 0.9657,
+      "step": 3715
+    },
+    {
+      "epoch": 0.3320302901690978,
+      "grad_norm": 0.46043315529823303,
+      "learning_rate": 7.792497807515317e-05,
+      "loss": 1.045,
+      "step": 3716
+    },
+    {
+      "epoch": 0.3321196417003596,
+      "grad_norm": 0.4285655617713928,
+      "learning_rate": 7.791297337639784e-05,
+      "loss": 1.0517,
+      "step": 3717
+    },
+    {
+      "epoch": 0.3322089932316215,
+      "grad_norm": 0.43057459592819214,
+      "learning_rate": 7.790096633963348e-05,
+      "loss": 0.9226,
+      "step": 3718
+    },
+    {
+      "epoch": 0.3322983447628834,
+      "grad_norm": 0.5528749823570251,
+      "learning_rate": 7.788895696586577e-05,
+      "loss": 0.9763,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3323876962941452,
+      "grad_norm": 0.49909940361976624,
+      "learning_rate": 7.787694525610066e-05,
+      "loss": 1.0276,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3324770478254071,
+      "grad_norm": 0.458188533782959,
+      "learning_rate": 7.786493121134423e-05,
+      "loss": 0.9415,
+      "step": 3721
+    },
+    {
+      "epoch": 0.332566399356669,
+      "grad_norm": 0.4523871839046478,
+      "learning_rate": 7.785291483260278e-05,
+      "loss": 0.9684,
+      "step": 3722
+    },
+    {
+      "epoch": 0.33265575088793087,
+      "grad_norm": 0.4228318929672241,
+      "learning_rate": 7.784089612088283e-05,
+      "loss": 1.042,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3327451024191927,
+      "grad_norm": 0.3901155889034271,
+      "learning_rate": 7.782887507719108e-05,
+      "loss": 1.0081,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3328344539504546,
+      "grad_norm": 0.39497750997543335,
+      "learning_rate": 7.781685170253439e-05,
+      "loss": 1.0522,
+      "step": 3725
+    },
+    {
+      "epoch": 0.33292380548171646,
+      "grad_norm": 0.43867218494415283,
+      "learning_rate": 7.780482599791987e-05,
+      "loss": 0.9623,
+      "step": 3726
+    },
+    {
+      "epoch": 0.3330131570129783,
+      "grad_norm": 0.43389105796813965,
+      "learning_rate": 7.779279796435479e-05,
+      "loss": 1.0464,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3331025085442402,
+      "grad_norm": 0.4637124538421631,
+      "learning_rate": 7.778076760284665e-05,
+      "loss": 0.9743,
+      "step": 3728
+    },
+    {
+      "epoch": 0.33319186007550206,
+      "grad_norm": 0.48631882667541504,
+      "learning_rate": 7.776873491440307e-05,
+      "loss": 1.0403,
+      "step": 3729
+    },
+    {
+      "epoch": 0.33328121160676394,
+      "grad_norm": 0.4544772505760193,
+      "learning_rate": 7.775669990003197e-05,
+      "loss": 1.0384,
+      "step": 3730
+    },
+    {
+      "epoch": 0.33337056313802577,
+      "grad_norm": 0.41751065850257874,
+      "learning_rate": 7.774466256074137e-05,
+      "loss": 0.9482,
+      "step": 3731
+    },
+    {
+      "epoch": 0.33345991466928765,
+      "grad_norm": 0.42404597997665405,
+      "learning_rate": 7.773262289753956e-05,
+      "loss": 0.9436,
+      "step": 3732
+    },
+    {
+      "epoch": 0.33354926620054953,
+      "grad_norm": 0.400020033121109,
+      "learning_rate": 7.772058091143497e-05,
+      "loss": 0.9523,
+      "step": 3733
+    },
+    {
+      "epoch": 0.33363861773181136,
+      "grad_norm": 0.4418000876903534,
+      "learning_rate": 7.770853660343625e-05,
+      "loss": 1.032,
+      "step": 3734
+    },
+    {
+      "epoch": 0.33372796926307324,
+      "grad_norm": 0.48235222697257996,
+      "learning_rate": 7.769648997455223e-05,
+      "loss": 0.9437,
+      "step": 3735
+    },
+    {
+      "epoch": 0.3338173207943351,
+      "grad_norm": 0.40464332699775696,
+      "learning_rate": 7.768444102579196e-05,
+      "loss": 0.9569,
+      "step": 3736
+    },
+    {
+      "epoch": 0.33390667232559695,
+      "grad_norm": 0.5996637940406799,
+      "learning_rate": 7.767238975816465e-05,
+      "loss": 0.9467,
+      "step": 3737
+    },
+    {
+      "epoch": 0.33399602385685884,
+      "grad_norm": 0.38040077686309814,
+      "learning_rate": 7.766033617267975e-05,
+      "loss": 0.9954,
+      "step": 3738
+    },
+    {
+      "epoch": 0.3340853753881207,
+      "grad_norm": 0.4809688925743103,
+      "learning_rate": 7.764828027034685e-05,
+      "loss": 0.8698,
+      "step": 3739
+    },
+    {
+      "epoch": 0.3341747269193826,
+      "grad_norm": 0.4978516399860382,
+      "learning_rate": 7.763622205217576e-05,
+      "loss": 0.9686,
+      "step": 3740
+    },
+    {
+      "epoch": 0.33426407845064443,
+      "grad_norm": 0.41253742575645447,
+      "learning_rate": 7.762416151917648e-05,
+      "loss": 0.9883,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3343534299819063,
+      "grad_norm": 0.35575124621391296,
+      "learning_rate": 7.761209867235924e-05,
+      "loss": 1.0379,
+      "step": 3742
+    },
+    {
+      "epoch": 0.3344427815131682,
+      "grad_norm": 0.3860713541507721,
+      "learning_rate": 7.760003351273442e-05,
+      "loss": 0.9995,
+      "step": 3743
+    },
+    {
+      "epoch": 0.33453213304443,
+      "grad_norm": 0.3986268639564514,
+      "learning_rate": 7.758796604131258e-05,
+      "loss": 0.986,
+      "step": 3744
+    },
+    {
+      "epoch": 0.3346214845756919,
+      "grad_norm": 0.4240453243255615,
+      "learning_rate": 7.757589625910452e-05,
+      "loss": 0.9722,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3347108361069538,
+      "grad_norm": 0.5454365015029907,
+      "learning_rate": 7.75638241671212e-05,
+      "loss": 0.8707,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3348001876382157,
+      "grad_norm": 0.4972113370895386,
+      "learning_rate": 7.755174976637381e-05,
+      "loss": 0.978,
+      "step": 3747
+    },
+    {
+      "epoch": 0.3348895391694775,
+      "grad_norm": 0.4331071972846985,
+      "learning_rate": 7.753967305787371e-05,
+      "loss": 0.9539,
+      "step": 3748
+    },
+    {
+      "epoch": 0.3349788907007394,
+      "grad_norm": 0.5387925505638123,
+      "learning_rate": 7.752759404263242e-05,
+      "loss": 0.8865,
+      "step": 3749
+    },
+    {
+      "epoch": 0.33506824223200127,
+      "grad_norm": 0.4231729805469513,
+      "learning_rate": 7.751551272166171e-05,
+      "loss": 0.9758,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3351575937632631,
+      "grad_norm": 0.4942473769187927,
+      "learning_rate": 7.750342909597352e-05,
+      "loss": 0.9943,
+      "step": 3751
+    },
+    {
+      "epoch": 0.335246945294525,
+      "grad_norm": 0.4464167356491089,
+      "learning_rate": 7.749134316657997e-05,
+      "loss": 1.0107,
+      "step": 3752
+    },
+    {
+      "epoch": 0.33533629682578686,
+      "grad_norm": 0.4313907027244568,
+      "learning_rate": 7.747925493449342e-05,
+      "loss": 1.0031,
+      "step": 3753
+    },
+    {
+      "epoch": 0.33542564835704874,
+      "grad_norm": 0.42571744322776794,
+      "learning_rate": 7.746716440072632e-05,
+      "loss": 1.003,
+      "step": 3754
+    },
+    {
+      "epoch": 0.33551499988831057,
+      "grad_norm": 0.3863215446472168,
+      "learning_rate": 7.745507156629145e-05,
+      "loss": 1.0386,
+      "step": 3755
+    },
+    {
+      "epoch": 0.33560435141957246,
+      "grad_norm": 0.4166812002658844,
+      "learning_rate": 7.744297643220168e-05,
+      "loss": 0.9372,
+      "step": 3756
+    },
+    {
+      "epoch": 0.33569370295083434,
+      "grad_norm": 0.4798368215560913,
+      "learning_rate": 7.74308789994701e-05,
+      "loss": 0.9966,
+      "step": 3757
+    },
+    {
+      "epoch": 0.33578305448209617,
+      "grad_norm": 0.3768024146556854,
+      "learning_rate": 7.741877926911003e-05,
+      "loss": 0.9845,
+      "step": 3758
+    },
+    {
+      "epoch": 0.33587240601335805,
+      "grad_norm": 0.46540406346321106,
+      "learning_rate": 7.740667724213493e-05,
+      "loss": 0.9618,
+      "step": 3759
+    },
+    {
+      "epoch": 0.33596175754461993,
+      "grad_norm": 0.5055358409881592,
+      "learning_rate": 7.739457291955847e-05,
+      "loss": 1.0283,
+      "step": 3760
+    },
+    {
+      "epoch": 0.3360511090758818,
+      "grad_norm": 0.5068111419677734,
+      "learning_rate": 7.738246630239452e-05,
+      "loss": 0.9803,
+      "step": 3761
+    },
+    {
+      "epoch": 0.33614046060714364,
+      "grad_norm": 0.43414464592933655,
+      "learning_rate": 7.737035739165715e-05,
+      "loss": 1.0182,
+      "step": 3762
+    },
+    {
+      "epoch": 0.3362298121384055,
+      "grad_norm": 0.47884586453437805,
+      "learning_rate": 7.73582461883606e-05,
+      "loss": 1.0077,
+      "step": 3763
+    },
+    {
+      "epoch": 0.3363191636696674,
+      "grad_norm": 0.3981468677520752,
+      "learning_rate": 7.734613269351931e-05,
+      "loss": 1.0856,
+      "step": 3764
+    },
+    {
+      "epoch": 0.33640851520092924,
+      "grad_norm": 0.4908856153488159,
+      "learning_rate": 7.733401690814793e-05,
+      "loss": 0.8595,
+      "step": 3765
+    },
+    {
+      "epoch": 0.3364978667321911,
+      "grad_norm": 0.3917779326438904,
+      "learning_rate": 7.732189883326125e-05,
+      "loss": 0.9844,
+      "step": 3766
+    },
+    {
+      "epoch": 0.336587218263453,
+      "grad_norm": 0.40213218331336975,
+      "learning_rate": 7.730977846987433e-05,
+      "loss": 0.968,
+      "step": 3767
+    },
+    {
+      "epoch": 0.33667656979471483,
+      "grad_norm": 0.3646174967288971,
+      "learning_rate": 7.729765581900235e-05,
+      "loss": 1.0258,
+      "step": 3768
+    },
+    {
+      "epoch": 0.3367659213259767,
+      "grad_norm": 0.41250017285346985,
+      "learning_rate": 7.728553088166075e-05,
+      "loss": 0.9605,
+      "step": 3769
+    },
+    {
+      "epoch": 0.3368552728572386,
+      "grad_norm": 0.384792685508728,
+      "learning_rate": 7.727340365886506e-05,
+      "loss": 1.0067,
+      "step": 3770
+    },
+    {
+      "epoch": 0.3369446243885005,
+      "grad_norm": 0.5328003168106079,
+      "learning_rate": 7.726127415163113e-05,
+      "loss": 0.9602,
+      "step": 3771
+    },
+    {
+      "epoch": 0.3370339759197623,
+      "grad_norm": 0.4145788550376892,
+      "learning_rate": 7.724914236097489e-05,
+      "loss": 1.0511,
+      "step": 3772
+    },
+    {
+      "epoch": 0.3371233274510242,
+      "grad_norm": 0.37632668018341064,
+      "learning_rate": 7.723700828791252e-05,
+      "loss": 0.972,
+      "step": 3773
+    },
+    {
+      "epoch": 0.3372126789822861,
+      "grad_norm": 0.37736862897872925,
+      "learning_rate": 7.722487193346039e-05,
+      "loss": 0.99,
+      "step": 3774
+    },
+    {
+      "epoch": 0.3373020305135479,
+      "grad_norm": 0.4212090075016022,
+      "learning_rate": 7.721273329863504e-05,
+      "loss": 0.9789,
+      "step": 3775
+    },
+    {
+      "epoch": 0.3373913820448098,
+      "grad_norm": 0.5166639089584351,
+      "learning_rate": 7.72005923844532e-05,
+      "loss": 0.9856,
+      "step": 3776
+    },
+    {
+      "epoch": 0.33748073357607167,
+      "grad_norm": 0.46971988677978516,
+      "learning_rate": 7.71884491919318e-05,
+      "loss": 1.048,
+      "step": 3777
+    },
+    {
+      "epoch": 0.33757008510733355,
+      "grad_norm": 0.4689841568470001,
+      "learning_rate": 7.7176303722088e-05,
+      "loss": 0.9496,
+      "step": 3778
+    },
+    {
+      "epoch": 0.3376594366385954,
+      "grad_norm": 0.5772603154182434,
+      "learning_rate": 7.716415597593907e-05,
+      "loss": 0.8967,
+      "step": 3779
+    },
+    {
+      "epoch": 0.33774878816985726,
+      "grad_norm": 0.4483826160430908,
+      "learning_rate": 7.715200595450253e-05,
+      "loss": 1.0697,
+      "step": 3780
+    },
+    {
+      "epoch": 0.33783813970111914,
+      "grad_norm": 0.4816476106643677,
+      "learning_rate": 7.713985365879606e-05,
+      "loss": 0.912,
+      "step": 3781
+    },
+    {
+      "epoch": 0.33792749123238097,
+      "grad_norm": 0.38734710216522217,
+      "learning_rate": 7.712769908983757e-05,
+      "loss": 0.9535,
+      "step": 3782
+    },
+    {
+      "epoch": 0.33801684276364286,
+      "grad_norm": 0.4487532079219818,
+      "learning_rate": 7.711554224864511e-05,
+      "loss": 0.9674,
+      "step": 3783
+    },
+    {
+      "epoch": 0.33810619429490474,
+      "grad_norm": 0.4268551170825958,
+      "learning_rate": 7.710338313623697e-05,
+      "loss": 0.9257,
+      "step": 3784
+    },
+    {
+      "epoch": 0.3381955458261666,
+      "grad_norm": 0.44561639428138733,
+      "learning_rate": 7.709122175363158e-05,
+      "loss": 0.9981,
+      "step": 3785
+    },
+    {
+      "epoch": 0.33828489735742845,
+      "grad_norm": 0.44499921798706055,
+      "learning_rate": 7.707905810184762e-05,
+      "loss": 0.9353,
+      "step": 3786
+    },
+    {
+      "epoch": 0.33837424888869033,
+      "grad_norm": 0.4151676595211029,
+      "learning_rate": 7.706689218190386e-05,
+      "loss": 0.9364,
+      "step": 3787
+    },
+    {
+      "epoch": 0.3384636004199522,
+      "grad_norm": 0.4505821764469147,
+      "learning_rate": 7.705472399481939e-05,
+      "loss": 0.9725,
+      "step": 3788
+    },
+    {
+      "epoch": 0.33855295195121404,
+      "grad_norm": 0.410693883895874,
+      "learning_rate": 7.704255354161341e-05,
+      "loss": 1.0559,
+      "step": 3789
+    },
+    {
+      "epoch": 0.3386423034824759,
+      "grad_norm": 0.38344308733940125,
+      "learning_rate": 7.703038082330532e-05,
+      "loss": 0.9554,
+      "step": 3790
+    },
+    {
+      "epoch": 0.3387316550137378,
+      "grad_norm": 0.3579278588294983,
+      "learning_rate": 7.70182058409147e-05,
+      "loss": 0.9861,
+      "step": 3791
+    },
+    {
+      "epoch": 0.3388210065449997,
+      "grad_norm": 0.4280867278575897,
+      "learning_rate": 7.700602859546134e-05,
+      "loss": 0.991,
+      "step": 3792
+    },
+    {
+      "epoch": 0.3389103580762615,
+      "grad_norm": 0.4199950695037842,
+      "learning_rate": 7.699384908796523e-05,
+      "loss": 0.9886,
+      "step": 3793
+    },
+    {
+      "epoch": 0.3389997096075234,
+      "grad_norm": 0.43950459361076355,
+      "learning_rate": 7.698166731944654e-05,
+      "loss": 0.9472,
+      "step": 3794
+    },
+    {
+      "epoch": 0.3390890611387853,
+      "grad_norm": 0.43579909205436707,
+      "learning_rate": 7.696948329092559e-05,
+      "loss": 0.955,
+      "step": 3795
+    },
+    {
+      "epoch": 0.3391784126700471,
+      "grad_norm": 0.4635678231716156,
+      "learning_rate": 7.695729700342294e-05,
+      "loss": 0.9009,
+      "step": 3796
+    },
+    {
+      "epoch": 0.339267764201309,
+      "grad_norm": 0.4732118248939514,
+      "learning_rate": 7.694510845795933e-05,
+      "loss": 0.9975,
+      "step": 3797
+    },
+    {
+      "epoch": 0.3393571157325709,
+      "grad_norm": 0.4001818299293518,
+      "learning_rate": 7.693291765555567e-05,
+      "loss": 0.9729,
+      "step": 3798
+    },
+    {
+      "epoch": 0.3394464672638327,
+      "grad_norm": 0.39679357409477234,
+      "learning_rate": 7.692072459723307e-05,
+      "loss": 1.0314,
+      "step": 3799
+    },
+    {
+      "epoch": 0.3395358187950946,
+      "grad_norm": 0.40091925859451294,
+      "learning_rate": 7.690852928401285e-05,
+      "loss": 0.9526,
+      "step": 3800
+    },
+    {
+      "epoch": 0.3396251703263565,
+      "grad_norm": 0.46458113193511963,
+      "learning_rate": 7.689633171691644e-05,
+      "loss": 0.9608,
+      "step": 3801
+    },
+    {
+      "epoch": 0.33971452185761836,
+      "grad_norm": 0.5359793305397034,
+      "learning_rate": 7.688413189696559e-05,
+      "loss": 0.9501,
+      "step": 3802
+    },
+    {
+      "epoch": 0.3398038733888802,
+      "grad_norm": 0.43518051505088806,
+      "learning_rate": 7.68719298251821e-05,
+      "loss": 0.9698,
+      "step": 3803
+    },
+    {
+      "epoch": 0.33989322492014207,
+      "grad_norm": 0.42377930879592896,
+      "learning_rate": 7.685972550258809e-05,
+      "loss": 0.974,
+      "step": 3804
+    },
+    {
+      "epoch": 0.33998257645140395,
+      "grad_norm": 0.40546149015426636,
+      "learning_rate": 7.684751893020574e-05,
+      "loss": 0.9878,
+      "step": 3805
+    },
+    {
+      "epoch": 0.3400719279826658,
+      "grad_norm": 0.43093767762184143,
+      "learning_rate": 7.683531010905748e-05,
+      "loss": 0.9468,
+      "step": 3806
+    },
+    {
+      "epoch": 0.34016127951392766,
+      "grad_norm": 0.4087904095649719,
+      "learning_rate": 7.682309904016601e-05,
+      "loss": 0.9218,
+      "step": 3807
+    },
+    {
+      "epoch": 0.34025063104518954,
+      "grad_norm": 0.4545688033103943,
+      "learning_rate": 7.681088572455405e-05,
+      "loss": 0.9562,
+      "step": 3808
+    },
+    {
+      "epoch": 0.3403399825764514,
+      "grad_norm": 0.4550299644470215,
+      "learning_rate": 7.679867016324465e-05,
+      "loss": 0.9438,
+      "step": 3809
+    },
+    {
+      "epoch": 0.34042933410771326,
+      "grad_norm": 0.4195230305194855,
+      "learning_rate": 7.678645235726094e-05,
+      "loss": 0.9424,
+      "step": 3810
+    },
+    {
+      "epoch": 0.34051868563897514,
+      "grad_norm": 0.5254753232002258,
+      "learning_rate": 7.677423230762632e-05,
+      "loss": 0.9047,
+      "step": 3811
+    },
+    {
+      "epoch": 0.340608037170237,
+      "grad_norm": 0.5194946527481079,
+      "learning_rate": 7.676201001536439e-05,
+      "loss": 0.9891,
+      "step": 3812
+    },
+    {
+      "epoch": 0.34069738870149885,
+      "grad_norm": 0.4409390985965729,
+      "learning_rate": 7.674978548149882e-05,
+      "loss": 1.0141,
+      "step": 3813
+    },
+    {
+      "epoch": 0.34078674023276073,
+      "grad_norm": 0.4274226725101471,
+      "learning_rate": 7.67375587070536e-05,
+      "loss": 1.0241,
+      "step": 3814
+    },
+    {
+      "epoch": 0.3408760917640226,
+      "grad_norm": 0.41632789373397827,
+      "learning_rate": 7.672532969305284e-05,
+      "loss": 1.006,
+      "step": 3815
+    },
+    {
+      "epoch": 0.3409654432952845,
+      "grad_norm": 0.41912585496902466,
+      "learning_rate": 7.671309844052084e-05,
+      "loss": 1.036,
+      "step": 3816
+    },
+    {
+      "epoch": 0.3410547948265463,
+      "grad_norm": 0.44865861535072327,
+      "learning_rate": 7.67008649504821e-05,
+      "loss": 0.9867,
+      "step": 3817
+    },
+    {
+      "epoch": 0.3411441463578082,
+      "grad_norm": 0.4446162283420563,
+      "learning_rate": 7.668862922396131e-05,
+      "loss": 0.9555,
+      "step": 3818
+    },
+    {
+      "epoch": 0.3412334978890701,
+      "grad_norm": 0.468001127243042,
+      "learning_rate": 7.667639126198337e-05,
+      "loss": 1.0173,
+      "step": 3819
+    },
+    {
+      "epoch": 0.3413228494203319,
+      "grad_norm": 0.38588184118270874,
+      "learning_rate": 7.666415106557327e-05,
+      "loss": 1.0348,
+      "step": 3820
+    },
+    {
+      "epoch": 0.3414122009515938,
+      "grad_norm": 0.4341317415237427,
+      "learning_rate": 7.665190863575633e-05,
+      "loss": 1.0196,
+      "step": 3821
+    },
+    {
+      "epoch": 0.3415015524828557,
+      "grad_norm": 0.4301292598247528,
+      "learning_rate": 7.663966397355793e-05,
+      "loss": 0.9469,
+      "step": 3822
+    },
+    {
+      "epoch": 0.34159090401411757,
+      "grad_norm": 0.49630647897720337,
+      "learning_rate": 7.662741708000374e-05,
+      "loss": 0.9003,
+      "step": 3823
+    },
+    {
+      "epoch": 0.3416802555453794,
+      "grad_norm": 0.46921518445014954,
+      "learning_rate": 7.661516795611951e-05,
+      "loss": 0.9352,
+      "step": 3824
+    },
+    {
+      "epoch": 0.3417696070766413,
+      "grad_norm": 0.43845120072364807,
+      "learning_rate": 7.66029166029313e-05,
+      "loss": 0.9933,
+      "step": 3825
+    },
+    {
+      "epoch": 0.34185895860790316,
+      "grad_norm": 0.3999157249927521,
+      "learning_rate": 7.659066302146524e-05,
+      "loss": 1.0288,
+      "step": 3826
+    },
+    {
+      "epoch": 0.341948310139165,
+      "grad_norm": 0.38439029455184937,
+      "learning_rate": 7.657840721274772e-05,
+      "loss": 1.0599,
+      "step": 3827
+    },
+    {
+      "epoch": 0.3420376616704269,
+      "grad_norm": 0.4651052951812744,
+      "learning_rate": 7.656614917780527e-05,
+      "loss": 1.0204,
+      "step": 3828
+    },
+    {
+      "epoch": 0.34212701320168876,
+      "grad_norm": 0.41157326102256775,
+      "learning_rate": 7.655388891766468e-05,
+      "loss": 0.9886,
+      "step": 3829
+    },
+    {
+      "epoch": 0.3422163647329506,
+      "grad_norm": 0.4303523302078247,
+      "learning_rate": 7.654162643335283e-05,
+      "loss": 0.9888,
+      "step": 3830
+    },
+    {
+      "epoch": 0.34230571626421247,
+      "grad_norm": 0.3977055549621582,
+      "learning_rate": 7.652936172589686e-05,
+      "loss": 1.0203,
+      "step": 3831
+    },
+    {
+      "epoch": 0.34239506779547435,
+      "grad_norm": 0.44146913290023804,
+      "learning_rate": 7.651709479632406e-05,
+      "loss": 1.0111,
+      "step": 3832
+    },
+    {
+      "epoch": 0.34248441932673623,
+      "grad_norm": 0.532141923904419,
+      "learning_rate": 7.650482564566193e-05,
+      "loss": 0.9072,
+      "step": 3833
+    },
+    {
+      "epoch": 0.34257377085799806,
+      "grad_norm": 0.49662163853645325,
+      "learning_rate": 7.649255427493812e-05,
+      "loss": 1.0138,
+      "step": 3834
+    },
+    {
+      "epoch": 0.34266312238925994,
+      "grad_norm": 0.4919273555278778,
+      "learning_rate": 7.64802806851805e-05,
+      "loss": 0.9425,
+      "step": 3835
+    },
+    {
+      "epoch": 0.3427524739205218,
+      "grad_norm": 0.4210042655467987,
+      "learning_rate": 7.646800487741711e-05,
+      "loss": 0.9302,
+      "step": 3836
+    },
+    {
+      "epoch": 0.34284182545178365,
+      "grad_norm": 0.45006346702575684,
+      "learning_rate": 7.645572685267619e-05,
+      "loss": 0.9837,
+      "step": 3837
+    },
+    {
+      "epoch": 0.34293117698304554,
+      "grad_norm": 0.5554122924804688,
+      "learning_rate": 7.644344661198615e-05,
+      "loss": 0.9685,
+      "step": 3838
+    },
+    {
+      "epoch": 0.3430205285143074,
+      "grad_norm": 0.4323464035987854,
+      "learning_rate": 7.643116415637559e-05,
+      "loss": 1.0146,
+      "step": 3839
+    },
+    {
+      "epoch": 0.3431098800455693,
+      "grad_norm": 0.5363249182701111,
+      "learning_rate": 7.64188794868733e-05,
+      "loss": 0.9732,
+      "step": 3840
+    },
+    {
+      "epoch": 0.34319923157683113,
+      "grad_norm": 0.4187726378440857,
+      "learning_rate": 7.640659260450823e-05,
+      "loss": 0.9552,
+      "step": 3841
+    },
+    {
+      "epoch": 0.343288583108093,
+      "grad_norm": 0.5245155692100525,
+      "learning_rate": 7.639430351030958e-05,
+      "loss": 0.8949,
+      "step": 3842
+    },
+    {
+      "epoch": 0.3433779346393549,
+      "grad_norm": 0.4653611481189728,
+      "learning_rate": 7.638201220530665e-05,
+      "loss": 0.9063,
+      "step": 3843
+    },
+    {
+      "epoch": 0.3434672861706167,
+      "grad_norm": 0.42201900482177734,
+      "learning_rate": 7.636971869052899e-05,
+      "loss": 0.9947,
+      "step": 3844
+    },
+    {
+      "epoch": 0.3435566377018786,
+      "grad_norm": 0.5015077590942383,
+      "learning_rate": 7.63574229670063e-05,
+      "loss": 0.9122,
+      "step": 3845
+    },
+    {
+      "epoch": 0.3436459892331405,
+      "grad_norm": 0.40728330612182617,
+      "learning_rate": 7.63451250357685e-05,
+      "loss": 0.9819,
+      "step": 3846
+    },
+    {
+      "epoch": 0.3437353407644024,
+      "grad_norm": 0.508651852607727,
+      "learning_rate": 7.633282489784564e-05,
+      "loss": 0.9947,
+      "step": 3847
+    },
+    {
+      "epoch": 0.3438246922956642,
+      "grad_norm": 0.47697994112968445,
+      "learning_rate": 7.632052255426803e-05,
+      "loss": 0.9991,
+      "step": 3848
+    },
+    {
+      "epoch": 0.3439140438269261,
+      "grad_norm": 0.566102147102356,
+      "learning_rate": 7.63082180060661e-05,
+      "loss": 0.9842,
+      "step": 3849
+    },
+    {
+      "epoch": 0.34400339535818797,
+      "grad_norm": 0.4584532380104065,
+      "learning_rate": 7.629591125427047e-05,
+      "loss": 0.9289,
+      "step": 3850
+    },
+    {
+      "epoch": 0.3440927468894498,
+      "grad_norm": 0.42579421401023865,
+      "learning_rate": 7.628360229991199e-05,
+      "loss": 0.9618,
+      "step": 3851
+    },
+    {
+      "epoch": 0.3441820984207117,
+      "grad_norm": 0.4023285508155823,
+      "learning_rate": 7.627129114402164e-05,
+      "loss": 1.0386,
+      "step": 3852
+    },
+    {
+      "epoch": 0.34427144995197356,
+      "grad_norm": 0.5076940655708313,
+      "learning_rate": 7.625897778763062e-05,
+      "loss": 0.9791,
+      "step": 3853
+    },
+    {
+      "epoch": 0.34436080148323545,
+      "grad_norm": 0.4704766273498535,
+      "learning_rate": 7.624666223177033e-05,
+      "loss": 0.9195,
+      "step": 3854
+    },
+    {
+      "epoch": 0.3444501530144973,
+      "grad_norm": 0.4094794690608978,
+      "learning_rate": 7.62343444774723e-05,
+      "loss": 0.9469,
+      "step": 3855
+    },
+    {
+      "epoch": 0.34453950454575916,
+      "grad_norm": 0.41486209630966187,
+      "learning_rate": 7.62220245257683e-05,
+      "loss": 1.0922,
+      "step": 3856
+    },
+    {
+      "epoch": 0.34462885607702104,
+      "grad_norm": 0.45725998282432556,
+      "learning_rate": 7.620970237769022e-05,
+      "loss": 0.9506,
+      "step": 3857
+    },
+    {
+      "epoch": 0.34471820760828287,
+      "grad_norm": 0.416765034198761,
+      "learning_rate": 7.619737803427019e-05,
+      "loss": 0.9818,
+      "step": 3858
+    },
+    {
+      "epoch": 0.34480755913954475,
+      "grad_norm": 0.3775465190410614,
+      "learning_rate": 7.618505149654051e-05,
+      "loss": 0.9946,
+      "step": 3859
+    },
+    {
+      "epoch": 0.34489691067080663,
+      "grad_norm": 0.43782955408096313,
+      "learning_rate": 7.617272276553366e-05,
+      "loss": 0.983,
+      "step": 3860
+    },
+    {
+      "epoch": 0.34498626220206846,
+      "grad_norm": 0.40604960918426514,
+      "learning_rate": 7.61603918422823e-05,
+      "loss": 1.008,
+      "step": 3861
+    },
+    {
+      "epoch": 0.34507561373333034,
+      "grad_norm": 0.4615291357040405,
+      "learning_rate": 7.614805872781926e-05,
+      "loss": 1.0312,
+      "step": 3862
+    },
+    {
+      "epoch": 0.3451649652645922,
+      "grad_norm": 0.41947072744369507,
+      "learning_rate": 7.613572342317758e-05,
+      "loss": 0.9949,
+      "step": 3863
+    },
+    {
+      "epoch": 0.3452543167958541,
+      "grad_norm": 0.4784375727176666,
+      "learning_rate": 7.612338592939049e-05,
+      "loss": 0.9847,
+      "step": 3864
+    },
+    {
+      "epoch": 0.34534366832711594,
+      "grad_norm": 0.38383907079696655,
+      "learning_rate": 7.611104624749137e-05,
+      "loss": 1.0202,
+      "step": 3865
+    },
+    {
+      "epoch": 0.3454330198583778,
+      "grad_norm": 0.4683041572570801,
+      "learning_rate": 7.609870437851381e-05,
+      "loss": 1.0224,
+      "step": 3866
+    },
+    {
+      "epoch": 0.3455223713896397,
+      "grad_norm": 0.35911422967910767,
+      "learning_rate": 7.608636032349155e-05,
+      "loss": 1.0381,
+      "step": 3867
+    },
+    {
+      "epoch": 0.34561172292090153,
+      "grad_norm": 0.4072626829147339,
+      "learning_rate": 7.607401408345855e-05,
+      "loss": 1.119,
+      "step": 3868
+    },
+    {
+      "epoch": 0.3457010744521634,
+      "grad_norm": 0.4566546082496643,
+      "learning_rate": 7.606166565944895e-05,
+      "loss": 0.9207,
+      "step": 3869
+    },
+    {
+      "epoch": 0.3457904259834253,
+      "grad_norm": 0.41566282510757446,
+      "learning_rate": 7.604931505249706e-05,
+      "loss": 0.9994,
+      "step": 3870
+    },
+    {
+      "epoch": 0.3458797775146872,
+      "grad_norm": 0.46840545535087585,
+      "learning_rate": 7.603696226363737e-05,
+      "loss": 1.0202,
+      "step": 3871
+    },
+    {
+      "epoch": 0.345969129045949,
+      "grad_norm": 0.5090840458869934,
+      "learning_rate": 7.602460729390455e-05,
+      "loss": 0.8464,
+      "step": 3872
+    },
+    {
+      "epoch": 0.3460584805772109,
+      "grad_norm": 0.39953356981277466,
+      "learning_rate": 7.601225014433346e-05,
+      "loss": 1.0418,
+      "step": 3873
+    },
+    {
+      "epoch": 0.3461478321084728,
+      "grad_norm": 0.48468056321144104,
+      "learning_rate": 7.599989081595915e-05,
+      "loss": 0.9746,
+      "step": 3874
+    },
+    {
+      "epoch": 0.3462371836397346,
+      "grad_norm": 0.45679304003715515,
+      "learning_rate": 7.598752930981686e-05,
+      "loss": 0.9614,
+      "step": 3875
+    },
+    {
+      "epoch": 0.3463265351709965,
+      "grad_norm": 0.39945322275161743,
+      "learning_rate": 7.597516562694197e-05,
+      "loss": 1.0286,
+      "step": 3876
+    },
+    {
+      "epoch": 0.34641588670225837,
+      "grad_norm": 0.37428709864616394,
+      "learning_rate": 7.59627997683701e-05,
+      "loss": 1.0062,
+      "step": 3877
+    },
+    {
+      "epoch": 0.34650523823352025,
+      "grad_norm": 0.36209729313850403,
+      "learning_rate": 7.595043173513698e-05,
+      "loss": 1.0071,
+      "step": 3878
+    },
+    {
+      "epoch": 0.3465945897647821,
+      "grad_norm": 0.3938021659851074,
+      "learning_rate": 7.593806152827861e-05,
+      "loss": 0.9865,
+      "step": 3879
+    },
+    {
+      "epoch": 0.34668394129604396,
+      "grad_norm": 0.3960947096347809,
+      "learning_rate": 7.592568914883112e-05,
+      "loss": 0.9646,
+      "step": 3880
+    },
+    {
+      "epoch": 0.34677329282730585,
+      "grad_norm": 0.4837634563446045,
+      "learning_rate": 7.591331459783078e-05,
+      "loss": 0.9538,
+      "step": 3881
+    },
+    {
+      "epoch": 0.3468626443585677,
+      "grad_norm": 0.44596779346466064,
+      "learning_rate": 7.590093787631414e-05,
+      "loss": 0.9657,
+      "step": 3882
+    },
+    {
+      "epoch": 0.34695199588982956,
+      "grad_norm": 0.39485910534858704,
+      "learning_rate": 7.588855898531787e-05,
+      "loss": 1.047,
+      "step": 3883
+    },
+    {
+      "epoch": 0.34704134742109144,
+      "grad_norm": 0.4514664113521576,
+      "learning_rate": 7.587617792587884e-05,
+      "loss": 1.0095,
+      "step": 3884
+    },
+    {
+      "epoch": 0.3471306989523533,
+      "grad_norm": 0.46183550357818604,
+      "learning_rate": 7.586379469903408e-05,
+      "loss": 0.8961,
+      "step": 3885
+    },
+    {
+      "epoch": 0.34722005048361515,
+      "grad_norm": 0.4102190136909485,
+      "learning_rate": 7.585140930582085e-05,
+      "loss": 0.8931,
+      "step": 3886
+    },
+    {
+      "epoch": 0.34730940201487703,
+      "grad_norm": 0.42231065034866333,
+      "learning_rate": 7.583902174727651e-05,
+      "loss": 0.9373,
+      "step": 3887
+    },
+    {
+      "epoch": 0.3473987535461389,
+      "grad_norm": 0.39157602190971375,
+      "learning_rate": 7.582663202443867e-05,
+      "loss": 1.0546,
+      "step": 3888
+    },
+    {
+      "epoch": 0.34748810507740074,
+      "grad_norm": 0.46377474069595337,
+      "learning_rate": 7.581424013834511e-05,
+      "loss": 0.928,
+      "step": 3889
+    },
+    {
+      "epoch": 0.3475774566086626,
+      "grad_norm": 0.3647017776966095,
+      "learning_rate": 7.580184609003378e-05,
+      "loss": 1.0289,
+      "step": 3890
+    },
+    {
+      "epoch": 0.3476668081399245,
+      "grad_norm": 0.46764108538627625,
+      "learning_rate": 7.578944988054281e-05,
+      "loss": 0.9646,
+      "step": 3891
+    },
+    {
+      "epoch": 0.34775615967118634,
+      "grad_norm": 0.37693339586257935,
+      "learning_rate": 7.577705151091053e-05,
+      "loss": 1.0112,
+      "step": 3892
+    },
+    {
+      "epoch": 0.3478455112024482,
+      "grad_norm": 0.39555221796035767,
+      "learning_rate": 7.576465098217542e-05,
+      "loss": 0.9854,
+      "step": 3893
+    },
+    {
+      "epoch": 0.3479348627337101,
+      "grad_norm": 0.5179402828216553,
+      "learning_rate": 7.575224829537615e-05,
+      "loss": 0.9444,
+      "step": 3894
+    },
+    {
+      "epoch": 0.348024214264972,
+      "grad_norm": 0.42474082112312317,
+      "learning_rate": 7.573984345155159e-05,
+      "loss": 1.0315,
+      "step": 3895
+    },
+    {
+      "epoch": 0.3481135657962338,
+      "grad_norm": 0.46355873346328735,
+      "learning_rate": 7.572743645174077e-05,
+      "loss": 0.9543,
+      "step": 3896
+    },
+    {
+      "epoch": 0.3482029173274957,
+      "grad_norm": 0.4335034489631653,
+      "learning_rate": 7.571502729698293e-05,
+      "loss": 0.9763,
+      "step": 3897
+    },
+    {
+      "epoch": 0.3482922688587576,
+      "grad_norm": 0.4546475112438202,
+      "learning_rate": 7.570261598831743e-05,
+      "loss": 0.9754,
+      "step": 3898
+    },
+    {
+      "epoch": 0.3483816203900194,
+      "grad_norm": 0.4445870518684387,
+      "learning_rate": 7.569020252678387e-05,
+      "loss": 0.9814,
+      "step": 3899
+    },
+    {
+      "epoch": 0.3484709719212813,
+      "grad_norm": 0.39527517557144165,
+      "learning_rate": 7.567778691342203e-05,
+      "loss": 1.0592,
+      "step": 3900
+    },
+    {
+      "epoch": 0.3485603234525432,
+      "grad_norm": 0.442655473947525,
+      "learning_rate": 7.566536914927181e-05,
+      "loss": 0.9154,
+      "step": 3901
+    },
+    {
+      "epoch": 0.34864967498380506,
+      "grad_norm": 0.40641334652900696,
+      "learning_rate": 7.565294923537336e-05,
+      "loss": 0.9864,
+      "step": 3902
+    },
+    {
+      "epoch": 0.3487390265150669,
+      "grad_norm": 0.42094290256500244,
+      "learning_rate": 7.564052717276696e-05,
+      "loss": 0.9796,
+      "step": 3903
+    },
+    {
+      "epoch": 0.34882837804632877,
+      "grad_norm": 0.4103735089302063,
+      "learning_rate": 7.56281029624931e-05,
+      "loss": 1.0549,
+      "step": 3904
+    },
+    {
+      "epoch": 0.34891772957759065,
+      "grad_norm": 0.5410390496253967,
+      "learning_rate": 7.561567660559246e-05,
+      "loss": 0.9967,
+      "step": 3905
+    },
+    {
+      "epoch": 0.3490070811088525,
+      "grad_norm": 0.4145830273628235,
+      "learning_rate": 7.560324810310586e-05,
+      "loss": 0.9267,
+      "step": 3906
+    },
+    {
+      "epoch": 0.34909643264011436,
+      "grad_norm": 0.47264495491981506,
+      "learning_rate": 7.559081745607431e-05,
+      "loss": 0.9579,
+      "step": 3907
+    },
+    {
+      "epoch": 0.34918578417137625,
+      "grad_norm": 0.5321443676948547,
+      "learning_rate": 7.557838466553902e-05,
+      "loss": 0.9407,
+      "step": 3908
+    },
+    {
+      "epoch": 0.34927513570263813,
+      "grad_norm": 0.465622216463089,
+      "learning_rate": 7.556594973254136e-05,
+      "loss": 0.9588,
+      "step": 3909
+    },
+    {
+      "epoch": 0.34936448723389996,
+      "grad_norm": 0.463840514421463,
+      "learning_rate": 7.555351265812292e-05,
+      "loss": 0.9034,
+      "step": 3910
+    },
+    {
+      "epoch": 0.34945383876516184,
+      "grad_norm": 0.4932054281234741,
+      "learning_rate": 7.55410734433254e-05,
+      "loss": 0.9503,
+      "step": 3911
+    },
+    {
+      "epoch": 0.3495431902964237,
+      "grad_norm": 0.3836749494075775,
+      "learning_rate": 7.552863208919073e-05,
+      "loss": 0.9432,
+      "step": 3912
+    },
+    {
+      "epoch": 0.34963254182768555,
+      "grad_norm": 0.43824368715286255,
+      "learning_rate": 7.551618859676101e-05,
+      "loss": 1.047,
+      "step": 3913
+    },
+    {
+      "epoch": 0.34972189335894743,
+      "grad_norm": 0.5028628706932068,
+      "learning_rate": 7.550374296707851e-05,
+      "loss": 0.9571,
+      "step": 3914
+    },
+    {
+      "epoch": 0.3498112448902093,
+      "grad_norm": 0.4159904718399048,
+      "learning_rate": 7.54912952011857e-05,
+      "loss": 0.9978,
+      "step": 3915
+    },
+    {
+      "epoch": 0.3499005964214712,
+      "grad_norm": 0.42895132303237915,
+      "learning_rate": 7.547884530012517e-05,
+      "loss": 0.9415,
+      "step": 3916
+    },
+    {
+      "epoch": 0.349989947952733,
+      "grad_norm": 0.4712909162044525,
+      "learning_rate": 7.546639326493978e-05,
+      "loss": 0.971,
+      "step": 3917
+    },
+    {
+      "epoch": 0.3500792994839949,
+      "grad_norm": 0.41875845193862915,
+      "learning_rate": 7.545393909667249e-05,
+      "loss": 1.0202,
+      "step": 3918
+    },
+    {
+      "epoch": 0.3501686510152568,
+      "grad_norm": 0.32883596420288086,
+      "learning_rate": 7.54414827963665e-05,
+      "loss": 1.0405,
+      "step": 3919
+    },
+    {
+      "epoch": 0.3502580025465186,
+      "grad_norm": 0.5156275033950806,
+      "learning_rate": 7.542902436506514e-05,
+      "loss": 0.8699,
+      "step": 3920
+    },
+    {
+      "epoch": 0.3503473540777805,
+      "grad_norm": 0.3887585997581482,
+      "learning_rate": 7.541656380381192e-05,
+      "loss": 0.9707,
+      "step": 3921
+    },
+    {
+      "epoch": 0.3504367056090424,
+      "grad_norm": 0.4152442216873169,
+      "learning_rate": 7.540410111365055e-05,
+      "loss": 0.9871,
+      "step": 3922
+    },
+    {
+      "epoch": 0.3505260571403042,
+      "grad_norm": 0.4276559054851532,
+      "learning_rate": 7.539163629562494e-05,
+      "loss": 0.9426,
+      "step": 3923
+    },
+    {
+      "epoch": 0.3506154086715661,
+      "grad_norm": 0.5117713212966919,
+      "learning_rate": 7.537916935077914e-05,
+      "loss": 0.9238,
+      "step": 3924
+    },
+    {
+      "epoch": 0.350704760202828,
+      "grad_norm": 0.4770560562610626,
+      "learning_rate": 7.536670028015737e-05,
+      "loss": 0.9661,
+      "step": 3925
+    },
+    {
+      "epoch": 0.35079411173408986,
+      "grad_norm": 0.49051031470298767,
+      "learning_rate": 7.535422908480408e-05,
+      "loss": 1.0228,
+      "step": 3926
+    },
+    {
+      "epoch": 0.3508834632653517,
+      "grad_norm": 0.3898465037345886,
+      "learning_rate": 7.534175576576384e-05,
+      "loss": 1.01,
+      "step": 3927
+    },
+    {
+      "epoch": 0.3509728147966136,
+      "grad_norm": 0.48158106207847595,
+      "learning_rate": 7.532928032408142e-05,
+      "loss": 1.0054,
+      "step": 3928
+    },
+    {
+      "epoch": 0.35106216632787546,
+      "grad_norm": 0.43918222188949585,
+      "learning_rate": 7.53168027608018e-05,
+      "loss": 0.9429,
+      "step": 3929
+    },
+    {
+      "epoch": 0.3511515178591373,
+      "grad_norm": 0.4521641433238983,
+      "learning_rate": 7.530432307697007e-05,
+      "loss": 0.8965,
+      "step": 3930
+    },
+    {
+      "epoch": 0.35124086939039917,
+      "grad_norm": 0.4660329520702362,
+      "learning_rate": 7.529184127363158e-05,
+      "loss": 0.9271,
+      "step": 3931
+    },
+    {
+      "epoch": 0.35133022092166105,
+      "grad_norm": 0.4383411705493927,
+      "learning_rate": 7.527935735183177e-05,
+      "loss": 1.0393,
+      "step": 3932
+    },
+    {
+      "epoch": 0.35141957245292293,
+      "grad_norm": 0.36236387491226196,
+      "learning_rate": 7.526687131261634e-05,
+      "loss": 1.0835,
+      "step": 3933
+    },
+    {
+      "epoch": 0.35150892398418476,
+      "grad_norm": 0.3911649286746979,
+      "learning_rate": 7.52543831570311e-05,
+      "loss": 1.0008,
+      "step": 3934
+    },
+    {
+      "epoch": 0.35159827551544665,
+      "grad_norm": 0.43021535873413086,
+      "learning_rate": 7.524189288612209e-05,
+      "loss": 1.0463,
+      "step": 3935
+    },
+    {
+      "epoch": 0.35168762704670853,
+      "grad_norm": 0.5328584909439087,
+      "learning_rate": 7.522940050093547e-05,
+      "loss": 1.0456,
+      "step": 3936
+    },
+    {
+      "epoch": 0.35177697857797036,
+      "grad_norm": 0.41291365027427673,
+      "learning_rate": 7.521690600251766e-05,
+      "loss": 0.9843,
+      "step": 3937
+    },
+    {
+      "epoch": 0.35186633010923224,
+      "grad_norm": 0.44431254267692566,
+      "learning_rate": 7.520440939191515e-05,
+      "loss": 0.9654,
+      "step": 3938
+    },
+    {
+      "epoch": 0.3519556816404941,
+      "grad_norm": 0.40238481760025024,
+      "learning_rate": 7.519191067017472e-05,
+      "loss": 1.0108,
+      "step": 3939
+    },
+    {
+      "epoch": 0.352045033171756,
+      "grad_norm": 0.41174423694610596,
+      "learning_rate": 7.517940983834323e-05,
+      "loss": 1.0082,
+      "step": 3940
+    },
+    {
+      "epoch": 0.35213438470301783,
+      "grad_norm": 0.44105100631713867,
+      "learning_rate": 7.516690689746779e-05,
+      "loss": 0.9419,
+      "step": 3941
+    },
+    {
+      "epoch": 0.3522237362342797,
+      "grad_norm": 0.4259895980358124,
+      "learning_rate": 7.515440184859561e-05,
+      "loss": 1.0009,
+      "step": 3942
+    },
+    {
+      "epoch": 0.3523130877655416,
+      "grad_norm": 0.45526599884033203,
+      "learning_rate": 7.514189469277418e-05,
+      "loss": 1.019,
+      "step": 3943
+    },
+    {
+      "epoch": 0.3524024392968034,
+      "grad_norm": 0.4207375943660736,
+      "learning_rate": 7.512938543105105e-05,
+      "loss": 0.9647,
+      "step": 3944
+    },
+    {
+      "epoch": 0.3524917908280653,
+      "grad_norm": 0.4691575765609741,
+      "learning_rate": 7.511687406447406e-05,
+      "loss": 0.984,
+      "step": 3945
+    },
+    {
+      "epoch": 0.3525811423593272,
+      "grad_norm": 0.4423462152481079,
+      "learning_rate": 7.51043605940911e-05,
+      "loss": 0.9842,
+      "step": 3946
+    },
+    {
+      "epoch": 0.3526704938905891,
+      "grad_norm": 0.44663453102111816,
+      "learning_rate": 7.509184502095038e-05,
+      "loss": 0.9933,
+      "step": 3947
+    },
+    {
+      "epoch": 0.3527598454218509,
+      "grad_norm": 0.4556397795677185,
+      "learning_rate": 7.507932734610017e-05,
+      "loss": 0.9362,
+      "step": 3948
+    },
+    {
+      "epoch": 0.3528491969531128,
+      "grad_norm": 0.3916218876838684,
+      "learning_rate": 7.506680757058896e-05,
+      "loss": 1.0302,
+      "step": 3949
+    },
+    {
+      "epoch": 0.35293854848437467,
+      "grad_norm": 0.39909112453460693,
+      "learning_rate": 7.505428569546542e-05,
+      "loss": 1.0345,
+      "step": 3950
+    },
+    {
+      "epoch": 0.3530279000156365,
+      "grad_norm": 0.4075121283531189,
+      "learning_rate": 7.504176172177841e-05,
+      "loss": 0.9996,
+      "step": 3951
+    },
+    {
+      "epoch": 0.3531172515468984,
+      "grad_norm": 0.46926334500312805,
+      "learning_rate": 7.502923565057692e-05,
+      "loss": 0.9457,
+      "step": 3952
+    },
+    {
+      "epoch": 0.35320660307816026,
+      "grad_norm": 0.4146152436733246,
+      "learning_rate": 7.501670748291016e-05,
+      "loss": 1.0144,
+      "step": 3953
+    },
+    {
+      "epoch": 0.3532959546094221,
+      "grad_norm": 0.4583793878555298,
+      "learning_rate": 7.500417721982748e-05,
+      "loss": 0.9683,
+      "step": 3954
+    },
+    {
+      "epoch": 0.353385306140684,
+      "grad_norm": 0.42835697531700134,
+      "learning_rate": 7.499164486237844e-05,
+      "loss": 0.9714,
+      "step": 3955
+    },
+    {
+      "epoch": 0.35347465767194586,
+      "grad_norm": 0.4047943949699402,
+      "learning_rate": 7.497911041161274e-05,
+      "loss": 1.0431,
+      "step": 3956
+    },
+    {
+      "epoch": 0.35356400920320774,
+      "grad_norm": 0.3960307836532593,
+      "learning_rate": 7.496657386858029e-05,
+      "loss": 0.9681,
+      "step": 3957
+    },
+    {
+      "epoch": 0.35365336073446957,
+      "grad_norm": 0.37818729877471924,
+      "learning_rate": 7.495403523433116e-05,
+      "loss": 1.0108,
+      "step": 3958
+    },
+    {
+      "epoch": 0.35374271226573145,
+      "grad_norm": 0.44863516092300415,
+      "learning_rate": 7.494149450991557e-05,
+      "loss": 0.9547,
+      "step": 3959
+    },
+    {
+      "epoch": 0.35383206379699333,
+      "grad_norm": 0.38477665185928345,
+      "learning_rate": 7.492895169638397e-05,
+      "loss": 1.0091,
+      "step": 3960
+    },
+    {
+      "epoch": 0.35392141532825516,
+      "grad_norm": 0.45500656962394714,
+      "learning_rate": 7.491640679478696e-05,
+      "loss": 0.9557,
+      "step": 3961
+    },
+    {
+      "epoch": 0.35401076685951705,
+      "grad_norm": 0.3944813013076782,
+      "learning_rate": 7.490385980617527e-05,
+      "loss": 1.0622,
+      "step": 3962
+    },
+    {
+      "epoch": 0.35410011839077893,
+      "grad_norm": 0.3850362300872803,
+      "learning_rate": 7.489131073159987e-05,
+      "loss": 0.96,
+      "step": 3963
+    },
+    {
+      "epoch": 0.3541894699220408,
+      "grad_norm": 0.402127742767334,
+      "learning_rate": 7.487875957211188e-05,
+      "loss": 0.8966,
+      "step": 3964
+    },
+    {
+      "epoch": 0.35427882145330264,
+      "grad_norm": 0.39359238743782043,
+      "learning_rate": 7.486620632876257e-05,
+      "loss": 1.0262,
+      "step": 3965
+    },
+    {
+      "epoch": 0.3543681729845645,
+      "grad_norm": 0.4493246078491211,
+      "learning_rate": 7.485365100260345e-05,
+      "loss": 0.9622,
+      "step": 3966
+    },
+    {
+      "epoch": 0.3544575245158264,
+      "grad_norm": 0.38160157203674316,
+      "learning_rate": 7.484109359468612e-05,
+      "loss": 0.9997,
+      "step": 3967
+    },
+    {
+      "epoch": 0.35454687604708823,
+      "grad_norm": 0.5229949355125427,
+      "learning_rate": 7.482853410606242e-05,
+      "loss": 0.8508,
+      "step": 3968
+    },
+    {
+      "epoch": 0.3546362275783501,
+      "grad_norm": 0.41521862149238586,
+      "learning_rate": 7.481597253778434e-05,
+      "loss": 1.0109,
+      "step": 3969
+    },
+    {
+      "epoch": 0.354725579109612,
+      "grad_norm": 0.4313722252845764,
+      "learning_rate": 7.480340889090403e-05,
+      "loss": 0.9793,
+      "step": 3970
+    },
+    {
+      "epoch": 0.3548149306408739,
+      "grad_norm": 0.47400006651878357,
+      "learning_rate": 7.479084316647385e-05,
+      "loss": 1.0458,
+      "step": 3971
+    },
+    {
+      "epoch": 0.3549042821721357,
+      "grad_norm": 0.44894692301750183,
+      "learning_rate": 7.477827536554629e-05,
+      "loss": 0.9617,
+      "step": 3972
+    },
+    {
+      "epoch": 0.3549936337033976,
+      "grad_norm": 0.40421560406684875,
+      "learning_rate": 7.476570548917406e-05,
+      "loss": 1.0115,
+      "step": 3973
+    },
+    {
+      "epoch": 0.3550829852346595,
+      "grad_norm": 0.4186607897281647,
+      "learning_rate": 7.475313353841e-05,
+      "loss": 1.0278,
+      "step": 3974
+    },
+    {
+      "epoch": 0.3551723367659213,
+      "grad_norm": 0.36971282958984375,
+      "learning_rate": 7.474055951430717e-05,
+      "loss": 0.9399,
+      "step": 3975
+    },
+    {
+      "epoch": 0.3552616882971832,
+      "grad_norm": 0.4785158634185791,
+      "learning_rate": 7.472798341791877e-05,
+      "loss": 0.9303,
+      "step": 3976
+    },
+    {
+      "epoch": 0.35535103982844507,
+      "grad_norm": 0.38244813680648804,
+      "learning_rate": 7.471540525029817e-05,
+      "loss": 0.9782,
+      "step": 3977
+    },
+    {
+      "epoch": 0.35544039135970695,
+      "grad_norm": 0.5407021641731262,
+      "learning_rate": 7.470282501249893e-05,
+      "loss": 0.948,
+      "step": 3978
+    },
+    {
+      "epoch": 0.3555297428909688,
+      "grad_norm": 0.4248984158039093,
+      "learning_rate": 7.469024270557477e-05,
+      "loss": 1.0419,
+      "step": 3979
+    },
+    {
+      "epoch": 0.35561909442223066,
+      "grad_norm": 0.38939395546913147,
+      "learning_rate": 7.467765833057964e-05,
+      "loss": 1.0081,
+      "step": 3980
+    },
+    {
+      "epoch": 0.35570844595349255,
+      "grad_norm": 0.45824161171913147,
+      "learning_rate": 7.466507188856755e-05,
+      "loss": 0.9602,
+      "step": 3981
+    },
+    {
+      "epoch": 0.3557977974847544,
+      "grad_norm": 0.45186036825180054,
+      "learning_rate": 7.46524833805928e-05,
+      "loss": 0.9459,
+      "step": 3982
+    },
+    {
+      "epoch": 0.35588714901601626,
+      "grad_norm": 0.4205573499202728,
+      "learning_rate": 7.463989280770978e-05,
+      "loss": 1.0347,
+      "step": 3983
+    },
+    {
+      "epoch": 0.35597650054727814,
+      "grad_norm": 0.5025462508201599,
+      "learning_rate": 7.462730017097308e-05,
+      "loss": 0.9495,
+      "step": 3984
+    },
+    {
+      "epoch": 0.35606585207853997,
+      "grad_norm": 0.45681893825531006,
+      "learning_rate": 7.46147054714375e-05,
+      "loss": 0.9172,
+      "step": 3985
+    },
+    {
+      "epoch": 0.35615520360980185,
+      "grad_norm": 0.43965548276901245,
+      "learning_rate": 7.460210871015796e-05,
+      "loss": 1.0745,
+      "step": 3986
+    },
+    {
+      "epoch": 0.35624455514106373,
+      "grad_norm": 0.4385167062282562,
+      "learning_rate": 7.458950988818957e-05,
+      "loss": 1.0192,
+      "step": 3987
+    },
+    {
+      "epoch": 0.3563339066723256,
+      "grad_norm": 0.4278493821620941,
+      "learning_rate": 7.457690900658762e-05,
+      "loss": 0.9996,
+      "step": 3988
+    },
+    {
+      "epoch": 0.35642325820358745,
+      "grad_norm": 0.45512884855270386,
+      "learning_rate": 7.456430606640757e-05,
+      "loss": 0.9825,
+      "step": 3989
+    },
+    {
+      "epoch": 0.35651260973484933,
+      "grad_norm": 0.4201413094997406,
+      "learning_rate": 7.455170106870505e-05,
+      "loss": 0.949,
+      "step": 3990
+    },
+    {
+      "epoch": 0.3566019612661112,
+      "grad_norm": 0.39953503012657166,
+      "learning_rate": 7.453909401453589e-05,
+      "loss": 0.9882,
+      "step": 3991
+    },
+    {
+      "epoch": 0.35669131279737304,
+      "grad_norm": 0.4012105464935303,
+      "learning_rate": 7.452648490495602e-05,
+      "loss": 0.9762,
+      "step": 3992
+    },
+    {
+      "epoch": 0.3567806643286349,
+      "grad_norm": 0.39870861172676086,
+      "learning_rate": 7.451387374102159e-05,
+      "loss": 0.9714,
+      "step": 3993
+    },
+    {
+      "epoch": 0.3568700158598968,
+      "grad_norm": 0.3774680197238922,
+      "learning_rate": 7.450126052378894e-05,
+      "loss": 0.9913,
+      "step": 3994
+    },
+    {
+      "epoch": 0.3569593673911587,
+      "grad_norm": 0.4141833186149597,
+      "learning_rate": 7.448864525431457e-05,
+      "loss": 0.9495,
+      "step": 3995
+    },
+    {
+      "epoch": 0.3570487189224205,
+      "grad_norm": 0.44811514019966125,
+      "learning_rate": 7.447602793365514e-05,
+      "loss": 1.0848,
+      "step": 3996
+    },
+    {
+      "epoch": 0.3571380704536824,
+      "grad_norm": 0.44921422004699707,
+      "learning_rate": 7.446340856286744e-05,
+      "loss": 1.0281,
+      "step": 3997
+    },
+    {
+      "epoch": 0.3572274219849443,
+      "grad_norm": 0.4648860991001129,
+      "learning_rate": 7.445078714300855e-05,
+      "loss": 1.0197,
+      "step": 3998
+    },
+    {
+      "epoch": 0.3573167735162061,
+      "grad_norm": 0.4424198865890503,
+      "learning_rate": 7.443816367513559e-05,
+      "loss": 1.0069,
+      "step": 3999
+    },
+    {
+      "epoch": 0.357406125047468,
+      "grad_norm": 0.5149994492530823,
+      "learning_rate": 7.442553816030592e-05,
+      "loss": 0.9001,
+      "step": 4000
+    },
+    {
+      "epoch": 0.3574954765787299,
+      "grad_norm": 0.3967181444168091,
+      "learning_rate": 7.441291059957708e-05,
+      "loss": 1.0558,
+      "step": 4001
+    },
+    {
+      "epoch": 0.35758482810999176,
+      "grad_norm": 0.4318191707134247,
+      "learning_rate": 7.440028099400677e-05,
+      "loss": 0.9383,
+      "step": 4002
+    },
+    {
+      "epoch": 0.3576741796412536,
+      "grad_norm": 0.4305358827114105,
+      "learning_rate": 7.438764934465283e-05,
+      "loss": 1.0499,
+      "step": 4003
+    },
+    {
+      "epoch": 0.35776353117251547,
+      "grad_norm": 0.38480517268180847,
+      "learning_rate": 7.437501565257329e-05,
+      "loss": 0.9881,
+      "step": 4004
+    },
+    {
+      "epoch": 0.35785288270377735,
+      "grad_norm": 0.5233545303344727,
+      "learning_rate": 7.436237991882637e-05,
+      "loss": 0.969,
+      "step": 4005
+    },
+    {
+      "epoch": 0.3579422342350392,
+      "grad_norm": 0.5495209693908691,
+      "learning_rate": 7.434974214447047e-05,
+      "loss": 0.9978,
+      "step": 4006
+    },
+    {
+      "epoch": 0.35803158576630106,
+      "grad_norm": 0.46904364228248596,
+      "learning_rate": 7.43371023305641e-05,
+      "loss": 0.9741,
+      "step": 4007
+    },
+    {
+      "epoch": 0.35812093729756295,
+      "grad_norm": 0.4199860990047455,
+      "learning_rate": 7.432446047816599e-05,
+      "loss": 0.9811,
+      "step": 4008
+    },
+    {
+      "epoch": 0.35821028882882483,
+      "grad_norm": 0.3669317364692688,
+      "learning_rate": 7.431181658833504e-05,
+      "loss": 1.07,
+      "step": 4009
+    },
+    {
+      "epoch": 0.35829964036008666,
+      "grad_norm": 0.41901111602783203,
+      "learning_rate": 7.42991706621303e-05,
+      "loss": 0.896,
+      "step": 4010
+    },
+    {
+      "epoch": 0.35838899189134854,
+      "grad_norm": 0.47777092456817627,
+      "learning_rate": 7.428652270061102e-05,
+      "loss": 1.0107,
+      "step": 4011
+    },
+    {
+      "epoch": 0.3584783434226104,
+      "grad_norm": 0.4377864897251129,
+      "learning_rate": 7.427387270483659e-05,
+      "loss": 0.9397,
+      "step": 4012
+    },
+    {
+      "epoch": 0.35856769495387225,
+      "grad_norm": 0.45756855607032776,
+      "learning_rate": 7.426122067586656e-05,
+      "loss": 0.9688,
+      "step": 4013
+    },
+    {
+      "epoch": 0.35865704648513413,
+      "grad_norm": 0.4156295359134674,
+      "learning_rate": 7.424856661476071e-05,
+      "loss": 1.0963,
+      "step": 4014
+    },
+    {
+      "epoch": 0.358746398016396,
+      "grad_norm": 0.43841129541397095,
+      "learning_rate": 7.423591052257893e-05,
+      "loss": 0.9589,
+      "step": 4015
+    },
+    {
+      "epoch": 0.3588357495476579,
+      "grad_norm": 0.35517293214797974,
+      "learning_rate": 7.42232524003813e-05,
+      "loss": 1.0089,
+      "step": 4016
+    },
+    {
+      "epoch": 0.35892510107891973,
+      "grad_norm": 0.5077205300331116,
+      "learning_rate": 7.42105922492281e-05,
+      "loss": 0.9804,
+      "step": 4017
+    },
+    {
+      "epoch": 0.3590144526101816,
+      "grad_norm": 0.480881929397583,
+      "learning_rate": 7.419793007017972e-05,
+      "loss": 1.0214,
+      "step": 4018
+    },
+    {
+      "epoch": 0.3591038041414435,
+      "grad_norm": 0.4301677942276001,
+      "learning_rate": 7.418526586429676e-05,
+      "loss": 1.0346,
+      "step": 4019
+    },
+    {
+      "epoch": 0.3591931556727053,
+      "grad_norm": 0.4578787684440613,
+      "learning_rate": 7.417259963263999e-05,
+      "loss": 0.9741,
+      "step": 4020
+    },
+    {
+      "epoch": 0.3592825072039672,
+      "grad_norm": 0.5144799947738647,
+      "learning_rate": 7.415993137627036e-05,
+      "loss": 0.9486,
+      "step": 4021
+    },
+    {
+      "epoch": 0.3593718587352291,
+      "grad_norm": 0.37972408533096313,
+      "learning_rate": 7.414726109624892e-05,
+      "loss": 1.0567,
+      "step": 4022
+    },
+    {
+      "epoch": 0.3594612102664909,
+      "grad_norm": 0.40619876980781555,
+      "learning_rate": 7.413458879363698e-05,
+      "loss": 0.9922,
+      "step": 4023
+    },
+    {
+      "epoch": 0.3595505617977528,
+      "grad_norm": 0.5421142578125,
+      "learning_rate": 7.412191446949598e-05,
+      "loss": 0.9754,
+      "step": 4024
+    },
+    {
+      "epoch": 0.3596399133290147,
+      "grad_norm": 0.46210673451423645,
+      "learning_rate": 7.410923812488752e-05,
+      "loss": 1.0145,
+      "step": 4025
+    },
+    {
+      "epoch": 0.35972926486027657,
+      "grad_norm": 0.45515942573547363,
+      "learning_rate": 7.409655976087339e-05,
+      "loss": 0.9441,
+      "step": 4026
+    },
+    {
+      "epoch": 0.3598186163915384,
+      "grad_norm": 0.40571707487106323,
+      "learning_rate": 7.408387937851551e-05,
+      "loss": 0.9343,
+      "step": 4027
+    },
+    {
+      "epoch": 0.3599079679228003,
+      "grad_norm": 0.5472614765167236,
+      "learning_rate": 7.407119697887602e-05,
+      "loss": 0.964,
+      "step": 4028
+    },
+    {
+      "epoch": 0.35999731945406216,
+      "grad_norm": 0.4565272629261017,
+      "learning_rate": 7.405851256301722e-05,
+      "loss": 0.9862,
+      "step": 4029
+    },
+    {
+      "epoch": 0.360086670985324,
+      "grad_norm": 0.5199412703514099,
+      "learning_rate": 7.404582613200153e-05,
+      "loss": 0.9162,
+      "step": 4030
+    },
+    {
+      "epoch": 0.36017602251658587,
+      "grad_norm": 0.38596537709236145,
+      "learning_rate": 7.403313768689159e-05,
+      "loss": 1.0374,
+      "step": 4031
+    },
+    {
+      "epoch": 0.36026537404784775,
+      "grad_norm": 0.4285348653793335,
+      "learning_rate": 7.402044722875021e-05,
+      "loss": 0.9533,
+      "step": 4032
+    },
+    {
+      "epoch": 0.36035472557910964,
+      "grad_norm": 0.4435160756111145,
+      "learning_rate": 7.400775475864032e-05,
+      "loss": 0.9475,
+      "step": 4033
+    },
+    {
+      "epoch": 0.36044407711037146,
+      "grad_norm": 0.40904703736305237,
+      "learning_rate": 7.399506027762507e-05,
+      "loss": 0.9486,
+      "step": 4034
+    },
+    {
+      "epoch": 0.36053342864163335,
+      "grad_norm": 0.42833709716796875,
+      "learning_rate": 7.398236378676776e-05,
+      "loss": 0.9862,
+      "step": 4035
+    },
+    {
+      "epoch": 0.36062278017289523,
+      "grad_norm": 0.44428741931915283,
+      "learning_rate": 7.396966528713184e-05,
+      "loss": 0.9491,
+      "step": 4036
+    },
+    {
+      "epoch": 0.36071213170415706,
+      "grad_norm": 0.37404051423072815,
+      "learning_rate": 7.395696477978096e-05,
+      "loss": 1.0652,
+      "step": 4037
+    },
+    {
+      "epoch": 0.36080148323541894,
+      "grad_norm": 0.5555472373962402,
+      "learning_rate": 7.394426226577891e-05,
+      "loss": 0.8922,
+      "step": 4038
+    },
+    {
+      "epoch": 0.3608908347666808,
+      "grad_norm": 0.40428173542022705,
+      "learning_rate": 7.393155774618967e-05,
+      "loss": 0.9652,
+      "step": 4039
+    },
+    {
+      "epoch": 0.3609801862979427,
+      "grad_norm": 0.5697705149650574,
+      "learning_rate": 7.391885122207738e-05,
+      "loss": 0.9272,
+      "step": 4040
+    },
+    {
+      "epoch": 0.36106953782920453,
+      "grad_norm": 0.38609790802001953,
+      "learning_rate": 7.390614269450634e-05,
+      "loss": 0.9702,
+      "step": 4041
+    },
+    {
+      "epoch": 0.3611588893604664,
+      "grad_norm": 0.401506245136261,
+      "learning_rate": 7.389343216454103e-05,
+      "loss": 1.0075,
+      "step": 4042
+    },
+    {
+      "epoch": 0.3612482408917283,
+      "grad_norm": 0.43169543147087097,
+      "learning_rate": 7.388071963324609e-05,
+      "loss": 0.9761,
+      "step": 4043
+    },
+    {
+      "epoch": 0.36133759242299013,
+      "grad_norm": 0.4852740168571472,
+      "learning_rate": 7.386800510168632e-05,
+      "loss": 0.9143,
+      "step": 4044
+    },
+    {
+      "epoch": 0.361426943954252,
+      "grad_norm": 0.4311829209327698,
+      "learning_rate": 7.385528857092672e-05,
+      "loss": 0.9728,
+      "step": 4045
+    },
+    {
+      "epoch": 0.3615162954855139,
+      "grad_norm": 0.4241611063480377,
+      "learning_rate": 7.384257004203242e-05,
+      "loss": 0.9631,
+      "step": 4046
+    },
+    {
+      "epoch": 0.3616056470167758,
+      "grad_norm": 0.435428649187088,
+      "learning_rate": 7.382984951606875e-05,
+      "loss": 0.9813,
+      "step": 4047
+    },
+    {
+      "epoch": 0.3616949985480376,
+      "grad_norm": 0.4283429980278015,
+      "learning_rate": 7.381712699410116e-05,
+      "loss": 0.9696,
+      "step": 4048
+    },
+    {
+      "epoch": 0.3617843500792995,
+      "grad_norm": 0.43169790506362915,
+      "learning_rate": 7.380440247719532e-05,
+      "loss": 0.968,
+      "step": 4049
+    },
+    {
+      "epoch": 0.36187370161056137,
+      "grad_norm": 0.44879093766212463,
+      "learning_rate": 7.379167596641702e-05,
+      "loss": 0.9826,
+      "step": 4050
+    },
+    {
+      "epoch": 0.3619630531418232,
+      "grad_norm": 0.3845369219779968,
+      "learning_rate": 7.377894746283227e-05,
+      "loss": 0.9373,
+      "step": 4051
+    },
+    {
+      "epoch": 0.3620524046730851,
+      "grad_norm": 0.42470452189445496,
+      "learning_rate": 7.37662169675072e-05,
+      "loss": 1.0286,
+      "step": 4052
+    },
+    {
+      "epoch": 0.36214175620434697,
+      "grad_norm": 0.5062727332115173,
+      "learning_rate": 7.375348448150814e-05,
+      "loss": 0.9409,
+      "step": 4053
+    },
+    {
+      "epoch": 0.3622311077356088,
+      "grad_norm": 0.41713082790374756,
+      "learning_rate": 7.374075000590155e-05,
+      "loss": 0.9857,
+      "step": 4054
+    },
+    {
+      "epoch": 0.3623204592668707,
+      "grad_norm": 0.5149965286254883,
+      "learning_rate": 7.372801354175409e-05,
+      "loss": 0.8709,
+      "step": 4055
+    },
+    {
+      "epoch": 0.36240981079813256,
+      "grad_norm": 0.49608248472213745,
+      "learning_rate": 7.371527509013257e-05,
+      "loss": 0.9298,
+      "step": 4056
+    },
+    {
+      "epoch": 0.36249916232939444,
+      "grad_norm": 0.5133218169212341,
+      "learning_rate": 7.370253465210398e-05,
+      "loss": 0.9234,
+      "step": 4057
+    },
+    {
+      "epoch": 0.36258851386065627,
+      "grad_norm": 0.4780612587928772,
+      "learning_rate": 7.368979222873547e-05,
+      "loss": 0.9725,
+      "step": 4058
+    },
+    {
+      "epoch": 0.36267786539191815,
+      "grad_norm": 0.5128650665283203,
+      "learning_rate": 7.367704782109433e-05,
+      "loss": 0.9635,
+      "step": 4059
+    },
+    {
+      "epoch": 0.36276721692318004,
+      "grad_norm": 0.41586834192276,
+      "learning_rate": 7.366430143024805e-05,
+      "loss": 0.9352,
+      "step": 4060
+    },
+    {
+      "epoch": 0.36285656845444186,
+      "grad_norm": 0.3711674213409424,
+      "learning_rate": 7.36515530572643e-05,
+      "loss": 1.0388,
+      "step": 4061
+    },
+    {
+      "epoch": 0.36294591998570375,
+      "grad_norm": 0.5341949462890625,
+      "learning_rate": 7.363880270321087e-05,
+      "loss": 0.9257,
+      "step": 4062
+    },
+    {
+      "epoch": 0.36303527151696563,
+      "grad_norm": 0.43725699186325073,
+      "learning_rate": 7.362605036915574e-05,
+      "loss": 0.9485,
+      "step": 4063
+    },
+    {
+      "epoch": 0.3631246230482275,
+      "grad_norm": 0.47559094429016113,
+      "learning_rate": 7.361329605616705e-05,
+      "loss": 1.0375,
+      "step": 4064
+    },
+    {
+      "epoch": 0.36321397457948934,
+      "grad_norm": 0.3746141493320465,
+      "learning_rate": 7.360053976531312e-05,
+      "loss": 0.9983,
+      "step": 4065
+    },
+    {
+      "epoch": 0.3633033261107512,
+      "grad_norm": 0.4194895625114441,
+      "learning_rate": 7.358778149766244e-05,
+      "loss": 0.9532,
+      "step": 4066
+    },
+    {
+      "epoch": 0.3633926776420131,
+      "grad_norm": 0.42524465918540955,
+      "learning_rate": 7.35750212542836e-05,
+      "loss": 1.0122,
+      "step": 4067
+    },
+    {
+      "epoch": 0.36348202917327493,
+      "grad_norm": 0.5586007237434387,
+      "learning_rate": 7.356225903624545e-05,
+      "loss": 0.9938,
+      "step": 4068
+    },
+    {
+      "epoch": 0.3635713807045368,
+      "grad_norm": 0.36692580580711365,
+      "learning_rate": 7.354949484461697e-05,
+      "loss": 1.0279,
+      "step": 4069
+    },
+    {
+      "epoch": 0.3636607322357987,
+      "grad_norm": 0.40066802501678467,
+      "learning_rate": 7.353672868046725e-05,
+      "loss": 1.0642,
+      "step": 4070
+    },
+    {
+      "epoch": 0.3637500837670606,
+      "grad_norm": 0.4100087285041809,
+      "learning_rate": 7.352396054486562e-05,
+      "loss": 1.0454,
+      "step": 4071
+    },
+    {
+      "epoch": 0.3638394352983224,
+      "grad_norm": 0.39286187291145325,
+      "learning_rate": 7.351119043888158e-05,
+      "loss": 1.0109,
+      "step": 4072
+    },
+    {
+      "epoch": 0.3639287868295843,
+      "grad_norm": 0.39239948987960815,
+      "learning_rate": 7.349841836358468e-05,
+      "loss": 1.0396,
+      "step": 4073
+    },
+    {
+      "epoch": 0.3640181383608462,
+      "grad_norm": 0.3697894513607025,
+      "learning_rate": 7.34856443200448e-05,
+      "loss": 1.02,
+      "step": 4074
+    },
+    {
+      "epoch": 0.364107489892108,
+      "grad_norm": 0.44793811440467834,
+      "learning_rate": 7.347286830933187e-05,
+      "loss": 0.9778,
+      "step": 4075
+    },
+    {
+      "epoch": 0.3641968414233699,
+      "grad_norm": 0.46395543217658997,
+      "learning_rate": 7.3460090332516e-05,
+      "loss": 0.9936,
+      "step": 4076
+    },
+    {
+      "epoch": 0.36428619295463177,
+      "grad_norm": 0.4393361806869507,
+      "learning_rate": 7.344731039066752e-05,
+      "loss": 0.9751,
+      "step": 4077
+    },
+    {
+      "epoch": 0.36437554448589365,
+      "grad_norm": 0.3656950294971466,
+      "learning_rate": 7.343452848485683e-05,
+      "loss": 1.0026,
+      "step": 4078
+    },
+    {
+      "epoch": 0.3644648960171555,
+      "grad_norm": 0.43100160360336304,
+      "learning_rate": 7.342174461615461e-05,
+      "loss": 0.9219,
+      "step": 4079
+    },
+    {
+      "epoch": 0.36455424754841737,
+      "grad_norm": 0.37089818716049194,
+      "learning_rate": 7.340895878563162e-05,
+      "loss": 1.0019,
+      "step": 4080
+    },
+    {
+      "epoch": 0.36464359907967925,
+      "grad_norm": 0.4475706219673157,
+      "learning_rate": 7.339617099435881e-05,
+      "loss": 0.9677,
+      "step": 4081
+    },
+    {
+      "epoch": 0.3647329506109411,
+      "grad_norm": 0.3892630636692047,
+      "learning_rate": 7.338338124340728e-05,
+      "loss": 1.0284,
+      "step": 4082
+    },
+    {
+      "epoch": 0.36482230214220296,
+      "grad_norm": 0.4408927857875824,
+      "learning_rate": 7.337058953384834e-05,
+      "loss": 0.9299,
+      "step": 4083
+    },
+    {
+      "epoch": 0.36491165367346484,
+      "grad_norm": 0.5656498670578003,
+      "learning_rate": 7.335779586675341e-05,
+      "loss": 0.9921,
+      "step": 4084
+    },
+    {
+      "epoch": 0.36500100520472667,
+      "grad_norm": 0.40620046854019165,
+      "learning_rate": 7.334500024319409e-05,
+      "loss": 1.0244,
+      "step": 4085
+    },
+    {
+      "epoch": 0.36509035673598855,
+      "grad_norm": 0.5039829611778259,
+      "learning_rate": 7.333220266424217e-05,
+      "loss": 0.944,
+      "step": 4086
+    },
+    {
+      "epoch": 0.36517970826725044,
+      "grad_norm": 0.4727097451686859,
+      "learning_rate": 7.331940313096957e-05,
+      "loss": 0.8922,
+      "step": 4087
+    },
+    {
+      "epoch": 0.3652690597985123,
+      "grad_norm": 0.44061312079429626,
+      "learning_rate": 7.33066016444484e-05,
+      "loss": 0.9798,
+      "step": 4088
+    },
+    {
+      "epoch": 0.36535841132977415,
+      "grad_norm": 0.45610326528549194,
+      "learning_rate": 7.329379820575089e-05,
+      "loss": 1.004,
+      "step": 4089
+    },
+    {
+      "epoch": 0.36544776286103603,
+      "grad_norm": 0.4043094515800476,
+      "learning_rate": 7.32809928159495e-05,
+      "loss": 1.0491,
+      "step": 4090
+    },
+    {
+      "epoch": 0.3655371143922979,
+      "grad_norm": 0.4471263885498047,
+      "learning_rate": 7.32681854761168e-05,
+      "loss": 0.9727,
+      "step": 4091
+    },
+    {
+      "epoch": 0.36562646592355974,
+      "grad_norm": 0.5790581703186035,
+      "learning_rate": 7.325537618732557e-05,
+      "loss": 0.9889,
+      "step": 4092
+    },
+    {
+      "epoch": 0.3657158174548216,
+      "grad_norm": 0.4110223650932312,
+      "learning_rate": 7.324256495064867e-05,
+      "loss": 1.0602,
+      "step": 4093
+    },
+    {
+      "epoch": 0.3658051689860835,
+      "grad_norm": 0.5571956634521484,
+      "learning_rate": 7.322975176715921e-05,
+      "loss": 1.0389,
+      "step": 4094
+    },
+    {
+      "epoch": 0.3658945205173454,
+      "grad_norm": 0.3689397871494293,
+      "learning_rate": 7.321693663793044e-05,
+      "loss": 0.9932,
+      "step": 4095
+    },
+    {
+      "epoch": 0.3659838720486072,
+      "grad_norm": 0.37520068883895874,
+      "learning_rate": 7.320411956403573e-05,
+      "loss": 1.0419,
+      "step": 4096
+    },
+    {
+      "epoch": 0.3660732235798691,
+      "grad_norm": 0.5426008105278015,
+      "learning_rate": 7.319130054654869e-05,
+      "loss": 0.9726,
+      "step": 4097
+    },
+    {
+      "epoch": 0.366162575111131,
+      "grad_norm": 0.41847026348114014,
+      "learning_rate": 7.317847958654303e-05,
+      "loss": 0.9663,
+      "step": 4098
+    },
+    {
+      "epoch": 0.3662519266423928,
+      "grad_norm": 0.39576879143714905,
+      "learning_rate": 7.316565668509262e-05,
+      "loss": 0.9269,
+      "step": 4099
+    },
+    {
+      "epoch": 0.3663412781736547,
+      "grad_norm": 0.46187660098075867,
+      "learning_rate": 7.315283184327156e-05,
+      "loss": 1.0216,
+      "step": 4100
+    },
+    {
+      "epoch": 0.3664306297049166,
+      "grad_norm": 0.4226139187812805,
+      "learning_rate": 7.314000506215402e-05,
+      "loss": 0.9553,
+      "step": 4101
+    },
+    {
+      "epoch": 0.36651998123617846,
+      "grad_norm": 0.41681933403015137,
+      "learning_rate": 7.312717634281441e-05,
+      "loss": 1.0133,
+      "step": 4102
+    },
+    {
+      "epoch": 0.3666093327674403,
+      "grad_norm": 0.43406426906585693,
+      "learning_rate": 7.311434568632725e-05,
+      "loss": 1.0148,
+      "step": 4103
+    },
+    {
+      "epoch": 0.36669868429870217,
+      "grad_norm": 0.6190657615661621,
+      "learning_rate": 7.310151309376728e-05,
+      "loss": 0.8836,
+      "step": 4104
+    },
+    {
+      "epoch": 0.36678803582996405,
+      "grad_norm": 0.43368059396743774,
+      "learning_rate": 7.308867856620933e-05,
+      "loss": 0.9831,
+      "step": 4105
+    },
+    {
+      "epoch": 0.3668773873612259,
+      "grad_norm": 0.41560328006744385,
+      "learning_rate": 7.307584210472844e-05,
+      "loss": 1.0052,
+      "step": 4106
+    },
+    {
+      "epoch": 0.36696673889248776,
+      "grad_norm": 0.45990949869155884,
+      "learning_rate": 7.306300371039983e-05,
+      "loss": 0.9839,
+      "step": 4107
+    },
+    {
+      "epoch": 0.36705609042374965,
+      "grad_norm": 0.4657072126865387,
+      "learning_rate": 7.30501633842988e-05,
+      "loss": 0.9505,
+      "step": 4108
+    },
+    {
+      "epoch": 0.36714544195501153,
+      "grad_norm": 0.39380112290382385,
+      "learning_rate": 7.303732112750089e-05,
+      "loss": 1.0012,
+      "step": 4109
+    },
+    {
+      "epoch": 0.36723479348627336,
+      "grad_norm": 0.4520123600959778,
+      "learning_rate": 7.302447694108177e-05,
+      "loss": 0.9982,
+      "step": 4110
+    },
+    {
+      "epoch": 0.36732414501753524,
+      "grad_norm": 0.5289416909217834,
+      "learning_rate": 7.301163082611729e-05,
+      "loss": 0.9575,
+      "step": 4111
+    },
+    {
+      "epoch": 0.3674134965487971,
+      "grad_norm": 0.5018066167831421,
+      "learning_rate": 7.299878278368345e-05,
+      "loss": 0.8618,
+      "step": 4112
+    },
+    {
+      "epoch": 0.36750284808005895,
+      "grad_norm": 0.3886435031890869,
+      "learning_rate": 7.29859328148564e-05,
+      "loss": 0.9727,
+      "step": 4113
+    },
+    {
+      "epoch": 0.36759219961132084,
+      "grad_norm": 0.45843201875686646,
+      "learning_rate": 7.297308092071245e-05,
+      "loss": 0.9968,
+      "step": 4114
+    },
+    {
+      "epoch": 0.3676815511425827,
+      "grad_norm": 0.4753422141075134,
+      "learning_rate": 7.296022710232812e-05,
+      "loss": 0.9434,
+      "step": 4115
+    },
+    {
+      "epoch": 0.36777090267384455,
+      "grad_norm": 0.4317519962787628,
+      "learning_rate": 7.294737136078001e-05,
+      "loss": 0.9388,
+      "step": 4116
+    },
+    {
+      "epoch": 0.36786025420510643,
+      "grad_norm": 0.388031005859375,
+      "learning_rate": 7.293451369714495e-05,
+      "loss": 1.079,
+      "step": 4117
+    },
+    {
+      "epoch": 0.3679496057363683,
+      "grad_norm": 0.42603445053100586,
+      "learning_rate": 7.292165411249993e-05,
+      "loss": 0.9291,
+      "step": 4118
+    },
+    {
+      "epoch": 0.3680389572676302,
+      "grad_norm": 0.4375905990600586,
+      "learning_rate": 7.290879260792203e-05,
+      "loss": 0.9836,
+      "step": 4119
+    },
+    {
+      "epoch": 0.368128308798892,
+      "grad_norm": 0.40802040696144104,
+      "learning_rate": 7.289592918448856e-05,
+      "loss": 0.9789,
+      "step": 4120
+    },
+    {
+      "epoch": 0.3682176603301539,
+      "grad_norm": 0.3883795142173767,
+      "learning_rate": 7.288306384327696e-05,
+      "loss": 1.0521,
+      "step": 4121
+    },
+    {
+      "epoch": 0.3683070118614158,
+      "grad_norm": 0.42812490463256836,
+      "learning_rate": 7.287019658536486e-05,
+      "loss": 1.0214,
+      "step": 4122
+    },
+    {
+      "epoch": 0.3683963633926776,
+      "grad_norm": 0.4556853771209717,
+      "learning_rate": 7.285732741183003e-05,
+      "loss": 0.9778,
+      "step": 4123
+    },
+    {
+      "epoch": 0.3684857149239395,
+      "grad_norm": 0.43323996663093567,
+      "learning_rate": 7.284445632375035e-05,
+      "loss": 0.9807,
+      "step": 4124
+    },
+    {
+      "epoch": 0.3685750664552014,
+      "grad_norm": 0.4459212124347687,
+      "learning_rate": 7.283158332220397e-05,
+      "loss": 0.9184,
+      "step": 4125
+    },
+    {
+      "epoch": 0.36866441798646327,
+      "grad_norm": 0.43171730637550354,
+      "learning_rate": 7.281870840826912e-05,
+      "loss": 0.9214,
+      "step": 4126
+    },
+    {
+      "epoch": 0.3687537695177251,
+      "grad_norm": 0.4195806086063385,
+      "learning_rate": 7.280583158302421e-05,
+      "loss": 0.9684,
+      "step": 4127
+    },
+    {
+      "epoch": 0.368843121048987,
+      "grad_norm": 0.4562622606754303,
+      "learning_rate": 7.279295284754782e-05,
+      "loss": 0.9569,
+      "step": 4128
+    },
+    {
+      "epoch": 0.36893247258024886,
+      "grad_norm": 0.5500267148017883,
+      "learning_rate": 7.278007220291866e-05,
+      "loss": 1.0506,
+      "step": 4129
+    },
+    {
+      "epoch": 0.3690218241115107,
+      "grad_norm": 0.5012460947036743,
+      "learning_rate": 7.276718965021563e-05,
+      "loss": 0.9673,
+      "step": 4130
+    },
+    {
+      "epoch": 0.36911117564277257,
+      "grad_norm": 0.46749043464660645,
+      "learning_rate": 7.27543051905178e-05,
+      "loss": 0.9777,
+      "step": 4131
+    },
+    {
+      "epoch": 0.36920052717403445,
+      "grad_norm": 0.49202150106430054,
+      "learning_rate": 7.274141882490435e-05,
+      "loss": 0.9335,
+      "step": 4132
+    },
+    {
+      "epoch": 0.36928987870529634,
+      "grad_norm": 0.47457700967788696,
+      "learning_rate": 7.27285305544547e-05,
+      "loss": 0.9617,
+      "step": 4133
+    },
+    {
+      "epoch": 0.36937923023655816,
+      "grad_norm": 0.46560418605804443,
+      "learning_rate": 7.271564038024831e-05,
+      "loss": 1.0336,
+      "step": 4134
+    },
+    {
+      "epoch": 0.36946858176782005,
+      "grad_norm": 0.41487690806388855,
+      "learning_rate": 7.270274830336493e-05,
+      "loss": 0.9539,
+      "step": 4135
+    },
+    {
+      "epoch": 0.36955793329908193,
+      "grad_norm": 0.4482000470161438,
+      "learning_rate": 7.268985432488438e-05,
+      "loss": 0.9841,
+      "step": 4136
+    },
+    {
+      "epoch": 0.36964728483034376,
+      "grad_norm": 0.44116854667663574,
+      "learning_rate": 7.267695844588668e-05,
+      "loss": 0.9894,
+      "step": 4137
+    },
+    {
+      "epoch": 0.36973663636160564,
+      "grad_norm": 0.3913993537425995,
+      "learning_rate": 7.266406066745199e-05,
+      "loss": 0.9203,
+      "step": 4138
+    },
+    {
+      "epoch": 0.3698259878928675,
+      "grad_norm": 0.43621063232421875,
+      "learning_rate": 7.265116099066063e-05,
+      "loss": 1.0129,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3699153394241294,
+      "grad_norm": 0.5264103412628174,
+      "learning_rate": 7.26382594165931e-05,
+      "loss": 0.9673,
+      "step": 4140
+    },
+    {
+      "epoch": 0.37000469095539124,
+      "grad_norm": 0.393413782119751,
+      "learning_rate": 7.262535594633002e-05,
+      "loss": 1.0037,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3700940424866531,
+      "grad_norm": 0.5121841430664062,
+      "learning_rate": 7.261245058095223e-05,
+      "loss": 0.8443,
+      "step": 4142
+    },
+    {
+      "epoch": 0.370183394017915,
+      "grad_norm": 0.4309183359146118,
+      "learning_rate": 7.259954332154066e-05,
+      "loss": 1.0241,
+      "step": 4143
+    },
+    {
+      "epoch": 0.37027274554917683,
+      "grad_norm": 0.3869653046131134,
+      "learning_rate": 7.258663416917645e-05,
+      "loss": 1.0111,
+      "step": 4144
+    },
+    {
+      "epoch": 0.3703620970804387,
+      "grad_norm": 0.5161389112472534,
+      "learning_rate": 7.257372312494088e-05,
+      "loss": 1.0608,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3704514486117006,
+      "grad_norm": 0.48384639620780945,
+      "learning_rate": 7.256081018991536e-05,
+      "loss": 0.9671,
+      "step": 4146
+    },
+    {
+      "epoch": 0.3705408001429624,
+      "grad_norm": 0.4092333912849426,
+      "learning_rate": 7.254789536518151e-05,
+      "loss": 1.0206,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3706301516742243,
+      "grad_norm": 0.4120895266532898,
+      "learning_rate": 7.25349786518211e-05,
+      "loss": 1.0183,
+      "step": 4148
+    },
+    {
+      "epoch": 0.3707195032054862,
+      "grad_norm": 0.47029179334640503,
+      "learning_rate": 7.2522060050916e-05,
+      "loss": 0.9786,
+      "step": 4149
+    },
+    {
+      "epoch": 0.3708088547367481,
+      "grad_norm": 0.38169971108436584,
+      "learning_rate": 7.25091395635483e-05,
+      "loss": 0.9691,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3708982062680099,
+      "grad_norm": 0.48329901695251465,
+      "learning_rate": 7.249621719080025e-05,
+      "loss": 0.9664,
+      "step": 4151
+    },
+    {
+      "epoch": 0.3709875577992718,
+      "grad_norm": 0.4072484076023102,
+      "learning_rate": 7.248329293375422e-05,
+      "loss": 1.0322,
+      "step": 4152
+    },
+    {
+      "epoch": 0.37107690933053367,
+      "grad_norm": 0.4380795955657959,
+      "learning_rate": 7.247036679349274e-05,
+      "loss": 0.9546,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3711662608617955,
+      "grad_norm": 0.515524685382843,
+      "learning_rate": 7.245743877109852e-05,
+      "loss": 0.9816,
+      "step": 4154
+    },
+    {
+      "epoch": 0.3712556123930574,
+      "grad_norm": 0.42287132143974304,
+      "learning_rate": 7.244450886765443e-05,
+      "loss": 0.9431,
+      "step": 4155
+    },
+    {
+      "epoch": 0.37134496392431926,
+      "grad_norm": 0.4340779483318329,
+      "learning_rate": 7.243157708424348e-05,
+      "loss": 0.9488,
+      "step": 4156
+    },
+    {
+      "epoch": 0.37143431545558114,
+      "grad_norm": 0.524341344833374,
+      "learning_rate": 7.241864342194886e-05,
+      "loss": 0.9756,
+      "step": 4157
+    },
+    {
+      "epoch": 0.37152366698684297,
+      "grad_norm": 0.6122943758964539,
+      "learning_rate": 7.240570788185388e-05,
+      "loss": 0.9382,
+      "step": 4158
+    },
+    {
+      "epoch": 0.37161301851810485,
+      "grad_norm": 0.5012384653091431,
+      "learning_rate": 7.239277046504202e-05,
+      "loss": 0.9809,
+      "step": 4159
+    },
+    {
+      "epoch": 0.37170237004936674,
+      "grad_norm": 0.3928990364074707,
+      "learning_rate": 7.237983117259696e-05,
+      "loss": 0.9839,
+      "step": 4160
+    },
+    {
+      "epoch": 0.37179172158062856,
+      "grad_norm": 0.4313662350177765,
+      "learning_rate": 7.236689000560248e-05,
+      "loss": 1.0517,
+      "step": 4161
+    },
+    {
+      "epoch": 0.37188107311189045,
+      "grad_norm": 0.3812869191169739,
+      "learning_rate": 7.235394696514255e-05,
+      "loss": 1.0109,
+      "step": 4162
+    },
+    {
+      "epoch": 0.37197042464315233,
+      "grad_norm": 0.3870796263217926,
+      "learning_rate": 7.23410020523013e-05,
+      "loss": 1.0234,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3720597761744142,
+      "grad_norm": 0.4438193142414093,
+      "learning_rate": 7.232805526816297e-05,
+      "loss": 0.9647,
+      "step": 4164
+    },
+    {
+      "epoch": 0.37214912770567604,
+      "grad_norm": 0.375704288482666,
+      "learning_rate": 7.231510661381202e-05,
+      "loss": 0.9969,
+      "step": 4165
+    },
+    {
+      "epoch": 0.3722384792369379,
+      "grad_norm": 0.38540956377983093,
+      "learning_rate": 7.230215609033301e-05,
+      "loss": 1.0308,
+      "step": 4166
+    },
+    {
+      "epoch": 0.3723278307681998,
+      "grad_norm": 0.3983640670776367,
+      "learning_rate": 7.228920369881073e-05,
+      "loss": 0.9823,
+      "step": 4167
+    },
+    {
+      "epoch": 0.37241718229946164,
+      "grad_norm": 0.4259108901023865,
+      "learning_rate": 7.227624944033006e-05,
+      "loss": 0.9739,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3725065338307235,
+      "grad_norm": 0.4795960783958435,
+      "learning_rate": 7.226329331597604e-05,
+      "loss": 0.9372,
+      "step": 4169
+    },
+    {
+      "epoch": 0.3725958853619854,
+      "grad_norm": 0.4027024209499359,
+      "learning_rate": 7.225033532683388e-05,
+      "loss": 0.9707,
+      "step": 4170
+    },
+    {
+      "epoch": 0.3726852368932473,
+      "grad_norm": 0.42690253257751465,
+      "learning_rate": 7.223737547398898e-05,
+      "loss": 0.973,
+      "step": 4171
+    },
+    {
+      "epoch": 0.3727745884245091,
+      "grad_norm": 0.42597901821136475,
+      "learning_rate": 7.222441375852685e-05,
+      "loss": 0.9487,
+      "step": 4172
+    },
+    {
+      "epoch": 0.372863939955771,
+      "grad_norm": 0.43113502860069275,
+      "learning_rate": 7.221145018153317e-05,
+      "loss": 1.0139,
+      "step": 4173
+    },
+    {
+      "epoch": 0.3729532914870329,
+      "grad_norm": 0.4594292938709259,
+      "learning_rate": 7.219848474409378e-05,
+      "loss": 1.0409,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3730426430182947,
+      "grad_norm": 0.39709535241127014,
+      "learning_rate": 7.218551744729468e-05,
+      "loss": 0.9839,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3731319945495566,
+      "grad_norm": 0.38931846618652344,
+      "learning_rate": 7.217254829222201e-05,
+      "loss": 0.9697,
+      "step": 4176
+    },
+    {
+      "epoch": 0.3732213460808185,
+      "grad_norm": 0.3806284964084625,
+      "learning_rate": 7.215957727996207e-05,
+      "loss": 1.0805,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3733106976120803,
+      "grad_norm": 0.4208787977695465,
+      "learning_rate": 7.214660441160134e-05,
+      "loss": 0.9579,
+      "step": 4178
+    },
+    {
+      "epoch": 0.3734000491433422,
+      "grad_norm": 0.4032423198223114,
+      "learning_rate": 7.213362968822643e-05,
+      "loss": 0.985,
+      "step": 4179
+    },
+    {
+      "epoch": 0.37348940067460407,
+      "grad_norm": 0.4312584400177002,
+      "learning_rate": 7.212065311092409e-05,
+      "loss": 0.9404,
+      "step": 4180
+    },
+    {
+      "epoch": 0.37357875220586595,
+      "grad_norm": 0.39697664976119995,
+      "learning_rate": 7.210767468078126e-05,
+      "loss": 0.9926,
+      "step": 4181
+    },
+    {
+      "epoch": 0.3736681037371278,
+      "grad_norm": 0.38140472769737244,
+      "learning_rate": 7.209469439888504e-05,
+      "loss": 0.9736,
+      "step": 4182
+    },
+    {
+      "epoch": 0.37375745526838966,
+      "grad_norm": 0.41029414534568787,
+      "learning_rate": 7.208171226632264e-05,
+      "loss": 1.0136,
+      "step": 4183
+    },
+    {
+      "epoch": 0.37384680679965154,
+      "grad_norm": 0.44529950618743896,
+      "learning_rate": 7.206872828418146e-05,
+      "loss": 0.9878,
+      "step": 4184
+    },
+    {
+      "epoch": 0.37393615833091337,
+      "grad_norm": 0.4840013086795807,
+      "learning_rate": 7.205574245354907e-05,
+      "loss": 0.9829,
+      "step": 4185
+    },
+    {
+      "epoch": 0.37402550986217525,
+      "grad_norm": 0.41455405950546265,
+      "learning_rate": 7.204275477551314e-05,
+      "loss": 0.9755,
+      "step": 4186
+    },
+    {
+      "epoch": 0.37411486139343714,
+      "grad_norm": 0.5246483087539673,
+      "learning_rate": 7.202976525116154e-05,
+      "loss": 0.9764,
+      "step": 4187
+    },
+    {
+      "epoch": 0.374204212924699,
+      "grad_norm": 0.484429270029068,
+      "learning_rate": 7.201677388158228e-05,
+      "loss": 0.9466,
+      "step": 4188
+    },
+    {
+      "epoch": 0.37429356445596085,
+      "grad_norm": 0.39618611335754395,
+      "learning_rate": 7.200378066786352e-05,
+      "loss": 1.0202,
+      "step": 4189
+    },
+    {
+      "epoch": 0.37438291598722273,
+      "grad_norm": 0.4614309072494507,
+      "learning_rate": 7.199078561109359e-05,
+      "loss": 0.9347,
+      "step": 4190
+    },
+    {
+      "epoch": 0.3744722675184846,
+      "grad_norm": 0.46450915932655334,
+      "learning_rate": 7.197778871236096e-05,
+      "loss": 0.9394,
+      "step": 4191
+    },
+    {
+      "epoch": 0.37456161904974644,
+      "grad_norm": 0.37364086508750916,
+      "learning_rate": 7.196478997275426e-05,
+      "loss": 1.0168,
+      "step": 4192
+    },
+    {
+      "epoch": 0.3746509705810083,
+      "grad_norm": 0.4108889400959015,
+      "learning_rate": 7.195178939336228e-05,
+      "loss": 0.9738,
+      "step": 4193
+    },
+    {
+      "epoch": 0.3747403221122702,
+      "grad_norm": 0.43093374371528625,
+      "learning_rate": 7.193878697527394e-05,
+      "loss": 0.9225,
+      "step": 4194
+    },
+    {
+      "epoch": 0.3748296736435321,
+      "grad_norm": 0.5023407340049744,
+      "learning_rate": 7.192578271957833e-05,
+      "loss": 0.9292,
+      "step": 4195
+    },
+    {
+      "epoch": 0.3749190251747939,
+      "grad_norm": 0.42425450682640076,
+      "learning_rate": 7.191277662736473e-05,
+      "loss": 0.9801,
+      "step": 4196
+    },
+    {
+      "epoch": 0.3750083767060558,
+      "grad_norm": 0.46366748213768005,
+      "learning_rate": 7.189976869972248e-05,
+      "loss": 0.9504,
+      "step": 4197
+    },
+    {
+      "epoch": 0.3750977282373177,
+      "grad_norm": 0.4702743589878082,
+      "learning_rate": 7.18867589377412e-05,
+      "loss": 0.9922,
+      "step": 4198
+    },
+    {
+      "epoch": 0.3751870797685795,
+      "grad_norm": 0.4722293019294739,
+      "learning_rate": 7.187374734251054e-05,
+      "loss": 1.0015,
+      "step": 4199
+    },
+    {
+      "epoch": 0.3752764312998414,
+      "grad_norm": 0.4281826615333557,
+      "learning_rate": 7.186073391512039e-05,
+      "loss": 0.9522,
+      "step": 4200
+    },
+    {
+      "epoch": 0.3753657828311033,
+      "grad_norm": 0.3947566747665405,
+      "learning_rate": 7.184771865666076e-05,
+      "loss": 1.0434,
+      "step": 4201
+    },
+    {
+      "epoch": 0.37545513436236516,
+      "grad_norm": 0.43979522585868835,
+      "learning_rate": 7.18347015682218e-05,
+      "loss": 0.9935,
+      "step": 4202
+    },
+    {
+      "epoch": 0.375544485893627,
+      "grad_norm": 0.41721320152282715,
+      "learning_rate": 7.182168265089386e-05,
+      "loss": 0.9861,
+      "step": 4203
+    },
+    {
+      "epoch": 0.3756338374248889,
+      "grad_norm": 0.46523186564445496,
+      "learning_rate": 7.18086619057674e-05,
+      "loss": 0.9951,
+      "step": 4204
+    },
+    {
+      "epoch": 0.37572318895615076,
+      "grad_norm": 0.43534207344055176,
+      "learning_rate": 7.179563933393303e-05,
+      "loss": 0.9288,
+      "step": 4205
+    },
+    {
+      "epoch": 0.3758125404874126,
+      "grad_norm": 0.4193892776966095,
+      "learning_rate": 7.178261493648154e-05,
+      "loss": 1.0575,
+      "step": 4206
+    },
+    {
+      "epoch": 0.37590189201867447,
+      "grad_norm": 0.48996466398239136,
+      "learning_rate": 7.176958871450386e-05,
+      "loss": 0.9262,
+      "step": 4207
+    },
+    {
+      "epoch": 0.37599124354993635,
+      "grad_norm": 0.4287664294242859,
+      "learning_rate": 7.175656066909109e-05,
+      "loss": 1.031,
+      "step": 4208
+    },
+    {
+      "epoch": 0.3760805950811982,
+      "grad_norm": 0.4364438056945801,
+      "learning_rate": 7.174353080133445e-05,
+      "loss": 1.0015,
+      "step": 4209
+    },
+    {
+      "epoch": 0.37616994661246006,
+      "grad_norm": 0.4443535804748535,
+      "learning_rate": 7.173049911232533e-05,
+      "loss": 0.9699,
+      "step": 4210
+    },
+    {
+      "epoch": 0.37625929814372194,
+      "grad_norm": 0.4485486149787903,
+      "learning_rate": 7.171746560315529e-05,
+      "loss": 0.996,
+      "step": 4211
+    },
+    {
+      "epoch": 0.3763486496749838,
+      "grad_norm": 0.43986693024635315,
+      "learning_rate": 7.1704430274916e-05,
+      "loss": 0.9912,
+      "step": 4212
+    },
+    {
+      "epoch": 0.37643800120624565,
+      "grad_norm": 0.42477118968963623,
+      "learning_rate": 7.169139312869933e-05,
+      "loss": 1.0281,
+      "step": 4213
+    },
+    {
+      "epoch": 0.37652735273750754,
+      "grad_norm": 0.35374677181243896,
+      "learning_rate": 7.167835416559727e-05,
+      "loss": 1.0282,
+      "step": 4214
+    },
+    {
+      "epoch": 0.3766167042687694,
+      "grad_norm": 0.5305443406105042,
+      "learning_rate": 7.166531338670198e-05,
+      "loss": 0.8789,
+      "step": 4215
+    },
+    {
+      "epoch": 0.37670605580003125,
+      "grad_norm": 0.4482033848762512,
+      "learning_rate": 7.165227079310575e-05,
+      "loss": 0.9928,
+      "step": 4216
+    },
+    {
+      "epoch": 0.37679540733129313,
+      "grad_norm": 0.4617825150489807,
+      "learning_rate": 7.163922638590104e-05,
+      "loss": 0.9478,
+      "step": 4217
+    },
+    {
+      "epoch": 0.376884758862555,
+      "grad_norm": 0.4732501804828644,
+      "learning_rate": 7.162618016618047e-05,
+      "loss": 0.9714,
+      "step": 4218
+    },
+    {
+      "epoch": 0.3769741103938169,
+      "grad_norm": 0.40239885449409485,
+      "learning_rate": 7.161313213503679e-05,
+      "loss": 0.9664,
+      "step": 4219
+    },
+    {
+      "epoch": 0.3770634619250787,
+      "grad_norm": 0.4415636658668518,
+      "learning_rate": 7.160008229356292e-05,
+      "loss": 1.0054,
+      "step": 4220
+    },
+    {
+      "epoch": 0.3771528134563406,
+      "grad_norm": 0.47628524899482727,
+      "learning_rate": 7.158703064285192e-05,
+      "loss": 0.9021,
+      "step": 4221
+    },
+    {
+      "epoch": 0.3772421649876025,
+      "grad_norm": 0.3998652398586273,
+      "learning_rate": 7.157397718399698e-05,
+      "loss": 1.046,
+      "step": 4222
+    },
+    {
+      "epoch": 0.3773315165188643,
+      "grad_norm": 0.3887675702571869,
+      "learning_rate": 7.156092191809152e-05,
+      "loss": 1.0419,
+      "step": 4223
+    },
+    {
+      "epoch": 0.3774208680501262,
+      "grad_norm": 0.3833785951137543,
+      "learning_rate": 7.154786484622901e-05,
+      "loss": 0.9874,
+      "step": 4224
+    },
+    {
+      "epoch": 0.3775102195813881,
+      "grad_norm": 0.4666437804698944,
+      "learning_rate": 7.153480596950314e-05,
+      "loss": 0.967,
+      "step": 4225
+    },
+    {
+      "epoch": 0.37759957111264997,
+      "grad_norm": 0.37928861379623413,
+      "learning_rate": 7.152174528900772e-05,
+      "loss": 0.9859,
+      "step": 4226
+    },
+    {
+      "epoch": 0.3776889226439118,
+      "grad_norm": 0.5285788178443909,
+      "learning_rate": 7.150868280583674e-05,
+      "loss": 0.9279,
+      "step": 4227
+    },
+    {
+      "epoch": 0.3777782741751737,
+      "grad_norm": 0.4263918697834015,
+      "learning_rate": 7.149561852108429e-05,
+      "loss": 0.9441,
+      "step": 4228
+    },
+    {
+      "epoch": 0.37786762570643556,
+      "grad_norm": 0.40622586011886597,
+      "learning_rate": 7.148255243584467e-05,
+      "loss": 0.9801,
+      "step": 4229
+    },
+    {
+      "epoch": 0.3779569772376974,
+      "grad_norm": 0.3894538879394531,
+      "learning_rate": 7.146948455121231e-05,
+      "loss": 1.0106,
+      "step": 4230
+    },
+    {
+      "epoch": 0.37804632876895927,
+      "grad_norm": 0.4317224621772766,
+      "learning_rate": 7.145641486828175e-05,
+      "loss": 0.941,
+      "step": 4231
+    },
+    {
+      "epoch": 0.37813568030022116,
+      "grad_norm": 0.4164870083332062,
+      "learning_rate": 7.144334338814774e-05,
+      "loss": 0.9747,
+      "step": 4232
+    },
+    {
+      "epoch": 0.37822503183148304,
+      "grad_norm": 0.5142220258712769,
+      "learning_rate": 7.143027011190515e-05,
+      "loss": 0.9172,
+      "step": 4233
+    },
+    {
+      "epoch": 0.37831438336274487,
+      "grad_norm": 0.3964780569076538,
+      "learning_rate": 7.141719504064902e-05,
+      "loss": 0.963,
+      "step": 4234
+    },
+    {
+      "epoch": 0.37840373489400675,
+      "grad_norm": 0.532341480255127,
+      "learning_rate": 7.140411817547452e-05,
+      "loss": 0.977,
+      "step": 4235
+    },
+    {
+      "epoch": 0.37849308642526863,
+      "grad_norm": 0.39835068583488464,
+      "learning_rate": 7.139103951747695e-05,
+      "loss": 0.9591,
+      "step": 4236
+    },
+    {
+      "epoch": 0.37858243795653046,
+      "grad_norm": 0.41592469811439514,
+      "learning_rate": 7.137795906775182e-05,
+      "loss": 0.9929,
+      "step": 4237
+    },
+    {
+      "epoch": 0.37867178948779234,
+      "grad_norm": 0.43503326177597046,
+      "learning_rate": 7.136487682739472e-05,
+      "loss": 0.9984,
+      "step": 4238
+    },
+    {
+      "epoch": 0.3787611410190542,
+      "grad_norm": 0.531616747379303,
+      "learning_rate": 7.135179279750149e-05,
+      "loss": 0.9221,
+      "step": 4239
+    },
+    {
+      "epoch": 0.37885049255031605,
+      "grad_norm": 0.4244007170200348,
+      "learning_rate": 7.1338706979168e-05,
+      "loss": 0.9679,
+      "step": 4240
+    },
+    {
+      "epoch": 0.37893984408157794,
+      "grad_norm": 0.4253380000591278,
+      "learning_rate": 7.132561937349035e-05,
+      "loss": 0.9562,
+      "step": 4241
+    },
+    {
+      "epoch": 0.3790291956128398,
+      "grad_norm": 0.4492589235305786,
+      "learning_rate": 7.131252998156476e-05,
+      "loss": 0.9848,
+      "step": 4242
+    },
+    {
+      "epoch": 0.3791185471441017,
+      "grad_norm": 0.41047170758247375,
+      "learning_rate": 7.129943880448762e-05,
+      "loss": 1.0768,
+      "step": 4243
+    },
+    {
+      "epoch": 0.37920789867536353,
+      "grad_norm": 0.43270984292030334,
+      "learning_rate": 7.128634584335545e-05,
+      "loss": 0.9937,
+      "step": 4244
+    },
+    {
+      "epoch": 0.3792972502066254,
+      "grad_norm": 0.45858004689216614,
+      "learning_rate": 7.127325109926491e-05,
+      "loss": 0.9881,
+      "step": 4245
+    },
+    {
+      "epoch": 0.3793866017378873,
+      "grad_norm": 0.4487616717815399,
+      "learning_rate": 7.126015457331281e-05,
+      "loss": 1.0136,
+      "step": 4246
+    },
+    {
+      "epoch": 0.3794759532691491,
+      "grad_norm": 0.41824638843536377,
+      "learning_rate": 7.124705626659616e-05,
+      "loss": 0.9809,
+      "step": 4247
+    },
+    {
+      "epoch": 0.379565304800411,
+      "grad_norm": 0.4919416606426239,
+      "learning_rate": 7.123395618021208e-05,
+      "loss": 0.9974,
+      "step": 4248
+    },
+    {
+      "epoch": 0.3796546563316729,
+      "grad_norm": 0.4910551905632019,
+      "learning_rate": 7.122085431525785e-05,
+      "loss": 1.0215,
+      "step": 4249
+    },
+    {
+      "epoch": 0.3797440078629348,
+      "grad_norm": 0.46813416481018066,
+      "learning_rate": 7.120775067283085e-05,
+      "loss": 0.951,
+      "step": 4250
+    },
+    {
+      "epoch": 0.3798333593941966,
+      "grad_norm": 0.4559837579727173,
+      "learning_rate": 7.119464525402867e-05,
+      "loss": 0.9815,
+      "step": 4251
+    },
+    {
+      "epoch": 0.3799227109254585,
+      "grad_norm": 0.38314858078956604,
+      "learning_rate": 7.118153805994902e-05,
+      "loss": 1.0214,
+      "step": 4252
+    },
+    {
+      "epoch": 0.38001206245672037,
+      "grad_norm": 0.4613505005836487,
+      "learning_rate": 7.11684290916898e-05,
+      "loss": 0.9771,
+      "step": 4253
+    },
+    {
+      "epoch": 0.3801014139879822,
+      "grad_norm": 0.43904268741607666,
+      "learning_rate": 7.1155318350349e-05,
+      "loss": 0.9689,
+      "step": 4254
+    },
+    {
+      "epoch": 0.3801907655192441,
+      "grad_norm": 0.4177344739437103,
+      "learning_rate": 7.11422058370248e-05,
+      "loss": 0.9386,
+      "step": 4255
+    },
+    {
+      "epoch": 0.38028011705050596,
+      "grad_norm": 0.44413110613822937,
+      "learning_rate": 7.112909155281549e-05,
+      "loss": 0.9454,
+      "step": 4256
+    },
+    {
+      "epoch": 0.38036946858176784,
+      "grad_norm": 0.4826868772506714,
+      "learning_rate": 7.111597549881953e-05,
+      "loss": 0.8906,
+      "step": 4257
+    },
+    {
+      "epoch": 0.38045882011302967,
+      "grad_norm": 0.517841100692749,
+      "learning_rate": 7.110285767613555e-05,
+      "loss": 0.9733,
+      "step": 4258
+    },
+    {
+      "epoch": 0.38054817164429156,
+      "grad_norm": 0.4552189111709595,
+      "learning_rate": 7.108973808586231e-05,
+      "loss": 0.969,
+      "step": 4259
+    },
+    {
+      "epoch": 0.38063752317555344,
+      "grad_norm": 0.40824398398399353,
+      "learning_rate": 7.107661672909871e-05,
+      "loss": 0.9928,
+      "step": 4260
+    },
+    {
+      "epoch": 0.38072687470681527,
+      "grad_norm": 0.45659610629081726,
+      "learning_rate": 7.106349360694379e-05,
+      "loss": 0.9898,
+      "step": 4261
+    },
+    {
+      "epoch": 0.38081622623807715,
+      "grad_norm": 0.39422810077667236,
+      "learning_rate": 7.105036872049675e-05,
+      "loss": 1.005,
+      "step": 4262
+    },
+    {
+      "epoch": 0.38090557776933903,
+      "grad_norm": 0.47162938117980957,
+      "learning_rate": 7.103724207085696e-05,
+      "loss": 0.995,
+      "step": 4263
+    },
+    {
+      "epoch": 0.3809949293006009,
+      "grad_norm": 0.45578041672706604,
+      "learning_rate": 7.10241136591239e-05,
+      "loss": 0.9659,
+      "step": 4264
+    },
+    {
+      "epoch": 0.38108428083186274,
+      "grad_norm": 0.35482344031333923,
+      "learning_rate": 7.101098348639722e-05,
+      "loss": 1.0144,
+      "step": 4265
+    },
+    {
+      "epoch": 0.3811736323631246,
+      "grad_norm": 0.4190642237663269,
+      "learning_rate": 7.099785155377672e-05,
+      "loss": 0.9442,
+      "step": 4266
+    },
+    {
+      "epoch": 0.3812629838943865,
+      "grad_norm": 0.4424975514411926,
+      "learning_rate": 7.098471786236231e-05,
+      "loss": 0.9564,
+      "step": 4267
+    },
+    {
+      "epoch": 0.38135233542564834,
+      "grad_norm": 0.4281081259250641,
+      "learning_rate": 7.097158241325411e-05,
+      "loss": 0.9608,
+      "step": 4268
+    },
+    {
+      "epoch": 0.3814416869569102,
+      "grad_norm": 0.46084991097450256,
+      "learning_rate": 7.095844520755235e-05,
+      "loss": 0.9865,
+      "step": 4269
+    },
+    {
+      "epoch": 0.3815310384881721,
+      "grad_norm": 0.3826462924480438,
+      "learning_rate": 7.094530624635741e-05,
+      "loss": 0.978,
+      "step": 4270
+    },
+    {
+      "epoch": 0.38162039001943393,
+      "grad_norm": 0.4165399372577667,
+      "learning_rate": 7.093216553076977e-05,
+      "loss": 0.9959,
+      "step": 4271
+    },
+    {
+      "epoch": 0.3817097415506958,
+      "grad_norm": 0.4049665331840515,
+      "learning_rate": 7.091902306189018e-05,
+      "loss": 1.0591,
+      "step": 4272
+    },
+    {
+      "epoch": 0.3817990930819577,
+      "grad_norm": 0.3788042664527893,
+      "learning_rate": 7.090587884081943e-05,
+      "loss": 1.0281,
+      "step": 4273
+    },
+    {
+      "epoch": 0.3818884446132196,
+      "grad_norm": 0.436847060918808,
+      "learning_rate": 7.089273286865849e-05,
+      "loss": 0.9862,
+      "step": 4274
+    },
+    {
+      "epoch": 0.3819777961444814,
+      "grad_norm": 0.4377835988998413,
+      "learning_rate": 7.087958514650846e-05,
+      "loss": 1.0327,
+      "step": 4275
+    },
+    {
+      "epoch": 0.3820671476757433,
+      "grad_norm": 0.43927526473999023,
+      "learning_rate": 7.08664356754706e-05,
+      "loss": 0.9563,
+      "step": 4276
+    },
+    {
+      "epoch": 0.3821564992070052,
+      "grad_norm": 0.4470997750759125,
+      "learning_rate": 7.085328445664636e-05,
+      "loss": 0.9565,
+      "step": 4277
+    },
+    {
+      "epoch": 0.382245850738267,
+      "grad_norm": 0.42849406599998474,
+      "learning_rate": 7.084013149113724e-05,
+      "loss": 0.9785,
+      "step": 4278
+    },
+    {
+      "epoch": 0.3823352022695289,
+      "grad_norm": 0.37449339032173157,
+      "learning_rate": 7.0826976780045e-05,
+      "loss": 0.9791,
+      "step": 4279
+    },
+    {
+      "epoch": 0.38242455380079077,
+      "grad_norm": 0.39009079337120056,
+      "learning_rate": 7.081382032447142e-05,
+      "loss": 0.9007,
+      "step": 4280
+    },
+    {
+      "epoch": 0.38251390533205265,
+      "grad_norm": 0.37804755568504333,
+      "learning_rate": 7.080066212551854e-05,
+      "loss": 1.0259,
+      "step": 4281
+    },
+    {
+      "epoch": 0.3826032568633145,
+      "grad_norm": 0.4352472722530365,
+      "learning_rate": 7.07875021842885e-05,
+      "loss": 0.9648,
+      "step": 4282
+    },
+    {
+      "epoch": 0.38269260839457636,
+      "grad_norm": 0.38885149359703064,
+      "learning_rate": 7.077434050188356e-05,
+      "loss": 1.0278,
+      "step": 4283
+    },
+    {
+      "epoch": 0.38278195992583824,
+      "grad_norm": 0.3463941514492035,
+      "learning_rate": 7.076117707940617e-05,
+      "loss": 1.0115,
+      "step": 4284
+    },
+    {
+      "epoch": 0.38287131145710007,
+      "grad_norm": 0.4116084575653076,
+      "learning_rate": 7.07480119179589e-05,
+      "loss": 0.9362,
+      "step": 4285
+    },
+    {
+      "epoch": 0.38296066298836196,
+      "grad_norm": 0.44609686732292175,
+      "learning_rate": 7.073484501864447e-05,
+      "loss": 0.9525,
+      "step": 4286
+    },
+    {
+      "epoch": 0.38305001451962384,
+      "grad_norm": 0.40134397149086,
+      "learning_rate": 7.072167638256576e-05,
+      "loss": 1.0264,
+      "step": 4287
+    },
+    {
+      "epoch": 0.3831393660508857,
+      "grad_norm": 0.38840845227241516,
+      "learning_rate": 7.070850601082577e-05,
+      "loss": 0.9817,
+      "step": 4288
+    },
+    {
+      "epoch": 0.38322871758214755,
+      "grad_norm": 0.43877261877059937,
+      "learning_rate": 7.069533390452766e-05,
+      "loss": 0.9695,
+      "step": 4289
+    },
+    {
+      "epoch": 0.38331806911340943,
+      "grad_norm": 0.42478203773498535,
+      "learning_rate": 7.068216006477476e-05,
+      "loss": 0.9468,
+      "step": 4290
+    },
+    {
+      "epoch": 0.3834074206446713,
+      "grad_norm": 0.5718162059783936,
+      "learning_rate": 7.066898449267047e-05,
+      "loss": 0.851,
+      "step": 4291
+    },
+    {
+      "epoch": 0.38349677217593314,
+      "grad_norm": 0.5737486481666565,
+      "learning_rate": 7.065580718931843e-05,
+      "loss": 0.852,
+      "step": 4292
+    },
+    {
+      "epoch": 0.383586123707195,
+      "grad_norm": 0.4374921917915344,
+      "learning_rate": 7.064262815582238e-05,
+      "loss": 0.8747,
+      "step": 4293
+    },
+    {
+      "epoch": 0.3836754752384569,
+      "grad_norm": 0.47579288482666016,
+      "learning_rate": 7.062944739328616e-05,
+      "loss": 1.02,
+      "step": 4294
+    },
+    {
+      "epoch": 0.3837648267697188,
+      "grad_norm": 0.4117971658706665,
+      "learning_rate": 7.061626490281384e-05,
+      "loss": 0.9578,
+      "step": 4295
+    },
+    {
+      "epoch": 0.3838541783009806,
+      "grad_norm": 0.38990846276283264,
+      "learning_rate": 7.060308068550959e-05,
+      "loss": 0.9143,
+      "step": 4296
+    },
+    {
+      "epoch": 0.3839435298322425,
+      "grad_norm": 0.45375731587409973,
+      "learning_rate": 7.05898947424777e-05,
+      "loss": 0.9272,
+      "step": 4297
+    },
+    {
+      "epoch": 0.3840328813635044,
+      "grad_norm": 0.496006041765213,
+      "learning_rate": 7.057670707482264e-05,
+      "loss": 0.8945,
+      "step": 4298
+    },
+    {
+      "epoch": 0.3841222328947662,
+      "grad_norm": 0.4508313536643982,
+      "learning_rate": 7.056351768364905e-05,
+      "loss": 1.0054,
+      "step": 4299
+    },
+    {
+      "epoch": 0.3842115844260281,
+      "grad_norm": 0.406969279050827,
+      "learning_rate": 7.055032657006166e-05,
+      "loss": 0.967,
+      "step": 4300
+    },
+    {
+      "epoch": 0.38430093595729,
+      "grad_norm": 0.432956725358963,
+      "learning_rate": 7.053713373516538e-05,
+      "loss": 0.9331,
+      "step": 4301
+    },
+    {
+      "epoch": 0.38439028748855186,
+      "grad_norm": 0.5145943760871887,
+      "learning_rate": 7.052393918006522e-05,
+      "loss": 0.9295,
+      "step": 4302
+    },
+    {
+      "epoch": 0.3844796390198137,
+      "grad_norm": 0.40598639845848083,
+      "learning_rate": 7.05107429058664e-05,
+      "loss": 0.99,
+      "step": 4303
+    },
+    {
+      "epoch": 0.3845689905510756,
+      "grad_norm": 0.42250970005989075,
+      "learning_rate": 7.049754491367421e-05,
+      "loss": 0.9821,
+      "step": 4304
+    },
+    {
+      "epoch": 0.38465834208233746,
+      "grad_norm": 0.48077642917633057,
+      "learning_rate": 7.048434520459418e-05,
+      "loss": 0.9702,
+      "step": 4305
+    },
+    {
+      "epoch": 0.3847476936135993,
+      "grad_norm": 0.41635996103286743,
+      "learning_rate": 7.047114377973188e-05,
+      "loss": 0.986,
+      "step": 4306
+    },
+    {
+      "epoch": 0.38483704514486117,
+      "grad_norm": 0.5170190930366516,
+      "learning_rate": 7.045794064019306e-05,
+      "loss": 0.9175,
+      "step": 4307
+    },
+    {
+      "epoch": 0.38492639667612305,
+      "grad_norm": 0.41324326395988464,
+      "learning_rate": 7.044473578708366e-05,
+      "loss": 1.0153,
+      "step": 4308
+    },
+    {
+      "epoch": 0.3850157482073849,
+      "grad_norm": 0.46481555700302124,
+      "learning_rate": 7.043152922150972e-05,
+      "loss": 0.9365,
+      "step": 4309
+    },
+    {
+      "epoch": 0.38510509973864676,
+      "grad_norm": 0.5050130486488342,
+      "learning_rate": 7.041832094457742e-05,
+      "loss": 1.0131,
+      "step": 4310
+    },
+    {
+      "epoch": 0.38519445126990864,
+      "grad_norm": 0.3722282648086548,
+      "learning_rate": 7.04051109573931e-05,
+      "loss": 0.9874,
+      "step": 4311
+    },
+    {
+      "epoch": 0.3852838028011705,
+      "grad_norm": 0.4793398082256317,
+      "learning_rate": 7.039189926106324e-05,
+      "loss": 1.0063,
+      "step": 4312
+    },
+    {
+      "epoch": 0.38537315433243235,
+      "grad_norm": 0.48876821994781494,
+      "learning_rate": 7.037868585669443e-05,
+      "loss": 0.9032,
+      "step": 4313
+    },
+    {
+      "epoch": 0.38546250586369424,
+      "grad_norm": 0.43197745084762573,
+      "learning_rate": 7.036547074539347e-05,
+      "loss": 1.0087,
+      "step": 4314
+    },
+    {
+      "epoch": 0.3855518573949561,
+      "grad_norm": 0.3676876127719879,
+      "learning_rate": 7.035225392826727e-05,
+      "loss": 1.0239,
+      "step": 4315
+    },
+    {
+      "epoch": 0.38564120892621795,
+      "grad_norm": 0.39950066804885864,
+      "learning_rate": 7.033903540642283e-05,
+      "loss": 0.9731,
+      "step": 4316
+    },
+    {
+      "epoch": 0.38573056045747983,
+      "grad_norm": 0.44849759340286255,
+      "learning_rate": 7.032581518096741e-05,
+      "loss": 0.9701,
+      "step": 4317
+    },
+    {
+      "epoch": 0.3858199119887417,
+      "grad_norm": 0.4369259476661682,
+      "learning_rate": 7.03125932530083e-05,
+      "loss": 1.032,
+      "step": 4318
+    },
+    {
+      "epoch": 0.3859092635200036,
+      "grad_norm": 0.40139544010162354,
+      "learning_rate": 7.029936962365297e-05,
+      "loss": 1.0624,
+      "step": 4319
+    },
+    {
+      "epoch": 0.3859986150512654,
+      "grad_norm": 0.4485412836074829,
+      "learning_rate": 7.028614429400908e-05,
+      "loss": 1.0387,
+      "step": 4320
+    },
+    {
+      "epoch": 0.3860879665825273,
+      "grad_norm": 0.44548967480659485,
+      "learning_rate": 7.027291726518436e-05,
+      "loss": 1.0029,
+      "step": 4321
+    },
+    {
+      "epoch": 0.3861773181137892,
+      "grad_norm": 0.40692946314811707,
+      "learning_rate": 7.025968853828674e-05,
+      "loss": 0.959,
+      "step": 4322
+    },
+    {
+      "epoch": 0.386266669645051,
+      "grad_norm": 0.4305645525455475,
+      "learning_rate": 7.024645811442423e-05,
+      "loss": 0.9873,
+      "step": 4323
+    },
+    {
+      "epoch": 0.3863560211763129,
+      "grad_norm": 0.5114787817001343,
+      "learning_rate": 7.023322599470505e-05,
+      "loss": 0.8899,
+      "step": 4324
+    },
+    {
+      "epoch": 0.3864453727075748,
+      "grad_norm": 0.41181930899620056,
+      "learning_rate": 7.021999218023753e-05,
+      "loss": 0.9755,
+      "step": 4325
+    },
+    {
+      "epoch": 0.38653472423883667,
+      "grad_norm": 0.3804915249347687,
+      "learning_rate": 7.020675667213014e-05,
+      "loss": 0.9616,
+      "step": 4326
+    },
+    {
+      "epoch": 0.3866240757700985,
+      "grad_norm": 0.41884955763816833,
+      "learning_rate": 7.019351947149148e-05,
+      "loss": 0.9835,
+      "step": 4327
+    },
+    {
+      "epoch": 0.3867134273013604,
+      "grad_norm": 0.44132474064826965,
+      "learning_rate": 7.018028057943032e-05,
+      "loss": 1.0083,
+      "step": 4328
+    },
+    {
+      "epoch": 0.38680277883262226,
+      "grad_norm": 0.3745604157447815,
+      "learning_rate": 7.016703999705556e-05,
+      "loss": 0.986,
+      "step": 4329
+    },
+    {
+      "epoch": 0.3868921303638841,
+      "grad_norm": 0.39664706587791443,
+      "learning_rate": 7.015379772547623e-05,
+      "loss": 0.9742,
+      "step": 4330
+    },
+    {
+      "epoch": 0.386981481895146,
+      "grad_norm": 0.510838508605957,
+      "learning_rate": 7.014055376580151e-05,
+      "loss": 1.0018,
+      "step": 4331
+    },
+    {
+      "epoch": 0.38707083342640786,
+      "grad_norm": 0.5416032075881958,
+      "learning_rate": 7.012730811914074e-05,
+      "loss": 0.9375,
+      "step": 4332
+    },
+    {
+      "epoch": 0.38716018495766974,
+      "grad_norm": 0.45843154191970825,
+      "learning_rate": 7.011406078660336e-05,
+      "loss": 0.9032,
+      "step": 4333
+    },
+    {
+      "epoch": 0.38724953648893157,
+      "grad_norm": 0.4215903878211975,
+      "learning_rate": 7.0100811769299e-05,
+      "loss": 0.9499,
+      "step": 4334
+    },
+    {
+      "epoch": 0.38733888802019345,
+      "grad_norm": 0.4728352725505829,
+      "learning_rate": 7.008756106833739e-05,
+      "loss": 0.9126,
+      "step": 4335
+    },
+    {
+      "epoch": 0.38742823955145533,
+      "grad_norm": 0.4127133786678314,
+      "learning_rate": 7.007430868482842e-05,
+      "loss": 1.0198,
+      "step": 4336
+    },
+    {
+      "epoch": 0.38751759108271716,
+      "grad_norm": 0.4411311745643616,
+      "learning_rate": 7.006105461988212e-05,
+      "loss": 0.9796,
+      "step": 4337
+    },
+    {
+      "epoch": 0.38760694261397904,
+      "grad_norm": 0.4794873297214508,
+      "learning_rate": 7.004779887460867e-05,
+      "loss": 0.9494,
+      "step": 4338
+    },
+    {
+      "epoch": 0.3876962941452409,
+      "grad_norm": 0.5127260088920593,
+      "learning_rate": 7.003454145011836e-05,
+      "loss": 0.9547,
+      "step": 4339
+    },
+    {
+      "epoch": 0.38778564567650275,
+      "grad_norm": 0.3873635530471802,
+      "learning_rate": 7.002128234752166e-05,
+      "loss": 1.028,
+      "step": 4340
+    },
+    {
+      "epoch": 0.38787499720776464,
+      "grad_norm": 0.4938105344772339,
+      "learning_rate": 7.000802156792915e-05,
+      "loss": 0.9775,
+      "step": 4341
+    },
+    {
+      "epoch": 0.3879643487390265,
+      "grad_norm": 0.43845218420028687,
+      "learning_rate": 6.999475911245156e-05,
+      "loss": 0.945,
+      "step": 4342
+    },
+    {
+      "epoch": 0.3880537002702884,
+      "grad_norm": 0.4035814106464386,
+      "learning_rate": 6.998149498219977e-05,
+      "loss": 1.0355,
+      "step": 4343
+    },
+    {
+      "epoch": 0.38814305180155023,
+      "grad_norm": 0.4540978670120239,
+      "learning_rate": 6.996822917828477e-05,
+      "loss": 0.9927,
+      "step": 4344
+    },
+    {
+      "epoch": 0.3882324033328121,
+      "grad_norm": 0.41308870911598206,
+      "learning_rate": 6.995496170181775e-05,
+      "loss": 0.963,
+      "step": 4345
+    },
+    {
+      "epoch": 0.388321754864074,
+      "grad_norm": 0.40299081802368164,
+      "learning_rate": 6.994169255390999e-05,
+      "loss": 0.9774,
+      "step": 4346
+    },
+    {
+      "epoch": 0.3884111063953358,
+      "grad_norm": 0.5228738784790039,
+      "learning_rate": 6.99284217356729e-05,
+      "loss": 0.9757,
+      "step": 4347
+    },
+    {
+      "epoch": 0.3885004579265977,
+      "grad_norm": 0.4321485757827759,
+      "learning_rate": 6.991514924821807e-05,
+      "loss": 0.9918,
+      "step": 4348
+    },
+    {
+      "epoch": 0.3885898094578596,
+      "grad_norm": 0.4189693331718445,
+      "learning_rate": 6.990187509265721e-05,
+      "loss": 0.9971,
+      "step": 4349
+    },
+    {
+      "epoch": 0.3886791609891215,
+      "grad_norm": 0.5382431745529175,
+      "learning_rate": 6.988859927010219e-05,
+      "loss": 1.0251,
+      "step": 4350
+    },
+    {
+      "epoch": 0.3887685125203833,
+      "grad_norm": 0.47325825691223145,
+      "learning_rate": 6.987532178166496e-05,
+      "loss": 0.9422,
+      "step": 4351
+    },
+    {
+      "epoch": 0.3888578640516452,
+      "grad_norm": 0.441567987203598,
+      "learning_rate": 6.986204262845768e-05,
+      "loss": 0.9758,
+      "step": 4352
+    },
+    {
+      "epoch": 0.38894721558290707,
+      "grad_norm": 0.44492974877357483,
+      "learning_rate": 6.984876181159261e-05,
+      "loss": 0.9484,
+      "step": 4353
+    },
+    {
+      "epoch": 0.3890365671141689,
+      "grad_norm": 0.4059096872806549,
+      "learning_rate": 6.983547933218218e-05,
+      "loss": 1.0436,
+      "step": 4354
+    },
+    {
+      "epoch": 0.3891259186454308,
+      "grad_norm": 0.4348817765712738,
+      "learning_rate": 6.982219519133892e-05,
+      "loss": 0.9641,
+      "step": 4355
+    },
+    {
+      "epoch": 0.38921527017669266,
+      "grad_norm": 0.4612578749656677,
+      "learning_rate": 6.980890939017551e-05,
+      "loss": 0.9583,
+      "step": 4356
+    },
+    {
+      "epoch": 0.38930462170795455,
+      "grad_norm": 0.4391173720359802,
+      "learning_rate": 6.979562192980481e-05,
+      "loss": 0.9592,
+      "step": 4357
+    },
+    {
+      "epoch": 0.3893939732392164,
+      "grad_norm": 0.45531633496284485,
+      "learning_rate": 6.978233281133976e-05,
+      "loss": 0.9537,
+      "step": 4358
+    },
+    {
+      "epoch": 0.38948332477047826,
+      "grad_norm": 0.39362677931785583,
+      "learning_rate": 6.976904203589346e-05,
+      "loss": 1.0051,
+      "step": 4359
+    },
+    {
+      "epoch": 0.38957267630174014,
+      "grad_norm": 0.5158856511116028,
+      "learning_rate": 6.975574960457919e-05,
+      "loss": 1.0024,
+      "step": 4360
+    },
+    {
+      "epoch": 0.38966202783300197,
+      "grad_norm": 0.5970863699913025,
+      "learning_rate": 6.974245551851029e-05,
+      "loss": 0.9359,
+      "step": 4361
+    },
+    {
+      "epoch": 0.38975137936426385,
+      "grad_norm": 0.40095895528793335,
+      "learning_rate": 6.972915977880031e-05,
+      "loss": 1.0245,
+      "step": 4362
+    },
+    {
+      "epoch": 0.38984073089552573,
+      "grad_norm": 0.46378135681152344,
+      "learning_rate": 6.971586238656291e-05,
+      "loss": 0.9918,
+      "step": 4363
+    },
+    {
+      "epoch": 0.3899300824267876,
+      "grad_norm": 0.5374437570571899,
+      "learning_rate": 6.970256334291187e-05,
+      "loss": 1.0324,
+      "step": 4364
+    },
+    {
+      "epoch": 0.39001943395804944,
+      "grad_norm": 0.5034531950950623,
+      "learning_rate": 6.968926264896114e-05,
+      "loss": 0.8832,
+      "step": 4365
+    },
+    {
+      "epoch": 0.3901087854893113,
+      "grad_norm": 0.3909834921360016,
+      "learning_rate": 6.967596030582478e-05,
+      "loss": 0.9664,
+      "step": 4366
+    },
+    {
+      "epoch": 0.3901981370205732,
+      "grad_norm": 0.4428032338619232,
+      "learning_rate": 6.966265631461703e-05,
+      "loss": 0.9607,
+      "step": 4367
+    },
+    {
+      "epoch": 0.39028748855183504,
+      "grad_norm": 0.6216595768928528,
+      "learning_rate": 6.96493506764522e-05,
+      "loss": 0.9773,
+      "step": 4368
+    },
+    {
+      "epoch": 0.3903768400830969,
+      "grad_norm": 0.44318845868110657,
+      "learning_rate": 6.963604339244481e-05,
+      "loss": 0.9698,
+      "step": 4369
+    },
+    {
+      "epoch": 0.3904661916143588,
+      "grad_norm": 0.4492654800415039,
+      "learning_rate": 6.96227344637095e-05,
+      "loss": 0.9126,
+      "step": 4370
+    },
+    {
+      "epoch": 0.39055554314562063,
+      "grad_norm": 0.46567076444625854,
+      "learning_rate": 6.960942389136101e-05,
+      "loss": 0.9868,
+      "step": 4371
+    },
+    {
+      "epoch": 0.3906448946768825,
+      "grad_norm": 0.4394150674343109,
+      "learning_rate": 6.959611167651423e-05,
+      "loss": 0.9878,
+      "step": 4372
+    },
+    {
+      "epoch": 0.3907342462081444,
+      "grad_norm": 0.44331878423690796,
+      "learning_rate": 6.958279782028423e-05,
+      "loss": 0.9312,
+      "step": 4373
+    },
+    {
+      "epoch": 0.3908235977394063,
+      "grad_norm": 0.4242664575576782,
+      "learning_rate": 6.956948232378617e-05,
+      "loss": 1.0962,
+      "step": 4374
+    },
+    {
+      "epoch": 0.3909129492706681,
+      "grad_norm": 0.44841963052749634,
+      "learning_rate": 6.955616518813538e-05,
+      "loss": 1.0338,
+      "step": 4375
+    },
+    {
+      "epoch": 0.39100230080193,
+      "grad_norm": 0.40618273615837097,
+      "learning_rate": 6.95428464144473e-05,
+      "loss": 0.9697,
+      "step": 4376
+    },
+    {
+      "epoch": 0.3910916523331919,
+      "grad_norm": 0.41985857486724854,
+      "learning_rate": 6.952952600383752e-05,
+      "loss": 0.9683,
+      "step": 4377
+    },
+    {
+      "epoch": 0.3911810038644537,
+      "grad_norm": 0.5133029222488403,
+      "learning_rate": 6.951620395742176e-05,
+      "loss": 0.953,
+      "step": 4378
+    },
+    {
+      "epoch": 0.3912703553957156,
+      "grad_norm": 0.42223209142684937,
+      "learning_rate": 6.950288027631588e-05,
+      "loss": 1.063,
+      "step": 4379
+    },
+    {
+      "epoch": 0.39135970692697747,
+      "grad_norm": 0.4248124659061432,
+      "learning_rate": 6.948955496163593e-05,
+      "loss": 1.0083,
+      "step": 4380
+    },
+    {
+      "epoch": 0.39144905845823935,
+      "grad_norm": 0.389886736869812,
+      "learning_rate": 6.947622801449799e-05,
+      "loss": 1.0032,
+      "step": 4381
+    },
+    {
+      "epoch": 0.3915384099895012,
+      "grad_norm": 0.4102005660533905,
+      "learning_rate": 6.946289943601833e-05,
+      "loss": 0.9879,
+      "step": 4382
+    },
+    {
+      "epoch": 0.39162776152076306,
+      "grad_norm": 0.4157712757587433,
+      "learning_rate": 6.94495692273134e-05,
+      "loss": 0.9817,
+      "step": 4383
+    },
+    {
+      "epoch": 0.39171711305202495,
+      "grad_norm": 0.42993542551994324,
+      "learning_rate": 6.943623738949973e-05,
+      "loss": 1.0497,
+      "step": 4384
+    },
+    {
+      "epoch": 0.3918064645832868,
+      "grad_norm": 0.46929094195365906,
+      "learning_rate": 6.942290392369401e-05,
+      "loss": 0.945,
+      "step": 4385
+    },
+    {
+      "epoch": 0.39189581611454866,
+      "grad_norm": 0.41299015283584595,
+      "learning_rate": 6.940956883101304e-05,
+      "loss": 0.9762,
+      "step": 4386
+    },
+    {
+      "epoch": 0.39198516764581054,
+      "grad_norm": 0.4164472222328186,
+      "learning_rate": 6.93962321125738e-05,
+      "loss": 0.8936,
+      "step": 4387
+    },
+    {
+      "epoch": 0.3920745191770724,
+      "grad_norm": 0.5273587107658386,
+      "learning_rate": 6.938289376949336e-05,
+      "loss": 0.9785,
+      "step": 4388
+    },
+    {
+      "epoch": 0.39216387070833425,
+      "grad_norm": 0.5239537954330444,
+      "learning_rate": 6.936955380288897e-05,
+      "loss": 0.9617,
+      "step": 4389
+    },
+    {
+      "epoch": 0.39225322223959613,
+      "grad_norm": 0.3496021628379822,
+      "learning_rate": 6.935621221387797e-05,
+      "loss": 1.0419,
+      "step": 4390
+    },
+    {
+      "epoch": 0.392342573770858,
+      "grad_norm": 0.5209729075431824,
+      "learning_rate": 6.934286900357789e-05,
+      "loss": 0.9904,
+      "step": 4391
+    },
+    {
+      "epoch": 0.39243192530211984,
+      "grad_norm": 0.4754767119884491,
+      "learning_rate": 6.932952417310634e-05,
+      "loss": 0.9543,
+      "step": 4392
+    },
+    {
+      "epoch": 0.3925212768333817,
+      "grad_norm": 0.4421217739582062,
+      "learning_rate": 6.931617772358112e-05,
+      "loss": 0.9329,
+      "step": 4393
+    },
+    {
+      "epoch": 0.3926106283646436,
+      "grad_norm": 0.40273427963256836,
+      "learning_rate": 6.93028296561201e-05,
+      "loss": 0.9904,
+      "step": 4394
+    },
+    {
+      "epoch": 0.3926999798959055,
+      "grad_norm": 0.4427558183670044,
+      "learning_rate": 6.928947997184134e-05,
+      "loss": 0.9429,
+      "step": 4395
+    },
+    {
+      "epoch": 0.3927893314271673,
+      "grad_norm": 0.46975377202033997,
+      "learning_rate": 6.927612867186305e-05,
+      "loss": 0.9122,
+      "step": 4396
+    },
+    {
+      "epoch": 0.3928786829584292,
+      "grad_norm": 0.3829523026943207,
+      "learning_rate": 6.926277575730349e-05,
+      "loss": 0.9974,
+      "step": 4397
+    },
+    {
+      "epoch": 0.3929680344896911,
+      "grad_norm": 0.5063918828964233,
+      "learning_rate": 6.924942122928115e-05,
+      "loss": 0.9954,
+      "step": 4398
+    },
+    {
+      "epoch": 0.3930573860209529,
+      "grad_norm": 0.4212724566459656,
+      "learning_rate": 6.923606508891459e-05,
+      "loss": 1.0462,
+      "step": 4399
+    },
+    {
+      "epoch": 0.3931467375522148,
+      "grad_norm": 0.44579607248306274,
+      "learning_rate": 6.922270733732254e-05,
+      "loss": 0.9423,
+      "step": 4400
+    },
+    {
+      "epoch": 0.3932360890834767,
+      "grad_norm": 0.45305055379867554,
+      "learning_rate": 6.920934797562385e-05,
+      "loss": 0.9498,
+      "step": 4401
+    },
+    {
+      "epoch": 0.3933254406147385,
+      "grad_norm": 0.39803916215896606,
+      "learning_rate": 6.91959870049375e-05,
+      "loss": 0.9772,
+      "step": 4402
+    },
+    {
+      "epoch": 0.3934147921460004,
+      "grad_norm": 0.5116259455680847,
+      "learning_rate": 6.918262442638263e-05,
+      "loss": 0.9202,
+      "step": 4403
+    },
+    {
+      "epoch": 0.3935041436772623,
+      "grad_norm": 0.39163100719451904,
+      "learning_rate": 6.91692602410785e-05,
+      "loss": 0.9933,
+      "step": 4404
+    },
+    {
+      "epoch": 0.39359349520852416,
+      "grad_norm": 0.386976420879364,
+      "learning_rate": 6.915589445014448e-05,
+      "loss": 1.006,
+      "step": 4405
+    },
+    {
+      "epoch": 0.393682846739786,
+      "grad_norm": 0.4204871356487274,
+      "learning_rate": 6.914252705470013e-05,
+      "loss": 0.9525,
+      "step": 4406
+    },
+    {
+      "epoch": 0.39377219827104787,
+      "grad_norm": 0.5166035294532776,
+      "learning_rate": 6.912915805586509e-05,
+      "loss": 0.8921,
+      "step": 4407
+    },
+    {
+      "epoch": 0.39386154980230975,
+      "grad_norm": 0.4778018891811371,
+      "learning_rate": 6.911578745475915e-05,
+      "loss": 0.9687,
+      "step": 4408
+    },
+    {
+      "epoch": 0.3939509013335716,
+      "grad_norm": 0.4855932295322418,
+      "learning_rate": 6.910241525250225e-05,
+      "loss": 0.9592,
+      "step": 4409
+    },
+    {
+      "epoch": 0.39404025286483346,
+      "grad_norm": 0.4324698746204376,
+      "learning_rate": 6.908904145021447e-05,
+      "loss": 0.9615,
+      "step": 4410
+    },
+    {
+      "epoch": 0.39412960439609535,
+      "grad_norm": 0.41898903250694275,
+      "learning_rate": 6.907566604901599e-05,
+      "loss": 0.9127,
+      "step": 4411
+    },
+    {
+      "epoch": 0.39421895592735723,
+      "grad_norm": 0.44606366753578186,
+      "learning_rate": 6.906228905002714e-05,
+      "loss": 0.9715,
+      "step": 4412
+    },
+    {
+      "epoch": 0.39430830745861906,
+      "grad_norm": 0.4419662356376648,
+      "learning_rate": 6.90489104543684e-05,
+      "loss": 0.9972,
+      "step": 4413
+    },
+    {
+      "epoch": 0.39439765898988094,
+      "grad_norm": 0.43655163049697876,
+      "learning_rate": 6.903553026316036e-05,
+      "loss": 0.9968,
+      "step": 4414
+    },
+    {
+      "epoch": 0.3944870105211428,
+      "grad_norm": 0.41895565390586853,
+      "learning_rate": 6.902214847752375e-05,
+      "loss": 0.9499,
+      "step": 4415
+    },
+    {
+      "epoch": 0.39457636205240465,
+      "grad_norm": 0.403507798910141,
+      "learning_rate": 6.900876509857945e-05,
+      "loss": 0.9934,
+      "step": 4416
+    },
+    {
+      "epoch": 0.39466571358366653,
+      "grad_norm": 0.4834029972553253,
+      "learning_rate": 6.899538012744847e-05,
+      "loss": 0.9242,
+      "step": 4417
+    },
+    {
+      "epoch": 0.3947550651149284,
+      "grad_norm": 0.46136564016342163,
+      "learning_rate": 6.89819935652519e-05,
+      "loss": 0.9524,
+      "step": 4418
+    },
+    {
+      "epoch": 0.3948444166461903,
+      "grad_norm": 0.4441271424293518,
+      "learning_rate": 6.896860541311105e-05,
+      "loss": 0.9916,
+      "step": 4419
+    },
+    {
+      "epoch": 0.3949337681774521,
+      "grad_norm": 0.5411950349807739,
+      "learning_rate": 6.89552156721473e-05,
+      "loss": 0.8352,
+      "step": 4420
+    },
+    {
+      "epoch": 0.395023119708714,
+      "grad_norm": 0.43596935272216797,
+      "learning_rate": 6.89418243434822e-05,
+      "loss": 0.9776,
+      "step": 4421
+    },
+    {
+      "epoch": 0.3951124712399759,
+      "grad_norm": 0.4356800317764282,
+      "learning_rate": 6.892843142823739e-05,
+      "loss": 0.9779,
+      "step": 4422
+    },
+    {
+      "epoch": 0.3952018227712377,
+      "grad_norm": 0.4590302109718323,
+      "learning_rate": 6.891503692753469e-05,
+      "loss": 1.0105,
+      "step": 4423
+    },
+    {
+      "epoch": 0.3952911743024996,
+      "grad_norm": 0.3544255793094635,
+      "learning_rate": 6.890164084249602e-05,
+      "loss": 1.0239,
+      "step": 4424
+    },
+    {
+      "epoch": 0.3953805258337615,
+      "grad_norm": 0.4187402129173279,
+      "learning_rate": 6.888824317424347e-05,
+      "loss": 1.0746,
+      "step": 4425
+    },
+    {
+      "epoch": 0.39546987736502337,
+      "grad_norm": 0.5059707760810852,
+      "learning_rate": 6.887484392389922e-05,
+      "loss": 0.7944,
+      "step": 4426
+    },
+    {
+      "epoch": 0.3955592288962852,
+      "grad_norm": 0.4193587601184845,
+      "learning_rate": 6.88614430925856e-05,
+      "loss": 1.0126,
+      "step": 4427
+    },
+    {
+      "epoch": 0.3956485804275471,
+      "grad_norm": 0.4689008891582489,
+      "learning_rate": 6.884804068142505e-05,
+      "loss": 0.9036,
+      "step": 4428
+    },
+    {
+      "epoch": 0.39573793195880896,
+      "grad_norm": 0.4160483479499817,
+      "learning_rate": 6.88346366915402e-05,
+      "loss": 0.9379,
+      "step": 4429
+    },
+    {
+      "epoch": 0.3958272834900708,
+      "grad_norm": 0.4379422068595886,
+      "learning_rate": 6.882123112405376e-05,
+      "loss": 0.9573,
+      "step": 4430
+    },
+    {
+      "epoch": 0.3959166350213327,
+      "grad_norm": 0.4520152509212494,
+      "learning_rate": 6.880782398008862e-05,
+      "loss": 0.9565,
+      "step": 4431
+    },
+    {
+      "epoch": 0.39600598655259456,
+      "grad_norm": 0.4579842686653137,
+      "learning_rate": 6.879441526076771e-05,
+      "loss": 0.9489,
+      "step": 4432
+    },
+    {
+      "epoch": 0.3960953380838564,
+      "grad_norm": 0.49286845326423645,
+      "learning_rate": 6.878100496721423e-05,
+      "loss": 0.9866,
+      "step": 4433
+    },
+    {
+      "epoch": 0.39618468961511827,
+      "grad_norm": 0.39000403881073,
+      "learning_rate": 6.876759310055135e-05,
+      "loss": 1.0161,
+      "step": 4434
+    },
+    {
+      "epoch": 0.39627404114638015,
+      "grad_norm": 0.3937504291534424,
+      "learning_rate": 6.875417966190251e-05,
+      "loss": 0.975,
+      "step": 4435
+    },
+    {
+      "epoch": 0.39636339267764203,
+      "grad_norm": 0.504807710647583,
+      "learning_rate": 6.874076465239123e-05,
+      "loss": 0.8956,
+      "step": 4436
+    },
+    {
+      "epoch": 0.39645274420890386,
+      "grad_norm": 0.38329121470451355,
+      "learning_rate": 6.872734807314115e-05,
+      "loss": 1.0308,
+      "step": 4437
+    },
+    {
+      "epoch": 0.39654209574016575,
+      "grad_norm": 0.4224548637866974,
+      "learning_rate": 6.871392992527603e-05,
+      "loss": 0.9944,
+      "step": 4438
+    },
+    {
+      "epoch": 0.39663144727142763,
+      "grad_norm": 0.44873788952827454,
+      "learning_rate": 6.870051020991981e-05,
+      "loss": 1.0696,
+      "step": 4439
+    },
+    {
+      "epoch": 0.39672079880268946,
+      "grad_norm": 0.49590742588043213,
+      "learning_rate": 6.868708892819653e-05,
+      "loss": 0.9732,
+      "step": 4440
+    },
+    {
+      "epoch": 0.39681015033395134,
+      "grad_norm": 0.42665350437164307,
+      "learning_rate": 6.867366608123038e-05,
+      "loss": 0.9137,
+      "step": 4441
+    },
+    {
+      "epoch": 0.3968995018652132,
+      "grad_norm": 0.3908413052558899,
+      "learning_rate": 6.866024167014562e-05,
+      "loss": 1.0205,
+      "step": 4442
+    },
+    {
+      "epoch": 0.3969888533964751,
+      "grad_norm": 0.4211508631706238,
+      "learning_rate": 6.864681569606673e-05,
+      "loss": 1.0103,
+      "step": 4443
+    },
+    {
+      "epoch": 0.39707820492773693,
+      "grad_norm": 0.4399520754814148,
+      "learning_rate": 6.863338816011826e-05,
+      "loss": 1.0127,
+      "step": 4444
+    },
+    {
+      "epoch": 0.3971675564589988,
+      "grad_norm": 0.37878918647766113,
+      "learning_rate": 6.861995906342491e-05,
+      "loss": 0.9763,
+      "step": 4445
+    },
+    {
+      "epoch": 0.3972569079902607,
+      "grad_norm": 0.47828400135040283,
+      "learning_rate": 6.860652840711154e-05,
+      "loss": 0.955,
+      "step": 4446
+    },
+    {
+      "epoch": 0.3973462595215225,
+      "grad_norm": 0.45073986053466797,
+      "learning_rate": 6.859309619230305e-05,
+      "loss": 0.9025,
+      "step": 4447
+    },
+    {
+      "epoch": 0.3974356110527844,
+      "grad_norm": 0.38527220487594604,
+      "learning_rate": 6.85796624201246e-05,
+      "loss": 0.9831,
+      "step": 4448
+    },
+    {
+      "epoch": 0.3975249625840463,
+      "grad_norm": 0.4932858347892761,
+      "learning_rate": 6.856622709170134e-05,
+      "loss": 0.9358,
+      "step": 4449
+    },
+    {
+      "epoch": 0.3976143141153082,
+      "grad_norm": 0.5208497047424316,
+      "learning_rate": 6.855279020815868e-05,
+      "loss": 0.9976,
+      "step": 4450
+    },
+    {
+      "epoch": 0.39770366564657,
+      "grad_norm": 0.40091952681541443,
+      "learning_rate": 6.853935177062209e-05,
+      "loss": 0.9934,
+      "step": 4451
+    },
+    {
+      "epoch": 0.3977930171778319,
+      "grad_norm": 0.4903776943683624,
+      "learning_rate": 6.852591178021716e-05,
+      "loss": 0.9286,
+      "step": 4452
+    },
+    {
+      "epoch": 0.39788236870909377,
+      "grad_norm": 0.45064613223075867,
+      "learning_rate": 6.851247023806964e-05,
+      "loss": 0.9525,
+      "step": 4453
+    },
+    {
+      "epoch": 0.3979717202403556,
+      "grad_norm": 0.49910131096839905,
+      "learning_rate": 6.849902714530542e-05,
+      "loss": 0.971,
+      "step": 4454
+    },
+    {
+      "epoch": 0.3980610717716175,
+      "grad_norm": 0.4719197452068329,
+      "learning_rate": 6.84855825030505e-05,
+      "loss": 0.9095,
+      "step": 4455
+    },
+    {
+      "epoch": 0.39815042330287936,
+      "grad_norm": 0.4351549744606018,
+      "learning_rate": 6.847213631243099e-05,
+      "loss": 0.8641,
+      "step": 4456
+    },
+    {
+      "epoch": 0.39823977483414125,
+      "grad_norm": 0.37931522727012634,
+      "learning_rate": 6.845868857457316e-05,
+      "loss": 0.9773,
+      "step": 4457
+    },
+    {
+      "epoch": 0.3983291263654031,
+      "grad_norm": 0.49004626274108887,
+      "learning_rate": 6.844523929060343e-05,
+      "loss": 1.012,
+      "step": 4458
+    },
+    {
+      "epoch": 0.39841847789666496,
+      "grad_norm": 0.36821407079696655,
+      "learning_rate": 6.843178846164826e-05,
+      "loss": 1.0009,
+      "step": 4459
+    },
+    {
+      "epoch": 0.39850782942792684,
+      "grad_norm": 0.42328307032585144,
+      "learning_rate": 6.841833608883437e-05,
+      "loss": 0.9568,
+      "step": 4460
+    },
+    {
+      "epoch": 0.39859718095918867,
+      "grad_norm": 0.4346807897090912,
+      "learning_rate": 6.840488217328848e-05,
+      "loss": 1.0407,
+      "step": 4461
+    },
+    {
+      "epoch": 0.39868653249045055,
+      "grad_norm": 0.35460466146469116,
+      "learning_rate": 6.839142671613757e-05,
+      "loss": 1.0353,
+      "step": 4462
+    },
+    {
+      "epoch": 0.39877588402171243,
+      "grad_norm": 0.43105101585388184,
+      "learning_rate": 6.837796971850859e-05,
+      "loss": 1.0479,
+      "step": 4463
+    },
+    {
+      "epoch": 0.39886523555297426,
+      "grad_norm": 0.43495675921440125,
+      "learning_rate": 6.836451118152877e-05,
+      "loss": 0.9813,
+      "step": 4464
+    },
+    {
+      "epoch": 0.39895458708423615,
+      "grad_norm": 0.4951534867286682,
+      "learning_rate": 6.835105110632539e-05,
+      "loss": 0.9718,
+      "step": 4465
+    },
+    {
+      "epoch": 0.39904393861549803,
+      "grad_norm": 0.48780739307403564,
+      "learning_rate": 6.833758949402587e-05,
+      "loss": 0.9653,
+      "step": 4466
+    },
+    {
+      "epoch": 0.3991332901467599,
+      "grad_norm": 0.3795831799507141,
+      "learning_rate": 6.832412634575774e-05,
+      "loss": 1.0119,
+      "step": 4467
+    },
+    {
+      "epoch": 0.39922264167802174,
+      "grad_norm": 0.36188191175460815,
+      "learning_rate": 6.831066166264874e-05,
+      "loss": 0.979,
+      "step": 4468
+    },
+    {
+      "epoch": 0.3993119932092836,
+      "grad_norm": 0.4243720769882202,
+      "learning_rate": 6.829719544582665e-05,
+      "loss": 1.0764,
+      "step": 4469
+    },
+    {
+      "epoch": 0.3994013447405455,
+      "grad_norm": 0.3986402750015259,
+      "learning_rate": 6.828372769641938e-05,
+      "loss": 1.0687,
+      "step": 4470
+    },
+    {
+      "epoch": 0.39949069627180733,
+      "grad_norm": 0.48252764344215393,
+      "learning_rate": 6.827025841555504e-05,
+      "loss": 0.9346,
+      "step": 4471
+    },
+    {
+      "epoch": 0.3995800478030692,
+      "grad_norm": 0.41776660084724426,
+      "learning_rate": 6.825678760436182e-05,
+      "loss": 0.9995,
+      "step": 4472
+    },
+    {
+      "epoch": 0.3996693993343311,
+      "grad_norm": 0.48631688952445984,
+      "learning_rate": 6.824331526396801e-05,
+      "loss": 1.0179,
+      "step": 4473
+    },
+    {
+      "epoch": 0.399758750865593,
+      "grad_norm": 0.39176318049430847,
+      "learning_rate": 6.82298413955021e-05,
+      "loss": 1.0147,
+      "step": 4474
+    },
+    {
+      "epoch": 0.3998481023968548,
+      "grad_norm": 0.43048596382141113,
+      "learning_rate": 6.821636600009266e-05,
+      "loss": 0.9092,
+      "step": 4475
+    },
+    {
+      "epoch": 0.3999374539281167,
+      "grad_norm": 0.445742130279541,
+      "learning_rate": 6.82028890788684e-05,
+      "loss": 0.9472,
+      "step": 4476
+    },
+    {
+      "epoch": 0.4000268054593786,
+      "grad_norm": 0.41645416617393494,
+      "learning_rate": 6.818941063295815e-05,
+      "loss": 0.9294,
+      "step": 4477
+    },
+    {
+      "epoch": 0.4001161569906404,
+      "grad_norm": 0.5943310260772705,
+      "learning_rate": 6.817593066349086e-05,
+      "loss": 0.9001,
+      "step": 4478
+    },
+    {
+      "epoch": 0.4002055085219023,
+      "grad_norm": 0.39284154772758484,
+      "learning_rate": 6.816244917159564e-05,
+      "loss": 0.9755,
+      "step": 4479
+    },
+    {
+      "epoch": 0.40029486005316417,
+      "grad_norm": 0.41957008838653564,
+      "learning_rate": 6.814896615840171e-05,
+      "loss": 0.9493,
+      "step": 4480
+    },
+    {
+      "epoch": 0.40038421158442605,
+      "grad_norm": 0.4775201082229614,
+      "learning_rate": 6.81354816250384e-05,
+      "loss": 1.0124,
+      "step": 4481
+    },
+    {
+      "epoch": 0.4004735631156879,
+      "grad_norm": 0.4292440116405487,
+      "learning_rate": 6.812199557263522e-05,
+      "loss": 0.9147,
+      "step": 4482
+    },
+    {
+      "epoch": 0.40056291464694976,
+      "grad_norm": 0.4357702434062958,
+      "learning_rate": 6.810850800232172e-05,
+      "loss": 0.9812,
+      "step": 4483
+    },
+    {
+      "epoch": 0.40065226617821165,
+      "grad_norm": 0.525195837020874,
+      "learning_rate": 6.809501891522765e-05,
+      "loss": 0.9529,
+      "step": 4484
+    },
+    {
+      "epoch": 0.4007416177094735,
+      "grad_norm": 0.42463192343711853,
+      "learning_rate": 6.808152831248287e-05,
+      "loss": 0.9269,
+      "step": 4485
+    },
+    {
+      "epoch": 0.40083096924073536,
+      "grad_norm": 0.4291436970233917,
+      "learning_rate": 6.806803619521737e-05,
+      "loss": 1.0478,
+      "step": 4486
+    },
+    {
+      "epoch": 0.40092032077199724,
+      "grad_norm": 0.4439351260662079,
+      "learning_rate": 6.805454256456125e-05,
+      "loss": 0.9217,
+      "step": 4487
+    },
+    {
+      "epoch": 0.4010096723032591,
+      "grad_norm": 0.5323197245597839,
+      "learning_rate": 6.804104742164472e-05,
+      "loss": 0.977,
+      "step": 4488
+    },
+    {
+      "epoch": 0.40109902383452095,
+      "grad_norm": 0.40645474195480347,
+      "learning_rate": 6.802755076759819e-05,
+      "loss": 0.9972,
+      "step": 4489
+    },
+    {
+      "epoch": 0.40118837536578283,
+      "grad_norm": 0.4209262728691101,
+      "learning_rate": 6.80140526035521e-05,
+      "loss": 0.9412,
+      "step": 4490
+    },
+    {
+      "epoch": 0.4012777268970447,
+      "grad_norm": 0.504423201084137,
+      "learning_rate": 6.80005529306371e-05,
+      "loss": 0.9029,
+      "step": 4491
+    },
+    {
+      "epoch": 0.40136707842830655,
+      "grad_norm": 0.4125538170337677,
+      "learning_rate": 6.798705174998392e-05,
+      "loss": 0.9586,
+      "step": 4492
+    },
+    {
+      "epoch": 0.40145642995956843,
+      "grad_norm": 0.36721163988113403,
+      "learning_rate": 6.797354906272342e-05,
+      "loss": 1.0099,
+      "step": 4493
+    },
+    {
+      "epoch": 0.4015457814908303,
+      "grad_norm": 0.4413255453109741,
+      "learning_rate": 6.796004486998661e-05,
+      "loss": 0.9163,
+      "step": 4494
+    },
+    {
+      "epoch": 0.40163513302209214,
+      "grad_norm": 0.44812944531440735,
+      "learning_rate": 6.79465391729046e-05,
+      "loss": 0.967,
+      "step": 4495
+    },
+    {
+      "epoch": 0.401724484553354,
+      "grad_norm": 0.38190940022468567,
+      "learning_rate": 6.793303197260864e-05,
+      "loss": 0.9877,
+      "step": 4496
+    },
+    {
+      "epoch": 0.4018138360846159,
+      "grad_norm": 0.44277167320251465,
+      "learning_rate": 6.79195232702301e-05,
+      "loss": 0.956,
+      "step": 4497
+    },
+    {
+      "epoch": 0.4019031876158778,
+      "grad_norm": 0.41372618079185486,
+      "learning_rate": 6.790601306690048e-05,
+      "loss": 1.003,
+      "step": 4498
+    },
+    {
+      "epoch": 0.4019925391471396,
+      "grad_norm": 0.39554059505462646,
+      "learning_rate": 6.78925013637514e-05,
+      "loss": 1.0137,
+      "step": 4499
+    },
+    {
+      "epoch": 0.4020818906784015,
+      "grad_norm": 0.3876018226146698,
+      "learning_rate": 6.78789881619146e-05,
+      "loss": 0.964,
+      "step": 4500
+    },
+    {
+      "epoch": 0.4021712422096634,
+      "grad_norm": 0.5051389932632446,
+      "learning_rate": 6.786547346252198e-05,
+      "loss": 0.9447,
+      "step": 4501
+    },
+    {
+      "epoch": 0.4022605937409252,
+      "grad_norm": 0.49276506900787354,
+      "learning_rate": 6.785195726670552e-05,
+      "loss": 0.949,
+      "step": 4502
+    },
+    {
+      "epoch": 0.4023499452721871,
+      "grad_norm": 0.5032357573509216,
+      "learning_rate": 6.783843957559734e-05,
+      "loss": 0.9475,
+      "step": 4503
+    },
+    {
+      "epoch": 0.402439296803449,
+      "grad_norm": 0.42340540885925293,
+      "learning_rate": 6.782492039032971e-05,
+      "loss": 0.9155,
+      "step": 4504
+    },
+    {
+      "epoch": 0.40252864833471086,
+      "grad_norm": 0.4485189914703369,
+      "learning_rate": 6.7811399712035e-05,
+      "loss": 0.8825,
+      "step": 4505
+    },
+    {
+      "epoch": 0.4026179998659727,
+      "grad_norm": 0.4616818130016327,
+      "learning_rate": 6.779787754184571e-05,
+      "loss": 0.9936,
+      "step": 4506
+    },
+    {
+      "epoch": 0.40270735139723457,
+      "grad_norm": 0.44533130526542664,
+      "learning_rate": 6.778435388089446e-05,
+      "loss": 0.9267,
+      "step": 4507
+    },
+    {
+      "epoch": 0.40279670292849645,
+      "grad_norm": 0.4395417869091034,
+      "learning_rate": 6.777082873031401e-05,
+      "loss": 0.9568,
+      "step": 4508
+    },
+    {
+      "epoch": 0.4028860544597583,
+      "grad_norm": 0.44981324672698975,
+      "learning_rate": 6.775730209123722e-05,
+      "loss": 0.9014,
+      "step": 4509
+    },
+    {
+      "epoch": 0.40297540599102016,
+      "grad_norm": 0.4732804596424103,
+      "learning_rate": 6.77437739647971e-05,
+      "loss": 0.9897,
+      "step": 4510
+    },
+    {
+      "epoch": 0.40306475752228205,
+      "grad_norm": 0.4026157557964325,
+      "learning_rate": 6.773024435212678e-05,
+      "loss": 1.0569,
+      "step": 4511
+    },
+    {
+      "epoch": 0.40315410905354393,
+      "grad_norm": 0.39510583877563477,
+      "learning_rate": 6.771671325435952e-05,
+      "loss": 0.9722,
+      "step": 4512
+    },
+    {
+      "epoch": 0.40324346058480576,
+      "grad_norm": 0.4581720232963562,
+      "learning_rate": 6.770318067262866e-05,
+      "loss": 0.9051,
+      "step": 4513
+    },
+    {
+      "epoch": 0.40333281211606764,
+      "grad_norm": 0.5355126261711121,
+      "learning_rate": 6.768964660806772e-05,
+      "loss": 0.9467,
+      "step": 4514
+    },
+    {
+      "epoch": 0.4034221636473295,
+      "grad_norm": 0.38436293601989746,
+      "learning_rate": 6.767611106181031e-05,
+      "loss": 0.9994,
+      "step": 4515
+    },
+    {
+      "epoch": 0.40351151517859135,
+      "grad_norm": 0.4458340108394623,
+      "learning_rate": 6.766257403499019e-05,
+      "loss": 0.9685,
+      "step": 4516
+    },
+    {
+      "epoch": 0.40360086670985323,
+      "grad_norm": 0.5055366158485413,
+      "learning_rate": 6.764903552874125e-05,
+      "loss": 0.9154,
+      "step": 4517
+    },
+    {
+      "epoch": 0.4036902182411151,
+      "grad_norm": 0.4424796998500824,
+      "learning_rate": 6.763549554419743e-05,
+      "loss": 0.9438,
+      "step": 4518
+    },
+    {
+      "epoch": 0.403779569772377,
+      "grad_norm": 0.4608469605445862,
+      "learning_rate": 6.762195408249288e-05,
+      "loss": 0.9699,
+      "step": 4519
+    },
+    {
+      "epoch": 0.40386892130363883,
+      "grad_norm": 0.4548738896846771,
+      "learning_rate": 6.760841114476185e-05,
+      "loss": 1.0052,
+      "step": 4520
+    },
+    {
+      "epoch": 0.4039582728349007,
+      "grad_norm": 0.5669234395027161,
+      "learning_rate": 6.759486673213869e-05,
+      "loss": 1.0287,
+      "step": 4521
+    },
+    {
+      "epoch": 0.4040476243661626,
+      "grad_norm": 0.4517204165458679,
+      "learning_rate": 6.75813208457579e-05,
+      "loss": 0.909,
+      "step": 4522
+    },
+    {
+      "epoch": 0.4041369758974244,
+      "grad_norm": 0.5192089080810547,
+      "learning_rate": 6.756777348675407e-05,
+      "loss": 0.9845,
+      "step": 4523
+    },
+    {
+      "epoch": 0.4042263274286863,
+      "grad_norm": 0.4186631739139557,
+      "learning_rate": 6.755422465626196e-05,
+      "loss": 0.9981,
+      "step": 4524
+    },
+    {
+      "epoch": 0.4043156789599482,
+      "grad_norm": 0.4013117551803589,
+      "learning_rate": 6.754067435541642e-05,
+      "loss": 0.9479,
+      "step": 4525
+    },
+    {
+      "epoch": 0.40440503049121,
+      "grad_norm": 0.4642598330974579,
+      "learning_rate": 6.752712258535244e-05,
+      "loss": 0.9661,
+      "step": 4526
+    },
+    {
+      "epoch": 0.4044943820224719,
+      "grad_norm": 0.5033729672431946,
+      "learning_rate": 6.751356934720511e-05,
+      "loss": 0.8962,
+      "step": 4527
+    },
+    {
+      "epoch": 0.4045837335537338,
+      "grad_norm": 0.4341753125190735,
+      "learning_rate": 6.750001464210967e-05,
+      "loss": 0.976,
+      "step": 4528
+    },
+    {
+      "epoch": 0.40467308508499567,
+      "grad_norm": 0.5763057470321655,
+      "learning_rate": 6.748645847120146e-05,
+      "loss": 0.9349,
+      "step": 4529
+    },
+    {
+      "epoch": 0.4047624366162575,
+      "grad_norm": 0.5385793447494507,
+      "learning_rate": 6.747290083561596e-05,
+      "loss": 0.9334,
+      "step": 4530
+    },
+    {
+      "epoch": 0.4048517881475194,
+      "grad_norm": 0.42370885610580444,
+      "learning_rate": 6.745934173648876e-05,
+      "loss": 0.9816,
+      "step": 4531
+    },
+    {
+      "epoch": 0.40494113967878126,
+      "grad_norm": 0.48537033796310425,
+      "learning_rate": 6.744578117495562e-05,
+      "loss": 0.9148,
+      "step": 4532
+    },
+    {
+      "epoch": 0.4050304912100431,
+      "grad_norm": 0.46244072914123535,
+      "learning_rate": 6.743221915215232e-05,
+      "loss": 0.9388,
+      "step": 4533
+    },
+    {
+      "epoch": 0.40511984274130497,
+      "grad_norm": 0.40350237488746643,
+      "learning_rate": 6.741865566921484e-05,
+      "loss": 0.9161,
+      "step": 4534
+    },
+    {
+      "epoch": 0.40520919427256685,
+      "grad_norm": 0.45283472537994385,
+      "learning_rate": 6.740509072727931e-05,
+      "loss": 1.0257,
+      "step": 4535
+    },
+    {
+      "epoch": 0.40529854580382874,
+      "grad_norm": 0.41299960017204285,
+      "learning_rate": 6.73915243274819e-05,
+      "loss": 1.0064,
+      "step": 4536
+    },
+    {
+      "epoch": 0.40538789733509056,
+      "grad_norm": 0.4359651207923889,
+      "learning_rate": 6.737795647095893e-05,
+      "loss": 0.9863,
+      "step": 4537
+    },
+    {
+      "epoch": 0.40547724886635245,
+      "grad_norm": 0.477792888879776,
+      "learning_rate": 6.736438715884688e-05,
+      "loss": 0.957,
+      "step": 4538
+    },
+    {
+      "epoch": 0.40556660039761433,
+      "grad_norm": 0.4058784544467926,
+      "learning_rate": 6.735081639228232e-05,
+      "loss": 0.971,
+      "step": 4539
+    },
+    {
+      "epoch": 0.40565595192887616,
+      "grad_norm": 0.45812395215034485,
+      "learning_rate": 6.733724417240194e-05,
+      "loss": 0.9408,
+      "step": 4540
+    },
+    {
+      "epoch": 0.40574530346013804,
+      "grad_norm": 0.47528955340385437,
+      "learning_rate": 6.732367050034254e-05,
+      "loss": 0.9003,
+      "step": 4541
+    },
+    {
+      "epoch": 0.4058346549913999,
+      "grad_norm": 0.3834351897239685,
+      "learning_rate": 6.731009537724112e-05,
+      "loss": 0.9874,
+      "step": 4542
+    },
+    {
+      "epoch": 0.4059240065226618,
+      "grad_norm": 0.3695882260799408,
+      "learning_rate": 6.729651880423465e-05,
+      "loss": 1.0228,
+      "step": 4543
+    },
+    {
+      "epoch": 0.40601335805392363,
+      "grad_norm": 0.4406605064868927,
+      "learning_rate": 6.728294078246038e-05,
+      "loss": 1.0354,
+      "step": 4544
+    },
+    {
+      "epoch": 0.4061027095851855,
+      "grad_norm": 0.4965038001537323,
+      "learning_rate": 6.72693613130556e-05,
+      "loss": 0.912,
+      "step": 4545
+    },
+    {
+      "epoch": 0.4061920611164474,
+      "grad_norm": 0.38543984293937683,
+      "learning_rate": 6.725578039715774e-05,
+      "loss": 1.0156,
+      "step": 4546
+    },
+    {
+      "epoch": 0.40628141264770923,
+      "grad_norm": 0.49438318610191345,
+      "learning_rate": 6.724219803590433e-05,
+      "loss": 0.9627,
+      "step": 4547
+    },
+    {
+      "epoch": 0.4063707641789711,
+      "grad_norm": 0.3538610637187958,
+      "learning_rate": 6.722861423043305e-05,
+      "loss": 1.0335,
+      "step": 4548
+    },
+    {
+      "epoch": 0.406460115710233,
+      "grad_norm": 0.3865319490432739,
+      "learning_rate": 6.721502898188167e-05,
+      "loss": 1.0594,
+      "step": 4549
+    },
+    {
+      "epoch": 0.4065494672414949,
+      "grad_norm": 0.395279198884964,
+      "learning_rate": 6.720144229138813e-05,
+      "loss": 0.9866,
+      "step": 4550
+    },
+    {
+      "epoch": 0.4066388187727567,
+      "grad_norm": 0.5369691252708435,
+      "learning_rate": 6.718785416009044e-05,
+      "loss": 0.9146,
+      "step": 4551
+    },
+    {
+      "epoch": 0.4067281703040186,
+      "grad_norm": 0.43310627341270447,
+      "learning_rate": 6.717426458912675e-05,
+      "loss": 0.9597,
+      "step": 4552
+    },
+    {
+      "epoch": 0.40681752183528047,
+      "grad_norm": 0.4922178387641907,
+      "learning_rate": 6.716067357963535e-05,
+      "loss": 0.9746,
+      "step": 4553
+    },
+    {
+      "epoch": 0.4069068733665423,
+      "grad_norm": 0.43535903096199036,
+      "learning_rate": 6.714708113275461e-05,
+      "loss": 0.967,
+      "step": 4554
+    },
+    {
+      "epoch": 0.4069962248978042,
+      "grad_norm": 0.44184789061546326,
+      "learning_rate": 6.713348724962305e-05,
+      "loss": 0.8968,
+      "step": 4555
+    },
+    {
+      "epoch": 0.40708557642906606,
+      "grad_norm": 0.4985772371292114,
+      "learning_rate": 6.711989193137929e-05,
+      "loss": 0.905,
+      "step": 4556
+    },
+    {
+      "epoch": 0.4071749279603279,
+      "grad_norm": 0.40521472692489624,
+      "learning_rate": 6.710629517916211e-05,
+      "loss": 1.0015,
+      "step": 4557
+    },
+    {
+      "epoch": 0.4072642794915898,
+      "grad_norm": 0.3867556154727936,
+      "learning_rate": 6.709269699411038e-05,
+      "loss": 0.9735,
+      "step": 4558
+    },
+    {
+      "epoch": 0.40735363102285166,
+      "grad_norm": 0.46226397156715393,
+      "learning_rate": 6.707909737736306e-05,
+      "loss": 0.8992,
+      "step": 4559
+    },
+    {
+      "epoch": 0.40744298255411354,
+      "grad_norm": 0.43801769614219666,
+      "learning_rate": 6.70654963300593e-05,
+      "loss": 0.9856,
+      "step": 4560
+    },
+    {
+      "epoch": 0.40753233408537537,
+      "grad_norm": 0.45655280351638794,
+      "learning_rate": 6.70518938533383e-05,
+      "loss": 0.9502,
+      "step": 4561
+    },
+    {
+      "epoch": 0.40762168561663725,
+      "grad_norm": 0.4761228561401367,
+      "learning_rate": 6.703828994833944e-05,
+      "loss": 0.9812,
+      "step": 4562
+    },
+    {
+      "epoch": 0.40771103714789914,
+      "grad_norm": 0.5070672631263733,
+      "learning_rate": 6.702468461620218e-05,
+      "loss": 0.8713,
+      "step": 4563
+    },
+    {
+      "epoch": 0.40780038867916096,
+      "grad_norm": 0.3671000599861145,
+      "learning_rate": 6.701107785806612e-05,
+      "loss": 1.0342,
+      "step": 4564
+    },
+    {
+      "epoch": 0.40788974021042285,
+      "grad_norm": 0.4318379759788513,
+      "learning_rate": 6.699746967507095e-05,
+      "loss": 0.9301,
+      "step": 4565
+    },
+    {
+      "epoch": 0.40797909174168473,
+      "grad_norm": 0.45091480016708374,
+      "learning_rate": 6.698386006835653e-05,
+      "loss": 1.0128,
+      "step": 4566
+    },
+    {
+      "epoch": 0.4080684432729466,
+      "grad_norm": 0.45846810936927795,
+      "learning_rate": 6.697024903906279e-05,
+      "loss": 1.044,
+      "step": 4567
+    },
+    {
+      "epoch": 0.40815779480420844,
+      "grad_norm": 0.42733606696128845,
+      "learning_rate": 6.695663658832981e-05,
+      "loss": 0.9401,
+      "step": 4568
+    },
+    {
+      "epoch": 0.4082471463354703,
+      "grad_norm": 0.48478513956069946,
+      "learning_rate": 6.694302271729774e-05,
+      "loss": 0.9547,
+      "step": 4569
+    },
+    {
+      "epoch": 0.4083364978667322,
+      "grad_norm": 0.42427879571914673,
+      "learning_rate": 6.692940742710694e-05,
+      "loss": 0.9735,
+      "step": 4570
+    },
+    {
+      "epoch": 0.40842584939799403,
+      "grad_norm": 0.40262219309806824,
+      "learning_rate": 6.691579071889782e-05,
+      "loss": 0.9945,
+      "step": 4571
+    },
+    {
+      "epoch": 0.4085152009292559,
+      "grad_norm": 0.445208877325058,
+      "learning_rate": 6.690217259381091e-05,
+      "loss": 0.9534,
+      "step": 4572
+    },
+    {
+      "epoch": 0.4086045524605178,
+      "grad_norm": 0.40912604331970215,
+      "learning_rate": 6.688855305298688e-05,
+      "loss": 0.9062,
+      "step": 4573
+    },
+    {
+      "epoch": 0.4086939039917797,
+      "grad_norm": 0.4119530916213989,
+      "learning_rate": 6.687493209756653e-05,
+      "loss": 0.9902,
+      "step": 4574
+    },
+    {
+      "epoch": 0.4087832555230415,
+      "grad_norm": 0.43412020802497864,
+      "learning_rate": 6.686130972869072e-05,
+      "loss": 0.9655,
+      "step": 4575
+    },
+    {
+      "epoch": 0.4088726070543034,
+      "grad_norm": 0.4143422842025757,
+      "learning_rate": 6.68476859475005e-05,
+      "loss": 0.9909,
+      "step": 4576
+    },
+    {
+      "epoch": 0.4089619585855653,
+      "grad_norm": 0.3809065520763397,
+      "learning_rate": 6.6834060755137e-05,
+      "loss": 1.0173,
+      "step": 4577
+    },
+    {
+      "epoch": 0.4090513101168271,
+      "grad_norm": 0.5703213810920715,
+      "learning_rate": 6.682043415274147e-05,
+      "loss": 0.9383,
+      "step": 4578
+    },
+    {
+      "epoch": 0.409140661648089,
+      "grad_norm": 0.37264561653137207,
+      "learning_rate": 6.680680614145529e-05,
+      "loss": 0.9856,
+      "step": 4579
+    },
+    {
+      "epoch": 0.40923001317935087,
+      "grad_norm": 0.4544839859008789,
+      "learning_rate": 6.679317672241994e-05,
+      "loss": 0.9899,
+      "step": 4580
+    },
+    {
+      "epoch": 0.40931936471061275,
+      "grad_norm": 0.3862614929676056,
+      "learning_rate": 6.677954589677705e-05,
+      "loss": 1.0016,
+      "step": 4581
+    },
+    {
+      "epoch": 0.4094087162418746,
+      "grad_norm": 0.4736924171447754,
+      "learning_rate": 6.676591366566831e-05,
+      "loss": 0.9096,
+      "step": 4582
+    },
+    {
+      "epoch": 0.40949806777313646,
+      "grad_norm": 0.4209919273853302,
+      "learning_rate": 6.675228003023561e-05,
+      "loss": 0.9689,
+      "step": 4583
+    },
+    {
+      "epoch": 0.40958741930439835,
+      "grad_norm": 0.416951060295105,
+      "learning_rate": 6.673864499162089e-05,
+      "loss": 0.9733,
+      "step": 4584
+    },
+    {
+      "epoch": 0.4096767708356602,
+      "grad_norm": 0.44141101837158203,
+      "learning_rate": 6.672500855096623e-05,
+      "loss": 0.9765,
+      "step": 4585
+    },
+    {
+      "epoch": 0.40976612236692206,
+      "grad_norm": 0.4138414263725281,
+      "learning_rate": 6.671137070941383e-05,
+      "loss": 0.9467,
+      "step": 4586
+    },
+    {
+      "epoch": 0.40985547389818394,
+      "grad_norm": 0.42416608333587646,
+      "learning_rate": 6.669773146810599e-05,
+      "loss": 0.9659,
+      "step": 4587
+    },
+    {
+      "epoch": 0.40994482542944577,
+      "grad_norm": 0.4376296401023865,
+      "learning_rate": 6.668409082818517e-05,
+      "loss": 1.0398,
+      "step": 4588
+    },
+    {
+      "epoch": 0.41003417696070765,
+      "grad_norm": 0.4650912880897522,
+      "learning_rate": 6.66704487907939e-05,
+      "loss": 0.9026,
+      "step": 4589
+    },
+    {
+      "epoch": 0.41012352849196954,
+      "grad_norm": 0.436990350484848,
+      "learning_rate": 6.665680535707485e-05,
+      "loss": 1.0163,
+      "step": 4590
+    },
+    {
+      "epoch": 0.4102128800232314,
+      "grad_norm": 0.42195725440979004,
+      "learning_rate": 6.664316052817079e-05,
+      "loss": 0.9237,
+      "step": 4591
+    },
+    {
+      "epoch": 0.41030223155449325,
+      "grad_norm": 0.4974232017993927,
+      "learning_rate": 6.662951430522464e-05,
+      "loss": 0.9441,
+      "step": 4592
+    },
+    {
+      "epoch": 0.41039158308575513,
+      "grad_norm": 0.43734729290008545,
+      "learning_rate": 6.661586668937943e-05,
+      "loss": 0.9482,
+      "step": 4593
+    },
+    {
+      "epoch": 0.410480934617017,
+      "grad_norm": 0.45577242970466614,
+      "learning_rate": 6.660221768177824e-05,
+      "loss": 0.937,
+      "step": 4594
+    },
+    {
+      "epoch": 0.41057028614827884,
+      "grad_norm": 0.4594924747943878,
+      "learning_rate": 6.658856728356437e-05,
+      "loss": 0.9181,
+      "step": 4595
+    },
+    {
+      "epoch": 0.4106596376795407,
+      "grad_norm": 0.441518098115921,
+      "learning_rate": 6.657491549588115e-05,
+      "loss": 0.9669,
+      "step": 4596
+    },
+    {
+      "epoch": 0.4107489892108026,
+      "grad_norm": 0.4046085774898529,
+      "learning_rate": 6.65612623198721e-05,
+      "loss": 0.9289,
+      "step": 4597
+    },
+    {
+      "epoch": 0.4108383407420645,
+      "grad_norm": 0.4187487065792084,
+      "learning_rate": 6.65476077566808e-05,
+      "loss": 1.0489,
+      "step": 4598
+    },
+    {
+      "epoch": 0.4109276922733263,
+      "grad_norm": 0.4384293854236603,
+      "learning_rate": 6.653395180745095e-05,
+      "loss": 0.9635,
+      "step": 4599
+    },
+    {
+      "epoch": 0.4110170438045882,
+      "grad_norm": 0.38338857889175415,
+      "learning_rate": 6.652029447332641e-05,
+      "loss": 1.0088,
+      "step": 4600
+    },
+    {
+      "epoch": 0.4111063953358501,
+      "grad_norm": 0.5203673839569092,
+      "learning_rate": 6.650663575545111e-05,
+      "loss": 0.9381,
+      "step": 4601
+    },
+    {
+      "epoch": 0.4111957468671119,
+      "grad_norm": 0.43095913529396057,
+      "learning_rate": 6.649297565496911e-05,
+      "loss": 1.0205,
+      "step": 4602
+    },
+    {
+      "epoch": 0.4112850983983738,
+      "grad_norm": 0.43971747159957886,
+      "learning_rate": 6.647931417302462e-05,
+      "loss": 0.9816,
+      "step": 4603
+    },
+    {
+      "epoch": 0.4113744499296357,
+      "grad_norm": 0.3989725708961487,
+      "learning_rate": 6.64656513107619e-05,
+      "loss": 0.9747,
+      "step": 4604
+    },
+    {
+      "epoch": 0.41146380146089756,
+      "grad_norm": 0.47779417037963867,
+      "learning_rate": 6.645198706932536e-05,
+      "loss": 0.9923,
+      "step": 4605
+    },
+    {
+      "epoch": 0.4115531529921594,
+      "grad_norm": 0.4292563199996948,
+      "learning_rate": 6.643832144985955e-05,
+      "loss": 0.944,
+      "step": 4606
+    },
+    {
+      "epoch": 0.41164250452342127,
+      "grad_norm": 0.5014485120773315,
+      "learning_rate": 6.64246544535091e-05,
+      "loss": 0.9265,
+      "step": 4607
+    },
+    {
+      "epoch": 0.41173185605468315,
+      "grad_norm": 0.43778491020202637,
+      "learning_rate": 6.641098608141874e-05,
+      "loss": 0.9643,
+      "step": 4608
+    },
+    {
+      "epoch": 0.411821207585945,
+      "grad_norm": 0.6082847714424133,
+      "learning_rate": 6.63973163347334e-05,
+      "loss": 0.9869,
+      "step": 4609
+    },
+    {
+      "epoch": 0.41191055911720686,
+      "grad_norm": 0.4783382713794708,
+      "learning_rate": 6.638364521459802e-05,
+      "loss": 1.0311,
+      "step": 4610
+    },
+    {
+      "epoch": 0.41199991064846875,
+      "grad_norm": 0.5080249905586243,
+      "learning_rate": 6.636997272215772e-05,
+      "loss": 0.9857,
+      "step": 4611
+    },
+    {
+      "epoch": 0.41208926217973063,
+      "grad_norm": 0.3644493520259857,
+      "learning_rate": 6.63562988585577e-05,
+      "loss": 1.006,
+      "step": 4612
+    },
+    {
+      "epoch": 0.41217861371099246,
+      "grad_norm": 0.4645083248615265,
+      "learning_rate": 6.634262362494332e-05,
+      "loss": 0.9701,
+      "step": 4613
+    },
+    {
+      "epoch": 0.41226796524225434,
+      "grad_norm": 0.48680824041366577,
+      "learning_rate": 6.632894702246001e-05,
+      "loss": 0.9591,
+      "step": 4614
+    },
+    {
+      "epoch": 0.4123573167735162,
+      "grad_norm": 0.4588204622268677,
+      "learning_rate": 6.631526905225333e-05,
+      "loss": 0.926,
+      "step": 4615
+    },
+    {
+      "epoch": 0.41244666830477805,
+      "grad_norm": 0.43005460500717163,
+      "learning_rate": 6.630158971546896e-05,
+      "loss": 1.0193,
+      "step": 4616
+    },
+    {
+      "epoch": 0.41253601983603994,
+      "grad_norm": 0.4423251152038574,
+      "learning_rate": 6.628790901325267e-05,
+      "loss": 0.8798,
+      "step": 4617
+    },
+    {
+      "epoch": 0.4126253713673018,
+      "grad_norm": 0.5219715237617493,
+      "learning_rate": 6.627422694675042e-05,
+      "loss": 1.0016,
+      "step": 4618
+    },
+    {
+      "epoch": 0.4127147228985637,
+      "grad_norm": 0.4232001304626465,
+      "learning_rate": 6.626054351710817e-05,
+      "loss": 0.9672,
+      "step": 4619
+    },
+    {
+      "epoch": 0.41280407442982553,
+      "grad_norm": 0.47682496905326843,
+      "learning_rate": 6.624685872547207e-05,
+      "loss": 0.9314,
+      "step": 4620
+    },
+    {
+      "epoch": 0.4128934259610874,
+      "grad_norm": 0.46166396141052246,
+      "learning_rate": 6.623317257298837e-05,
+      "loss": 1.015,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4129827774923493,
+      "grad_norm": 0.39371028542518616,
+      "learning_rate": 6.621948506080345e-05,
+      "loss": 0.9835,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4130721290236111,
+      "grad_norm": 0.41618743538856506,
+      "learning_rate": 6.620579619006377e-05,
+      "loss": 0.9857,
+      "step": 4623
+    },
+    {
+      "epoch": 0.413161480554873,
+      "grad_norm": 0.4997164011001587,
+      "learning_rate": 6.619210596191592e-05,
+      "loss": 0.8842,
+      "step": 4624
+    },
+    {
+      "epoch": 0.4132508320861349,
+      "grad_norm": 0.45205265283584595,
+      "learning_rate": 6.61784143775066e-05,
+      "loss": 0.9845,
+      "step": 4625
+    },
+    {
+      "epoch": 0.4133401836173967,
+      "grad_norm": 0.49516791105270386,
+      "learning_rate": 6.616472143798261e-05,
+      "loss": 0.8691,
+      "step": 4626
+    },
+    {
+      "epoch": 0.4134295351486586,
+      "grad_norm": 0.4403949975967407,
+      "learning_rate": 6.615102714449089e-05,
+      "loss": 0.9634,
+      "step": 4627
+    },
+    {
+      "epoch": 0.4135188866799205,
+      "grad_norm": 0.43604418635368347,
+      "learning_rate": 6.613733149817852e-05,
+      "loss": 1.0116,
+      "step": 4628
+    },
+    {
+      "epoch": 0.41360823821118237,
+      "grad_norm": 0.4321603775024414,
+      "learning_rate": 6.612363450019261e-05,
+      "loss": 0.924,
+      "step": 4629
+    },
+    {
+      "epoch": 0.4136975897424442,
+      "grad_norm": 0.40170714259147644,
+      "learning_rate": 6.610993615168044e-05,
+      "loss": 1.0031,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4137869412737061,
+      "grad_norm": 0.4581966996192932,
+      "learning_rate": 6.60962364537894e-05,
+      "loss": 0.9848,
+      "step": 4631
+    },
+    {
+      "epoch": 0.41387629280496796,
+      "grad_norm": 0.42870965600013733,
+      "learning_rate": 6.608253540766698e-05,
+      "loss": 1.0401,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4139656443362298,
+      "grad_norm": 0.4049166738986969,
+      "learning_rate": 6.60688330144608e-05,
+      "loss": 0.9862,
+      "step": 4633
+    },
+    {
+      "epoch": 0.41405499586749167,
+      "grad_norm": 0.44598478078842163,
+      "learning_rate": 6.605512927531858e-05,
+      "loss": 1.0188,
+      "step": 4634
+    },
+    {
+      "epoch": 0.41414434739875355,
+      "grad_norm": 0.4548451602458954,
+      "learning_rate": 6.604142419138812e-05,
+      "loss": 0.9463,
+      "step": 4635
+    },
+    {
+      "epoch": 0.41423369893001544,
+      "grad_norm": 0.41547542810440063,
+      "learning_rate": 6.602771776381743e-05,
+      "loss": 1.0352,
+      "step": 4636
+    },
+    {
+      "epoch": 0.41432305046127726,
+      "grad_norm": 0.42148005962371826,
+      "learning_rate": 6.60140099937545e-05,
+      "loss": 0.944,
+      "step": 4637
+    },
+    {
+      "epoch": 0.41441240199253915,
+      "grad_norm": 0.41155093908309937,
+      "learning_rate": 6.600030088234755e-05,
+      "loss": 0.9639,
+      "step": 4638
+    },
+    {
+      "epoch": 0.41450175352380103,
+      "grad_norm": 0.4119633436203003,
+      "learning_rate": 6.598659043074487e-05,
+      "loss": 0.9254,
+      "step": 4639
+    },
+    {
+      "epoch": 0.41459110505506286,
+      "grad_norm": 0.4401843547821045,
+      "learning_rate": 6.59728786400948e-05,
+      "loss": 0.9798,
+      "step": 4640
+    },
+    {
+      "epoch": 0.41468045658632474,
+      "grad_norm": 0.4861004054546356,
+      "learning_rate": 6.595916551154591e-05,
+      "loss": 0.9584,
+      "step": 4641
+    },
+    {
+      "epoch": 0.4147698081175866,
+      "grad_norm": 0.44102394580841064,
+      "learning_rate": 6.59454510462468e-05,
+      "loss": 0.9108,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4148591596488485,
+      "grad_norm": 0.4617249667644501,
+      "learning_rate": 6.593173524534619e-05,
+      "loss": 0.9601,
+      "step": 4643
+    },
+    {
+      "epoch": 0.41494851118011034,
+      "grad_norm": 0.5489480495452881,
+      "learning_rate": 6.591801810999294e-05,
+      "loss": 0.9668,
+      "step": 4644
+    },
+    {
+      "epoch": 0.4150378627113722,
+      "grad_norm": 0.5144347548484802,
+      "learning_rate": 6.590429964133599e-05,
+      "loss": 0.9139,
+      "step": 4645
+    },
+    {
+      "epoch": 0.4151272142426341,
+      "grad_norm": 0.4274512827396393,
+      "learning_rate": 6.589057984052441e-05,
+      "loss": 1.001,
+      "step": 4646
+    },
+    {
+      "epoch": 0.41521656577389593,
+      "grad_norm": 0.4437926709651947,
+      "learning_rate": 6.58768587087074e-05,
+      "loss": 0.9184,
+      "step": 4647
+    },
+    {
+      "epoch": 0.4153059173051578,
+      "grad_norm": 0.46083325147628784,
+      "learning_rate": 6.586313624703423e-05,
+      "loss": 1.0409,
+      "step": 4648
+    },
+    {
+      "epoch": 0.4153952688364197,
+      "grad_norm": 0.46094614267349243,
+      "learning_rate": 6.584941245665432e-05,
+      "loss": 0.9456,
+      "step": 4649
+    },
+    {
+      "epoch": 0.4154846203676816,
+      "grad_norm": 0.3888949751853943,
+      "learning_rate": 6.583568733871716e-05,
+      "loss": 0.9999,
+      "step": 4650
+    },
+    {
+      "epoch": 0.4155739718989434,
+      "grad_norm": 0.4531100392341614,
+      "learning_rate": 6.58219608943724e-05,
+      "loss": 1.0092,
+      "step": 4651
+    },
+    {
+      "epoch": 0.4156633234302053,
+      "grad_norm": 0.44666728377342224,
+      "learning_rate": 6.580823312476976e-05,
+      "loss": 0.961,
+      "step": 4652
+    },
+    {
+      "epoch": 0.4157526749614672,
+      "grad_norm": 0.39835554361343384,
+      "learning_rate": 6.579450403105909e-05,
+      "loss": 0.9973,
+      "step": 4653
+    },
+    {
+      "epoch": 0.415842026492729,
+      "grad_norm": 0.4402940571308136,
+      "learning_rate": 6.578077361439037e-05,
+      "loss": 0.9221,
+      "step": 4654
+    },
+    {
+      "epoch": 0.4159313780239909,
+      "grad_norm": 0.43528056144714355,
+      "learning_rate": 6.576704187591362e-05,
+      "loss": 0.9335,
+      "step": 4655
+    },
+    {
+      "epoch": 0.41602072955525277,
+      "grad_norm": 0.48147714138031006,
+      "learning_rate": 6.575330881677907e-05,
+      "loss": 0.9158,
+      "step": 4656
+    },
+    {
+      "epoch": 0.4161100810865146,
+      "grad_norm": 0.4357036352157593,
+      "learning_rate": 6.573957443813698e-05,
+      "loss": 0.9688,
+      "step": 4657
+    },
+    {
+      "epoch": 0.4161994326177765,
+      "grad_norm": 0.41116756200790405,
+      "learning_rate": 6.572583874113777e-05,
+      "loss": 0.9678,
+      "step": 4658
+    },
+    {
+      "epoch": 0.41628878414903836,
+      "grad_norm": 0.3910764753818512,
+      "learning_rate": 6.571210172693192e-05,
+      "loss": 1.0419,
+      "step": 4659
+    },
+    {
+      "epoch": 0.41637813568030024,
+      "grad_norm": 0.5164288878440857,
+      "learning_rate": 6.569836339667009e-05,
+      "loss": 0.8935,
+      "step": 4660
+    },
+    {
+      "epoch": 0.41646748721156207,
+      "grad_norm": 0.48171505331993103,
+      "learning_rate": 6.568462375150298e-05,
+      "loss": 1.0256,
+      "step": 4661
+    },
+    {
+      "epoch": 0.41655683874282395,
+      "grad_norm": 0.43852394819259644,
+      "learning_rate": 6.567088279258144e-05,
+      "loss": 0.925,
+      "step": 4662
+    },
+    {
+      "epoch": 0.41664619027408584,
+      "grad_norm": 0.42417702078819275,
+      "learning_rate": 6.565714052105645e-05,
+      "loss": 0.9589,
+      "step": 4663
+    },
+    {
+      "epoch": 0.41673554180534766,
+      "grad_norm": 0.42217710614204407,
+      "learning_rate": 6.564339693807904e-05,
+      "loss": 0.9391,
+      "step": 4664
+    },
+    {
+      "epoch": 0.41682489333660955,
+      "grad_norm": 0.48488372564315796,
+      "learning_rate": 6.562965204480039e-05,
+      "loss": 0.9816,
+      "step": 4665
+    },
+    {
+      "epoch": 0.41691424486787143,
+      "grad_norm": 0.424485981464386,
+      "learning_rate": 6.561590584237176e-05,
+      "loss": 1.0454,
+      "step": 4666
+    },
+    {
+      "epoch": 0.4170035963991333,
+      "grad_norm": 0.40469837188720703,
+      "learning_rate": 6.560215833194457e-05,
+      "loss": 1.0021,
+      "step": 4667
+    },
+    {
+      "epoch": 0.41709294793039514,
+      "grad_norm": 0.4954060912132263,
+      "learning_rate": 6.558840951467033e-05,
+      "loss": 1.0377,
+      "step": 4668
+    },
+    {
+      "epoch": 0.417182299461657,
+      "grad_norm": 0.47317901253700256,
+      "learning_rate": 6.557465939170062e-05,
+      "loss": 0.9776,
+      "step": 4669
+    },
+    {
+      "epoch": 0.4172716509929189,
+      "grad_norm": 0.4012228846549988,
+      "learning_rate": 6.556090796418717e-05,
+      "loss": 0.9594,
+      "step": 4670
+    },
+    {
+      "epoch": 0.41736100252418074,
+      "grad_norm": 0.4025217890739441,
+      "learning_rate": 6.554715523328181e-05,
+      "loss": 0.9784,
+      "step": 4671
+    },
+    {
+      "epoch": 0.4174503540554426,
+      "grad_norm": 0.4811991751194,
+      "learning_rate": 6.553340120013649e-05,
+      "loss": 0.9726,
+      "step": 4672
+    },
+    {
+      "epoch": 0.4175397055867045,
+      "grad_norm": 0.4033472239971161,
+      "learning_rate": 6.551964586590323e-05,
+      "loss": 0.9814,
+      "step": 4673
+    },
+    {
+      "epoch": 0.4176290571179664,
+      "grad_norm": 0.45485714077949524,
+      "learning_rate": 6.550588923173422e-05,
+      "loss": 0.9099,
+      "step": 4674
+    },
+    {
+      "epoch": 0.4177184086492282,
+      "grad_norm": 0.48728662729263306,
+      "learning_rate": 6.549213129878169e-05,
+      "loss": 0.9082,
+      "step": 4675
+    },
+    {
+      "epoch": 0.4178077601804901,
+      "grad_norm": 0.46427708864212036,
+      "learning_rate": 6.547837206819804e-05,
+      "loss": 0.9797,
+      "step": 4676
+    },
+    {
+      "epoch": 0.417897111711752,
+      "grad_norm": 0.5304161310195923,
+      "learning_rate": 6.546461154113575e-05,
+      "loss": 0.9526,
+      "step": 4677
+    },
+    {
+      "epoch": 0.4179864632430138,
+      "grad_norm": 0.5012792348861694,
+      "learning_rate": 6.545084971874738e-05,
+      "loss": 0.9665,
+      "step": 4678
+    },
+    {
+      "epoch": 0.4180758147742757,
+      "grad_norm": 0.41334888339042664,
+      "learning_rate": 6.543708660218566e-05,
+      "loss": 0.945,
+      "step": 4679
+    },
+    {
+      "epoch": 0.4181651663055376,
+      "grad_norm": 0.38999998569488525,
+      "learning_rate": 6.54233221926034e-05,
+      "loss": 0.9889,
+      "step": 4680
+    },
+    {
+      "epoch": 0.41825451783679946,
+      "grad_norm": 0.43242332339286804,
+      "learning_rate": 6.540955649115349e-05,
+      "loss": 0.9914,
+      "step": 4681
+    },
+    {
+      "epoch": 0.4183438693680613,
+      "grad_norm": 0.46687716245651245,
+      "learning_rate": 6.539578949898896e-05,
+      "loss": 0.986,
+      "step": 4682
+    },
+    {
+      "epoch": 0.41843322089932317,
+      "grad_norm": 0.5817546844482422,
+      "learning_rate": 6.538202121726298e-05,
+      "loss": 0.8811,
+      "step": 4683
+    },
+    {
+      "epoch": 0.41852257243058505,
+      "grad_norm": 0.4378966987133026,
+      "learning_rate": 6.536825164712876e-05,
+      "loss": 0.9343,
+      "step": 4684
+    },
+    {
+      "epoch": 0.4186119239618469,
+      "grad_norm": 0.45147213339805603,
+      "learning_rate": 6.535448078973963e-05,
+      "loss": 0.928,
+      "step": 4685
+    },
+    {
+      "epoch": 0.41870127549310876,
+      "grad_norm": 0.5144475102424622,
+      "learning_rate": 6.534070864624908e-05,
+      "loss": 0.9623,
+      "step": 4686
+    },
+    {
+      "epoch": 0.41879062702437064,
+      "grad_norm": 0.44257742166519165,
+      "learning_rate": 6.532693521781066e-05,
+      "loss": 0.9797,
+      "step": 4687
+    },
+    {
+      "epoch": 0.41887997855563247,
+      "grad_norm": 0.4004027247428894,
+      "learning_rate": 6.531316050557803e-05,
+      "loss": 1.0237,
+      "step": 4688
+    },
+    {
+      "epoch": 0.41896933008689435,
+      "grad_norm": 0.5068597793579102,
+      "learning_rate": 6.529938451070501e-05,
+      "loss": 0.9238,
+      "step": 4689
+    },
+    {
+      "epoch": 0.41905868161815624,
+      "grad_norm": 0.4261419177055359,
+      "learning_rate": 6.528560723434543e-05,
+      "loss": 0.9867,
+      "step": 4690
+    },
+    {
+      "epoch": 0.4191480331494181,
+      "grad_norm": 0.5222342014312744,
+      "learning_rate": 6.527182867765332e-05,
+      "loss": 0.9864,
+      "step": 4691
+    },
+    {
+      "epoch": 0.41923738468067995,
+      "grad_norm": 0.5611239075660706,
+      "learning_rate": 6.525804884178277e-05,
+      "loss": 0.9771,
+      "step": 4692
+    },
+    {
+      "epoch": 0.41932673621194183,
+      "grad_norm": 0.3744679391384125,
+      "learning_rate": 6.524426772788801e-05,
+      "loss": 0.9883,
+      "step": 4693
+    },
+    {
+      "epoch": 0.4194160877432037,
+      "grad_norm": 0.4100448787212372,
+      "learning_rate": 6.523048533712331e-05,
+      "loss": 1.0148,
+      "step": 4694
+    },
+    {
+      "epoch": 0.41950543927446554,
+      "grad_norm": 0.42226895689964294,
+      "learning_rate": 6.521670167064313e-05,
+      "loss": 1.027,
+      "step": 4695
+    },
+    {
+      "epoch": 0.4195947908057274,
+      "grad_norm": 0.450785756111145,
+      "learning_rate": 6.5202916729602e-05,
+      "loss": 0.9611,
+      "step": 4696
+    },
+    {
+      "epoch": 0.4196841423369893,
+      "grad_norm": 0.4674597382545471,
+      "learning_rate": 6.518913051515451e-05,
+      "loss": 0.9714,
+      "step": 4697
+    },
+    {
+      "epoch": 0.4197734938682512,
+      "grad_norm": 0.5814161896705627,
+      "learning_rate": 6.517534302845545e-05,
+      "loss": 0.8825,
+      "step": 4698
+    },
+    {
+      "epoch": 0.419862845399513,
+      "grad_norm": 0.4287194609642029,
+      "learning_rate": 6.516155427065967e-05,
+      "loss": 1.0032,
+      "step": 4699
+    },
+    {
+      "epoch": 0.4199521969307749,
+      "grad_norm": 0.466623991727829,
+      "learning_rate": 6.514776424292208e-05,
+      "loss": 0.9162,
+      "step": 4700
+    },
+    {
+      "epoch": 0.4200415484620368,
+      "grad_norm": 0.5070553421974182,
+      "learning_rate": 6.513397294639778e-05,
+      "loss": 1.0066,
+      "step": 4701
+    },
+    {
+      "epoch": 0.4201308999932986,
+      "grad_norm": 0.42103180289268494,
+      "learning_rate": 6.512018038224194e-05,
+      "loss": 0.9602,
+      "step": 4702
+    },
+    {
+      "epoch": 0.4202202515245605,
+      "grad_norm": 0.49392175674438477,
+      "learning_rate": 6.510638655160981e-05,
+      "loss": 0.9859,
+      "step": 4703
+    },
+    {
+      "epoch": 0.4203096030558224,
+      "grad_norm": 0.46998798847198486,
+      "learning_rate": 6.509259145565681e-05,
+      "loss": 0.9445,
+      "step": 4704
+    },
+    {
+      "epoch": 0.42039895458708426,
+      "grad_norm": 0.47235095500946045,
+      "learning_rate": 6.507879509553837e-05,
+      "loss": 0.942,
+      "step": 4705
+    },
+    {
+      "epoch": 0.4204883061183461,
+      "grad_norm": 0.4976077973842621,
+      "learning_rate": 6.506499747241013e-05,
+      "loss": 0.9331,
+      "step": 4706
+    },
+    {
+      "epoch": 0.42057765764960797,
+      "grad_norm": 0.45665648579597473,
+      "learning_rate": 6.505119858742775e-05,
+      "loss": 1.0242,
+      "step": 4707
+    },
+    {
+      "epoch": 0.42066700918086986,
+      "grad_norm": 0.5178200602531433,
+      "learning_rate": 6.503739844174708e-05,
+      "loss": 0.9287,
+      "step": 4708
+    },
+    {
+      "epoch": 0.4207563607121317,
+      "grad_norm": 0.41020286083221436,
+      "learning_rate": 6.502359703652398e-05,
+      "loss": 1.0093,
+      "step": 4709
+    },
+    {
+      "epoch": 0.42084571224339357,
+      "grad_norm": 0.44288718700408936,
+      "learning_rate": 6.500979437291451e-05,
+      "loss": 1.0115,
+      "step": 4710
+    },
+    {
+      "epoch": 0.42093506377465545,
+      "grad_norm": 0.4006830155849457,
+      "learning_rate": 6.499599045207475e-05,
+      "loss": 0.9736,
+      "step": 4711
+    },
+    {
+      "epoch": 0.42102441530591733,
+      "grad_norm": 0.4733505845069885,
+      "learning_rate": 6.498218527516097e-05,
+      "loss": 0.9286,
+      "step": 4712
+    },
+    {
+      "epoch": 0.42111376683717916,
+      "grad_norm": 0.48651424050331116,
+      "learning_rate": 6.496837884332945e-05,
+      "loss": 1.0209,
+      "step": 4713
+    },
+    {
+      "epoch": 0.42120311836844104,
+      "grad_norm": 0.38982510566711426,
+      "learning_rate": 6.495457115773667e-05,
+      "loss": 1.0108,
+      "step": 4714
+    },
+    {
+      "epoch": 0.4212924698997029,
+      "grad_norm": 0.47377124428749084,
+      "learning_rate": 6.494076221953912e-05,
+      "loss": 0.986,
+      "step": 4715
+    },
+    {
+      "epoch": 0.42138182143096475,
+      "grad_norm": 0.3638954758644104,
+      "learning_rate": 6.492695202989351e-05,
+      "loss": 1.0072,
+      "step": 4716
+    },
+    {
+      "epoch": 0.42147117296222664,
+      "grad_norm": 0.49875038862228394,
+      "learning_rate": 6.491314058995654e-05,
+      "loss": 0.9364,
+      "step": 4717
+    },
+    {
+      "epoch": 0.4215605244934885,
+      "grad_norm": 0.41740208864212036,
+      "learning_rate": 6.489932790088508e-05,
+      "loss": 0.9213,
+      "step": 4718
+    },
+    {
+      "epoch": 0.42164987602475035,
+      "grad_norm": 0.40949100255966187,
+      "learning_rate": 6.48855139638361e-05,
+      "loss": 0.9908,
+      "step": 4719
+    },
+    {
+      "epoch": 0.42173922755601223,
+      "grad_norm": 0.4184902608394623,
+      "learning_rate": 6.487169877996667e-05,
+      "loss": 0.9252,
+      "step": 4720
+    },
+    {
+      "epoch": 0.4218285790872741,
+      "grad_norm": 0.3919154405593872,
+      "learning_rate": 6.485788235043392e-05,
+      "loss": 1.0046,
+      "step": 4721
+    },
+    {
+      "epoch": 0.421917930618536,
+      "grad_norm": 0.44090232253074646,
+      "learning_rate": 6.484406467639516e-05,
+      "loss": 1.0108,
+      "step": 4722
+    },
+    {
+      "epoch": 0.4220072821497978,
+      "grad_norm": 0.407052606344223,
+      "learning_rate": 6.483024575900776e-05,
+      "loss": 1.01,
+      "step": 4723
+    },
+    {
+      "epoch": 0.4220966336810597,
+      "grad_norm": 0.5126084089279175,
+      "learning_rate": 6.481642559942919e-05,
+      "loss": 0.9421,
+      "step": 4724
+    },
+    {
+      "epoch": 0.4221859852123216,
+      "grad_norm": 0.3954399824142456,
+      "learning_rate": 6.480260419881706e-05,
+      "loss": 1.0176,
+      "step": 4725
+    },
+    {
+      "epoch": 0.4222753367435834,
+      "grad_norm": 0.4803306758403778,
+      "learning_rate": 6.478878155832903e-05,
+      "loss": 0.8829,
+      "step": 4726
+    },
+    {
+      "epoch": 0.4223646882748453,
+      "grad_norm": 0.4403332769870758,
+      "learning_rate": 6.477495767912292e-05,
+      "loss": 0.9965,
+      "step": 4727
+    },
+    {
+      "epoch": 0.4224540398061072,
+      "grad_norm": 0.4345102906227112,
+      "learning_rate": 6.476113256235661e-05,
+      "loss": 0.9472,
+      "step": 4728
+    },
+    {
+      "epoch": 0.42254339133736907,
+      "grad_norm": 0.4194163978099823,
+      "learning_rate": 6.47473062091881e-05,
+      "loss": 1.0142,
+      "step": 4729
+    },
+    {
+      "epoch": 0.4226327428686309,
+      "grad_norm": 0.5548313856124878,
+      "learning_rate": 6.473347862077552e-05,
+      "loss": 0.9398,
+      "step": 4730
+    },
+    {
+      "epoch": 0.4227220943998928,
+      "grad_norm": 0.49894118309020996,
+      "learning_rate": 6.471964979827702e-05,
+      "loss": 0.9525,
+      "step": 4731
+    },
+    {
+      "epoch": 0.42281144593115466,
+      "grad_norm": 0.5956845879554749,
+      "learning_rate": 6.470581974285098e-05,
+      "loss": 0.9364,
+      "step": 4732
+    },
+    {
+      "epoch": 0.4229007974624165,
+      "grad_norm": 0.5391216278076172,
+      "learning_rate": 6.469198845565577e-05,
+      "loss": 0.9362,
+      "step": 4733
+    },
+    {
+      "epoch": 0.42299014899367837,
+      "grad_norm": 0.4035119414329529,
+      "learning_rate": 6.467815593784993e-05,
+      "loss": 0.9826,
+      "step": 4734
+    },
+    {
+      "epoch": 0.42307950052494026,
+      "grad_norm": 0.43336036801338196,
+      "learning_rate": 6.466432219059208e-05,
+      "loss": 0.9782,
+      "step": 4735
+    },
+    {
+      "epoch": 0.42316885205620214,
+      "grad_norm": 0.43682193756103516,
+      "learning_rate": 6.465048721504091e-05,
+      "loss": 0.9663,
+      "step": 4736
+    },
+    {
+      "epoch": 0.42325820358746397,
+      "grad_norm": 0.49515190720558167,
+      "learning_rate": 6.46366510123553e-05,
+      "loss": 0.9055,
+      "step": 4737
+    },
+    {
+      "epoch": 0.42334755511872585,
+      "grad_norm": 0.4037051200866699,
+      "learning_rate": 6.462281358369413e-05,
+      "loss": 0.9359,
+      "step": 4738
+    },
+    {
+      "epoch": 0.42343690664998773,
+      "grad_norm": 0.4590768814086914,
+      "learning_rate": 6.460897493021646e-05,
+      "loss": 0.9446,
+      "step": 4739
+    },
+    {
+      "epoch": 0.42352625818124956,
+      "grad_norm": 0.45074954628944397,
+      "learning_rate": 6.459513505308142e-05,
+      "loss": 0.993,
+      "step": 4740
+    },
+    {
+      "epoch": 0.42361560971251144,
+      "grad_norm": 0.4910895824432373,
+      "learning_rate": 6.458129395344825e-05,
+      "loss": 0.9115,
+      "step": 4741
+    },
+    {
+      "epoch": 0.4237049612437733,
+      "grad_norm": 0.4874398410320282,
+      "learning_rate": 6.456745163247628e-05,
+      "loss": 0.9896,
+      "step": 4742
+    },
+    {
+      "epoch": 0.4237943127750352,
+      "grad_norm": 0.4627344012260437,
+      "learning_rate": 6.455360809132496e-05,
+      "loss": 0.9819,
+      "step": 4743
+    },
+    {
+      "epoch": 0.42388366430629704,
+      "grad_norm": 0.3980332314968109,
+      "learning_rate": 6.453976333115383e-05,
+      "loss": 0.9962,
+      "step": 4744
+    },
+    {
+      "epoch": 0.4239730158375589,
+      "grad_norm": 0.527260422706604,
+      "learning_rate": 6.452591735312257e-05,
+      "loss": 0.8636,
+      "step": 4745
+    },
+    {
+      "epoch": 0.4240623673688208,
+      "grad_norm": 0.4000941514968872,
+      "learning_rate": 6.451207015839086e-05,
+      "loss": 0.9904,
+      "step": 4746
+    },
+    {
+      "epoch": 0.42415171890008263,
+      "grad_norm": 0.4481007158756256,
+      "learning_rate": 6.44982217481186e-05,
+      "loss": 0.9877,
+      "step": 4747
+    },
+    {
+      "epoch": 0.4242410704313445,
+      "grad_norm": 0.49271902441978455,
+      "learning_rate": 6.448437212346572e-05,
+      "loss": 0.8871,
+      "step": 4748
+    },
+    {
+      "epoch": 0.4243304219626064,
+      "grad_norm": 0.4154813587665558,
+      "learning_rate": 6.447052128559229e-05,
+      "loss": 0.9812,
+      "step": 4749
+    },
+    {
+      "epoch": 0.4244197734938682,
+      "grad_norm": 0.5907759666442871,
+      "learning_rate": 6.445666923565846e-05,
+      "loss": 0.9432,
+      "step": 4750
+    },
+    {
+      "epoch": 0.4245091250251301,
+      "grad_norm": 0.4116978049278259,
+      "learning_rate": 6.444281597482448e-05,
+      "loss": 0.9952,
+      "step": 4751
+    },
+    {
+      "epoch": 0.424598476556392,
+      "grad_norm": 0.4625799357891083,
+      "learning_rate": 6.44289615042507e-05,
+      "loss": 1.0082,
+      "step": 4752
+    },
+    {
+      "epoch": 0.4246878280876539,
+      "grad_norm": 0.4559314548969269,
+      "learning_rate": 6.441510582509761e-05,
+      "loss": 1.0048,
+      "step": 4753
+    },
+    {
+      "epoch": 0.4247771796189157,
+      "grad_norm": 0.4704115390777588,
+      "learning_rate": 6.440124893852575e-05,
+      "loss": 0.9848,
+      "step": 4754
+    },
+    {
+      "epoch": 0.4248665311501776,
+      "grad_norm": 0.40942203998565674,
+      "learning_rate": 6.438739084569579e-05,
+      "loss": 0.9624,
+      "step": 4755
+    },
+    {
+      "epoch": 0.42495588268143947,
+      "grad_norm": 0.3844430446624756,
+      "learning_rate": 6.437353154776849e-05,
+      "loss": 1.0214,
+      "step": 4756
+    },
+    {
+      "epoch": 0.4250452342127013,
+      "grad_norm": 0.4514857232570648,
+      "learning_rate": 6.435967104590469e-05,
+      "loss": 0.9475,
+      "step": 4757
+    },
+    {
+      "epoch": 0.4251345857439632,
+      "grad_norm": 0.44414305686950684,
+      "learning_rate": 6.43458093412654e-05,
+      "loss": 0.9882,
+      "step": 4758
+    },
+    {
+      "epoch": 0.42522393727522506,
+      "grad_norm": 0.5354793667793274,
+      "learning_rate": 6.433194643501164e-05,
+      "loss": 0.9163,
+      "step": 4759
+    },
+    {
+      "epoch": 0.42531328880648694,
+      "grad_norm": 0.4825092554092407,
+      "learning_rate": 6.431808232830462e-05,
+      "loss": 0.96,
+      "step": 4760
+    },
+    {
+      "epoch": 0.42540264033774877,
+      "grad_norm": 0.44994306564331055,
+      "learning_rate": 6.430421702230556e-05,
+      "loss": 0.9369,
+      "step": 4761
+    },
+    {
+      "epoch": 0.42549199186901066,
+      "grad_norm": 0.4433254599571228,
+      "learning_rate": 6.429035051817588e-05,
+      "loss": 0.9945,
+      "step": 4762
+    },
+    {
+      "epoch": 0.42558134340027254,
+      "grad_norm": 0.49267831444740295,
+      "learning_rate": 6.427648281707701e-05,
+      "loss": 1.0097,
+      "step": 4763
+    },
+    {
+      "epoch": 0.42567069493153437,
+      "grad_norm": 0.3941679000854492,
+      "learning_rate": 6.426261392017052e-05,
+      "loss": 1.0503,
+      "step": 4764
+    },
+    {
+      "epoch": 0.42576004646279625,
+      "grad_norm": 0.3821266293525696,
+      "learning_rate": 6.424874382861811e-05,
+      "loss": 0.9521,
+      "step": 4765
+    },
+    {
+      "epoch": 0.42584939799405813,
+      "grad_norm": 0.44064003229141235,
+      "learning_rate": 6.42348725435815e-05,
+      "loss": 0.9589,
+      "step": 4766
+    },
+    {
+      "epoch": 0.42593874952532,
+      "grad_norm": 0.437092661857605,
+      "learning_rate": 6.422100006622257e-05,
+      "loss": 1.004,
+      "step": 4767
+    },
+    {
+      "epoch": 0.42602810105658184,
+      "grad_norm": 0.5033923387527466,
+      "learning_rate": 6.420712639770333e-05,
+      "loss": 0.8967,
+      "step": 4768
+    },
+    {
+      "epoch": 0.4261174525878437,
+      "grad_norm": 0.4463227093219757,
+      "learning_rate": 6.41932515391858e-05,
+      "loss": 0.9387,
+      "step": 4769
+    },
+    {
+      "epoch": 0.4262068041191056,
+      "grad_norm": 0.4098617732524872,
+      "learning_rate": 6.417937549183218e-05,
+      "loss": 0.9521,
+      "step": 4770
+    },
+    {
+      "epoch": 0.42629615565036744,
+      "grad_norm": 0.4326515197753906,
+      "learning_rate": 6.41654982568047e-05,
+      "loss": 1.0335,
+      "step": 4771
+    },
+    {
+      "epoch": 0.4263855071816293,
+      "grad_norm": 0.42155736684799194,
+      "learning_rate": 6.415161983526576e-05,
+      "loss": 0.9386,
+      "step": 4772
+    },
+    {
+      "epoch": 0.4264748587128912,
+      "grad_norm": 0.407709538936615,
+      "learning_rate": 6.41377402283778e-05,
+      "loss": 1.0556,
+      "step": 4773
+    },
+    {
+      "epoch": 0.4265642102441531,
+      "grad_norm": 0.4312109649181366,
+      "learning_rate": 6.412385943730341e-05,
+      "loss": 1.004,
+      "step": 4774
+    },
+    {
+      "epoch": 0.4266535617754149,
+      "grad_norm": 0.5315758585929871,
+      "learning_rate": 6.410997746320524e-05,
+      "loss": 0.8835,
+      "step": 4775
+    },
+    {
+      "epoch": 0.4267429133066768,
+      "grad_norm": 0.40640825033187866,
+      "learning_rate": 6.409609430724607e-05,
+      "loss": 1.0284,
+      "step": 4776
+    },
+    {
+      "epoch": 0.4268322648379387,
+      "grad_norm": 0.3886488676071167,
+      "learning_rate": 6.408220997058873e-05,
+      "loss": 0.9606,
+      "step": 4777
+    },
+    {
+      "epoch": 0.4269216163692005,
+      "grad_norm": 0.47869306802749634,
+      "learning_rate": 6.40683244543962e-05,
+      "loss": 1.0092,
+      "step": 4778
+    },
+    {
+      "epoch": 0.4270109679004624,
+      "grad_norm": 0.5204102396965027,
+      "learning_rate": 6.405443775983154e-05,
+      "loss": 0.8834,
+      "step": 4779
+    },
+    {
+      "epoch": 0.4271003194317243,
+      "grad_norm": 0.5049982666969299,
+      "learning_rate": 6.404054988805792e-05,
+      "loss": 0.8793,
+      "step": 4780
+    },
+    {
+      "epoch": 0.4271896709629861,
+      "grad_norm": 0.48872700333595276,
+      "learning_rate": 6.402666084023858e-05,
+      "loss": 0.9817,
+      "step": 4781
+    },
+    {
+      "epoch": 0.427279022494248,
+      "grad_norm": 0.428631991147995,
+      "learning_rate": 6.401277061753689e-05,
+      "loss": 0.9423,
+      "step": 4782
+    },
+    {
+      "epoch": 0.42736837402550987,
+      "grad_norm": 0.4668726623058319,
+      "learning_rate": 6.399887922111627e-05,
+      "loss": 0.9722,
+      "step": 4783
+    },
+    {
+      "epoch": 0.42745772555677175,
+      "grad_norm": 0.5166489481925964,
+      "learning_rate": 6.398498665214032e-05,
+      "loss": 0.9832,
+      "step": 4784
+    },
+    {
+      "epoch": 0.4275470770880336,
+      "grad_norm": 0.4075476825237274,
+      "learning_rate": 6.397109291177266e-05,
+      "loss": 1.0145,
+      "step": 4785
+    },
+    {
+      "epoch": 0.42763642861929546,
+      "grad_norm": 0.40599527955055237,
+      "learning_rate": 6.395719800117706e-05,
+      "loss": 0.9469,
+      "step": 4786
+    },
+    {
+      "epoch": 0.42772578015055734,
+      "grad_norm": 0.4204799234867096,
+      "learning_rate": 6.394330192151732e-05,
+      "loss": 0.9147,
+      "step": 4787
+    },
+    {
+      "epoch": 0.42781513168181917,
+      "grad_norm": 0.3994778096675873,
+      "learning_rate": 6.392940467395745e-05,
+      "loss": 1.0195,
+      "step": 4788
+    },
+    {
+      "epoch": 0.42790448321308105,
+      "grad_norm": 0.4340113401412964,
+      "learning_rate": 6.391550625966144e-05,
+      "loss": 1.0036,
+      "step": 4789
+    },
+    {
+      "epoch": 0.42799383474434294,
+      "grad_norm": 0.45802706480026245,
+      "learning_rate": 6.390160667979348e-05,
+      "loss": 0.9356,
+      "step": 4790
+    },
+    {
+      "epoch": 0.4280831862756048,
+      "grad_norm": 0.4832572937011719,
+      "learning_rate": 6.388770593551777e-05,
+      "loss": 1.0298,
+      "step": 4791
+    },
+    {
+      "epoch": 0.42817253780686665,
+      "grad_norm": 0.45630741119384766,
+      "learning_rate": 6.387380402799866e-05,
+      "loss": 0.9754,
+      "step": 4792
+    },
+    {
+      "epoch": 0.42826188933812853,
+      "grad_norm": 0.42788875102996826,
+      "learning_rate": 6.385990095840055e-05,
+      "loss": 1.0313,
+      "step": 4793
+    },
+    {
+      "epoch": 0.4283512408693904,
+      "grad_norm": 0.4116227328777313,
+      "learning_rate": 6.384599672788802e-05,
+      "loss": 0.9863,
+      "step": 4794
+    },
+    {
+      "epoch": 0.42844059240065224,
+      "grad_norm": 0.3941948115825653,
+      "learning_rate": 6.383209133762569e-05,
+      "loss": 1.0179,
+      "step": 4795
+    },
+    {
+      "epoch": 0.4285299439319141,
+      "grad_norm": 0.41168758273124695,
+      "learning_rate": 6.381818478877825e-05,
+      "loss": 0.979,
+      "step": 4796
+    },
+    {
+      "epoch": 0.428619295463176,
+      "grad_norm": 0.5500960946083069,
+      "learning_rate": 6.380427708251054e-05,
+      "loss": 0.9703,
+      "step": 4797
+    },
+    {
+      "epoch": 0.4287086469944379,
+      "grad_norm": 0.42650794982910156,
+      "learning_rate": 6.379036821998751e-05,
+      "loss": 0.9895,
+      "step": 4798
+    },
+    {
+      "epoch": 0.4287979985256997,
+      "grad_norm": 0.4761294424533844,
+      "learning_rate": 6.377645820237412e-05,
+      "loss": 0.9629,
+      "step": 4799
+    },
+    {
+      "epoch": 0.4288873500569616,
+      "grad_norm": 0.45492222905158997,
+      "learning_rate": 6.376254703083552e-05,
+      "loss": 0.9479,
+      "step": 4800
+    },
+    {
+      "epoch": 0.4289767015882235,
+      "grad_norm": 0.44997844099998474,
+      "learning_rate": 6.374863470653691e-05,
+      "loss": 0.9265,
+      "step": 4801
+    },
+    {
+      "epoch": 0.4290660531194853,
+      "grad_norm": 0.5663099884986877,
+      "learning_rate": 6.373472123064358e-05,
+      "loss": 0.8915,
+      "step": 4802
+    },
+    {
+      "epoch": 0.4291554046507472,
+      "grad_norm": 0.41858309507369995,
+      "learning_rate": 6.372080660432095e-05,
+      "loss": 0.9733,
+      "step": 4803
+    },
+    {
+      "epoch": 0.4292447561820091,
+      "grad_norm": 0.42884135246276855,
+      "learning_rate": 6.370689082873451e-05,
+      "loss": 0.9741,
+      "step": 4804
+    },
+    {
+      "epoch": 0.42933410771327096,
+      "grad_norm": 0.49144938588142395,
+      "learning_rate": 6.369297390504987e-05,
+      "loss": 0.9202,
+      "step": 4805
+    },
+    {
+      "epoch": 0.4294234592445328,
+      "grad_norm": 0.43993905186653137,
+      "learning_rate": 6.36790558344327e-05,
+      "loss": 0.9182,
+      "step": 4806
+    },
+    {
+      "epoch": 0.4295128107757947,
+      "grad_norm": 0.41477641463279724,
+      "learning_rate": 6.36651366180488e-05,
+      "loss": 1.0702,
+      "step": 4807
+    },
+    {
+      "epoch": 0.42960216230705656,
+      "grad_norm": 0.48536887764930725,
+      "learning_rate": 6.365121625706405e-05,
+      "loss": 1.009,
+      "step": 4808
+    },
+    {
+      "epoch": 0.4296915138383184,
+      "grad_norm": 0.4555813670158386,
+      "learning_rate": 6.363729475264441e-05,
+      "loss": 0.9185,
+      "step": 4809
+    },
+    {
+      "epoch": 0.42978086536958027,
+      "grad_norm": 0.4251463711261749,
+      "learning_rate": 6.362337210595599e-05,
+      "loss": 0.9602,
+      "step": 4810
+    },
+    {
+      "epoch": 0.42987021690084215,
+      "grad_norm": 0.5072605013847351,
+      "learning_rate": 6.360944831816495e-05,
+      "loss": 0.9825,
+      "step": 4811
+    },
+    {
+      "epoch": 0.429959568432104,
+      "grad_norm": 0.39637628197669983,
+      "learning_rate": 6.359552339043753e-05,
+      "loss": 1.0062,
+      "step": 4812
+    },
+    {
+      "epoch": 0.43004891996336586,
+      "grad_norm": 0.45429277420043945,
+      "learning_rate": 6.35815973239401e-05,
+      "loss": 0.9989,
+      "step": 4813
+    },
+    {
+      "epoch": 0.43013827149462774,
+      "grad_norm": 0.4280487298965454,
+      "learning_rate": 6.356767011983915e-05,
+      "loss": 0.9106,
+      "step": 4814
+    },
+    {
+      "epoch": 0.4302276230258896,
+      "grad_norm": 0.45695221424102783,
+      "learning_rate": 6.355374177930118e-05,
+      "loss": 0.977,
+      "step": 4815
+    },
+    {
+      "epoch": 0.43031697455715145,
+      "grad_norm": 0.40981754660606384,
+      "learning_rate": 6.353981230349289e-05,
+      "loss": 1.0272,
+      "step": 4816
+    },
+    {
+      "epoch": 0.43040632608841334,
+      "grad_norm": 0.4471674859523773,
+      "learning_rate": 6.352588169358099e-05,
+      "loss": 0.9212,
+      "step": 4817
+    },
+    {
+      "epoch": 0.4304956776196752,
+      "grad_norm": 0.5394784808158875,
+      "learning_rate": 6.35119499507323e-05,
+      "loss": 0.9334,
+      "step": 4818
+    },
+    {
+      "epoch": 0.43058502915093705,
+      "grad_norm": 0.5647401213645935,
+      "learning_rate": 6.34980170761138e-05,
+      "loss": 0.9176,
+      "step": 4819
+    },
+    {
+      "epoch": 0.43067438068219893,
+      "grad_norm": 0.5009137392044067,
+      "learning_rate": 6.348408307089248e-05,
+      "loss": 0.9756,
+      "step": 4820
+    },
+    {
+      "epoch": 0.4307637322134608,
+      "grad_norm": 0.38283008337020874,
+      "learning_rate": 6.347014793623547e-05,
+      "loss": 1.0153,
+      "step": 4821
+    },
+    {
+      "epoch": 0.4308530837447227,
+      "grad_norm": 0.4172475337982178,
+      "learning_rate": 6.345621167331e-05,
+      "loss": 1.0625,
+      "step": 4822
+    },
+    {
+      "epoch": 0.4309424352759845,
+      "grad_norm": 0.3502405285835266,
+      "learning_rate": 6.344227428328335e-05,
+      "loss": 0.9735,
+      "step": 4823
+    },
+    {
+      "epoch": 0.4310317868072464,
+      "grad_norm": 0.43267232179641724,
+      "learning_rate": 6.342833576732297e-05,
+      "loss": 0.9561,
+      "step": 4824
+    },
+    {
+      "epoch": 0.4311211383385083,
+      "grad_norm": 0.41811031103134155,
+      "learning_rate": 6.341439612659631e-05,
+      "loss": 0.9465,
+      "step": 4825
+    },
+    {
+      "epoch": 0.4312104898697701,
+      "grad_norm": 0.4469202756881714,
+      "learning_rate": 6.340045536227101e-05,
+      "loss": 0.9639,
+      "step": 4826
+    },
+    {
+      "epoch": 0.431299841401032,
+      "grad_norm": 0.43867889046669006,
+      "learning_rate": 6.338651347551472e-05,
+      "loss": 0.9826,
+      "step": 4827
+    },
+    {
+      "epoch": 0.4313891929322939,
+      "grad_norm": 0.4113059341907501,
+      "learning_rate": 6.337257046749523e-05,
+      "loss": 1.0002,
+      "step": 4828
+    },
+    {
+      "epoch": 0.43147854446355577,
+      "grad_norm": 0.47658199071884155,
+      "learning_rate": 6.335862633938044e-05,
+      "loss": 0.9545,
+      "step": 4829
+    },
+    {
+      "epoch": 0.4315678959948176,
+      "grad_norm": 0.3787316083908081,
+      "learning_rate": 6.334468109233827e-05,
+      "loss": 1.033,
+      "step": 4830
+    },
+    {
+      "epoch": 0.4316572475260795,
+      "grad_norm": 0.5350416898727417,
+      "learning_rate": 6.333073472753686e-05,
+      "loss": 0.9514,
+      "step": 4831
+    },
+    {
+      "epoch": 0.43174659905734136,
+      "grad_norm": 0.40219366550445557,
+      "learning_rate": 6.331678724614429e-05,
+      "loss": 0.9928,
+      "step": 4832
+    },
+    {
+      "epoch": 0.4318359505886032,
+      "grad_norm": 0.457727313041687,
+      "learning_rate": 6.330283864932885e-05,
+      "loss": 0.9433,
+      "step": 4833
+    },
+    {
+      "epoch": 0.4319253021198651,
+      "grad_norm": 0.4098197817802429,
+      "learning_rate": 6.328888893825888e-05,
+      "loss": 1.0253,
+      "step": 4834
+    },
+    {
+      "epoch": 0.43201465365112696,
+      "grad_norm": 0.4684712886810303,
+      "learning_rate": 6.32749381141028e-05,
+      "loss": 0.9667,
+      "step": 4835
+    },
+    {
+      "epoch": 0.43210400518238884,
+      "grad_norm": 0.45592206716537476,
+      "learning_rate": 6.326098617802917e-05,
+      "loss": 0.9459,
+      "step": 4836
+    },
+    {
+      "epoch": 0.43219335671365067,
+      "grad_norm": 0.4041098952293396,
+      "learning_rate": 6.324703313120659e-05,
+      "loss": 1.0305,
+      "step": 4837
+    },
+    {
+      "epoch": 0.43228270824491255,
+      "grad_norm": 0.44690701365470886,
+      "learning_rate": 6.323307897480376e-05,
+      "loss": 0.9919,
+      "step": 4838
+    },
+    {
+      "epoch": 0.43237205977617443,
+      "grad_norm": 0.4994511008262634,
+      "learning_rate": 6.321912370998952e-05,
+      "loss": 0.921,
+      "step": 4839
+    },
+    {
+      "epoch": 0.43246141130743626,
+      "grad_norm": 0.4134766757488251,
+      "learning_rate": 6.320516733793278e-05,
+      "loss": 0.9395,
+      "step": 4840
+    },
+    {
+      "epoch": 0.43255076283869814,
+      "grad_norm": 0.517279863357544,
+      "learning_rate": 6.319120985980251e-05,
+      "loss": 0.9349,
+      "step": 4841
+    },
+    {
+      "epoch": 0.43264011436996,
+      "grad_norm": 0.5020187497138977,
+      "learning_rate": 6.317725127676781e-05,
+      "loss": 0.961,
+      "step": 4842
+    },
+    {
+      "epoch": 0.43272946590122185,
+      "grad_norm": 0.4880422353744507,
+      "learning_rate": 6.316329158999784e-05,
+      "loss": 0.9669,
+      "step": 4843
+    },
+    {
+      "epoch": 0.43281881743248374,
+      "grad_norm": 0.3964580297470093,
+      "learning_rate": 6.31493308006619e-05,
+      "loss": 0.9649,
+      "step": 4844
+    },
+    {
+      "epoch": 0.4329081689637456,
+      "grad_norm": 0.37571898102760315,
+      "learning_rate": 6.313536890992935e-05,
+      "loss": 1.0031,
+      "step": 4845
+    },
+    {
+      "epoch": 0.4329975204950075,
+      "grad_norm": 0.45982518792152405,
+      "learning_rate": 6.312140591896964e-05,
+      "loss": 0.9339,
+      "step": 4846
+    },
+    {
+      "epoch": 0.43308687202626933,
+      "grad_norm": 0.4254016876220703,
+      "learning_rate": 6.310744182895231e-05,
+      "loss": 1.0641,
+      "step": 4847
+    },
+    {
+      "epoch": 0.4331762235575312,
+      "grad_norm": 0.43494075536727905,
+      "learning_rate": 6.309347664104701e-05,
+      "loss": 0.9164,
+      "step": 4848
+    },
+    {
+      "epoch": 0.4332655750887931,
+      "grad_norm": 0.451556921005249,
+      "learning_rate": 6.307951035642349e-05,
+      "loss": 1.0389,
+      "step": 4849
+    },
+    {
+      "epoch": 0.4333549266200549,
+      "grad_norm": 0.34367331862449646,
+      "learning_rate": 6.306554297625156e-05,
+      "loss": 1.0079,
+      "step": 4850
+    },
+    {
+      "epoch": 0.4334442781513168,
+      "grad_norm": 0.5219529271125793,
+      "learning_rate": 6.305157450170111e-05,
+      "loss": 0.8767,
+      "step": 4851
+    },
+    {
+      "epoch": 0.4335336296825787,
+      "grad_norm": 0.36903488636016846,
+      "learning_rate": 6.303760493394221e-05,
+      "loss": 0.9979,
+      "step": 4852
+    },
+    {
+      "epoch": 0.4336229812138406,
+      "grad_norm": 0.37834039330482483,
+      "learning_rate": 6.302363427414491e-05,
+      "loss": 0.9771,
+      "step": 4853
+    },
+    {
+      "epoch": 0.4337123327451024,
+      "grad_norm": 0.4161636233329773,
+      "learning_rate": 6.300966252347942e-05,
+      "loss": 0.9805,
+      "step": 4854
+    },
+    {
+      "epoch": 0.4338016842763643,
+      "grad_norm": 0.5334766507148743,
+      "learning_rate": 6.299568968311601e-05,
+      "loss": 0.9568,
+      "step": 4855
+    },
+    {
+      "epoch": 0.43389103580762617,
+      "grad_norm": 0.4917714595794678,
+      "learning_rate": 6.298171575422508e-05,
+      "loss": 0.917,
+      "step": 4856
+    },
+    {
+      "epoch": 0.433980387338888,
+      "grad_norm": 0.42037853598594666,
+      "learning_rate": 6.296774073797708e-05,
+      "loss": 0.9624,
+      "step": 4857
+    },
+    {
+      "epoch": 0.4340697388701499,
+      "grad_norm": 0.3785925805568695,
+      "learning_rate": 6.295376463554255e-05,
+      "loss": 1.0437,
+      "step": 4858
+    },
+    {
+      "epoch": 0.43415909040141176,
+      "grad_norm": 0.47149381041526794,
+      "learning_rate": 6.293978744809217e-05,
+      "loss": 0.9401,
+      "step": 4859
+    },
+    {
+      "epoch": 0.43424844193267365,
+      "grad_norm": 0.4850710928440094,
+      "learning_rate": 6.292580917679665e-05,
+      "loss": 0.9874,
+      "step": 4860
+    },
+    {
+      "epoch": 0.4343377934639355,
+      "grad_norm": 0.43360236287117004,
+      "learning_rate": 6.291182982282685e-05,
+      "loss": 0.9758,
+      "step": 4861
+    },
+    {
+      "epoch": 0.43442714499519736,
+      "grad_norm": 0.41151541471481323,
+      "learning_rate": 6.289784938735366e-05,
+      "loss": 1.0028,
+      "step": 4862
+    },
+    {
+      "epoch": 0.43451649652645924,
+      "grad_norm": 0.46998703479766846,
+      "learning_rate": 6.28838678715481e-05,
+      "loss": 0.9591,
+      "step": 4863
+    },
+    {
+      "epoch": 0.43460584805772107,
+      "grad_norm": 0.5863698124885559,
+      "learning_rate": 6.286988527658129e-05,
+      "loss": 0.9661,
+      "step": 4864
+    },
+    {
+      "epoch": 0.43469519958898295,
+      "grad_norm": 0.434731662273407,
+      "learning_rate": 6.285590160362438e-05,
+      "loss": 0.9454,
+      "step": 4865
+    },
+    {
+      "epoch": 0.43478455112024483,
+      "grad_norm": 0.5044207572937012,
+      "learning_rate": 6.28419168538487e-05,
+      "loss": 0.9308,
+      "step": 4866
+    },
+    {
+      "epoch": 0.4348739026515067,
+      "grad_norm": 0.4645664095878601,
+      "learning_rate": 6.282793102842559e-05,
+      "loss": 0.9077,
+      "step": 4867
+    },
+    {
+      "epoch": 0.43496325418276854,
+      "grad_norm": 0.43170252442359924,
+      "learning_rate": 6.281394412852652e-05,
+      "loss": 1.0509,
+      "step": 4868
+    },
+    {
+      "epoch": 0.4350526057140304,
+      "grad_norm": 0.41620826721191406,
+      "learning_rate": 6.279995615532304e-05,
+      "loss": 0.9653,
+      "step": 4869
+    },
+    {
+      "epoch": 0.4351419572452923,
+      "grad_norm": 0.3965005874633789,
+      "learning_rate": 6.27859671099868e-05,
+      "loss": 0.9601,
+      "step": 4870
+    },
+    {
+      "epoch": 0.43523130877655414,
+      "grad_norm": 0.3978107273578644,
+      "learning_rate": 6.277197699368954e-05,
+      "loss": 1.0369,
+      "step": 4871
+    },
+    {
+      "epoch": 0.435320660307816,
+      "grad_norm": 0.42217400670051575,
+      "learning_rate": 6.275798580760304e-05,
+      "loss": 0.9626,
+      "step": 4872
+    },
+    {
+      "epoch": 0.4354100118390779,
+      "grad_norm": 0.4375147223472595,
+      "learning_rate": 6.274399355289923e-05,
+      "loss": 0.9481,
+      "step": 4873
+    },
+    {
+      "epoch": 0.43549936337033973,
+      "grad_norm": 0.4031033515930176,
+      "learning_rate": 6.273000023075014e-05,
+      "loss": 0.9599,
+      "step": 4874
+    },
+    {
+      "epoch": 0.4355887149016016,
+      "grad_norm": 0.50283282995224,
+      "learning_rate": 6.271600584232784e-05,
+      "loss": 0.9854,
+      "step": 4875
+    },
+    {
+      "epoch": 0.4356780664328635,
+      "grad_norm": 0.5138799548149109,
+      "learning_rate": 6.27020103888045e-05,
+      "loss": 1.0223,
+      "step": 4876
+    },
+    {
+      "epoch": 0.4357674179641254,
+      "grad_norm": 0.37269532680511475,
+      "learning_rate": 6.26880138713524e-05,
+      "loss": 0.9626,
+      "step": 4877
+    },
+    {
+      "epoch": 0.4358567694953872,
+      "grad_norm": 0.4185316264629364,
+      "learning_rate": 6.267401629114389e-05,
+      "loss": 0.9613,
+      "step": 4878
+    },
+    {
+      "epoch": 0.4359461210266491,
+      "grad_norm": 0.41371724009513855,
+      "learning_rate": 6.266001764935144e-05,
+      "loss": 0.9528,
+      "step": 4879
+    },
+    {
+      "epoch": 0.436035472557911,
+      "grad_norm": 0.392198383808136,
+      "learning_rate": 6.264601794714753e-05,
+      "loss": 0.9722,
+      "step": 4880
+    },
+    {
+      "epoch": 0.4361248240891728,
+      "grad_norm": 0.39201459288597107,
+      "learning_rate": 6.263201718570485e-05,
+      "loss": 0.9175,
+      "step": 4881
+    },
+    {
+      "epoch": 0.4362141756204347,
+      "grad_norm": 0.5743230581283569,
+      "learning_rate": 6.261801536619607e-05,
+      "loss": 0.9564,
+      "step": 4882
+    },
+    {
+      "epoch": 0.43630352715169657,
+      "grad_norm": 0.5308464765548706,
+      "learning_rate": 6.2604012489794e-05,
+      "loss": 0.9499,
+      "step": 4883
+    },
+    {
+      "epoch": 0.43639287868295845,
+      "grad_norm": 0.4258040487766266,
+      "learning_rate": 6.259000855767155e-05,
+      "loss": 0.9583,
+      "step": 4884
+    },
+    {
+      "epoch": 0.4364822302142203,
+      "grad_norm": 0.45301637053489685,
+      "learning_rate": 6.257600357100167e-05,
+      "loss": 0.9621,
+      "step": 4885
+    },
+    {
+      "epoch": 0.43657158174548216,
+      "grad_norm": 0.5080031752586365,
+      "learning_rate": 6.256199753095745e-05,
+      "loss": 0.9353,
+      "step": 4886
+    },
+    {
+      "epoch": 0.43666093327674405,
+      "grad_norm": 0.41666001081466675,
+      "learning_rate": 6.254799043871204e-05,
+      "loss": 1.0159,
+      "step": 4887
+    },
+    {
+      "epoch": 0.4367502848080059,
+      "grad_norm": 0.4304046928882599,
+      "learning_rate": 6.253398229543867e-05,
+      "loss": 0.9442,
+      "step": 4888
+    },
+    {
+      "epoch": 0.43683963633926776,
+      "grad_norm": 0.41285210847854614,
+      "learning_rate": 6.251997310231067e-05,
+      "loss": 0.9862,
+      "step": 4889
+    },
+    {
+      "epoch": 0.43692898787052964,
+      "grad_norm": 0.5112424492835999,
+      "learning_rate": 6.250596286050148e-05,
+      "loss": 0.9551,
+      "step": 4890
+    },
+    {
+      "epoch": 0.4370183394017915,
+      "grad_norm": 0.46234673261642456,
+      "learning_rate": 6.249195157118461e-05,
+      "loss": 0.9087,
+      "step": 4891
+    },
+    {
+      "epoch": 0.43710769093305335,
+      "grad_norm": 0.43776413798332214,
+      "learning_rate": 6.247793923553362e-05,
+      "loss": 0.9494,
+      "step": 4892
+    },
+    {
+      "epoch": 0.43719704246431523,
+      "grad_norm": 0.42224618792533875,
+      "learning_rate": 6.246392585472222e-05,
+      "loss": 1.0397,
+      "step": 4893
+    },
+    {
+      "epoch": 0.4372863939955771,
+      "grad_norm": 0.4742605984210968,
+      "learning_rate": 6.244991142992417e-05,
+      "loss": 0.9794,
+      "step": 4894
+    },
+    {
+      "epoch": 0.43737574552683894,
+      "grad_norm": 0.5097817778587341,
+      "learning_rate": 6.243589596231333e-05,
+      "loss": 0.9319,
+      "step": 4895
+    },
+    {
+      "epoch": 0.4374650970581008,
+      "grad_norm": 0.4433581829071045,
+      "learning_rate": 6.242187945306364e-05,
+      "loss": 0.9704,
+      "step": 4896
+    },
+    {
+      "epoch": 0.4375544485893627,
+      "grad_norm": 0.3601015508174896,
+      "learning_rate": 6.240786190334918e-05,
+      "loss": 1.0209,
+      "step": 4897
+    },
+    {
+      "epoch": 0.4376438001206246,
+      "grad_norm": 0.6255925893783569,
+      "learning_rate": 6.239384331434399e-05,
+      "loss": 0.986,
+      "step": 4898
+    },
+    {
+      "epoch": 0.4377331516518864,
+      "grad_norm": 0.45729881525039673,
+      "learning_rate": 6.237982368722232e-05,
+      "loss": 1.0179,
+      "step": 4899
+    },
+    {
+      "epoch": 0.4378225031831483,
+      "grad_norm": 0.41129839420318604,
+      "learning_rate": 6.236580302315844e-05,
+      "loss": 0.9502,
+      "step": 4900
+    },
+    {
+      "epoch": 0.4379118547144102,
+      "grad_norm": 0.4457799792289734,
+      "learning_rate": 6.235178132332677e-05,
+      "loss": 0.9722,
+      "step": 4901
+    },
+    {
+      "epoch": 0.438001206245672,
+      "grad_norm": 0.492878794670105,
+      "learning_rate": 6.233775858890175e-05,
+      "loss": 0.9167,
+      "step": 4902
+    },
+    {
+      "epoch": 0.4380905577769339,
+      "grad_norm": 0.43004390597343445,
+      "learning_rate": 6.232373482105794e-05,
+      "loss": 0.9728,
+      "step": 4903
+    },
+    {
+      "epoch": 0.4381799093081958,
+      "grad_norm": 0.4913460314273834,
+      "learning_rate": 6.230971002096999e-05,
+      "loss": 1.05,
+      "step": 4904
+    },
+    {
+      "epoch": 0.43826926083945766,
+      "grad_norm": 0.47525662183761597,
+      "learning_rate": 6.229568418981258e-05,
+      "loss": 0.9822,
+      "step": 4905
+    },
+    {
+      "epoch": 0.4383586123707195,
+      "grad_norm": 0.4017292857170105,
+      "learning_rate": 6.228165732876056e-05,
+      "loss": 0.9573,
+      "step": 4906
+    },
+    {
+      "epoch": 0.4384479639019814,
+      "grad_norm": 0.41514596343040466,
+      "learning_rate": 6.226762943898886e-05,
+      "loss": 0.9435,
+      "step": 4907
+    },
+    {
+      "epoch": 0.43853731543324326,
+      "grad_norm": 0.4426291286945343,
+      "learning_rate": 6.22536005216724e-05,
+      "loss": 1.0117,
+      "step": 4908
+    },
+    {
+      "epoch": 0.4386266669645051,
+      "grad_norm": 0.42675042152404785,
+      "learning_rate": 6.223957057798629e-05,
+      "loss": 0.9363,
+      "step": 4909
+    },
+    {
+      "epoch": 0.43871601849576697,
+      "grad_norm": 0.5054933428764343,
+      "learning_rate": 6.222553960910567e-05,
+      "loss": 0.961,
+      "step": 4910
+    },
+    {
+      "epoch": 0.43880537002702885,
+      "grad_norm": 0.37240302562713623,
+      "learning_rate": 6.221150761620581e-05,
+      "loss": 1.098,
+      "step": 4911
+    },
+    {
+      "epoch": 0.4388947215582907,
+      "grad_norm": 0.37764713168144226,
+      "learning_rate": 6.219747460046203e-05,
+      "loss": 0.9538,
+      "step": 4912
+    },
+    {
+      "epoch": 0.43898407308955256,
+      "grad_norm": 0.49377158284187317,
+      "learning_rate": 6.218344056304972e-05,
+      "loss": 0.9257,
+      "step": 4913
+    },
+    {
+      "epoch": 0.43907342462081445,
+      "grad_norm": 0.44281005859375,
+      "learning_rate": 6.216940550514439e-05,
+      "loss": 0.9663,
+      "step": 4914
+    },
+    {
+      "epoch": 0.43916277615207633,
+      "grad_norm": 0.42006000876426697,
+      "learning_rate": 6.215536942792163e-05,
+      "loss": 0.9435,
+      "step": 4915
+    },
+    {
+      "epoch": 0.43925212768333816,
+      "grad_norm": 0.468666136264801,
+      "learning_rate": 6.214133233255713e-05,
+      "loss": 0.9804,
+      "step": 4916
+    },
+    {
+      "epoch": 0.43934147921460004,
+      "grad_norm": 0.3761788308620453,
+      "learning_rate": 6.212729422022664e-05,
+      "loss": 1.0003,
+      "step": 4917
+    },
+    {
+      "epoch": 0.4394308307458619,
+      "grad_norm": 0.4250943064689636,
+      "learning_rate": 6.211325509210597e-05,
+      "loss": 0.9809,
+      "step": 4918
+    },
+    {
+      "epoch": 0.43952018227712375,
+      "grad_norm": 0.40891769528388977,
+      "learning_rate": 6.209921494937108e-05,
+      "loss": 0.9841,
+      "step": 4919
+    },
+    {
+      "epoch": 0.43960953380838563,
+      "grad_norm": 0.3647308945655823,
+      "learning_rate": 6.208517379319796e-05,
+      "loss": 0.9747,
+      "step": 4920
+    },
+    {
+      "epoch": 0.4396988853396475,
+      "grad_norm": 0.4176364541053772,
+      "learning_rate": 6.207113162476272e-05,
+      "loss": 0.9336,
+      "step": 4921
+    },
+    {
+      "epoch": 0.4397882368709094,
+      "grad_norm": 0.40200045704841614,
+      "learning_rate": 6.205708844524153e-05,
+      "loss": 1.0223,
+      "step": 4922
+    },
+    {
+      "epoch": 0.4398775884021712,
+      "grad_norm": 0.4894162714481354,
+      "learning_rate": 6.204304425581069e-05,
+      "loss": 0.9348,
+      "step": 4923
+    },
+    {
+      "epoch": 0.4399669399334331,
+      "grad_norm": 0.4525613784790039,
+      "learning_rate": 6.20289990576465e-05,
+      "loss": 0.941,
+      "step": 4924
+    },
+    {
+      "epoch": 0.440056291464695,
+      "grad_norm": 0.4009491205215454,
+      "learning_rate": 6.201495285192542e-05,
+      "loss": 1.0821,
+      "step": 4925
+    },
+    {
+      "epoch": 0.4401456429959568,
+      "grad_norm": 0.4592028558254242,
+      "learning_rate": 6.200090563982397e-05,
+      "loss": 0.9359,
+      "step": 4926
+    },
+    {
+      "epoch": 0.4402349945272187,
+      "grad_norm": 0.45449405908584595,
+      "learning_rate": 6.198685742251877e-05,
+      "loss": 0.9356,
+      "step": 4927
+    },
+    {
+      "epoch": 0.4403243460584806,
+      "grad_norm": 0.6728967428207397,
+      "learning_rate": 6.197280820118646e-05,
+      "loss": 1.0374,
+      "step": 4928
+    },
+    {
+      "epoch": 0.44041369758974247,
+      "grad_norm": 0.5596892833709717,
+      "learning_rate": 6.195875797700385e-05,
+      "loss": 0.9524,
+      "step": 4929
+    },
+    {
+      "epoch": 0.4405030491210043,
+      "grad_norm": 0.4790952801704407,
+      "learning_rate": 6.19447067511478e-05,
+      "loss": 1.0068,
+      "step": 4930
+    },
+    {
+      "epoch": 0.4405924006522662,
+      "grad_norm": 0.4227498173713684,
+      "learning_rate": 6.193065452479523e-05,
+      "loss": 1.0209,
+      "step": 4931
+    },
+    {
+      "epoch": 0.44068175218352806,
+      "grad_norm": 0.4277952015399933,
+      "learning_rate": 6.191660129912317e-05,
+      "loss": 0.8967,
+      "step": 4932
+    },
+    {
+      "epoch": 0.4407711037147899,
+      "grad_norm": 0.4384140074253082,
+      "learning_rate": 6.190254707530874e-05,
+      "loss": 0.9636,
+      "step": 4933
+    },
+    {
+      "epoch": 0.4408604552460518,
+      "grad_norm": 0.4057149589061737,
+      "learning_rate": 6.188849185452911e-05,
+      "loss": 1.0107,
+      "step": 4934
+    },
+    {
+      "epoch": 0.44094980677731366,
+      "grad_norm": 0.43641364574432373,
+      "learning_rate": 6.187443563796157e-05,
+      "loss": 0.9296,
+      "step": 4935
+    },
+    {
+      "epoch": 0.44103915830857554,
+      "grad_norm": 0.4788879156112671,
+      "learning_rate": 6.186037842678349e-05,
+      "loss": 0.9666,
+      "step": 4936
+    },
+    {
+      "epoch": 0.44112850983983737,
+      "grad_norm": 0.3350755572319031,
+      "learning_rate": 6.184632022217227e-05,
+      "loss": 1.0117,
+      "step": 4937
+    },
+    {
+      "epoch": 0.44121786137109925,
+      "grad_norm": 0.4167011082172394,
+      "learning_rate": 6.183226102530547e-05,
+      "loss": 1.023,
+      "step": 4938
+    },
+    {
+      "epoch": 0.44130721290236113,
+      "grad_norm": 0.47220227122306824,
+      "learning_rate": 6.181820083736067e-05,
+      "loss": 1.0494,
+      "step": 4939
+    },
+    {
+      "epoch": 0.44139656443362296,
+      "grad_norm": 0.4266274571418762,
+      "learning_rate": 6.18041396595156e-05,
+      "loss": 0.9815,
+      "step": 4940
+    },
+    {
+      "epoch": 0.44148591596488485,
+      "grad_norm": 0.4164573848247528,
+      "learning_rate": 6.1790077492948e-05,
+      "loss": 1.0067,
+      "step": 4941
+    },
+    {
+      "epoch": 0.44157526749614673,
+      "grad_norm": 0.4291037917137146,
+      "learning_rate": 6.177601433883573e-05,
+      "loss": 0.9725,
+      "step": 4942
+    },
+    {
+      "epoch": 0.44166461902740856,
+      "grad_norm": 0.49879229068756104,
+      "learning_rate": 6.176195019835674e-05,
+      "loss": 0.9782,
+      "step": 4943
+    },
+    {
+      "epoch": 0.44175397055867044,
+      "grad_norm": 0.5074213743209839,
+      "learning_rate": 6.174788507268905e-05,
+      "loss": 0.9614,
+      "step": 4944
+    },
+    {
+      "epoch": 0.4418433220899323,
+      "grad_norm": 0.39095085859298706,
+      "learning_rate": 6.173381896301076e-05,
+      "loss": 0.9515,
+      "step": 4945
+    },
+    {
+      "epoch": 0.4419326736211942,
+      "grad_norm": 0.40660542249679565,
+      "learning_rate": 6.171975187050005e-05,
+      "loss": 0.9768,
+      "step": 4946
+    },
+    {
+      "epoch": 0.44202202515245603,
+      "grad_norm": 0.6411069631576538,
+      "learning_rate": 6.170568379633522e-05,
+      "loss": 1.0163,
+      "step": 4947
+    },
+    {
+      "epoch": 0.4421113766837179,
+      "grad_norm": 0.4861661195755005,
+      "learning_rate": 6.169161474169458e-05,
+      "loss": 0.9521,
+      "step": 4948
+    },
+    {
+      "epoch": 0.4422007282149798,
+      "grad_norm": 0.4027434289455414,
+      "learning_rate": 6.167754470775659e-05,
+      "loss": 0.9752,
+      "step": 4949
+    },
+    {
+      "epoch": 0.4422900797462416,
+      "grad_norm": 0.46598201990127563,
+      "learning_rate": 6.166347369569975e-05,
+      "loss": 0.9576,
+      "step": 4950
+    },
+    {
+      "epoch": 0.4423794312775035,
+      "grad_norm": 0.4167962372303009,
+      "learning_rate": 6.164940170670266e-05,
+      "loss": 0.9705,
+      "step": 4951
+    },
+    {
+      "epoch": 0.4424687828087654,
+      "grad_norm": 0.4604121744632721,
+      "learning_rate": 6.163532874194401e-05,
+      "loss": 0.9491,
+      "step": 4952
+    },
+    {
+      "epoch": 0.4425581343400273,
+      "grad_norm": 0.4519181549549103,
+      "learning_rate": 6.162125480260257e-05,
+      "loss": 0.9529,
+      "step": 4953
+    },
+    {
+      "epoch": 0.4426474858712891,
+      "grad_norm": 0.3763716220855713,
+      "learning_rate": 6.160717988985714e-05,
+      "loss": 0.962,
+      "step": 4954
+    },
+    {
+      "epoch": 0.442736837402551,
+      "grad_norm": 0.3946007788181305,
+      "learning_rate": 6.15931040048867e-05,
+      "loss": 1.0416,
+      "step": 4955
+    },
+    {
+      "epoch": 0.44282618893381287,
+      "grad_norm": 0.4197639226913452,
+      "learning_rate": 6.15790271488702e-05,
+      "loss": 1.05,
+      "step": 4956
+    },
+    {
+      "epoch": 0.4429155404650747,
+      "grad_norm": 0.40052515268325806,
+      "learning_rate": 6.156494932298678e-05,
+      "loss": 0.95,
+      "step": 4957
+    },
+    {
+      "epoch": 0.4430048919963366,
+      "grad_norm": 0.4258505702018738,
+      "learning_rate": 6.155087052841555e-05,
+      "loss": 0.9829,
+      "step": 4958
+    },
+    {
+      "epoch": 0.44309424352759846,
+      "grad_norm": 0.41184738278388977,
+      "learning_rate": 6.153679076633581e-05,
+      "loss": 0.9795,
+      "step": 4959
+    },
+    {
+      "epoch": 0.44318359505886035,
+      "grad_norm": 0.5452486872673035,
+      "learning_rate": 6.152271003792686e-05,
+      "loss": 0.8949,
+      "step": 4960
+    },
+    {
+      "epoch": 0.4432729465901222,
+      "grad_norm": 0.47022590041160583,
+      "learning_rate": 6.150862834436811e-05,
+      "loss": 0.9618,
+      "step": 4961
+    },
+    {
+      "epoch": 0.44336229812138406,
+      "grad_norm": 0.4404926300048828,
+      "learning_rate": 6.149454568683909e-05,
+      "loss": 1.0751,
+      "step": 4962
+    },
+    {
+      "epoch": 0.44345164965264594,
+      "grad_norm": 0.4911143481731415,
+      "learning_rate": 6.148046206651932e-05,
+      "loss": 0.9599,
+      "step": 4963
+    },
+    {
+      "epoch": 0.44354100118390777,
+      "grad_norm": 0.4653252959251404,
+      "learning_rate": 6.146637748458849e-05,
+      "loss": 1.0376,
+      "step": 4964
+    },
+    {
+      "epoch": 0.44363035271516965,
+      "grad_norm": 0.5367299914360046,
+      "learning_rate": 6.145229194222633e-05,
+      "loss": 0.9954,
+      "step": 4965
+    },
+    {
+      "epoch": 0.44371970424643153,
+      "grad_norm": 0.39032191038131714,
+      "learning_rate": 6.143820544061263e-05,
+      "loss": 0.9678,
+      "step": 4966
+    },
+    {
+      "epoch": 0.4438090557776934,
+      "grad_norm": 0.46807265281677246,
+      "learning_rate": 6.142411798092731e-05,
+      "loss": 0.9434,
+      "step": 4967
+    },
+    {
+      "epoch": 0.44389840730895525,
+      "grad_norm": 0.3878948390483856,
+      "learning_rate": 6.141002956435034e-05,
+      "loss": 0.9584,
+      "step": 4968
+    },
+    {
+      "epoch": 0.44398775884021713,
+      "grad_norm": 0.4154819846153259,
+      "learning_rate": 6.139594019206178e-05,
+      "loss": 0.9433,
+      "step": 4969
+    },
+    {
+      "epoch": 0.444077110371479,
+      "grad_norm": 0.4845220744609833,
+      "learning_rate": 6.138184986524175e-05,
+      "loss": 1.0408,
+      "step": 4970
+    },
+    {
+      "epoch": 0.44416646190274084,
+      "grad_norm": 0.4787101149559021,
+      "learning_rate": 6.136775858507046e-05,
+      "loss": 0.9348,
+      "step": 4971
+    },
+    {
+      "epoch": 0.4442558134340027,
+      "grad_norm": 0.4312731921672821,
+      "learning_rate": 6.135366635272824e-05,
+      "loss": 0.9757,
+      "step": 4972
+    },
+    {
+      "epoch": 0.4443451649652646,
+      "grad_norm": 0.4321444034576416,
+      "learning_rate": 6.133957316939543e-05,
+      "loss": 0.9491,
+      "step": 4973
+    },
+    {
+      "epoch": 0.44443451649652643,
+      "grad_norm": 0.5200006365776062,
+      "learning_rate": 6.132547903625249e-05,
+      "loss": 0.897,
+      "step": 4974
+    },
+    {
+      "epoch": 0.4445238680277883,
+      "grad_norm": 0.44951632618904114,
+      "learning_rate": 6.131138395447997e-05,
+      "loss": 0.991,
+      "step": 4975
+    },
+    {
+      "epoch": 0.4446132195590502,
+      "grad_norm": 0.556250274181366,
+      "learning_rate": 6.129728792525846e-05,
+      "loss": 0.9118,
+      "step": 4976
+    },
+    {
+      "epoch": 0.4447025710903121,
+      "grad_norm": 0.5766691565513611,
+      "learning_rate": 6.128319094976868e-05,
+      "loss": 0.9496,
+      "step": 4977
+    },
+    {
+      "epoch": 0.4447919226215739,
+      "grad_norm": 0.472476601600647,
+      "learning_rate": 6.126909302919138e-05,
+      "loss": 0.9091,
+      "step": 4978
+    },
+    {
+      "epoch": 0.4448812741528358,
+      "grad_norm": 0.41769498586654663,
+      "learning_rate": 6.125499416470742e-05,
+      "loss": 1.0127,
+      "step": 4979
+    },
+    {
+      "epoch": 0.4449706256840977,
+      "grad_norm": 0.44769975543022156,
+      "learning_rate": 6.124089435749772e-05,
+      "loss": 0.9084,
+      "step": 4980
+    },
+    {
+      "epoch": 0.4450599772153595,
+      "grad_norm": 0.42755383253097534,
+      "learning_rate": 6.122679360874331e-05,
+      "loss": 1.0314,
+      "step": 4981
+    },
+    {
+      "epoch": 0.4451493287466214,
+      "grad_norm": 0.4499891698360443,
+      "learning_rate": 6.121269191962527e-05,
+      "loss": 0.9204,
+      "step": 4982
+    },
+    {
+      "epoch": 0.44523868027788327,
+      "grad_norm": 0.5186624526977539,
+      "learning_rate": 6.119858929132475e-05,
+      "loss": 0.973,
+      "step": 4983
+    },
+    {
+      "epoch": 0.44532803180914515,
+      "grad_norm": 0.43319734930992126,
+      "learning_rate": 6.118448572502302e-05,
+      "loss": 1.0126,
+      "step": 4984
+    },
+    {
+      "epoch": 0.445417383340407,
+      "grad_norm": 0.41849610209465027,
+      "learning_rate": 6.117038122190139e-05,
+      "loss": 1.0017,
+      "step": 4985
+    },
+    {
+      "epoch": 0.44550673487166886,
+      "grad_norm": 0.5016182065010071,
+      "learning_rate": 6.115627578314125e-05,
+      "loss": 0.9203,
+      "step": 4986
+    },
+    {
+      "epoch": 0.44559608640293075,
+      "grad_norm": 0.47967931628227234,
+      "learning_rate": 6.114216940992411e-05,
+      "loss": 0.9842,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4456854379341926,
+      "grad_norm": 0.4858255088329315,
+      "learning_rate": 6.112806210343152e-05,
+      "loss": 0.9719,
+      "step": 4988
+    },
+    {
+      "epoch": 0.44577478946545446,
+      "grad_norm": 0.5705578327178955,
+      "learning_rate": 6.111395386484511e-05,
+      "loss": 0.8807,
+      "step": 4989
+    },
+    {
+      "epoch": 0.44586414099671634,
+      "grad_norm": 0.4477182626724243,
+      "learning_rate": 6.109984469534659e-05,
+      "loss": 0.9386,
+      "step": 4990
+    },
+    {
+      "epoch": 0.4459534925279782,
+      "grad_norm": 0.4489213824272156,
+      "learning_rate": 6.108573459611776e-05,
+      "loss": 1.0481,
+      "step": 4991
+    },
+    {
+      "epoch": 0.44604284405924005,
+      "grad_norm": 0.36626997590065,
+      "learning_rate": 6.10716235683405e-05,
+      "loss": 0.908,
+      "step": 4992
+    },
+    {
+      "epoch": 0.44613219559050193,
+      "grad_norm": 0.409940630197525,
+      "learning_rate": 6.105751161319675e-05,
+      "loss": 0.9784,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4462215471217638,
+      "grad_norm": 0.42009782791137695,
+      "learning_rate": 6.104339873186855e-05,
+      "loss": 1.0009,
+      "step": 4994
+    },
+    {
+      "epoch": 0.44631089865302565,
+      "grad_norm": 0.4435548186302185,
+      "learning_rate": 6.102928492553796e-05,
+      "loss": 1.0153,
+      "step": 4995
+    },
+    {
+      "epoch": 0.44640025018428753,
+      "grad_norm": 0.574257493019104,
+      "learning_rate": 6.101517019538721e-05,
+      "loss": 0.8759,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4464896017155494,
+      "grad_norm": 0.43628671765327454,
+      "learning_rate": 6.1001054542598534e-05,
+      "loss": 0.9219,
+      "step": 4997
+    },
+    {
+      "epoch": 0.4465789532468113,
+      "grad_norm": 0.3821120262145996,
+      "learning_rate": 6.0986937968354295e-05,
+      "loss": 1.0096,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4466683047780731,
+      "grad_norm": 0.41857802867889404,
+      "learning_rate": 6.097282047383688e-05,
+      "loss": 0.9917,
+      "step": 4999
+    },
+    {
+      "epoch": 0.446757656309335,
+      "grad_norm": 0.6450624465942383,
+      "learning_rate": 6.095870206022879e-05,
+      "loss": 0.8897,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4468470078405969,
+      "grad_norm": 0.4506264925003052,
+      "learning_rate": 6.0944582728712585e-05,
+      "loss": 0.9337,
+      "step": 5001
+    },
+    {
+      "epoch": 0.4469363593718587,
+      "grad_norm": 0.5254955887794495,
+      "learning_rate": 6.093046248047092e-05,
+      "loss": 0.9374,
+      "step": 5002
+    },
+    {
+      "epoch": 0.4470257109031206,
+      "grad_norm": 0.4589659869670868,
+      "learning_rate": 6.091634131668652e-05,
+      "loss": 0.9522,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4471150624343825,
+      "grad_norm": 0.5653790831565857,
+      "learning_rate": 6.090221923854217e-05,
+      "loss": 0.9442,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4472044139656443,
+      "grad_norm": 0.4403340220451355,
+      "learning_rate": 6.088809624722074e-05,
+      "loss": 0.9674,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4472937654969062,
+      "grad_norm": 0.5026565790176392,
+      "learning_rate": 6.0873972343905206e-05,
+      "loss": 0.9168,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4473831170281681,
+      "grad_norm": 0.4866418242454529,
+      "learning_rate": 6.085984752977857e-05,
+      "loss": 0.9227,
+      "step": 5007
+    },
+    {
+      "epoch": 0.44747246855942996,
+      "grad_norm": 0.43236756324768066,
+      "learning_rate": 6.0845721806023945e-05,
+      "loss": 0.9651,
+      "step": 5008
+    },
+    {
+      "epoch": 0.4475618200906918,
+      "grad_norm": 0.49595317244529724,
+      "learning_rate": 6.083159517382452e-05,
+      "loss": 0.8973,
+      "step": 5009
+    },
+    {
+      "epoch": 0.44765117162195367,
+      "grad_norm": 0.3765192925930023,
+      "learning_rate": 6.0817467634363535e-05,
+      "loss": 0.9991,
+      "step": 5010
+    },
+    {
+      "epoch": 0.44774052315321555,
+      "grad_norm": 0.49171602725982666,
+      "learning_rate": 6.0803339188824326e-05,
+      "loss": 1.0389,
+      "step": 5011
+    },
+    {
+      "epoch": 0.4478298746844774,
+      "grad_norm": 0.47733286023139954,
+      "learning_rate": 6.078920983839031e-05,
+      "loss": 0.974,
+      "step": 5012
+    },
+    {
+      "epoch": 0.44791922621573926,
+      "grad_norm": 0.5364671945571899,
+      "learning_rate": 6.0775079584244976e-05,
+      "loss": 0.957,
+      "step": 5013
+    },
+    {
+      "epoch": 0.44800857774700115,
+      "grad_norm": 0.5192524790763855,
+      "learning_rate": 6.076094842757185e-05,
+      "loss": 0.8715,
+      "step": 5014
+    },
+    {
+      "epoch": 0.44809792927826303,
+      "grad_norm": 0.38916611671447754,
+      "learning_rate": 6.07468163695546e-05,
+      "loss": 1.0494,
+      "step": 5015
+    },
+    {
+      "epoch": 0.44818728080952486,
+      "grad_norm": 0.4378674030303955,
+      "learning_rate": 6.0732683411376935e-05,
+      "loss": 1.0302,
+      "step": 5016
+    },
+    {
+      "epoch": 0.44827663234078674,
+      "grad_norm": 0.4529155492782593,
+      "learning_rate": 6.0718549554222614e-05,
+      "loss": 0.9895,
+      "step": 5017
+    },
+    {
+      "epoch": 0.4483659838720486,
+      "grad_norm": 0.4706246256828308,
+      "learning_rate": 6.070441479927554e-05,
+      "loss": 0.9995,
+      "step": 5018
+    },
+    {
+      "epoch": 0.44845533540331045,
+      "grad_norm": 0.515957772731781,
+      "learning_rate": 6.069027914771961e-05,
+      "loss": 0.898,
+      "step": 5019
+    },
+    {
+      "epoch": 0.44854468693457233,
+      "grad_norm": 0.43126967549324036,
+      "learning_rate": 6.067614260073885e-05,
+      "loss": 1.0187,
+      "step": 5020
+    },
+    {
+      "epoch": 0.4486340384658342,
+      "grad_norm": 0.44768694043159485,
+      "learning_rate": 6.0662005159517354e-05,
+      "loss": 1.047,
+      "step": 5021
+    },
+    {
+      "epoch": 0.4487233899970961,
+      "grad_norm": 0.4717482328414917,
+      "learning_rate": 6.064786682523928e-05,
+      "loss": 0.9254,
+      "step": 5022
+    },
+    {
+      "epoch": 0.44881274152835793,
+      "grad_norm": 0.40896281599998474,
+      "learning_rate": 6.063372759908885e-05,
+      "loss": 0.9374,
+      "step": 5023
+    },
+    {
+      "epoch": 0.4489020930596198,
+      "grad_norm": 0.43339207768440247,
+      "learning_rate": 6.06195874822504e-05,
+      "loss": 0.9764,
+      "step": 5024
+    },
+    {
+      "epoch": 0.4489914445908817,
+      "grad_norm": 0.4062994718551636,
+      "learning_rate": 6.06054464759083e-05,
+      "loss": 0.9822,
+      "step": 5025
+    },
+    {
+      "epoch": 0.4490807961221435,
+      "grad_norm": 0.4177139401435852,
+      "learning_rate": 6.0591304581247e-05,
+      "loss": 1.0045,
+      "step": 5026
+    },
+    {
+      "epoch": 0.4491701476534054,
+      "grad_norm": 0.46026477217674255,
+      "learning_rate": 6.0577161799451054e-05,
+      "loss": 0.9932,
+      "step": 5027
+    },
+    {
+      "epoch": 0.4492594991846673,
+      "grad_norm": 0.3944912552833557,
+      "learning_rate": 6.0563018131705063e-05,
+      "loss": 1.0112,
+      "step": 5028
+    },
+    {
+      "epoch": 0.44934885071592917,
+      "grad_norm": 0.49842584133148193,
+      "learning_rate": 6.054887357919371e-05,
+      "loss": 1.0421,
+      "step": 5029
+    },
+    {
+      "epoch": 0.449438202247191,
+      "grad_norm": 0.4486824572086334,
+      "learning_rate": 6.053472814310175e-05,
+      "loss": 0.9671,
+      "step": 5030
+    },
+    {
+      "epoch": 0.4495275537784529,
+      "grad_norm": 0.41824930906295776,
+      "learning_rate": 6.052058182461401e-05,
+      "loss": 0.9581,
+      "step": 5031
+    },
+    {
+      "epoch": 0.44961690530971476,
+      "grad_norm": 0.500913143157959,
+      "learning_rate": 6.0506434624915396e-05,
+      "loss": 0.9364,
+      "step": 5032
+    },
+    {
+      "epoch": 0.4497062568409766,
+      "grad_norm": 0.4475401043891907,
+      "learning_rate": 6.049228654519091e-05,
+      "loss": 0.9434,
+      "step": 5033
+    },
+    {
+      "epoch": 0.4497956083722385,
+      "grad_norm": 0.47360408306121826,
+      "learning_rate": 6.0478137586625584e-05,
+      "loss": 0.959,
+      "step": 5034
+    },
+    {
+      "epoch": 0.44988495990350036,
+      "grad_norm": 0.4704830050468445,
+      "learning_rate": 6.046398775040454e-05,
+      "loss": 0.9246,
+      "step": 5035
+    },
+    {
+      "epoch": 0.4499743114347622,
+      "grad_norm": 0.41631588339805603,
+      "learning_rate": 6.0449837037712976e-05,
+      "loss": 1.0194,
+      "step": 5036
+    },
+    {
+      "epoch": 0.45006366296602407,
+      "grad_norm": 0.47200682759284973,
+      "learning_rate": 6.043568544973618e-05,
+      "loss": 0.9078,
+      "step": 5037
+    },
+    {
+      "epoch": 0.45015301449728595,
+      "grad_norm": 0.4732605516910553,
+      "learning_rate": 6.0421532987659504e-05,
+      "loss": 1.0336,
+      "step": 5038
+    },
+    {
+      "epoch": 0.45024236602854784,
+      "grad_norm": 0.4652157425880432,
+      "learning_rate": 6.040737965266834e-05,
+      "loss": 0.9257,
+      "step": 5039
+    },
+    {
+      "epoch": 0.45033171755980966,
+      "grad_norm": 0.5224878191947937,
+      "learning_rate": 6.039322544594819e-05,
+      "loss": 0.9199,
+      "step": 5040
+    },
+    {
+      "epoch": 0.45042106909107155,
+      "grad_norm": 0.4996034801006317,
+      "learning_rate": 6.037907036868464e-05,
+      "loss": 1.0375,
+      "step": 5041
+    },
+    {
+      "epoch": 0.45051042062233343,
+      "grad_norm": 0.4124327600002289,
+      "learning_rate": 6.0364914422063304e-05,
+      "loss": 0.9245,
+      "step": 5042
+    },
+    {
+      "epoch": 0.45059977215359526,
+      "grad_norm": 0.441354900598526,
+      "learning_rate": 6.0350757607269904e-05,
+      "loss": 0.9574,
+      "step": 5043
+    },
+    {
+      "epoch": 0.45068912368485714,
+      "grad_norm": 0.4461125433444977,
+      "learning_rate": 6.033659992549023e-05,
+      "loss": 0.8827,
+      "step": 5044
+    },
+    {
+      "epoch": 0.450778475216119,
+      "grad_norm": 0.3978963792324066,
+      "learning_rate": 6.0322441377910135e-05,
+      "loss": 0.9883,
+      "step": 5045
+    },
+    {
+      "epoch": 0.4508678267473809,
+      "grad_norm": 0.4818345010280609,
+      "learning_rate": 6.030828196571553e-05,
+      "loss": 0.9646,
+      "step": 5046
+    },
+    {
+      "epoch": 0.45095717827864273,
+      "grad_norm": 0.5546659231185913,
+      "learning_rate": 6.029412169009243e-05,
+      "loss": 0.9554,
+      "step": 5047
+    },
+    {
+      "epoch": 0.4510465298099046,
+      "grad_norm": 0.4250444173812866,
+      "learning_rate": 6.0279960552226934e-05,
+      "loss": 0.9748,
+      "step": 5048
+    },
+    {
+      "epoch": 0.4511358813411665,
+      "grad_norm": 0.44978564977645874,
+      "learning_rate": 6.026579855330515e-05,
+      "loss": 0.9755,
+      "step": 5049
+    },
+    {
+      "epoch": 0.45122523287242833,
+      "grad_norm": 0.4644257426261902,
+      "learning_rate": 6.025163569451331e-05,
+      "loss": 1.003,
+      "step": 5050
+    },
+    {
+      "epoch": 0.4513145844036902,
+      "grad_norm": 0.41228753328323364,
+      "learning_rate": 6.0237471977037706e-05,
+      "loss": 1.0315,
+      "step": 5051
+    },
+    {
+      "epoch": 0.4514039359349521,
+      "grad_norm": 0.5227345824241638,
+      "learning_rate": 6.02233074020647e-05,
+      "loss": 0.8891,
+      "step": 5052
+    },
+    {
+      "epoch": 0.451493287466214,
+      "grad_norm": 0.46746036410331726,
+      "learning_rate": 6.020914197078074e-05,
+      "loss": 0.9664,
+      "step": 5053
+    },
+    {
+      "epoch": 0.4515826389974758,
+      "grad_norm": 0.3485402762889862,
+      "learning_rate": 6.01949756843723e-05,
+      "loss": 1.0188,
+      "step": 5054
+    },
+    {
+      "epoch": 0.4516719905287377,
+      "grad_norm": 0.4057181179523468,
+      "learning_rate": 6.018080854402599e-05,
+      "loss": 0.9448,
+      "step": 5055
+    },
+    {
+      "epoch": 0.45176134205999957,
+      "grad_norm": 0.39723464846611023,
+      "learning_rate": 6.0166640550928434e-05,
+      "loss": 0.97,
+      "step": 5056
+    },
+    {
+      "epoch": 0.4518506935912614,
+      "grad_norm": 0.3836243152618408,
+      "learning_rate": 6.015247170626637e-05,
+      "loss": 0.9494,
+      "step": 5057
+    },
+    {
+      "epoch": 0.4519400451225233,
+      "grad_norm": 0.4795580208301544,
+      "learning_rate": 6.013830201122659e-05,
+      "loss": 0.9054,
+      "step": 5058
+    },
+    {
+      "epoch": 0.45202939665378516,
+      "grad_norm": 0.4811798930168152,
+      "learning_rate": 6.0124131466995936e-05,
+      "loss": 0.9832,
+      "step": 5059
+    },
+    {
+      "epoch": 0.45211874818504705,
+      "grad_norm": 0.41265878081321716,
+      "learning_rate": 6.0109960074761374e-05,
+      "loss": 0.9763,
+      "step": 5060
+    },
+    {
+      "epoch": 0.4522080997163089,
+      "grad_norm": 0.4164207875728607,
+      "learning_rate": 6.009578783570987e-05,
+      "loss": 0.9297,
+      "step": 5061
+    },
+    {
+      "epoch": 0.45229745124757076,
+      "grad_norm": 0.392967164516449,
+      "learning_rate": 6.008161475102853e-05,
+      "loss": 0.953,
+      "step": 5062
+    },
+    {
+      "epoch": 0.45238680277883264,
+      "grad_norm": 0.42946505546569824,
+      "learning_rate": 6.00674408219045e-05,
+      "loss": 0.979,
+      "step": 5063
+    },
+    {
+      "epoch": 0.45247615431009447,
+      "grad_norm": 0.47703489661216736,
+      "learning_rate": 6.005326604952498e-05,
+      "loss": 0.9656,
+      "step": 5064
+    },
+    {
+      "epoch": 0.45256550584135635,
+      "grad_norm": 0.4116286635398865,
+      "learning_rate": 6.003909043507727e-05,
+      "loss": 0.943,
+      "step": 5065
+    },
+    {
+      "epoch": 0.45265485737261824,
+      "grad_norm": 0.4779094159603119,
+      "learning_rate": 6.002491397974872e-05,
+      "loss": 0.9066,
+      "step": 5066
+    },
+    {
+      "epoch": 0.45274420890388006,
+      "grad_norm": 0.4313408136367798,
+      "learning_rate": 6.0010736684726774e-05,
+      "loss": 1.01,
+      "step": 5067
+    },
+    {
+      "epoch": 0.45283356043514195,
+      "grad_norm": 0.4300564229488373,
+      "learning_rate": 5.999655855119893e-05,
+      "loss": 0.9609,
+      "step": 5068
+    },
+    {
+      "epoch": 0.45292291196640383,
+      "grad_norm": 0.4585869610309601,
+      "learning_rate": 5.998237958035274e-05,
+      "loss": 0.9914,
+      "step": 5069
+    },
+    {
+      "epoch": 0.4530122634976657,
+      "grad_norm": 0.5256367325782776,
+      "learning_rate": 5.996819977337587e-05,
+      "loss": 0.9232,
+      "step": 5070
+    },
+    {
+      "epoch": 0.45310161502892754,
+      "grad_norm": 0.38605356216430664,
+      "learning_rate": 5.9954019131456e-05,
+      "loss": 1.0481,
+      "step": 5071
+    },
+    {
+      "epoch": 0.4531909665601894,
+      "grad_norm": 0.41398924589157104,
+      "learning_rate": 5.993983765578093e-05,
+      "loss": 1.0571,
+      "step": 5072
+    },
+    {
+      "epoch": 0.4532803180914513,
+      "grad_norm": 0.45383667945861816,
+      "learning_rate": 5.9925655347538504e-05,
+      "loss": 0.9514,
+      "step": 5073
+    },
+    {
+      "epoch": 0.45336966962271313,
+      "grad_norm": 0.5158712863922119,
+      "learning_rate": 5.991147220791665e-05,
+      "loss": 0.8985,
+      "step": 5074
+    },
+    {
+      "epoch": 0.453459021153975,
+      "grad_norm": 0.4722534716129303,
+      "learning_rate": 5.989728823810335e-05,
+      "loss": 0.9492,
+      "step": 5075
+    },
+    {
+      "epoch": 0.4535483726852369,
+      "grad_norm": 0.517153799533844,
+      "learning_rate": 5.9883103439286646e-05,
+      "loss": 0.9614,
+      "step": 5076
+    },
+    {
+      "epoch": 0.4536377242164988,
+      "grad_norm": 0.4288327097892761,
+      "learning_rate": 5.986891781265471e-05,
+      "loss": 0.9977,
+      "step": 5077
+    },
+    {
+      "epoch": 0.4537270757477606,
+      "grad_norm": 0.3851867616176605,
+      "learning_rate": 5.98547313593957e-05,
+      "loss": 0.9741,
+      "step": 5078
+    },
+    {
+      "epoch": 0.4538164272790225,
+      "grad_norm": 0.41952255368232727,
+      "learning_rate": 5.9840544080697904e-05,
+      "loss": 0.901,
+      "step": 5079
+    },
+    {
+      "epoch": 0.4539057788102844,
+      "grad_norm": 0.4248369634151459,
+      "learning_rate": 5.9826355977749624e-05,
+      "loss": 1.004,
+      "step": 5080
+    },
+    {
+      "epoch": 0.4539951303415462,
+      "grad_norm": 0.3857657015323639,
+      "learning_rate": 5.98121670517393e-05,
+      "loss": 1.0764,
+      "step": 5081
+    },
+    {
+      "epoch": 0.4540844818728081,
+      "grad_norm": 0.4897959232330322,
+      "learning_rate": 5.97979773038554e-05,
+      "loss": 0.9433,
+      "step": 5082
+    },
+    {
+      "epoch": 0.45417383340406997,
+      "grad_norm": 0.4853098392486572,
+      "learning_rate": 5.978378673528645e-05,
+      "loss": 1.0143,
+      "step": 5083
+    },
+    {
+      "epoch": 0.45426318493533185,
+      "grad_norm": 0.3813380300998688,
+      "learning_rate": 5.9769595347221096e-05,
+      "loss": 0.991,
+      "step": 5084
+    },
+    {
+      "epoch": 0.4543525364665937,
+      "grad_norm": 0.3379877507686615,
+      "learning_rate": 5.9755403140847974e-05,
+      "loss": 1.0139,
+      "step": 5085
+    },
+    {
+      "epoch": 0.45444188799785556,
+      "grad_norm": 0.41478732228279114,
+      "learning_rate": 5.974121011735586e-05,
+      "loss": 0.9286,
+      "step": 5086
+    },
+    {
+      "epoch": 0.45453123952911745,
+      "grad_norm": 0.4362654685974121,
+      "learning_rate": 5.972701627793357e-05,
+      "loss": 0.9713,
+      "step": 5087
+    },
+    {
+      "epoch": 0.4546205910603793,
+      "grad_norm": 0.44205188751220703,
+      "learning_rate": 5.9712821623769976e-05,
+      "loss": 0.9442,
+      "step": 5088
+    },
+    {
+      "epoch": 0.45470994259164116,
+      "grad_norm": 0.4520505666732788,
+      "learning_rate": 5.969862615605405e-05,
+      "loss": 0.9221,
+      "step": 5089
+    },
+    {
+      "epoch": 0.45479929412290304,
+      "grad_norm": 0.3804233968257904,
+      "learning_rate": 5.96844298759748e-05,
+      "loss": 0.9983,
+      "step": 5090
+    },
+    {
+      "epoch": 0.4548886456541649,
+      "grad_norm": 0.45884522795677185,
+      "learning_rate": 5.967023278472131e-05,
+      "loss": 0.9408,
+      "step": 5091
+    },
+    {
+      "epoch": 0.45497799718542675,
+      "grad_norm": 0.5373643040657043,
+      "learning_rate": 5.965603488348276e-05,
+      "loss": 0.9206,
+      "step": 5092
+    },
+    {
+      "epoch": 0.45506734871668864,
+      "grad_norm": 0.4332393705844879,
+      "learning_rate": 5.964183617344836e-05,
+      "loss": 0.9796,
+      "step": 5093
+    },
+    {
+      "epoch": 0.4551567002479505,
+      "grad_norm": 0.5239064693450928,
+      "learning_rate": 5.962763665580741e-05,
+      "loss": 0.9123,
+      "step": 5094
+    },
+    {
+      "epoch": 0.45524605177921235,
+      "grad_norm": 0.41141557693481445,
+      "learning_rate": 5.961343633174926e-05,
+      "loss": 0.9346,
+      "step": 5095
+    },
+    {
+      "epoch": 0.45533540331047423,
+      "grad_norm": 0.43024352192878723,
+      "learning_rate": 5.9599235202463344e-05,
+      "loss": 0.9898,
+      "step": 5096
+    },
+    {
+      "epoch": 0.4554247548417361,
+      "grad_norm": 0.4978499412536621,
+      "learning_rate": 5.958503326913916e-05,
+      "loss": 0.9262,
+      "step": 5097
+    },
+    {
+      "epoch": 0.45551410637299794,
+      "grad_norm": 0.37708181142807007,
+      "learning_rate": 5.957083053296626e-05,
+      "loss": 1.0786,
+      "step": 5098
+    },
+    {
+      "epoch": 0.4556034579042598,
+      "grad_norm": 0.42339766025543213,
+      "learning_rate": 5.955662699513431e-05,
+      "loss": 0.9934,
+      "step": 5099
+    },
+    {
+      "epoch": 0.4556928094355217,
+      "grad_norm": 0.515396773815155,
+      "learning_rate": 5.954242265683296e-05,
+      "loss": 0.9361,
+      "step": 5100
+    },
+    {
+      "epoch": 0.4557821609667836,
+      "grad_norm": 0.502129316329956,
+      "learning_rate": 5.9528217519252003e-05,
+      "loss": 0.9834,
+      "step": 5101
+    },
+    {
+      "epoch": 0.4558715124980454,
+      "grad_norm": 0.44378775358200073,
+      "learning_rate": 5.9514011583581275e-05,
+      "loss": 0.9662,
+      "step": 5102
+    },
+    {
+      "epoch": 0.4559608640293073,
+      "grad_norm": 0.4403262436389923,
+      "learning_rate": 5.9499804851010655e-05,
+      "loss": 0.986,
+      "step": 5103
+    },
+    {
+      "epoch": 0.4560502155605692,
+      "grad_norm": 0.38599881529808044,
+      "learning_rate": 5.948559732273013e-05,
+      "loss": 1.0328,
+      "step": 5104
+    },
+    {
+      "epoch": 0.456139567091831,
+      "grad_norm": 0.5018365979194641,
+      "learning_rate": 5.9471388999929714e-05,
+      "loss": 0.9439,
+      "step": 5105
+    },
+    {
+      "epoch": 0.4562289186230929,
+      "grad_norm": 0.4095374047756195,
+      "learning_rate": 5.9457179883799496e-05,
+      "loss": 0.9718,
+      "step": 5106
+    },
+    {
+      "epoch": 0.4563182701543548,
+      "grad_norm": 0.49552807211875916,
+      "learning_rate": 5.944296997552967e-05,
+      "loss": 0.9708,
+      "step": 5107
+    },
+    {
+      "epoch": 0.45640762168561666,
+      "grad_norm": 0.45015573501586914,
+      "learning_rate": 5.942875927631045e-05,
+      "loss": 0.9136,
+      "step": 5108
+    },
+    {
+      "epoch": 0.4564969732168785,
+      "grad_norm": 0.4981837868690491,
+      "learning_rate": 5.9414547787332155e-05,
+      "loss": 0.8726,
+      "step": 5109
+    },
+    {
+      "epoch": 0.45658632474814037,
+      "grad_norm": 0.38596221804618835,
+      "learning_rate": 5.940033550978511e-05,
+      "loss": 1.0169,
+      "step": 5110
+    },
+    {
+      "epoch": 0.45667567627940225,
+      "grad_norm": 0.4662768840789795,
+      "learning_rate": 5.938612244485977e-05,
+      "loss": 0.8954,
+      "step": 5111
+    },
+    {
+      "epoch": 0.4567650278106641,
+      "grad_norm": 0.5224105715751648,
+      "learning_rate": 5.937190859374664e-05,
+      "loss": 0.9889,
+      "step": 5112
+    },
+    {
+      "epoch": 0.45685437934192596,
+      "grad_norm": 0.5646452307701111,
+      "learning_rate": 5.935769395763625e-05,
+      "loss": 0.8719,
+      "step": 5113
+    },
+    {
+      "epoch": 0.45694373087318785,
+      "grad_norm": 0.4161957800388336,
+      "learning_rate": 5.934347853771927e-05,
+      "loss": 0.9331,
+      "step": 5114
+    },
+    {
+      "epoch": 0.45703308240444973,
+      "grad_norm": 0.394827276468277,
+      "learning_rate": 5.9329262335186344e-05,
+      "loss": 0.954,
+      "step": 5115
+    },
+    {
+      "epoch": 0.45712243393571156,
+      "grad_norm": 0.5495721697807312,
+      "learning_rate": 5.931504535122825e-05,
+      "loss": 1.0314,
+      "step": 5116
+    },
+    {
+      "epoch": 0.45721178546697344,
+      "grad_norm": 0.48053717613220215,
+      "learning_rate": 5.9300827587035825e-05,
+      "loss": 0.9705,
+      "step": 5117
+    },
+    {
+      "epoch": 0.4573011369982353,
+      "grad_norm": 0.48041045665740967,
+      "learning_rate": 5.9286609043799945e-05,
+      "loss": 0.941,
+      "step": 5118
+    },
+    {
+      "epoch": 0.45739048852949715,
+      "grad_norm": 0.46086958050727844,
+      "learning_rate": 5.9272389722711586e-05,
+      "loss": 0.9866,
+      "step": 5119
+    },
+    {
+      "epoch": 0.45747984006075904,
+      "grad_norm": 0.4236505925655365,
+      "learning_rate": 5.925816962496175e-05,
+      "loss": 0.9627,
+      "step": 5120
+    },
+    {
+      "epoch": 0.4575691915920209,
+      "grad_norm": 0.41899070143699646,
+      "learning_rate": 5.9243948751741505e-05,
+      "loss": 0.9541,
+      "step": 5121
+    },
+    {
+      "epoch": 0.4576585431232828,
+      "grad_norm": 0.37787696719169617,
+      "learning_rate": 5.922972710424203e-05,
+      "loss": 0.9989,
+      "step": 5122
+    },
+    {
+      "epoch": 0.45774789465454463,
+      "grad_norm": 0.40664857625961304,
+      "learning_rate": 5.921550468365452e-05,
+      "loss": 1.0442,
+      "step": 5123
+    },
+    {
+      "epoch": 0.4578372461858065,
+      "grad_norm": 0.4115360975265503,
+      "learning_rate": 5.920128149117028e-05,
+      "loss": 0.9699,
+      "step": 5124
+    },
+    {
+      "epoch": 0.4579265977170684,
+      "grad_norm": 0.44247210025787354,
+      "learning_rate": 5.918705752798062e-05,
+      "loss": 0.9151,
+      "step": 5125
+    },
+    {
+      "epoch": 0.4580159492483302,
+      "grad_norm": 0.4786369800567627,
+      "learning_rate": 5.917283279527697e-05,
+      "loss": 0.9834,
+      "step": 5126
+    },
+    {
+      "epoch": 0.4581053007795921,
+      "grad_norm": 0.5580417513847351,
+      "learning_rate": 5.915860729425079e-05,
+      "loss": 0.7882,
+      "step": 5127
+    },
+    {
+      "epoch": 0.458194652310854,
+      "grad_norm": 0.4049915373325348,
+      "learning_rate": 5.914438102609364e-05,
+      "loss": 0.9563,
+      "step": 5128
+    },
+    {
+      "epoch": 0.4582840038421158,
+      "grad_norm": 0.4272023141384125,
+      "learning_rate": 5.91301539919971e-05,
+      "loss": 0.9359,
+      "step": 5129
+    },
+    {
+      "epoch": 0.4583733553733777,
+      "grad_norm": 0.4311048686504364,
+      "learning_rate": 5.911592619315286e-05,
+      "loss": 0.9599,
+      "step": 5130
+    },
+    {
+      "epoch": 0.4584627069046396,
+      "grad_norm": 0.4287171959877014,
+      "learning_rate": 5.910169763075261e-05,
+      "loss": 0.9668,
+      "step": 5131
+    },
+    {
+      "epoch": 0.45855205843590147,
+      "grad_norm": 0.4313339293003082,
+      "learning_rate": 5.908746830598817e-05,
+      "loss": 0.9181,
+      "step": 5132
+    },
+    {
+      "epoch": 0.4586414099671633,
+      "grad_norm": 0.4437636435031891,
+      "learning_rate": 5.9073238220051394e-05,
+      "loss": 1.0319,
+      "step": 5133
+    },
+    {
+      "epoch": 0.4587307614984252,
+      "grad_norm": 0.4116639196872711,
+      "learning_rate": 5.905900737413421e-05,
+      "loss": 0.9471,
+      "step": 5134
+    },
+    {
+      "epoch": 0.45882011302968706,
+      "grad_norm": 0.5594172477722168,
+      "learning_rate": 5.9044775769428594e-05,
+      "loss": 0.8384,
+      "step": 5135
+    },
+    {
+      "epoch": 0.4589094645609489,
+      "grad_norm": 0.4292183816432953,
+      "learning_rate": 5.903054340712659e-05,
+      "loss": 1.0041,
+      "step": 5136
+    },
+    {
+      "epoch": 0.45899881609221077,
+      "grad_norm": 0.5219025015830994,
+      "learning_rate": 5.9016310288420316e-05,
+      "loss": 0.8519,
+      "step": 5137
+    },
+    {
+      "epoch": 0.45908816762347265,
+      "grad_norm": 0.4652097821235657,
+      "learning_rate": 5.900207641450195e-05,
+      "loss": 0.9249,
+      "step": 5138
+    },
+    {
+      "epoch": 0.45917751915473454,
+      "grad_norm": 0.41842982172966003,
+      "learning_rate": 5.898784178656375e-05,
+      "loss": 1.0074,
+      "step": 5139
+    },
+    {
+      "epoch": 0.45926687068599636,
+      "grad_norm": 0.41296178102493286,
+      "learning_rate": 5.897360640579798e-05,
+      "loss": 0.9087,
+      "step": 5140
+    },
+    {
+      "epoch": 0.45935622221725825,
+      "grad_norm": 0.49167054891586304,
+      "learning_rate": 5.895937027339701e-05,
+      "loss": 0.9117,
+      "step": 5141
+    },
+    {
+      "epoch": 0.45944557374852013,
+      "grad_norm": 0.3634302020072937,
+      "learning_rate": 5.894513339055329e-05,
+      "loss": 1.0143,
+      "step": 5142
+    },
+    {
+      "epoch": 0.45953492527978196,
+      "grad_norm": 0.522591233253479,
+      "learning_rate": 5.89308957584593e-05,
+      "loss": 0.9263,
+      "step": 5143
+    },
+    {
+      "epoch": 0.45962427681104384,
+      "grad_norm": 0.49898669123649597,
+      "learning_rate": 5.89166573783076e-05,
+      "loss": 0.9532,
+      "step": 5144
+    },
+    {
+      "epoch": 0.4597136283423057,
+      "grad_norm": 0.44929975271224976,
+      "learning_rate": 5.890241825129079e-05,
+      "loss": 0.9635,
+      "step": 5145
+    },
+    {
+      "epoch": 0.4598029798735676,
+      "grad_norm": 0.4347250461578369,
+      "learning_rate": 5.8888178378601565e-05,
+      "loss": 0.9352,
+      "step": 5146
+    },
+    {
+      "epoch": 0.45989233140482944,
+      "grad_norm": 0.44016188383102417,
+      "learning_rate": 5.887393776143267e-05,
+      "loss": 0.9607,
+      "step": 5147
+    },
+    {
+      "epoch": 0.4599816829360913,
+      "grad_norm": 0.5058707594871521,
+      "learning_rate": 5.8859696400976884e-05,
+      "loss": 0.965,
+      "step": 5148
+    },
+    {
+      "epoch": 0.4600710344673532,
+      "grad_norm": 0.40825730562210083,
+      "learning_rate": 5.8845454298427094e-05,
+      "loss": 1.0115,
+      "step": 5149
+    },
+    {
+      "epoch": 0.46016038599861503,
+      "grad_norm": 0.4351537227630615,
+      "learning_rate": 5.883121145497622e-05,
+      "loss": 0.9622,
+      "step": 5150
+    },
+    {
+      "epoch": 0.4602497375298769,
+      "grad_norm": 0.4099435806274414,
+      "learning_rate": 5.881696787181724e-05,
+      "loss": 0.9724,
+      "step": 5151
+    },
+    {
+      "epoch": 0.4603390890611388,
+      "grad_norm": 0.44913604855537415,
+      "learning_rate": 5.8802723550143225e-05,
+      "loss": 0.9009,
+      "step": 5152
+    },
+    {
+      "epoch": 0.4604284405924007,
+      "grad_norm": 0.417484313249588,
+      "learning_rate": 5.878847849114728e-05,
+      "loss": 0.9643,
+      "step": 5153
+    },
+    {
+      "epoch": 0.4605177921236625,
+      "grad_norm": 0.43109041452407837,
+      "learning_rate": 5.8774232696022604e-05,
+      "loss": 0.9845,
+      "step": 5154
+    },
+    {
+      "epoch": 0.4606071436549244,
+      "grad_norm": 0.432075172662735,
+      "learning_rate": 5.8759986165962386e-05,
+      "loss": 1.0115,
+      "step": 5155
+    },
+    {
+      "epoch": 0.4606964951861863,
+      "grad_norm": 0.49210742115974426,
+      "learning_rate": 5.8745738902159965e-05,
+      "loss": 0.8991,
+      "step": 5156
+    },
+    {
+      "epoch": 0.4607858467174481,
+      "grad_norm": 0.41765424609184265,
+      "learning_rate": 5.873149090580868e-05,
+      "loss": 0.9901,
+      "step": 5157
+    },
+    {
+      "epoch": 0.46087519824871,
+      "grad_norm": 0.5431698560714722,
+      "learning_rate": 5.8717242178101975e-05,
+      "loss": 0.9571,
+      "step": 5158
+    },
+    {
+      "epoch": 0.46096454977997187,
+      "grad_norm": 0.5281333923339844,
+      "learning_rate": 5.8702992720233296e-05,
+      "loss": 0.9096,
+      "step": 5159
+    },
+    {
+      "epoch": 0.4610539013112337,
+      "grad_norm": 0.4179060757160187,
+      "learning_rate": 5.868874253339624e-05,
+      "loss": 0.937,
+      "step": 5160
+    },
+    {
+      "epoch": 0.4611432528424956,
+      "grad_norm": 0.4784322679042816,
+      "learning_rate": 5.867449161878435e-05,
+      "loss": 0.989,
+      "step": 5161
+    },
+    {
+      "epoch": 0.46123260437375746,
+      "grad_norm": 0.431211918592453,
+      "learning_rate": 5.8660239977591344e-05,
+      "loss": 1.0558,
+      "step": 5162
+    },
+    {
+      "epoch": 0.46132195590501934,
+      "grad_norm": 0.4160225987434387,
+      "learning_rate": 5.8645987611010935e-05,
+      "loss": 0.9614,
+      "step": 5163
+    },
+    {
+      "epoch": 0.46141130743628117,
+      "grad_norm": 0.41637277603149414,
+      "learning_rate": 5.863173452023688e-05,
+      "loss": 0.9515,
+      "step": 5164
+    },
+    {
+      "epoch": 0.46150065896754305,
+      "grad_norm": 0.35038280487060547,
+      "learning_rate": 5.861748070646307e-05,
+      "loss": 0.9557,
+      "step": 5165
+    },
+    {
+      "epoch": 0.46159001049880494,
+      "grad_norm": 0.49989181756973267,
+      "learning_rate": 5.860322617088339e-05,
+      "loss": 0.9531,
+      "step": 5166
+    },
+    {
+      "epoch": 0.46167936203006676,
+      "grad_norm": 0.3543355166912079,
+      "learning_rate": 5.858897091469181e-05,
+      "loss": 0.971,
+      "step": 5167
+    },
+    {
+      "epoch": 0.46176871356132865,
+      "grad_norm": 0.4254325032234192,
+      "learning_rate": 5.857471493908237e-05,
+      "loss": 1.009,
+      "step": 5168
+    },
+    {
+      "epoch": 0.46185806509259053,
+      "grad_norm": 0.4708648920059204,
+      "learning_rate": 5.856045824524914e-05,
+      "loss": 0.9253,
+      "step": 5169
+    },
+    {
+      "epoch": 0.4619474166238524,
+      "grad_norm": 0.48328697681427,
+      "learning_rate": 5.85462008343863e-05,
+      "loss": 0.8906,
+      "step": 5170
+    },
+    {
+      "epoch": 0.46203676815511424,
+      "grad_norm": 0.42431074380874634,
+      "learning_rate": 5.8531942707688034e-05,
+      "loss": 0.9893,
+      "step": 5171
+    },
+    {
+      "epoch": 0.4621261196863761,
+      "grad_norm": 0.37147825956344604,
+      "learning_rate": 5.851768386634863e-05,
+      "loss": 1.0255,
+      "step": 5172
+    },
+    {
+      "epoch": 0.462215471217638,
+      "grad_norm": 0.47785985469818115,
+      "learning_rate": 5.8503424311562406e-05,
+      "loss": 0.9222,
+      "step": 5173
+    },
+    {
+      "epoch": 0.46230482274889984,
+      "grad_norm": 0.4230659604072571,
+      "learning_rate": 5.848916404452375e-05,
+      "loss": 1.0169,
+      "step": 5174
+    },
+    {
+      "epoch": 0.4623941742801617,
+      "grad_norm": 0.45284929871559143,
+      "learning_rate": 5.8474903066427136e-05,
+      "loss": 1.0053,
+      "step": 5175
+    },
+    {
+      "epoch": 0.4624835258114236,
+      "grad_norm": 0.4324350357055664,
+      "learning_rate": 5.846064137846704e-05,
+      "loss": 0.9689,
+      "step": 5176
+    },
+    {
+      "epoch": 0.4625728773426855,
+      "grad_norm": 0.4609506130218506,
+      "learning_rate": 5.844637898183805e-05,
+      "loss": 0.9361,
+      "step": 5177
+    },
+    {
+      "epoch": 0.4626622288739473,
+      "grad_norm": 0.45454010367393494,
+      "learning_rate": 5.8432115877734775e-05,
+      "loss": 0.8882,
+      "step": 5178
+    },
+    {
+      "epoch": 0.4627515804052092,
+      "grad_norm": 0.47145554423332214,
+      "learning_rate": 5.841785206735192e-05,
+      "loss": 0.9578,
+      "step": 5179
+    },
+    {
+      "epoch": 0.4628409319364711,
+      "grad_norm": 0.5458246469497681,
+      "learning_rate": 5.8403587551884244e-05,
+      "loss": 0.9361,
+      "step": 5180
+    },
+    {
+      "epoch": 0.4629302834677329,
+      "grad_norm": 0.46121159195899963,
+      "learning_rate": 5.838932233252652e-05,
+      "loss": 0.8609,
+      "step": 5181
+    },
+    {
+      "epoch": 0.4630196349989948,
+      "grad_norm": 0.40931928157806396,
+      "learning_rate": 5.837505641047364e-05,
+      "loss": 1.0475,
+      "step": 5182
+    },
+    {
+      "epoch": 0.46310898653025667,
+      "grad_norm": 0.3603355288505554,
+      "learning_rate": 5.83607897869205e-05,
+      "loss": 0.9752,
+      "step": 5183
+    },
+    {
+      "epoch": 0.46319833806151856,
+      "grad_norm": 0.40766477584838867,
+      "learning_rate": 5.834652246306209e-05,
+      "loss": 0.9826,
+      "step": 5184
+    },
+    {
+      "epoch": 0.4632876895927804,
+      "grad_norm": 0.43302759528160095,
+      "learning_rate": 5.833225444009348e-05,
+      "loss": 0.954,
+      "step": 5185
+    },
+    {
+      "epoch": 0.46337704112404227,
+      "grad_norm": 0.518700122833252,
+      "learning_rate": 5.831798571920973e-05,
+      "loss": 1.0633,
+      "step": 5186
+    },
+    {
+      "epoch": 0.46346639265530415,
+      "grad_norm": 0.4842970371246338,
+      "learning_rate": 5.8303716301606e-05,
+      "loss": 0.9383,
+      "step": 5187
+    },
+    {
+      "epoch": 0.463555744186566,
+      "grad_norm": 0.42370370030403137,
+      "learning_rate": 5.828944618847753e-05,
+      "loss": 0.9952,
+      "step": 5188
+    },
+    {
+      "epoch": 0.46364509571782786,
+      "grad_norm": 0.47464433312416077,
+      "learning_rate": 5.827517538101959e-05,
+      "loss": 0.9543,
+      "step": 5189
+    },
+    {
+      "epoch": 0.46373444724908974,
+      "grad_norm": 0.48624545335769653,
+      "learning_rate": 5.826090388042751e-05,
+      "loss": 0.9396,
+      "step": 5190
+    },
+    {
+      "epoch": 0.4638237987803516,
+      "grad_norm": 0.45458412170410156,
+      "learning_rate": 5.8246631687896656e-05,
+      "loss": 0.9381,
+      "step": 5191
+    },
+    {
+      "epoch": 0.46391315031161345,
+      "grad_norm": 0.43564853072166443,
+      "learning_rate": 5.823235880462251e-05,
+      "loss": 0.9548,
+      "step": 5192
+    },
+    {
+      "epoch": 0.46400250184287534,
+      "grad_norm": 0.46724554896354675,
+      "learning_rate": 5.8218085231800544e-05,
+      "loss": 0.9358,
+      "step": 5193
+    },
+    {
+      "epoch": 0.4640918533741372,
+      "grad_norm": 0.4100496172904968,
+      "learning_rate": 5.820381097062636e-05,
+      "loss": 0.9394,
+      "step": 5194
+    },
+    {
+      "epoch": 0.46418120490539905,
+      "grad_norm": 0.5092671513557434,
+      "learning_rate": 5.818953602229557e-05,
+      "loss": 0.9381,
+      "step": 5195
+    },
+    {
+      "epoch": 0.46427055643666093,
+      "grad_norm": 0.410299688577652,
+      "learning_rate": 5.8175260388003813e-05,
+      "loss": 1.0375,
+      "step": 5196
+    },
+    {
+      "epoch": 0.4643599079679228,
+      "grad_norm": 0.4275777041912079,
+      "learning_rate": 5.816098406894688e-05,
+      "loss": 0.9305,
+      "step": 5197
+    },
+    {
+      "epoch": 0.46444925949918464,
+      "grad_norm": 0.4333481192588806,
+      "learning_rate": 5.814670706632054e-05,
+      "loss": 0.9786,
+      "step": 5198
+    },
+    {
+      "epoch": 0.4645386110304465,
+      "grad_norm": 0.5642595291137695,
+      "learning_rate": 5.813242938132064e-05,
+      "loss": 0.9754,
+      "step": 5199
+    },
+    {
+      "epoch": 0.4646279625617084,
+      "grad_norm": 0.5947056412696838,
+      "learning_rate": 5.8118151015143094e-05,
+      "loss": 0.9364,
+      "step": 5200
+    },
+    {
+      "epoch": 0.4647173140929703,
+      "grad_norm": 0.4831881523132324,
+      "learning_rate": 5.810387196898387e-05,
+      "loss": 0.9705,
+      "step": 5201
+    },
+    {
+      "epoch": 0.4648066656242321,
+      "grad_norm": 0.4384908676147461,
+      "learning_rate": 5.808959224403898e-05,
+      "loss": 0.9407,
+      "step": 5202
+    },
+    {
+      "epoch": 0.464896017155494,
+      "grad_norm": 0.4423673748970032,
+      "learning_rate": 5.807531184150452e-05,
+      "loss": 1.0036,
+      "step": 5203
+    },
+    {
+      "epoch": 0.4649853686867559,
+      "grad_norm": 0.4656173288822174,
+      "learning_rate": 5.806103076257661e-05,
+      "loss": 0.9284,
+      "step": 5204
+    },
+    {
+      "epoch": 0.4650747202180177,
+      "grad_norm": 0.4020898938179016,
+      "learning_rate": 5.804674900845145e-05,
+      "loss": 0.9451,
+      "step": 5205
+    },
+    {
+      "epoch": 0.4651640717492796,
+      "grad_norm": 0.3898525536060333,
+      "learning_rate": 5.803246658032528e-05,
+      "loss": 0.9848,
+      "step": 5206
+    },
+    {
+      "epoch": 0.4652534232805415,
+      "grad_norm": 0.41617026925086975,
+      "learning_rate": 5.801818347939443e-05,
+      "loss": 0.9434,
+      "step": 5207
+    },
+    {
+      "epoch": 0.46534277481180336,
+      "grad_norm": 0.3756409287452698,
+      "learning_rate": 5.8003899706855235e-05,
+      "loss": 1.0008,
+      "step": 5208
+    },
+    {
+      "epoch": 0.4654321263430652,
+      "grad_norm": 0.48038768768310547,
+      "learning_rate": 5.798961526390413e-05,
+      "loss": 0.9255,
+      "step": 5209
+    },
+    {
+      "epoch": 0.46552147787432707,
+      "grad_norm": 0.42943793535232544,
+      "learning_rate": 5.797533015173759e-05,
+      "loss": 1.0199,
+      "step": 5210
+    },
+    {
+      "epoch": 0.46561082940558896,
+      "grad_norm": 0.5280621647834778,
+      "learning_rate": 5.796104437155213e-05,
+      "loss": 0.9118,
+      "step": 5211
+    },
+    {
+      "epoch": 0.4657001809368508,
+      "grad_norm": 0.4121812582015991,
+      "learning_rate": 5.794675792454434e-05,
+      "loss": 0.9665,
+      "step": 5212
+    },
+    {
+      "epoch": 0.46578953246811267,
+      "grad_norm": 0.43791842460632324,
+      "learning_rate": 5.793247081191087e-05,
+      "loss": 0.9876,
+      "step": 5213
+    },
+    {
+      "epoch": 0.46587888399937455,
+      "grad_norm": 0.38042962551116943,
+      "learning_rate": 5.791818303484842e-05,
+      "loss": 0.916,
+      "step": 5214
+    },
+    {
+      "epoch": 0.46596823553063643,
+      "grad_norm": 0.4212323725223541,
+      "learning_rate": 5.790389459455374e-05,
+      "loss": 0.9963,
+      "step": 5215
+    },
+    {
+      "epoch": 0.46605758706189826,
+      "grad_norm": 0.4514927566051483,
+      "learning_rate": 5.788960549222364e-05,
+      "loss": 0.9769,
+      "step": 5216
+    },
+    {
+      "epoch": 0.46614693859316014,
+      "grad_norm": 0.4762234687805176,
+      "learning_rate": 5.787531572905498e-05,
+      "loss": 1.0555,
+      "step": 5217
+    },
+    {
+      "epoch": 0.466236290124422,
+      "grad_norm": 0.3893480598926544,
+      "learning_rate": 5.786102530624469e-05,
+      "loss": 0.9298,
+      "step": 5218
+    },
+    {
+      "epoch": 0.46632564165568385,
+      "grad_norm": 0.4439602494239807,
+      "learning_rate": 5.784673422498972e-05,
+      "loss": 0.9701,
+      "step": 5219
+    },
+    {
+      "epoch": 0.46641499318694574,
+      "grad_norm": 0.44786638021469116,
+      "learning_rate": 5.783244248648714e-05,
+      "loss": 0.9189,
+      "step": 5220
+    },
+    {
+      "epoch": 0.4665043447182076,
+      "grad_norm": 0.4381410479545593,
+      "learning_rate": 5.7818150091934e-05,
+      "loss": 1.0335,
+      "step": 5221
+    },
+    {
+      "epoch": 0.4665936962494695,
+      "grad_norm": 0.4737434685230255,
+      "learning_rate": 5.780385704252746e-05,
+      "loss": 0.9773,
+      "step": 5222
+    },
+    {
+      "epoch": 0.46668304778073133,
+      "grad_norm": 0.41768383979797363,
+      "learning_rate": 5.7789563339464695e-05,
+      "loss": 0.9492,
+      "step": 5223
+    },
+    {
+      "epoch": 0.4667723993119932,
+      "grad_norm": 0.5461154580116272,
+      "learning_rate": 5.777526898394298e-05,
+      "loss": 0.8902,
+      "step": 5224
+    },
+    {
+      "epoch": 0.4668617508432551,
+      "grad_norm": 0.46472278237342834,
+      "learning_rate": 5.7760973977159615e-05,
+      "loss": 0.9915,
+      "step": 5225
+    },
+    {
+      "epoch": 0.4669511023745169,
+      "grad_norm": 0.4949001669883728,
+      "learning_rate": 5.7746678320311954e-05,
+      "loss": 0.8738,
+      "step": 5226
+    },
+    {
+      "epoch": 0.4670404539057788,
+      "grad_norm": 0.5098089575767517,
+      "learning_rate": 5.77323820145974e-05,
+      "loss": 0.9527,
+      "step": 5227
+    },
+    {
+      "epoch": 0.4671298054370407,
+      "grad_norm": 0.4948689341545105,
+      "learning_rate": 5.7718085061213436e-05,
+      "loss": 0.9125,
+      "step": 5228
+    },
+    {
+      "epoch": 0.4672191569683025,
+      "grad_norm": 0.4423523247241974,
+      "learning_rate": 5.7703787461357575e-05,
+      "loss": 1.0395,
+      "step": 5229
+    },
+    {
+      "epoch": 0.4673085084995644,
+      "grad_norm": 0.4000280499458313,
+      "learning_rate": 5.7689489216227413e-05,
+      "loss": 0.9811,
+      "step": 5230
+    },
+    {
+      "epoch": 0.4673978600308263,
+      "grad_norm": 0.4417012333869934,
+      "learning_rate": 5.767519032702055e-05,
+      "loss": 0.9685,
+      "step": 5231
+    },
+    {
+      "epoch": 0.46748721156208817,
+      "grad_norm": 0.5120064616203308,
+      "learning_rate": 5.7660890794934676e-05,
+      "loss": 0.8708,
+      "step": 5232
+    },
+    {
+      "epoch": 0.46757656309335,
+      "grad_norm": 0.4820534288883209,
+      "learning_rate": 5.7646590621167564e-05,
+      "loss": 0.9201,
+      "step": 5233
+    },
+    {
+      "epoch": 0.4676659146246119,
+      "grad_norm": 0.46045050024986267,
+      "learning_rate": 5.763228980691696e-05,
+      "loss": 0.9647,
+      "step": 5234
+    },
+    {
+      "epoch": 0.46775526615587376,
+      "grad_norm": 0.448191374540329,
+      "learning_rate": 5.7617988353380746e-05,
+      "loss": 0.9384,
+      "step": 5235
+    },
+    {
+      "epoch": 0.4678446176871356,
+      "grad_norm": 0.5090450644493103,
+      "learning_rate": 5.76036862617568e-05,
+      "loss": 0.8526,
+      "step": 5236
+    },
+    {
+      "epoch": 0.46793396921839747,
+      "grad_norm": 0.4289693236351013,
+      "learning_rate": 5.758938353324308e-05,
+      "loss": 0.9773,
+      "step": 5237
+    },
+    {
+      "epoch": 0.46802332074965936,
+      "grad_norm": 0.522016704082489,
+      "learning_rate": 5.757508016903759e-05,
+      "loss": 0.867,
+      "step": 5238
+    },
+    {
+      "epoch": 0.46811267228092124,
+      "grad_norm": 0.48837369680404663,
+      "learning_rate": 5.756077617033838e-05,
+      "loss": 0.9976,
+      "step": 5239
+    },
+    {
+      "epoch": 0.46820202381218307,
+      "grad_norm": 0.41904178261756897,
+      "learning_rate": 5.75464715383436e-05,
+      "loss": 1.0333,
+      "step": 5240
+    },
+    {
+      "epoch": 0.46829137534344495,
+      "grad_norm": 0.4596676230430603,
+      "learning_rate": 5.7532166274251364e-05,
+      "loss": 0.9308,
+      "step": 5241
+    },
+    {
+      "epoch": 0.46838072687470683,
+      "grad_norm": 0.40436744689941406,
+      "learning_rate": 5.751786037925993e-05,
+      "loss": 1.0501,
+      "step": 5242
+    },
+    {
+      "epoch": 0.46847007840596866,
+      "grad_norm": 0.4699536859989166,
+      "learning_rate": 5.750355385456755e-05,
+      "loss": 0.9808,
+      "step": 5243
+    },
+    {
+      "epoch": 0.46855942993723054,
+      "grad_norm": 0.4330219030380249,
+      "learning_rate": 5.7489246701372545e-05,
+      "loss": 0.9724,
+      "step": 5244
+    },
+    {
+      "epoch": 0.4686487814684924,
+      "grad_norm": 0.421366810798645,
+      "learning_rate": 5.74749389208733e-05,
+      "loss": 1.0166,
+      "step": 5245
+    },
+    {
+      "epoch": 0.4687381329997543,
+      "grad_norm": 0.39331406354904175,
+      "learning_rate": 5.746063051426827e-05,
+      "loss": 0.9864,
+      "step": 5246
+    },
+    {
+      "epoch": 0.46882748453101614,
+      "grad_norm": 0.5093923807144165,
+      "learning_rate": 5.744632148275588e-05,
+      "loss": 0.9477,
+      "step": 5247
+    },
+    {
+      "epoch": 0.468916836062278,
+      "grad_norm": 0.480220228433609,
+      "learning_rate": 5.74320118275347e-05,
+      "loss": 0.934,
+      "step": 5248
+    },
+    {
+      "epoch": 0.4690061875935399,
+      "grad_norm": 0.4359627664089203,
+      "learning_rate": 5.741770154980331e-05,
+      "loss": 0.9851,
+      "step": 5249
+    },
+    {
+      "epoch": 0.46909553912480173,
+      "grad_norm": 0.4476899802684784,
+      "learning_rate": 5.7403390650760356e-05,
+      "loss": 0.8952,
+      "step": 5250
+    },
+    {
+      "epoch": 0.4691848906560636,
+      "grad_norm": 0.42261365056037903,
+      "learning_rate": 5.738907913160452e-05,
+      "loss": 0.9202,
+      "step": 5251
+    },
+    {
+      "epoch": 0.4692742421873255,
+      "grad_norm": 0.4985414743423462,
+      "learning_rate": 5.7374766993534545e-05,
+      "loss": 0.8766,
+      "step": 5252
+    },
+    {
+      "epoch": 0.4693635937185874,
+      "grad_norm": 0.5040563941001892,
+      "learning_rate": 5.736045423774923e-05,
+      "loss": 0.9752,
+      "step": 5253
+    },
+    {
+      "epoch": 0.4694529452498492,
+      "grad_norm": 0.4096861779689789,
+      "learning_rate": 5.7346140865447395e-05,
+      "loss": 1.0323,
+      "step": 5254
+    },
+    {
+      "epoch": 0.4695422967811111,
+      "grad_norm": 0.4268275499343872,
+      "learning_rate": 5.7331826877827965e-05,
+      "loss": 1.0152,
+      "step": 5255
+    },
+    {
+      "epoch": 0.469631648312373,
+      "grad_norm": 0.4520736336708069,
+      "learning_rate": 5.73175122760899e-05,
+      "loss": 0.9099,
+      "step": 5256
+    },
+    {
+      "epoch": 0.4697209998436348,
+      "grad_norm": 0.4920952618122101,
+      "learning_rate": 5.7303197061432165e-05,
+      "loss": 0.9575,
+      "step": 5257
+    },
+    {
+      "epoch": 0.4698103513748967,
+      "grad_norm": 0.44144368171691895,
+      "learning_rate": 5.728888123505384e-05,
+      "loss": 0.9656,
+      "step": 5258
+    },
+    {
+      "epoch": 0.46989970290615857,
+      "grad_norm": 0.39283469319343567,
+      "learning_rate": 5.727456479815401e-05,
+      "loss": 1.0424,
+      "step": 5259
+    },
+    {
+      "epoch": 0.4699890544374204,
+      "grad_norm": 0.4499633312225342,
+      "learning_rate": 5.726024775193184e-05,
+      "loss": 0.9702,
+      "step": 5260
+    },
+    {
+      "epoch": 0.4700784059686823,
+      "grad_norm": 0.48507824540138245,
+      "learning_rate": 5.724593009758653e-05,
+      "loss": 0.9632,
+      "step": 5261
+    },
+    {
+      "epoch": 0.47016775749994416,
+      "grad_norm": 0.4047746956348419,
+      "learning_rate": 5.723161183631734e-05,
+      "loss": 0.9593,
+      "step": 5262
+    },
+    {
+      "epoch": 0.47025710903120604,
+      "grad_norm": 0.41394850611686707,
+      "learning_rate": 5.7217292969323576e-05,
+      "loss": 0.9266,
+      "step": 5263
+    },
+    {
+      "epoch": 0.47034646056246787,
+      "grad_norm": 0.45513802766799927,
+      "learning_rate": 5.720297349780459e-05,
+      "loss": 1.0043,
+      "step": 5264
+    },
+    {
+      "epoch": 0.47043581209372975,
+      "grad_norm": 0.4691244959831238,
+      "learning_rate": 5.718865342295979e-05,
+      "loss": 0.9583,
+      "step": 5265
+    },
+    {
+      "epoch": 0.47052516362499164,
+      "grad_norm": 0.49911266565322876,
+      "learning_rate": 5.7174332745988666e-05,
+      "loss": 0.924,
+      "step": 5266
+    },
+    {
+      "epoch": 0.47061451515625347,
+      "grad_norm": 0.441734254360199,
+      "learning_rate": 5.716001146809068e-05,
+      "loss": 0.9989,
+      "step": 5267
+    },
+    {
+      "epoch": 0.47070386668751535,
+      "grad_norm": 0.40593159198760986,
+      "learning_rate": 5.714568959046542e-05,
+      "loss": 0.936,
+      "step": 5268
+    },
+    {
+      "epoch": 0.47079321821877723,
+      "grad_norm": 0.3953614830970764,
+      "learning_rate": 5.713136711431249e-05,
+      "loss": 0.9676,
+      "step": 5269
+    },
+    {
+      "epoch": 0.4708825697500391,
+      "grad_norm": 0.48885035514831543,
+      "learning_rate": 5.7117044040831544e-05,
+      "loss": 0.9002,
+      "step": 5270
+    },
+    {
+      "epoch": 0.47097192128130094,
+      "grad_norm": 0.4319499433040619,
+      "learning_rate": 5.7102720371222305e-05,
+      "loss": 1.0525,
+      "step": 5271
+    },
+    {
+      "epoch": 0.4710612728125628,
+      "grad_norm": 0.48925381898880005,
+      "learning_rate": 5.708839610668453e-05,
+      "loss": 1.023,
+      "step": 5272
+    },
+    {
+      "epoch": 0.4711506243438247,
+      "grad_norm": 0.39478790760040283,
+      "learning_rate": 5.7074071248418e-05,
+      "loss": 0.9983,
+      "step": 5273
+    },
+    {
+      "epoch": 0.47123997587508654,
+      "grad_norm": 0.5163027048110962,
+      "learning_rate": 5.7059745797622624e-05,
+      "loss": 0.9257,
+      "step": 5274
+    },
+    {
+      "epoch": 0.4713293274063484,
+      "grad_norm": 0.5084983110427856,
+      "learning_rate": 5.7045419755498264e-05,
+      "loss": 0.9036,
+      "step": 5275
+    },
+    {
+      "epoch": 0.4714186789376103,
+      "grad_norm": 0.5043264031410217,
+      "learning_rate": 5.7031093123244925e-05,
+      "loss": 1.005,
+      "step": 5276
+    },
+    {
+      "epoch": 0.4715080304688722,
+      "grad_norm": 0.4621896743774414,
+      "learning_rate": 5.7016765902062574e-05,
+      "loss": 0.9614,
+      "step": 5277
+    },
+    {
+      "epoch": 0.471597382000134,
+      "grad_norm": 0.45665669441223145,
+      "learning_rate": 5.700243809315129e-05,
+      "loss": 0.9324,
+      "step": 5278
+    },
+    {
+      "epoch": 0.4716867335313959,
+      "grad_norm": 0.4607732892036438,
+      "learning_rate": 5.6988109697711164e-05,
+      "loss": 1.0348,
+      "step": 5279
+    },
+    {
+      "epoch": 0.4717760850626578,
+      "grad_norm": 0.3779040277004242,
+      "learning_rate": 5.697378071694238e-05,
+      "loss": 1.0202,
+      "step": 5280
+    },
+    {
+      "epoch": 0.4718654365939196,
+      "grad_norm": 0.4088393449783325,
+      "learning_rate": 5.6959451152045126e-05,
+      "loss": 0.9944,
+      "step": 5281
+    },
+    {
+      "epoch": 0.4719547881251815,
+      "grad_norm": 0.5540783405303955,
+      "learning_rate": 5.694512100421965e-05,
+      "loss": 0.9225,
+      "step": 5282
+    },
+    {
+      "epoch": 0.4720441396564434,
+      "grad_norm": 0.4990440011024475,
+      "learning_rate": 5.6930790274666256e-05,
+      "loss": 0.9773,
+      "step": 5283
+    },
+    {
+      "epoch": 0.47213349118770526,
+      "grad_norm": 0.4320768713951111,
+      "learning_rate": 5.691645896458529e-05,
+      "loss": 1.0435,
+      "step": 5284
+    },
+    {
+      "epoch": 0.4722228427189671,
+      "grad_norm": 0.47398531436920166,
+      "learning_rate": 5.690212707517719e-05,
+      "loss": 0.9129,
+      "step": 5285
+    },
+    {
+      "epoch": 0.47231219425022897,
+      "grad_norm": 0.468569815158844,
+      "learning_rate": 5.688779460764235e-05,
+      "loss": 0.9181,
+      "step": 5286
+    },
+    {
+      "epoch": 0.47240154578149085,
+      "grad_norm": 0.37630656361579895,
+      "learning_rate": 5.687346156318132e-05,
+      "loss": 1.0331,
+      "step": 5287
+    },
+    {
+      "epoch": 0.4724908973127527,
+      "grad_norm": 0.49889349937438965,
+      "learning_rate": 5.68591279429946e-05,
+      "loss": 0.9137,
+      "step": 5288
+    },
+    {
+      "epoch": 0.47258024884401456,
+      "grad_norm": 0.4064541161060333,
+      "learning_rate": 5.68447937482828e-05,
+      "loss": 1.0404,
+      "step": 5289
+    },
+    {
+      "epoch": 0.47266960037527644,
+      "grad_norm": 0.416142076253891,
+      "learning_rate": 5.6830458980246574e-05,
+      "loss": 0.9972,
+      "step": 5290
+    },
+    {
+      "epoch": 0.47275895190653827,
+      "grad_norm": 0.5175201892852783,
+      "learning_rate": 5.681612364008659e-05,
+      "loss": 0.9116,
+      "step": 5291
+    },
+    {
+      "epoch": 0.47284830343780015,
+      "grad_norm": 0.5164797902107239,
+      "learning_rate": 5.6801787729003595e-05,
+      "loss": 0.941,
+      "step": 5292
+    },
+    {
+      "epoch": 0.47293765496906204,
+      "grad_norm": 0.4773990213871002,
+      "learning_rate": 5.6787451248198384e-05,
+      "loss": 0.8495,
+      "step": 5293
+    },
+    {
+      "epoch": 0.4730270065003239,
+      "grad_norm": 0.4284185469150543,
+      "learning_rate": 5.677311419887178e-05,
+      "loss": 1.0105,
+      "step": 5294
+    },
+    {
+      "epoch": 0.47311635803158575,
+      "grad_norm": 0.39999738335609436,
+      "learning_rate": 5.675877658222466e-05,
+      "loss": 0.9562,
+      "step": 5295
+    },
+    {
+      "epoch": 0.47320570956284763,
+      "grad_norm": 0.537138044834137,
+      "learning_rate": 5.674443839945797e-05,
+      "loss": 1.0542,
+      "step": 5296
+    },
+    {
+      "epoch": 0.4732950610941095,
+      "grad_norm": 0.40088164806365967,
+      "learning_rate": 5.673009965177266e-05,
+      "loss": 0.9687,
+      "step": 5297
+    },
+    {
+      "epoch": 0.47338441262537134,
+      "grad_norm": 0.4030899405479431,
+      "learning_rate": 5.6715760340369775e-05,
+      "loss": 1.0254,
+      "step": 5298
+    },
+    {
+      "epoch": 0.4734737641566332,
+      "grad_norm": 0.41325297951698303,
+      "learning_rate": 5.670142046645037e-05,
+      "loss": 0.9592,
+      "step": 5299
+    },
+    {
+      "epoch": 0.4735631156878951,
+      "grad_norm": 0.5127687454223633,
+      "learning_rate": 5.668708003121557e-05,
+      "loss": 0.9605,
+      "step": 5300
+    },
+    {
+      "epoch": 0.473652467219157,
+      "grad_norm": 0.4479494094848633,
+      "learning_rate": 5.667273903586655e-05,
+      "loss": 0.9453,
+      "step": 5301
+    },
+    {
+      "epoch": 0.4737418187504188,
+      "grad_norm": 0.41398364305496216,
+      "learning_rate": 5.66583974816045e-05,
+      "loss": 0.8632,
+      "step": 5302
+    },
+    {
+      "epoch": 0.4738311702816807,
+      "grad_norm": 0.3753226101398468,
+      "learning_rate": 5.66440553696307e-05,
+      "loss": 1.002,
+      "step": 5303
+    },
+    {
+      "epoch": 0.4739205218129426,
+      "grad_norm": 0.4559895694255829,
+      "learning_rate": 5.662971270114642e-05,
+      "loss": 0.9741,
+      "step": 5304
+    },
+    {
+      "epoch": 0.4740098733442044,
+      "grad_norm": 0.4614042043685913,
+      "learning_rate": 5.661536947735304e-05,
+      "loss": 0.9813,
+      "step": 5305
+    },
+    {
+      "epoch": 0.4740992248754663,
+      "grad_norm": 0.4653113782405853,
+      "learning_rate": 5.660102569945197e-05,
+      "loss": 0.9023,
+      "step": 5306
+    },
+    {
+      "epoch": 0.4741885764067282,
+      "grad_norm": 0.4278392493724823,
+      "learning_rate": 5.658668136864462e-05,
+      "loss": 0.9543,
+      "step": 5307
+    },
+    {
+      "epoch": 0.47427792793799006,
+      "grad_norm": 0.4447217583656311,
+      "learning_rate": 5.6572336486132495e-05,
+      "loss": 0.9646,
+      "step": 5308
+    },
+    {
+      "epoch": 0.4743672794692519,
+      "grad_norm": 0.4308806359767914,
+      "learning_rate": 5.655799105311713e-05,
+      "loss": 0.8999,
+      "step": 5309
+    },
+    {
+      "epoch": 0.4744566310005138,
+      "grad_norm": 0.401542603969574,
+      "learning_rate": 5.654364507080011e-05,
+      "loss": 0.9943,
+      "step": 5310
+    },
+    {
+      "epoch": 0.47454598253177566,
+      "grad_norm": 0.4810407757759094,
+      "learning_rate": 5.652929854038309e-05,
+      "loss": 0.8805,
+      "step": 5311
+    },
+    {
+      "epoch": 0.4746353340630375,
+      "grad_norm": 0.3643689453601837,
+      "learning_rate": 5.65149514630677e-05,
+      "loss": 0.9883,
+      "step": 5312
+    },
+    {
+      "epoch": 0.47472468559429937,
+      "grad_norm": 0.34783321619033813,
+      "learning_rate": 5.650060384005568e-05,
+      "loss": 0.9739,
+      "step": 5313
+    },
+    {
+      "epoch": 0.47481403712556125,
+      "grad_norm": 0.45785385370254517,
+      "learning_rate": 5.6486255672548794e-05,
+      "loss": 0.9802,
+      "step": 5314
+    },
+    {
+      "epoch": 0.47490338865682313,
+      "grad_norm": 0.5383105278015137,
+      "learning_rate": 5.6471906961748856e-05,
+      "loss": 0.9742,
+      "step": 5315
+    },
+    {
+      "epoch": 0.47499274018808496,
+      "grad_norm": 0.5157361626625061,
+      "learning_rate": 5.6457557708857745e-05,
+      "loss": 0.9478,
+      "step": 5316
+    },
+    {
+      "epoch": 0.47508209171934684,
+      "grad_norm": 0.43339022994041443,
+      "learning_rate": 5.644320791507732e-05,
+      "loss": 0.965,
+      "step": 5317
+    },
+    {
+      "epoch": 0.4751714432506087,
+      "grad_norm": 0.4871036112308502,
+      "learning_rate": 5.642885758160956e-05,
+      "loss": 0.9936,
+      "step": 5318
+    },
+    {
+      "epoch": 0.47526079478187055,
+      "grad_norm": 0.4251110553741455,
+      "learning_rate": 5.641450670965646e-05,
+      "loss": 1.0055,
+      "step": 5319
+    },
+    {
+      "epoch": 0.47535014631313244,
+      "grad_norm": 0.46318215131759644,
+      "learning_rate": 5.640015530042004e-05,
+      "loss": 0.9703,
+      "step": 5320
+    },
+    {
+      "epoch": 0.4754394978443943,
+      "grad_norm": 0.4937099814414978,
+      "learning_rate": 5.6385803355102395e-05,
+      "loss": 0.9328,
+      "step": 5321
+    },
+    {
+      "epoch": 0.47552884937565615,
+      "grad_norm": 0.4998278021812439,
+      "learning_rate": 5.637145087490566e-05,
+      "loss": 0.9198,
+      "step": 5322
+    },
+    {
+      "epoch": 0.47561820090691803,
+      "grad_norm": 0.43673455715179443,
+      "learning_rate": 5.6357097861031984e-05,
+      "loss": 0.9822,
+      "step": 5323
+    },
+    {
+      "epoch": 0.4757075524381799,
+      "grad_norm": 0.5582587122917175,
+      "learning_rate": 5.63427443146836e-05,
+      "loss": 0.9003,
+      "step": 5324
+    },
+    {
+      "epoch": 0.4757969039694418,
+      "grad_norm": 0.47949889302253723,
+      "learning_rate": 5.632839023706278e-05,
+      "loss": 0.9357,
+      "step": 5325
+    },
+    {
+      "epoch": 0.4758862555007036,
+      "grad_norm": 0.480672150850296,
+      "learning_rate": 5.631403562937183e-05,
+      "loss": 0.9649,
+      "step": 5326
+    },
+    {
+      "epoch": 0.4759756070319655,
+      "grad_norm": 0.44741761684417725,
+      "learning_rate": 5.629968049281308e-05,
+      "loss": 1.0556,
+      "step": 5327
+    },
+    {
+      "epoch": 0.4760649585632274,
+      "grad_norm": 0.5157933235168457,
+      "learning_rate": 5.6285324828588934e-05,
+      "loss": 1.0001,
+      "step": 5328
+    },
+    {
+      "epoch": 0.4761543100944892,
+      "grad_norm": 0.45148682594299316,
+      "learning_rate": 5.6270968637901844e-05,
+      "loss": 0.9291,
+      "step": 5329
+    },
+    {
+      "epoch": 0.4762436616257511,
+      "grad_norm": 0.43948331475257874,
+      "learning_rate": 5.625661192195428e-05,
+      "loss": 0.9562,
+      "step": 5330
+    },
+    {
+      "epoch": 0.476333013157013,
+      "grad_norm": 0.49232932925224304,
+      "learning_rate": 5.624225468194879e-05,
+      "loss": 0.8804,
+      "step": 5331
+    },
+    {
+      "epoch": 0.47642236468827487,
+      "grad_norm": 0.44230684638023376,
+      "learning_rate": 5.622789691908794e-05,
+      "loss": 0.9769,
+      "step": 5332
+    },
+    {
+      "epoch": 0.4765117162195367,
+      "grad_norm": 0.46092236042022705,
+      "learning_rate": 5.621353863457431e-05,
+      "loss": 0.9966,
+      "step": 5333
+    },
+    {
+      "epoch": 0.4766010677507986,
+      "grad_norm": 0.45058542490005493,
+      "learning_rate": 5.61991798296106e-05,
+      "loss": 0.921,
+      "step": 5334
+    },
+    {
+      "epoch": 0.47669041928206046,
+      "grad_norm": 0.4571652114391327,
+      "learning_rate": 5.6184820505399514e-05,
+      "loss": 1.0159,
+      "step": 5335
+    },
+    {
+      "epoch": 0.4767797708133223,
+      "grad_norm": 0.45455148816108704,
+      "learning_rate": 5.617046066314377e-05,
+      "loss": 0.9191,
+      "step": 5336
+    },
+    {
+      "epoch": 0.4768691223445842,
+      "grad_norm": 0.5318629741668701,
+      "learning_rate": 5.6156100304046186e-05,
+      "loss": 0.8992,
+      "step": 5337
+    },
+    {
+      "epoch": 0.47695847387584606,
+      "grad_norm": 0.42745161056518555,
+      "learning_rate": 5.614173942930958e-05,
+      "loss": 1.0188,
+      "step": 5338
+    },
+    {
+      "epoch": 0.47704782540710794,
+      "grad_norm": 0.4684041738510132,
+      "learning_rate": 5.6127378040136826e-05,
+      "loss": 1.0166,
+      "step": 5339
+    },
+    {
+      "epoch": 0.47713717693836977,
+      "grad_norm": 0.5044588446617126,
+      "learning_rate": 5.6113016137730844e-05,
+      "loss": 0.9859,
+      "step": 5340
+    },
+    {
+      "epoch": 0.47722652846963165,
+      "grad_norm": 0.41825881600379944,
+      "learning_rate": 5.6098653723294604e-05,
+      "loss": 0.9519,
+      "step": 5341
+    },
+    {
+      "epoch": 0.47731588000089353,
+      "grad_norm": 0.5564168095588684,
+      "learning_rate": 5.6084290798031116e-05,
+      "loss": 0.9854,
+      "step": 5342
+    },
+    {
+      "epoch": 0.47740523153215536,
+      "grad_norm": 0.45696184039115906,
+      "learning_rate": 5.6069927363143405e-05,
+      "loss": 0.9567,
+      "step": 5343
+    },
+    {
+      "epoch": 0.47749458306341724,
+      "grad_norm": 0.49420660734176636,
+      "learning_rate": 5.6055563419834575e-05,
+      "loss": 0.8842,
+      "step": 5344
+    },
+    {
+      "epoch": 0.4775839345946791,
+      "grad_norm": 0.4271746873855591,
+      "learning_rate": 5.604119896930777e-05,
+      "loss": 0.9697,
+      "step": 5345
+    },
+    {
+      "epoch": 0.477673286125941,
+      "grad_norm": 0.4472793936729431,
+      "learning_rate": 5.602683401276615e-05,
+      "loss": 0.9975,
+      "step": 5346
+    },
+    {
+      "epoch": 0.47776263765720284,
+      "grad_norm": 0.5239841341972351,
+      "learning_rate": 5.601246855141296e-05,
+      "loss": 0.9098,
+      "step": 5347
+    },
+    {
+      "epoch": 0.4778519891884647,
+      "grad_norm": 0.4218825697898865,
+      "learning_rate": 5.599810258645142e-05,
+      "loss": 0.9914,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4779413407197266,
+      "grad_norm": 0.4484994113445282,
+      "learning_rate": 5.598373611908485e-05,
+      "loss": 0.953,
+      "step": 5349
+    },
+    {
+      "epoch": 0.47803069225098843,
+      "grad_norm": 0.49346399307250977,
+      "learning_rate": 5.596936915051662e-05,
+      "loss": 0.9513,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4781200437822503,
+      "grad_norm": 0.5029492974281311,
+      "learning_rate": 5.595500168195007e-05,
+      "loss": 0.9715,
+      "step": 5351
+    },
+    {
+      "epoch": 0.4782093953135122,
+      "grad_norm": 0.4443117678165436,
+      "learning_rate": 5.5940633714588675e-05,
+      "loss": 0.9222,
+      "step": 5352
+    },
+    {
+      "epoch": 0.478298746844774,
+      "grad_norm": 0.4409370720386505,
+      "learning_rate": 5.592626524963587e-05,
+      "loss": 0.9049,
+      "step": 5353
+    },
+    {
+      "epoch": 0.4783880983760359,
+      "grad_norm": 0.44364142417907715,
+      "learning_rate": 5.5911896288295183e-05,
+      "loss": 0.9859,
+      "step": 5354
+    },
+    {
+      "epoch": 0.4784774499072978,
+      "grad_norm": 0.52082759141922,
+      "learning_rate": 5.5897526831770175e-05,
+      "loss": 0.9021,
+      "step": 5355
+    },
+    {
+      "epoch": 0.4785668014385597,
+      "grad_norm": 0.38146278262138367,
+      "learning_rate": 5.588315688126442e-05,
+      "loss": 1.0069,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4786561529698215,
+      "grad_norm": 0.44152483344078064,
+      "learning_rate": 5.586878643798158e-05,
+      "loss": 1.0141,
+      "step": 5357
+    },
+    {
+      "epoch": 0.4787455045010834,
+      "grad_norm": 0.5023549795150757,
+      "learning_rate": 5.58544155031253e-05,
+      "loss": 0.964,
+      "step": 5358
+    },
+    {
+      "epoch": 0.47883485603234527,
+      "grad_norm": 0.43117037415504456,
+      "learning_rate": 5.584004407789933e-05,
+      "loss": 1.0089,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4789242075636071,
+      "grad_norm": 0.44485437870025635,
+      "learning_rate": 5.582567216350741e-05,
+      "loss": 0.9615,
+      "step": 5360
+    },
+    {
+      "epoch": 0.479013559094869,
+      "grad_norm": 0.4267984628677368,
+      "learning_rate": 5.581129976115336e-05,
+      "loss": 0.9292,
+      "step": 5361
+    },
+    {
+      "epoch": 0.47910291062613086,
+      "grad_norm": 0.4499671459197998,
+      "learning_rate": 5.579692687204101e-05,
+      "loss": 0.9711,
+      "step": 5362
+    },
+    {
+      "epoch": 0.47919226215739275,
+      "grad_norm": 0.5573053956031799,
+      "learning_rate": 5.578255349737423e-05,
+      "loss": 0.965,
+      "step": 5363
+    },
+    {
+      "epoch": 0.4792816136886546,
+      "grad_norm": 0.41541776061058044,
+      "learning_rate": 5.5768179638356963e-05,
+      "loss": 0.9682,
+      "step": 5364
+    },
+    {
+      "epoch": 0.47937096521991646,
+      "grad_norm": 0.4482407569885254,
+      "learning_rate": 5.575380529619317e-05,
+      "loss": 0.9534,
+      "step": 5365
+    },
+    {
+      "epoch": 0.47946031675117834,
+      "grad_norm": 0.4427946209907532,
+      "learning_rate": 5.573943047208685e-05,
+      "loss": 1.0162,
+      "step": 5366
+    },
+    {
+      "epoch": 0.47954966828244017,
+      "grad_norm": 0.5755056738853455,
+      "learning_rate": 5.572505516724207e-05,
+      "loss": 0.9427,
+      "step": 5367
+    },
+    {
+      "epoch": 0.47963901981370205,
+      "grad_norm": 0.45547792315483093,
+      "learning_rate": 5.5710679382862874e-05,
+      "loss": 0.9308,
+      "step": 5368
+    },
+    {
+      "epoch": 0.47972837134496393,
+      "grad_norm": 0.5226430892944336,
+      "learning_rate": 5.5696303120153405e-05,
+      "loss": 0.9416,
+      "step": 5369
+    },
+    {
+      "epoch": 0.4798177228762258,
+      "grad_norm": 0.4200889468193054,
+      "learning_rate": 5.568192638031784e-05,
+      "loss": 1.042,
+      "step": 5370
+    },
+    {
+      "epoch": 0.47990707440748764,
+      "grad_norm": 0.4149283468723297,
+      "learning_rate": 5.566754916456037e-05,
+      "loss": 0.9486,
+      "step": 5371
+    },
+    {
+      "epoch": 0.4799964259387495,
+      "grad_norm": 0.5608735680580139,
+      "learning_rate": 5.5653171474085254e-05,
+      "loss": 0.8783,
+      "step": 5372
+    },
+    {
+      "epoch": 0.4800857774700114,
+      "grad_norm": 0.47211307287216187,
+      "learning_rate": 5.563879331009677e-05,
+      "loss": 0.9475,
+      "step": 5373
+    },
+    {
+      "epoch": 0.48017512900127324,
+      "grad_norm": 0.38663092255592346,
+      "learning_rate": 5.562441467379923e-05,
+      "loss": 0.9769,
+      "step": 5374
+    },
+    {
+      "epoch": 0.4802644805325351,
+      "grad_norm": 0.36634334921836853,
+      "learning_rate": 5.5610035566397014e-05,
+      "loss": 0.9618,
+      "step": 5375
+    },
+    {
+      "epoch": 0.480353832063797,
+      "grad_norm": 0.4378538131713867,
+      "learning_rate": 5.559565598909452e-05,
+      "loss": 0.9778,
+      "step": 5376
+    },
+    {
+      "epoch": 0.4804431835950589,
+      "grad_norm": 0.40313366055488586,
+      "learning_rate": 5.55812759430962e-05,
+      "loss": 0.9487,
+      "step": 5377
+    },
+    {
+      "epoch": 0.4805325351263207,
+      "grad_norm": 0.4929332733154297,
+      "learning_rate": 5.5566895429606516e-05,
+      "loss": 0.9624,
+      "step": 5378
+    },
+    {
+      "epoch": 0.4806218866575826,
+      "grad_norm": 0.5455753803253174,
+      "learning_rate": 5.5552514449830005e-05,
+      "loss": 0.9789,
+      "step": 5379
+    },
+    {
+      "epoch": 0.4807112381888445,
+      "grad_norm": 0.3865823745727539,
+      "learning_rate": 5.5538133004971216e-05,
+      "loss": 0.9508,
+      "step": 5380
+    },
+    {
+      "epoch": 0.4808005897201063,
+      "grad_norm": 0.48918184638023376,
+      "learning_rate": 5.5523751096234765e-05,
+      "loss": 0.951,
+      "step": 5381
+    },
+    {
+      "epoch": 0.4808899412513682,
+      "grad_norm": 0.49451667070388794,
+      "learning_rate": 5.550936872482528e-05,
+      "loss": 0.9803,
+      "step": 5382
+    },
+    {
+      "epoch": 0.4809792927826301,
+      "grad_norm": 0.40212544798851013,
+      "learning_rate": 5.549498589194744e-05,
+      "loss": 0.9332,
+      "step": 5383
+    },
+    {
+      "epoch": 0.4810686443138919,
+      "grad_norm": 0.44129249453544617,
+      "learning_rate": 5.5480602598805955e-05,
+      "loss": 0.9288,
+      "step": 5384
+    },
+    {
+      "epoch": 0.4811579958451538,
+      "grad_norm": 0.4733429551124573,
+      "learning_rate": 5.546621884660558e-05,
+      "loss": 0.9687,
+      "step": 5385
+    },
+    {
+      "epoch": 0.48124734737641567,
+      "grad_norm": 0.44276171922683716,
+      "learning_rate": 5.545183463655112e-05,
+      "loss": 0.9343,
+      "step": 5386
+    },
+    {
+      "epoch": 0.48133669890767755,
+      "grad_norm": 0.505408525466919,
+      "learning_rate": 5.54374499698474e-05,
+      "loss": 0.8595,
+      "step": 5387
+    },
+    {
+      "epoch": 0.4814260504389394,
+      "grad_norm": 0.4253001809120178,
+      "learning_rate": 5.542306484769927e-05,
+      "loss": 0.9035,
+      "step": 5388
+    },
+    {
+      "epoch": 0.48151540197020126,
+      "grad_norm": 0.5153298377990723,
+      "learning_rate": 5.540867927131166e-05,
+      "loss": 0.9208,
+      "step": 5389
+    },
+    {
+      "epoch": 0.48160475350146315,
+      "grad_norm": 0.443081796169281,
+      "learning_rate": 5.5394293241889516e-05,
+      "loss": 0.9694,
+      "step": 5390
+    },
+    {
+      "epoch": 0.481694105032725,
+      "grad_norm": 0.4136885404586792,
+      "learning_rate": 5.5379906760637814e-05,
+      "loss": 0.9473,
+      "step": 5391
+    },
+    {
+      "epoch": 0.48178345656398686,
+      "grad_norm": 0.3998454809188843,
+      "learning_rate": 5.536551982876157e-05,
+      "loss": 0.9494,
+      "step": 5392
+    },
+    {
+      "epoch": 0.48187280809524874,
+      "grad_norm": 0.410869836807251,
+      "learning_rate": 5.535113244746585e-05,
+      "loss": 0.979,
+      "step": 5393
+    },
+    {
+      "epoch": 0.4819621596265106,
+      "grad_norm": 0.4814499318599701,
+      "learning_rate": 5.5336744617955735e-05,
+      "loss": 0.9871,
+      "step": 5394
+    },
+    {
+      "epoch": 0.48205151115777245,
+      "grad_norm": 0.44015729427337646,
+      "learning_rate": 5.532235634143639e-05,
+      "loss": 1.0168,
+      "step": 5395
+    },
+    {
+      "epoch": 0.48214086268903433,
+      "grad_norm": 0.38507404923439026,
+      "learning_rate": 5.530796761911295e-05,
+      "loss": 0.9503,
+      "step": 5396
+    },
+    {
+      "epoch": 0.4822302142202962,
+      "grad_norm": 0.5117488503456116,
+      "learning_rate": 5.5293578452190675e-05,
+      "loss": 0.9804,
+      "step": 5397
+    },
+    {
+      "epoch": 0.48231956575155804,
+      "grad_norm": 0.4578893780708313,
+      "learning_rate": 5.527918884187475e-05,
+      "loss": 1.038,
+      "step": 5398
+    },
+    {
+      "epoch": 0.4824089172828199,
+      "grad_norm": 0.5064082145690918,
+      "learning_rate": 5.5264798789370496e-05,
+      "loss": 0.8768,
+      "step": 5399
+    },
+    {
+      "epoch": 0.4824982688140818,
+      "grad_norm": 0.42315277457237244,
+      "learning_rate": 5.525040829588323e-05,
+      "loss": 0.9454,
+      "step": 5400
+    },
+    {
+      "epoch": 0.4825876203453437,
+      "grad_norm": 0.3969995379447937,
+      "learning_rate": 5.52360173626183e-05,
+      "loss": 0.9996,
+      "step": 5401
+    },
+    {
+      "epoch": 0.4826769718766055,
+      "grad_norm": 0.5402255654335022,
+      "learning_rate": 5.5221625990781115e-05,
+      "loss": 0.9203,
+      "step": 5402
+    },
+    {
+      "epoch": 0.4827663234078674,
+      "grad_norm": 0.4401646554470062,
+      "learning_rate": 5.520723418157707e-05,
+      "loss": 0.9595,
+      "step": 5403
+    },
+    {
+      "epoch": 0.4828556749391293,
+      "grad_norm": 0.4190303683280945,
+      "learning_rate": 5.519284193621167e-05,
+      "loss": 1.0173,
+      "step": 5404
+    },
+    {
+      "epoch": 0.4829450264703911,
+      "grad_norm": 0.38738173246383667,
+      "learning_rate": 5.517844925589041e-05,
+      "loss": 0.9696,
+      "step": 5405
+    },
+    {
+      "epoch": 0.483034378001653,
+      "grad_norm": 0.4196470081806183,
+      "learning_rate": 5.516405614181883e-05,
+      "loss": 0.9762,
+      "step": 5406
+    },
+    {
+      "epoch": 0.4831237295329149,
+      "grad_norm": 0.4423119127750397,
+      "learning_rate": 5.514966259520249e-05,
+      "loss": 0.9455,
+      "step": 5407
+    },
+    {
+      "epoch": 0.48321308106417676,
+      "grad_norm": 0.45127880573272705,
+      "learning_rate": 5.513526861724703e-05,
+      "loss": 0.994,
+      "step": 5408
+    },
+    {
+      "epoch": 0.4833024325954386,
+      "grad_norm": 0.3754526674747467,
+      "learning_rate": 5.512087420915808e-05,
+      "loss": 0.9263,
+      "step": 5409
+    },
+    {
+      "epoch": 0.4833917841267005,
+      "grad_norm": 0.4244709014892578,
+      "learning_rate": 5.510647937214133e-05,
+      "loss": 1.0555,
+      "step": 5410
+    },
+    {
+      "epoch": 0.48348113565796236,
+      "grad_norm": 0.4102063477039337,
+      "learning_rate": 5.509208410740249e-05,
+      "loss": 0.9678,
+      "step": 5411
+    },
+    {
+      "epoch": 0.4835704871892242,
+      "grad_norm": 0.4029648005962372,
+      "learning_rate": 5.5077688416147345e-05,
+      "loss": 1.0338,
+      "step": 5412
+    },
+    {
+      "epoch": 0.48365983872048607,
+      "grad_norm": 0.4508571922779083,
+      "learning_rate": 5.506329229958166e-05,
+      "loss": 1.035,
+      "step": 5413
+    },
+    {
+      "epoch": 0.48374919025174795,
+      "grad_norm": 0.38453125953674316,
+      "learning_rate": 5.504889575891128e-05,
+      "loss": 1.0721,
+      "step": 5414
+    },
+    {
+      "epoch": 0.4838385417830098,
+      "grad_norm": 0.4159301221370697,
+      "learning_rate": 5.503449879534206e-05,
+      "loss": 0.9116,
+      "step": 5415
+    },
+    {
+      "epoch": 0.48392789331427166,
+      "grad_norm": 0.5297693610191345,
+      "learning_rate": 5.5020101410079896e-05,
+      "loss": 0.9292,
+      "step": 5416
+    },
+    {
+      "epoch": 0.48401724484553355,
+      "grad_norm": 0.3714810907840729,
+      "learning_rate": 5.5005703604330736e-05,
+      "loss": 0.8904,
+      "step": 5417
+    },
+    {
+      "epoch": 0.48410659637679543,
+      "grad_norm": 0.4217725694179535,
+      "learning_rate": 5.499130537930055e-05,
+      "loss": 0.9627,
+      "step": 5418
+    },
+    {
+      "epoch": 0.48419594790805726,
+      "grad_norm": 0.5042391419410706,
+      "learning_rate": 5.497690673619532e-05,
+      "loss": 0.8922,
+      "step": 5419
+    },
+    {
+      "epoch": 0.48428529943931914,
+      "grad_norm": 0.4610849916934967,
+      "learning_rate": 5.496250767622111e-05,
+      "loss": 0.9797,
+      "step": 5420
+    },
+    {
+      "epoch": 0.484374650970581,
+      "grad_norm": 0.4403340518474579,
+      "learning_rate": 5.494810820058398e-05,
+      "loss": 0.9383,
+      "step": 5421
+    },
+    {
+      "epoch": 0.48446400250184285,
+      "grad_norm": 0.38612040877342224,
+      "learning_rate": 5.493370831049004e-05,
+      "loss": 0.9801,
+      "step": 5422
+    },
+    {
+      "epoch": 0.48455335403310473,
+      "grad_norm": 0.4851481318473816,
+      "learning_rate": 5.4919308007145465e-05,
+      "loss": 0.9842,
+      "step": 5423
+    },
+    {
+      "epoch": 0.4846427055643666,
+      "grad_norm": 0.4784315824508667,
+      "learning_rate": 5.490490729175638e-05,
+      "loss": 0.9389,
+      "step": 5424
+    },
+    {
+      "epoch": 0.4847320570956285,
+      "grad_norm": 0.39346399903297424,
+      "learning_rate": 5.489050616552905e-05,
+      "loss": 1.0562,
+      "step": 5425
+    },
+    {
+      "epoch": 0.4848214086268903,
+      "grad_norm": 0.5939106941223145,
+      "learning_rate": 5.487610462966969e-05,
+      "loss": 0.9138,
+      "step": 5426
+    },
+    {
+      "epoch": 0.4849107601581522,
+      "grad_norm": 0.4027441442012787,
+      "learning_rate": 5.4861702685384586e-05,
+      "loss": 1.0134,
+      "step": 5427
+    },
+    {
+      "epoch": 0.4850001116894141,
+      "grad_norm": 0.5542457699775696,
+      "learning_rate": 5.484730033388007e-05,
+      "loss": 0.9148,
+      "step": 5428
+    },
+    {
+      "epoch": 0.4850894632206759,
+      "grad_norm": 0.40890124440193176,
+      "learning_rate": 5.483289757636247e-05,
+      "loss": 1.011,
+      "step": 5429
+    },
+    {
+      "epoch": 0.4851788147519378,
+      "grad_norm": 0.3923655152320862,
+      "learning_rate": 5.481849441403819e-05,
+      "loss": 0.9849,
+      "step": 5430
+    },
+    {
+      "epoch": 0.4852681662831997,
+      "grad_norm": 0.444996178150177,
+      "learning_rate": 5.480409084811363e-05,
+      "loss": 0.9099,
+      "step": 5431
+    },
+    {
+      "epoch": 0.48535751781446157,
+      "grad_norm": 0.5601452589035034,
+      "learning_rate": 5.4789686879795263e-05,
+      "loss": 0.8231,
+      "step": 5432
+    },
+    {
+      "epoch": 0.4854468693457234,
+      "grad_norm": 0.4075312316417694,
+      "learning_rate": 5.477528251028957e-05,
+      "loss": 0.9326,
+      "step": 5433
+    },
+    {
+      "epoch": 0.4855362208769853,
+      "grad_norm": 0.4870973825454712,
+      "learning_rate": 5.4760877740803066e-05,
+      "loss": 0.9654,
+      "step": 5434
+    },
+    {
+      "epoch": 0.48562557240824716,
+      "grad_norm": 0.5276638269424438,
+      "learning_rate": 5.47464725725423e-05,
+      "loss": 0.9561,
+      "step": 5435
+    },
+    {
+      "epoch": 0.485714923939509,
+      "grad_norm": 0.39273542165756226,
+      "learning_rate": 5.473206700671386e-05,
+      "loss": 0.9866,
+      "step": 5436
+    },
+    {
+      "epoch": 0.4858042754707709,
+      "grad_norm": 0.48121076822280884,
+      "learning_rate": 5.471766104452436e-05,
+      "loss": 0.9523,
+      "step": 5437
+    },
+    {
+      "epoch": 0.48589362700203276,
+      "grad_norm": 0.5641562938690186,
+      "learning_rate": 5.4703254687180486e-05,
+      "loss": 0.9552,
+      "step": 5438
+    },
+    {
+      "epoch": 0.48598297853329464,
+      "grad_norm": 0.5305027961730957,
+      "learning_rate": 5.468884793588888e-05,
+      "loss": 0.953,
+      "step": 5439
+    },
+    {
+      "epoch": 0.48607233006455647,
+      "grad_norm": 0.40917256474494934,
+      "learning_rate": 5.467444079185629e-05,
+      "loss": 0.9913,
+      "step": 5440
+    },
+    {
+      "epoch": 0.48616168159581835,
+      "grad_norm": 0.5030852556228638,
+      "learning_rate": 5.466003325628945e-05,
+      "loss": 0.9434,
+      "step": 5441
+    },
+    {
+      "epoch": 0.48625103312708023,
+      "grad_norm": 0.3704419434070587,
+      "learning_rate": 5.464562533039515e-05,
+      "loss": 1.023,
+      "step": 5442
+    },
+    {
+      "epoch": 0.48634038465834206,
+      "grad_norm": 0.46701717376708984,
+      "learning_rate": 5.463121701538021e-05,
+      "loss": 0.8984,
+      "step": 5443
+    },
+    {
+      "epoch": 0.48642973618960395,
+      "grad_norm": 0.43691638112068176,
+      "learning_rate": 5.46168083124515e-05,
+      "loss": 1.0049,
+      "step": 5444
+    },
+    {
+      "epoch": 0.48651908772086583,
+      "grad_norm": 0.4441946744918823,
+      "learning_rate": 5.460239922281586e-05,
+      "loss": 0.9694,
+      "step": 5445
+    },
+    {
+      "epoch": 0.48660843925212766,
+      "grad_norm": 0.42461204528808594,
+      "learning_rate": 5.458798974768022e-05,
+      "loss": 0.9686,
+      "step": 5446
+    },
+    {
+      "epoch": 0.48669779078338954,
+      "grad_norm": 0.44621115922927856,
+      "learning_rate": 5.457357988825155e-05,
+      "loss": 1.0075,
+      "step": 5447
+    },
+    {
+      "epoch": 0.4867871423146514,
+      "grad_norm": 0.5380200743675232,
+      "learning_rate": 5.455916964573682e-05,
+      "loss": 0.8571,
+      "step": 5448
+    },
+    {
+      "epoch": 0.4868764938459133,
+      "grad_norm": 0.5656547546386719,
+      "learning_rate": 5.4544759021343016e-05,
+      "loss": 0.8964,
+      "step": 5449
+    },
+    {
+      "epoch": 0.48696584537717513,
+      "grad_norm": 0.43907156586647034,
+      "learning_rate": 5.453034801627721e-05,
+      "loss": 1.0135,
+      "step": 5450
+    },
+    {
+      "epoch": 0.487055196908437,
+      "grad_norm": 0.4475903809070587,
+      "learning_rate": 5.451593663174647e-05,
+      "loss": 0.8743,
+      "step": 5451
+    },
+    {
+      "epoch": 0.4871445484396989,
+      "grad_norm": 0.42074811458587646,
+      "learning_rate": 5.4501524868957886e-05,
+      "loss": 0.9357,
+      "step": 5452
+    },
+    {
+      "epoch": 0.4872338999709607,
+      "grad_norm": 0.47343945503234863,
+      "learning_rate": 5.448711272911864e-05,
+      "loss": 0.9863,
+      "step": 5453
+    },
+    {
+      "epoch": 0.4873232515022226,
+      "grad_norm": 0.5782976746559143,
+      "learning_rate": 5.4472700213435854e-05,
+      "loss": 0.9197,
+      "step": 5454
+    },
+    {
+      "epoch": 0.4874126030334845,
+      "grad_norm": 0.36520230770111084,
+      "learning_rate": 5.4458287323116755e-05,
+      "loss": 0.9455,
+      "step": 5455
+    },
+    {
+      "epoch": 0.4875019545647464,
+      "grad_norm": 0.5519242882728577,
+      "learning_rate": 5.444387405936856e-05,
+      "loss": 0.8343,
+      "step": 5456
+    },
+    {
+      "epoch": 0.4875913060960082,
+      "grad_norm": 0.4189887046813965,
+      "learning_rate": 5.442946042339855e-05,
+      "loss": 0.9955,
+      "step": 5457
+    },
+    {
+      "epoch": 0.4876806576272701,
+      "grad_norm": 0.46009689569473267,
+      "learning_rate": 5.441504641641403e-05,
+      "loss": 0.9444,
+      "step": 5458
+    },
+    {
+      "epoch": 0.48777000915853197,
+      "grad_norm": 0.5175769925117493,
+      "learning_rate": 5.4400632039622293e-05,
+      "loss": 0.9562,
+      "step": 5459
+    },
+    {
+      "epoch": 0.4878593606897938,
+      "grad_norm": 0.41818684339523315,
+      "learning_rate": 5.438621729423072e-05,
+      "loss": 0.9337,
+      "step": 5460
+    },
+    {
+      "epoch": 0.4879487122210557,
+      "grad_norm": 0.42660924792289734,
+      "learning_rate": 5.4371802181446694e-05,
+      "loss": 1.0022,
+      "step": 5461
+    },
+    {
+      "epoch": 0.48803806375231756,
+      "grad_norm": 0.4058513939380646,
+      "learning_rate": 5.4357386702477645e-05,
+      "loss": 0.99,
+      "step": 5462
+    },
+    {
+      "epoch": 0.48812741528357945,
+      "grad_norm": 0.43901243805885315,
+      "learning_rate": 5.434297085853101e-05,
+      "loss": 0.8644,
+      "step": 5463
+    },
+    {
+      "epoch": 0.4882167668148413,
+      "grad_norm": 0.410495787858963,
+      "learning_rate": 5.432855465081427e-05,
+      "loss": 0.9906,
+      "step": 5464
+    },
+    {
+      "epoch": 0.48830611834610316,
+      "grad_norm": 0.5046043395996094,
+      "learning_rate": 5.431413808053492e-05,
+      "loss": 1.0235,
+      "step": 5465
+    },
+    {
+      "epoch": 0.48839546987736504,
+      "grad_norm": 0.4128507077693939,
+      "learning_rate": 5.4299721148900554e-05,
+      "loss": 1.0187,
+      "step": 5466
+    },
+    {
+      "epoch": 0.48848482140862687,
+      "grad_norm": 0.4370626211166382,
+      "learning_rate": 5.4285303857118685e-05,
+      "loss": 0.935,
+      "step": 5467
+    },
+    {
+      "epoch": 0.48857417293988875,
+      "grad_norm": 0.41134530305862427,
+      "learning_rate": 5.4270886206396956e-05,
+      "loss": 0.9916,
+      "step": 5468
+    },
+    {
+      "epoch": 0.48866352447115063,
+      "grad_norm": 0.6058627367019653,
+      "learning_rate": 5.4256468197942964e-05,
+      "loss": 0.9549,
+      "step": 5469
+    },
+    {
+      "epoch": 0.4887528760024125,
+      "grad_norm": 0.47504645586013794,
+      "learning_rate": 5.42420498329644e-05,
+      "loss": 1.0599,
+      "step": 5470
+    },
+    {
+      "epoch": 0.48884222753367434,
+      "grad_norm": 0.42550885677337646,
+      "learning_rate": 5.4227631112668955e-05,
+      "loss": 0.9998,
+      "step": 5471
+    },
+    {
+      "epoch": 0.48893157906493623,
+      "grad_norm": 0.4544508755207062,
+      "learning_rate": 5.4213212038264325e-05,
+      "loss": 0.9259,
+      "step": 5472
+    },
+    {
+      "epoch": 0.4890209305961981,
+      "grad_norm": 0.41492947936058044,
+      "learning_rate": 5.419879261095829e-05,
+      "loss": 0.9139,
+      "step": 5473
+    },
+    {
+      "epoch": 0.48911028212745994,
+      "grad_norm": 0.3988552391529083,
+      "learning_rate": 5.418437283195862e-05,
+      "loss": 1.0173,
+      "step": 5474
+    },
+    {
+      "epoch": 0.4891996336587218,
+      "grad_norm": 0.48259106278419495,
+      "learning_rate": 5.4169952702473114e-05,
+      "loss": 0.9367,
+      "step": 5475
+    },
+    {
+      "epoch": 0.4892889851899837,
+      "grad_norm": 0.38662248849868774,
+      "learning_rate": 5.415553222370963e-05,
+      "loss": 1.0017,
+      "step": 5476
+    },
+    {
+      "epoch": 0.48937833672124553,
+      "grad_norm": 0.4289376139640808,
+      "learning_rate": 5.414111139687601e-05,
+      "loss": 0.9943,
+      "step": 5477
+    },
+    {
+      "epoch": 0.4894676882525074,
+      "grad_norm": 0.42582669854164124,
+      "learning_rate": 5.4126690223180175e-05,
+      "loss": 1.0044,
+      "step": 5478
+    },
+    {
+      "epoch": 0.4895570397837693,
+      "grad_norm": 0.3854537606239319,
+      "learning_rate": 5.411226870383006e-05,
+      "loss": 0.9932,
+      "step": 5479
+    },
+    {
+      "epoch": 0.4896463913150312,
+      "grad_norm": 0.41988447308540344,
+      "learning_rate": 5.4097846840033604e-05,
+      "loss": 0.9692,
+      "step": 5480
+    },
+    {
+      "epoch": 0.489735742846293,
+      "grad_norm": 0.424517959356308,
+      "learning_rate": 5.408342463299878e-05,
+      "loss": 0.9933,
+      "step": 5481
+    },
+    {
+      "epoch": 0.4898250943775549,
+      "grad_norm": 0.524263858795166,
+      "learning_rate": 5.4069002083933625e-05,
+      "loss": 0.921,
+      "step": 5482
+    },
+    {
+      "epoch": 0.4899144459088168,
+      "grad_norm": 0.5599950551986694,
+      "learning_rate": 5.405457919404619e-05,
+      "loss": 0.9442,
+      "step": 5483
+    },
+    {
+      "epoch": 0.4900037974400786,
+      "grad_norm": 0.3951818645000458,
+      "learning_rate": 5.4040155964544506e-05,
+      "loss": 0.9275,
+      "step": 5484
+    },
+    {
+      "epoch": 0.4900931489713405,
+      "grad_norm": 0.5091139674186707,
+      "learning_rate": 5.40257323966367e-05,
+      "loss": 1.0093,
+      "step": 5485
+    },
+    {
+      "epoch": 0.49018250050260237,
+      "grad_norm": 0.46656838059425354,
+      "learning_rate": 5.40113084915309e-05,
+      "loss": 0.9953,
+      "step": 5486
+    },
+    {
+      "epoch": 0.49027185203386425,
+      "grad_norm": 0.41893553733825684,
+      "learning_rate": 5.399688425043524e-05,
+      "loss": 0.9857,
+      "step": 5487
+    },
+    {
+      "epoch": 0.4903612035651261,
+      "grad_norm": 0.5395665764808655,
+      "learning_rate": 5.398245967455795e-05,
+      "loss": 0.9104,
+      "step": 5488
+    },
+    {
+      "epoch": 0.49045055509638796,
+      "grad_norm": 0.5200545191764832,
+      "learning_rate": 5.396803476510719e-05,
+      "loss": 0.8928,
+      "step": 5489
+    },
+    {
+      "epoch": 0.49053990662764985,
+      "grad_norm": 0.5178142786026001,
+      "learning_rate": 5.3953609523291225e-05,
+      "loss": 0.9587,
+      "step": 5490
+    },
+    {
+      "epoch": 0.4906292581589117,
+      "grad_norm": 0.46901631355285645,
+      "learning_rate": 5.3939183950318316e-05,
+      "loss": 0.9875,
+      "step": 5491
+    },
+    {
+      "epoch": 0.49071860969017356,
+      "grad_norm": 0.3886154294013977,
+      "learning_rate": 5.392475804739677e-05,
+      "loss": 0.9903,
+      "step": 5492
+    },
+    {
+      "epoch": 0.49080796122143544,
+      "grad_norm": 0.4371306598186493,
+      "learning_rate": 5.391033181573491e-05,
+      "loss": 0.9537,
+      "step": 5493
+    },
+    {
+      "epoch": 0.4908973127526973,
+      "grad_norm": 0.5079873204231262,
+      "learning_rate": 5.3895905256541066e-05,
+      "loss": 0.934,
+      "step": 5494
+    },
+    {
+      "epoch": 0.49098666428395915,
+      "grad_norm": 0.46429815888404846,
+      "learning_rate": 5.3881478371023655e-05,
+      "loss": 0.9762,
+      "step": 5495
+    },
+    {
+      "epoch": 0.49107601581522103,
+      "grad_norm": 0.44685813784599304,
+      "learning_rate": 5.3867051160391025e-05,
+      "loss": 1.0016,
+      "step": 5496
+    },
+    {
+      "epoch": 0.4911653673464829,
+      "grad_norm": 0.48742127418518066,
+      "learning_rate": 5.385262362585165e-05,
+      "loss": 0.8865,
+      "step": 5497
+    },
+    {
+      "epoch": 0.49125471887774474,
+      "grad_norm": 0.410982221364975,
+      "learning_rate": 5.3838195768613995e-05,
+      "loss": 0.9393,
+      "step": 5498
+    },
+    {
+      "epoch": 0.49134407040900663,
+      "grad_norm": 0.4197956919670105,
+      "learning_rate": 5.382376758988652e-05,
+      "loss": 0.9513,
+      "step": 5499
+    },
+    {
+      "epoch": 0.4914334219402685,
+      "grad_norm": 0.48656147718429565,
+      "learning_rate": 5.3809339090877745e-05,
+      "loss": 0.8835,
+      "step": 5500
+    },
+    {
+      "epoch": 0.4915227734715304,
+      "grad_norm": 0.43064260482788086,
+      "learning_rate": 5.379491027279622e-05,
+      "loss": 1.0597,
+      "step": 5501
+    },
+    {
+      "epoch": 0.4916121250027922,
+      "grad_norm": 0.39117228984832764,
+      "learning_rate": 5.3780481136850505e-05,
+      "loss": 0.9828,
+      "step": 5502
+    },
+    {
+      "epoch": 0.4917014765340541,
+      "grad_norm": 0.5010277032852173,
+      "learning_rate": 5.37660516842492e-05,
+      "loss": 1.0373,
+      "step": 5503
+    },
+    {
+      "epoch": 0.491790828065316,
+      "grad_norm": 0.367341011762619,
+      "learning_rate": 5.375162191620093e-05,
+      "loss": 0.9854,
+      "step": 5504
+    },
+    {
+      "epoch": 0.4918801795965778,
+      "grad_norm": 0.5271230340003967,
+      "learning_rate": 5.373719183391434e-05,
+      "loss": 0.8852,
+      "step": 5505
+    },
+    {
+      "epoch": 0.4919695311278397,
+      "grad_norm": 0.4072398543357849,
+      "learning_rate": 5.372276143859809e-05,
+      "loss": 0.9513,
+      "step": 5506
+    },
+    {
+      "epoch": 0.4920588826591016,
+      "grad_norm": 0.4533540606498718,
+      "learning_rate": 5.3708330731460876e-05,
+      "loss": 0.9483,
+      "step": 5507
+    },
+    {
+      "epoch": 0.49214823419036346,
+      "grad_norm": 0.48236024379730225,
+      "learning_rate": 5.369389971371145e-05,
+      "loss": 1.027,
+      "step": 5508
+    },
+    {
+      "epoch": 0.4922375857216253,
+      "grad_norm": 0.4861671030521393,
+      "learning_rate": 5.367946838655855e-05,
+      "loss": 0.905,
+      "step": 5509
+    },
+    {
+      "epoch": 0.4923269372528872,
+      "grad_norm": 0.5689902305603027,
+      "learning_rate": 5.366503675121095e-05,
+      "loss": 0.9575,
+      "step": 5510
+    },
+    {
+      "epoch": 0.49241628878414906,
+      "grad_norm": 0.5556491613388062,
+      "learning_rate": 5.3650604808877456e-05,
+      "loss": 0.9894,
+      "step": 5511
+    },
+    {
+      "epoch": 0.4925056403154109,
+      "grad_norm": 0.3819589912891388,
+      "learning_rate": 5.36361725607669e-05,
+      "loss": 0.9288,
+      "step": 5512
+    },
+    {
+      "epoch": 0.49259499184667277,
+      "grad_norm": 0.4342455267906189,
+      "learning_rate": 5.3621740008088126e-05,
+      "loss": 1.0173,
+      "step": 5513
+    },
+    {
+      "epoch": 0.49268434337793465,
+      "grad_norm": 0.5042043924331665,
+      "learning_rate": 5.360730715205002e-05,
+      "loss": 1.0367,
+      "step": 5514
+    },
+    {
+      "epoch": 0.4927736949091965,
+      "grad_norm": 0.4823912978172302,
+      "learning_rate": 5.35928739938615e-05,
+      "loss": 0.9668,
+      "step": 5515
+    },
+    {
+      "epoch": 0.49286304644045836,
+      "grad_norm": 0.4222562313079834,
+      "learning_rate": 5.357844053473148e-05,
+      "loss": 0.9825,
+      "step": 5516
+    },
+    {
+      "epoch": 0.49295239797172025,
+      "grad_norm": 0.48596012592315674,
+      "learning_rate": 5.356400677586891e-05,
+      "loss": 0.9142,
+      "step": 5517
+    },
+    {
+      "epoch": 0.49304174950298213,
+      "grad_norm": 0.4577832818031311,
+      "learning_rate": 5.3549572718482785e-05,
+      "loss": 0.9555,
+      "step": 5518
+    },
+    {
+      "epoch": 0.49313110103424396,
+      "grad_norm": 0.4585660696029663,
+      "learning_rate": 5.353513836378213e-05,
+      "loss": 0.9421,
+      "step": 5519
+    },
+    {
+      "epoch": 0.49322045256550584,
+      "grad_norm": 0.44608667492866516,
+      "learning_rate": 5.352070371297594e-05,
+      "loss": 1.0034,
+      "step": 5520
+    },
+    {
+      "epoch": 0.4933098040967677,
+      "grad_norm": 0.4560176432132721,
+      "learning_rate": 5.350626876727328e-05,
+      "loss": 0.9055,
+      "step": 5521
+    },
+    {
+      "epoch": 0.49339915562802955,
+      "grad_norm": 0.4542362689971924,
+      "learning_rate": 5.349183352788325e-05,
+      "loss": 0.976,
+      "step": 5522
+    },
+    {
+      "epoch": 0.49348850715929143,
+      "grad_norm": 0.4418618679046631,
+      "learning_rate": 5.347739799601493e-05,
+      "loss": 0.9314,
+      "step": 5523
+    },
+    {
+      "epoch": 0.4935778586905533,
+      "grad_norm": 0.4947437345981598,
+      "learning_rate": 5.3462962172877475e-05,
+      "loss": 0.9936,
+      "step": 5524
+    },
+    {
+      "epoch": 0.4936672102218152,
+      "grad_norm": 0.47864702343940735,
+      "learning_rate": 5.3448526059680016e-05,
+      "loss": 0.9143,
+      "step": 5525
+    },
+    {
+      "epoch": 0.49375656175307703,
+      "grad_norm": 0.40519431233406067,
+      "learning_rate": 5.343408965763174e-05,
+      "loss": 0.9579,
+      "step": 5526
+    },
+    {
+      "epoch": 0.4938459132843389,
+      "grad_norm": 0.4214836061000824,
+      "learning_rate": 5.341965296794185e-05,
+      "loss": 1.0252,
+      "step": 5527
+    },
+    {
+      "epoch": 0.4939352648156008,
+      "grad_norm": 0.4515914022922516,
+      "learning_rate": 5.340521599181959e-05,
+      "loss": 0.9455,
+      "step": 5528
+    },
+    {
+      "epoch": 0.4940246163468626,
+      "grad_norm": 0.39432477951049805,
+      "learning_rate": 5.339077873047418e-05,
+      "loss": 0.9309,
+      "step": 5529
+    },
+    {
+      "epoch": 0.4941139678781245,
+      "grad_norm": 0.4477103054523468,
+      "learning_rate": 5.337634118511491e-05,
+      "loss": 0.9542,
+      "step": 5530
+    },
+    {
+      "epoch": 0.4942033194093864,
+      "grad_norm": 0.48979735374450684,
+      "learning_rate": 5.336190335695107e-05,
+      "loss": 0.9266,
+      "step": 5531
+    },
+    {
+      "epoch": 0.49429267094064827,
+      "grad_norm": 0.470674991607666,
+      "learning_rate": 5.3347465247192e-05,
+      "loss": 0.9846,
+      "step": 5532
+    },
+    {
+      "epoch": 0.4943820224719101,
+      "grad_norm": 0.5746048092842102,
+      "learning_rate": 5.333302685704702e-05,
+      "loss": 0.9958,
+      "step": 5533
+    },
+    {
+      "epoch": 0.494471374003172,
+      "grad_norm": 0.44571205973625183,
+      "learning_rate": 5.331858818772553e-05,
+      "loss": 0.9015,
+      "step": 5534
+    },
+    {
+      "epoch": 0.49456072553443386,
+      "grad_norm": 0.39946070313453674,
+      "learning_rate": 5.330414924043689e-05,
+      "loss": 1.0127,
+      "step": 5535
+    },
+    {
+      "epoch": 0.4946500770656957,
+      "grad_norm": 0.3653550148010254,
+      "learning_rate": 5.3289710016390535e-05,
+      "loss": 0.9527,
+      "step": 5536
+    },
+    {
+      "epoch": 0.4947394285969576,
+      "grad_norm": 0.4395412802696228,
+      "learning_rate": 5.327527051679591e-05,
+      "loss": 0.9398,
+      "step": 5537
+    },
+    {
+      "epoch": 0.49482878012821946,
+      "grad_norm": 0.5096445679664612,
+      "learning_rate": 5.3260830742862456e-05,
+      "loss": 0.9331,
+      "step": 5538
+    },
+    {
+      "epoch": 0.49491813165948134,
+      "grad_norm": 0.5146742463111877,
+      "learning_rate": 5.324639069579968e-05,
+      "loss": 0.9474,
+      "step": 5539
+    },
+    {
+      "epoch": 0.49500748319074317,
+      "grad_norm": 0.41445231437683105,
+      "learning_rate": 5.3231950376817065e-05,
+      "loss": 0.9408,
+      "step": 5540
+    },
+    {
+      "epoch": 0.49509683472200505,
+      "grad_norm": 0.4815199077129364,
+      "learning_rate": 5.3217509787124165e-05,
+      "loss": 0.9785,
+      "step": 5541
+    },
+    {
+      "epoch": 0.49518618625326694,
+      "grad_norm": 0.43417826294898987,
+      "learning_rate": 5.320306892793052e-05,
+      "loss": 0.9017,
+      "step": 5542
+    },
+    {
+      "epoch": 0.49527553778452876,
+      "grad_norm": 0.4291759133338928,
+      "learning_rate": 5.318862780044571e-05,
+      "loss": 0.9912,
+      "step": 5543
+    },
+    {
+      "epoch": 0.49536488931579065,
+      "grad_norm": 0.41811853647232056,
+      "learning_rate": 5.317418640587934e-05,
+      "loss": 0.9695,
+      "step": 5544
+    },
+    {
+      "epoch": 0.49545424084705253,
+      "grad_norm": 0.4489332139492035,
+      "learning_rate": 5.315974474544102e-05,
+      "loss": 0.9603,
+      "step": 5545
+    },
+    {
+      "epoch": 0.49554359237831436,
+      "grad_norm": 0.4664028584957123,
+      "learning_rate": 5.3145302820340404e-05,
+      "loss": 0.9628,
+      "step": 5546
+    },
+    {
+      "epoch": 0.49563294390957624,
+      "grad_norm": 0.5637416243553162,
+      "learning_rate": 5.313086063178715e-05,
+      "loss": 1.0238,
+      "step": 5547
+    },
+    {
+      "epoch": 0.4957222954408381,
+      "grad_norm": 0.4180266559123993,
+      "learning_rate": 5.311641818099093e-05,
+      "loss": 0.9883,
+      "step": 5548
+    },
+    {
+      "epoch": 0.4958116469721,
+      "grad_norm": 0.3877975344657898,
+      "learning_rate": 5.310197546916149e-05,
+      "loss": 0.93,
+      "step": 5549
+    },
+    {
+      "epoch": 0.49590099850336183,
+      "grad_norm": 0.3823062777519226,
+      "learning_rate": 5.308753249750853e-05,
+      "loss": 1.0135,
+      "step": 5550
+    },
+    {
+      "epoch": 0.4959903500346237,
+      "grad_norm": 0.4974936246871948,
+      "learning_rate": 5.3073089267241804e-05,
+      "loss": 0.9422,
+      "step": 5551
+    },
+    {
+      "epoch": 0.4960797015658856,
+      "grad_norm": 0.44719165563583374,
+      "learning_rate": 5.305864577957111e-05,
+      "loss": 0.9379,
+      "step": 5552
+    },
+    {
+      "epoch": 0.49616905309714743,
+      "grad_norm": 0.4038889706134796,
+      "learning_rate": 5.30442020357062e-05,
+      "loss": 0.9951,
+      "step": 5553
+    },
+    {
+      "epoch": 0.4962584046284093,
+      "grad_norm": 0.497601181268692,
+      "learning_rate": 5.3029758036856955e-05,
+      "loss": 0.9611,
+      "step": 5554
+    },
+    {
+      "epoch": 0.4963477561596712,
+      "grad_norm": 0.4878460764884949,
+      "learning_rate": 5.301531378423317e-05,
+      "loss": 0.9078,
+      "step": 5555
+    },
+    {
+      "epoch": 0.4964371076909331,
+      "grad_norm": 0.5067799687385559,
+      "learning_rate": 5.300086927904471e-05,
+      "loss": 0.9569,
+      "step": 5556
+    },
+    {
+      "epoch": 0.4965264592221949,
+      "grad_norm": 0.3875921368598938,
+      "learning_rate": 5.298642452250145e-05,
+      "loss": 0.9997,
+      "step": 5557
+    },
+    {
+      "epoch": 0.4966158107534568,
+      "grad_norm": 0.3980264365673065,
+      "learning_rate": 5.297197951581332e-05,
+      "loss": 1.0059,
+      "step": 5558
+    },
+    {
+      "epoch": 0.49670516228471867,
+      "grad_norm": 0.42988121509552,
+      "learning_rate": 5.295753426019022e-05,
+      "loss": 0.9343,
+      "step": 5559
+    },
+    {
+      "epoch": 0.4967945138159805,
+      "grad_norm": 0.4407704174518585,
+      "learning_rate": 5.294308875684211e-05,
+      "loss": 0.9433,
+      "step": 5560
+    },
+    {
+      "epoch": 0.4968838653472424,
+      "grad_norm": 0.4002476930618286,
+      "learning_rate": 5.2928643006978926e-05,
+      "loss": 1.0739,
+      "step": 5561
+    },
+    {
+      "epoch": 0.49697321687850426,
+      "grad_norm": 0.44479769468307495,
+      "learning_rate": 5.291419701181068e-05,
+      "loss": 0.9789,
+      "step": 5562
+    },
+    {
+      "epoch": 0.49706256840976615,
+      "grad_norm": 0.4247623383998871,
+      "learning_rate": 5.2899750772547385e-05,
+      "loss": 0.9605,
+      "step": 5563
+    },
+    {
+      "epoch": 0.497151919941028,
+      "grad_norm": 0.4519825279712677,
+      "learning_rate": 5.288530429039904e-05,
+      "loss": 0.9244,
+      "step": 5564
+    },
+    {
+      "epoch": 0.49724127147228986,
+      "grad_norm": 0.42883196473121643,
+      "learning_rate": 5.287085756657573e-05,
+      "loss": 0.955,
+      "step": 5565
+    },
+    {
+      "epoch": 0.49733062300355174,
+      "grad_norm": 0.3723878562450409,
+      "learning_rate": 5.2856410602287465e-05,
+      "loss": 0.9387,
+      "step": 5566
+    },
+    {
+      "epoch": 0.49741997453481357,
+      "grad_norm": 0.4619128704071045,
+      "learning_rate": 5.284196339874438e-05,
+      "loss": 0.9301,
+      "step": 5567
+    },
+    {
+      "epoch": 0.49750932606607545,
+      "grad_norm": 0.42548492550849915,
+      "learning_rate": 5.282751595715656e-05,
+      "loss": 1.0291,
+      "step": 5568
+    },
+    {
+      "epoch": 0.49759867759733734,
+      "grad_norm": 0.43272992968559265,
+      "learning_rate": 5.281306827873416e-05,
+      "loss": 1.0186,
+      "step": 5569
+    },
+    {
+      "epoch": 0.4976880291285992,
+      "grad_norm": 0.4682878851890564,
+      "learning_rate": 5.279862036468728e-05,
+      "loss": 0.9629,
+      "step": 5570
+    },
+    {
+      "epoch": 0.49777738065986105,
+      "grad_norm": 0.5456759929656982,
+      "learning_rate": 5.2784172216226124e-05,
+      "loss": 0.8809,
+      "step": 5571
+    },
+    {
+      "epoch": 0.49786673219112293,
+      "grad_norm": 0.38989564776420593,
+      "learning_rate": 5.2769723834560855e-05,
+      "loss": 0.9556,
+      "step": 5572
+    },
+    {
+      "epoch": 0.4979560837223848,
+      "grad_norm": 0.4226441979408264,
+      "learning_rate": 5.275527522090171e-05,
+      "loss": 0.9417,
+      "step": 5573
+    },
+    {
+      "epoch": 0.49804543525364664,
+      "grad_norm": 0.4417705833911896,
+      "learning_rate": 5.2740826376458894e-05,
+      "loss": 0.9943,
+      "step": 5574
+    },
+    {
+      "epoch": 0.4981347867849085,
+      "grad_norm": 0.4658608138561249,
+      "learning_rate": 5.2726377302442656e-05,
+      "loss": 0.907,
+      "step": 5575
+    },
+    {
+      "epoch": 0.4982241383161704,
+      "grad_norm": 0.516185998916626,
+      "learning_rate": 5.2711928000063245e-05,
+      "loss": 0.9793,
+      "step": 5576
+    },
+    {
+      "epoch": 0.49831348984743223,
+      "grad_norm": 0.4596862494945526,
+      "learning_rate": 5.269747847053096e-05,
+      "loss": 0.9666,
+      "step": 5577
+    },
+    {
+      "epoch": 0.4984028413786941,
+      "grad_norm": 0.43760859966278076,
+      "learning_rate": 5.268302871505611e-05,
+      "loss": 0.9403,
+      "step": 5578
+    },
+    {
+      "epoch": 0.498492192909956,
+      "grad_norm": 0.4713875949382782,
+      "learning_rate": 5.2668578734849e-05,
+      "loss": 0.867,
+      "step": 5579
+    },
+    {
+      "epoch": 0.4985815444412179,
+      "grad_norm": 0.4082961678504944,
+      "learning_rate": 5.265412853111997e-05,
+      "loss": 1.002,
+      "step": 5580
+    },
+    {
+      "epoch": 0.4986708959724797,
+      "grad_norm": 0.4304913282394409,
+      "learning_rate": 5.2639678105079394e-05,
+      "loss": 0.9176,
+      "step": 5581
+    },
+    {
+      "epoch": 0.4987602475037416,
+      "grad_norm": 0.5440099835395813,
+      "learning_rate": 5.262522745793764e-05,
+      "loss": 0.8735,
+      "step": 5582
+    },
+    {
+      "epoch": 0.4988495990350035,
+      "grad_norm": 0.43295422196388245,
+      "learning_rate": 5.26107765909051e-05,
+      "loss": 0.9699,
+      "step": 5583
+    },
+    {
+      "epoch": 0.4989389505662653,
+      "grad_norm": 0.5452165603637695,
+      "learning_rate": 5.25963255051922e-05,
+      "loss": 1.0142,
+      "step": 5584
+    },
+    {
+      "epoch": 0.4990283020975272,
+      "grad_norm": 0.4139503240585327,
+      "learning_rate": 5.258187420200935e-05,
+      "loss": 0.9684,
+      "step": 5585
+    },
+    {
+      "epoch": 0.49911765362878907,
+      "grad_norm": 0.4647497832775116,
+      "learning_rate": 5.256742268256703e-05,
+      "loss": 0.9009,
+      "step": 5586
+    },
+    {
+      "epoch": 0.49920700516005095,
+      "grad_norm": 0.41619575023651123,
+      "learning_rate": 5.255297094807568e-05,
+      "loss": 0.9714,
+      "step": 5587
+    },
+    {
+      "epoch": 0.4992963566913128,
+      "grad_norm": 0.46064916253089905,
+      "learning_rate": 5.25385189997458e-05,
+      "loss": 0.9339,
+      "step": 5588
+    },
+    {
+      "epoch": 0.49938570822257466,
+      "grad_norm": 0.46546754240989685,
+      "learning_rate": 5.252406683878791e-05,
+      "loss": 0.9525,
+      "step": 5589
+    },
+    {
+      "epoch": 0.49947505975383655,
+      "grad_norm": 0.5033959150314331,
+      "learning_rate": 5.250961446641251e-05,
+      "loss": 0.9203,
+      "step": 5590
+    },
+    {
+      "epoch": 0.4995644112850984,
+      "grad_norm": 0.44631263613700867,
+      "learning_rate": 5.249516188383014e-05,
+      "loss": 0.9998,
+      "step": 5591
+    },
+    {
+      "epoch": 0.49965376281636026,
+      "grad_norm": 0.4394984543323517,
+      "learning_rate": 5.248070909225136e-05,
+      "loss": 0.9694,
+      "step": 5592
+    },
+    {
+      "epoch": 0.49974311434762214,
+      "grad_norm": 0.5696101784706116,
+      "learning_rate": 5.2466256092886754e-05,
+      "loss": 0.93,
+      "step": 5593
+    },
+    {
+      "epoch": 0.499832465878884,
+      "grad_norm": 0.4666324555873871,
+      "learning_rate": 5.245180288694692e-05,
+      "loss": 0.8524,
+      "step": 5594
+    },
+    {
+      "epoch": 0.49992181741014585,
+      "grad_norm": 0.5041062235832214,
+      "learning_rate": 5.243734947564247e-05,
+      "loss": 0.9306,
+      "step": 5595
+    },
+    {
+      "epoch": 0.5000111689414077,
+      "grad_norm": 0.4425108730792999,
+      "learning_rate": 5.242289586018401e-05,
+      "loss": 0.9774,
+      "step": 5596
+    },
+    {
+      "epoch": 0.5001005204726696,
+      "grad_norm": 0.38095951080322266,
+      "learning_rate": 5.240844204178219e-05,
+      "loss": 1.0063,
+      "step": 5597
+    },
+    {
+      "epoch": 0.5001898720039315,
+      "grad_norm": 0.4874595105648041,
+      "learning_rate": 5.23939880216477e-05,
+      "loss": 0.8665,
+      "step": 5598
+    },
+    {
+      "epoch": 0.5002792235351934,
+      "grad_norm": 0.4805736839771271,
+      "learning_rate": 5.237953380099116e-05,
+      "loss": 0.9152,
+      "step": 5599
+    },
+    {
+      "epoch": 0.5003685750664552,
+      "grad_norm": 0.60560142993927,
+      "learning_rate": 5.236507938102334e-05,
+      "loss": 0.8751,
+      "step": 5600
+    },
+    {
+      "epoch": 0.500457926597717,
+      "grad_norm": 0.37698978185653687,
+      "learning_rate": 5.2350624762954884e-05,
+      "loss": 0.9901,
+      "step": 5601
+    },
+    {
+      "epoch": 0.5005472781289789,
+      "grad_norm": 0.5520381331443787,
+      "learning_rate": 5.233616994799655e-05,
+      "loss": 0.9581,
+      "step": 5602
+    },
+    {
+      "epoch": 0.5006366296602408,
+      "grad_norm": 0.4423448443412781,
+      "learning_rate": 5.232171493735909e-05,
+      "loss": 0.936,
+      "step": 5603
+    },
+    {
+      "epoch": 0.5007259811915027,
+      "grad_norm": 0.4027252495288849,
+      "learning_rate": 5.230725973225324e-05,
+      "loss": 1.0058,
+      "step": 5604
+    },
+    {
+      "epoch": 0.5008153327227646,
+      "grad_norm": 0.39916396141052246,
+      "learning_rate": 5.229280433388982e-05,
+      "loss": 1.1366,
+      "step": 5605
+    },
+    {
+      "epoch": 0.5009046842540265,
+      "grad_norm": 0.49851128458976746,
+      "learning_rate": 5.227834874347958e-05,
+      "loss": 0.8975,
+      "step": 5606
+    },
+    {
+      "epoch": 0.5009940357852882,
+      "grad_norm": 0.37986597418785095,
+      "learning_rate": 5.226389296223335e-05,
+      "loss": 1.0004,
+      "step": 5607
+    },
+    {
+      "epoch": 0.5010833873165501,
+      "grad_norm": 0.41239768266677856,
+      "learning_rate": 5.224943699136195e-05,
+      "loss": 0.9784,
+      "step": 5608
+    },
+    {
+      "epoch": 0.501172738847812,
+      "grad_norm": 0.40920108556747437,
+      "learning_rate": 5.223498083207622e-05,
+      "loss": 0.9932,
+      "step": 5609
+    },
+    {
+      "epoch": 0.5012620903790739,
+      "grad_norm": 0.4581597149372101,
+      "learning_rate": 5.222052448558704e-05,
+      "loss": 0.914,
+      "step": 5610
+    },
+    {
+      "epoch": 0.5013514419103358,
+      "grad_norm": 0.39035099744796753,
+      "learning_rate": 5.2206067953105255e-05,
+      "loss": 0.992,
+      "step": 5611
+    },
+    {
+      "epoch": 0.5014407934415976,
+      "grad_norm": 0.4067721962928772,
+      "learning_rate": 5.219161123584176e-05,
+      "loss": 0.9821,
+      "step": 5612
+    },
+    {
+      "epoch": 0.5015301449728595,
+      "grad_norm": 0.5137872695922852,
+      "learning_rate": 5.217715433500747e-05,
+      "loss": 0.9398,
+      "step": 5613
+    },
+    {
+      "epoch": 0.5016194965041213,
+      "grad_norm": 0.5111674070358276,
+      "learning_rate": 5.21626972518133e-05,
+      "loss": 0.9266,
+      "step": 5614
+    },
+    {
+      "epoch": 0.5017088480353832,
+      "grad_norm": 0.41246017813682556,
+      "learning_rate": 5.214823998747017e-05,
+      "loss": 0.965,
+      "step": 5615
+    },
+    {
+      "epoch": 0.5017981995666451,
+      "grad_norm": 0.48276659846305847,
+      "learning_rate": 5.213378254318906e-05,
+      "loss": 0.9673,
+      "step": 5616
+    },
+    {
+      "epoch": 0.501887551097907,
+      "grad_norm": 0.48614341020584106,
+      "learning_rate": 5.211932492018091e-05,
+      "loss": 0.874,
+      "step": 5617
+    },
+    {
+      "epoch": 0.5019769026291688,
+      "grad_norm": 0.42692437767982483,
+      "learning_rate": 5.21048671196567e-05,
+      "loss": 0.9371,
+      "step": 5618
+    },
+    {
+      "epoch": 0.5020662541604307,
+      "grad_norm": 0.41222038865089417,
+      "learning_rate": 5.209040914282744e-05,
+      "loss": 0.9588,
+      "step": 5619
+    },
+    {
+      "epoch": 0.5021556056916925,
+      "grad_norm": 0.4768015146255493,
+      "learning_rate": 5.2075950990904133e-05,
+      "loss": 0.943,
+      "step": 5620
+    },
+    {
+      "epoch": 0.5022449572229544,
+      "grad_norm": 0.37785643339157104,
+      "learning_rate": 5.20614926650978e-05,
+      "loss": 0.9968,
+      "step": 5621
+    },
+    {
+      "epoch": 0.5023343087542163,
+      "grad_norm": 0.4530482590198517,
+      "learning_rate": 5.2047034166619476e-05,
+      "loss": 0.9125,
+      "step": 5622
+    },
+    {
+      "epoch": 0.5024236602854781,
+      "grad_norm": 0.4255998432636261,
+      "learning_rate": 5.2032575496680224e-05,
+      "loss": 0.9464,
+      "step": 5623
+    },
+    {
+      "epoch": 0.50251301181674,
+      "grad_norm": 0.41948202252388,
+      "learning_rate": 5.201811665649111e-05,
+      "loss": 0.9442,
+      "step": 5624
+    },
+    {
+      "epoch": 0.5026023633480019,
+      "grad_norm": 0.38996177911758423,
+      "learning_rate": 5.200365764726323e-05,
+      "loss": 0.9536,
+      "step": 5625
+    },
+    {
+      "epoch": 0.5026917148792638,
+      "grad_norm": 0.4734954833984375,
+      "learning_rate": 5.198919847020765e-05,
+      "loss": 0.9781,
+      "step": 5626
+    },
+    {
+      "epoch": 0.5027810664105256,
+      "grad_norm": 0.4030005633831024,
+      "learning_rate": 5.197473912653549e-05,
+      "loss": 0.9611,
+      "step": 5627
+    },
+    {
+      "epoch": 0.5028704179417874,
+      "grad_norm": 0.4445195496082306,
+      "learning_rate": 5.1960279617457874e-05,
+      "loss": 0.9721,
+      "step": 5628
+    },
+    {
+      "epoch": 0.5029597694730493,
+      "grad_norm": 0.3933766186237335,
+      "learning_rate": 5.1945819944185944e-05,
+      "loss": 0.9375,
+      "step": 5629
+    },
+    {
+      "epoch": 0.5030491210043112,
+      "grad_norm": 0.40946388244628906,
+      "learning_rate": 5.193136010793088e-05,
+      "loss": 1.0194,
+      "step": 5630
+    },
+    {
+      "epoch": 0.5031384725355731,
+      "grad_norm": 0.49432265758514404,
+      "learning_rate": 5.1916900109903796e-05,
+      "loss": 0.9872,
+      "step": 5631
+    },
+    {
+      "epoch": 0.503227824066835,
+      "grad_norm": 0.5955692529678345,
+      "learning_rate": 5.190243995131591e-05,
+      "loss": 0.9548,
+      "step": 5632
+    },
+    {
+      "epoch": 0.5033171755980969,
+      "grad_norm": 0.47443050146102905,
+      "learning_rate": 5.188797963337839e-05,
+      "loss": 0.9318,
+      "step": 5633
+    },
+    {
+      "epoch": 0.5034065271293586,
+      "grad_norm": 0.5481427907943726,
+      "learning_rate": 5.187351915730245e-05,
+      "loss": 0.8878,
+      "step": 5634
+    },
+    {
+      "epoch": 0.5034958786606205,
+      "grad_norm": 0.5294390320777893,
+      "learning_rate": 5.185905852429933e-05,
+      "loss": 0.8749,
+      "step": 5635
+    },
+    {
+      "epoch": 0.5035852301918824,
+      "grad_norm": 0.4480739235877991,
+      "learning_rate": 5.184459773558022e-05,
+      "loss": 1.0006,
+      "step": 5636
+    },
+    {
+      "epoch": 0.5036745817231443,
+      "grad_norm": 0.4594273865222931,
+      "learning_rate": 5.183013679235639e-05,
+      "loss": 1.0186,
+      "step": 5637
+    },
+    {
+      "epoch": 0.5037639332544062,
+      "grad_norm": 0.4672705829143524,
+      "learning_rate": 5.1815675695839095e-05,
+      "loss": 0.8973,
+      "step": 5638
+    },
+    {
+      "epoch": 0.503853284785668,
+      "grad_norm": 0.405490517616272,
+      "learning_rate": 5.180121444723961e-05,
+      "loss": 0.9966,
+      "step": 5639
+    },
+    {
+      "epoch": 0.5039426363169299,
+      "grad_norm": 0.4211828410625458,
+      "learning_rate": 5.178675304776923e-05,
+      "loss": 0.9086,
+      "step": 5640
+    },
+    {
+      "epoch": 0.5040319878481917,
+      "grad_norm": 0.4052455723285675,
+      "learning_rate": 5.177229149863922e-05,
+      "loss": 0.9668,
+      "step": 5641
+    },
+    {
+      "epoch": 0.5041213393794536,
+      "grad_norm": 0.4231189787387848,
+      "learning_rate": 5.1757829801060905e-05,
+      "loss": 0.9387,
+      "step": 5642
+    },
+    {
+      "epoch": 0.5042106909107155,
+      "grad_norm": 0.48749008774757385,
+      "learning_rate": 5.1743367956245614e-05,
+      "loss": 0.9537,
+      "step": 5643
+    },
+    {
+      "epoch": 0.5043000424419773,
+      "grad_norm": 0.45534324645996094,
+      "learning_rate": 5.172890596540466e-05,
+      "loss": 0.9535,
+      "step": 5644
+    },
+    {
+      "epoch": 0.5043893939732392,
+      "grad_norm": 0.39456576108932495,
+      "learning_rate": 5.171444382974943e-05,
+      "loss": 1.0038,
+      "step": 5645
+    },
+    {
+      "epoch": 0.5044787455045011,
+      "grad_norm": 0.4487841725349426,
+      "learning_rate": 5.169998155049123e-05,
+      "loss": 0.9386,
+      "step": 5646
+    },
+    {
+      "epoch": 0.504568097035763,
+      "grad_norm": 0.4985732138156891,
+      "learning_rate": 5.168551912884147e-05,
+      "loss": 0.9678,
+      "step": 5647
+    },
+    {
+      "epoch": 0.5046574485670248,
+      "grad_norm": 0.42822030186653137,
+      "learning_rate": 5.16710565660115e-05,
+      "loss": 1.0048,
+      "step": 5648
+    },
+    {
+      "epoch": 0.5047468000982867,
+      "grad_norm": 0.4911450445652008,
+      "learning_rate": 5.165659386321273e-05,
+      "loss": 0.9387,
+      "step": 5649
+    },
+    {
+      "epoch": 0.5048361516295485,
+      "grad_norm": 0.43765896558761597,
+      "learning_rate": 5.164213102165659e-05,
+      "loss": 0.9642,
+      "step": 5650
+    },
+    {
+      "epoch": 0.5049255031608104,
+      "grad_norm": 0.4351087212562561,
+      "learning_rate": 5.1627668042554454e-05,
+      "loss": 0.9759,
+      "step": 5651
+    },
+    {
+      "epoch": 0.5050148546920723,
+      "grad_norm": 0.49687519669532776,
+      "learning_rate": 5.1613204927117785e-05,
+      "loss": 0.889,
+      "step": 5652
+    },
+    {
+      "epoch": 0.5051042062233342,
+      "grad_norm": 0.4558013379573822,
+      "learning_rate": 5.1598741676557995e-05,
+      "loss": 1.0162,
+      "step": 5653
+    },
+    {
+      "epoch": 0.5051935577545961,
+      "grad_norm": 0.5322263240814209,
+      "learning_rate": 5.158427829208655e-05,
+      "loss": 0.929,
+      "step": 5654
+    },
+    {
+      "epoch": 0.5052829092858578,
+      "grad_norm": 0.4086034893989563,
+      "learning_rate": 5.1569814774914916e-05,
+      "loss": 0.9369,
+      "step": 5655
+    },
+    {
+      "epoch": 0.5053722608171197,
+      "grad_norm": 0.44374528527259827,
+      "learning_rate": 5.155535112625456e-05,
+      "loss": 0.9484,
+      "step": 5656
+    },
+    {
+      "epoch": 0.5054616123483816,
+      "grad_norm": 0.495838463306427,
+      "learning_rate": 5.154088734731696e-05,
+      "loss": 0.9707,
+      "step": 5657
+    },
+    {
+      "epoch": 0.5055509638796435,
+      "grad_norm": 0.44730693101882935,
+      "learning_rate": 5.152642343931362e-05,
+      "loss": 0.956,
+      "step": 5658
+    },
+    {
+      "epoch": 0.5056403154109054,
+      "grad_norm": 0.4501069188117981,
+      "learning_rate": 5.1511959403456055e-05,
+      "loss": 0.9622,
+      "step": 5659
+    },
+    {
+      "epoch": 0.5057296669421673,
+      "grad_norm": 0.41660791635513306,
+      "learning_rate": 5.149749524095577e-05,
+      "loss": 0.9941,
+      "step": 5660
+    },
+    {
+      "epoch": 0.5058190184734291,
+      "grad_norm": 0.4883604347705841,
+      "learning_rate": 5.1483030953024305e-05,
+      "loss": 1.0128,
+      "step": 5661
+    },
+    {
+      "epoch": 0.5059083700046909,
+      "grad_norm": 0.4125032424926758,
+      "learning_rate": 5.146856654087318e-05,
+      "loss": 1.0034,
+      "step": 5662
+    },
+    {
+      "epoch": 0.5059977215359528,
+      "grad_norm": 0.4337124824523926,
+      "learning_rate": 5.145410200571395e-05,
+      "loss": 0.9983,
+      "step": 5663
+    },
+    {
+      "epoch": 0.5060870730672147,
+      "grad_norm": 0.42638298869132996,
+      "learning_rate": 5.1439637348758175e-05,
+      "loss": 0.9504,
+      "step": 5664
+    },
+    {
+      "epoch": 0.5061764245984766,
+      "grad_norm": 0.5545623302459717,
+      "learning_rate": 5.142517257121744e-05,
+      "loss": 0.9618,
+      "step": 5665
+    },
+    {
+      "epoch": 0.5062657761297384,
+      "grad_norm": 0.3784050941467285,
+      "learning_rate": 5.14107076743033e-05,
+      "loss": 1.0023,
+      "step": 5666
+    },
+    {
+      "epoch": 0.5063551276610003,
+      "grad_norm": 0.5684899687767029,
+      "learning_rate": 5.139624265922737e-05,
+      "loss": 0.9455,
+      "step": 5667
+    },
+    {
+      "epoch": 0.5064444791922622,
+      "grad_norm": 0.5042484402656555,
+      "learning_rate": 5.138177752720122e-05,
+      "loss": 0.9237,
+      "step": 5668
+    },
+    {
+      "epoch": 0.506533830723524,
+      "grad_norm": 0.4414721727371216,
+      "learning_rate": 5.1367312279436486e-05,
+      "loss": 0.9426,
+      "step": 5669
+    },
+    {
+      "epoch": 0.5066231822547859,
+      "grad_norm": 0.5522531867027283,
+      "learning_rate": 5.135284691714477e-05,
+      "loss": 0.9738,
+      "step": 5670
+    },
+    {
+      "epoch": 0.5067125337860477,
+      "grad_norm": 0.45710471272468567,
+      "learning_rate": 5.13383814415377e-05,
+      "loss": 0.9844,
+      "step": 5671
+    },
+    {
+      "epoch": 0.5068018853173096,
+      "grad_norm": 0.47510063648223877,
+      "learning_rate": 5.132391585382692e-05,
+      "loss": 1.033,
+      "step": 5672
+    },
+    {
+      "epoch": 0.5068912368485715,
+      "grad_norm": 0.4167405366897583,
+      "learning_rate": 5.130945015522407e-05,
+      "loss": 0.9195,
+      "step": 5673
+    },
+    {
+      "epoch": 0.5069805883798334,
+      "grad_norm": 0.4073796570301056,
+      "learning_rate": 5.129498434694081e-05,
+      "loss": 0.9487,
+      "step": 5674
+    },
+    {
+      "epoch": 0.5070699399110953,
+      "grad_norm": 0.37733787298202515,
+      "learning_rate": 5.128051843018882e-05,
+      "loss": 0.9742,
+      "step": 5675
+    },
+    {
+      "epoch": 0.507159291442357,
+      "grad_norm": 0.3933600187301636,
+      "learning_rate": 5.126605240617975e-05,
+      "loss": 0.9361,
+      "step": 5676
+    },
+    {
+      "epoch": 0.5072486429736189,
+      "grad_norm": 0.41629257798194885,
+      "learning_rate": 5.1251586276125305e-05,
+      "loss": 0.9511,
+      "step": 5677
+    },
+    {
+      "epoch": 0.5073379945048808,
+      "grad_norm": 0.47717419266700745,
+      "learning_rate": 5.123712004123716e-05,
+      "loss": 0.9537,
+      "step": 5678
+    },
+    {
+      "epoch": 0.5074273460361427,
+      "grad_norm": 0.5077764391899109,
+      "learning_rate": 5.122265370272703e-05,
+      "loss": 0.9123,
+      "step": 5679
+    },
+    {
+      "epoch": 0.5075166975674046,
+      "grad_norm": 0.4773484170436859,
+      "learning_rate": 5.1208187261806615e-05,
+      "loss": 0.9243,
+      "step": 5680
+    },
+    {
+      "epoch": 0.5076060490986665,
+      "grad_norm": 0.46315327286720276,
+      "learning_rate": 5.119372071968764e-05,
+      "loss": 0.962,
+      "step": 5681
+    },
+    {
+      "epoch": 0.5076954006299284,
+      "grad_norm": 0.4640251100063324,
+      "learning_rate": 5.117925407758184e-05,
+      "loss": 0.9902,
+      "step": 5682
+    },
+    {
+      "epoch": 0.5077847521611901,
+      "grad_norm": 0.5840420722961426,
+      "learning_rate": 5.116478733670092e-05,
+      "loss": 0.9444,
+      "step": 5683
+    },
+    {
+      "epoch": 0.507874103692452,
+      "grad_norm": 0.46668434143066406,
+      "learning_rate": 5.115032049825667e-05,
+      "loss": 0.9198,
+      "step": 5684
+    },
+    {
+      "epoch": 0.5079634552237139,
+      "grad_norm": 0.4504847526550293,
+      "learning_rate": 5.113585356346081e-05,
+      "loss": 1.0119,
+      "step": 5685
+    },
+    {
+      "epoch": 0.5080528067549758,
+      "grad_norm": 0.531743586063385,
+      "learning_rate": 5.1121386533525105e-05,
+      "loss": 0.8559,
+      "step": 5686
+    },
+    {
+      "epoch": 0.5081421582862377,
+      "grad_norm": 0.47331151366233826,
+      "learning_rate": 5.110691940966133e-05,
+      "loss": 0.9896,
+      "step": 5687
+    },
+    {
+      "epoch": 0.5082315098174995,
+      "grad_norm": 0.42783498764038086,
+      "learning_rate": 5.109245219308124e-05,
+      "loss": 0.9287,
+      "step": 5688
+    },
+    {
+      "epoch": 0.5083208613487613,
+      "grad_norm": 0.47757160663604736,
+      "learning_rate": 5.107798488499664e-05,
+      "loss": 1.0162,
+      "step": 5689
+    },
+    {
+      "epoch": 0.5084102128800232,
+      "grad_norm": 0.44240617752075195,
+      "learning_rate": 5.106351748661932e-05,
+      "loss": 0.9406,
+      "step": 5690
+    },
+    {
+      "epoch": 0.5084995644112851,
+      "grad_norm": 0.4076976776123047,
+      "learning_rate": 5.104904999916108e-05,
+      "loss": 1.0159,
+      "step": 5691
+    },
+    {
+      "epoch": 0.508588915942547,
+      "grad_norm": 0.45423173904418945,
+      "learning_rate": 5.103458242383371e-05,
+      "loss": 0.948,
+      "step": 5692
+    },
+    {
+      "epoch": 0.5086782674738088,
+      "grad_norm": 0.42676976323127747,
+      "learning_rate": 5.102011476184903e-05,
+      "loss": 0.9556,
+      "step": 5693
+    },
+    {
+      "epoch": 0.5087676190050707,
+      "grad_norm": 0.4629276990890503,
+      "learning_rate": 5.1005647014418866e-05,
+      "loss": 0.9159,
+      "step": 5694
+    },
+    {
+      "epoch": 0.5088569705363326,
+      "grad_norm": 0.47380349040031433,
+      "learning_rate": 5.0991179182755046e-05,
+      "loss": 0.9767,
+      "step": 5695
+    },
+    {
+      "epoch": 0.5089463220675944,
+      "grad_norm": 0.563545823097229,
+      "learning_rate": 5.097671126806942e-05,
+      "loss": 0.9585,
+      "step": 5696
+    },
+    {
+      "epoch": 0.5090356735988563,
+      "grad_norm": 0.48258113861083984,
+      "learning_rate": 5.0962243271573795e-05,
+      "loss": 0.9296,
+      "step": 5697
+    },
+    {
+      "epoch": 0.5091250251301181,
+      "grad_norm": 0.42884641885757446,
+      "learning_rate": 5.094777519448005e-05,
+      "loss": 0.9912,
+      "step": 5698
+    },
+    {
+      "epoch": 0.50921437666138,
+      "grad_norm": 0.47126665711402893,
+      "learning_rate": 5.093330703800002e-05,
+      "loss": 1.0121,
+      "step": 5699
+    },
+    {
+      "epoch": 0.5093037281926419,
+      "grad_norm": 0.4382041096687317,
+      "learning_rate": 5.091883880334558e-05,
+      "loss": 0.9886,
+      "step": 5700
+    },
+    {
+      "epoch": 0.5093930797239038,
+      "grad_norm": 0.4659852087497711,
+      "learning_rate": 5.090437049172862e-05,
+      "loss": 0.9174,
+      "step": 5701
+    },
+    {
+      "epoch": 0.5094824312551657,
+      "grad_norm": 0.5495339632034302,
+      "learning_rate": 5.0889902104360965e-05,
+      "loss": 0.9337,
+      "step": 5702
+    },
+    {
+      "epoch": 0.5095717827864275,
+      "grad_norm": 0.42241787910461426,
+      "learning_rate": 5.087543364245454e-05,
+      "loss": 0.9787,
+      "step": 5703
+    },
+    {
+      "epoch": 0.5096611343176893,
+      "grad_norm": 0.47585710883140564,
+      "learning_rate": 5.08609651072212e-05,
+      "loss": 0.9195,
+      "step": 5704
+    },
+    {
+      "epoch": 0.5097504858489512,
+      "grad_norm": 0.49689507484436035,
+      "learning_rate": 5.084649649987285e-05,
+      "loss": 0.9603,
+      "step": 5705
+    },
+    {
+      "epoch": 0.5098398373802131,
+      "grad_norm": 0.4226880669593811,
+      "learning_rate": 5.083202782162142e-05,
+      "loss": 1.0116,
+      "step": 5706
+    },
+    {
+      "epoch": 0.509929188911475,
+      "grad_norm": 0.4046030640602112,
+      "learning_rate": 5.0817559073678775e-05,
+      "loss": 0.9555,
+      "step": 5707
+    },
+    {
+      "epoch": 0.5100185404427369,
+      "grad_norm": 0.4440669119358063,
+      "learning_rate": 5.080309025725685e-05,
+      "loss": 0.9408,
+      "step": 5708
+    },
+    {
+      "epoch": 0.5101078919739988,
+      "grad_norm": 0.5712856650352478,
+      "learning_rate": 5.0788621373567545e-05,
+      "loss": 0.922,
+      "step": 5709
+    },
+    {
+      "epoch": 0.5101972435052605,
+      "grad_norm": 0.39347535371780396,
+      "learning_rate": 5.0774152423822796e-05,
+      "loss": 1.0552,
+      "step": 5710
+    },
+    {
+      "epoch": 0.5102865950365224,
+      "grad_norm": 0.45367705821990967,
+      "learning_rate": 5.075968340923454e-05,
+      "loss": 0.9869,
+      "step": 5711
+    },
+    {
+      "epoch": 0.5103759465677843,
+      "grad_norm": 0.48188331723213196,
+      "learning_rate": 5.074521433101469e-05,
+      "loss": 0.9637,
+      "step": 5712
+    },
+    {
+      "epoch": 0.5104652980990462,
+      "grad_norm": 0.4748827815055847,
+      "learning_rate": 5.07307451903752e-05,
+      "loss": 0.9533,
+      "step": 5713
+    },
+    {
+      "epoch": 0.510554649630308,
+      "grad_norm": 0.4055231809616089,
+      "learning_rate": 5.0716275988528005e-05,
+      "loss": 0.9954,
+      "step": 5714
+    },
+    {
+      "epoch": 0.5106440011615699,
+      "grad_norm": 0.4364885091781616,
+      "learning_rate": 5.0701806726685076e-05,
+      "loss": 0.9841,
+      "step": 5715
+    },
+    {
+      "epoch": 0.5107333526928318,
+      "grad_norm": 0.3934551775455475,
+      "learning_rate": 5.0687337406058345e-05,
+      "loss": 0.9707,
+      "step": 5716
+    },
+    {
+      "epoch": 0.5108227042240936,
+      "grad_norm": 0.505912721157074,
+      "learning_rate": 5.0672868027859774e-05,
+      "loss": 0.9605,
+      "step": 5717
+    },
+    {
+      "epoch": 0.5109120557553555,
+      "grad_norm": 0.4747815728187561,
+      "learning_rate": 5.065839859330134e-05,
+      "loss": 0.9746,
+      "step": 5718
+    },
+    {
+      "epoch": 0.5110014072866174,
+      "grad_norm": 0.5368762016296387,
+      "learning_rate": 5.0643929103595e-05,
+      "loss": 0.9514,
+      "step": 5719
+    },
+    {
+      "epoch": 0.5110907588178792,
+      "grad_norm": 0.45678213238716125,
+      "learning_rate": 5.0629459559952754e-05,
+      "loss": 0.9288,
+      "step": 5720
+    },
+    {
+      "epoch": 0.5111801103491411,
+      "grad_norm": 0.407548725605011,
+      "learning_rate": 5.0614989963586536e-05,
+      "loss": 0.9745,
+      "step": 5721
+    },
+    {
+      "epoch": 0.511269461880403,
+      "grad_norm": 0.4418816864490509,
+      "learning_rate": 5.060052031570837e-05,
+      "loss": 0.9858,
+      "step": 5722
+    },
+    {
+      "epoch": 0.5113588134116649,
+      "grad_norm": 0.4085575342178345,
+      "learning_rate": 5.0586050617530214e-05,
+      "loss": 1.0397,
+      "step": 5723
+    },
+    {
+      "epoch": 0.5114481649429267,
+      "grad_norm": 0.4091111719608307,
+      "learning_rate": 5.057158087026408e-05,
+      "loss": 1.0209,
+      "step": 5724
+    },
+    {
+      "epoch": 0.5115375164741885,
+      "grad_norm": 0.42167025804519653,
+      "learning_rate": 5.0557111075121944e-05,
+      "loss": 0.9675,
+      "step": 5725
+    },
+    {
+      "epoch": 0.5116268680054504,
+      "grad_norm": 0.5113321542739868,
+      "learning_rate": 5.0542641233315834e-05,
+      "loss": 0.8551,
+      "step": 5726
+    },
+    {
+      "epoch": 0.5117162195367123,
+      "grad_norm": 0.4538382291793823,
+      "learning_rate": 5.0528171346057725e-05,
+      "loss": 0.9376,
+      "step": 5727
+    },
+    {
+      "epoch": 0.5118055710679742,
+      "grad_norm": 0.4386774003505707,
+      "learning_rate": 5.0513701414559635e-05,
+      "loss": 0.9178,
+      "step": 5728
+    },
+    {
+      "epoch": 0.5118949225992361,
+      "grad_norm": 0.45510563254356384,
+      "learning_rate": 5.049923144003358e-05,
+      "loss": 0.9116,
+      "step": 5729
+    },
+    {
+      "epoch": 0.511984274130498,
+      "grad_norm": 0.39837151765823364,
+      "learning_rate": 5.048476142369156e-05,
+      "loss": 0.9675,
+      "step": 5730
+    },
+    {
+      "epoch": 0.5120736256617597,
+      "grad_norm": 0.5683768391609192,
+      "learning_rate": 5.047029136674563e-05,
+      "loss": 0.9143,
+      "step": 5731
+    },
+    {
+      "epoch": 0.5121629771930216,
+      "grad_norm": 0.44197937846183777,
+      "learning_rate": 5.0455821270407763e-05,
+      "loss": 0.9469,
+      "step": 5732
+    },
+    {
+      "epoch": 0.5122523287242835,
+      "grad_norm": 0.4030378758907318,
+      "learning_rate": 5.0441351135890004e-05,
+      "loss": 0.9871,
+      "step": 5733
+    },
+    {
+      "epoch": 0.5123416802555454,
+      "grad_norm": 0.40224945545196533,
+      "learning_rate": 5.042688096440439e-05,
+      "loss": 0.9674,
+      "step": 5734
+    },
+    {
+      "epoch": 0.5124310317868073,
+      "grad_norm": 0.43630731105804443,
+      "learning_rate": 5.041241075716294e-05,
+      "loss": 0.9477,
+      "step": 5735
+    },
+    {
+      "epoch": 0.5125203833180692,
+      "grad_norm": 0.4436815679073334,
+      "learning_rate": 5.039794051537771e-05,
+      "loss": 0.9675,
+      "step": 5736
+    },
+    {
+      "epoch": 0.512609734849331,
+      "grad_norm": 0.45859208703041077,
+      "learning_rate": 5.0383470240260713e-05,
+      "loss": 0.9066,
+      "step": 5737
+    },
+    {
+      "epoch": 0.5126990863805928,
+      "grad_norm": 0.4863572120666504,
+      "learning_rate": 5.0368999933024e-05,
+      "loss": 0.9758,
+      "step": 5738
+    },
+    {
+      "epoch": 0.5127884379118547,
+      "grad_norm": 0.4681277871131897,
+      "learning_rate": 5.035452959487959e-05,
+      "loss": 0.9519,
+      "step": 5739
+    },
+    {
+      "epoch": 0.5128777894431166,
+      "grad_norm": 0.5145890712738037,
+      "learning_rate": 5.034005922703956e-05,
+      "loss": 0.8922,
+      "step": 5740
+    },
+    {
+      "epoch": 0.5129671409743785,
+      "grad_norm": 0.46464183926582336,
+      "learning_rate": 5.032558883071594e-05,
+      "loss": 0.9765,
+      "step": 5741
+    },
+    {
+      "epoch": 0.5130564925056403,
+      "grad_norm": 0.4565260410308838,
+      "learning_rate": 5.031111840712079e-05,
+      "loss": 0.9139,
+      "step": 5742
+    },
+    {
+      "epoch": 0.5131458440369022,
+      "grad_norm": 0.4922895133495331,
+      "learning_rate": 5.029664795746616e-05,
+      "loss": 0.8892,
+      "step": 5743
+    },
+    {
+      "epoch": 0.5132351955681641,
+      "grad_norm": 0.4101276993751526,
+      "learning_rate": 5.028217748296409e-05,
+      "loss": 1.0292,
+      "step": 5744
+    },
+    {
+      "epoch": 0.5133245470994259,
+      "grad_norm": 0.44919613003730774,
+      "learning_rate": 5.0267706984826644e-05,
+      "loss": 0.9039,
+      "step": 5745
+    },
+    {
+      "epoch": 0.5134138986306878,
+      "grad_norm": 0.403489887714386,
+      "learning_rate": 5.025323646426591e-05,
+      "loss": 0.9693,
+      "step": 5746
+    },
+    {
+      "epoch": 0.5135032501619496,
+      "grad_norm": 0.4634898602962494,
+      "learning_rate": 5.02387659224939e-05,
+      "loss": 0.9674,
+      "step": 5747
+    },
+    {
+      "epoch": 0.5135926016932115,
+      "grad_norm": 0.5770649909973145,
+      "learning_rate": 5.022429536072271e-05,
+      "loss": 0.9314,
+      "step": 5748
+    },
+    {
+      "epoch": 0.5136819532244734,
+      "grad_norm": 0.36833301186561584,
+      "learning_rate": 5.0209824780164386e-05,
+      "loss": 1.0579,
+      "step": 5749
+    },
+    {
+      "epoch": 0.5137713047557353,
+      "grad_norm": 0.44532787799835205,
+      "learning_rate": 5.0195354182031e-05,
+      "loss": 1.0662,
+      "step": 5750
+    },
+    {
+      "epoch": 0.5138606562869971,
+      "grad_norm": 0.40112441778182983,
+      "learning_rate": 5.018088356753463e-05,
+      "loss": 0.9897,
+      "step": 5751
+    },
+    {
+      "epoch": 0.513950007818259,
+      "grad_norm": 0.3955027759075165,
+      "learning_rate": 5.016641293788732e-05,
+      "loss": 1.0451,
+      "step": 5752
+    },
+    {
+      "epoch": 0.5140393593495208,
+      "grad_norm": 0.4612029194831848,
+      "learning_rate": 5.0151942294301155e-05,
+      "loss": 0.951,
+      "step": 5753
+    },
+    {
+      "epoch": 0.5141287108807827,
+      "grad_norm": 0.5001698136329651,
+      "learning_rate": 5.013747163798821e-05,
+      "loss": 0.9077,
+      "step": 5754
+    },
+    {
+      "epoch": 0.5142180624120446,
+      "grad_norm": 0.46220487356185913,
+      "learning_rate": 5.012300097016055e-05,
+      "loss": 0.9542,
+      "step": 5755
+    },
+    {
+      "epoch": 0.5143074139433065,
+      "grad_norm": 0.5143389701843262,
+      "learning_rate": 5.010853029203024e-05,
+      "loss": 0.9437,
+      "step": 5756
+    },
+    {
+      "epoch": 0.5143967654745684,
+      "grad_norm": 0.45577070116996765,
+      "learning_rate": 5.0094059604809375e-05,
+      "loss": 0.977,
+      "step": 5757
+    },
+    {
+      "epoch": 0.5144861170058301,
+      "grad_norm": 0.4429572522640228,
+      "learning_rate": 5.0079588909710005e-05,
+      "loss": 0.9611,
+      "step": 5758
+    },
+    {
+      "epoch": 0.514575468537092,
+      "grad_norm": 0.3987106382846832,
+      "learning_rate": 5.0065118207944216e-05,
+      "loss": 0.9967,
+      "step": 5759
+    },
+    {
+      "epoch": 0.5146648200683539,
+      "grad_norm": 0.37299370765686035,
+      "learning_rate": 5.005064750072408e-05,
+      "loss": 0.9641,
+      "step": 5760
+    },
+    {
+      "epoch": 0.5147541715996158,
+      "grad_norm": 0.5281698107719421,
+      "learning_rate": 5.003617678926169e-05,
+      "loss": 0.967,
+      "step": 5761
+    },
+    {
+      "epoch": 0.5148435231308777,
+      "grad_norm": 0.5716055631637573,
+      "learning_rate": 5.0021706074769095e-05,
+      "loss": 0.8772,
+      "step": 5762
+    },
+    {
+      "epoch": 0.5149328746621396,
+      "grad_norm": 0.5003583431243896,
+      "learning_rate": 5.000723535845838e-05,
+      "loss": 1.0233,
+      "step": 5763
+    },
+    {
+      "epoch": 0.5150222261934014,
+      "grad_norm": 0.40355581045150757,
+      "learning_rate": 4.999276464154164e-05,
+      "loss": 0.9501,
+      "step": 5764
+    },
+    {
+      "epoch": 0.5151115777246632,
+      "grad_norm": 0.4532722234725952,
+      "learning_rate": 4.997829392523093e-05,
+      "loss": 0.9233,
+      "step": 5765
+    },
+    {
+      "epoch": 0.5152009292559251,
+      "grad_norm": 0.46640169620513916,
+      "learning_rate": 4.9963823210738335e-05,
+      "loss": 0.9473,
+      "step": 5766
+    },
+    {
+      "epoch": 0.515290280787187,
+      "grad_norm": 0.4269860088825226,
+      "learning_rate": 4.994935249927593e-05,
+      "loss": 0.9754,
+      "step": 5767
+    },
+    {
+      "epoch": 0.5153796323184489,
+      "grad_norm": 0.45868611335754395,
+      "learning_rate": 4.9934881792055796e-05,
+      "loss": 1.0073,
+      "step": 5768
+    },
+    {
+      "epoch": 0.5154689838497107,
+      "grad_norm": 0.4609023332595825,
+      "learning_rate": 4.9920411090290006e-05,
+      "loss": 0.9759,
+      "step": 5769
+    },
+    {
+      "epoch": 0.5155583353809726,
+      "grad_norm": 0.4061686098575592,
+      "learning_rate": 4.990594039519063e-05,
+      "loss": 0.9157,
+      "step": 5770
+    },
+    {
+      "epoch": 0.5156476869122345,
+      "grad_norm": 0.4474005103111267,
+      "learning_rate": 4.9891469707969765e-05,
+      "loss": 0.9675,
+      "step": 5771
+    },
+    {
+      "epoch": 0.5157370384434963,
+      "grad_norm": 0.41910743713378906,
+      "learning_rate": 4.987699902983946e-05,
+      "loss": 0.8943,
+      "step": 5772
+    },
+    {
+      "epoch": 0.5158263899747582,
+      "grad_norm": 0.5283576250076294,
+      "learning_rate": 4.9862528362011794e-05,
+      "loss": 0.8617,
+      "step": 5773
+    },
+    {
+      "epoch": 0.51591574150602,
+      "grad_norm": 0.6106252670288086,
+      "learning_rate": 4.9848057705698856e-05,
+      "loss": 0.9183,
+      "step": 5774
+    },
+    {
+      "epoch": 0.5160050930372819,
+      "grad_norm": 0.40918898582458496,
+      "learning_rate": 4.98335870621127e-05,
+      "loss": 0.9387,
+      "step": 5775
+    },
+    {
+      "epoch": 0.5160944445685438,
+      "grad_norm": 0.582862138748169,
+      "learning_rate": 4.981911643246539e-05,
+      "loss": 0.9047,
+      "step": 5776
+    },
+    {
+      "epoch": 0.5161837960998057,
+      "grad_norm": 0.4943019449710846,
+      "learning_rate": 4.9804645817969015e-05,
+      "loss": 0.8898,
+      "step": 5777
+    },
+    {
+      "epoch": 0.5162731476310676,
+      "grad_norm": 0.41431719064712524,
+      "learning_rate": 4.9790175219835626e-05,
+      "loss": 0.9774,
+      "step": 5778
+    },
+    {
+      "epoch": 0.5163624991623293,
+      "grad_norm": 0.5077488422393799,
+      "learning_rate": 4.97757046392773e-05,
+      "loss": 0.985,
+      "step": 5779
+    },
+    {
+      "epoch": 0.5164518506935912,
+      "grad_norm": 0.5016276240348816,
+      "learning_rate": 4.976123407750611e-05,
+      "loss": 0.9377,
+      "step": 5780
+    },
+    {
+      "epoch": 0.5165412022248531,
+      "grad_norm": 0.3875533938407898,
+      "learning_rate": 4.9746763535734104e-05,
+      "loss": 1.008,
+      "step": 5781
+    },
+    {
+      "epoch": 0.516630553756115,
+      "grad_norm": 0.5634098649024963,
+      "learning_rate": 4.973229301517335e-05,
+      "loss": 0.9303,
+      "step": 5782
+    },
+    {
+      "epoch": 0.5167199052873769,
+      "grad_norm": 0.446728378534317,
+      "learning_rate": 4.971782251703591e-05,
+      "loss": 0.9349,
+      "step": 5783
+    },
+    {
+      "epoch": 0.5168092568186388,
+      "grad_norm": 0.3732495903968811,
+      "learning_rate": 4.9703352042533866e-05,
+      "loss": 0.9877,
+      "step": 5784
+    },
+    {
+      "epoch": 0.5168986083499006,
+      "grad_norm": 0.4483281075954437,
+      "learning_rate": 4.9688881592879233e-05,
+      "loss": 0.875,
+      "step": 5785
+    },
+    {
+      "epoch": 0.5169879598811624,
+      "grad_norm": 0.4210112392902374,
+      "learning_rate": 4.967441116928407e-05,
+      "loss": 0.9735,
+      "step": 5786
+    },
+    {
+      "epoch": 0.5170773114124243,
+      "grad_norm": 0.40223428606987,
+      "learning_rate": 4.9659940772960456e-05,
+      "loss": 0.946,
+      "step": 5787
+    },
+    {
+      "epoch": 0.5171666629436862,
+      "grad_norm": 0.5056480169296265,
+      "learning_rate": 4.964547040512042e-05,
+      "loss": 0.8455,
+      "step": 5788
+    },
+    {
+      "epoch": 0.5172560144749481,
+      "grad_norm": 0.4198668599128723,
+      "learning_rate": 4.963100006697602e-05,
+      "loss": 1.0444,
+      "step": 5789
+    },
+    {
+      "epoch": 0.51734536600621,
+      "grad_norm": 0.4606395363807678,
+      "learning_rate": 4.9616529759739305e-05,
+      "loss": 0.9313,
+      "step": 5790
+    },
+    {
+      "epoch": 0.5174347175374718,
+      "grad_norm": 0.47824013233184814,
+      "learning_rate": 4.96020594846223e-05,
+      "loss": 0.9325,
+      "step": 5791
+    },
+    {
+      "epoch": 0.5175240690687337,
+      "grad_norm": 0.46849849820137024,
+      "learning_rate": 4.958758924283706e-05,
+      "loss": 0.8359,
+      "step": 5792
+    },
+    {
+      "epoch": 0.5176134205999955,
+      "grad_norm": 0.39913827180862427,
+      "learning_rate": 4.957311903559561e-05,
+      "loss": 0.9767,
+      "step": 5793
+    },
+    {
+      "epoch": 0.5177027721312574,
+      "grad_norm": 0.5645284652709961,
+      "learning_rate": 4.9558648864110014e-05,
+      "loss": 0.8838,
+      "step": 5794
+    },
+    {
+      "epoch": 0.5177921236625193,
+      "grad_norm": 0.4400821924209595,
+      "learning_rate": 4.954417872959226e-05,
+      "loss": 0.9193,
+      "step": 5795
+    },
+    {
+      "epoch": 0.5178814751937811,
+      "grad_norm": 0.4306231439113617,
+      "learning_rate": 4.95297086332544e-05,
+      "loss": 0.9235,
+      "step": 5796
+    },
+    {
+      "epoch": 0.517970826725043,
+      "grad_norm": 0.5578770637512207,
+      "learning_rate": 4.951523857630845e-05,
+      "loss": 0.9793,
+      "step": 5797
+    },
+    {
+      "epoch": 0.5180601782563049,
+      "grad_norm": 0.4247242510318756,
+      "learning_rate": 4.950076855996643e-05,
+      "loss": 1.0176,
+      "step": 5798
+    },
+    {
+      "epoch": 0.5181495297875668,
+      "grad_norm": 0.5067040920257568,
+      "learning_rate": 4.9486298585440376e-05,
+      "loss": 0.9595,
+      "step": 5799
+    },
+    {
+      "epoch": 0.5182388813188286,
+      "grad_norm": 0.5062249898910522,
+      "learning_rate": 4.9471828653942286e-05,
+      "loss": 0.8926,
+      "step": 5800
+    },
+    {
+      "epoch": 0.5183282328500904,
+      "grad_norm": 0.43885526061058044,
+      "learning_rate": 4.945735876668418e-05,
+      "loss": 0.9922,
+      "step": 5801
+    },
+    {
+      "epoch": 0.5184175843813523,
+      "grad_norm": 0.36411014199256897,
+      "learning_rate": 4.9442888924878054e-05,
+      "loss": 0.9627,
+      "step": 5802
+    },
+    {
+      "epoch": 0.5185069359126142,
+      "grad_norm": 0.45766767859458923,
+      "learning_rate": 4.9428419129735925e-05,
+      "loss": 0.9285,
+      "step": 5803
+    },
+    {
+      "epoch": 0.5185962874438761,
+      "grad_norm": 0.4283965826034546,
+      "learning_rate": 4.9413949382469805e-05,
+      "loss": 1.0057,
+      "step": 5804
+    },
+    {
+      "epoch": 0.518685638975138,
+      "grad_norm": 0.4837194085121155,
+      "learning_rate": 4.939947968429165e-05,
+      "loss": 0.9082,
+      "step": 5805
+    },
+    {
+      "epoch": 0.5187749905063999,
+      "grad_norm": 0.459768146276474,
+      "learning_rate": 4.9385010036413475e-05,
+      "loss": 0.8664,
+      "step": 5806
+    },
+    {
+      "epoch": 0.5188643420376616,
+      "grad_norm": 0.39023587107658386,
+      "learning_rate": 4.9370540440047264e-05,
+      "loss": 1.0252,
+      "step": 5807
+    },
+    {
+      "epoch": 0.5189536935689235,
+      "grad_norm": 0.44945186376571655,
+      "learning_rate": 4.935607089640501e-05,
+      "loss": 0.919,
+      "step": 5808
+    },
+    {
+      "epoch": 0.5190430451001854,
+      "grad_norm": 0.4950411915779114,
+      "learning_rate": 4.934160140669867e-05,
+      "loss": 0.9833,
+      "step": 5809
+    },
+    {
+      "epoch": 0.5191323966314473,
+      "grad_norm": 0.42366206645965576,
+      "learning_rate": 4.932713197214023e-05,
+      "loss": 0.9545,
+      "step": 5810
+    },
+    {
+      "epoch": 0.5192217481627092,
+      "grad_norm": 0.5626574158668518,
+      "learning_rate": 4.931266259394167e-05,
+      "loss": 0.8383,
+      "step": 5811
+    },
+    {
+      "epoch": 0.519311099693971,
+      "grad_norm": 0.49013814330101013,
+      "learning_rate": 4.929819327331493e-05,
+      "loss": 0.9492,
+      "step": 5812
+    },
+    {
+      "epoch": 0.5194004512252328,
+      "grad_norm": 0.48132896423339844,
+      "learning_rate": 4.928372401147199e-05,
+      "loss": 0.8303,
+      "step": 5813
+    },
+    {
+      "epoch": 0.5194898027564947,
+      "grad_norm": 0.5105422735214233,
+      "learning_rate": 4.92692548096248e-05,
+      "loss": 0.9641,
+      "step": 5814
+    },
+    {
+      "epoch": 0.5195791542877566,
+      "grad_norm": 0.4313611388206482,
+      "learning_rate": 4.9254785668985335e-05,
+      "loss": 0.9816,
+      "step": 5815
+    },
+    {
+      "epoch": 0.5196685058190185,
+      "grad_norm": 0.3787804841995239,
+      "learning_rate": 4.924031659076548e-05,
+      "loss": 0.9836,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5197578573502804,
+      "grad_norm": 0.519477903842926,
+      "learning_rate": 4.9225847576177216e-05,
+      "loss": 0.9174,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5198472088815422,
+      "grad_norm": 0.40833377838134766,
+      "learning_rate": 4.921137862643247e-05,
+      "loss": 0.9769,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5199365604128041,
+      "grad_norm": 0.5591320991516113,
+      "learning_rate": 4.919690974274316e-05,
+      "loss": 1.0388,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5200259119440659,
+      "grad_norm": 0.44129371643066406,
+      "learning_rate": 4.918244092632123e-05,
+      "loss": 0.9731,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5201152634753278,
+      "grad_norm": 0.3888798654079437,
+      "learning_rate": 4.916797217837859e-05,
+      "loss": 0.9446,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5202046150065897,
+      "grad_norm": 0.4500011205673218,
+      "learning_rate": 4.915350350012714e-05,
+      "loss": 1.0109,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5202939665378515,
+      "grad_norm": 0.5037985444068909,
+      "learning_rate": 4.913903489277881e-05,
+      "loss": 0.9591,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5203833180691134,
+      "grad_norm": 0.49185624718666077,
+      "learning_rate": 4.9124566357545476e-05,
+      "loss": 1.0417,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5204726696003753,
+      "grad_norm": 0.38457438349723816,
+      "learning_rate": 4.9110097895639046e-05,
+      "loss": 1.0718,
+      "step": 5825
+    },
+    {
+      "epoch": 0.5205620211316372,
+      "grad_norm": 0.4443393051624298,
+      "learning_rate": 4.9095629508271394e-05,
+      "loss": 1.0176,
+      "step": 5826
+    },
+    {
+      "epoch": 0.520651372662899,
+      "grad_norm": 0.39499226212501526,
+      "learning_rate": 4.908116119665443e-05,
+      "loss": 0.9926,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5207407241941608,
+      "grad_norm": 0.563543975353241,
+      "learning_rate": 4.906669296199999e-05,
+      "loss": 0.87,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5208300757254227,
+      "grad_norm": 0.5549851655960083,
+      "learning_rate": 4.905222480551996e-05,
+      "loss": 0.9655,
+      "step": 5829
+    },
+    {
+      "epoch": 0.5209194272566846,
+      "grad_norm": 0.5066598653793335,
+      "learning_rate": 4.903775672842621e-05,
+      "loss": 0.968,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5210087787879465,
+      "grad_norm": 0.38776955008506775,
+      "learning_rate": 4.902328873193059e-05,
+      "loss": 1.0964,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5210981303192084,
+      "grad_norm": 0.45412904024124146,
+      "learning_rate": 4.900882081724495e-05,
+      "loss": 0.9324,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5211874818504703,
+      "grad_norm": 0.5597448945045471,
+      "learning_rate": 4.899435298558113e-05,
+      "loss": 0.9376,
+      "step": 5833
+    },
+    {
+      "epoch": 0.521276833381732,
+      "grad_norm": 0.413144588470459,
+      "learning_rate": 4.897988523815097e-05,
+      "loss": 0.9871,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5213661849129939,
+      "grad_norm": 0.44255807995796204,
+      "learning_rate": 4.8965417576166316e-05,
+      "loss": 0.9833,
+      "step": 5835
+    },
+    {
+      "epoch": 0.5214555364442558,
+      "grad_norm": 0.43993690609931946,
+      "learning_rate": 4.8950950000838945e-05,
+      "loss": 0.972,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5215448879755177,
+      "grad_norm": 0.4028208553791046,
+      "learning_rate": 4.893648251338069e-05,
+      "loss": 0.9361,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5216342395067796,
+      "grad_norm": 0.4599561095237732,
+      "learning_rate": 4.8922015115003375e-05,
+      "loss": 0.9747,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5217235910380414,
+      "grad_norm": 0.5245079398155212,
+      "learning_rate": 4.890754780691877e-05,
+      "loss": 0.8898,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5218129425693033,
+      "grad_norm": 0.5127478837966919,
+      "learning_rate": 4.8893080590338684e-05,
+      "loss": 0.9127,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5219022941005651,
+      "grad_norm": 0.37643924355506897,
+      "learning_rate": 4.887861346647491e-05,
+      "loss": 0.9822,
+      "step": 5841
+    },
+    {
+      "epoch": 0.521991645631827,
+      "grad_norm": 0.4069046974182129,
+      "learning_rate": 4.8864146436539196e-05,
+      "loss": 0.9837,
+      "step": 5842
+    },
+    {
+      "epoch": 0.5220809971630889,
+      "grad_norm": 0.4322601854801178,
+      "learning_rate": 4.884967950174333e-05,
+      "loss": 0.9865,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5221703486943507,
+      "grad_norm": 0.4148218333721161,
+      "learning_rate": 4.883521266329907e-05,
+      "loss": 0.9269,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5222597002256126,
+      "grad_norm": 0.3982529640197754,
+      "learning_rate": 4.882074592241817e-05,
+      "loss": 0.9791,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5223490517568745,
+      "grad_norm": 0.48719900846481323,
+      "learning_rate": 4.880627928031237e-05,
+      "loss": 0.9728,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5224384032881364,
+      "grad_norm": 0.5783781409263611,
+      "learning_rate": 4.87918127381934e-05,
+      "loss": 0.9212,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5225277548193982,
+      "grad_norm": 0.4238271713256836,
+      "learning_rate": 4.8777346297272986e-05,
+      "loss": 1.0067,
+      "step": 5848
+    },
+    {
+      "epoch": 0.52261710635066,
+      "grad_norm": 0.433936208486557,
+      "learning_rate": 4.876287995876285e-05,
+      "loss": 0.9611,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5227064578819219,
+      "grad_norm": 0.4108850657939911,
+      "learning_rate": 4.874841372387471e-05,
+      "loss": 0.9153,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5227958094131838,
+      "grad_norm": 0.4847879111766815,
+      "learning_rate": 4.8733947593820254e-05,
+      "loss": 0.9725,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5228851609444457,
+      "grad_norm": 0.508726954460144,
+      "learning_rate": 4.871948156981119e-05,
+      "loss": 0.8959,
+      "step": 5852
+    },
+    {
+      "epoch": 0.5229745124757076,
+      "grad_norm": 0.4954078495502472,
+      "learning_rate": 4.870501565305919e-05,
+      "loss": 0.9549,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5230638640069695,
+      "grad_norm": 0.4881262183189392,
+      "learning_rate": 4.8690549844775935e-05,
+      "loss": 1.066,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5231532155382312,
+      "grad_norm": 0.6973787546157837,
+      "learning_rate": 4.867608414617311e-05,
+      "loss": 0.9015,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5232425670694931,
+      "grad_norm": 0.4565022587776184,
+      "learning_rate": 4.8661618558462325e-05,
+      "loss": 0.9478,
+      "step": 5856
+    },
+    {
+      "epoch": 0.523331918600755,
+      "grad_norm": 0.48407623171806335,
+      "learning_rate": 4.8647153082855254e-05,
+      "loss": 0.9787,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5234212701320169,
+      "grad_norm": 0.4646012783050537,
+      "learning_rate": 4.863268772056353e-05,
+      "loss": 0.9203,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5235106216632788,
+      "grad_norm": 0.4288366436958313,
+      "learning_rate": 4.8618222472798783e-05,
+      "loss": 0.9631,
+      "step": 5859
+    },
+    {
+      "epoch": 0.5235999731945407,
+      "grad_norm": 0.5206550359725952,
+      "learning_rate": 4.860375734077265e-05,
+      "loss": 0.9404,
+      "step": 5860
+    },
+    {
+      "epoch": 0.5236893247258025,
+      "grad_norm": 0.36840370297431946,
+      "learning_rate": 4.858929232569671e-05,
+      "loss": 1.0194,
+      "step": 5861
+    },
+    {
+      "epoch": 0.5237786762570643,
+      "grad_norm": 0.4584810435771942,
+      "learning_rate": 4.857482742878257e-05,
+      "loss": 0.9383,
+      "step": 5862
+    },
+    {
+      "epoch": 0.5238680277883262,
+      "grad_norm": 0.3916776478290558,
+      "learning_rate": 4.856036265124182e-05,
+      "loss": 1.0608,
+      "step": 5863
+    },
+    {
+      "epoch": 0.5239573793195881,
+      "grad_norm": 0.4396377503871918,
+      "learning_rate": 4.854589799428606e-05,
+      "loss": 0.9877,
+      "step": 5864
+    },
+    {
+      "epoch": 0.52404673085085,
+      "grad_norm": 0.5073574781417847,
+      "learning_rate": 4.853143345912685e-05,
+      "loss": 0.8528,
+      "step": 5865
+    },
+    {
+      "epoch": 0.5241360823821118,
+      "grad_norm": 0.5877308249473572,
+      "learning_rate": 4.851696904697573e-05,
+      "loss": 0.9014,
+      "step": 5866
+    },
+    {
+      "epoch": 0.5242254339133737,
+      "grad_norm": 0.470747172832489,
+      "learning_rate": 4.8502504759044245e-05,
+      "loss": 0.922,
+      "step": 5867
+    },
+    {
+      "epoch": 0.5243147854446356,
+      "grad_norm": 0.4053327143192291,
+      "learning_rate": 4.848804059654396e-05,
+      "loss": 0.93,
+      "step": 5868
+    },
+    {
+      "epoch": 0.5244041369758974,
+      "grad_norm": 0.44682514667510986,
+      "learning_rate": 4.847357656068639e-05,
+      "loss": 1.0523,
+      "step": 5869
+    },
+    {
+      "epoch": 0.5244934885071593,
+      "grad_norm": 0.3955923914909363,
+      "learning_rate": 4.845911265268305e-05,
+      "loss": 1.0694,
+      "step": 5870
+    },
+    {
+      "epoch": 0.5245828400384211,
+      "grad_norm": 0.46006014943122864,
+      "learning_rate": 4.844464887374545e-05,
+      "loss": 0.9943,
+      "step": 5871
+    },
+    {
+      "epoch": 0.524672191569683,
+      "grad_norm": 0.4770232141017914,
+      "learning_rate": 4.8430185225085096e-05,
+      "loss": 0.9395,
+      "step": 5872
+    },
+    {
+      "epoch": 0.5247615431009449,
+      "grad_norm": 0.5369167327880859,
+      "learning_rate": 4.841572170791345e-05,
+      "loss": 0.9318,
+      "step": 5873
+    },
+    {
+      "epoch": 0.5248508946322068,
+      "grad_norm": 0.4109400808811188,
+      "learning_rate": 4.8401258323442016e-05,
+      "loss": 1.0272,
+      "step": 5874
+    },
+    {
+      "epoch": 0.5249402461634686,
+      "grad_norm": 0.42871788144111633,
+      "learning_rate": 4.8386795072882234e-05,
+      "loss": 0.9692,
+      "step": 5875
+    },
+    {
+      "epoch": 0.5250295976947305,
+      "grad_norm": 0.4132499098777771,
+      "learning_rate": 4.837233195744556e-05,
+      "loss": 0.9554,
+      "step": 5876
+    },
+    {
+      "epoch": 0.5251189492259923,
+      "grad_norm": 0.40216875076293945,
+      "learning_rate": 4.835786897834342e-05,
+      "loss": 0.9684,
+      "step": 5877
+    },
+    {
+      "epoch": 0.5252083007572542,
+      "grad_norm": 0.3839309513568878,
+      "learning_rate": 4.8343406136787274e-05,
+      "loss": 0.9613,
+      "step": 5878
+    },
+    {
+      "epoch": 0.5252976522885161,
+      "grad_norm": 0.4431716799736023,
+      "learning_rate": 4.832894343398851e-05,
+      "loss": 0.9824,
+      "step": 5879
+    },
+    {
+      "epoch": 0.525387003819778,
+      "grad_norm": 0.5026467442512512,
+      "learning_rate": 4.831448087115855e-05,
+      "loss": 1.0332,
+      "step": 5880
+    },
+    {
+      "epoch": 0.5254763553510399,
+      "grad_norm": 0.431949257850647,
+      "learning_rate": 4.8300018449508775e-05,
+      "loss": 0.9471,
+      "step": 5881
+    },
+    {
+      "epoch": 0.5255657068823016,
+      "grad_norm": 0.5039017796516418,
+      "learning_rate": 4.8285556170250584e-05,
+      "loss": 0.9319,
+      "step": 5882
+    },
+    {
+      "epoch": 0.5256550584135635,
+      "grad_norm": 0.44245362281799316,
+      "learning_rate": 4.8271094034595336e-05,
+      "loss": 0.9568,
+      "step": 5883
+    },
+    {
+      "epoch": 0.5257444099448254,
+      "grad_norm": 0.4410870671272278,
+      "learning_rate": 4.825663204375439e-05,
+      "loss": 0.9713,
+      "step": 5884
+    },
+    {
+      "epoch": 0.5258337614760873,
+      "grad_norm": 0.5179281234741211,
+      "learning_rate": 4.824217019893912e-05,
+      "loss": 0.8289,
+      "step": 5885
+    },
+    {
+      "epoch": 0.5259231130073492,
+      "grad_norm": 0.4670274555683136,
+      "learning_rate": 4.82277085013608e-05,
+      "loss": 0.9096,
+      "step": 5886
+    },
+    {
+      "epoch": 0.5260124645386111,
+      "grad_norm": 0.5021420121192932,
+      "learning_rate": 4.8213246952230794e-05,
+      "loss": 0.9992,
+      "step": 5887
+    },
+    {
+      "epoch": 0.5261018160698729,
+      "grad_norm": 0.4476109445095062,
+      "learning_rate": 4.81987855527604e-05,
+      "loss": 0.9169,
+      "step": 5888
+    },
+    {
+      "epoch": 0.5261911676011347,
+      "grad_norm": 0.4486454725265503,
+      "learning_rate": 4.818432430416091e-05,
+      "loss": 1.0424,
+      "step": 5889
+    },
+    {
+      "epoch": 0.5262805191323966,
+      "grad_norm": 0.5046273469924927,
+      "learning_rate": 4.8169863207643615e-05,
+      "loss": 0.9194,
+      "step": 5890
+    },
+    {
+      "epoch": 0.5263698706636585,
+      "grad_norm": 0.4156638979911804,
+      "learning_rate": 4.815540226441979e-05,
+      "loss": 0.9668,
+      "step": 5891
+    },
+    {
+      "epoch": 0.5264592221949204,
+      "grad_norm": 0.4498765468597412,
+      "learning_rate": 4.814094147570069e-05,
+      "loss": 0.9449,
+      "step": 5892
+    },
+    {
+      "epoch": 0.5265485737261822,
+      "grad_norm": 0.4794931411743164,
+      "learning_rate": 4.812648084269756e-05,
+      "loss": 1.0374,
+      "step": 5893
+    },
+    {
+      "epoch": 0.5266379252574441,
+      "grad_norm": 0.4733235836029053,
+      "learning_rate": 4.811202036662162e-05,
+      "loss": 0.9983,
+      "step": 5894
+    },
+    {
+      "epoch": 0.526727276788706,
+      "grad_norm": 0.46899041533470154,
+      "learning_rate": 4.8097560048684095e-05,
+      "loss": 0.9764,
+      "step": 5895
+    },
+    {
+      "epoch": 0.5268166283199678,
+      "grad_norm": 0.44023454189300537,
+      "learning_rate": 4.808309989009621e-05,
+      "loss": 0.935,
+      "step": 5896
+    },
+    {
+      "epoch": 0.5269059798512297,
+      "grad_norm": 0.43503814935684204,
+      "learning_rate": 4.806863989206914e-05,
+      "loss": 0.8848,
+      "step": 5897
+    },
+    {
+      "epoch": 0.5269953313824915,
+      "grad_norm": 0.4301520884037018,
+      "learning_rate": 4.805418005581406e-05,
+      "loss": 0.999,
+      "step": 5898
+    },
+    {
+      "epoch": 0.5270846829137534,
+      "grad_norm": 0.3994184136390686,
+      "learning_rate": 4.803972038254214e-05,
+      "loss": 0.9203,
+      "step": 5899
+    },
+    {
+      "epoch": 0.5271740344450153,
+      "grad_norm": 0.4339260756969452,
+      "learning_rate": 4.802526087346453e-05,
+      "loss": 1.0067,
+      "step": 5900
+    },
+    {
+      "epoch": 0.5272633859762772,
+      "grad_norm": 0.4898630678653717,
+      "learning_rate": 4.801080152979237e-05,
+      "loss": 0.9212,
+      "step": 5901
+    },
+    {
+      "epoch": 0.5273527375075391,
+      "grad_norm": 0.49684223532676697,
+      "learning_rate": 4.799634235273679e-05,
+      "loss": 0.9273,
+      "step": 5902
+    },
+    {
+      "epoch": 0.5274420890388009,
+      "grad_norm": 0.420746386051178,
+      "learning_rate": 4.798188334350889e-05,
+      "loss": 0.9468,
+      "step": 5903
+    },
+    {
+      "epoch": 0.5275314405700627,
+      "grad_norm": 0.4721129536628723,
+      "learning_rate": 4.7967424503319774e-05,
+      "loss": 0.9378,
+      "step": 5904
+    },
+    {
+      "epoch": 0.5276207921013246,
+      "grad_norm": 0.4775635004043579,
+      "learning_rate": 4.795296583338052e-05,
+      "loss": 0.996,
+      "step": 5905
+    },
+    {
+      "epoch": 0.5277101436325865,
+      "grad_norm": 0.4336440861225128,
+      "learning_rate": 4.793850733490222e-05,
+      "loss": 1.0088,
+      "step": 5906
+    },
+    {
+      "epoch": 0.5277994951638484,
+      "grad_norm": 0.4374808669090271,
+      "learning_rate": 4.792404900909589e-05,
+      "loss": 0.9745,
+      "step": 5907
+    },
+    {
+      "epoch": 0.5278888466951103,
+      "grad_norm": 0.4299390912055969,
+      "learning_rate": 4.7909590857172574e-05,
+      "loss": 0.9602,
+      "step": 5908
+    },
+    {
+      "epoch": 0.5279781982263722,
+      "grad_norm": 0.4343792498111725,
+      "learning_rate": 4.7895132880343306e-05,
+      "loss": 0.9381,
+      "step": 5909
+    },
+    {
+      "epoch": 0.5280675497576339,
+      "grad_norm": 0.536439836025238,
+      "learning_rate": 4.78806750798191e-05,
+      "loss": 0.8658,
+      "step": 5910
+    },
+    {
+      "epoch": 0.5281569012888958,
+      "grad_norm": 0.46042898297309875,
+      "learning_rate": 4.786621745681096e-05,
+      "loss": 0.9794,
+      "step": 5911
+    },
+    {
+      "epoch": 0.5282462528201577,
+      "grad_norm": 0.4773177206516266,
+      "learning_rate": 4.785176001252983e-05,
+      "loss": 0.9424,
+      "step": 5912
+    },
+    {
+      "epoch": 0.5283356043514196,
+      "grad_norm": 0.47220951318740845,
+      "learning_rate": 4.7837302748186705e-05,
+      "loss": 0.9412,
+      "step": 5913
+    },
+    {
+      "epoch": 0.5284249558826815,
+      "grad_norm": 0.46493634581565857,
+      "learning_rate": 4.782284566499253e-05,
+      "loss": 0.9811,
+      "step": 5914
+    },
+    {
+      "epoch": 0.5285143074139433,
+      "grad_norm": 0.5822499990463257,
+      "learning_rate": 4.780838876415824e-05,
+      "loss": 0.8253,
+      "step": 5915
+    },
+    {
+      "epoch": 0.5286036589452052,
+      "grad_norm": 0.46834489703178406,
+      "learning_rate": 4.779393204689477e-05,
+      "loss": 0.9326,
+      "step": 5916
+    },
+    {
+      "epoch": 0.528693010476467,
+      "grad_norm": 0.5302138328552246,
+      "learning_rate": 4.777947551441298e-05,
+      "loss": 0.8613,
+      "step": 5917
+    },
+    {
+      "epoch": 0.5287823620077289,
+      "grad_norm": 0.4419107735157013,
+      "learning_rate": 4.776501916792379e-05,
+      "loss": 0.9798,
+      "step": 5918
+    },
+    {
+      "epoch": 0.5288717135389908,
+      "grad_norm": 0.4270457923412323,
+      "learning_rate": 4.775056300863806e-05,
+      "loss": 1.0351,
+      "step": 5919
+    },
+    {
+      "epoch": 0.5289610650702526,
+      "grad_norm": 0.4577907621860504,
+      "learning_rate": 4.773610703776666e-05,
+      "loss": 0.9121,
+      "step": 5920
+    },
+    {
+      "epoch": 0.5290504166015145,
+      "grad_norm": 0.4939514398574829,
+      "learning_rate": 4.772165125652043e-05,
+      "loss": 0.9367,
+      "step": 5921
+    },
+    {
+      "epoch": 0.5291397681327764,
+      "grad_norm": 0.45236796140670776,
+      "learning_rate": 4.7707195666110195e-05,
+      "loss": 0.9506,
+      "step": 5922
+    },
+    {
+      "epoch": 0.5292291196640383,
+      "grad_norm": 0.4021003842353821,
+      "learning_rate": 4.7692740267746757e-05,
+      "loss": 0.9241,
+      "step": 5923
+    },
+    {
+      "epoch": 0.5293184711953001,
+      "grad_norm": 0.44145897030830383,
+      "learning_rate": 4.767828506264091e-05,
+      "loss": 0.937,
+      "step": 5924
+    },
+    {
+      "epoch": 0.529407822726562,
+      "grad_norm": 0.4989675283432007,
+      "learning_rate": 4.766383005200345e-05,
+      "loss": 1.0619,
+      "step": 5925
+    },
+    {
+      "epoch": 0.5294971742578238,
+      "grad_norm": 0.45392560958862305,
+      "learning_rate": 4.7649375237045135e-05,
+      "loss": 0.9564,
+      "step": 5926
+    },
+    {
+      "epoch": 0.5295865257890857,
+      "grad_norm": 0.40658605098724365,
+      "learning_rate": 4.763492061897669e-05,
+      "loss": 0.9595,
+      "step": 5927
+    },
+    {
+      "epoch": 0.5296758773203476,
+      "grad_norm": 0.47459685802459717,
+      "learning_rate": 4.762046619900884e-05,
+      "loss": 0.9083,
+      "step": 5928
+    },
+    {
+      "epoch": 0.5297652288516095,
+      "grad_norm": 0.5040339827537537,
+      "learning_rate": 4.7606011978352315e-05,
+      "loss": 0.9873,
+      "step": 5929
+    },
+    {
+      "epoch": 0.5298545803828714,
+      "grad_norm": 0.5010223388671875,
+      "learning_rate": 4.759155795821782e-05,
+      "loss": 0.8799,
+      "step": 5930
+    },
+    {
+      "epoch": 0.5299439319141331,
+      "grad_norm": 0.4148063063621521,
+      "learning_rate": 4.7577104139816e-05,
+      "loss": 0.9431,
+      "step": 5931
+    },
+    {
+      "epoch": 0.530033283445395,
+      "grad_norm": 0.469046413898468,
+      "learning_rate": 4.7562650524357536e-05,
+      "loss": 0.9503,
+      "step": 5932
+    },
+    {
+      "epoch": 0.5301226349766569,
+      "grad_norm": 0.5078368186950684,
+      "learning_rate": 4.754819711305308e-05,
+      "loss": 0.9369,
+      "step": 5933
+    },
+    {
+      "epoch": 0.5302119865079188,
+      "grad_norm": 0.5710771083831787,
+      "learning_rate": 4.753374390711324e-05,
+      "loss": 0.8195,
+      "step": 5934
+    },
+    {
+      "epoch": 0.5303013380391807,
+      "grad_norm": 0.4148007035255432,
+      "learning_rate": 4.751929090774864e-05,
+      "loss": 0.9669,
+      "step": 5935
+    },
+    {
+      "epoch": 0.5303906895704426,
+      "grad_norm": 0.4209470748901367,
+      "learning_rate": 4.7504838116169895e-05,
+      "loss": 0.9673,
+      "step": 5936
+    },
+    {
+      "epoch": 0.5304800411017044,
+      "grad_norm": 0.45647767186164856,
+      "learning_rate": 4.7490385533587525e-05,
+      "loss": 0.9706,
+      "step": 5937
+    },
+    {
+      "epoch": 0.5305693926329662,
+      "grad_norm": 0.4158954620361328,
+      "learning_rate": 4.7475933161212116e-05,
+      "loss": 1.0093,
+      "step": 5938
+    },
+    {
+      "epoch": 0.5306587441642281,
+      "grad_norm": 0.4575401842594147,
+      "learning_rate": 4.746148100025421e-05,
+      "loss": 1.0107,
+      "step": 5939
+    },
+    {
+      "epoch": 0.53074809569549,
+      "grad_norm": 0.4388822019100189,
+      "learning_rate": 4.7447029051924334e-05,
+      "loss": 0.9565,
+      "step": 5940
+    },
+    {
+      "epoch": 0.5308374472267519,
+      "grad_norm": 0.5215067267417908,
+      "learning_rate": 4.7432577317432984e-05,
+      "loss": 0.8885,
+      "step": 5941
+    },
+    {
+      "epoch": 0.5309267987580137,
+      "grad_norm": 0.5837841033935547,
+      "learning_rate": 4.7418125797990655e-05,
+      "loss": 0.9078,
+      "step": 5942
+    },
+    {
+      "epoch": 0.5310161502892756,
+      "grad_norm": 0.4431391656398773,
+      "learning_rate": 4.740367449480781e-05,
+      "loss": 0.9465,
+      "step": 5943
+    },
+    {
+      "epoch": 0.5311055018205374,
+      "grad_norm": 0.4966404438018799,
+      "learning_rate": 4.7389223409094904e-05,
+      "loss": 0.9855,
+      "step": 5944
+    },
+    {
+      "epoch": 0.5311948533517993,
+      "grad_norm": 0.4595504403114319,
+      "learning_rate": 4.737477254206236e-05,
+      "loss": 0.9012,
+      "step": 5945
+    },
+    {
+      "epoch": 0.5312842048830612,
+      "grad_norm": 0.4891195595264435,
+      "learning_rate": 4.736032189492062e-05,
+      "loss": 0.865,
+      "step": 5946
+    },
+    {
+      "epoch": 0.531373556414323,
+      "grad_norm": 0.38100045919418335,
+      "learning_rate": 4.734587146888003e-05,
+      "loss": 0.9545,
+      "step": 5947
+    },
+    {
+      "epoch": 0.5314629079455849,
+      "grad_norm": 0.4812198877334595,
+      "learning_rate": 4.733142126515101e-05,
+      "loss": 1.0111,
+      "step": 5948
+    },
+    {
+      "epoch": 0.5315522594768468,
+      "grad_norm": 0.5166088938713074,
+      "learning_rate": 4.7316971284943905e-05,
+      "loss": 0.9222,
+      "step": 5949
+    },
+    {
+      "epoch": 0.5316416110081087,
+      "grad_norm": 0.42743030190467834,
+      "learning_rate": 4.730252152946905e-05,
+      "loss": 0.9176,
+      "step": 5950
+    },
+    {
+      "epoch": 0.5317309625393705,
+      "grad_norm": 0.4856008291244507,
+      "learning_rate": 4.7288071999936766e-05,
+      "loss": 0.9253,
+      "step": 5951
+    },
+    {
+      "epoch": 0.5318203140706323,
+      "grad_norm": 0.40241360664367676,
+      "learning_rate": 4.7273622697557356e-05,
+      "loss": 1.0319,
+      "step": 5952
+    },
+    {
+      "epoch": 0.5319096656018942,
+      "grad_norm": 0.46790629625320435,
+      "learning_rate": 4.725917362354111e-05,
+      "loss": 0.9318,
+      "step": 5953
+    },
+    {
+      "epoch": 0.5319990171331561,
+      "grad_norm": 0.4300851821899414,
+      "learning_rate": 4.7244724779098293e-05,
+      "loss": 0.9938,
+      "step": 5954
+    },
+    {
+      "epoch": 0.532088368664418,
+      "grad_norm": 0.5142315626144409,
+      "learning_rate": 4.7230276165439136e-05,
+      "loss": 0.9487,
+      "step": 5955
+    },
+    {
+      "epoch": 0.5321777201956799,
+      "grad_norm": 0.38115739822387695,
+      "learning_rate": 4.72158277837739e-05,
+      "loss": 0.9286,
+      "step": 5956
+    },
+    {
+      "epoch": 0.5322670717269418,
+      "grad_norm": 0.7321306467056274,
+      "learning_rate": 4.720137963531274e-05,
+      "loss": 0.9289,
+      "step": 5957
+    },
+    {
+      "epoch": 0.5323564232582035,
+      "grad_norm": 0.4938461184501648,
+      "learning_rate": 4.718693172126587e-05,
+      "loss": 0.973,
+      "step": 5958
+    },
+    {
+      "epoch": 0.5324457747894654,
+      "grad_norm": 0.5693181157112122,
+      "learning_rate": 4.7172484042843454e-05,
+      "loss": 0.9181,
+      "step": 5959
+    },
+    {
+      "epoch": 0.5325351263207273,
+      "grad_norm": 0.5643350481987,
+      "learning_rate": 4.7158036601255634e-05,
+      "loss": 1.017,
+      "step": 5960
+    },
+    {
+      "epoch": 0.5326244778519892,
+      "grad_norm": 0.5340102910995483,
+      "learning_rate": 4.714358939771255e-05,
+      "loss": 0.8804,
+      "step": 5961
+    },
+    {
+      "epoch": 0.5327138293832511,
+      "grad_norm": 0.5467627048492432,
+      "learning_rate": 4.712914243342429e-05,
+      "loss": 0.8569,
+      "step": 5962
+    },
+    {
+      "epoch": 0.532803180914513,
+      "grad_norm": 0.5655173659324646,
+      "learning_rate": 4.711469570960096e-05,
+      "loss": 0.9432,
+      "step": 5963
+    },
+    {
+      "epoch": 0.5328925324457748,
+      "grad_norm": 0.41005784273147583,
+      "learning_rate": 4.7100249227452627e-05,
+      "loss": 1.0204,
+      "step": 5964
+    },
+    {
+      "epoch": 0.5329818839770366,
+      "grad_norm": 0.4703254997730255,
+      "learning_rate": 4.7085802988189315e-05,
+      "loss": 0.9528,
+      "step": 5965
+    },
+    {
+      "epoch": 0.5330712355082985,
+      "grad_norm": 0.4015927314758301,
+      "learning_rate": 4.707135699302108e-05,
+      "loss": 0.955,
+      "step": 5966
+    },
+    {
+      "epoch": 0.5331605870395604,
+      "grad_norm": 0.4751605689525604,
+      "learning_rate": 4.705691124315792e-05,
+      "loss": 0.9394,
+      "step": 5967
+    },
+    {
+      "epoch": 0.5332499385708223,
+      "grad_norm": 0.4073632061481476,
+      "learning_rate": 4.70424657398098e-05,
+      "loss": 0.943,
+      "step": 5968
+    },
+    {
+      "epoch": 0.5333392901020841,
+      "grad_norm": 0.43543925881385803,
+      "learning_rate": 4.70280204841867e-05,
+      "loss": 0.9864,
+      "step": 5969
+    },
+    {
+      "epoch": 0.533428641633346,
+      "grad_norm": 0.4593610465526581,
+      "learning_rate": 4.701357547749856e-05,
+      "loss": 0.9586,
+      "step": 5970
+    },
+    {
+      "epoch": 0.5335179931646079,
+      "grad_norm": 0.4604717493057251,
+      "learning_rate": 4.699913072095531e-05,
+      "loss": 0.9719,
+      "step": 5971
+    },
+    {
+      "epoch": 0.5336073446958697,
+      "grad_norm": 0.522756814956665,
+      "learning_rate": 4.698468621576685e-05,
+      "loss": 0.9178,
+      "step": 5972
+    },
+    {
+      "epoch": 0.5336966962271316,
+      "grad_norm": 0.43565860390663147,
+      "learning_rate": 4.697024196314305e-05,
+      "loss": 0.9208,
+      "step": 5973
+    },
+    {
+      "epoch": 0.5337860477583934,
+      "grad_norm": 0.45407941937446594,
+      "learning_rate": 4.695579796429379e-05,
+      "loss": 0.9655,
+      "step": 5974
+    },
+    {
+      "epoch": 0.5338753992896553,
+      "grad_norm": 0.524958074092865,
+      "learning_rate": 4.69413542204289e-05,
+      "loss": 0.9391,
+      "step": 5975
+    },
+    {
+      "epoch": 0.5339647508209172,
+      "grad_norm": 0.4323229193687439,
+      "learning_rate": 4.6926910732758215e-05,
+      "loss": 0.937,
+      "step": 5976
+    },
+    {
+      "epoch": 0.5340541023521791,
+      "grad_norm": 0.48769766092300415,
+      "learning_rate": 4.69124675024915e-05,
+      "loss": 0.9422,
+      "step": 5977
+    },
+    {
+      "epoch": 0.534143453883441,
+      "grad_norm": 0.47193393111228943,
+      "learning_rate": 4.689802453083853e-05,
+      "loss": 0.9762,
+      "step": 5978
+    },
+    {
+      "epoch": 0.5342328054147027,
+      "grad_norm": 0.574012041091919,
+      "learning_rate": 4.688358181900907e-05,
+      "loss": 0.867,
+      "step": 5979
+    },
+    {
+      "epoch": 0.5343221569459646,
+      "grad_norm": 0.43543365597724915,
+      "learning_rate": 4.686913936821287e-05,
+      "loss": 0.9602,
+      "step": 5980
+    },
+    {
+      "epoch": 0.5344115084772265,
+      "grad_norm": 0.45189157128334045,
+      "learning_rate": 4.6854697179659614e-05,
+      "loss": 0.9971,
+      "step": 5981
+    },
+    {
+      "epoch": 0.5345008600084884,
+      "grad_norm": 0.4232499599456787,
+      "learning_rate": 4.684025525455899e-05,
+      "loss": 1.0099,
+      "step": 5982
+    },
+    {
+      "epoch": 0.5345902115397503,
+      "grad_norm": 0.42236101627349854,
+      "learning_rate": 4.682581359412066e-05,
+      "loss": 0.9674,
+      "step": 5983
+    },
+    {
+      "epoch": 0.5346795630710122,
+      "grad_norm": 0.4021371603012085,
+      "learning_rate": 4.681137219955429e-05,
+      "loss": 0.932,
+      "step": 5984
+    },
+    {
+      "epoch": 0.534768914602274,
+      "grad_norm": 0.39467480778694153,
+      "learning_rate": 4.6796931072069484e-05,
+      "loss": 0.9422,
+      "step": 5985
+    },
+    {
+      "epoch": 0.5348582661335358,
+      "grad_norm": 0.40237951278686523,
+      "learning_rate": 4.678249021287583e-05,
+      "loss": 0.9538,
+      "step": 5986
+    },
+    {
+      "epoch": 0.5349476176647977,
+      "grad_norm": 0.5124891996383667,
+      "learning_rate": 4.6768049623182953e-05,
+      "loss": 0.9768,
+      "step": 5987
+    },
+    {
+      "epoch": 0.5350369691960596,
+      "grad_norm": 0.4601878225803375,
+      "learning_rate": 4.675360930420035e-05,
+      "loss": 0.9826,
+      "step": 5988
+    },
+    {
+      "epoch": 0.5351263207273215,
+      "grad_norm": 0.4513419568538666,
+      "learning_rate": 4.673916925713756e-05,
+      "loss": 0.9567,
+      "step": 5989
+    },
+    {
+      "epoch": 0.5352156722585834,
+      "grad_norm": 0.38111451268196106,
+      "learning_rate": 4.672472948320411e-05,
+      "loss": 1.0469,
+      "step": 5990
+    },
+    {
+      "epoch": 0.5353050237898452,
+      "grad_norm": 0.5757367610931396,
+      "learning_rate": 4.671028998360947e-05,
+      "loss": 0.9822,
+      "step": 5991
+    },
+    {
+      "epoch": 0.5353943753211071,
+      "grad_norm": 0.5025779008865356,
+      "learning_rate": 4.669585075956312e-05,
+      "loss": 0.9309,
+      "step": 5992
+    },
+    {
+      "epoch": 0.5354837268523689,
+      "grad_norm": 0.4102911651134491,
+      "learning_rate": 4.668141181227448e-05,
+      "loss": 0.9255,
+      "step": 5993
+    },
+    {
+      "epoch": 0.5355730783836308,
+      "grad_norm": 0.4956420958042145,
+      "learning_rate": 4.666697314295298e-05,
+      "loss": 0.8852,
+      "step": 5994
+    },
+    {
+      "epoch": 0.5356624299148927,
+      "grad_norm": 0.4618532061576843,
+      "learning_rate": 4.665253475280801e-05,
+      "loss": 0.9496,
+      "step": 5995
+    },
+    {
+      "epoch": 0.5357517814461545,
+      "grad_norm": 0.5225943922996521,
+      "learning_rate": 4.663809664304894e-05,
+      "loss": 0.9285,
+      "step": 5996
+    },
+    {
+      "epoch": 0.5358411329774164,
+      "grad_norm": 0.5119801759719849,
+      "learning_rate": 4.662365881488511e-05,
+      "loss": 0.9436,
+      "step": 5997
+    },
+    {
+      "epoch": 0.5359304845086783,
+      "grad_norm": 0.5475390553474426,
+      "learning_rate": 4.6609221269525835e-05,
+      "loss": 0.9381,
+      "step": 5998
+    },
+    {
+      "epoch": 0.5360198360399402,
+      "grad_norm": 0.48015183210372925,
+      "learning_rate": 4.659478400818043e-05,
+      "loss": 0.9396,
+      "step": 5999
+    },
+    {
+      "epoch": 0.536109187571202,
+      "grad_norm": 0.5057094693183899,
+      "learning_rate": 4.658034703205816e-05,
+      "loss": 0.9935,
+      "step": 6000
+    },
+    {
+      "epoch": 0.5361985391024638,
+      "grad_norm": 0.48920387029647827,
+      "learning_rate": 4.6565910342368266e-05,
+      "loss": 0.9402,
+      "step": 6001
+    },
+    {
+      "epoch": 0.5362878906337257,
+      "grad_norm": 0.4365025758743286,
+      "learning_rate": 4.6551473940319995e-05,
+      "loss": 0.9472,
+      "step": 6002
+    },
+    {
+      "epoch": 0.5363772421649876,
+      "grad_norm": 0.408589243888855,
+      "learning_rate": 4.6537037827122536e-05,
+      "loss": 0.9435,
+      "step": 6003
+    },
+    {
+      "epoch": 0.5364665936962495,
+      "grad_norm": 0.5231300592422485,
+      "learning_rate": 4.652260200398507e-05,
+      "loss": 0.9997,
+      "step": 6004
+    },
+    {
+      "epoch": 0.5365559452275114,
+      "grad_norm": 0.435147225856781,
+      "learning_rate": 4.6508166472116754e-05,
+      "loss": 0.9694,
+      "step": 6005
+    },
+    {
+      "epoch": 0.5366452967587731,
+      "grad_norm": 0.42764097452163696,
+      "learning_rate": 4.649373123272672e-05,
+      "loss": 0.9844,
+      "step": 6006
+    },
+    {
+      "epoch": 0.536734648290035,
+      "grad_norm": 0.5183307528495789,
+      "learning_rate": 4.647929628702408e-05,
+      "loss": 0.9385,
+      "step": 6007
+    },
+    {
+      "epoch": 0.5368239998212969,
+      "grad_norm": 0.4516974091529846,
+      "learning_rate": 4.6464861636217895e-05,
+      "loss": 0.9214,
+      "step": 6008
+    },
+    {
+      "epoch": 0.5369133513525588,
+      "grad_norm": 0.39056047797203064,
+      "learning_rate": 4.645042728151722e-05,
+      "loss": 0.9394,
+      "step": 6009
+    },
+    {
+      "epoch": 0.5370027028838207,
+      "grad_norm": 0.3742974102497101,
+      "learning_rate": 4.64359932241311e-05,
+      "loss": 0.9829,
+      "step": 6010
+    },
+    {
+      "epoch": 0.5370920544150826,
+      "grad_norm": 0.491948664188385,
+      "learning_rate": 4.642155946526854e-05,
+      "loss": 0.8914,
+      "step": 6011
+    },
+    {
+      "epoch": 0.5371814059463444,
+      "grad_norm": 0.4805702567100525,
+      "learning_rate": 4.640712600613851e-05,
+      "loss": 0.9744,
+      "step": 6012
+    },
+    {
+      "epoch": 0.5372707574776062,
+      "grad_norm": 0.48853474855422974,
+      "learning_rate": 4.6392692847949984e-05,
+      "loss": 0.985,
+      "step": 6013
+    },
+    {
+      "epoch": 0.5373601090088681,
+      "grad_norm": 0.414133220911026,
+      "learning_rate": 4.6378259991911886e-05,
+      "loss": 0.9825,
+      "step": 6014
+    },
+    {
+      "epoch": 0.53744946054013,
+      "grad_norm": 0.3783036470413208,
+      "learning_rate": 4.6363827439233114e-05,
+      "loss": 0.9994,
+      "step": 6015
+    },
+    {
+      "epoch": 0.5375388120713919,
+      "grad_norm": 0.5566924214363098,
+      "learning_rate": 4.634939519112255e-05,
+      "loss": 0.9119,
+      "step": 6016
+    },
+    {
+      "epoch": 0.5376281636026538,
+      "grad_norm": 0.49489137530326843,
+      "learning_rate": 4.633496324878906e-05,
+      "loss": 0.952,
+      "step": 6017
+    },
+    {
+      "epoch": 0.5377175151339156,
+      "grad_norm": 0.45840591192245483,
+      "learning_rate": 4.632053161344146e-05,
+      "loss": 0.8933,
+      "step": 6018
+    },
+    {
+      "epoch": 0.5378068666651775,
+      "grad_norm": 0.5176952481269836,
+      "learning_rate": 4.6306100286288565e-05,
+      "loss": 0.9606,
+      "step": 6019
+    },
+    {
+      "epoch": 0.5378962181964393,
+      "grad_norm": 0.484392374753952,
+      "learning_rate": 4.629166926853913e-05,
+      "loss": 0.9356,
+      "step": 6020
+    },
+    {
+      "epoch": 0.5379855697277012,
+      "grad_norm": 0.5076186656951904,
+      "learning_rate": 4.6277238561401927e-05,
+      "loss": 0.9744,
+      "step": 6021
+    },
+    {
+      "epoch": 0.5380749212589631,
+      "grad_norm": 0.4623713791370392,
+      "learning_rate": 4.6262808166085674e-05,
+      "loss": 0.9705,
+      "step": 6022
+    },
+    {
+      "epoch": 0.5381642727902249,
+      "grad_norm": 0.419096440076828,
+      "learning_rate": 4.624837808379907e-05,
+      "loss": 1.0033,
+      "step": 6023
+    },
+    {
+      "epoch": 0.5382536243214868,
+      "grad_norm": 0.3825317323207855,
+      "learning_rate": 4.62339483157508e-05,
+      "loss": 1.0228,
+      "step": 6024
+    },
+    {
+      "epoch": 0.5383429758527487,
+      "grad_norm": 0.5461272597312927,
+      "learning_rate": 4.6219518863149493e-05,
+      "loss": 0.9623,
+      "step": 6025
+    },
+    {
+      "epoch": 0.5384323273840106,
+      "grad_norm": 0.4370659589767456,
+      "learning_rate": 4.6205089727203785e-05,
+      "loss": 0.9881,
+      "step": 6026
+    },
+    {
+      "epoch": 0.5385216789152724,
+      "grad_norm": 0.48183608055114746,
+      "learning_rate": 4.619066090912228e-05,
+      "loss": 0.867,
+      "step": 6027
+    },
+    {
+      "epoch": 0.5386110304465342,
+      "grad_norm": 0.4971342980861664,
+      "learning_rate": 4.6176232410113506e-05,
+      "loss": 0.9406,
+      "step": 6028
+    },
+    {
+      "epoch": 0.5387003819777961,
+      "grad_norm": 0.44625434279441833,
+      "learning_rate": 4.616180423138603e-05,
+      "loss": 0.9701,
+      "step": 6029
+    },
+    {
+      "epoch": 0.538789733509058,
+      "grad_norm": 0.4783417880535126,
+      "learning_rate": 4.614737637414836e-05,
+      "loss": 0.8333,
+      "step": 6030
+    },
+    {
+      "epoch": 0.5388790850403199,
+      "grad_norm": 0.39774414896965027,
+      "learning_rate": 4.613294883960898e-05,
+      "loss": 0.9892,
+      "step": 6031
+    },
+    {
+      "epoch": 0.5389684365715818,
+      "grad_norm": 0.449542373418808,
+      "learning_rate": 4.611852162897636e-05,
+      "loss": 0.9164,
+      "step": 6032
+    },
+    {
+      "epoch": 0.5390577881028437,
+      "grad_norm": 0.38516610860824585,
+      "learning_rate": 4.610409474345894e-05,
+      "loss": 0.9848,
+      "step": 6033
+    },
+    {
+      "epoch": 0.5391471396341054,
+      "grad_norm": 0.43511533737182617,
+      "learning_rate": 4.60896681842651e-05,
+      "loss": 0.9101,
+      "step": 6034
+    },
+    {
+      "epoch": 0.5392364911653673,
+      "grad_norm": 0.4192296862602234,
+      "learning_rate": 4.6075241952603225e-05,
+      "loss": 1.0098,
+      "step": 6035
+    },
+    {
+      "epoch": 0.5393258426966292,
+      "grad_norm": 0.38819652795791626,
+      "learning_rate": 4.6060816049681676e-05,
+      "loss": 0.9525,
+      "step": 6036
+    },
+    {
+      "epoch": 0.5394151942278911,
+      "grad_norm": 0.4016992449760437,
+      "learning_rate": 4.6046390476708794e-05,
+      "loss": 0.993,
+      "step": 6037
+    },
+    {
+      "epoch": 0.539504545759153,
+      "grad_norm": 0.4347737431526184,
+      "learning_rate": 4.6031965234892834e-05,
+      "loss": 0.9847,
+      "step": 6038
+    },
+    {
+      "epoch": 0.5395938972904148,
+      "grad_norm": 0.4962137043476105,
+      "learning_rate": 4.601754032544208e-05,
+      "loss": 0.908,
+      "step": 6039
+    },
+    {
+      "epoch": 0.5396832488216767,
+      "grad_norm": 0.4483336806297302,
+      "learning_rate": 4.6003115749564765e-05,
+      "loss": 0.9726,
+      "step": 6040
+    },
+    {
+      "epoch": 0.5397726003529385,
+      "grad_norm": 0.5861220359802246,
+      "learning_rate": 4.598869150846912e-05,
+      "loss": 0.8876,
+      "step": 6041
+    },
+    {
+      "epoch": 0.5398619518842004,
+      "grad_norm": 0.4730405807495117,
+      "learning_rate": 4.597426760336331e-05,
+      "loss": 0.9645,
+      "step": 6042
+    },
+    {
+      "epoch": 0.5399513034154623,
+      "grad_norm": 0.48867109417915344,
+      "learning_rate": 4.59598440354555e-05,
+      "loss": 0.9377,
+      "step": 6043
+    },
+    {
+      "epoch": 0.5400406549467242,
+      "grad_norm": 0.4301815927028656,
+      "learning_rate": 4.5945420805953825e-05,
+      "loss": 0.9474,
+      "step": 6044
+    },
+    {
+      "epoch": 0.540130006477986,
+      "grad_norm": 0.4912492334842682,
+      "learning_rate": 4.593099791606637e-05,
+      "loss": 0.955,
+      "step": 6045
+    },
+    {
+      "epoch": 0.5402193580092479,
+      "grad_norm": 0.5850039124488831,
+      "learning_rate": 4.5916575367001214e-05,
+      "loss": 0.904,
+      "step": 6046
+    },
+    {
+      "epoch": 0.5403087095405098,
+      "grad_norm": 0.4344959557056427,
+      "learning_rate": 4.590215315996642e-05,
+      "loss": 0.9921,
+      "step": 6047
+    },
+    {
+      "epoch": 0.5403980610717716,
+      "grad_norm": 0.5166114568710327,
+      "learning_rate": 4.588773129616996e-05,
+      "loss": 0.941,
+      "step": 6048
+    },
+    {
+      "epoch": 0.5404874126030335,
+      "grad_norm": 0.5347388386726379,
+      "learning_rate": 4.587330977681983e-05,
+      "loss": 0.893,
+      "step": 6049
+    },
+    {
+      "epoch": 0.5405767641342953,
+      "grad_norm": 0.398970365524292,
+      "learning_rate": 4.585888860312399e-05,
+      "loss": 0.9573,
+      "step": 6050
+    },
+    {
+      "epoch": 0.5406661156655572,
+      "grad_norm": 0.4769752621650696,
+      "learning_rate": 4.584446777629038e-05,
+      "loss": 0.9554,
+      "step": 6051
+    },
+    {
+      "epoch": 0.5407554671968191,
+      "grad_norm": 0.5892907381057739,
+      "learning_rate": 4.5830047297526904e-05,
+      "loss": 1.025,
+      "step": 6052
+    },
+    {
+      "epoch": 0.540844818728081,
+      "grad_norm": 0.4411056637763977,
+      "learning_rate": 4.58156271680414e-05,
+      "loss": 0.9451,
+      "step": 6053
+    },
+    {
+      "epoch": 0.5409341702593429,
+      "grad_norm": 0.4067056477069855,
+      "learning_rate": 4.5801207389041715e-05,
+      "loss": 0.993,
+      "step": 6054
+    },
+    {
+      "epoch": 0.5410235217906046,
+      "grad_norm": 0.44847869873046875,
+      "learning_rate": 4.5786787961735673e-05,
+      "loss": 0.9894,
+      "step": 6055
+    },
+    {
+      "epoch": 0.5411128733218665,
+      "grad_norm": 0.43127429485321045,
+      "learning_rate": 4.577236888733105e-05,
+      "loss": 0.9412,
+      "step": 6056
+    },
+    {
+      "epoch": 0.5412022248531284,
+      "grad_norm": 0.4345093369483948,
+      "learning_rate": 4.575795016703561e-05,
+      "loss": 1.0033,
+      "step": 6057
+    },
+    {
+      "epoch": 0.5412915763843903,
+      "grad_norm": 0.44015347957611084,
+      "learning_rate": 4.574353180205705e-05,
+      "loss": 0.9976,
+      "step": 6058
+    },
+    {
+      "epoch": 0.5413809279156522,
+      "grad_norm": 0.45786052942276,
+      "learning_rate": 4.572911379360307e-05,
+      "loss": 0.9713,
+      "step": 6059
+    },
+    {
+      "epoch": 0.5414702794469141,
+      "grad_norm": 0.47783318161964417,
+      "learning_rate": 4.571469614288133e-05,
+      "loss": 0.9236,
+      "step": 6060
+    },
+    {
+      "epoch": 0.5415596309781759,
+      "grad_norm": 0.5037944316864014,
+      "learning_rate": 4.5700278851099464e-05,
+      "loss": 0.9076,
+      "step": 6061
+    },
+    {
+      "epoch": 0.5416489825094377,
+      "grad_norm": 0.402452290058136,
+      "learning_rate": 4.568586191946508e-05,
+      "loss": 0.9454,
+      "step": 6062
+    },
+    {
+      "epoch": 0.5417383340406996,
+      "grad_norm": 0.45303410291671753,
+      "learning_rate": 4.567144534918574e-05,
+      "loss": 0.9148,
+      "step": 6063
+    },
+    {
+      "epoch": 0.5418276855719615,
+      "grad_norm": 0.4717887341976166,
+      "learning_rate": 4.5657029141468996e-05,
+      "loss": 0.9045,
+      "step": 6064
+    },
+    {
+      "epoch": 0.5419170371032234,
+      "grad_norm": 0.4819827079772949,
+      "learning_rate": 4.564261329752236e-05,
+      "loss": 0.8857,
+      "step": 6065
+    },
+    {
+      "epoch": 0.5420063886344852,
+      "grad_norm": 0.4282495975494385,
+      "learning_rate": 4.562819781855331e-05,
+      "loss": 0.9649,
+      "step": 6066
+    },
+    {
+      "epoch": 0.5420957401657471,
+      "grad_norm": 0.43561655282974243,
+      "learning_rate": 4.561378270576929e-05,
+      "loss": 0.9829,
+      "step": 6067
+    },
+    {
+      "epoch": 0.5421850916970089,
+      "grad_norm": 0.47019922733306885,
+      "learning_rate": 4.559936796037772e-05,
+      "loss": 0.9926,
+      "step": 6068
+    },
+    {
+      "epoch": 0.5422744432282708,
+      "grad_norm": 0.5216493606567383,
+      "learning_rate": 4.5584953583585985e-05,
+      "loss": 0.8186,
+      "step": 6069
+    },
+    {
+      "epoch": 0.5423637947595327,
+      "grad_norm": 0.5065385699272156,
+      "learning_rate": 4.5570539576601463e-05,
+      "loss": 0.9491,
+      "step": 6070
+    },
+    {
+      "epoch": 0.5424531462907946,
+      "grad_norm": 0.473476380109787,
+      "learning_rate": 4.5556125940631454e-05,
+      "loss": 0.9613,
+      "step": 6071
+    },
+    {
+      "epoch": 0.5425424978220564,
+      "grad_norm": 0.4875657558441162,
+      "learning_rate": 4.5541712676883263e-05,
+      "loss": 0.9939,
+      "step": 6072
+    },
+    {
+      "epoch": 0.5426318493533183,
+      "grad_norm": 0.5284254550933838,
+      "learning_rate": 4.552729978656416e-05,
+      "loss": 0.8902,
+      "step": 6073
+    },
+    {
+      "epoch": 0.5427212008845802,
+      "grad_norm": 0.5050989985466003,
+      "learning_rate": 4.5512887270881374e-05,
+      "loss": 0.8749,
+      "step": 6074
+    },
+    {
+      "epoch": 0.542810552415842,
+      "grad_norm": 0.5607906579971313,
+      "learning_rate": 4.5498475131042106e-05,
+      "loss": 0.9388,
+      "step": 6075
+    },
+    {
+      "epoch": 0.5428999039471039,
+      "grad_norm": 0.5271878838539124,
+      "learning_rate": 4.548406336825353e-05,
+      "loss": 1.0645,
+      "step": 6076
+    },
+    {
+      "epoch": 0.5429892554783657,
+      "grad_norm": 0.41676685214042664,
+      "learning_rate": 4.546965198372279e-05,
+      "loss": 0.9279,
+      "step": 6077
+    },
+    {
+      "epoch": 0.5430786070096276,
+      "grad_norm": 0.47751128673553467,
+      "learning_rate": 4.5455240978656996e-05,
+      "loss": 0.9398,
+      "step": 6078
+    },
+    {
+      "epoch": 0.5431679585408895,
+      "grad_norm": 0.4766964912414551,
+      "learning_rate": 4.5440830354263205e-05,
+      "loss": 0.9324,
+      "step": 6079
+    },
+    {
+      "epoch": 0.5432573100721514,
+      "grad_norm": 0.4669797718524933,
+      "learning_rate": 4.542642011174846e-05,
+      "loss": 0.9492,
+      "step": 6080
+    },
+    {
+      "epoch": 0.5433466616034133,
+      "grad_norm": 0.4376409947872162,
+      "learning_rate": 4.5412010252319784e-05,
+      "loss": 1.0007,
+      "step": 6081
+    },
+    {
+      "epoch": 0.543436013134675,
+      "grad_norm": 0.42658764123916626,
+      "learning_rate": 4.539760077718416e-05,
+      "loss": 0.9542,
+      "step": 6082
+    },
+    {
+      "epoch": 0.5435253646659369,
+      "grad_norm": 0.4167364835739136,
+      "learning_rate": 4.5383191687548513e-05,
+      "loss": 1.0401,
+      "step": 6083
+    },
+    {
+      "epoch": 0.5436147161971988,
+      "grad_norm": 0.4156143367290497,
+      "learning_rate": 4.53687829846198e-05,
+      "loss": 1.0165,
+      "step": 6084
+    },
+    {
+      "epoch": 0.5437040677284607,
+      "grad_norm": 0.5895379781723022,
+      "learning_rate": 4.535437466960486e-05,
+      "loss": 0.9548,
+      "step": 6085
+    },
+    {
+      "epoch": 0.5437934192597226,
+      "grad_norm": 0.44905561208724976,
+      "learning_rate": 4.533996674371056e-05,
+      "loss": 0.9401,
+      "step": 6086
+    },
+    {
+      "epoch": 0.5438827707909845,
+      "grad_norm": 0.43985962867736816,
+      "learning_rate": 4.5325559208143717e-05,
+      "loss": 0.9516,
+      "step": 6087
+    },
+    {
+      "epoch": 0.5439721223222463,
+      "grad_norm": 0.5113573670387268,
+      "learning_rate": 4.5311152064111134e-05,
+      "loss": 0.9881,
+      "step": 6088
+    },
+    {
+      "epoch": 0.5440614738535081,
+      "grad_norm": 0.37436649203300476,
+      "learning_rate": 4.529674531281954e-05,
+      "loss": 0.9438,
+      "step": 6089
+    },
+    {
+      "epoch": 0.54415082538477,
+      "grad_norm": 0.5028565526008606,
+      "learning_rate": 4.5282338955475644e-05,
+      "loss": 1.0298,
+      "step": 6090
+    },
+    {
+      "epoch": 0.5442401769160319,
+      "grad_norm": 0.4592767059803009,
+      "learning_rate": 4.526793299328616e-05,
+      "loss": 0.9478,
+      "step": 6091
+    },
+    {
+      "epoch": 0.5443295284472938,
+      "grad_norm": 0.4168972671031952,
+      "learning_rate": 4.5253527427457715e-05,
+      "loss": 0.9624,
+      "step": 6092
+    },
+    {
+      "epoch": 0.5444188799785556,
+      "grad_norm": 0.4067929685115814,
+      "learning_rate": 4.523912225919694e-05,
+      "loss": 0.8968,
+      "step": 6093
+    },
+    {
+      "epoch": 0.5445082315098175,
+      "grad_norm": 0.5613819360733032,
+      "learning_rate": 4.522471748971043e-05,
+      "loss": 0.8793,
+      "step": 6094
+    },
+    {
+      "epoch": 0.5445975830410794,
+      "grad_norm": 0.4839697778224945,
+      "learning_rate": 4.5210313120204735e-05,
+      "loss": 0.9269,
+      "step": 6095
+    },
+    {
+      "epoch": 0.5446869345723412,
+      "grad_norm": 0.42698168754577637,
+      "learning_rate": 4.519590915188637e-05,
+      "loss": 0.9873,
+      "step": 6096
+    },
+    {
+      "epoch": 0.5447762861036031,
+      "grad_norm": 0.5424548983573914,
+      "learning_rate": 4.5181505585961816e-05,
+      "loss": 0.9847,
+      "step": 6097
+    },
+    {
+      "epoch": 0.544865637634865,
+      "grad_norm": 0.4275175929069519,
+      "learning_rate": 4.5167102423637554e-05,
+      "loss": 1.0299,
+      "step": 6098
+    },
+    {
+      "epoch": 0.5449549891661268,
+      "grad_norm": 0.4584430158138275,
+      "learning_rate": 4.515269966611996e-05,
+      "loss": 0.9439,
+      "step": 6099
+    },
+    {
+      "epoch": 0.5450443406973887,
+      "grad_norm": 0.4883447289466858,
+      "learning_rate": 4.513829731461543e-05,
+      "loss": 0.9921,
+      "step": 6100
+    },
+    {
+      "epoch": 0.5451336922286506,
+      "grad_norm": 0.44181644916534424,
+      "learning_rate": 4.512389537033033e-05,
+      "loss": 0.9704,
+      "step": 6101
+    },
+    {
+      "epoch": 0.5452230437599125,
+      "grad_norm": 0.47616517543792725,
+      "learning_rate": 4.510949383447096e-05,
+      "loss": 0.9071,
+      "step": 6102
+    },
+    {
+      "epoch": 0.5453123952911743,
+      "grad_norm": 0.6123970746994019,
+      "learning_rate": 4.5095092708243623e-05,
+      "loss": 0.8817,
+      "step": 6103
+    },
+    {
+      "epoch": 0.5454017468224361,
+      "grad_norm": 0.4854317009449005,
+      "learning_rate": 4.5080691992854554e-05,
+      "loss": 0.8724,
+      "step": 6104
+    },
+    {
+      "epoch": 0.545491098353698,
+      "grad_norm": 0.47287046909332275,
+      "learning_rate": 4.5066291689509953e-05,
+      "loss": 0.9679,
+      "step": 6105
+    },
+    {
+      "epoch": 0.5455804498849599,
+      "grad_norm": 0.4474678933620453,
+      "learning_rate": 4.5051891799416025e-05,
+      "loss": 0.943,
+      "step": 6106
+    },
+    {
+      "epoch": 0.5456698014162218,
+      "grad_norm": 0.46921437978744507,
+      "learning_rate": 4.503749232377889e-05,
+      "loss": 0.9584,
+      "step": 6107
+    },
+    {
+      "epoch": 0.5457591529474837,
+      "grad_norm": 0.5251668691635132,
+      "learning_rate": 4.50230932638047e-05,
+      "loss": 0.9205,
+      "step": 6108
+    },
+    {
+      "epoch": 0.5458485044787456,
+      "grad_norm": 0.41507506370544434,
+      "learning_rate": 4.5008694620699474e-05,
+      "loss": 0.9704,
+      "step": 6109
+    },
+    {
+      "epoch": 0.5459378560100073,
+      "grad_norm": 0.43505141139030457,
+      "learning_rate": 4.4994296395669276e-05,
+      "loss": 0.926,
+      "step": 6110
+    },
+    {
+      "epoch": 0.5460272075412692,
+      "grad_norm": 0.5086851716041565,
+      "learning_rate": 4.497989858992011e-05,
+      "loss": 0.905,
+      "step": 6111
+    },
+    {
+      "epoch": 0.5461165590725311,
+      "grad_norm": 0.4543953537940979,
+      "learning_rate": 4.496550120465795e-05,
+      "loss": 0.9689,
+      "step": 6112
+    },
+    {
+      "epoch": 0.546205910603793,
+      "grad_norm": 0.4054141938686371,
+      "learning_rate": 4.495110424108873e-05,
+      "loss": 1.0004,
+      "step": 6113
+    },
+    {
+      "epoch": 0.5462952621350549,
+      "grad_norm": 0.39775604009628296,
+      "learning_rate": 4.4936707700418346e-05,
+      "loss": 1.021,
+      "step": 6114
+    },
+    {
+      "epoch": 0.5463846136663167,
+      "grad_norm": 0.5066169500350952,
+      "learning_rate": 4.492231158385266e-05,
+      "loss": 0.9796,
+      "step": 6115
+    },
+    {
+      "epoch": 0.5464739651975786,
+      "grad_norm": 0.43312814831733704,
+      "learning_rate": 4.4907915892597504e-05,
+      "loss": 0.9808,
+      "step": 6116
+    },
+    {
+      "epoch": 0.5465633167288404,
+      "grad_norm": 0.4440118372440338,
+      "learning_rate": 4.489352062785869e-05,
+      "loss": 0.9781,
+      "step": 6117
+    },
+    {
+      "epoch": 0.5466526682601023,
+      "grad_norm": 0.45043429732322693,
+      "learning_rate": 4.4879125790841944e-05,
+      "loss": 1.0155,
+      "step": 6118
+    },
+    {
+      "epoch": 0.5467420197913642,
+      "grad_norm": 0.44489559531211853,
+      "learning_rate": 4.486473138275299e-05,
+      "loss": 0.9426,
+      "step": 6119
+    },
+    {
+      "epoch": 0.546831371322626,
+      "grad_norm": 0.45286840200424194,
+      "learning_rate": 4.485033740479752e-05,
+      "loss": 0.982,
+      "step": 6120
+    },
+    {
+      "epoch": 0.5469207228538879,
+      "grad_norm": 0.4958401322364807,
+      "learning_rate": 4.483594385818118e-05,
+      "loss": 0.9419,
+      "step": 6121
+    },
+    {
+      "epoch": 0.5470100743851498,
+      "grad_norm": 0.4510517120361328,
+      "learning_rate": 4.482155074410961e-05,
+      "loss": 0.9684,
+      "step": 6122
+    },
+    {
+      "epoch": 0.5470994259164117,
+      "grad_norm": 0.3880564272403717,
+      "learning_rate": 4.480715806378834e-05,
+      "loss": 1.0172,
+      "step": 6123
+    },
+    {
+      "epoch": 0.5471887774476735,
+      "grad_norm": 0.5004435181617737,
+      "learning_rate": 4.479276581842294e-05,
+      "loss": 1.0378,
+      "step": 6124
+    },
+    {
+      "epoch": 0.5472781289789354,
+      "grad_norm": 0.43363669514656067,
+      "learning_rate": 4.4778374009218904e-05,
+      "loss": 1.0019,
+      "step": 6125
+    },
+    {
+      "epoch": 0.5473674805101972,
+      "grad_norm": 0.4868524670600891,
+      "learning_rate": 4.4763982637381706e-05,
+      "loss": 0.9947,
+      "step": 6126
+    },
+    {
+      "epoch": 0.5474568320414591,
+      "grad_norm": 0.43527916073799133,
+      "learning_rate": 4.474959170411677e-05,
+      "loss": 1.0288,
+      "step": 6127
+    },
+    {
+      "epoch": 0.547546183572721,
+      "grad_norm": 0.5454886555671692,
+      "learning_rate": 4.473520121062952e-05,
+      "loss": 0.9882,
+      "step": 6128
+    },
+    {
+      "epoch": 0.5476355351039829,
+      "grad_norm": 0.4397745132446289,
+      "learning_rate": 4.4720811158125267e-05,
+      "loss": 1.0003,
+      "step": 6129
+    },
+    {
+      "epoch": 0.5477248866352447,
+      "grad_norm": 0.5059059262275696,
+      "learning_rate": 4.470642154780935e-05,
+      "loss": 0.9265,
+      "step": 6130
+    },
+    {
+      "epoch": 0.5478142381665065,
+      "grad_norm": 0.5163365006446838,
+      "learning_rate": 4.469203238088705e-05,
+      "loss": 0.9582,
+      "step": 6131
+    },
+    {
+      "epoch": 0.5479035896977684,
+      "grad_norm": 0.48326799273490906,
+      "learning_rate": 4.467764365856362e-05,
+      "loss": 0.9396,
+      "step": 6132
+    },
+    {
+      "epoch": 0.5479929412290303,
+      "grad_norm": 0.46925976872444153,
+      "learning_rate": 4.466325538204427e-05,
+      "loss": 1.0036,
+      "step": 6133
+    },
+    {
+      "epoch": 0.5480822927602922,
+      "grad_norm": 0.4270949363708496,
+      "learning_rate": 4.464886755253416e-05,
+      "loss": 1.0472,
+      "step": 6134
+    },
+    {
+      "epoch": 0.5481716442915541,
+      "grad_norm": 0.49136438965797424,
+      "learning_rate": 4.463448017123844e-05,
+      "loss": 0.8925,
+      "step": 6135
+    },
+    {
+      "epoch": 0.548260995822816,
+      "grad_norm": 0.41291165351867676,
+      "learning_rate": 4.4620093239362204e-05,
+      "loss": 0.9923,
+      "step": 6136
+    },
+    {
+      "epoch": 0.5483503473540777,
+      "grad_norm": 0.40838754177093506,
+      "learning_rate": 4.460570675811049e-05,
+      "loss": 0.9903,
+      "step": 6137
+    },
+    {
+      "epoch": 0.5484396988853396,
+      "grad_norm": 0.5193032622337341,
+      "learning_rate": 4.459132072868835e-05,
+      "loss": 0.9049,
+      "step": 6138
+    },
+    {
+      "epoch": 0.5485290504166015,
+      "grad_norm": 0.4100017547607422,
+      "learning_rate": 4.457693515230074e-05,
+      "loss": 1.0034,
+      "step": 6139
+    },
+    {
+      "epoch": 0.5486184019478634,
+      "grad_norm": 0.5860205888748169,
+      "learning_rate": 4.456255003015263e-05,
+      "loss": 0.9951,
+      "step": 6140
+    },
+    {
+      "epoch": 0.5487077534791253,
+      "grad_norm": 0.49297282099723816,
+      "learning_rate": 4.4548165363448894e-05,
+      "loss": 0.9113,
+      "step": 6141
+    },
+    {
+      "epoch": 0.5487971050103871,
+      "grad_norm": 0.41039666533470154,
+      "learning_rate": 4.4533781153394426e-05,
+      "loss": 0.9713,
+      "step": 6142
+    },
+    {
+      "epoch": 0.548886456541649,
+      "grad_norm": 0.40572866797447205,
+      "learning_rate": 4.4519397401194056e-05,
+      "loss": 0.9805,
+      "step": 6143
+    },
+    {
+      "epoch": 0.5489758080729108,
+      "grad_norm": 0.43796059489250183,
+      "learning_rate": 4.4505014108052564e-05,
+      "loss": 1.0003,
+      "step": 6144
+    },
+    {
+      "epoch": 0.5490651596041727,
+      "grad_norm": 0.4492766857147217,
+      "learning_rate": 4.449063127517472e-05,
+      "loss": 0.9586,
+      "step": 6145
+    },
+    {
+      "epoch": 0.5491545111354346,
+      "grad_norm": 0.3965035676956177,
+      "learning_rate": 4.447624890376523e-05,
+      "loss": 0.9954,
+      "step": 6146
+    },
+    {
+      "epoch": 0.5492438626666964,
+      "grad_norm": 0.45780253410339355,
+      "learning_rate": 4.4461866995028776e-05,
+      "loss": 0.9275,
+      "step": 6147
+    },
+    {
+      "epoch": 0.5493332141979583,
+      "grad_norm": 0.4715741276741028,
+      "learning_rate": 4.4447485550170013e-05,
+      "loss": 0.9706,
+      "step": 6148
+    },
+    {
+      "epoch": 0.5494225657292202,
+      "grad_norm": 0.43292608857154846,
+      "learning_rate": 4.44331045703935e-05,
+      "loss": 0.9596,
+      "step": 6149
+    },
+    {
+      "epoch": 0.5495119172604821,
+      "grad_norm": 0.4346010982990265,
+      "learning_rate": 4.4418724056903824e-05,
+      "loss": 1.0092,
+      "step": 6150
+    },
+    {
+      "epoch": 0.5496012687917439,
+      "grad_norm": 0.5378443002700806,
+      "learning_rate": 4.440434401090549e-05,
+      "loss": 0.9605,
+      "step": 6151
+    },
+    {
+      "epoch": 0.5496906203230058,
+      "grad_norm": 0.5118531584739685,
+      "learning_rate": 4.438996443360299e-05,
+      "loss": 0.8896,
+      "step": 6152
+    },
+    {
+      "epoch": 0.5497799718542676,
+      "grad_norm": 0.4279499351978302,
+      "learning_rate": 4.437558532620077e-05,
+      "loss": 0.8621,
+      "step": 6153
+    },
+    {
+      "epoch": 0.5498693233855295,
+      "grad_norm": 0.4221118092536926,
+      "learning_rate": 4.436120668990324e-05,
+      "loss": 0.9983,
+      "step": 6154
+    },
+    {
+      "epoch": 0.5499586749167914,
+      "grad_norm": 0.42579370737075806,
+      "learning_rate": 4.434682852591476e-05,
+      "loss": 0.9634,
+      "step": 6155
+    },
+    {
+      "epoch": 0.5500480264480533,
+      "grad_norm": 0.47894287109375,
+      "learning_rate": 4.4332450835439634e-05,
+      "loss": 0.9609,
+      "step": 6156
+    },
+    {
+      "epoch": 0.5501373779793152,
+      "grad_norm": 0.5039592385292053,
+      "learning_rate": 4.431807361968217e-05,
+      "loss": 0.9188,
+      "step": 6157
+    },
+    {
+      "epoch": 0.5502267295105769,
+      "grad_norm": 0.4221189618110657,
+      "learning_rate": 4.4303696879846593e-05,
+      "loss": 0.9687,
+      "step": 6158
+    },
+    {
+      "epoch": 0.5503160810418388,
+      "grad_norm": 0.454010009765625,
+      "learning_rate": 4.428932061713715e-05,
+      "loss": 0.9647,
+      "step": 6159
+    },
+    {
+      "epoch": 0.5504054325731007,
+      "grad_norm": 0.409974604845047,
+      "learning_rate": 4.427494483275796e-05,
+      "loss": 0.9301,
+      "step": 6160
+    },
+    {
+      "epoch": 0.5504947841043626,
+      "grad_norm": 0.4791439473628998,
+      "learning_rate": 4.426056952791316e-05,
+      "loss": 0.9908,
+      "step": 6161
+    },
+    {
+      "epoch": 0.5505841356356245,
+      "grad_norm": 0.4021007716655731,
+      "learning_rate": 4.424619470380684e-05,
+      "loss": 1.0427,
+      "step": 6162
+    },
+    {
+      "epoch": 0.5506734871668864,
+      "grad_norm": 0.48248547315597534,
+      "learning_rate": 4.423182036164304e-05,
+      "loss": 0.9028,
+      "step": 6163
+    },
+    {
+      "epoch": 0.5507628386981482,
+      "grad_norm": 0.4851128160953522,
+      "learning_rate": 4.4217446502625773e-05,
+      "loss": 0.9881,
+      "step": 6164
+    },
+    {
+      "epoch": 0.55085219022941,
+      "grad_norm": 0.42037060856819153,
+      "learning_rate": 4.420307312795901e-05,
+      "loss": 0.9585,
+      "step": 6165
+    },
+    {
+      "epoch": 0.5509415417606719,
+      "grad_norm": 0.47032657265663147,
+      "learning_rate": 4.418870023884665e-05,
+      "loss": 0.9322,
+      "step": 6166
+    },
+    {
+      "epoch": 0.5510308932919338,
+      "grad_norm": 0.6283010244369507,
+      "learning_rate": 4.4174327836492587e-05,
+      "loss": 0.9798,
+      "step": 6167
+    },
+    {
+      "epoch": 0.5511202448231957,
+      "grad_norm": 0.45526832342147827,
+      "learning_rate": 4.4159955922100674e-05,
+      "loss": 0.9473,
+      "step": 6168
+    },
+    {
+      "epoch": 0.5512095963544575,
+      "grad_norm": 0.4486526548862457,
+      "learning_rate": 4.414558449687471e-05,
+      "loss": 0.9134,
+      "step": 6169
+    },
+    {
+      "epoch": 0.5512989478857194,
+      "grad_norm": 0.38991108536720276,
+      "learning_rate": 4.413121356201844e-05,
+      "loss": 0.9947,
+      "step": 6170
+    },
+    {
+      "epoch": 0.5513882994169813,
+      "grad_norm": 0.4334878623485565,
+      "learning_rate": 4.411684311873559e-05,
+      "loss": 0.9613,
+      "step": 6171
+    },
+    {
+      "epoch": 0.5514776509482431,
+      "grad_norm": 0.49451541900634766,
+      "learning_rate": 4.4102473168229837e-05,
+      "loss": 0.9073,
+      "step": 6172
+    },
+    {
+      "epoch": 0.551567002479505,
+      "grad_norm": 0.41775211691856384,
+      "learning_rate": 4.408810371170483e-05,
+      "loss": 0.9729,
+      "step": 6173
+    },
+    {
+      "epoch": 0.5516563540107668,
+      "grad_norm": 0.4632229506969452,
+      "learning_rate": 4.4073734750364144e-05,
+      "loss": 0.9096,
+      "step": 6174
+    },
+    {
+      "epoch": 0.5517457055420287,
+      "grad_norm": 0.5613375306129456,
+      "learning_rate": 4.4059366285411344e-05,
+      "loss": 0.8467,
+      "step": 6175
+    },
+    {
+      "epoch": 0.5518350570732906,
+      "grad_norm": 0.4471941590309143,
+      "learning_rate": 4.404499831804993e-05,
+      "loss": 0.9103,
+      "step": 6176
+    },
+    {
+      "epoch": 0.5519244086045525,
+      "grad_norm": 0.4910937547683716,
+      "learning_rate": 4.403063084948339e-05,
+      "loss": 1.0049,
+      "step": 6177
+    },
+    {
+      "epoch": 0.5520137601358144,
+      "grad_norm": 0.551129162311554,
+      "learning_rate": 4.4016263880915146e-05,
+      "loss": 0.9435,
+      "step": 6178
+    },
+    {
+      "epoch": 0.5521031116670762,
+      "grad_norm": 0.44475892186164856,
+      "learning_rate": 4.4001897413548605e-05,
+      "loss": 0.9798,
+      "step": 6179
+    },
+    {
+      "epoch": 0.552192463198338,
+      "grad_norm": 0.549535870552063,
+      "learning_rate": 4.398753144858707e-05,
+      "loss": 0.878,
+      "step": 6180
+    },
+    {
+      "epoch": 0.5522818147295999,
+      "grad_norm": 0.4066050052642822,
+      "learning_rate": 4.397316598723385e-05,
+      "loss": 1.008,
+      "step": 6181
+    },
+    {
+      "epoch": 0.5523711662608618,
+      "grad_norm": 0.4381326138973236,
+      "learning_rate": 4.3958801030692245e-05,
+      "loss": 0.9418,
+      "step": 6182
+    },
+    {
+      "epoch": 0.5524605177921237,
+      "grad_norm": 0.4462524354457855,
+      "learning_rate": 4.394443658016543e-05,
+      "loss": 0.932,
+      "step": 6183
+    },
+    {
+      "epoch": 0.5525498693233856,
+      "grad_norm": 0.4263027608394623,
+      "learning_rate": 4.393007263685661e-05,
+      "loss": 0.968,
+      "step": 6184
+    },
+    {
+      "epoch": 0.5526392208546475,
+      "grad_norm": 0.3910346031188965,
+      "learning_rate": 4.3915709201968896e-05,
+      "loss": 1.014,
+      "step": 6185
+    },
+    {
+      "epoch": 0.5527285723859092,
+      "grad_norm": 0.45934584736824036,
+      "learning_rate": 4.39013462767054e-05,
+      "loss": 0.9602,
+      "step": 6186
+    },
+    {
+      "epoch": 0.5528179239171711,
+      "grad_norm": 0.48381033539772034,
+      "learning_rate": 4.388698386226917e-05,
+      "loss": 0.9204,
+      "step": 6187
+    },
+    {
+      "epoch": 0.552907275448433,
+      "grad_norm": 0.4025076627731323,
+      "learning_rate": 4.3872621959863185e-05,
+      "loss": 0.9068,
+      "step": 6188
+    },
+    {
+      "epoch": 0.5529966269796949,
+      "grad_norm": 0.5276347398757935,
+      "learning_rate": 4.385826057069044e-05,
+      "loss": 0.9264,
+      "step": 6189
+    },
+    {
+      "epoch": 0.5530859785109568,
+      "grad_norm": 0.42190733551979065,
+      "learning_rate": 4.3843899695953826e-05,
+      "loss": 0.941,
+      "step": 6190
+    },
+    {
+      "epoch": 0.5531753300422186,
+      "grad_norm": 0.45186877250671387,
+      "learning_rate": 4.382953933685623e-05,
+      "loss": 0.9799,
+      "step": 6191
+    },
+    {
+      "epoch": 0.5532646815734804,
+      "grad_norm": 0.4708520472049713,
+      "learning_rate": 4.38151794946005e-05,
+      "loss": 0.9633,
+      "step": 6192
+    },
+    {
+      "epoch": 0.5533540331047423,
+      "grad_norm": 0.44095146656036377,
+      "learning_rate": 4.38008201703894e-05,
+      "loss": 0.9377,
+      "step": 6193
+    },
+    {
+      "epoch": 0.5534433846360042,
+      "grad_norm": 0.4277689754962921,
+      "learning_rate": 4.378646136542569e-05,
+      "loss": 0.9737,
+      "step": 6194
+    },
+    {
+      "epoch": 0.5535327361672661,
+      "grad_norm": 0.4015730321407318,
+      "learning_rate": 4.377210308091207e-05,
+      "loss": 0.9792,
+      "step": 6195
+    },
+    {
+      "epoch": 0.5536220876985279,
+      "grad_norm": 0.4145139753818512,
+      "learning_rate": 4.375774531805121e-05,
+      "loss": 0.9336,
+      "step": 6196
+    },
+    {
+      "epoch": 0.5537114392297898,
+      "grad_norm": 0.42907387018203735,
+      "learning_rate": 4.374338807804571e-05,
+      "loss": 0.9789,
+      "step": 6197
+    },
+    {
+      "epoch": 0.5538007907610517,
+      "grad_norm": 0.4848146140575409,
+      "learning_rate": 4.372903136209815e-05,
+      "loss": 0.9244,
+      "step": 6198
+    },
+    {
+      "epoch": 0.5538901422923135,
+      "grad_norm": 0.4261986017227173,
+      "learning_rate": 4.371467517141108e-05,
+      "loss": 0.9835,
+      "step": 6199
+    },
+    {
+      "epoch": 0.5539794938235754,
+      "grad_norm": 0.430756539106369,
+      "learning_rate": 4.3700319507186935e-05,
+      "loss": 1.0051,
+      "step": 6200
+    },
+    {
+      "epoch": 0.5540688453548372,
+      "grad_norm": 0.3806210160255432,
+      "learning_rate": 4.3685964370628193e-05,
+      "loss": 0.9518,
+      "step": 6201
+    },
+    {
+      "epoch": 0.5541581968860991,
+      "grad_norm": 0.44627901911735535,
+      "learning_rate": 4.367160976293723e-05,
+      "loss": 0.9854,
+      "step": 6202
+    },
+    {
+      "epoch": 0.554247548417361,
+      "grad_norm": 0.492567241191864,
+      "learning_rate": 4.3657255685316404e-05,
+      "loss": 0.9096,
+      "step": 6203
+    },
+    {
+      "epoch": 0.5543368999486229,
+      "grad_norm": 0.3693414330482483,
+      "learning_rate": 4.364290213896802e-05,
+      "loss": 0.9909,
+      "step": 6204
+    },
+    {
+      "epoch": 0.5544262514798848,
+      "grad_norm": 0.4952232539653778,
+      "learning_rate": 4.362854912509435e-05,
+      "loss": 0.9532,
+      "step": 6205
+    },
+    {
+      "epoch": 0.5545156030111466,
+      "grad_norm": 0.44463467597961426,
+      "learning_rate": 4.361419664489762e-05,
+      "loss": 0.9138,
+      "step": 6206
+    },
+    {
+      "epoch": 0.5546049545424084,
+      "grad_norm": 0.4538500905036926,
+      "learning_rate": 4.3599844699579964e-05,
+      "loss": 0.9869,
+      "step": 6207
+    },
+    {
+      "epoch": 0.5546943060736703,
+      "grad_norm": 0.4345034658908844,
+      "learning_rate": 4.358549329034355e-05,
+      "loss": 0.8805,
+      "step": 6208
+    },
+    {
+      "epoch": 0.5547836576049322,
+      "grad_norm": 0.48444312810897827,
+      "learning_rate": 4.357114241839045e-05,
+      "loss": 0.9715,
+      "step": 6209
+    },
+    {
+      "epoch": 0.5548730091361941,
+      "grad_norm": 0.46041545271873474,
+      "learning_rate": 4.35567920849227e-05,
+      "loss": 0.9387,
+      "step": 6210
+    },
+    {
+      "epoch": 0.554962360667456,
+      "grad_norm": 0.49429547786712646,
+      "learning_rate": 4.354244229114228e-05,
+      "loss": 0.9641,
+      "step": 6211
+    },
+    {
+      "epoch": 0.5550517121987179,
+      "grad_norm": 0.41394224762916565,
+      "learning_rate": 4.352809303825115e-05,
+      "loss": 0.9933,
+      "step": 6212
+    },
+    {
+      "epoch": 0.5551410637299796,
+      "grad_norm": 0.4146006405353546,
+      "learning_rate": 4.351374432745122e-05,
+      "loss": 0.8814,
+      "step": 6213
+    },
+    {
+      "epoch": 0.5552304152612415,
+      "grad_norm": 0.5136590600013733,
+      "learning_rate": 4.349939615994433e-05,
+      "loss": 0.9119,
+      "step": 6214
+    },
+    {
+      "epoch": 0.5553197667925034,
+      "grad_norm": 0.5504260659217834,
+      "learning_rate": 4.3485048536932314e-05,
+      "loss": 0.9742,
+      "step": 6215
+    },
+    {
+      "epoch": 0.5554091183237653,
+      "grad_norm": 0.5039506554603577,
+      "learning_rate": 4.347070145961692e-05,
+      "loss": 0.9296,
+      "step": 6216
+    },
+    {
+      "epoch": 0.5554984698550272,
+      "grad_norm": 0.46146827936172485,
+      "learning_rate": 4.345635492919988e-05,
+      "loss": 0.9909,
+      "step": 6217
+    },
+    {
+      "epoch": 0.555587821386289,
+      "grad_norm": 0.4409083127975464,
+      "learning_rate": 4.344200894688287e-05,
+      "loss": 0.9877,
+      "step": 6218
+    },
+    {
+      "epoch": 0.5556771729175509,
+      "grad_norm": 0.49056732654571533,
+      "learning_rate": 4.342766351386753e-05,
+      "loss": 0.9094,
+      "step": 6219
+    },
+    {
+      "epoch": 0.5557665244488127,
+      "grad_norm": 0.4934116303920746,
+      "learning_rate": 4.3413318631355403e-05,
+      "loss": 0.9019,
+      "step": 6220
+    },
+    {
+      "epoch": 0.5558558759800746,
+      "grad_norm": 0.4715670943260193,
+      "learning_rate": 4.339897430054806e-05,
+      "loss": 0.9102,
+      "step": 6221
+    },
+    {
+      "epoch": 0.5559452275113365,
+      "grad_norm": 0.41046231985092163,
+      "learning_rate": 4.338463052264697e-05,
+      "loss": 0.9674,
+      "step": 6222
+    },
+    {
+      "epoch": 0.5560345790425983,
+      "grad_norm": 0.39487752318382263,
+      "learning_rate": 4.3370287298853585e-05,
+      "loss": 1.0071,
+      "step": 6223
+    },
+    {
+      "epoch": 0.5561239305738602,
+      "grad_norm": 0.510352611541748,
+      "learning_rate": 4.3355944630369315e-05,
+      "loss": 0.9072,
+      "step": 6224
+    },
+    {
+      "epoch": 0.5562132821051221,
+      "grad_norm": 0.409078985452652,
+      "learning_rate": 4.334160251839551e-05,
+      "loss": 0.9709,
+      "step": 6225
+    },
+    {
+      "epoch": 0.556302633636384,
+      "grad_norm": 0.4232335388660431,
+      "learning_rate": 4.332726096413346e-05,
+      "loss": 0.9738,
+      "step": 6226
+    },
+    {
+      "epoch": 0.5563919851676458,
+      "grad_norm": 0.44725194573402405,
+      "learning_rate": 4.331291996878443e-05,
+      "loss": 0.8612,
+      "step": 6227
+    },
+    {
+      "epoch": 0.5564813366989076,
+      "grad_norm": 0.44963935017585754,
+      "learning_rate": 4.329857953354963e-05,
+      "loss": 0.9369,
+      "step": 6228
+    },
+    {
+      "epoch": 0.5565706882301695,
+      "grad_norm": 0.48645251989364624,
+      "learning_rate": 4.328423965963025e-05,
+      "loss": 0.9099,
+      "step": 6229
+    },
+    {
+      "epoch": 0.5566600397614314,
+      "grad_norm": 0.5652446746826172,
+      "learning_rate": 4.326990034822736e-05,
+      "loss": 0.8716,
+      "step": 6230
+    },
+    {
+      "epoch": 0.5567493912926933,
+      "grad_norm": 0.3996671438217163,
+      "learning_rate": 4.325556160054205e-05,
+      "loss": 0.9745,
+      "step": 6231
+    },
+    {
+      "epoch": 0.5568387428239552,
+      "grad_norm": 0.5085249543190002,
+      "learning_rate": 4.324122341777535e-05,
+      "loss": 1.0347,
+      "step": 6232
+    },
+    {
+      "epoch": 0.5569280943552171,
+      "grad_norm": 0.3927776515483856,
+      "learning_rate": 4.322688580112824e-05,
+      "loss": 1.0099,
+      "step": 6233
+    },
+    {
+      "epoch": 0.5570174458864788,
+      "grad_norm": 0.5127935409545898,
+      "learning_rate": 4.321254875180163e-05,
+      "loss": 1.0354,
+      "step": 6234
+    },
+    {
+      "epoch": 0.5571067974177407,
+      "grad_norm": 0.4555366039276123,
+      "learning_rate": 4.319821227099641e-05,
+      "loss": 0.9839,
+      "step": 6235
+    },
+    {
+      "epoch": 0.5571961489490026,
+      "grad_norm": 0.4541626274585724,
+      "learning_rate": 4.318387635991342e-05,
+      "loss": 0.9915,
+      "step": 6236
+    },
+    {
+      "epoch": 0.5572855004802645,
+      "grad_norm": 0.4412585198879242,
+      "learning_rate": 4.316954101975343e-05,
+      "loss": 1.0239,
+      "step": 6237
+    },
+    {
+      "epoch": 0.5573748520115264,
+      "grad_norm": 0.42731690406799316,
+      "learning_rate": 4.31552062517172e-05,
+      "loss": 0.9534,
+      "step": 6238
+    },
+    {
+      "epoch": 0.5574642035427883,
+      "grad_norm": 0.4002136290073395,
+      "learning_rate": 4.314087205700542e-05,
+      "loss": 0.9607,
+      "step": 6239
+    },
+    {
+      "epoch": 0.5575535550740501,
+      "grad_norm": 0.4313288629055023,
+      "learning_rate": 4.3126538436818704e-05,
+      "loss": 0.926,
+      "step": 6240
+    },
+    {
+      "epoch": 0.5576429066053119,
+      "grad_norm": 0.4170110821723938,
+      "learning_rate": 4.311220539235765e-05,
+      "loss": 0.9685,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5577322581365738,
+      "grad_norm": 0.4707728624343872,
+      "learning_rate": 4.3097872924822816e-05,
+      "loss": 0.9372,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5578216096678357,
+      "grad_norm": 0.5320422649383545,
+      "learning_rate": 4.308354103541471e-05,
+      "loss": 0.99,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5579109611990976,
+      "grad_norm": 0.44399821758270264,
+      "learning_rate": 4.3069209725333756e-05,
+      "loss": 0.9685,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5580003127303594,
+      "grad_norm": 0.5755261182785034,
+      "learning_rate": 4.305487899578036e-05,
+      "loss": 0.9169,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5580896642616213,
+      "grad_norm": 0.4340269863605499,
+      "learning_rate": 4.3040548847954885e-05,
+      "loss": 0.9954,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5581790157928832,
+      "grad_norm": 0.4525034725666046,
+      "learning_rate": 4.3026219283057625e-05,
+      "loss": 0.9127,
+      "step": 6247
+    },
+    {
+      "epoch": 0.558268367324145,
+      "grad_norm": 0.45277920365333557,
+      "learning_rate": 4.301189030228883e-05,
+      "loss": 0.9261,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5583577188554069,
+      "grad_norm": 0.5205567479133606,
+      "learning_rate": 4.299756190684871e-05,
+      "loss": 0.9047,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5584470703866687,
+      "grad_norm": 0.5122022032737732,
+      "learning_rate": 4.2983234097937444e-05,
+      "loss": 0.9429,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5585364219179306,
+      "grad_norm": 0.5380150675773621,
+      "learning_rate": 4.29689068767551e-05,
+      "loss": 0.974,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5586257734491925,
+      "grad_norm": 0.45227405428886414,
+      "learning_rate": 4.295458024450174e-05,
+      "loss": 0.9372,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5587151249804544,
+      "grad_norm": 0.47980716824531555,
+      "learning_rate": 4.2940254202377395e-05,
+      "loss": 0.9399,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5588044765117163,
+      "grad_norm": 0.5136576890945435,
+      "learning_rate": 4.2925928751582e-05,
+      "loss": 0.8619,
+      "step": 6254
+    },
+    {
+      "epoch": 0.558893828042978,
+      "grad_norm": 0.4292888045310974,
+      "learning_rate": 4.291160389331549e-05,
+      "loss": 0.9736,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5589831795742399,
+      "grad_norm": 0.4159403443336487,
+      "learning_rate": 4.28972796287777e-05,
+      "loss": 0.9077,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5590725311055018,
+      "grad_norm": 0.43436843156814575,
+      "learning_rate": 4.2882955959168454e-05,
+      "loss": 0.965,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5591618826367637,
+      "grad_norm": 0.4158000349998474,
+      "learning_rate": 4.286863288568752e-05,
+      "loss": 1.0099,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5592512341680256,
+      "grad_norm": 0.4041937589645386,
+      "learning_rate": 4.2854310409534583e-05,
+      "loss": 0.9982,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5593405856992875,
+      "grad_norm": 0.5290785431861877,
+      "learning_rate": 4.283998853190933e-05,
+      "loss": 0.9658,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5594299372305492,
+      "grad_norm": 0.4550890326499939,
+      "learning_rate": 4.2825667254011346e-05,
+      "loss": 0.9857,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5595192887618111,
+      "grad_norm": 0.5154798030853271,
+      "learning_rate": 4.281134657704022e-05,
+      "loss": 0.8585,
+      "step": 6262
+    },
+    {
+      "epoch": 0.559608640293073,
+      "grad_norm": 0.46254241466522217,
+      "learning_rate": 4.279702650219543e-05,
+      "loss": 0.9572,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5596979918243349,
+      "grad_norm": 0.41614094376564026,
+      "learning_rate": 4.278270703067644e-05,
+      "loss": 0.9475,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5597873433555968,
+      "grad_norm": 0.39605072140693665,
+      "learning_rate": 4.276838816368267e-05,
+      "loss": 0.984,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5598766948868587,
+      "grad_norm": 0.4425526261329651,
+      "learning_rate": 4.275406990241348e-05,
+      "loss": 1.0881,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5599660464181205,
+      "grad_norm": 0.43981900811195374,
+      "learning_rate": 4.273975224806816e-05,
+      "loss": 0.9565,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5600553979493823,
+      "grad_norm": 0.48525822162628174,
+      "learning_rate": 4.272543520184599e-05,
+      "loss": 0.8835,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5601447494806442,
+      "grad_norm": 0.43978825211524963,
+      "learning_rate": 4.271111876494616e-05,
+      "loss": 0.8794,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5602341010119061,
+      "grad_norm": 0.4795457124710083,
+      "learning_rate": 4.2696802938567854e-05,
+      "loss": 0.9744,
+      "step": 6270
+    },
+    {
+      "epoch": 0.560323452543168,
+      "grad_norm": 0.46985960006713867,
+      "learning_rate": 4.2682487723910116e-05,
+      "loss": 0.9602,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5604128040744298,
+      "grad_norm": 0.4838345944881439,
+      "learning_rate": 4.266817312217204e-05,
+      "loss": 1.0128,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5605021556056917,
+      "grad_norm": 0.4208124279975891,
+      "learning_rate": 4.2653859134552616e-05,
+      "loss": 0.9531,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5605915071369536,
+      "grad_norm": 0.36559927463531494,
+      "learning_rate": 4.263954576225079e-05,
+      "loss": 0.9299,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5606808586682154,
+      "grad_norm": 0.6015887260437012,
+      "learning_rate": 4.262523300646546e-05,
+      "loss": 0.8584,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5607702101994773,
+      "grad_norm": 0.6308769583702087,
+      "learning_rate": 4.261092086839549e-05,
+      "loss": 0.8135,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5608595617307391,
+      "grad_norm": 0.46543291211128235,
+      "learning_rate": 4.259660934923965e-05,
+      "loss": 0.9407,
+      "step": 6277
+    },
+    {
+      "epoch": 0.560948913262001,
+      "grad_norm": 0.4775448739528656,
+      "learning_rate": 4.258229845019669e-05,
+      "loss": 0.9693,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5610382647932629,
+      "grad_norm": 0.5649442672729492,
+      "learning_rate": 4.2567988172465304e-05,
+      "loss": 1.0061,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5611276163245248,
+      "grad_norm": 0.5074711441993713,
+      "learning_rate": 4.2553678517244144e-05,
+      "loss": 0.9523,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5612169678557867,
+      "grad_norm": 0.49485981464385986,
+      "learning_rate": 4.253936948573176e-05,
+      "loss": 0.9723,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5613063193870484,
+      "grad_norm": 0.41250553727149963,
+      "learning_rate": 4.2525061079126705e-05,
+      "loss": 0.9542,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5613956709183103,
+      "grad_norm": 0.42612266540527344,
+      "learning_rate": 4.251075329862747e-05,
+      "loss": 1.0109,
+      "step": 6283
+    },
+    {
+      "epoch": 0.5614850224495722,
+      "grad_norm": 0.41712063550949097,
+      "learning_rate": 4.249644614543247e-05,
+      "loss": 0.9561,
+      "step": 6284
+    },
+    {
+      "epoch": 0.5615743739808341,
+      "grad_norm": 0.42624518275260925,
+      "learning_rate": 4.2482139620740084e-05,
+      "loss": 0.9082,
+      "step": 6285
+    },
+    {
+      "epoch": 0.561663725512096,
+      "grad_norm": 0.4930400848388672,
+      "learning_rate": 4.246783372574864e-05,
+      "loss": 0.9108,
+      "step": 6286
+    },
+    {
+      "epoch": 0.5617530770433579,
+      "grad_norm": 0.46004951000213623,
+      "learning_rate": 4.245352846165641e-05,
+      "loss": 1.0204,
+      "step": 6287
+    },
+    {
+      "epoch": 0.5618424285746197,
+      "grad_norm": 0.46289774775505066,
+      "learning_rate": 4.243922382966162e-05,
+      "loss": 0.9439,
+      "step": 6288
+    },
+    {
+      "epoch": 0.5619317801058815,
+      "grad_norm": 0.39655157923698425,
+      "learning_rate": 4.2424919830962414e-05,
+      "loss": 0.9299,
+      "step": 6289
+    },
+    {
+      "epoch": 0.5620211316371434,
+      "grad_norm": 0.53416508436203,
+      "learning_rate": 4.241061646675695e-05,
+      "loss": 1.0126,
+      "step": 6290
+    },
+    {
+      "epoch": 0.5621104831684053,
+      "grad_norm": 0.5086248517036438,
+      "learning_rate": 4.239631373824322e-05,
+      "loss": 0.9606,
+      "step": 6291
+    },
+    {
+      "epoch": 0.5621998346996672,
+      "grad_norm": 0.4238837659358978,
+      "learning_rate": 4.2382011646619265e-05,
+      "loss": 1.0183,
+      "step": 6292
+    },
+    {
+      "epoch": 0.562289186230929,
+      "grad_norm": 0.4114820063114166,
+      "learning_rate": 4.236771019308304e-05,
+      "loss": 0.9271,
+      "step": 6293
+    },
+    {
+      "epoch": 0.5623785377621909,
+      "grad_norm": 0.5694311261177063,
+      "learning_rate": 4.235340937883245e-05,
+      "loss": 0.9144,
+      "step": 6294
+    },
+    {
+      "epoch": 0.5624678892934528,
+      "grad_norm": 0.41458985209465027,
+      "learning_rate": 4.233910920506533e-05,
+      "loss": 0.9604,
+      "step": 6295
+    },
+    {
+      "epoch": 0.5625572408247146,
+      "grad_norm": 0.5260997414588928,
+      "learning_rate": 4.232480967297947e-05,
+      "loss": 0.9441,
+      "step": 6296
+    },
+    {
+      "epoch": 0.5626465923559765,
+      "grad_norm": 0.4400598406791687,
+      "learning_rate": 4.2310510783772605e-05,
+      "loss": 0.9635,
+      "step": 6297
+    },
+    {
+      "epoch": 0.5627359438872384,
+      "grad_norm": 0.6165459156036377,
+      "learning_rate": 4.229621253864243e-05,
+      "loss": 0.8765,
+      "step": 6298
+    },
+    {
+      "epoch": 0.5628252954185002,
+      "grad_norm": 0.535810649394989,
+      "learning_rate": 4.228191493878657e-05,
+      "loss": 0.8623,
+      "step": 6299
+    },
+    {
+      "epoch": 0.5629146469497621,
+      "grad_norm": 0.556769073009491,
+      "learning_rate": 4.2267617985402625e-05,
+      "loss": 0.9318,
+      "step": 6300
+    },
+    {
+      "epoch": 0.563003998481024,
+      "grad_norm": 0.44776806235313416,
+      "learning_rate": 4.225332167968807e-05,
+      "loss": 0.9505,
+      "step": 6301
+    },
+    {
+      "epoch": 0.5630933500122859,
+      "grad_norm": 0.5459929704666138,
+      "learning_rate": 4.2239026022840404e-05,
+      "loss": 0.9728,
+      "step": 6302
+    },
+    {
+      "epoch": 0.5631827015435477,
+      "grad_norm": 0.4663999378681183,
+      "learning_rate": 4.222473101605703e-05,
+      "loss": 0.9274,
+      "step": 6303
+    },
+    {
+      "epoch": 0.5632720530748095,
+      "grad_norm": 0.4985668957233429,
+      "learning_rate": 4.221043666053531e-05,
+      "loss": 0.9777,
+      "step": 6304
+    },
+    {
+      "epoch": 0.5633614046060714,
+      "grad_norm": 0.4315873086452484,
+      "learning_rate": 4.2196142957472554e-05,
+      "loss": 0.9759,
+      "step": 6305
+    },
+    {
+      "epoch": 0.5634507561373333,
+      "grad_norm": 0.45682254433631897,
+      "learning_rate": 4.218184990806601e-05,
+      "loss": 0.9806,
+      "step": 6306
+    },
+    {
+      "epoch": 0.5635401076685952,
+      "grad_norm": 0.5467448234558105,
+      "learning_rate": 4.216755751351287e-05,
+      "loss": 0.8838,
+      "step": 6307
+    },
+    {
+      "epoch": 0.5636294591998571,
+      "grad_norm": 0.4684566557407379,
+      "learning_rate": 4.215326577501028e-05,
+      "loss": 0.8951,
+      "step": 6308
+    },
+    {
+      "epoch": 0.563718810731119,
+      "grad_norm": 0.4865800142288208,
+      "learning_rate": 4.213897469375533e-05,
+      "loss": 0.9647,
+      "step": 6309
+    },
+    {
+      "epoch": 0.5638081622623807,
+      "grad_norm": 0.5342085361480713,
+      "learning_rate": 4.212468427094503e-05,
+      "loss": 0.9468,
+      "step": 6310
+    },
+    {
+      "epoch": 0.5638975137936426,
+      "grad_norm": 0.38983017206192017,
+      "learning_rate": 4.2110394507776377e-05,
+      "loss": 1.0299,
+      "step": 6311
+    },
+    {
+      "epoch": 0.5639868653249045,
+      "grad_norm": 0.42452338337898254,
+      "learning_rate": 4.2096105405446264e-05,
+      "loss": 0.9127,
+      "step": 6312
+    },
+    {
+      "epoch": 0.5640762168561664,
+      "grad_norm": 0.5090858936309814,
+      "learning_rate": 4.2081816965151595e-05,
+      "loss": 0.858,
+      "step": 6313
+    },
+    {
+      "epoch": 0.5641655683874283,
+      "grad_norm": 0.4234965443611145,
+      "learning_rate": 4.206752918808914e-05,
+      "loss": 0.9866,
+      "step": 6314
+    },
+    {
+      "epoch": 0.5642549199186901,
+      "grad_norm": 0.5019885897636414,
+      "learning_rate": 4.205324207545567e-05,
+      "loss": 0.9501,
+      "step": 6315
+    },
+    {
+      "epoch": 0.564344271449952,
+      "grad_norm": 0.5135182738304138,
+      "learning_rate": 4.203895562844789e-05,
+      "loss": 1.013,
+      "step": 6316
+    },
+    {
+      "epoch": 0.5644336229812138,
+      "grad_norm": 0.42812684178352356,
+      "learning_rate": 4.202466984826242e-05,
+      "loss": 1.0301,
+      "step": 6317
+    },
+    {
+      "epoch": 0.5645229745124757,
+      "grad_norm": 0.5810215473175049,
+      "learning_rate": 4.201038473609587e-05,
+      "loss": 0.938,
+      "step": 6318
+    },
+    {
+      "epoch": 0.5646123260437376,
+      "grad_norm": 0.4419316351413727,
+      "learning_rate": 4.199610029314476e-05,
+      "loss": 0.9318,
+      "step": 6319
+    },
+    {
+      "epoch": 0.5647016775749994,
+      "grad_norm": 0.48427048325538635,
+      "learning_rate": 4.198181652060559e-05,
+      "loss": 0.9813,
+      "step": 6320
+    },
+    {
+      "epoch": 0.5647910291062613,
+      "grad_norm": 0.4527822732925415,
+      "learning_rate": 4.196753341967473e-05,
+      "loss": 0.9851,
+      "step": 6321
+    },
+    {
+      "epoch": 0.5648803806375232,
+      "grad_norm": 0.45797282457351685,
+      "learning_rate": 4.195325099154857e-05,
+      "loss": 0.9875,
+      "step": 6322
+    },
+    {
+      "epoch": 0.564969732168785,
+      "grad_norm": 0.5200628042221069,
+      "learning_rate": 4.193896923742341e-05,
+      "loss": 0.9328,
+      "step": 6323
+    },
+    {
+      "epoch": 0.5650590837000469,
+      "grad_norm": 0.4386134147644043,
+      "learning_rate": 4.192468815849549e-05,
+      "loss": 0.9706,
+      "step": 6324
+    },
+    {
+      "epoch": 0.5651484352313088,
+      "grad_norm": 0.4257854223251343,
+      "learning_rate": 4.1910407755961025e-05,
+      "loss": 0.9642,
+      "step": 6325
+    },
+    {
+      "epoch": 0.5652377867625706,
+      "grad_norm": 0.42326584458351135,
+      "learning_rate": 4.189612803101614e-05,
+      "loss": 0.9783,
+      "step": 6326
+    },
+    {
+      "epoch": 0.5653271382938325,
+      "grad_norm": 0.38882023096084595,
+      "learning_rate": 4.188184898485691e-05,
+      "loss": 0.9474,
+      "step": 6327
+    },
+    {
+      "epoch": 0.5654164898250944,
+      "grad_norm": 0.4262528717517853,
+      "learning_rate": 4.186757061867937e-05,
+      "loss": 1.0289,
+      "step": 6328
+    },
+    {
+      "epoch": 0.5655058413563563,
+      "grad_norm": 0.5065363049507141,
+      "learning_rate": 4.185329293367947e-05,
+      "loss": 0.8959,
+      "step": 6329
+    },
+    {
+      "epoch": 0.5655951928876181,
+      "grad_norm": 0.4666295051574707,
+      "learning_rate": 4.1839015931053125e-05,
+      "loss": 0.9883,
+      "step": 6330
+    },
+    {
+      "epoch": 0.5656845444188799,
+      "grad_norm": 0.4304862916469574,
+      "learning_rate": 4.182473961199619e-05,
+      "loss": 0.9752,
+      "step": 6331
+    },
+    {
+      "epoch": 0.5657738959501418,
+      "grad_norm": 0.4546220600605011,
+      "learning_rate": 4.1810463977704464e-05,
+      "loss": 0.9584,
+      "step": 6332
+    },
+    {
+      "epoch": 0.5658632474814037,
+      "grad_norm": 0.4094926714897156,
+      "learning_rate": 4.179618902937365e-05,
+      "loss": 0.957,
+      "step": 6333
+    },
+    {
+      "epoch": 0.5659525990126656,
+      "grad_norm": 0.41204431653022766,
+      "learning_rate": 4.178191476819946e-05,
+      "loss": 1.0065,
+      "step": 6334
+    },
+    {
+      "epoch": 0.5660419505439275,
+      "grad_norm": 0.5892501473426819,
+      "learning_rate": 4.17676411953775e-05,
+      "loss": 1.0111,
+      "step": 6335
+    },
+    {
+      "epoch": 0.5661313020751894,
+      "grad_norm": 0.5121837854385376,
+      "learning_rate": 4.175336831210335e-05,
+      "loss": 0.8916,
+      "step": 6336
+    },
+    {
+      "epoch": 0.5662206536064511,
+      "grad_norm": 0.5115182995796204,
+      "learning_rate": 4.17390961195725e-05,
+      "loss": 0.9181,
+      "step": 6337
+    },
+    {
+      "epoch": 0.566310005137713,
+      "grad_norm": 0.4587537348270416,
+      "learning_rate": 4.172482461898041e-05,
+      "loss": 0.9743,
+      "step": 6338
+    },
+    {
+      "epoch": 0.5663993566689749,
+      "grad_norm": 0.533208966255188,
+      "learning_rate": 4.171055381152246e-05,
+      "loss": 0.9295,
+      "step": 6339
+    },
+    {
+      "epoch": 0.5664887082002368,
+      "grad_norm": 0.4383775591850281,
+      "learning_rate": 4.169628369839399e-05,
+      "loss": 0.9257,
+      "step": 6340
+    },
+    {
+      "epoch": 0.5665780597314987,
+      "grad_norm": 0.572481632232666,
+      "learning_rate": 4.1682014280790294e-05,
+      "loss": 0.9358,
+      "step": 6341
+    },
+    {
+      "epoch": 0.5666674112627605,
+      "grad_norm": 0.4642484486103058,
+      "learning_rate": 4.166774555990654e-05,
+      "loss": 0.9556,
+      "step": 6342
+    },
+    {
+      "epoch": 0.5667567627940224,
+      "grad_norm": 0.4034402668476105,
+      "learning_rate": 4.165347753693791e-05,
+      "loss": 1.0468,
+      "step": 6343
+    },
+    {
+      "epoch": 0.5668461143252842,
+      "grad_norm": 0.5274239778518677,
+      "learning_rate": 4.1639210213079513e-05,
+      "loss": 0.9687,
+      "step": 6344
+    },
+    {
+      "epoch": 0.5669354658565461,
+      "grad_norm": 0.4413708746433258,
+      "learning_rate": 4.162494358952637e-05,
+      "loss": 0.9749,
+      "step": 6345
+    },
+    {
+      "epoch": 0.567024817387808,
+      "grad_norm": 0.4011813700199127,
+      "learning_rate": 4.161067766747349e-05,
+      "loss": 1.012,
+      "step": 6346
+    },
+    {
+      "epoch": 0.5671141689190698,
+      "grad_norm": 0.4661290943622589,
+      "learning_rate": 4.159641244811577e-05,
+      "loss": 0.9554,
+      "step": 6347
+    },
+    {
+      "epoch": 0.5672035204503317,
+      "grad_norm": 0.40306004881858826,
+      "learning_rate": 4.1582147932648074e-05,
+      "loss": 0.9564,
+      "step": 6348
+    },
+    {
+      "epoch": 0.5672928719815936,
+      "grad_norm": 0.4345554709434509,
+      "learning_rate": 4.156788412226522e-05,
+      "loss": 0.9711,
+      "step": 6349
+    },
+    {
+      "epoch": 0.5673822235128555,
+      "grad_norm": 0.4540269076824188,
+      "learning_rate": 4.155362101816196e-05,
+      "loss": 0.91,
+      "step": 6350
+    },
+    {
+      "epoch": 0.5674715750441173,
+      "grad_norm": 0.44446346163749695,
+      "learning_rate": 4.153935862153298e-05,
+      "loss": 0.9575,
+      "step": 6351
+    },
+    {
+      "epoch": 0.5675609265753792,
+      "grad_norm": 0.5166485905647278,
+      "learning_rate": 4.152509693357289e-05,
+      "loss": 0.9601,
+      "step": 6352
+    },
+    {
+      "epoch": 0.567650278106641,
+      "grad_norm": 0.39662766456604004,
+      "learning_rate": 4.1510835955476256e-05,
+      "loss": 1.0281,
+      "step": 6353
+    },
+    {
+      "epoch": 0.5677396296379029,
+      "grad_norm": 0.4909988045692444,
+      "learning_rate": 4.1496575688437605e-05,
+      "loss": 0.9623,
+      "step": 6354
+    },
+    {
+      "epoch": 0.5678289811691648,
+      "grad_norm": 0.5233615040779114,
+      "learning_rate": 4.1482316133651375e-05,
+      "loss": 0.9088,
+      "step": 6355
+    },
+    {
+      "epoch": 0.5679183327004267,
+      "grad_norm": 0.48598557710647583,
+      "learning_rate": 4.146805729231197e-05,
+      "loss": 0.9457,
+      "step": 6356
+    },
+    {
+      "epoch": 0.5680076842316886,
+      "grad_norm": 0.4184345304965973,
+      "learning_rate": 4.145379916561371e-05,
+      "loss": 1.0264,
+      "step": 6357
+    },
+    {
+      "epoch": 0.5680970357629503,
+      "grad_norm": 0.48690488934516907,
+      "learning_rate": 4.143954175475086e-05,
+      "loss": 0.9859,
+      "step": 6358
+    },
+    {
+      "epoch": 0.5681863872942122,
+      "grad_norm": 0.5679425001144409,
+      "learning_rate": 4.142528506091764e-05,
+      "loss": 1.0039,
+      "step": 6359
+    },
+    {
+      "epoch": 0.5682757388254741,
+      "grad_norm": 0.44501417875289917,
+      "learning_rate": 4.141102908530819e-05,
+      "loss": 0.974,
+      "step": 6360
+    },
+    {
+      "epoch": 0.568365090356736,
+      "grad_norm": 0.4163120687007904,
+      "learning_rate": 4.139677382911663e-05,
+      "loss": 0.8965,
+      "step": 6361
+    },
+    {
+      "epoch": 0.5684544418879979,
+      "grad_norm": 0.46128639578819275,
+      "learning_rate": 4.138251929353695e-05,
+      "loss": 0.9975,
+      "step": 6362
+    },
+    {
+      "epoch": 0.5685437934192598,
+      "grad_norm": 0.42836207151412964,
+      "learning_rate": 4.1368265479763127e-05,
+      "loss": 0.9683,
+      "step": 6363
+    },
+    {
+      "epoch": 0.5686331449505216,
+      "grad_norm": 0.4289848208427429,
+      "learning_rate": 4.135401238898908e-05,
+      "loss": 0.9996,
+      "step": 6364
+    },
+    {
+      "epoch": 0.5687224964817834,
+      "grad_norm": 0.5396338105201721,
+      "learning_rate": 4.133976002240867e-05,
+      "loss": 0.929,
+      "step": 6365
+    },
+    {
+      "epoch": 0.5688118480130453,
+      "grad_norm": 0.5357837080955505,
+      "learning_rate": 4.132550838121565e-05,
+      "loss": 0.9172,
+      "step": 6366
+    },
+    {
+      "epoch": 0.5689011995443072,
+      "grad_norm": 0.48109525442123413,
+      "learning_rate": 4.1311257466603774e-05,
+      "loss": 0.9564,
+      "step": 6367
+    },
+    {
+      "epoch": 0.5689905510755691,
+      "grad_norm": 0.5908039212226868,
+      "learning_rate": 4.12970072797667e-05,
+      "loss": 0.9478,
+      "step": 6368
+    },
+    {
+      "epoch": 0.569079902606831,
+      "grad_norm": 0.4044991731643677,
+      "learning_rate": 4.128275782189803e-05,
+      "loss": 0.9789,
+      "step": 6369
+    },
+    {
+      "epoch": 0.5691692541380928,
+      "grad_norm": 0.44325581192970276,
+      "learning_rate": 4.1268509094191315e-05,
+      "loss": 0.9043,
+      "step": 6370
+    },
+    {
+      "epoch": 0.5692586056693547,
+      "grad_norm": 0.37263184785842896,
+      "learning_rate": 4.125426109784006e-05,
+      "loss": 1.0072,
+      "step": 6371
+    },
+    {
+      "epoch": 0.5693479572006165,
+      "grad_norm": 0.4318254888057709,
+      "learning_rate": 4.1240013834037626e-05,
+      "loss": 0.8824,
+      "step": 6372
+    },
+    {
+      "epoch": 0.5694373087318784,
+      "grad_norm": 0.49909618496894836,
+      "learning_rate": 4.122576730397742e-05,
+      "loss": 0.9341,
+      "step": 6373
+    },
+    {
+      "epoch": 0.5695266602631402,
+      "grad_norm": 0.6334441304206848,
+      "learning_rate": 4.1211521508852726e-05,
+      "loss": 0.8383,
+      "step": 6374
+    },
+    {
+      "epoch": 0.5696160117944021,
+      "grad_norm": 0.5850768685340881,
+      "learning_rate": 4.119727644985678e-05,
+      "loss": 0.88,
+      "step": 6375
+    },
+    {
+      "epoch": 0.569705363325664,
+      "grad_norm": 0.48001495003700256,
+      "learning_rate": 4.1183032128182766e-05,
+      "loss": 0.9366,
+      "step": 6376
+    },
+    {
+      "epoch": 0.5697947148569259,
+      "grad_norm": 0.504117488861084,
+      "learning_rate": 4.1168788545023796e-05,
+      "loss": 0.9801,
+      "step": 6377
+    },
+    {
+      "epoch": 0.5698840663881878,
+      "grad_norm": 0.4841177761554718,
+      "learning_rate": 4.115454570157291e-05,
+      "loss": 0.9103,
+      "step": 6378
+    },
+    {
+      "epoch": 0.5699734179194496,
+      "grad_norm": 0.475758820772171,
+      "learning_rate": 4.114030359902313e-05,
+      "loss": 0.9221,
+      "step": 6379
+    },
+    {
+      "epoch": 0.5700627694507114,
+      "grad_norm": 0.4464789927005768,
+      "learning_rate": 4.112606223856734e-05,
+      "loss": 1.0044,
+      "step": 6380
+    },
+    {
+      "epoch": 0.5701521209819733,
+      "grad_norm": 0.4405708312988281,
+      "learning_rate": 4.1111821621398446e-05,
+      "loss": 0.9311,
+      "step": 6381
+    },
+    {
+      "epoch": 0.5702414725132352,
+      "grad_norm": 0.49968060851097107,
+      "learning_rate": 4.109758174870921e-05,
+      "loss": 0.898,
+      "step": 6382
+    },
+    {
+      "epoch": 0.5703308240444971,
+      "grad_norm": 0.4581187963485718,
+      "learning_rate": 4.10833426216924e-05,
+      "loss": 0.95,
+      "step": 6383
+    },
+    {
+      "epoch": 0.570420175575759,
+      "grad_norm": 0.46304047107696533,
+      "learning_rate": 4.1069104241540715e-05,
+      "loss": 0.9737,
+      "step": 6384
+    },
+    {
+      "epoch": 0.5705095271070207,
+      "grad_norm": 0.5021114349365234,
+      "learning_rate": 4.105486660944672e-05,
+      "loss": 0.9034,
+      "step": 6385
+    },
+    {
+      "epoch": 0.5705988786382826,
+      "grad_norm": 0.5141803622245789,
+      "learning_rate": 4.1040629726602996e-05,
+      "loss": 0.9224,
+      "step": 6386
+    },
+    {
+      "epoch": 0.5706882301695445,
+      "grad_norm": 0.4306526482105255,
+      "learning_rate": 4.102639359420204e-05,
+      "loss": 0.9798,
+      "step": 6387
+    },
+    {
+      "epoch": 0.5707775817008064,
+      "grad_norm": 0.5190479755401611,
+      "learning_rate": 4.101215821343626e-05,
+      "loss": 0.9899,
+      "step": 6388
+    },
+    {
+      "epoch": 0.5708669332320683,
+      "grad_norm": 0.5050314664840698,
+      "learning_rate": 4.0997923585498046e-05,
+      "loss": 0.935,
+      "step": 6389
+    },
+    {
+      "epoch": 0.5709562847633302,
+      "grad_norm": 0.47217750549316406,
+      "learning_rate": 4.098368971157968e-05,
+      "loss": 0.9683,
+      "step": 6390
+    },
+    {
+      "epoch": 0.571045636294592,
+      "grad_norm": 0.43386927247047424,
+      "learning_rate": 4.0969456592873435e-05,
+      "loss": 0.925,
+      "step": 6391
+    },
+    {
+      "epoch": 0.5711349878258538,
+      "grad_norm": 0.577601969242096,
+      "learning_rate": 4.095522423057143e-05,
+      "loss": 0.9045,
+      "step": 6392
+    },
+    {
+      "epoch": 0.5712243393571157,
+      "grad_norm": 0.38935279846191406,
+      "learning_rate": 4.094099262586581e-05,
+      "loss": 0.9945,
+      "step": 6393
+    },
+    {
+      "epoch": 0.5713136908883776,
+      "grad_norm": 0.4525635838508606,
+      "learning_rate": 4.092676177994862e-05,
+      "loss": 1.0153,
+      "step": 6394
+    },
+    {
+      "epoch": 0.5714030424196395,
+      "grad_norm": 0.4267630875110626,
+      "learning_rate": 4.091253169401184e-05,
+      "loss": 1.0024,
+      "step": 6395
+    },
+    {
+      "epoch": 0.5714923939509013,
+      "grad_norm": 0.47713613510131836,
+      "learning_rate": 4.0898302369247405e-05,
+      "loss": 0.9021,
+      "step": 6396
+    },
+    {
+      "epoch": 0.5715817454821632,
+      "grad_norm": 0.4627811908721924,
+      "learning_rate": 4.088407380684715e-05,
+      "loss": 0.9642,
+      "step": 6397
+    },
+    {
+      "epoch": 0.5716710970134251,
+      "grad_norm": 0.5233446955680847,
+      "learning_rate": 4.086984600800291e-05,
+      "loss": 0.9229,
+      "step": 6398
+    },
+    {
+      "epoch": 0.5717604485446869,
+      "grad_norm": 0.4403021037578583,
+      "learning_rate": 4.0855618973906365e-05,
+      "loss": 0.9546,
+      "step": 6399
+    },
+    {
+      "epoch": 0.5718498000759488,
+      "grad_norm": 0.42531147599220276,
+      "learning_rate": 4.08413927057492e-05,
+      "loss": 1.0129,
+      "step": 6400
+    },
+    {
+      "epoch": 0.5719391516072106,
+      "grad_norm": 0.46862322092056274,
+      "learning_rate": 4.082716720472304e-05,
+      "loss": 0.9922,
+      "step": 6401
+    },
+    {
+      "epoch": 0.5720285031384725,
+      "grad_norm": 0.43281012773513794,
+      "learning_rate": 4.08129424720194e-05,
+      "loss": 0.9253,
+      "step": 6402
+    },
+    {
+      "epoch": 0.5721178546697344,
+      "grad_norm": 0.44874048233032227,
+      "learning_rate": 4.079871850882975e-05,
+      "loss": 0.9576,
+      "step": 6403
+    },
+    {
+      "epoch": 0.5722072062009963,
+      "grad_norm": 0.3677184283733368,
+      "learning_rate": 4.0784495316345496e-05,
+      "loss": 1.0069,
+      "step": 6404
+    },
+    {
+      "epoch": 0.5722965577322582,
+      "grad_norm": 0.5108903050422668,
+      "learning_rate": 4.077027289575799e-05,
+      "loss": 0.9036,
+      "step": 6405
+    },
+    {
+      "epoch": 0.57238590926352,
+      "grad_norm": 0.5099184513092041,
+      "learning_rate": 4.0756051248258506e-05,
+      "loss": 0.9907,
+      "step": 6406
+    },
+    {
+      "epoch": 0.5724752607947818,
+      "grad_norm": 0.506807804107666,
+      "learning_rate": 4.074183037503827e-05,
+      "loss": 0.8992,
+      "step": 6407
+    },
+    {
+      "epoch": 0.5725646123260437,
+      "grad_norm": 0.4727827310562134,
+      "learning_rate": 4.072761027728842e-05,
+      "loss": 0.9714,
+      "step": 6408
+    },
+    {
+      "epoch": 0.5726539638573056,
+      "grad_norm": 0.41330844163894653,
+      "learning_rate": 4.0713390956200046e-05,
+      "loss": 0.9732,
+      "step": 6409
+    },
+    {
+      "epoch": 0.5727433153885675,
+      "grad_norm": 0.5296694040298462,
+      "learning_rate": 4.069917241296417e-05,
+      "loss": 1.0646,
+      "step": 6410
+    },
+    {
+      "epoch": 0.5728326669198294,
+      "grad_norm": 0.5255220532417297,
+      "learning_rate": 4.068495464877177e-05,
+      "loss": 0.967,
+      "step": 6411
+    },
+    {
+      "epoch": 0.5729220184510913,
+      "grad_norm": 0.5186451077461243,
+      "learning_rate": 4.067073766481369e-05,
+      "loss": 0.9345,
+      "step": 6412
+    },
+    {
+      "epoch": 0.573011369982353,
+      "grad_norm": 0.5088133215904236,
+      "learning_rate": 4.0656521462280764e-05,
+      "loss": 0.8783,
+      "step": 6413
+    },
+    {
+      "epoch": 0.5731007215136149,
+      "grad_norm": 0.4002193510532379,
+      "learning_rate": 4.064230604236376e-05,
+      "loss": 0.975,
+      "step": 6414
+    },
+    {
+      "epoch": 0.5731900730448768,
+      "grad_norm": 0.4866524040699005,
+      "learning_rate": 4.062809140625338e-05,
+      "loss": 0.8875,
+      "step": 6415
+    },
+    {
+      "epoch": 0.5732794245761387,
+      "grad_norm": 0.45396822690963745,
+      "learning_rate": 4.061387755514024e-05,
+      "loss": 0.9482,
+      "step": 6416
+    },
+    {
+      "epoch": 0.5733687761074006,
+      "grad_norm": 0.5568546056747437,
+      "learning_rate": 4.05996644902149e-05,
+      "loss": 0.9384,
+      "step": 6417
+    },
+    {
+      "epoch": 0.5734581276386624,
+      "grad_norm": 0.4871857464313507,
+      "learning_rate": 4.0585452212667864e-05,
+      "loss": 0.9945,
+      "step": 6418
+    },
+    {
+      "epoch": 0.5735474791699243,
+      "grad_norm": 0.4038003385066986,
+      "learning_rate": 4.0571240723689546e-05,
+      "loss": 0.9829,
+      "step": 6419
+    },
+    {
+      "epoch": 0.5736368307011861,
+      "grad_norm": 0.39206719398498535,
+      "learning_rate": 4.055703002447033e-05,
+      "loss": 1.0198,
+      "step": 6420
+    },
+    {
+      "epoch": 0.573726182232448,
+      "grad_norm": 0.37577176094055176,
+      "learning_rate": 4.0542820116200495e-05,
+      "loss": 0.9572,
+      "step": 6421
+    },
+    {
+      "epoch": 0.5738155337637099,
+      "grad_norm": 0.3899545669555664,
+      "learning_rate": 4.052861100007032e-05,
+      "loss": 1.035,
+      "step": 6422
+    },
+    {
+      "epoch": 0.5739048852949717,
+      "grad_norm": 0.6146190762519836,
+      "learning_rate": 4.051440267726989e-05,
+      "loss": 0.91,
+      "step": 6423
+    },
+    {
+      "epoch": 0.5739942368262336,
+      "grad_norm": 0.4568648338317871,
+      "learning_rate": 4.050019514898936e-05,
+      "loss": 0.9665,
+      "step": 6424
+    },
+    {
+      "epoch": 0.5740835883574955,
+      "grad_norm": 0.4437832534313202,
+      "learning_rate": 4.048598841641874e-05,
+      "loss": 0.9733,
+      "step": 6425
+    },
+    {
+      "epoch": 0.5741729398887574,
+      "grad_norm": 0.4725315272808075,
+      "learning_rate": 4.0471782480748e-05,
+      "loss": 0.947,
+      "step": 6426
+    },
+    {
+      "epoch": 0.5742622914200192,
+      "grad_norm": 0.5983218550682068,
+      "learning_rate": 4.0457577343167044e-05,
+      "loss": 0.9638,
+      "step": 6427
+    },
+    {
+      "epoch": 0.574351642951281,
+      "grad_norm": 0.42528533935546875,
+      "learning_rate": 4.04433730048657e-05,
+      "loss": 0.9599,
+      "step": 6428
+    },
+    {
+      "epoch": 0.5744409944825429,
+      "grad_norm": 0.4161036014556885,
+      "learning_rate": 4.042916946703373e-05,
+      "loss": 0.9122,
+      "step": 6429
+    },
+    {
+      "epoch": 0.5745303460138048,
+      "grad_norm": 0.418058305978775,
+      "learning_rate": 4.0414966730860846e-05,
+      "loss": 0.9734,
+      "step": 6430
+    },
+    {
+      "epoch": 0.5746196975450667,
+      "grad_norm": 0.5074496865272522,
+      "learning_rate": 4.0400764797536675e-05,
+      "loss": 0.8858,
+      "step": 6431
+    },
+    {
+      "epoch": 0.5747090490763286,
+      "grad_norm": 0.5379347801208496,
+      "learning_rate": 4.038656366825076e-05,
+      "loss": 0.9721,
+      "step": 6432
+    },
+    {
+      "epoch": 0.5747984006075905,
+      "grad_norm": 0.5281144380569458,
+      "learning_rate": 4.037236334419261e-05,
+      "loss": 1.0213,
+      "step": 6433
+    },
+    {
+      "epoch": 0.5748877521388522,
+      "grad_norm": 0.4760182201862335,
+      "learning_rate": 4.035816382655165e-05,
+      "loss": 0.8549,
+      "step": 6434
+    },
+    {
+      "epoch": 0.5749771036701141,
+      "grad_norm": 0.4302056133747101,
+      "learning_rate": 4.034396511651726e-05,
+      "loss": 0.9301,
+      "step": 6435
+    },
+    {
+      "epoch": 0.575066455201376,
+      "grad_norm": 0.4959314167499542,
+      "learning_rate": 4.032976721527869e-05,
+      "loss": 0.889,
+      "step": 6436
+    },
+    {
+      "epoch": 0.5751558067326379,
+      "grad_norm": 0.5147891044616699,
+      "learning_rate": 4.0315570124025216e-05,
+      "loss": 0.9727,
+      "step": 6437
+    },
+    {
+      "epoch": 0.5752451582638998,
+      "grad_norm": 0.42699047923088074,
+      "learning_rate": 4.030137384394595e-05,
+      "loss": 1.0242,
+      "step": 6438
+    },
+    {
+      "epoch": 0.5753345097951617,
+      "grad_norm": 0.5570961833000183,
+      "learning_rate": 4.028717837623002e-05,
+      "loss": 0.8377,
+      "step": 6439
+    },
+    {
+      "epoch": 0.5754238613264235,
+      "grad_norm": 0.5209365487098694,
+      "learning_rate": 4.0272983722066435e-05,
+      "loss": 0.9127,
+      "step": 6440
+    },
+    {
+      "epoch": 0.5755132128576853,
+      "grad_norm": 0.4663841128349304,
+      "learning_rate": 4.0258789882644135e-05,
+      "loss": 0.9139,
+      "step": 6441
+    },
+    {
+      "epoch": 0.5756025643889472,
+      "grad_norm": 0.5318526029586792,
+      "learning_rate": 4.024459685915204e-05,
+      "loss": 0.9117,
+      "step": 6442
+    },
+    {
+      "epoch": 0.5756919159202091,
+      "grad_norm": 0.455567866563797,
+      "learning_rate": 4.023040465277892e-05,
+      "loss": 0.8577,
+      "step": 6443
+    },
+    {
+      "epoch": 0.575781267451471,
+      "grad_norm": 0.4229848086833954,
+      "learning_rate": 4.0216213264713556e-05,
+      "loss": 0.9288,
+      "step": 6444
+    },
+    {
+      "epoch": 0.5758706189827328,
+      "grad_norm": 0.38377058506011963,
+      "learning_rate": 4.020202269614461e-05,
+      "loss": 0.9909,
+      "step": 6445
+    },
+    {
+      "epoch": 0.5759599705139947,
+      "grad_norm": 0.48710566759109497,
+      "learning_rate": 4.0187832948260705e-05,
+      "loss": 0.9682,
+      "step": 6446
+    },
+    {
+      "epoch": 0.5760493220452565,
+      "grad_norm": 0.3837445378303528,
+      "learning_rate": 4.017364402225038e-05,
+      "loss": 0.9965,
+      "step": 6447
+    },
+    {
+      "epoch": 0.5761386735765184,
+      "grad_norm": 0.4128133952617645,
+      "learning_rate": 4.0159455919302114e-05,
+      "loss": 0.9991,
+      "step": 6448
+    },
+    {
+      "epoch": 0.5762280251077803,
+      "grad_norm": 0.44364961981773376,
+      "learning_rate": 4.014526864060432e-05,
+      "loss": 1.0147,
+      "step": 6449
+    },
+    {
+      "epoch": 0.5763173766390421,
+      "grad_norm": 0.47115516662597656,
+      "learning_rate": 4.01310821873453e-05,
+      "loss": 0.9597,
+      "step": 6450
+    },
+    {
+      "epoch": 0.576406728170304,
+      "grad_norm": 0.4687146842479706,
+      "learning_rate": 4.0116896560713346e-05,
+      "loss": 0.9166,
+      "step": 6451
+    },
+    {
+      "epoch": 0.5764960797015659,
+      "grad_norm": 0.46051499247550964,
+      "learning_rate": 4.010271176189666e-05,
+      "loss": 0.9735,
+      "step": 6452
+    },
+    {
+      "epoch": 0.5765854312328278,
+      "grad_norm": 0.45504701137542725,
+      "learning_rate": 4.008852779208336e-05,
+      "loss": 0.9643,
+      "step": 6453
+    },
+    {
+      "epoch": 0.5766747827640896,
+      "grad_norm": 0.5997002124786377,
+      "learning_rate": 4.007434465246151e-05,
+      "loss": 0.9214,
+      "step": 6454
+    },
+    {
+      "epoch": 0.5767641342953514,
+      "grad_norm": 0.5700611472129822,
+      "learning_rate": 4.006016234421908e-05,
+      "loss": 1.0074,
+      "step": 6455
+    },
+    {
+      "epoch": 0.5768534858266133,
+      "grad_norm": 0.5041797757148743,
+      "learning_rate": 4.0045980868544014e-05,
+      "loss": 0.9945,
+      "step": 6456
+    },
+    {
+      "epoch": 0.5769428373578752,
+      "grad_norm": 0.39958614110946655,
+      "learning_rate": 4.003180022662415e-05,
+      "loss": 1.0317,
+      "step": 6457
+    },
+    {
+      "epoch": 0.5770321888891371,
+      "grad_norm": 0.5589715242385864,
+      "learning_rate": 4.001762041964727e-05,
+      "loss": 0.8559,
+      "step": 6458
+    },
+    {
+      "epoch": 0.577121540420399,
+      "grad_norm": 0.4533728361129761,
+      "learning_rate": 4.000344144880108e-05,
+      "loss": 0.9256,
+      "step": 6459
+    },
+    {
+      "epoch": 0.5772108919516609,
+      "grad_norm": 0.5269187688827515,
+      "learning_rate": 3.998926331527323e-05,
+      "loss": 0.9339,
+      "step": 6460
+    },
+    {
+      "epoch": 0.5773002434829226,
+      "grad_norm": 0.4693765938282013,
+      "learning_rate": 3.997508602025128e-05,
+      "loss": 0.9535,
+      "step": 6461
+    },
+    {
+      "epoch": 0.5773895950141845,
+      "grad_norm": 0.42026007175445557,
+      "learning_rate": 3.996090956492275e-05,
+      "loss": 0.9109,
+      "step": 6462
+    },
+    {
+      "epoch": 0.5774789465454464,
+      "grad_norm": 0.41388946771621704,
+      "learning_rate": 3.994673395047505e-05,
+      "loss": 0.9696,
+      "step": 6463
+    },
+    {
+      "epoch": 0.5775682980767083,
+      "grad_norm": 0.3982580304145813,
+      "learning_rate": 3.993255917809553e-05,
+      "loss": 1.0228,
+      "step": 6464
+    },
+    {
+      "epoch": 0.5776576496079702,
+      "grad_norm": 0.4337417483329773,
+      "learning_rate": 3.9918385248971484e-05,
+      "loss": 0.996,
+      "step": 6465
+    },
+    {
+      "epoch": 0.577747001139232,
+      "grad_norm": 0.5209522843360901,
+      "learning_rate": 3.990421216429014e-05,
+      "loss": 0.8832,
+      "step": 6466
+    },
+    {
+      "epoch": 0.5778363526704939,
+      "grad_norm": 0.4163426160812378,
+      "learning_rate": 3.9890039925238645e-05,
+      "loss": 0.9445,
+      "step": 6467
+    },
+    {
+      "epoch": 0.5779257042017557,
+      "grad_norm": 0.4440252482891083,
+      "learning_rate": 3.987586853300408e-05,
+      "loss": 0.9476,
+      "step": 6468
+    },
+    {
+      "epoch": 0.5780150557330176,
+      "grad_norm": 0.41356053948402405,
+      "learning_rate": 3.9861697988773425e-05,
+      "loss": 0.9586,
+      "step": 6469
+    },
+    {
+      "epoch": 0.5781044072642795,
+      "grad_norm": 0.46509677171707153,
+      "learning_rate": 3.9847528293733636e-05,
+      "loss": 0.956,
+      "step": 6470
+    },
+    {
+      "epoch": 0.5781937587955414,
+      "grad_norm": 0.4374508261680603,
+      "learning_rate": 3.9833359449071564e-05,
+      "loss": 0.9895,
+      "step": 6471
+    },
+    {
+      "epoch": 0.5782831103268032,
+      "grad_norm": 0.4301753044128418,
+      "learning_rate": 3.981919145597404e-05,
+      "loss": 1.0006,
+      "step": 6472
+    },
+    {
+      "epoch": 0.5783724618580651,
+      "grad_norm": 0.4043073356151581,
+      "learning_rate": 3.9805024315627714e-05,
+      "loss": 0.9939,
+      "step": 6473
+    },
+    {
+      "epoch": 0.578461813389327,
+      "grad_norm": 0.4929652512073517,
+      "learning_rate": 3.979085802921928e-05,
+      "loss": 0.808,
+      "step": 6474
+    },
+    {
+      "epoch": 0.5785511649205888,
+      "grad_norm": 0.3793524205684662,
+      "learning_rate": 3.977669259793531e-05,
+      "loss": 0.9753,
+      "step": 6475
+    },
+    {
+      "epoch": 0.5786405164518507,
+      "grad_norm": 0.4306744635105133,
+      "learning_rate": 3.9762528022962305e-05,
+      "loss": 0.9424,
+      "step": 6476
+    },
+    {
+      "epoch": 0.5787298679831125,
+      "grad_norm": 0.5743783712387085,
+      "learning_rate": 3.9748364305486703e-05,
+      "loss": 0.8385,
+      "step": 6477
+    },
+    {
+      "epoch": 0.5788192195143744,
+      "grad_norm": 0.46976256370544434,
+      "learning_rate": 3.9734201446694865e-05,
+      "loss": 0.9779,
+      "step": 6478
+    },
+    {
+      "epoch": 0.5789085710456363,
+      "grad_norm": 0.3878982365131378,
+      "learning_rate": 3.972003944777308e-05,
+      "loss": 1.0064,
+      "step": 6479
+    },
+    {
+      "epoch": 0.5789979225768982,
+      "grad_norm": 0.3860369324684143,
+      "learning_rate": 3.9705878309907565e-05,
+      "loss": 0.9581,
+      "step": 6480
+    },
+    {
+      "epoch": 0.5790872741081601,
+      "grad_norm": 0.582404375076294,
+      "learning_rate": 3.969171803428447e-05,
+      "loss": 0.8623,
+      "step": 6481
+    },
+    {
+      "epoch": 0.5791766256394218,
+      "grad_norm": 0.5933336019515991,
+      "learning_rate": 3.96775586220899e-05,
+      "loss": 0.8301,
+      "step": 6482
+    },
+    {
+      "epoch": 0.5792659771706837,
+      "grad_norm": 0.4246273934841156,
+      "learning_rate": 3.9663400074509786e-05,
+      "loss": 0.9505,
+      "step": 6483
+    },
+    {
+      "epoch": 0.5793553287019456,
+      "grad_norm": 0.4642765522003174,
+      "learning_rate": 3.964924239273011e-05,
+      "loss": 0.9685,
+      "step": 6484
+    },
+    {
+      "epoch": 0.5794446802332075,
+      "grad_norm": 0.4278883635997772,
+      "learning_rate": 3.96350855779367e-05,
+      "loss": 1.0149,
+      "step": 6485
+    },
+    {
+      "epoch": 0.5795340317644694,
+      "grad_norm": 0.427097350358963,
+      "learning_rate": 3.962092963131537e-05,
+      "loss": 0.9337,
+      "step": 6486
+    },
+    {
+      "epoch": 0.5796233832957313,
+      "grad_norm": 0.5278881788253784,
+      "learning_rate": 3.9606774554051824e-05,
+      "loss": 0.9693,
+      "step": 6487
+    },
+    {
+      "epoch": 0.5797127348269931,
+      "grad_norm": 0.42778128385543823,
+      "learning_rate": 3.959262034733168e-05,
+      "loss": 0.984,
+      "step": 6488
+    },
+    {
+      "epoch": 0.5798020863582549,
+      "grad_norm": 0.6493330597877502,
+      "learning_rate": 3.9578467012340515e-05,
+      "loss": 0.8847,
+      "step": 6489
+    },
+    {
+      "epoch": 0.5798914378895168,
+      "grad_norm": 0.4391600787639618,
+      "learning_rate": 3.956431455026382e-05,
+      "loss": 1.0071,
+      "step": 6490
+    },
+    {
+      "epoch": 0.5799807894207787,
+      "grad_norm": 0.4158203899860382,
+      "learning_rate": 3.955016296228702e-05,
+      "loss": 0.967,
+      "step": 6491
+    },
+    {
+      "epoch": 0.5800701409520406,
+      "grad_norm": 0.41016045212745667,
+      "learning_rate": 3.953601224959549e-05,
+      "loss": 0.9616,
+      "step": 6492
+    },
+    {
+      "epoch": 0.5801594924833025,
+      "grad_norm": 0.420639306306839,
+      "learning_rate": 3.952186241337444e-05,
+      "loss": 0.9555,
+      "step": 6493
+    },
+    {
+      "epoch": 0.5802488440145643,
+      "grad_norm": 0.4078323543071747,
+      "learning_rate": 3.9507713454809106e-05,
+      "loss": 0.9645,
+      "step": 6494
+    },
+    {
+      "epoch": 0.5803381955458262,
+      "grad_norm": 0.4655108153820038,
+      "learning_rate": 3.949356537508461e-05,
+      "loss": 0.8071,
+      "step": 6495
+    },
+    {
+      "epoch": 0.580427547077088,
+      "grad_norm": 0.4758930504322052,
+      "learning_rate": 3.947941817538601e-05,
+      "loss": 0.8597,
+      "step": 6496
+    },
+    {
+      "epoch": 0.5805168986083499,
+      "grad_norm": 0.41579219698905945,
+      "learning_rate": 3.946527185689827e-05,
+      "loss": 0.9475,
+      "step": 6497
+    },
+    {
+      "epoch": 0.5806062501396118,
+      "grad_norm": 0.4709773659706116,
+      "learning_rate": 3.9451126420806304e-05,
+      "loss": 0.9533,
+      "step": 6498
+    },
+    {
+      "epoch": 0.5806956016708736,
+      "grad_norm": 0.572155773639679,
+      "learning_rate": 3.943698186829495e-05,
+      "loss": 0.932,
+      "step": 6499
+    },
+    {
+      "epoch": 0.5807849532021355,
+      "grad_norm": 0.4728442132472992,
+      "learning_rate": 3.942283820054895e-05,
+      "loss": 1.0003,
+      "step": 6500
+    },
+    {
+      "epoch": 0.5808743047333974,
+      "grad_norm": 0.504787266254425,
+      "learning_rate": 3.940869541875301e-05,
+      "loss": 0.9066,
+      "step": 6501
+    },
+    {
+      "epoch": 0.5809636562646593,
+      "grad_norm": 0.4003005921840668,
+      "learning_rate": 3.939455352409172e-05,
+      "loss": 0.9711,
+      "step": 6502
+    },
+    {
+      "epoch": 0.5810530077959211,
+      "grad_norm": 0.4955715537071228,
+      "learning_rate": 3.9380412517749613e-05,
+      "loss": 0.9386,
+      "step": 6503
+    },
+    {
+      "epoch": 0.581142359327183,
+      "grad_norm": 0.4974922835826874,
+      "learning_rate": 3.9366272400911156e-05,
+      "loss": 0.9126,
+      "step": 6504
+    },
+    {
+      "epoch": 0.5812317108584448,
+      "grad_norm": 0.444976806640625,
+      "learning_rate": 3.935213317476074e-05,
+      "loss": 1.0139,
+      "step": 6505
+    },
+    {
+      "epoch": 0.5813210623897067,
+      "grad_norm": 0.47083789110183716,
+      "learning_rate": 3.9337994840482664e-05,
+      "loss": 0.9698,
+      "step": 6506
+    },
+    {
+      "epoch": 0.5814104139209686,
+      "grad_norm": 0.3796704411506653,
+      "learning_rate": 3.932385739926116e-05,
+      "loss": 0.9525,
+      "step": 6507
+    },
+    {
+      "epoch": 0.5814997654522305,
+      "grad_norm": 0.4334143400192261,
+      "learning_rate": 3.93097208522804e-05,
+      "loss": 0.9909,
+      "step": 6508
+    },
+    {
+      "epoch": 0.5815891169834922,
+      "grad_norm": 0.47649437189102173,
+      "learning_rate": 3.929558520072447e-05,
+      "loss": 0.9077,
+      "step": 6509
+    },
+    {
+      "epoch": 0.5816784685147541,
+      "grad_norm": 0.6175134181976318,
+      "learning_rate": 3.928145044577738e-05,
+      "loss": 0.9106,
+      "step": 6510
+    },
+    {
+      "epoch": 0.581767820046016,
+      "grad_norm": 0.4393305480480194,
+      "learning_rate": 3.926731658862307e-05,
+      "loss": 0.9112,
+      "step": 6511
+    },
+    {
+      "epoch": 0.5818571715772779,
+      "grad_norm": 0.46963876485824585,
+      "learning_rate": 3.9253183630445395e-05,
+      "loss": 0.9079,
+      "step": 6512
+    },
+    {
+      "epoch": 0.5819465231085398,
+      "grad_norm": 0.4847874343395233,
+      "learning_rate": 3.923905157242817e-05,
+      "loss": 0.9349,
+      "step": 6513
+    },
+    {
+      "epoch": 0.5820358746398017,
+      "grad_norm": 0.40541452169418335,
+      "learning_rate": 3.922492041575505e-05,
+      "loss": 0.9665,
+      "step": 6514
+    },
+    {
+      "epoch": 0.5821252261710635,
+      "grad_norm": 0.4827271103858948,
+      "learning_rate": 3.92107901616097e-05,
+      "loss": 0.9835,
+      "step": 6515
+    },
+    {
+      "epoch": 0.5822145777023253,
+      "grad_norm": 0.5301873087882996,
+      "learning_rate": 3.9196660811175685e-05,
+      "loss": 0.8416,
+      "step": 6516
+    },
+    {
+      "epoch": 0.5823039292335872,
+      "grad_norm": 0.557282030582428,
+      "learning_rate": 3.918253236563648e-05,
+      "loss": 0.8892,
+      "step": 6517
+    },
+    {
+      "epoch": 0.5823932807648491,
+      "grad_norm": 0.584327220916748,
+      "learning_rate": 3.9168404826175486e-05,
+      "loss": 0.9241,
+      "step": 6518
+    },
+    {
+      "epoch": 0.582482632296111,
+      "grad_norm": 0.42827340960502625,
+      "learning_rate": 3.9154278193976066e-05,
+      "loss": 0.92,
+      "step": 6519
+    },
+    {
+      "epoch": 0.5825719838273729,
+      "grad_norm": 0.43465957045555115,
+      "learning_rate": 3.914015247022144e-05,
+      "loss": 0.9636,
+      "step": 6520
+    },
+    {
+      "epoch": 0.5826613353586347,
+      "grad_norm": 0.4111988842487335,
+      "learning_rate": 3.9126027656094806e-05,
+      "loss": 0.9871,
+      "step": 6521
+    },
+    {
+      "epoch": 0.5827506868898966,
+      "grad_norm": 0.4078396260738373,
+      "learning_rate": 3.9111903752779263e-05,
+      "loss": 0.9392,
+      "step": 6522
+    },
+    {
+      "epoch": 0.5828400384211584,
+      "grad_norm": 0.49271008372306824,
+      "learning_rate": 3.909778076145785e-05,
+      "loss": 0.9823,
+      "step": 6523
+    },
+    {
+      "epoch": 0.5829293899524203,
+      "grad_norm": 0.4981904625892639,
+      "learning_rate": 3.90836586833135e-05,
+      "loss": 0.9286,
+      "step": 6524
+    },
+    {
+      "epoch": 0.5830187414836822,
+      "grad_norm": 0.5224156379699707,
+      "learning_rate": 3.906953751952909e-05,
+      "loss": 0.9063,
+      "step": 6525
+    },
+    {
+      "epoch": 0.583108093014944,
+      "grad_norm": 0.5273329615592957,
+      "learning_rate": 3.9055417271287426e-05,
+      "loss": 0.999,
+      "step": 6526
+    },
+    {
+      "epoch": 0.5831974445462059,
+      "grad_norm": 0.5131478905677795,
+      "learning_rate": 3.9041297939771224e-05,
+      "loss": 0.9669,
+      "step": 6527
+    },
+    {
+      "epoch": 0.5832867960774678,
+      "grad_norm": 0.5106601119041443,
+      "learning_rate": 3.9027179526163125e-05,
+      "loss": 0.8704,
+      "step": 6528
+    },
+    {
+      "epoch": 0.5833761476087297,
+      "grad_norm": 0.5162898302078247,
+      "learning_rate": 3.901306203164571e-05,
+      "loss": 0.9613,
+      "step": 6529
+    },
+    {
+      "epoch": 0.5834654991399915,
+      "grad_norm": 0.42621272802352905,
+      "learning_rate": 3.899894545740146e-05,
+      "loss": 0.9468,
+      "step": 6530
+    },
+    {
+      "epoch": 0.5835548506712533,
+      "grad_norm": 0.47468671202659607,
+      "learning_rate": 3.898482980461279e-05,
+      "loss": 0.9432,
+      "step": 6531
+    },
+    {
+      "epoch": 0.5836442022025152,
+      "grad_norm": 0.4372991919517517,
+      "learning_rate": 3.897071507446204e-05,
+      "loss": 0.9773,
+      "step": 6532
+    },
+    {
+      "epoch": 0.5837335537337771,
+      "grad_norm": 0.4610428810119629,
+      "learning_rate": 3.8956601268131486e-05,
+      "loss": 0.9766,
+      "step": 6533
+    },
+    {
+      "epoch": 0.583822905265039,
+      "grad_norm": 0.4634798765182495,
+      "learning_rate": 3.894248838680327e-05,
+      "loss": 1.0252,
+      "step": 6534
+    },
+    {
+      "epoch": 0.5839122567963009,
+      "grad_norm": 0.46993133425712585,
+      "learning_rate": 3.8928376431659516e-05,
+      "loss": 0.8756,
+      "step": 6535
+    },
+    {
+      "epoch": 0.5840016083275628,
+      "grad_norm": 0.4452194273471832,
+      "learning_rate": 3.891426540388224e-05,
+      "loss": 0.9596,
+      "step": 6536
+    },
+    {
+      "epoch": 0.5840909598588245,
+      "grad_norm": 0.47519785165786743,
+      "learning_rate": 3.890015530465342e-05,
+      "loss": 0.9268,
+      "step": 6537
+    },
+    {
+      "epoch": 0.5841803113900864,
+      "grad_norm": 0.42138373851776123,
+      "learning_rate": 3.888604613515491e-05,
+      "loss": 0.9785,
+      "step": 6538
+    },
+    {
+      "epoch": 0.5842696629213483,
+      "grad_norm": 0.429235577583313,
+      "learning_rate": 3.887193789656849e-05,
+      "loss": 0.9077,
+      "step": 6539
+    },
+    {
+      "epoch": 0.5843590144526102,
+      "grad_norm": 0.4701876640319824,
+      "learning_rate": 3.8857830590075895e-05,
+      "loss": 0.9449,
+      "step": 6540
+    },
+    {
+      "epoch": 0.5844483659838721,
+      "grad_norm": 0.45711401104927063,
+      "learning_rate": 3.8843724216858745e-05,
+      "loss": 1.0787,
+      "step": 6541
+    },
+    {
+      "epoch": 0.584537717515134,
+      "grad_norm": 0.4365895390510559,
+      "learning_rate": 3.882961877809862e-05,
+      "loss": 0.9814,
+      "step": 6542
+    },
+    {
+      "epoch": 0.5846270690463958,
+      "grad_norm": 0.5129074454307556,
+      "learning_rate": 3.881551427497701e-05,
+      "loss": 0.9376,
+      "step": 6543
+    },
+    {
+      "epoch": 0.5847164205776576,
+      "grad_norm": 0.45288604497909546,
+      "learning_rate": 3.880141070867527e-05,
+      "loss": 0.9004,
+      "step": 6544
+    },
+    {
+      "epoch": 0.5848057721089195,
+      "grad_norm": 0.42734140157699585,
+      "learning_rate": 3.878730808037475e-05,
+      "loss": 1.0315,
+      "step": 6545
+    },
+    {
+      "epoch": 0.5848951236401814,
+      "grad_norm": 0.47614336013793945,
+      "learning_rate": 3.87732063912567e-05,
+      "loss": 1.0126,
+      "step": 6546
+    },
+    {
+      "epoch": 0.5849844751714433,
+      "grad_norm": 0.475925475358963,
+      "learning_rate": 3.875910564250229e-05,
+      "loss": 0.9627,
+      "step": 6547
+    },
+    {
+      "epoch": 0.5850738267027051,
+      "grad_norm": 0.4257443845272064,
+      "learning_rate": 3.874500583529259e-05,
+      "loss": 0.9535,
+      "step": 6548
+    },
+    {
+      "epoch": 0.585163178233967,
+      "grad_norm": 0.4719519317150116,
+      "learning_rate": 3.873090697080863e-05,
+      "loss": 0.9882,
+      "step": 6549
+    },
+    {
+      "epoch": 0.5852525297652289,
+      "grad_norm": 0.5359987020492554,
+      "learning_rate": 3.871680905023133e-05,
+      "loss": 0.9496,
+      "step": 6550
+    },
+    {
+      "epoch": 0.5853418812964907,
+      "grad_norm": 0.5468908548355103,
+      "learning_rate": 3.8702712074741534e-05,
+      "loss": 0.9608,
+      "step": 6551
+    },
+    {
+      "epoch": 0.5854312328277526,
+      "grad_norm": 0.5251081585884094,
+      "learning_rate": 3.868861604552004e-05,
+      "loss": 0.8878,
+      "step": 6552
+    },
+    {
+      "epoch": 0.5855205843590144,
+      "grad_norm": 0.4280673861503601,
+      "learning_rate": 3.8674520963747526e-05,
+      "loss": 0.916,
+      "step": 6553
+    },
+    {
+      "epoch": 0.5856099358902763,
+      "grad_norm": 0.4354718029499054,
+      "learning_rate": 3.866042683060459e-05,
+      "loss": 0.952,
+      "step": 6554
+    },
+    {
+      "epoch": 0.5856992874215382,
+      "grad_norm": 0.43345123529434204,
+      "learning_rate": 3.864633364727177e-05,
+      "loss": 0.9935,
+      "step": 6555
+    },
+    {
+      "epoch": 0.5857886389528001,
+      "grad_norm": 0.3917457163333893,
+      "learning_rate": 3.8632241414929536e-05,
+      "loss": 0.9337,
+      "step": 6556
+    },
+    {
+      "epoch": 0.585877990484062,
+      "grad_norm": 0.45168983936309814,
+      "learning_rate": 3.861815013475827e-05,
+      "loss": 0.9561,
+      "step": 6557
+    },
+    {
+      "epoch": 0.5859673420153237,
+      "grad_norm": 0.47507405281066895,
+      "learning_rate": 3.860405980793823e-05,
+      "loss": 0.9946,
+      "step": 6558
+    },
+    {
+      "epoch": 0.5860566935465856,
+      "grad_norm": 0.5725569725036621,
+      "learning_rate": 3.858997043564966e-05,
+      "loss": 0.9315,
+      "step": 6559
+    },
+    {
+      "epoch": 0.5861460450778475,
+      "grad_norm": 0.4958851635456085,
+      "learning_rate": 3.8575882019072683e-05,
+      "loss": 0.8849,
+      "step": 6560
+    },
+    {
+      "epoch": 0.5862353966091094,
+      "grad_norm": 0.4789509177207947,
+      "learning_rate": 3.8561794559387366e-05,
+      "loss": 0.9942,
+      "step": 6561
+    },
+    {
+      "epoch": 0.5863247481403713,
+      "grad_norm": 0.4200628697872162,
+      "learning_rate": 3.854770805777368e-05,
+      "loss": 0.9415,
+      "step": 6562
+    },
+    {
+      "epoch": 0.5864140996716332,
+      "grad_norm": 0.4809480607509613,
+      "learning_rate": 3.8533622515411525e-05,
+      "loss": 0.9314,
+      "step": 6563
+    },
+    {
+      "epoch": 0.586503451202895,
+      "grad_norm": 0.5276997685432434,
+      "learning_rate": 3.85195379334807e-05,
+      "loss": 0.9348,
+      "step": 6564
+    },
+    {
+      "epoch": 0.5865928027341568,
+      "grad_norm": 0.5005677938461304,
+      "learning_rate": 3.8505454313160935e-05,
+      "loss": 0.8954,
+      "step": 6565
+    },
+    {
+      "epoch": 0.5866821542654187,
+      "grad_norm": 0.6186678409576416,
+      "learning_rate": 3.8491371655631897e-05,
+      "loss": 0.9265,
+      "step": 6566
+    },
+    {
+      "epoch": 0.5867715057966806,
+      "grad_norm": 0.45182543992996216,
+      "learning_rate": 3.847728996207316e-05,
+      "loss": 1.0398,
+      "step": 6567
+    },
+    {
+      "epoch": 0.5868608573279425,
+      "grad_norm": 0.5082495212554932,
+      "learning_rate": 3.846320923366421e-05,
+      "loss": 0.9212,
+      "step": 6568
+    },
+    {
+      "epoch": 0.5869502088592043,
+      "grad_norm": 0.4350298345088959,
+      "learning_rate": 3.844912947158446e-05,
+      "loss": 0.9523,
+      "step": 6569
+    },
+    {
+      "epoch": 0.5870395603904662,
+      "grad_norm": 0.5526973605155945,
+      "learning_rate": 3.843505067701324e-05,
+      "loss": 0.8776,
+      "step": 6570
+    },
+    {
+      "epoch": 0.5871289119217281,
+      "grad_norm": 0.4889020323753357,
+      "learning_rate": 3.84209728511298e-05,
+      "loss": 1.0305,
+      "step": 6571
+    },
+    {
+      "epoch": 0.5872182634529899,
+      "grad_norm": 0.5451194643974304,
+      "learning_rate": 3.840689599511331e-05,
+      "loss": 0.9137,
+      "step": 6572
+    },
+    {
+      "epoch": 0.5873076149842518,
+      "grad_norm": 0.44288796186447144,
+      "learning_rate": 3.839282011014286e-05,
+      "loss": 0.8818,
+      "step": 6573
+    },
+    {
+      "epoch": 0.5873969665155137,
+      "grad_norm": 0.42640823125839233,
+      "learning_rate": 3.837874519739744e-05,
+      "loss": 0.9603,
+      "step": 6574
+    },
+    {
+      "epoch": 0.5874863180467755,
+      "grad_norm": 0.43930867314338684,
+      "learning_rate": 3.8364671258056e-05,
+      "loss": 1.0335,
+      "step": 6575
+    },
+    {
+      "epoch": 0.5875756695780374,
+      "grad_norm": 0.5153297185897827,
+      "learning_rate": 3.835059829329735e-05,
+      "loss": 1.0377,
+      "step": 6576
+    },
+    {
+      "epoch": 0.5876650211092993,
+      "grad_norm": 0.47636160254478455,
+      "learning_rate": 3.8336526304300265e-05,
+      "loss": 0.9774,
+      "step": 6577
+    },
+    {
+      "epoch": 0.5877543726405611,
+      "grad_norm": 0.44680356979370117,
+      "learning_rate": 3.832245529224342e-05,
+      "loss": 1.0178,
+      "step": 6578
+    },
+    {
+      "epoch": 0.587843724171823,
+      "grad_norm": 0.44048798084259033,
+      "learning_rate": 3.830838525830542e-05,
+      "loss": 1.0009,
+      "step": 6579
+    },
+    {
+      "epoch": 0.5879330757030848,
+      "grad_norm": 0.4684334993362427,
+      "learning_rate": 3.829431620366479e-05,
+      "loss": 0.8922,
+      "step": 6580
+    },
+    {
+      "epoch": 0.5880224272343467,
+      "grad_norm": 0.5170097351074219,
+      "learning_rate": 3.828024812949994e-05,
+      "loss": 0.9479,
+      "step": 6581
+    },
+    {
+      "epoch": 0.5881117787656086,
+      "grad_norm": 0.4309813976287842,
+      "learning_rate": 3.826618103698924e-05,
+      "loss": 0.9732,
+      "step": 6582
+    },
+    {
+      "epoch": 0.5882011302968705,
+      "grad_norm": 0.4656025469303131,
+      "learning_rate": 3.825211492731097e-05,
+      "loss": 0.9387,
+      "step": 6583
+    },
+    {
+      "epoch": 0.5882904818281324,
+      "grad_norm": 0.4610069990158081,
+      "learning_rate": 3.823804980164328e-05,
+      "loss": 1.0072,
+      "step": 6584
+    },
+    {
+      "epoch": 0.5883798333593941,
+      "grad_norm": 0.49006903171539307,
+      "learning_rate": 3.8223985661164284e-05,
+      "loss": 0.9517,
+      "step": 6585
+    },
+    {
+      "epoch": 0.588469184890656,
+      "grad_norm": 0.5057728290557861,
+      "learning_rate": 3.820992250705202e-05,
+      "loss": 0.9777,
+      "step": 6586
+    },
+    {
+      "epoch": 0.5885585364219179,
+      "grad_norm": 0.45757678151130676,
+      "learning_rate": 3.819586034048441e-05,
+      "loss": 0.9111,
+      "step": 6587
+    },
+    {
+      "epoch": 0.5886478879531798,
+      "grad_norm": 0.4634867012500763,
+      "learning_rate": 3.818179916263933e-05,
+      "loss": 0.9293,
+      "step": 6588
+    },
+    {
+      "epoch": 0.5887372394844417,
+      "grad_norm": 0.42738625407218933,
+      "learning_rate": 3.816773897469454e-05,
+      "loss": 1.005,
+      "step": 6589
+    },
+    {
+      "epoch": 0.5888265910157036,
+      "grad_norm": 0.496150404214859,
+      "learning_rate": 3.815367977782774e-05,
+      "loss": 0.878,
+      "step": 6590
+    },
+    {
+      "epoch": 0.5889159425469654,
+      "grad_norm": 0.4651472866535187,
+      "learning_rate": 3.813962157321653e-05,
+      "loss": 0.991,
+      "step": 6591
+    },
+    {
+      "epoch": 0.5890052940782272,
+      "grad_norm": 0.49362361431121826,
+      "learning_rate": 3.812556436203843e-05,
+      "loss": 0.9222,
+      "step": 6592
+    },
+    {
+      "epoch": 0.5890946456094891,
+      "grad_norm": 0.4723232686519623,
+      "learning_rate": 3.8111508145470886e-05,
+      "loss": 0.9071,
+      "step": 6593
+    },
+    {
+      "epoch": 0.589183997140751,
+      "grad_norm": 0.4556770622730255,
+      "learning_rate": 3.809745292469128e-05,
+      "loss": 0.9447,
+      "step": 6594
+    },
+    {
+      "epoch": 0.5892733486720129,
+      "grad_norm": 0.49390918016433716,
+      "learning_rate": 3.808339870087684e-05,
+      "loss": 0.8855,
+      "step": 6595
+    },
+    {
+      "epoch": 0.5893627002032747,
+      "grad_norm": 0.4614836275577545,
+      "learning_rate": 3.8069345475204784e-05,
+      "loss": 0.9299,
+      "step": 6596
+    },
+    {
+      "epoch": 0.5894520517345366,
+      "grad_norm": 0.42079588770866394,
+      "learning_rate": 3.805529324885222e-05,
+      "loss": 0.9729,
+      "step": 6597
+    },
+    {
+      "epoch": 0.5895414032657985,
+      "grad_norm": 0.4124428629875183,
+      "learning_rate": 3.804124202299615e-05,
+      "loss": 0.9736,
+      "step": 6598
+    },
+    {
+      "epoch": 0.5896307547970603,
+      "grad_norm": 0.45447784662246704,
+      "learning_rate": 3.8027191798813546e-05,
+      "loss": 0.937,
+      "step": 6599
+    },
+    {
+      "epoch": 0.5897201063283222,
+      "grad_norm": 0.5072098970413208,
+      "learning_rate": 3.801314257748125e-05,
+      "loss": 0.9483,
+      "step": 6600
+    },
+    {
+      "epoch": 0.589809457859584,
+      "grad_norm": 0.42272502183914185,
+      "learning_rate": 3.799909436017604e-05,
+      "loss": 1.0475,
+      "step": 6601
+    },
+    {
+      "epoch": 0.5898988093908459,
+      "grad_norm": 0.44008708000183105,
+      "learning_rate": 3.7985047148074585e-05,
+      "loss": 0.9706,
+      "step": 6602
+    },
+    {
+      "epoch": 0.5899881609221078,
+      "grad_norm": 0.4524151086807251,
+      "learning_rate": 3.797100094235351e-05,
+      "loss": 0.9213,
+      "step": 6603
+    },
+    {
+      "epoch": 0.5900775124533697,
+      "grad_norm": 0.4076539874076843,
+      "learning_rate": 3.795695574418934e-05,
+      "loss": 1.0978,
+      "step": 6604
+    },
+    {
+      "epoch": 0.5901668639846316,
+      "grad_norm": 0.46805551648139954,
+      "learning_rate": 3.794291155475848e-05,
+      "loss": 0.8995,
+      "step": 6605
+    },
+    {
+      "epoch": 0.5902562155158934,
+      "grad_norm": 0.4466945230960846,
+      "learning_rate": 3.792886837523729e-05,
+      "loss": 0.9412,
+      "step": 6606
+    },
+    {
+      "epoch": 0.5903455670471552,
+      "grad_norm": 0.4632209837436676,
+      "learning_rate": 3.7914826206802047e-05,
+      "loss": 0.9332,
+      "step": 6607
+    },
+    {
+      "epoch": 0.5904349185784171,
+      "grad_norm": 0.49044013023376465,
+      "learning_rate": 3.790078505062894e-05,
+      "loss": 1.0105,
+      "step": 6608
+    },
+    {
+      "epoch": 0.590524270109679,
+      "grad_norm": 0.45329976081848145,
+      "learning_rate": 3.788674490789404e-05,
+      "loss": 1.0149,
+      "step": 6609
+    },
+    {
+      "epoch": 0.5906136216409409,
+      "grad_norm": 0.45417705178260803,
+      "learning_rate": 3.7872705779773376e-05,
+      "loss": 0.9508,
+      "step": 6610
+    },
+    {
+      "epoch": 0.5907029731722028,
+      "grad_norm": 0.4554515480995178,
+      "learning_rate": 3.785866766744287e-05,
+      "loss": 0.9459,
+      "step": 6611
+    },
+    {
+      "epoch": 0.5907923247034647,
+      "grad_norm": 0.4931671619415283,
+      "learning_rate": 3.784463057207836e-05,
+      "loss": 0.9422,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5908816762347264,
+      "grad_norm": 0.5159240961074829,
+      "learning_rate": 3.783059449485561e-05,
+      "loss": 0.9961,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5909710277659883,
+      "grad_norm": 0.4163471758365631,
+      "learning_rate": 3.78165594369503e-05,
+      "loss": 1.0271,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5910603792972502,
+      "grad_norm": 0.4183323383331299,
+      "learning_rate": 3.7802525399538e-05,
+      "loss": 1.0124,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5911497308285121,
+      "grad_norm": 0.46458837389945984,
+      "learning_rate": 3.77884923837942e-05,
+      "loss": 0.952,
+      "step": 6616
+    },
+    {
+      "epoch": 0.591239082359774,
+      "grad_norm": 0.3890257775783539,
+      "learning_rate": 3.777446039089433e-05,
+      "loss": 0.9898,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5913284338910358,
+      "grad_norm": 0.4664503335952759,
+      "learning_rate": 3.776042942201372e-05,
+      "loss": 0.9048,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5914177854222977,
+      "grad_norm": 0.5847722887992859,
+      "learning_rate": 3.774639947832761e-05,
+      "loss": 0.8766,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5915071369535595,
+      "grad_norm": 0.4752833843231201,
+      "learning_rate": 3.7732370561011154e-05,
+      "loss": 0.9276,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5915964884848214,
+      "grad_norm": 0.4764302670955658,
+      "learning_rate": 3.771834267123943e-05,
+      "loss": 0.9594,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5916858400160833,
+      "grad_norm": 0.40147051215171814,
+      "learning_rate": 3.770431581018743e-05,
+      "loss": 0.9262,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5917751915473451,
+      "grad_norm": 0.4563705623149872,
+      "learning_rate": 3.769028997903003e-05,
+      "loss": 0.9904,
+      "step": 6623
+    },
+    {
+      "epoch": 0.591864543078607,
+      "grad_norm": 0.4160234034061432,
+      "learning_rate": 3.7676265178942074e-05,
+      "loss": 1.0111,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5919538946098689,
+      "grad_norm": 0.426800400018692,
+      "learning_rate": 3.766224141109825e-05,
+      "loss": 0.9869,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5920432461411308,
+      "grad_norm": 0.43789252638816833,
+      "learning_rate": 3.764821867667323e-05,
+      "loss": 0.9818,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5921325976723926,
+      "grad_norm": 0.5048993229866028,
+      "learning_rate": 3.763419697684156e-05,
+      "loss": 0.8763,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5922219492036545,
+      "grad_norm": 0.5642315745353699,
+      "learning_rate": 3.76201763127777e-05,
+      "loss": 0.9,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5923113007349163,
+      "grad_norm": 0.4938600957393646,
+      "learning_rate": 3.7606156685656026e-05,
+      "loss": 0.924,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5924006522661782,
+      "grad_norm": 0.6127682328224182,
+      "learning_rate": 3.759213809665084e-05,
+      "loss": 0.9511,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5924900037974401,
+      "grad_norm": 0.5626466870307922,
+      "learning_rate": 3.757812054693634e-05,
+      "loss": 0.9138,
+      "step": 6631
+    },
+    {
+      "epoch": 0.592579355328702,
+      "grad_norm": 0.4270266592502594,
+      "learning_rate": 3.756410403768667e-05,
+      "loss": 0.9399,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5926687068599639,
+      "grad_norm": 0.4422089159488678,
+      "learning_rate": 3.755008857007583e-05,
+      "loss": 0.9363,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5927580583912256,
+      "grad_norm": 0.4473762512207031,
+      "learning_rate": 3.75360741452778e-05,
+      "loss": 0.9541,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5928474099224875,
+      "grad_norm": 0.44996803998947144,
+      "learning_rate": 3.752206076446641e-05,
+      "loss": 0.9283,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5929367614537494,
+      "grad_norm": 0.544922411441803,
+      "learning_rate": 3.7508048428815416e-05,
+      "loss": 0.893,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5930261129850113,
+      "grad_norm": 0.38203221559524536,
+      "learning_rate": 3.7494037139498525e-05,
+      "loss": 0.9555,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5931154645162732,
+      "grad_norm": 0.47850242257118225,
+      "learning_rate": 3.748002689768934e-05,
+      "loss": 0.939,
+      "step": 6638
+    },
+    {
+      "epoch": 0.593204816047535,
+      "grad_norm": 0.43377867341041565,
+      "learning_rate": 3.7466017704561345e-05,
+      "loss": 0.9497,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5932941675787968,
+      "grad_norm": 0.4623333215713501,
+      "learning_rate": 3.745200956128797e-05,
+      "loss": 0.8625,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5933835191100587,
+      "grad_norm": 0.4614277482032776,
+      "learning_rate": 3.7438002469042565e-05,
+      "loss": 0.9202,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5934728706413206,
+      "grad_norm": 0.4852299988269806,
+      "learning_rate": 3.742399642899833e-05,
+      "loss": 0.9428,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5935622221725825,
+      "grad_norm": 0.46920087933540344,
+      "learning_rate": 3.740999144232846e-05,
+      "loss": 0.9912,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5936515737038444,
+      "grad_norm": 0.4522823393344879,
+      "learning_rate": 3.739598751020601e-05,
+      "loss": 0.9731,
+      "step": 6644
+    },
+    {
+      "epoch": 0.5937409252351062,
+      "grad_norm": 0.4409060478210449,
+      "learning_rate": 3.7381984633803955e-05,
+      "loss": 0.9657,
+      "step": 6645
+    },
+    {
+      "epoch": 0.5938302767663681,
+      "grad_norm": 0.45765984058380127,
+      "learning_rate": 3.7367982814295174e-05,
+      "loss": 0.9724,
+      "step": 6646
+    },
+    {
+      "epoch": 0.5939196282976299,
+      "grad_norm": 0.5184260606765747,
+      "learning_rate": 3.735398205285248e-05,
+      "loss": 0.9308,
+      "step": 6647
+    },
+    {
+      "epoch": 0.5940089798288918,
+      "grad_norm": 0.44960641860961914,
+      "learning_rate": 3.733998235064858e-05,
+      "loss": 0.9715,
+      "step": 6648
+    },
+    {
+      "epoch": 0.5940983313601537,
+      "grad_norm": 0.416469931602478,
+      "learning_rate": 3.732598370885612e-05,
+      "loss": 0.9493,
+      "step": 6649
+    },
+    {
+      "epoch": 0.5941876828914155,
+      "grad_norm": 0.4398280382156372,
+      "learning_rate": 3.73119861286476e-05,
+      "loss": 0.996,
+      "step": 6650
+    },
+    {
+      "epoch": 0.5942770344226774,
+      "grad_norm": 0.4765319526195526,
+      "learning_rate": 3.7297989611195506e-05,
+      "loss": 0.9283,
+      "step": 6651
+    },
+    {
+      "epoch": 0.5943663859539393,
+      "grad_norm": 0.4116184115409851,
+      "learning_rate": 3.728399415767216e-05,
+      "loss": 0.9913,
+      "step": 6652
+    },
+    {
+      "epoch": 0.5944557374852012,
+      "grad_norm": 0.43389296531677246,
+      "learning_rate": 3.7269999769249855e-05,
+      "loss": 0.9189,
+      "step": 6653
+    },
+    {
+      "epoch": 0.594545089016463,
+      "grad_norm": 0.46604517102241516,
+      "learning_rate": 3.725600644710078e-05,
+      "loss": 0.8957,
+      "step": 6654
+    },
+    {
+      "epoch": 0.5946344405477249,
+      "grad_norm": 0.509748101234436,
+      "learning_rate": 3.724201419239699e-05,
+      "loss": 0.895,
+      "step": 6655
+    },
+    {
+      "epoch": 0.5947237920789867,
+      "grad_norm": 0.4493167996406555,
+      "learning_rate": 3.722802300631049e-05,
+      "loss": 0.9533,
+      "step": 6656
+    },
+    {
+      "epoch": 0.5948131436102486,
+      "grad_norm": 0.5172920823097229,
+      "learning_rate": 3.721403289001321e-05,
+      "loss": 0.9407,
+      "step": 6657
+    },
+    {
+      "epoch": 0.5949024951415105,
+      "grad_norm": 0.45403075218200684,
+      "learning_rate": 3.720004384467697e-05,
+      "loss": 0.98,
+      "step": 6658
+    },
+    {
+      "epoch": 0.5949918466727724,
+      "grad_norm": 0.5167436003684998,
+      "learning_rate": 3.718605587147348e-05,
+      "loss": 0.972,
+      "step": 6659
+    },
+    {
+      "epoch": 0.5950811982040343,
+      "grad_norm": 0.4543740153312683,
+      "learning_rate": 3.7172068971574426e-05,
+      "loss": 0.9288,
+      "step": 6660
+    },
+    {
+      "epoch": 0.595170549735296,
+      "grad_norm": 0.401083767414093,
+      "learning_rate": 3.715808314615131e-05,
+      "loss": 0.9727,
+      "step": 6661
+    },
+    {
+      "epoch": 0.5952599012665579,
+      "grad_norm": 0.4421873986721039,
+      "learning_rate": 3.714409839637562e-05,
+      "loss": 0.96,
+      "step": 6662
+    },
+    {
+      "epoch": 0.5953492527978198,
+      "grad_norm": 0.436882883310318,
+      "learning_rate": 3.713011472341872e-05,
+      "loss": 0.947,
+      "step": 6663
+    },
+    {
+      "epoch": 0.5954386043290817,
+      "grad_norm": 0.5699417591094971,
+      "learning_rate": 3.711613212845192e-05,
+      "loss": 0.9443,
+      "step": 6664
+    },
+    {
+      "epoch": 0.5955279558603436,
+      "grad_norm": 0.48340317606925964,
+      "learning_rate": 3.7102150612646356e-05,
+      "loss": 0.9616,
+      "step": 6665
+    },
+    {
+      "epoch": 0.5956173073916055,
+      "grad_norm": 0.6746427416801453,
+      "learning_rate": 3.708817017717317e-05,
+      "loss": 0.8092,
+      "step": 6666
+    },
+    {
+      "epoch": 0.5957066589228673,
+      "grad_norm": 0.5548288822174072,
+      "learning_rate": 3.707419082320336e-05,
+      "loss": 0.9948,
+      "step": 6667
+    },
+    {
+      "epoch": 0.5957960104541291,
+      "grad_norm": 0.4507887363433838,
+      "learning_rate": 3.7060212551907845e-05,
+      "loss": 0.9182,
+      "step": 6668
+    },
+    {
+      "epoch": 0.595885361985391,
+      "grad_norm": 0.43728700280189514,
+      "learning_rate": 3.704623536445746e-05,
+      "loss": 0.9271,
+      "step": 6669
+    },
+    {
+      "epoch": 0.5959747135166529,
+      "grad_norm": 0.43481287360191345,
+      "learning_rate": 3.7032259262022936e-05,
+      "loss": 1.0612,
+      "step": 6670
+    },
+    {
+      "epoch": 0.5960640650479148,
+      "grad_norm": 0.4465053081512451,
+      "learning_rate": 3.7018284245774925e-05,
+      "loss": 0.9987,
+      "step": 6671
+    },
+    {
+      "epoch": 0.5961534165791766,
+      "grad_norm": 0.50724196434021,
+      "learning_rate": 3.700431031688399e-05,
+      "loss": 0.9271,
+      "step": 6672
+    },
+    {
+      "epoch": 0.5962427681104385,
+      "grad_norm": 0.3783160150051117,
+      "learning_rate": 3.699033747652059e-05,
+      "loss": 0.9373,
+      "step": 6673
+    },
+    {
+      "epoch": 0.5963321196417004,
+      "grad_norm": 0.4265550971031189,
+      "learning_rate": 3.697636572585511e-05,
+      "loss": 0.923,
+      "step": 6674
+    },
+    {
+      "epoch": 0.5964214711729622,
+      "grad_norm": 0.4511755406856537,
+      "learning_rate": 3.6962395066057806e-05,
+      "loss": 0.9411,
+      "step": 6675
+    },
+    {
+      "epoch": 0.5965108227042241,
+      "grad_norm": 0.398965984582901,
+      "learning_rate": 3.694842549829889e-05,
+      "loss": 1.0101,
+      "step": 6676
+    },
+    {
+      "epoch": 0.596600174235486,
+      "grad_norm": 0.48364296555519104,
+      "learning_rate": 3.693445702374846e-05,
+      "loss": 0.9362,
+      "step": 6677
+    },
+    {
+      "epoch": 0.5966895257667478,
+      "grad_norm": 0.4517565071582794,
+      "learning_rate": 3.692048964357653e-05,
+      "loss": 0.9031,
+      "step": 6678
+    },
+    {
+      "epoch": 0.5967788772980097,
+      "grad_norm": 0.42774462699890137,
+      "learning_rate": 3.690652335895299e-05,
+      "loss": 0.9305,
+      "step": 6679
+    },
+    {
+      "epoch": 0.5968682288292716,
+      "grad_norm": 0.5115528702735901,
+      "learning_rate": 3.68925581710477e-05,
+      "loss": 0.8811,
+      "step": 6680
+    },
+    {
+      "epoch": 0.5969575803605335,
+      "grad_norm": 0.4454442262649536,
+      "learning_rate": 3.687859408103037e-05,
+      "loss": 0.9036,
+      "step": 6681
+    },
+    {
+      "epoch": 0.5970469318917953,
+      "grad_norm": 0.5473392009735107,
+      "learning_rate": 3.6864631090070655e-05,
+      "loss": 0.8625,
+      "step": 6682
+    },
+    {
+      "epoch": 0.5971362834230571,
+      "grad_norm": 0.4324687719345093,
+      "learning_rate": 3.6850669199338096e-05,
+      "loss": 0.9759,
+      "step": 6683
+    },
+    {
+      "epoch": 0.597225634954319,
+      "grad_norm": 0.4886661767959595,
+      "learning_rate": 3.683670841000215e-05,
+      "loss": 0.9272,
+      "step": 6684
+    },
+    {
+      "epoch": 0.5973149864855809,
+      "grad_norm": 0.4877077341079712,
+      "learning_rate": 3.682274872323221e-05,
+      "loss": 0.8533,
+      "step": 6685
+    },
+    {
+      "epoch": 0.5974043380168428,
+      "grad_norm": 0.42951878905296326,
+      "learning_rate": 3.680879014019751e-05,
+      "loss": 0.9555,
+      "step": 6686
+    },
+    {
+      "epoch": 0.5974936895481047,
+      "grad_norm": 0.5074982047080994,
+      "learning_rate": 3.679483266206723e-05,
+      "loss": 0.9338,
+      "step": 6687
+    },
+    {
+      "epoch": 0.5975830410793666,
+      "grad_norm": 0.4952092170715332,
+      "learning_rate": 3.678087629001048e-05,
+      "loss": 1.0184,
+      "step": 6688
+    },
+    {
+      "epoch": 0.5976723926106283,
+      "grad_norm": 0.45204105973243713,
+      "learning_rate": 3.676692102519625e-05,
+      "loss": 0.9575,
+      "step": 6689
+    },
+    {
+      "epoch": 0.5977617441418902,
+      "grad_norm": 0.5234025120735168,
+      "learning_rate": 3.675296686879343e-05,
+      "loss": 0.9418,
+      "step": 6690
+    },
+    {
+      "epoch": 0.5978510956731521,
+      "grad_norm": 0.4052625298500061,
+      "learning_rate": 3.6739013821970846e-05,
+      "loss": 0.949,
+      "step": 6691
+    },
+    {
+      "epoch": 0.597940447204414,
+      "grad_norm": 0.5295931696891785,
+      "learning_rate": 3.67250618858972e-05,
+      "loss": 0.9215,
+      "step": 6692
+    },
+    {
+      "epoch": 0.5980297987356759,
+      "grad_norm": 0.4193294942378998,
+      "learning_rate": 3.671111106174113e-05,
+      "loss": 0.9675,
+      "step": 6693
+    },
+    {
+      "epoch": 0.5981191502669377,
+      "grad_norm": 0.4514252841472626,
+      "learning_rate": 3.669716135067116e-05,
+      "loss": 0.9491,
+      "step": 6694
+    },
+    {
+      "epoch": 0.5982085017981996,
+      "grad_norm": 0.38525697588920593,
+      "learning_rate": 3.6683212753855726e-05,
+      "loss": 0.962,
+      "step": 6695
+    },
+    {
+      "epoch": 0.5982978533294614,
+      "grad_norm": 0.49608129262924194,
+      "learning_rate": 3.666926527246316e-05,
+      "loss": 0.9742,
+      "step": 6696
+    },
+    {
+      "epoch": 0.5983872048607233,
+      "grad_norm": 0.5682451128959656,
+      "learning_rate": 3.6655318907661726e-05,
+      "loss": 0.879,
+      "step": 6697
+    },
+    {
+      "epoch": 0.5984765563919852,
+      "grad_norm": 0.3935823142528534,
+      "learning_rate": 3.664137366061958e-05,
+      "loss": 0.9591,
+      "step": 6698
+    },
+    {
+      "epoch": 0.598565907923247,
+      "grad_norm": 0.4220222234725952,
+      "learning_rate": 3.662742953250478e-05,
+      "loss": 1.0057,
+      "step": 6699
+    },
+    {
+      "epoch": 0.5986552594545089,
+      "grad_norm": 0.4496472179889679,
+      "learning_rate": 3.6613486524485294e-05,
+      "loss": 0.9168,
+      "step": 6700
+    },
+    {
+      "epoch": 0.5987446109857708,
+      "grad_norm": 0.40350624918937683,
+      "learning_rate": 3.6599544637729007e-05,
+      "loss": 1.0385,
+      "step": 6701
+    },
+    {
+      "epoch": 0.5988339625170326,
+      "grad_norm": 0.42951273918151855,
+      "learning_rate": 3.65856038734037e-05,
+      "loss": 0.9669,
+      "step": 6702
+    },
+    {
+      "epoch": 0.5989233140482945,
+      "grad_norm": 0.4338366687297821,
+      "learning_rate": 3.657166423267704e-05,
+      "loss": 0.9676,
+      "step": 6703
+    },
+    {
+      "epoch": 0.5990126655795563,
+      "grad_norm": 0.5033963918685913,
+      "learning_rate": 3.655772571671664e-05,
+      "loss": 0.9289,
+      "step": 6704
+    },
+    {
+      "epoch": 0.5991020171108182,
+      "grad_norm": 0.48832231760025024,
+      "learning_rate": 3.654378832669002e-05,
+      "loss": 0.8811,
+      "step": 6705
+    },
+    {
+      "epoch": 0.5991913686420801,
+      "grad_norm": 0.5550376176834106,
+      "learning_rate": 3.6529852063764545e-05,
+      "loss": 0.8869,
+      "step": 6706
+    },
+    {
+      "epoch": 0.599280720173342,
+      "grad_norm": 0.46830353140830994,
+      "learning_rate": 3.651591692910754e-05,
+      "loss": 0.9608,
+      "step": 6707
+    },
+    {
+      "epoch": 0.5993700717046039,
+      "grad_norm": 0.4385646879673004,
+      "learning_rate": 3.650198292388621e-05,
+      "loss": 1.0007,
+      "step": 6708
+    },
+    {
+      "epoch": 0.5994594232358657,
+      "grad_norm": 0.5016947984695435,
+      "learning_rate": 3.64880500492677e-05,
+      "loss": 1.0079,
+      "step": 6709
+    },
+    {
+      "epoch": 0.5995487747671275,
+      "grad_norm": 0.38246116042137146,
+      "learning_rate": 3.647411830641903e-05,
+      "loss": 0.9701,
+      "step": 6710
+    },
+    {
+      "epoch": 0.5996381262983894,
+      "grad_norm": 0.6153517365455627,
+      "learning_rate": 3.646018769650713e-05,
+      "loss": 0.8733,
+      "step": 6711
+    },
+    {
+      "epoch": 0.5997274778296513,
+      "grad_norm": 0.5294399857521057,
+      "learning_rate": 3.6446258220698814e-05,
+      "loss": 0.9023,
+      "step": 6712
+    },
+    {
+      "epoch": 0.5998168293609132,
+      "grad_norm": 0.4735589623451233,
+      "learning_rate": 3.643232988016086e-05,
+      "loss": 0.9632,
+      "step": 6713
+    },
+    {
+      "epoch": 0.5999061808921751,
+      "grad_norm": 0.4517965018749237,
+      "learning_rate": 3.641840267605989e-05,
+      "loss": 0.95,
+      "step": 6714
+    },
+    {
+      "epoch": 0.599995532423437,
+      "grad_norm": 0.5445354580879211,
+      "learning_rate": 3.640447660956249e-05,
+      "loss": 0.8805,
+      "step": 6715
+    },
+    {
+      "epoch": 0.6000848839546987,
+      "grad_norm": 0.4683813452720642,
+      "learning_rate": 3.639055168183507e-05,
+      "loss": 0.9601,
+      "step": 6716
+    },
+    {
+      "epoch": 0.6001742354859606,
+      "grad_norm": 0.5692836046218872,
+      "learning_rate": 3.637662789404402e-05,
+      "loss": 0.9754,
+      "step": 6717
+    },
+    {
+      "epoch": 0.6002635870172225,
+      "grad_norm": 0.47742897272109985,
+      "learning_rate": 3.636270524735559e-05,
+      "loss": 0.956,
+      "step": 6718
+    },
+    {
+      "epoch": 0.6003529385484844,
+      "grad_norm": 0.5341329574584961,
+      "learning_rate": 3.6348783742935966e-05,
+      "loss": 0.9581,
+      "step": 6719
+    },
+    {
+      "epoch": 0.6004422900797463,
+      "grad_norm": 0.45011308789253235,
+      "learning_rate": 3.6334863381951214e-05,
+      "loss": 0.9189,
+      "step": 6720
+    },
+    {
+      "epoch": 0.6005316416110081,
+      "grad_norm": 0.4406468868255615,
+      "learning_rate": 3.632094416556731e-05,
+      "loss": 0.9414,
+      "step": 6721
+    },
+    {
+      "epoch": 0.60062099314227,
+      "grad_norm": 0.46647706627845764,
+      "learning_rate": 3.630702609495014e-05,
+      "loss": 0.893,
+      "step": 6722
+    },
+    {
+      "epoch": 0.6007103446735318,
+      "grad_norm": 0.3957505226135254,
+      "learning_rate": 3.6293109171265486e-05,
+      "loss": 0.9984,
+      "step": 6723
+    },
+    {
+      "epoch": 0.6007996962047937,
+      "grad_norm": 0.5789626836776733,
+      "learning_rate": 3.627919339567906e-05,
+      "loss": 0.9168,
+      "step": 6724
+    },
+    {
+      "epoch": 0.6008890477360556,
+      "grad_norm": 0.4705089032649994,
+      "learning_rate": 3.626527876935645e-05,
+      "loss": 0.9538,
+      "step": 6725
+    },
+    {
+      "epoch": 0.6009783992673174,
+      "grad_norm": 0.4237965941429138,
+      "learning_rate": 3.625136529346312e-05,
+      "loss": 1.0059,
+      "step": 6726
+    },
+    {
+      "epoch": 0.6010677507985793,
+      "grad_norm": 0.423532634973526,
+      "learning_rate": 3.6237452969164495e-05,
+      "loss": 0.9481,
+      "step": 6727
+    },
+    {
+      "epoch": 0.6011571023298412,
+      "grad_norm": 0.5018135905265808,
+      "learning_rate": 3.622354179762589e-05,
+      "loss": 0.9048,
+      "step": 6728
+    },
+    {
+      "epoch": 0.6012464538611031,
+      "grad_norm": 0.4970499873161316,
+      "learning_rate": 3.620963178001251e-05,
+      "loss": 0.8817,
+      "step": 6729
+    },
+    {
+      "epoch": 0.6013358053923649,
+      "grad_norm": 0.5816099047660828,
+      "learning_rate": 3.619572291748947e-05,
+      "loss": 0.9101,
+      "step": 6730
+    },
+    {
+      "epoch": 0.6014251569236267,
+      "grad_norm": 0.43934929370880127,
+      "learning_rate": 3.618181521122176e-05,
+      "loss": 0.9431,
+      "step": 6731
+    },
+    {
+      "epoch": 0.6015145084548886,
+      "grad_norm": 0.503680408000946,
+      "learning_rate": 3.616790866237433e-05,
+      "loss": 0.9065,
+      "step": 6732
+    },
+    {
+      "epoch": 0.6016038599861505,
+      "grad_norm": 0.6071944832801819,
+      "learning_rate": 3.615400327211198e-05,
+      "loss": 0.9336,
+      "step": 6733
+    },
+    {
+      "epoch": 0.6016932115174124,
+      "grad_norm": 0.4700120687484741,
+      "learning_rate": 3.614009904159945e-05,
+      "loss": 0.9705,
+      "step": 6734
+    },
+    {
+      "epoch": 0.6017825630486743,
+      "grad_norm": 0.4359018802642822,
+      "learning_rate": 3.6126195972001376e-05,
+      "loss": 0.968,
+      "step": 6735
+    },
+    {
+      "epoch": 0.6018719145799362,
+      "grad_norm": 0.3945593237876892,
+      "learning_rate": 3.6112294064482253e-05,
+      "loss": 1.0053,
+      "step": 6736
+    },
+    {
+      "epoch": 0.6019612661111979,
+      "grad_norm": 0.3968038558959961,
+      "learning_rate": 3.6098393320206536e-05,
+      "loss": 0.9179,
+      "step": 6737
+    },
+    {
+      "epoch": 0.6020506176424598,
+      "grad_norm": 0.4622810184955597,
+      "learning_rate": 3.608449374033856e-05,
+      "loss": 0.9345,
+      "step": 6738
+    },
+    {
+      "epoch": 0.6021399691737217,
+      "grad_norm": 0.45781803131103516,
+      "learning_rate": 3.607059532604256e-05,
+      "loss": 0.9804,
+      "step": 6739
+    },
+    {
+      "epoch": 0.6022293207049836,
+      "grad_norm": 0.43766266107559204,
+      "learning_rate": 3.6056698078482676e-05,
+      "loss": 0.9481,
+      "step": 6740
+    },
+    {
+      "epoch": 0.6023186722362455,
+      "grad_norm": 0.3953046202659607,
+      "learning_rate": 3.604280199882296e-05,
+      "loss": 0.9496,
+      "step": 6741
+    },
+    {
+      "epoch": 0.6024080237675074,
+      "grad_norm": 0.5054026246070862,
+      "learning_rate": 3.602890708822735e-05,
+      "loss": 0.9,
+      "step": 6742
+    },
+    {
+      "epoch": 0.6024973752987692,
+      "grad_norm": 0.43824490904808044,
+      "learning_rate": 3.601501334785968e-05,
+      "loss": 0.9556,
+      "step": 6743
+    },
+    {
+      "epoch": 0.602586726830031,
+      "grad_norm": 0.4914816915988922,
+      "learning_rate": 3.600112077888374e-05,
+      "loss": 1.0054,
+      "step": 6744
+    },
+    {
+      "epoch": 0.6026760783612929,
+      "grad_norm": 0.4403778910636902,
+      "learning_rate": 3.598722938246314e-05,
+      "loss": 0.9259,
+      "step": 6745
+    },
+    {
+      "epoch": 0.6027654298925548,
+      "grad_norm": 0.4326126277446747,
+      "learning_rate": 3.5973339159761435e-05,
+      "loss": 0.9524,
+      "step": 6746
+    },
+    {
+      "epoch": 0.6028547814238167,
+      "grad_norm": 0.4220033586025238,
+      "learning_rate": 3.595945011194208e-05,
+      "loss": 1.0042,
+      "step": 6747
+    },
+    {
+      "epoch": 0.6029441329550785,
+      "grad_norm": 0.5944996476173401,
+      "learning_rate": 3.594556224016847e-05,
+      "loss": 0.9395,
+      "step": 6748
+    },
+    {
+      "epoch": 0.6030334844863404,
+      "grad_norm": 0.4896973967552185,
+      "learning_rate": 3.593167554560381e-05,
+      "loss": 0.9024,
+      "step": 6749
+    },
+    {
+      "epoch": 0.6031228360176023,
+      "grad_norm": 0.5966628789901733,
+      "learning_rate": 3.591779002941128e-05,
+      "loss": 0.9577,
+      "step": 6750
+    },
+    {
+      "epoch": 0.6032121875488641,
+      "grad_norm": 0.4684157073497772,
+      "learning_rate": 3.590390569275395e-05,
+      "loss": 0.9349,
+      "step": 6751
+    },
+    {
+      "epoch": 0.603301539080126,
+      "grad_norm": 0.5347898006439209,
+      "learning_rate": 3.589002253679476e-05,
+      "loss": 0.9619,
+      "step": 6752
+    },
+    {
+      "epoch": 0.6033908906113878,
+      "grad_norm": 0.4841513931751251,
+      "learning_rate": 3.5876140562696594e-05,
+      "loss": 0.8933,
+      "step": 6753
+    },
+    {
+      "epoch": 0.6034802421426497,
+      "grad_norm": 0.48916760087013245,
+      "learning_rate": 3.58622597716222e-05,
+      "loss": 0.9148,
+      "step": 6754
+    },
+    {
+      "epoch": 0.6035695936739116,
+      "grad_norm": 0.440901517868042,
+      "learning_rate": 3.584838016473426e-05,
+      "loss": 0.9739,
+      "step": 6755
+    },
+    {
+      "epoch": 0.6036589452051735,
+      "grad_norm": 0.39379745721817017,
+      "learning_rate": 3.5834501743195314e-05,
+      "loss": 1.0049,
+      "step": 6756
+    },
+    {
+      "epoch": 0.6037482967364354,
+      "grad_norm": 0.6075735688209534,
+      "learning_rate": 3.582062450816784e-05,
+      "loss": 0.8999,
+      "step": 6757
+    },
+    {
+      "epoch": 0.6038376482676971,
+      "grad_norm": 0.4395672082901001,
+      "learning_rate": 3.580674846081421e-05,
+      "loss": 0.894,
+      "step": 6758
+    },
+    {
+      "epoch": 0.603926999798959,
+      "grad_norm": 0.5487232208251953,
+      "learning_rate": 3.579287360229668e-05,
+      "loss": 0.9693,
+      "step": 6759
+    },
+    {
+      "epoch": 0.6040163513302209,
+      "grad_norm": 0.44997963309288025,
+      "learning_rate": 3.5778999933777423e-05,
+      "loss": 0.9385,
+      "step": 6760
+    },
+    {
+      "epoch": 0.6041057028614828,
+      "grad_norm": 0.49523553252220154,
+      "learning_rate": 3.5765127456418514e-05,
+      "loss": 0.8897,
+      "step": 6761
+    },
+    {
+      "epoch": 0.6041950543927447,
+      "grad_norm": 0.4439547657966614,
+      "learning_rate": 3.57512561713819e-05,
+      "loss": 0.9414,
+      "step": 6762
+    },
+    {
+      "epoch": 0.6042844059240066,
+      "grad_norm": 0.5505682826042175,
+      "learning_rate": 3.5737386079829484e-05,
+      "loss": 0.8705,
+      "step": 6763
+    },
+    {
+      "epoch": 0.6043737574552683,
+      "grad_norm": 0.42850804328918457,
+      "learning_rate": 3.5723517182923e-05,
+      "loss": 0.9114,
+      "step": 6764
+    },
+    {
+      "epoch": 0.6044631089865302,
+      "grad_norm": 0.40083956718444824,
+      "learning_rate": 3.570964948182412e-05,
+      "loss": 0.9795,
+      "step": 6765
+    },
+    {
+      "epoch": 0.6045524605177921,
+      "grad_norm": 0.4469483196735382,
+      "learning_rate": 3.5695782977694436e-05,
+      "loss": 1.0354,
+      "step": 6766
+    },
+    {
+      "epoch": 0.604641812049054,
+      "grad_norm": 0.5799636840820312,
+      "learning_rate": 3.56819176716954e-05,
+      "loss": 0.8358,
+      "step": 6767
+    },
+    {
+      "epoch": 0.6047311635803159,
+      "grad_norm": 0.4821333587169647,
+      "learning_rate": 3.566805356498837e-05,
+      "loss": 0.8875,
+      "step": 6768
+    },
+    {
+      "epoch": 0.6048205151115778,
+      "grad_norm": 0.4493705928325653,
+      "learning_rate": 3.5654190658734624e-05,
+      "loss": 0.971,
+      "step": 6769
+    },
+    {
+      "epoch": 0.6049098666428396,
+      "grad_norm": 0.4657124876976013,
+      "learning_rate": 3.564032895409532e-05,
+      "loss": 0.9372,
+      "step": 6770
+    },
+    {
+      "epoch": 0.6049992181741014,
+      "grad_norm": 0.4851822555065155,
+      "learning_rate": 3.562646845223153e-05,
+      "loss": 0.9399,
+      "step": 6771
+    },
+    {
+      "epoch": 0.6050885697053633,
+      "grad_norm": 0.4866061210632324,
+      "learning_rate": 3.561260915430422e-05,
+      "loss": 0.9057,
+      "step": 6772
+    },
+    {
+      "epoch": 0.6051779212366252,
+      "grad_norm": 0.4425218403339386,
+      "learning_rate": 3.559875106147425e-05,
+      "loss": 0.9446,
+      "step": 6773
+    },
+    {
+      "epoch": 0.605267272767887,
+      "grad_norm": 0.49788007140159607,
+      "learning_rate": 3.5584894174902386e-05,
+      "loss": 0.934,
+      "step": 6774
+    },
+    {
+      "epoch": 0.6053566242991489,
+      "grad_norm": 0.4979552626609802,
+      "learning_rate": 3.557103849574929e-05,
+      "loss": 0.8783,
+      "step": 6775
+    },
+    {
+      "epoch": 0.6054459758304108,
+      "grad_norm": 0.4978952407836914,
+      "learning_rate": 3.5557184025175536e-05,
+      "loss": 0.9696,
+      "step": 6776
+    },
+    {
+      "epoch": 0.6055353273616727,
+      "grad_norm": 0.5217999219894409,
+      "learning_rate": 3.554333076434156e-05,
+      "loss": 0.8745,
+      "step": 6777
+    },
+    {
+      "epoch": 0.6056246788929345,
+      "grad_norm": 0.4879658818244934,
+      "learning_rate": 3.552947871440772e-05,
+      "loss": 0.9278,
+      "step": 6778
+    },
+    {
+      "epoch": 0.6057140304241964,
+      "grad_norm": 0.3954866826534271,
+      "learning_rate": 3.551562787653429e-05,
+      "loss": 0.943,
+      "step": 6779
+    },
+    {
+      "epoch": 0.6058033819554582,
+      "grad_norm": 0.41415131092071533,
+      "learning_rate": 3.550177825188141e-05,
+      "loss": 0.9986,
+      "step": 6780
+    },
+    {
+      "epoch": 0.6058927334867201,
+      "grad_norm": 0.4643884003162384,
+      "learning_rate": 3.5487929841609154e-05,
+      "loss": 0.8917,
+      "step": 6781
+    },
+    {
+      "epoch": 0.605982085017982,
+      "grad_norm": 0.5743858218193054,
+      "learning_rate": 3.5474082646877446e-05,
+      "loss": 0.9231,
+      "step": 6782
+    },
+    {
+      "epoch": 0.6060714365492439,
+      "grad_norm": 0.46073269844055176,
+      "learning_rate": 3.546023666884616e-05,
+      "loss": 0.9074,
+      "step": 6783
+    },
+    {
+      "epoch": 0.6061607880805058,
+      "grad_norm": 0.3921644389629364,
+      "learning_rate": 3.5446391908675033e-05,
+      "loss": 0.9552,
+      "step": 6784
+    },
+    {
+      "epoch": 0.6062501396117675,
+      "grad_norm": 0.4658121168613434,
+      "learning_rate": 3.543254836752371e-05,
+      "loss": 0.9736,
+      "step": 6785
+    },
+    {
+      "epoch": 0.6063394911430294,
+      "grad_norm": 0.44751229882240295,
+      "learning_rate": 3.541870604655176e-05,
+      "loss": 1.0575,
+      "step": 6786
+    },
+    {
+      "epoch": 0.6064288426742913,
+      "grad_norm": 0.534199595451355,
+      "learning_rate": 3.5404864946918595e-05,
+      "loss": 0.8935,
+      "step": 6787
+    },
+    {
+      "epoch": 0.6065181942055532,
+      "grad_norm": 0.41165226697921753,
+      "learning_rate": 3.539102506978356e-05,
+      "loss": 0.9994,
+      "step": 6788
+    },
+    {
+      "epoch": 0.6066075457368151,
+      "grad_norm": 0.5048467516899109,
+      "learning_rate": 3.5377186416305884e-05,
+      "loss": 0.9057,
+      "step": 6789
+    },
+    {
+      "epoch": 0.606696897268077,
+      "grad_norm": 0.44976770877838135,
+      "learning_rate": 3.5363348987644725e-05,
+      "loss": 0.9591,
+      "step": 6790
+    },
+    {
+      "epoch": 0.6067862487993388,
+      "grad_norm": 0.4342833161354065,
+      "learning_rate": 3.534951278495909e-05,
+      "loss": 0.945,
+      "step": 6791
+    },
+    {
+      "epoch": 0.6068756003306006,
+      "grad_norm": 0.4819537401199341,
+      "learning_rate": 3.533567780940794e-05,
+      "loss": 0.9135,
+      "step": 6792
+    },
+    {
+      "epoch": 0.6069649518618625,
+      "grad_norm": 0.4745539426803589,
+      "learning_rate": 3.532184406215008e-05,
+      "loss": 0.9931,
+      "step": 6793
+    },
+    {
+      "epoch": 0.6070543033931244,
+      "grad_norm": 0.4145744740962982,
+      "learning_rate": 3.5308011544344224e-05,
+      "loss": 1.0892,
+      "step": 6794
+    },
+    {
+      "epoch": 0.6071436549243863,
+      "grad_norm": 0.5059308409690857,
+      "learning_rate": 3.529418025714902e-05,
+      "loss": 0.9136,
+      "step": 6795
+    },
+    {
+      "epoch": 0.6072330064556481,
+      "grad_norm": 0.4676026403903961,
+      "learning_rate": 3.528035020172299e-05,
+      "loss": 0.9068,
+      "step": 6796
+    },
+    {
+      "epoch": 0.60732235798691,
+      "grad_norm": 0.4627551734447479,
+      "learning_rate": 3.526652137922451e-05,
+      "loss": 0.9546,
+      "step": 6797
+    },
+    {
+      "epoch": 0.6074117095181719,
+      "grad_norm": 0.5230789184570312,
+      "learning_rate": 3.525269379081191e-05,
+      "loss": 0.959,
+      "step": 6798
+    },
+    {
+      "epoch": 0.6075010610494337,
+      "grad_norm": 0.46442854404449463,
+      "learning_rate": 3.52388674376434e-05,
+      "loss": 0.9959,
+      "step": 6799
+    },
+    {
+      "epoch": 0.6075904125806956,
+      "grad_norm": 0.4610006511211395,
+      "learning_rate": 3.52250423208771e-05,
+      "loss": 0.8957,
+      "step": 6800
+    },
+    {
+      "epoch": 0.6076797641119575,
+      "grad_norm": 0.48097458481788635,
+      "learning_rate": 3.521121844167098e-05,
+      "loss": 0.9721,
+      "step": 6801
+    },
+    {
+      "epoch": 0.6077691156432193,
+      "grad_norm": 0.539371132850647,
+      "learning_rate": 3.5197395801182955e-05,
+      "loss": 0.8974,
+      "step": 6802
+    },
+    {
+      "epoch": 0.6078584671744812,
+      "grad_norm": 0.4365449845790863,
+      "learning_rate": 3.518357440057081e-05,
+      "loss": 0.9525,
+      "step": 6803
+    },
+    {
+      "epoch": 0.6079478187057431,
+      "grad_norm": 0.4736482501029968,
+      "learning_rate": 3.516975424099225e-05,
+      "loss": 0.8908,
+      "step": 6804
+    },
+    {
+      "epoch": 0.608037170237005,
+      "grad_norm": 0.4505004286766052,
+      "learning_rate": 3.515593532360484e-05,
+      "loss": 0.9166,
+      "step": 6805
+    },
+    {
+      "epoch": 0.6081265217682668,
+      "grad_norm": 0.5180654525756836,
+      "learning_rate": 3.5142117649566104e-05,
+      "loss": 0.9138,
+      "step": 6806
+    },
+    {
+      "epoch": 0.6082158732995286,
+      "grad_norm": 0.3982846140861511,
+      "learning_rate": 3.5128301220033366e-05,
+      "loss": 0.9941,
+      "step": 6807
+    },
+    {
+      "epoch": 0.6083052248307905,
+      "grad_norm": 0.4852418005466461,
+      "learning_rate": 3.511448603616392e-05,
+      "loss": 0.9798,
+      "step": 6808
+    },
+    {
+      "epoch": 0.6083945763620524,
+      "grad_norm": 0.4647953510284424,
+      "learning_rate": 3.510067209911493e-05,
+      "loss": 0.8897,
+      "step": 6809
+    },
+    {
+      "epoch": 0.6084839278933143,
+      "grad_norm": 0.4955390691757202,
+      "learning_rate": 3.508685941004348e-05,
+      "loss": 0.9398,
+      "step": 6810
+    },
+    {
+      "epoch": 0.6085732794245762,
+      "grad_norm": 0.45322465896606445,
+      "learning_rate": 3.507304797010651e-05,
+      "loss": 0.9526,
+      "step": 6811
+    },
+    {
+      "epoch": 0.6086626309558381,
+      "grad_norm": 0.47306427359580994,
+      "learning_rate": 3.505923778046088e-05,
+      "loss": 0.9454,
+      "step": 6812
+    },
+    {
+      "epoch": 0.6087519824870998,
+      "grad_norm": 0.5509787201881409,
+      "learning_rate": 3.5045428842263344e-05,
+      "loss": 0.87,
+      "step": 6813
+    },
+    {
+      "epoch": 0.6088413340183617,
+      "grad_norm": 0.4736871123313904,
+      "learning_rate": 3.503162115667056e-05,
+      "loss": 0.9705,
+      "step": 6814
+    },
+    {
+      "epoch": 0.6089306855496236,
+      "grad_norm": 0.6895001530647278,
+      "learning_rate": 3.5017814724839046e-05,
+      "loss": 0.9558,
+      "step": 6815
+    },
+    {
+      "epoch": 0.6090200370808855,
+      "grad_norm": 0.47932013869285583,
+      "learning_rate": 3.5004009547925255e-05,
+      "loss": 0.9453,
+      "step": 6816
+    },
+    {
+      "epoch": 0.6091093886121474,
+      "grad_norm": 0.49553731083869934,
+      "learning_rate": 3.4990205627085504e-05,
+      "loss": 0.9632,
+      "step": 6817
+    },
+    {
+      "epoch": 0.6091987401434092,
+      "grad_norm": 0.49811407923698425,
+      "learning_rate": 3.497640296347603e-05,
+      "loss": 0.898,
+      "step": 6818
+    },
+    {
+      "epoch": 0.6092880916746711,
+      "grad_norm": 0.42561087012290955,
+      "learning_rate": 3.496260155825294e-05,
+      "loss": 0.9824,
+      "step": 6819
+    },
+    {
+      "epoch": 0.6093774432059329,
+      "grad_norm": 0.4683232009410858,
+      "learning_rate": 3.4948801412572255e-05,
+      "loss": 0.9188,
+      "step": 6820
+    },
+    {
+      "epoch": 0.6094667947371948,
+      "grad_norm": 0.4638214409351349,
+      "learning_rate": 3.493500252758989e-05,
+      "loss": 0.9847,
+      "step": 6821
+    },
+    {
+      "epoch": 0.6095561462684567,
+      "grad_norm": 0.47168654203414917,
+      "learning_rate": 3.492120490446164e-05,
+      "loss": 0.9237,
+      "step": 6822
+    },
+    {
+      "epoch": 0.6096454977997185,
+      "grad_norm": 0.5087332725524902,
+      "learning_rate": 3.490740854434321e-05,
+      "loss": 0.9413,
+      "step": 6823
+    },
+    {
+      "epoch": 0.6097348493309804,
+      "grad_norm": 0.5125972628593445,
+      "learning_rate": 3.489361344839018e-05,
+      "loss": 1.0025,
+      "step": 6824
+    },
+    {
+      "epoch": 0.6098242008622423,
+      "grad_norm": 0.4349963068962097,
+      "learning_rate": 3.487981961775806e-05,
+      "loss": 0.9474,
+      "step": 6825
+    },
+    {
+      "epoch": 0.6099135523935042,
+      "grad_norm": 0.5097174644470215,
+      "learning_rate": 3.486602705360223e-05,
+      "loss": 0.9926,
+      "step": 6826
+    },
+    {
+      "epoch": 0.610002903924766,
+      "grad_norm": 0.43980687856674194,
+      "learning_rate": 3.4852235757077934e-05,
+      "loss": 0.9322,
+      "step": 6827
+    },
+    {
+      "epoch": 0.6100922554560279,
+      "grad_norm": 0.48493608832359314,
+      "learning_rate": 3.483844572934036e-05,
+      "loss": 0.992,
+      "step": 6828
+    },
+    {
+      "epoch": 0.6101816069872897,
+      "grad_norm": 0.43996307253837585,
+      "learning_rate": 3.482465697154456e-05,
+      "loss": 0.9379,
+      "step": 6829
+    },
+    {
+      "epoch": 0.6102709585185516,
+      "grad_norm": 0.5646868944168091,
+      "learning_rate": 3.48108694848455e-05,
+      "loss": 0.9061,
+      "step": 6830
+    },
+    {
+      "epoch": 0.6103603100498135,
+      "grad_norm": 0.4832056164741516,
+      "learning_rate": 3.479708327039802e-05,
+      "loss": 0.9026,
+      "step": 6831
+    },
+    {
+      "epoch": 0.6104496615810754,
+      "grad_norm": 0.5810287594795227,
+      "learning_rate": 3.478329832935687e-05,
+      "loss": 0.825,
+      "step": 6832
+    },
+    {
+      "epoch": 0.6105390131123372,
+      "grad_norm": 0.5271215438842773,
+      "learning_rate": 3.47695146628767e-05,
+      "loss": 0.9671,
+      "step": 6833
+    },
+    {
+      "epoch": 0.610628364643599,
+      "grad_norm": 0.461401104927063,
+      "learning_rate": 3.475573227211201e-05,
+      "loss": 0.9199,
+      "step": 6834
+    },
+    {
+      "epoch": 0.6107177161748609,
+      "grad_norm": 0.46049854159355164,
+      "learning_rate": 3.474195115821723e-05,
+      "loss": 0.9179,
+      "step": 6835
+    },
+    {
+      "epoch": 0.6108070677061228,
+      "grad_norm": 0.521156370639801,
+      "learning_rate": 3.4728171322346694e-05,
+      "loss": 0.8643,
+      "step": 6836
+    },
+    {
+      "epoch": 0.6108964192373847,
+      "grad_norm": 0.4350111782550812,
+      "learning_rate": 3.471439276565459e-05,
+      "loss": 0.9443,
+      "step": 6837
+    },
+    {
+      "epoch": 0.6109857707686466,
+      "grad_norm": 0.4543260931968689,
+      "learning_rate": 3.470061548929502e-05,
+      "loss": 0.9858,
+      "step": 6838
+    },
+    {
+      "epoch": 0.6110751222999085,
+      "grad_norm": 0.43811121582984924,
+      "learning_rate": 3.4686839494421976e-05,
+      "loss": 0.9243,
+      "step": 6839
+    },
+    {
+      "epoch": 0.6111644738311702,
+      "grad_norm": 0.45619046688079834,
+      "learning_rate": 3.4673064782189356e-05,
+      "loss": 0.9676,
+      "step": 6840
+    },
+    {
+      "epoch": 0.6112538253624321,
+      "grad_norm": 0.39327651262283325,
+      "learning_rate": 3.4659291353750934e-05,
+      "loss": 0.9336,
+      "step": 6841
+    },
+    {
+      "epoch": 0.611343176893694,
+      "grad_norm": 0.5329780578613281,
+      "learning_rate": 3.464551921026038e-05,
+      "loss": 0.9395,
+      "step": 6842
+    },
+    {
+      "epoch": 0.6114325284249559,
+      "grad_norm": 0.5322310924530029,
+      "learning_rate": 3.463174835287125e-05,
+      "loss": 0.9203,
+      "step": 6843
+    },
+    {
+      "epoch": 0.6115218799562178,
+      "grad_norm": 0.5095454454421997,
+      "learning_rate": 3.461797878273703e-05,
+      "loss": 0.9522,
+      "step": 6844
+    },
+    {
+      "epoch": 0.6116112314874796,
+      "grad_norm": 0.5680137276649475,
+      "learning_rate": 3.460421050101103e-05,
+      "loss": 0.9741,
+      "step": 6845
+    },
+    {
+      "epoch": 0.6117005830187415,
+      "grad_norm": 0.4388297498226166,
+      "learning_rate": 3.4590443508846536e-05,
+      "loss": 0.9311,
+      "step": 6846
+    },
+    {
+      "epoch": 0.6117899345500033,
+      "grad_norm": 0.3977706730365753,
+      "learning_rate": 3.457667780739663e-05,
+      "loss": 0.9636,
+      "step": 6847
+    },
+    {
+      "epoch": 0.6118792860812652,
+      "grad_norm": 0.42693883180618286,
+      "learning_rate": 3.456291339781435e-05,
+      "loss": 0.8826,
+      "step": 6848
+    },
+    {
+      "epoch": 0.6119686376125271,
+      "grad_norm": 0.4118557870388031,
+      "learning_rate": 3.4549150281252636e-05,
+      "loss": 0.9483,
+      "step": 6849
+    },
+    {
+      "epoch": 0.612057989143789,
+      "grad_norm": 0.5658688545227051,
+      "learning_rate": 3.453538845886427e-05,
+      "loss": 0.909,
+      "step": 6850
+    },
+    {
+      "epoch": 0.6121473406750508,
+      "grad_norm": 0.48689889907836914,
+      "learning_rate": 3.452162793180198e-05,
+      "loss": 1.0,
+      "step": 6851
+    },
+    {
+      "epoch": 0.6122366922063127,
+      "grad_norm": 0.5138941407203674,
+      "learning_rate": 3.4507868701218314e-05,
+      "loss": 1.0012,
+      "step": 6852
+    },
+    {
+      "epoch": 0.6123260437375746,
+      "grad_norm": 0.5919985771179199,
+      "learning_rate": 3.4494110768265795e-05,
+      "loss": 0.9056,
+      "step": 6853
+    },
+    {
+      "epoch": 0.6124153952688364,
+      "grad_norm": 0.5183218121528625,
+      "learning_rate": 3.448035413409677e-05,
+      "loss": 0.9247,
+      "step": 6854
+    },
+    {
+      "epoch": 0.6125047468000983,
+      "grad_norm": 0.455565869808197,
+      "learning_rate": 3.446659879986351e-05,
+      "loss": 1.024,
+      "step": 6855
+    },
+    {
+      "epoch": 0.6125940983313601,
+      "grad_norm": 0.5125479698181152,
+      "learning_rate": 3.445284476671818e-05,
+      "loss": 1.0608,
+      "step": 6856
+    },
+    {
+      "epoch": 0.612683449862622,
+      "grad_norm": 0.39517349004745483,
+      "learning_rate": 3.443909203581285e-05,
+      "loss": 0.9636,
+      "step": 6857
+    },
+    {
+      "epoch": 0.6127728013938839,
+      "grad_norm": 0.5921311974525452,
+      "learning_rate": 3.44253406082994e-05,
+      "loss": 0.9022,
+      "step": 6858
+    },
+    {
+      "epoch": 0.6128621529251458,
+      "grad_norm": 0.5381450057029724,
+      "learning_rate": 3.441159048532969e-05,
+      "loss": 0.9532,
+      "step": 6859
+    },
+    {
+      "epoch": 0.6129515044564077,
+      "grad_norm": 0.4705429971218109,
+      "learning_rate": 3.439784166805544e-05,
+      "loss": 0.9147,
+      "step": 6860
+    },
+    {
+      "epoch": 0.6130408559876694,
+      "grad_norm": 0.4696353077888489,
+      "learning_rate": 3.438409415762825e-05,
+      "loss": 0.9437,
+      "step": 6861
+    },
+    {
+      "epoch": 0.6131302075189313,
+      "grad_norm": 0.5168521404266357,
+      "learning_rate": 3.437034795519963e-05,
+      "loss": 0.9752,
+      "step": 6862
+    },
+    {
+      "epoch": 0.6132195590501932,
+      "grad_norm": 0.5304763317108154,
+      "learning_rate": 3.435660306192098e-05,
+      "loss": 0.9091,
+      "step": 6863
+    },
+    {
+      "epoch": 0.6133089105814551,
+      "grad_norm": 0.460254043340683,
+      "learning_rate": 3.434285947894356e-05,
+      "loss": 0.9877,
+      "step": 6864
+    },
+    {
+      "epoch": 0.613398262112717,
+      "grad_norm": 0.5005228519439697,
+      "learning_rate": 3.432911720741855e-05,
+      "loss": 0.9853,
+      "step": 6865
+    },
+    {
+      "epoch": 0.6134876136439789,
+      "grad_norm": 0.4847668409347534,
+      "learning_rate": 3.4315376248497025e-05,
+      "loss": 0.9965,
+      "step": 6866
+    },
+    {
+      "epoch": 0.6135769651752407,
+      "grad_norm": 0.4120919704437256,
+      "learning_rate": 3.430163660332993e-05,
+      "loss": 0.9123,
+      "step": 6867
+    },
+    {
+      "epoch": 0.6136663167065025,
+      "grad_norm": 0.5527864694595337,
+      "learning_rate": 3.428789827306809e-05,
+      "loss": 0.9233,
+      "step": 6868
+    },
+    {
+      "epoch": 0.6137556682377644,
+      "grad_norm": 0.5144837498664856,
+      "learning_rate": 3.4274161258862245e-05,
+      "loss": 0.9465,
+      "step": 6869
+    },
+    {
+      "epoch": 0.6138450197690263,
+      "grad_norm": 0.4069208800792694,
+      "learning_rate": 3.426042556186303e-05,
+      "loss": 0.9784,
+      "step": 6870
+    },
+    {
+      "epoch": 0.6139343713002882,
+      "grad_norm": 0.47290289402008057,
+      "learning_rate": 3.424669118322094e-05,
+      "loss": 0.9245,
+      "step": 6871
+    },
+    {
+      "epoch": 0.61402372283155,
+      "grad_norm": 0.5234418511390686,
+      "learning_rate": 3.423295812408638e-05,
+      "loss": 0.9324,
+      "step": 6872
+    },
+    {
+      "epoch": 0.6141130743628119,
+      "grad_norm": 0.42255857586860657,
+      "learning_rate": 3.421922638560964e-05,
+      "loss": 0.9565,
+      "step": 6873
+    },
+    {
+      "epoch": 0.6142024258940738,
+      "grad_norm": 0.4056835472583771,
+      "learning_rate": 3.4205495968940907e-05,
+      "loss": 0.983,
+      "step": 6874
+    },
+    {
+      "epoch": 0.6142917774253356,
+      "grad_norm": 0.42643529176712036,
+      "learning_rate": 3.4191766875230234e-05,
+      "loss": 0.906,
+      "step": 6875
+    },
+    {
+      "epoch": 0.6143811289565975,
+      "grad_norm": 0.42716148495674133,
+      "learning_rate": 3.4178039105627594e-05,
+      "loss": 1.0189,
+      "step": 6876
+    },
+    {
+      "epoch": 0.6144704804878593,
+      "grad_norm": 0.4740234315395355,
+      "learning_rate": 3.416431266128286e-05,
+      "loss": 0.9249,
+      "step": 6877
+    },
+    {
+      "epoch": 0.6145598320191212,
+      "grad_norm": 0.41565650701522827,
+      "learning_rate": 3.4150587543345705e-05,
+      "loss": 0.9317,
+      "step": 6878
+    },
+    {
+      "epoch": 0.6146491835503831,
+      "grad_norm": 0.46336111426353455,
+      "learning_rate": 3.413686375296578e-05,
+      "loss": 0.9369,
+      "step": 6879
+    },
+    {
+      "epoch": 0.614738535081645,
+      "grad_norm": 0.4771583676338196,
+      "learning_rate": 3.4123141291292615e-05,
+      "loss": 0.9388,
+      "step": 6880
+    },
+    {
+      "epoch": 0.6148278866129069,
+      "grad_norm": 0.4211154878139496,
+      "learning_rate": 3.41094201594756e-05,
+      "loss": 0.8495,
+      "step": 6881
+    },
+    {
+      "epoch": 0.6149172381441687,
+      "grad_norm": 0.4368746280670166,
+      "learning_rate": 3.4095700358664026e-05,
+      "loss": 0.9002,
+      "step": 6882
+    },
+    {
+      "epoch": 0.6150065896754305,
+      "grad_norm": 0.4566292464733124,
+      "learning_rate": 3.4081981890007074e-05,
+      "loss": 0.9938,
+      "step": 6883
+    },
+    {
+      "epoch": 0.6150959412066924,
+      "grad_norm": 0.43355435132980347,
+      "learning_rate": 3.406826475465382e-05,
+      "loss": 0.9271,
+      "step": 6884
+    },
+    {
+      "epoch": 0.6151852927379543,
+      "grad_norm": 0.502873420715332,
+      "learning_rate": 3.4054548953753205e-05,
+      "loss": 0.9655,
+      "step": 6885
+    },
+    {
+      "epoch": 0.6152746442692162,
+      "grad_norm": 0.5166613459587097,
+      "learning_rate": 3.4040834488454086e-05,
+      "loss": 0.8589,
+      "step": 6886
+    },
+    {
+      "epoch": 0.6153639958004781,
+      "grad_norm": 0.47696536779403687,
+      "learning_rate": 3.40271213599052e-05,
+      "loss": 0.8985,
+      "step": 6887
+    },
+    {
+      "epoch": 0.61545334733174,
+      "grad_norm": 0.47705844044685364,
+      "learning_rate": 3.401340956925515e-05,
+      "loss": 0.9217,
+      "step": 6888
+    },
+    {
+      "epoch": 0.6155426988630017,
+      "grad_norm": 0.5489514470100403,
+      "learning_rate": 3.399969911765246e-05,
+      "loss": 0.9599,
+      "step": 6889
+    },
+    {
+      "epoch": 0.6156320503942636,
+      "grad_norm": 0.47389769554138184,
+      "learning_rate": 3.398599000624551e-05,
+      "loss": 0.9344,
+      "step": 6890
+    },
+    {
+      "epoch": 0.6157214019255255,
+      "grad_norm": 0.41707125306129456,
+      "learning_rate": 3.39722822361826e-05,
+      "loss": 1.0065,
+      "step": 6891
+    },
+    {
+      "epoch": 0.6158107534567874,
+      "grad_norm": 0.4215580224990845,
+      "learning_rate": 3.395857580861188e-05,
+      "loss": 0.9582,
+      "step": 6892
+    },
+    {
+      "epoch": 0.6159001049880493,
+      "grad_norm": 0.5935801863670349,
+      "learning_rate": 3.394487072468144e-05,
+      "loss": 0.8802,
+      "step": 6893
+    },
+    {
+      "epoch": 0.6159894565193111,
+      "grad_norm": 0.40050363540649414,
+      "learning_rate": 3.393116698553921e-05,
+      "loss": 0.9529,
+      "step": 6894
+    },
+    {
+      "epoch": 0.6160788080505729,
+      "grad_norm": 0.41118553280830383,
+      "learning_rate": 3.391746459233302e-05,
+      "loss": 0.9639,
+      "step": 6895
+    },
+    {
+      "epoch": 0.6161681595818348,
+      "grad_norm": 0.47675037384033203,
+      "learning_rate": 3.39037635462106e-05,
+      "loss": 0.9067,
+      "step": 6896
+    },
+    {
+      "epoch": 0.6162575111130967,
+      "grad_norm": 0.5176188945770264,
+      "learning_rate": 3.3890063848319585e-05,
+      "loss": 0.8955,
+      "step": 6897
+    },
+    {
+      "epoch": 0.6163468626443586,
+      "grad_norm": 0.4446241557598114,
+      "learning_rate": 3.3876365499807414e-05,
+      "loss": 0.9455,
+      "step": 6898
+    },
+    {
+      "epoch": 0.6164362141756204,
+      "grad_norm": 0.42455655336380005,
+      "learning_rate": 3.38626685018215e-05,
+      "loss": 0.8833,
+      "step": 6899
+    },
+    {
+      "epoch": 0.6165255657068823,
+      "grad_norm": 0.4535345733165741,
+      "learning_rate": 3.384897285550911e-05,
+      "loss": 0.9354,
+      "step": 6900
+    },
+    {
+      "epoch": 0.6166149172381442,
+      "grad_norm": 0.3765280544757843,
+      "learning_rate": 3.38352785620174e-05,
+      "loss": 0.9248,
+      "step": 6901
+    },
+    {
+      "epoch": 0.616704268769406,
+      "grad_norm": 0.42387163639068604,
+      "learning_rate": 3.382158562249342e-05,
+      "loss": 0.945,
+      "step": 6902
+    },
+    {
+      "epoch": 0.6167936203006679,
+      "grad_norm": 0.4898373484611511,
+      "learning_rate": 3.38078940380841e-05,
+      "loss": 0.9428,
+      "step": 6903
+    },
+    {
+      "epoch": 0.6168829718319297,
+      "grad_norm": 0.5467028021812439,
+      "learning_rate": 3.3794203809936235e-05,
+      "loss": 0.891,
+      "step": 6904
+    },
+    {
+      "epoch": 0.6169723233631916,
+      "grad_norm": 0.4893369972705841,
+      "learning_rate": 3.3780514939196554e-05,
+      "loss": 0.9611,
+      "step": 6905
+    },
+    {
+      "epoch": 0.6170616748944535,
+      "grad_norm": 0.6234750747680664,
+      "learning_rate": 3.376682742701161e-05,
+      "loss": 0.946,
+      "step": 6906
+    },
+    {
+      "epoch": 0.6171510264257154,
+      "grad_norm": 0.49757829308509827,
+      "learning_rate": 3.375314127452795e-05,
+      "loss": 0.8848,
+      "step": 6907
+    },
+    {
+      "epoch": 0.6172403779569773,
+      "grad_norm": 0.49189379811286926,
+      "learning_rate": 3.3739456482891854e-05,
+      "loss": 0.901,
+      "step": 6908
+    },
+    {
+      "epoch": 0.617329729488239,
+      "grad_norm": 0.49801400303840637,
+      "learning_rate": 3.37257730532496e-05,
+      "loss": 1.0374,
+      "step": 6909
+    },
+    {
+      "epoch": 0.6174190810195009,
+      "grad_norm": 0.5089427828788757,
+      "learning_rate": 3.371209098674734e-05,
+      "loss": 0.9399,
+      "step": 6910
+    },
+    {
+      "epoch": 0.6175084325507628,
+      "grad_norm": 0.43401315808296204,
+      "learning_rate": 3.3698410284531055e-05,
+      "loss": 0.9137,
+      "step": 6911
+    },
+    {
+      "epoch": 0.6175977840820247,
+      "grad_norm": 0.43453726172447205,
+      "learning_rate": 3.368473094774668e-05,
+      "loss": 0.9569,
+      "step": 6912
+    },
+    {
+      "epoch": 0.6176871356132866,
+      "grad_norm": 0.5029697418212891,
+      "learning_rate": 3.3671052977539995e-05,
+      "loss": 0.9513,
+      "step": 6913
+    },
+    {
+      "epoch": 0.6177764871445485,
+      "grad_norm": 0.3645218312740326,
+      "learning_rate": 3.3657376375056684e-05,
+      "loss": 1.0016,
+      "step": 6914
+    },
+    {
+      "epoch": 0.6178658386758104,
+      "grad_norm": 0.5805116891860962,
+      "learning_rate": 3.36437011414423e-05,
+      "loss": 1.0221,
+      "step": 6915
+    },
+    {
+      "epoch": 0.6179551902070721,
+      "grad_norm": 0.4427826702594757,
+      "learning_rate": 3.363002727784228e-05,
+      "loss": 0.9114,
+      "step": 6916
+    },
+    {
+      "epoch": 0.618044541738334,
+      "grad_norm": 0.41726526618003845,
+      "learning_rate": 3.3616354785401996e-05,
+      "loss": 1.0393,
+      "step": 6917
+    },
+    {
+      "epoch": 0.6181338932695959,
+      "grad_norm": 0.6137450337409973,
+      "learning_rate": 3.360268366526662e-05,
+      "loss": 0.9014,
+      "step": 6918
+    },
+    {
+      "epoch": 0.6182232448008578,
+      "grad_norm": 0.4480533301830292,
+      "learning_rate": 3.358901391858126e-05,
+      "loss": 0.9186,
+      "step": 6919
+    },
+    {
+      "epoch": 0.6183125963321197,
+      "grad_norm": 0.4977112114429474,
+      "learning_rate": 3.357534554649092e-05,
+      "loss": 0.8681,
+      "step": 6920
+    },
+    {
+      "epoch": 0.6184019478633815,
+      "grad_norm": 0.44784173369407654,
+      "learning_rate": 3.3561678550140466e-05,
+      "loss": 0.9717,
+      "step": 6921
+    },
+    {
+      "epoch": 0.6184912993946434,
+      "grad_norm": 0.5142949223518372,
+      "learning_rate": 3.3548012930674656e-05,
+      "loss": 0.924,
+      "step": 6922
+    },
+    {
+      "epoch": 0.6185806509259052,
+      "grad_norm": 0.4716286361217499,
+      "learning_rate": 3.3534348689238115e-05,
+      "loss": 1.0101,
+      "step": 6923
+    },
+    {
+      "epoch": 0.6186700024571671,
+      "grad_norm": 0.4497227668762207,
+      "learning_rate": 3.352068582697539e-05,
+      "loss": 1.0712,
+      "step": 6924
+    },
+    {
+      "epoch": 0.618759353988429,
+      "grad_norm": 0.4125528633594513,
+      "learning_rate": 3.3507024345030884e-05,
+      "loss": 0.9306,
+      "step": 6925
+    },
+    {
+      "epoch": 0.6188487055196908,
+      "grad_norm": 0.47352614998817444,
+      "learning_rate": 3.349336424454889e-05,
+      "loss": 0.9388,
+      "step": 6926
+    },
+    {
+      "epoch": 0.6189380570509527,
+      "grad_norm": 0.4631035029888153,
+      "learning_rate": 3.347970552667361e-05,
+      "loss": 0.9669,
+      "step": 6927
+    },
+    {
+      "epoch": 0.6190274085822146,
+      "grad_norm": 0.42280247807502747,
+      "learning_rate": 3.346604819254907e-05,
+      "loss": 0.9841,
+      "step": 6928
+    },
+    {
+      "epoch": 0.6191167601134765,
+      "grad_norm": 0.4958139955997467,
+      "learning_rate": 3.3452392243319216e-05,
+      "loss": 1.0159,
+      "step": 6929
+    },
+    {
+      "epoch": 0.6192061116447383,
+      "grad_norm": 0.4598698019981384,
+      "learning_rate": 3.343873768012792e-05,
+      "loss": 0.9439,
+      "step": 6930
+    },
+    {
+      "epoch": 0.6192954631760001,
+      "grad_norm": 0.4202096462249756,
+      "learning_rate": 3.342508450411886e-05,
+      "loss": 1.0049,
+      "step": 6931
+    },
+    {
+      "epoch": 0.619384814707262,
+      "grad_norm": 0.43892961740493774,
+      "learning_rate": 3.341143271643565e-05,
+      "loss": 0.8916,
+      "step": 6932
+    },
+    {
+      "epoch": 0.6194741662385239,
+      "grad_norm": 0.4676835834980011,
+      "learning_rate": 3.339778231822177e-05,
+      "loss": 0.9904,
+      "step": 6933
+    },
+    {
+      "epoch": 0.6195635177697858,
+      "grad_norm": 0.4813263416290283,
+      "learning_rate": 3.338413331062059e-05,
+      "loss": 0.91,
+      "step": 6934
+    },
+    {
+      "epoch": 0.6196528693010477,
+      "grad_norm": 0.442055344581604,
+      "learning_rate": 3.3370485694775354e-05,
+      "loss": 0.9984,
+      "step": 6935
+    },
+    {
+      "epoch": 0.6197422208323096,
+      "grad_norm": 0.5169895887374878,
+      "learning_rate": 3.335683947182921e-05,
+      "loss": 0.9339,
+      "step": 6936
+    },
+    {
+      "epoch": 0.6198315723635713,
+      "grad_norm": 0.4888537526130676,
+      "learning_rate": 3.3343194642925166e-05,
+      "loss": 0.9264,
+      "step": 6937
+    },
+    {
+      "epoch": 0.6199209238948332,
+      "grad_norm": 0.5011920928955078,
+      "learning_rate": 3.332955120920612e-05,
+      "loss": 0.9406,
+      "step": 6938
+    },
+    {
+      "epoch": 0.6200102754260951,
+      "grad_norm": 0.4616535007953644,
+      "learning_rate": 3.331590917181484e-05,
+      "loss": 0.945,
+      "step": 6939
+    },
+    {
+      "epoch": 0.620099626957357,
+      "grad_norm": 0.5329020619392395,
+      "learning_rate": 3.330226853189402e-05,
+      "loss": 0.8971,
+      "step": 6940
+    },
+    {
+      "epoch": 0.6201889784886189,
+      "grad_norm": 0.4462106227874756,
+      "learning_rate": 3.328862929058619e-05,
+      "loss": 0.9299,
+      "step": 6941
+    },
+    {
+      "epoch": 0.6202783300198808,
+      "grad_norm": 0.6361902952194214,
+      "learning_rate": 3.327499144903378e-05,
+      "loss": 0.9665,
+      "step": 6942
+    },
+    {
+      "epoch": 0.6203676815511426,
+      "grad_norm": 0.4172038733959198,
+      "learning_rate": 3.326135500837911e-05,
+      "loss": 0.9784,
+      "step": 6943
+    },
+    {
+      "epoch": 0.6204570330824044,
+      "grad_norm": 0.4322963058948517,
+      "learning_rate": 3.324771996976439e-05,
+      "loss": 0.9376,
+      "step": 6944
+    },
+    {
+      "epoch": 0.6205463846136663,
+      "grad_norm": 0.4417525827884674,
+      "learning_rate": 3.323408633433168e-05,
+      "loss": 0.923,
+      "step": 6945
+    },
+    {
+      "epoch": 0.6206357361449282,
+      "grad_norm": 0.5419497489929199,
+      "learning_rate": 3.322045410322296e-05,
+      "loss": 0.7987,
+      "step": 6946
+    },
+    {
+      "epoch": 0.6207250876761901,
+      "grad_norm": 0.5265266299247742,
+      "learning_rate": 3.3206823277580054e-05,
+      "loss": 0.9383,
+      "step": 6947
+    },
+    {
+      "epoch": 0.6208144392074519,
+      "grad_norm": 0.3826887905597687,
+      "learning_rate": 3.3193193858544735e-05,
+      "loss": 1.0365,
+      "step": 6948
+    },
+    {
+      "epoch": 0.6209037907387138,
+      "grad_norm": 0.44361376762390137,
+      "learning_rate": 3.317956584725855e-05,
+      "loss": 0.9833,
+      "step": 6949
+    },
+    {
+      "epoch": 0.6209931422699757,
+      "grad_norm": 0.4837126135826111,
+      "learning_rate": 3.316593924486302e-05,
+      "loss": 0.8984,
+      "step": 6950
+    },
+    {
+      "epoch": 0.6210824938012375,
+      "grad_norm": 0.4288838505744934,
+      "learning_rate": 3.315231405249951e-05,
+      "loss": 0.9326,
+      "step": 6951
+    },
+    {
+      "epoch": 0.6211718453324994,
+      "grad_norm": 0.5369590520858765,
+      "learning_rate": 3.313869027130929e-05,
+      "loss": 0.9524,
+      "step": 6952
+    },
+    {
+      "epoch": 0.6212611968637612,
+      "grad_norm": 0.5428600907325745,
+      "learning_rate": 3.3125067902433485e-05,
+      "loss": 0.9537,
+      "step": 6953
+    },
+    {
+      "epoch": 0.6213505483950231,
+      "grad_norm": 0.43962809443473816,
+      "learning_rate": 3.311144694701313e-05,
+      "loss": 0.9584,
+      "step": 6954
+    },
+    {
+      "epoch": 0.621439899926285,
+      "grad_norm": 0.40536513924598694,
+      "learning_rate": 3.3097827406189094e-05,
+      "loss": 0.96,
+      "step": 6955
+    },
+    {
+      "epoch": 0.6215292514575469,
+      "grad_norm": 0.4061243236064911,
+      "learning_rate": 3.3084209281102184e-05,
+      "loss": 0.9844,
+      "step": 6956
+    },
+    {
+      "epoch": 0.6216186029888087,
+      "grad_norm": 0.44544127583503723,
+      "learning_rate": 3.307059257289306e-05,
+      "loss": 0.9206,
+      "step": 6957
+    },
+    {
+      "epoch": 0.6217079545200705,
+      "grad_norm": 0.4827932119369507,
+      "learning_rate": 3.305697728270226e-05,
+      "loss": 0.9254,
+      "step": 6958
+    },
+    {
+      "epoch": 0.6217973060513324,
+      "grad_norm": 0.5052177906036377,
+      "learning_rate": 3.3043363411670225e-05,
+      "loss": 0.9132,
+      "step": 6959
+    },
+    {
+      "epoch": 0.6218866575825943,
+      "grad_norm": 0.5290608406066895,
+      "learning_rate": 3.302975096093723e-05,
+      "loss": 0.9137,
+      "step": 6960
+    },
+    {
+      "epoch": 0.6219760091138562,
+      "grad_norm": 0.6665788292884827,
+      "learning_rate": 3.3016139931643486e-05,
+      "loss": 0.8932,
+      "step": 6961
+    },
+    {
+      "epoch": 0.6220653606451181,
+      "grad_norm": 0.44664615392684937,
+      "learning_rate": 3.300253032492906e-05,
+      "loss": 0.9808,
+      "step": 6962
+    },
+    {
+      "epoch": 0.62215471217638,
+      "grad_norm": 0.44915464520454407,
+      "learning_rate": 3.29889221419339e-05,
+      "loss": 0.9504,
+      "step": 6963
+    },
+    {
+      "epoch": 0.6222440637076417,
+      "grad_norm": 0.5096574425697327,
+      "learning_rate": 3.297531538379782e-05,
+      "loss": 0.9388,
+      "step": 6964
+    },
+    {
+      "epoch": 0.6223334152389036,
+      "grad_norm": 0.5032405853271484,
+      "learning_rate": 3.296171005166057e-05,
+      "loss": 0.8991,
+      "step": 6965
+    },
+    {
+      "epoch": 0.6224227667701655,
+      "grad_norm": 0.43075719475746155,
+      "learning_rate": 3.29481061466617e-05,
+      "loss": 0.9858,
+      "step": 6966
+    },
+    {
+      "epoch": 0.6225121183014274,
+      "grad_norm": 0.5037357211112976,
+      "learning_rate": 3.293450366994071e-05,
+      "loss": 0.8802,
+      "step": 6967
+    },
+    {
+      "epoch": 0.6226014698326893,
+      "grad_norm": 0.3918449878692627,
+      "learning_rate": 3.292090262263696e-05,
+      "loss": 1.063,
+      "step": 6968
+    },
+    {
+      "epoch": 0.6226908213639512,
+      "grad_norm": 0.5222644209861755,
+      "learning_rate": 3.290730300588965e-05,
+      "loss": 0.8485,
+      "step": 6969
+    },
+    {
+      "epoch": 0.622780172895213,
+      "grad_norm": 0.46805500984191895,
+      "learning_rate": 3.28937048208379e-05,
+      "loss": 0.9382,
+      "step": 6970
+    },
+    {
+      "epoch": 0.6228695244264748,
+      "grad_norm": 0.5328028798103333,
+      "learning_rate": 3.288010806862071e-05,
+      "loss": 0.8772,
+      "step": 6971
+    },
+    {
+      "epoch": 0.6229588759577367,
+      "grad_norm": 0.45489615201950073,
+      "learning_rate": 3.286651275037697e-05,
+      "loss": 0.9407,
+      "step": 6972
+    },
+    {
+      "epoch": 0.6230482274889986,
+      "grad_norm": 0.4729410409927368,
+      "learning_rate": 3.285291886724541e-05,
+      "loss": 0.9488,
+      "step": 6973
+    },
+    {
+      "epoch": 0.6231375790202605,
+      "grad_norm": 0.5226532816886902,
+      "learning_rate": 3.2839326420364664e-05,
+      "loss": 0.9016,
+      "step": 6974
+    },
+    {
+      "epoch": 0.6232269305515223,
+      "grad_norm": 0.5036323070526123,
+      "learning_rate": 3.282573541087325e-05,
+      "loss": 0.89,
+      "step": 6975
+    },
+    {
+      "epoch": 0.6233162820827842,
+      "grad_norm": 0.42593827843666077,
+      "learning_rate": 3.281214583990956e-05,
+      "loss": 1.067,
+      "step": 6976
+    },
+    {
+      "epoch": 0.6234056336140461,
+      "grad_norm": 0.46937912702560425,
+      "learning_rate": 3.2798557708611864e-05,
+      "loss": 0.9958,
+      "step": 6977
+    },
+    {
+      "epoch": 0.6234949851453079,
+      "grad_norm": 0.4938536286354065,
+      "learning_rate": 3.2784971018118346e-05,
+      "loss": 0.9676,
+      "step": 6978
+    },
+    {
+      "epoch": 0.6235843366765698,
+      "grad_norm": 0.5184669494628906,
+      "learning_rate": 3.2771385769566975e-05,
+      "loss": 0.9144,
+      "step": 6979
+    },
+    {
+      "epoch": 0.6236736882078316,
+      "grad_norm": 0.46924924850463867,
+      "learning_rate": 3.275780196409569e-05,
+      "loss": 0.9918,
+      "step": 6980
+    },
+    {
+      "epoch": 0.6237630397390935,
+      "grad_norm": 0.4476311504840851,
+      "learning_rate": 3.2744219602842276e-05,
+      "loss": 0.9499,
+      "step": 6981
+    },
+    {
+      "epoch": 0.6238523912703554,
+      "grad_norm": 0.4355534315109253,
+      "learning_rate": 3.27306386869444e-05,
+      "loss": 0.912,
+      "step": 6982
+    },
+    {
+      "epoch": 0.6239417428016173,
+      "grad_norm": 0.532875120639801,
+      "learning_rate": 3.271705921753962e-05,
+      "loss": 0.9276,
+      "step": 6983
+    },
+    {
+      "epoch": 0.6240310943328792,
+      "grad_norm": 0.4715759754180908,
+      "learning_rate": 3.270348119576536e-05,
+      "loss": 0.9415,
+      "step": 6984
+    },
+    {
+      "epoch": 0.624120445864141,
+      "grad_norm": 0.5215761065483093,
+      "learning_rate": 3.26899046227589e-05,
+      "loss": 0.9365,
+      "step": 6985
+    },
+    {
+      "epoch": 0.6242097973954028,
+      "grad_norm": 0.5269849896430969,
+      "learning_rate": 3.2676329499657455e-05,
+      "loss": 0.8021,
+      "step": 6986
+    },
+    {
+      "epoch": 0.6242991489266647,
+      "grad_norm": 0.43522122502326965,
+      "learning_rate": 3.266275582759808e-05,
+      "loss": 0.9598,
+      "step": 6987
+    },
+    {
+      "epoch": 0.6243885004579266,
+      "grad_norm": 0.45015889406204224,
+      "learning_rate": 3.2649183607717706e-05,
+      "loss": 0.9508,
+      "step": 6988
+    },
+    {
+      "epoch": 0.6244778519891885,
+      "grad_norm": 0.4154300093650818,
+      "learning_rate": 3.263561284115313e-05,
+      "loss": 0.9569,
+      "step": 6989
+    },
+    {
+      "epoch": 0.6245672035204504,
+      "grad_norm": 0.39608290791511536,
+      "learning_rate": 3.262204352904108e-05,
+      "loss": 0.9678,
+      "step": 6990
+    },
+    {
+      "epoch": 0.6246565550517122,
+      "grad_norm": 0.4975006878376007,
+      "learning_rate": 3.2608475672518115e-05,
+      "loss": 0.9861,
+      "step": 6991
+    },
+    {
+      "epoch": 0.624745906582974,
+      "grad_norm": 0.507334291934967,
+      "learning_rate": 3.259490927272071e-05,
+      "loss": 0.8913,
+      "step": 6992
+    },
+    {
+      "epoch": 0.6248352581142359,
+      "grad_norm": 0.46734029054641724,
+      "learning_rate": 3.2581344330785156e-05,
+      "loss": 0.9106,
+      "step": 6993
+    },
+    {
+      "epoch": 0.6249246096454978,
+      "grad_norm": 0.45459601283073425,
+      "learning_rate": 3.2567780847847693e-05,
+      "loss": 0.9296,
+      "step": 6994
+    },
+    {
+      "epoch": 0.6250139611767597,
+      "grad_norm": 0.45823296904563904,
+      "learning_rate": 3.25542188250444e-05,
+      "loss": 0.9551,
+      "step": 6995
+    },
+    {
+      "epoch": 0.6251033127080216,
+      "grad_norm": 0.4765163064002991,
+      "learning_rate": 3.2540658263511235e-05,
+      "loss": 0.9167,
+      "step": 6996
+    },
+    {
+      "epoch": 0.6251926642392834,
+      "grad_norm": 0.4322914183139801,
+      "learning_rate": 3.252709916438404e-05,
+      "loss": 0.9762,
+      "step": 6997
+    },
+    {
+      "epoch": 0.6252820157705453,
+      "grad_norm": 0.4407638907432556,
+      "learning_rate": 3.251354152879856e-05,
+      "loss": 0.9413,
+      "step": 6998
+    },
+    {
+      "epoch": 0.6253713673018071,
+      "grad_norm": 0.44821932911872864,
+      "learning_rate": 3.2499985357890356e-05,
+      "loss": 0.9412,
+      "step": 6999
+    },
+    {
+      "epoch": 0.625460718833069,
+      "grad_norm": 0.5897961854934692,
+      "learning_rate": 3.2486430652794906e-05,
+      "loss": 1.0047,
+      "step": 7000
+    },
+    {
+      "epoch": 0.6255500703643309,
+      "grad_norm": 0.37673619389533997,
+      "learning_rate": 3.247287741464758e-05,
+      "loss": 1.0576,
+      "step": 7001
+    },
+    {
+      "epoch": 0.6256394218955927,
+      "grad_norm": 0.43584850430488586,
+      "learning_rate": 3.245932564458359e-05,
+      "loss": 0.9994,
+      "step": 7002
+    },
+    {
+      "epoch": 0.6257287734268546,
+      "grad_norm": 0.4906632602214813,
+      "learning_rate": 3.244577534373805e-05,
+      "loss": 0.9025,
+      "step": 7003
+    },
+    {
+      "epoch": 0.6258181249581165,
+      "grad_norm": 0.4197581112384796,
+      "learning_rate": 3.2432226513245935e-05,
+      "loss": 0.9456,
+      "step": 7004
+    },
+    {
+      "epoch": 0.6259074764893784,
+      "grad_norm": 0.4052625000476837,
+      "learning_rate": 3.241867915424211e-05,
+      "loss": 0.9911,
+      "step": 7005
+    },
+    {
+      "epoch": 0.6259968280206402,
+      "grad_norm": 0.46664437651634216,
+      "learning_rate": 3.240513326786132e-05,
+      "loss": 0.9425,
+      "step": 7006
+    },
+    {
+      "epoch": 0.626086179551902,
+      "grad_norm": 0.5681549310684204,
+      "learning_rate": 3.239158885523815e-05,
+      "loss": 0.9522,
+      "step": 7007
+    },
+    {
+      "epoch": 0.6261755310831639,
+      "grad_norm": 0.4695439040660858,
+      "learning_rate": 3.237804591750713e-05,
+      "loss": 1.0061,
+      "step": 7008
+    },
+    {
+      "epoch": 0.6262648826144258,
+      "grad_norm": 0.4879624545574188,
+      "learning_rate": 3.236450445580258e-05,
+      "loss": 1.0038,
+      "step": 7009
+    },
+    {
+      "epoch": 0.6263542341456877,
+      "grad_norm": 0.4483213424682617,
+      "learning_rate": 3.2350964471258785e-05,
+      "loss": 0.9624,
+      "step": 7010
+    },
+    {
+      "epoch": 0.6264435856769496,
+      "grad_norm": 0.5487602949142456,
+      "learning_rate": 3.233742596500982e-05,
+      "loss": 0.8609,
+      "step": 7011
+    },
+    {
+      "epoch": 0.6265329372082115,
+      "grad_norm": 0.6317974925041199,
+      "learning_rate": 3.2323888938189696e-05,
+      "loss": 0.8496,
+      "step": 7012
+    },
+    {
+      "epoch": 0.6266222887394732,
+      "grad_norm": 0.5183166861534119,
+      "learning_rate": 3.231035339193229e-05,
+      "loss": 0.9314,
+      "step": 7013
+    },
+    {
+      "epoch": 0.6267116402707351,
+      "grad_norm": 0.4140419363975525,
+      "learning_rate": 3.2296819327371354e-05,
+      "loss": 1.0071,
+      "step": 7014
+    },
+    {
+      "epoch": 0.626800991801997,
+      "grad_norm": 0.4753478467464447,
+      "learning_rate": 3.228328674564049e-05,
+      "loss": 0.8971,
+      "step": 7015
+    },
+    {
+      "epoch": 0.6268903433332589,
+      "grad_norm": 0.4986949861049652,
+      "learning_rate": 3.226975564787322e-05,
+      "loss": 0.9599,
+      "step": 7016
+    },
+    {
+      "epoch": 0.6269796948645208,
+      "grad_norm": 0.4630643427371979,
+      "learning_rate": 3.2256226035202895e-05,
+      "loss": 0.9859,
+      "step": 7017
+    },
+    {
+      "epoch": 0.6270690463957826,
+      "grad_norm": 0.45809265971183777,
+      "learning_rate": 3.22426979087628e-05,
+      "loss": 0.9974,
+      "step": 7018
+    },
+    {
+      "epoch": 0.6271583979270444,
+      "grad_norm": 0.5661429762840271,
+      "learning_rate": 3.222917126968601e-05,
+      "loss": 0.8188,
+      "step": 7019
+    },
+    {
+      "epoch": 0.6272477494583063,
+      "grad_norm": 0.4159087538719177,
+      "learning_rate": 3.221564611910556e-05,
+      "loss": 0.941,
+      "step": 7020
+    },
+    {
+      "epoch": 0.6273371009895682,
+      "grad_norm": 0.4369848966598511,
+      "learning_rate": 3.22021224581543e-05,
+      "loss": 0.9944,
+      "step": 7021
+    },
+    {
+      "epoch": 0.6274264525208301,
+      "grad_norm": 0.43189677596092224,
+      "learning_rate": 3.218860028796501e-05,
+      "loss": 1.0337,
+      "step": 7022
+    },
+    {
+      "epoch": 0.627515804052092,
+      "grad_norm": 0.4517824947834015,
+      "learning_rate": 3.2175079609670286e-05,
+      "loss": 0.9424,
+      "step": 7023
+    },
+    {
+      "epoch": 0.6276051555833538,
+      "grad_norm": 0.42732521891593933,
+      "learning_rate": 3.216156042440267e-05,
+      "loss": 0.9269,
+      "step": 7024
+    },
+    {
+      "epoch": 0.6276945071146157,
+      "grad_norm": 0.5549626350402832,
+      "learning_rate": 3.2148042733294494e-05,
+      "loss": 0.898,
+      "step": 7025
+    },
+    {
+      "epoch": 0.6277838586458775,
+      "grad_norm": 0.4612114429473877,
+      "learning_rate": 3.2134526537478034e-05,
+      "loss": 0.9135,
+      "step": 7026
+    },
+    {
+      "epoch": 0.6278732101771394,
+      "grad_norm": 0.42315712571144104,
+      "learning_rate": 3.21210118380854e-05,
+      "loss": 0.9782,
+      "step": 7027
+    },
+    {
+      "epoch": 0.6279625617084013,
+      "grad_norm": 0.5238030552864075,
+      "learning_rate": 3.210749863624861e-05,
+      "loss": 0.923,
+      "step": 7028
+    },
+    {
+      "epoch": 0.6280519132396631,
+      "grad_norm": 0.39960044622421265,
+      "learning_rate": 3.209398693309954e-05,
+      "loss": 0.944,
+      "step": 7029
+    },
+    {
+      "epoch": 0.628141264770925,
+      "grad_norm": 0.5192845463752747,
+      "learning_rate": 3.2080476729769916e-05,
+      "loss": 0.9699,
+      "step": 7030
+    },
+    {
+      "epoch": 0.6282306163021869,
+      "grad_norm": 0.5337046980857849,
+      "learning_rate": 3.2066968027391374e-05,
+      "loss": 0.8788,
+      "step": 7031
+    },
+    {
+      "epoch": 0.6283199678334488,
+      "grad_norm": 0.4468163549900055,
+      "learning_rate": 3.20534608270954e-05,
+      "loss": 0.9746,
+      "step": 7032
+    },
+    {
+      "epoch": 0.6284093193647106,
+      "grad_norm": 0.4410332441329956,
+      "learning_rate": 3.20399551300134e-05,
+      "loss": 0.9883,
+      "step": 7033
+    },
+    {
+      "epoch": 0.6284986708959724,
+      "grad_norm": 0.4546487629413605,
+      "learning_rate": 3.202645093727659e-05,
+      "loss": 0.9566,
+      "step": 7034
+    },
+    {
+      "epoch": 0.6285880224272343,
+      "grad_norm": 0.5063059329986572,
+      "learning_rate": 3.2012948250016084e-05,
+      "loss": 0.865,
+      "step": 7035
+    },
+    {
+      "epoch": 0.6286773739584962,
+      "grad_norm": 0.432156503200531,
+      "learning_rate": 3.1999447069362904e-05,
+      "loss": 0.9478,
+      "step": 7036
+    },
+    {
+      "epoch": 0.6287667254897581,
+      "grad_norm": 0.5517258644104004,
+      "learning_rate": 3.19859473964479e-05,
+      "loss": 0.9291,
+      "step": 7037
+    },
+    {
+      "epoch": 0.62885607702102,
+      "grad_norm": 0.44080060720443726,
+      "learning_rate": 3.197244923240182e-05,
+      "loss": 1.0247,
+      "step": 7038
+    },
+    {
+      "epoch": 0.6289454285522819,
+      "grad_norm": 0.7278925180435181,
+      "learning_rate": 3.1958952578355295e-05,
+      "loss": 0.8903,
+      "step": 7039
+    },
+    {
+      "epoch": 0.6290347800835436,
+      "grad_norm": 0.44589099287986755,
+      "learning_rate": 3.194545743543878e-05,
+      "loss": 0.9401,
+      "step": 7040
+    },
+    {
+      "epoch": 0.6291241316148055,
+      "grad_norm": 0.5441766381263733,
+      "learning_rate": 3.193196380478264e-05,
+      "loss": 0.9553,
+      "step": 7041
+    },
+    {
+      "epoch": 0.6292134831460674,
+      "grad_norm": 0.4079403579235077,
+      "learning_rate": 3.191847168751714e-05,
+      "loss": 0.9159,
+      "step": 7042
+    },
+    {
+      "epoch": 0.6293028346773293,
+      "grad_norm": 0.5653771758079529,
+      "learning_rate": 3.190498108477237e-05,
+      "loss": 0.9289,
+      "step": 7043
+    },
+    {
+      "epoch": 0.6293921862085912,
+      "grad_norm": 0.48930737376213074,
+      "learning_rate": 3.18914919976783e-05,
+      "loss": 0.9681,
+      "step": 7044
+    },
+    {
+      "epoch": 0.629481537739853,
+      "grad_norm": 0.5332466959953308,
+      "learning_rate": 3.187800442736481e-05,
+      "loss": 0.9396,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6295708892711149,
+      "grad_norm": 0.4455600678920746,
+      "learning_rate": 3.1864518374961606e-05,
+      "loss": 0.9359,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6296602408023767,
+      "grad_norm": 0.47383755445480347,
+      "learning_rate": 3.1851033841598297e-05,
+      "loss": 0.9179,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6297495923336386,
+      "grad_norm": 0.4770072102546692,
+      "learning_rate": 3.183755082840436e-05,
+      "loss": 0.9212,
+      "step": 7048
+    },
+    {
+      "epoch": 0.6298389438649005,
+      "grad_norm": 0.48469939827919006,
+      "learning_rate": 3.182406933650917e-05,
+      "loss": 0.9359,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6299282953961624,
+      "grad_norm": 0.6045317053794861,
+      "learning_rate": 3.181058936704187e-05,
+      "loss": 0.9362,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6300176469274242,
+      "grad_norm": 0.41952088475227356,
+      "learning_rate": 3.179711092113162e-05,
+      "loss": 0.9626,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6301069984586861,
+      "grad_norm": 0.5453062653541565,
+      "learning_rate": 3.178363399990735e-05,
+      "loss": 0.8763,
+      "step": 7052
+    },
+    {
+      "epoch": 0.630196349989948,
+      "grad_norm": 0.47446975111961365,
+      "learning_rate": 3.1770158604497905e-05,
+      "loss": 0.9664,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6302857015212098,
+      "grad_norm": 0.521207869052887,
+      "learning_rate": 3.175668473603199e-05,
+      "loss": 0.9685,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6303750530524717,
+      "grad_norm": 0.47734490036964417,
+      "learning_rate": 3.17432123956382e-05,
+      "loss": 0.9631,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6304644045837335,
+      "grad_norm": 0.5704666376113892,
+      "learning_rate": 3.172974158444496e-05,
+      "loss": 0.9473,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6305537561149954,
+      "grad_norm": 0.5392587184906006,
+      "learning_rate": 3.171627230358063e-05,
+      "loss": 0.8843,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6306431076462573,
+      "grad_norm": 0.5207484364509583,
+      "learning_rate": 3.1702804554173374e-05,
+      "loss": 0.9672,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6307324591775192,
+      "grad_norm": 0.3899401128292084,
+      "learning_rate": 3.1689338337351273e-05,
+      "loss": 0.9844,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6308218107087811,
+      "grad_norm": 0.4320841431617737,
+      "learning_rate": 3.1675873654242264e-05,
+      "loss": 0.9221,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6309111622400428,
+      "grad_norm": 0.4138674736022949,
+      "learning_rate": 3.1662410505974146e-05,
+      "loss": 1.0483,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6310005137713047,
+      "grad_norm": 0.5958953499794006,
+      "learning_rate": 3.164894889367463e-05,
+      "loss": 0.9235,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6310898653025666,
+      "grad_norm": 0.4245629608631134,
+      "learning_rate": 3.1635488818471246e-05,
+      "loss": 0.9503,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6311792168338285,
+      "grad_norm": 0.47923198342323303,
+      "learning_rate": 3.162203028149142e-05,
+      "loss": 0.9579,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6312685683650904,
+      "grad_norm": 0.4923160672187805,
+      "learning_rate": 3.160857328386245e-05,
+      "loss": 0.9471,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6313579198963523,
+      "grad_norm": 0.3958245515823364,
+      "learning_rate": 3.1595117826711514e-05,
+      "loss": 0.9887,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6314472714276141,
+      "grad_norm": 0.4653833508491516,
+      "learning_rate": 3.1581663911165635e-05,
+      "loss": 0.9628,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6315366229588759,
+      "grad_norm": 0.43516650795936584,
+      "learning_rate": 3.1568211538351736e-05,
+      "loss": 0.876,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6316259744901378,
+      "grad_norm": 0.5455069541931152,
+      "learning_rate": 3.15547607093966e-05,
+      "loss": 0.8408,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6317153260213997,
+      "grad_norm": 0.6274839639663696,
+      "learning_rate": 3.154131142542686e-05,
+      "loss": 0.9539,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6318046775526616,
+      "grad_norm": 0.5620449781417847,
+      "learning_rate": 3.1527863687569026e-05,
+      "loss": 0.9296,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6318940290839234,
+      "grad_norm": 0.5150429010391235,
+      "learning_rate": 3.1514417496949525e-05,
+      "loss": 0.8938,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6319833806151853,
+      "grad_norm": 0.3932952880859375,
+      "learning_rate": 3.150097285469459e-05,
+      "loss": 0.9809,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6320727321464472,
+      "grad_norm": 0.4048050045967102,
+      "learning_rate": 3.148752976193036e-05,
+      "loss": 0.9128,
+      "step": 7074
+    },
+    {
+      "epoch": 0.632162083677709,
+      "grad_norm": 0.45377394556999207,
+      "learning_rate": 3.147408821978285e-05,
+      "loss": 0.9428,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6322514352089709,
+      "grad_norm": 0.47857969999313354,
+      "learning_rate": 3.146064822937793e-05,
+      "loss": 0.9388,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6323407867402328,
+      "grad_norm": 0.42985495924949646,
+      "learning_rate": 3.144720979184133e-05,
+      "loss": 0.9978,
+      "step": 7077
+    },
+    {
+      "epoch": 0.6324301382714946,
+      "grad_norm": 0.4862077236175537,
+      "learning_rate": 3.1433772908298665e-05,
+      "loss": 0.9369,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6325194898027565,
+      "grad_norm": 0.5292137861251831,
+      "learning_rate": 3.1420337579875424e-05,
+      "loss": 0.8707,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6326088413340184,
+      "grad_norm": 0.5022360682487488,
+      "learning_rate": 3.140690380769696e-05,
+      "loss": 0.968,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6326981928652802,
+      "grad_norm": 0.45004457235336304,
+      "learning_rate": 3.139347159288849e-05,
+      "loss": 0.9509,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6327875443965421,
+      "grad_norm": 0.3994084894657135,
+      "learning_rate": 3.1380040936575094e-05,
+      "loss": 1.0183,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6328768959278039,
+      "grad_norm": 0.4302981495857239,
+      "learning_rate": 3.136661183988175e-05,
+      "loss": 1.0077,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6329662474590658,
+      "grad_norm": 0.5247629284858704,
+      "learning_rate": 3.135318430393328e-05,
+      "loss": 0.9228,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6330555989903277,
+      "grad_norm": 0.44027116894721985,
+      "learning_rate": 3.133975832985438e-05,
+      "loss": 0.9233,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6331449505215896,
+      "grad_norm": 0.4487362205982208,
+      "learning_rate": 3.1326333918769633e-05,
+      "loss": 0.9578,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6332343020528515,
+      "grad_norm": 0.4788406789302826,
+      "learning_rate": 3.1312911071803464e-05,
+      "loss": 0.9458,
+      "step": 7087
+    },
+    {
+      "epoch": 0.6333236535841132,
+      "grad_norm": 0.5990272164344788,
+      "learning_rate": 3.1299489790080184e-05,
+      "loss": 0.8684,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6334130051153751,
+      "grad_norm": 0.4570746123790741,
+      "learning_rate": 3.128607007472398e-05,
+      "loss": 1.0058,
+      "step": 7089
+    },
+    {
+      "epoch": 0.633502356646637,
+      "grad_norm": 0.45381224155426025,
+      "learning_rate": 3.127265192685887e-05,
+      "loss": 0.9837,
+      "step": 7090
+    },
+    {
+      "epoch": 0.6335917081778989,
+      "grad_norm": 0.46698102355003357,
+      "learning_rate": 3.1259235347608786e-05,
+      "loss": 0.9462,
+      "step": 7091
+    },
+    {
+      "epoch": 0.6336810597091608,
+      "grad_norm": 0.39173898100852966,
+      "learning_rate": 3.12458203380975e-05,
+      "loss": 0.9532,
+      "step": 7092
+    },
+    {
+      "epoch": 0.6337704112404227,
+      "grad_norm": 0.5905160307884216,
+      "learning_rate": 3.123240689944866e-05,
+      "loss": 0.8619,
+      "step": 7093
+    },
+    {
+      "epoch": 0.6338597627716845,
+      "grad_norm": 0.4507081210613251,
+      "learning_rate": 3.121899503278579e-05,
+      "loss": 0.9821,
+      "step": 7094
+    },
+    {
+      "epoch": 0.6339491143029463,
+      "grad_norm": 0.4796884059906006,
+      "learning_rate": 3.120558473923229e-05,
+      "loss": 0.9545,
+      "step": 7095
+    },
+    {
+      "epoch": 0.6340384658342082,
+      "grad_norm": 0.4841066896915436,
+      "learning_rate": 3.119217601991139e-05,
+      "loss": 0.8731,
+      "step": 7096
+    },
+    {
+      "epoch": 0.6341278173654701,
+      "grad_norm": 0.4773595631122589,
+      "learning_rate": 3.117876887594623e-05,
+      "loss": 0.9447,
+      "step": 7097
+    },
+    {
+      "epoch": 0.634217168896732,
+      "grad_norm": 0.5614939332008362,
+      "learning_rate": 3.116536330845979e-05,
+      "loss": 0.9185,
+      "step": 7098
+    },
+    {
+      "epoch": 0.6343065204279938,
+      "grad_norm": 0.440338671207428,
+      "learning_rate": 3.1151959318574964e-05,
+      "loss": 0.9498,
+      "step": 7099
+    },
+    {
+      "epoch": 0.6343958719592557,
+      "grad_norm": 0.43341484665870667,
+      "learning_rate": 3.113855690741443e-05,
+      "loss": 1.0054,
+      "step": 7100
+    },
+    {
+      "epoch": 0.6344852234905176,
+      "grad_norm": 0.5022625923156738,
+      "learning_rate": 3.1125156076100805e-05,
+      "loss": 0.8804,
+      "step": 7101
+    },
+    {
+      "epoch": 0.6345745750217794,
+      "grad_norm": 0.4659278392791748,
+      "learning_rate": 3.1111756825756546e-05,
+      "loss": 0.93,
+      "step": 7102
+    },
+    {
+      "epoch": 0.6346639265530413,
+      "grad_norm": 0.39844343066215515,
+      "learning_rate": 3.109835915750398e-05,
+      "loss": 1.0028,
+      "step": 7103
+    },
+    {
+      "epoch": 0.6347532780843032,
+      "grad_norm": 0.4734951853752136,
+      "learning_rate": 3.108496307246532e-05,
+      "loss": 0.9552,
+      "step": 7104
+    },
+    {
+      "epoch": 0.634842629615565,
+      "grad_norm": 0.4761861264705658,
+      "learning_rate": 3.107156857176262e-05,
+      "loss": 0.976,
+      "step": 7105
+    },
+    {
+      "epoch": 0.6349319811468269,
+      "grad_norm": 0.5420219302177429,
+      "learning_rate": 3.105817565651782e-05,
+      "loss": 0.8865,
+      "step": 7106
+    },
+    {
+      "epoch": 0.6350213326780888,
+      "grad_norm": 0.4390241205692291,
+      "learning_rate": 3.10447843278527e-05,
+      "loss": 0.9958,
+      "step": 7107
+    },
+    {
+      "epoch": 0.6351106842093507,
+      "grad_norm": 0.5415240526199341,
+      "learning_rate": 3.103139458688895e-05,
+      "loss": 0.8849,
+      "step": 7108
+    },
+    {
+      "epoch": 0.6352000357406125,
+      "grad_norm": 0.6151022911071777,
+      "learning_rate": 3.1018006434748113e-05,
+      "loss": 0.9167,
+      "step": 7109
+    },
+    {
+      "epoch": 0.6352893872718743,
+      "grad_norm": 0.4660266935825348,
+      "learning_rate": 3.100461987255155e-05,
+      "loss": 0.9687,
+      "step": 7110
+    },
+    {
+      "epoch": 0.6353787388031362,
+      "grad_norm": 0.4258476495742798,
+      "learning_rate": 3.0991234901420555e-05,
+      "loss": 0.9189,
+      "step": 7111
+    },
+    {
+      "epoch": 0.6354680903343981,
+      "grad_norm": 0.47770798206329346,
+      "learning_rate": 3.0977851522476254e-05,
+      "loss": 0.9047,
+      "step": 7112
+    },
+    {
+      "epoch": 0.63555744186566,
+      "grad_norm": 0.473283976316452,
+      "learning_rate": 3.096446973683966e-05,
+      "loss": 0.9983,
+      "step": 7113
+    },
+    {
+      "epoch": 0.6356467933969219,
+      "grad_norm": 0.4351825714111328,
+      "learning_rate": 3.0951089545631614e-05,
+      "loss": 0.9757,
+      "step": 7114
+    },
+    {
+      "epoch": 0.6357361449281838,
+      "grad_norm": 0.49671927094459534,
+      "learning_rate": 3.093771094997286e-05,
+      "loss": 0.8834,
+      "step": 7115
+    },
+    {
+      "epoch": 0.6358254964594455,
+      "grad_norm": 0.45592954754829407,
+      "learning_rate": 3.092433395098402e-05,
+      "loss": 0.9736,
+      "step": 7116
+    },
+    {
+      "epoch": 0.6359148479907074,
+      "grad_norm": 0.46876510977745056,
+      "learning_rate": 3.091095854978553e-05,
+      "loss": 0.9785,
+      "step": 7117
+    },
+    {
+      "epoch": 0.6360041995219693,
+      "grad_norm": 0.5314716696739197,
+      "learning_rate": 3.089758474749774e-05,
+      "loss": 0.9481,
+      "step": 7118
+    },
+    {
+      "epoch": 0.6360935510532312,
+      "grad_norm": 0.4798244833946228,
+      "learning_rate": 3.088421254524085e-05,
+      "loss": 0.8803,
+      "step": 7119
+    },
+    {
+      "epoch": 0.6361829025844931,
+      "grad_norm": 0.5197677612304688,
+      "learning_rate": 3.087084194413493e-05,
+      "loss": 0.8943,
+      "step": 7120
+    },
+    {
+      "epoch": 0.6362722541157549,
+      "grad_norm": 0.4947774112224579,
+      "learning_rate": 3.085747294529989e-05,
+      "loss": 0.9543,
+      "step": 7121
+    },
+    {
+      "epoch": 0.6363616056470168,
+      "grad_norm": 0.4913742244243622,
+      "learning_rate": 3.084410554985553e-05,
+      "loss": 0.9508,
+      "step": 7122
+    },
+    {
+      "epoch": 0.6364509571782786,
+      "grad_norm": 0.5018487572669983,
+      "learning_rate": 3.083073975892151e-05,
+      "loss": 0.8714,
+      "step": 7123
+    },
+    {
+      "epoch": 0.6365403087095405,
+      "grad_norm": 0.4363705813884735,
+      "learning_rate": 3.081737557361737e-05,
+      "loss": 0.946,
+      "step": 7124
+    },
+    {
+      "epoch": 0.6366296602408024,
+      "grad_norm": 0.3925338387489319,
+      "learning_rate": 3.0804012995062503e-05,
+      "loss": 0.988,
+      "step": 7125
+    },
+    {
+      "epoch": 0.6367190117720642,
+      "grad_norm": 0.4123278260231018,
+      "learning_rate": 3.0790652024376157e-05,
+      "loss": 0.9768,
+      "step": 7126
+    },
+    {
+      "epoch": 0.6368083633033261,
+      "grad_norm": 0.5473465323448181,
+      "learning_rate": 3.077729266267748e-05,
+      "loss": 0.9486,
+      "step": 7127
+    },
+    {
+      "epoch": 0.636897714834588,
+      "grad_norm": 0.5201464295387268,
+      "learning_rate": 3.076393491108542e-05,
+      "loss": 0.9381,
+      "step": 7128
+    },
+    {
+      "epoch": 0.6369870663658499,
+      "grad_norm": 0.4789254069328308,
+      "learning_rate": 3.075057877071886e-05,
+      "loss": 0.9006,
+      "step": 7129
+    },
+    {
+      "epoch": 0.6370764178971117,
+      "grad_norm": 0.4798673391342163,
+      "learning_rate": 3.0737224242696515e-05,
+      "loss": 0.9128,
+      "step": 7130
+    },
+    {
+      "epoch": 0.6371657694283736,
+      "grad_norm": 0.42166826128959656,
+      "learning_rate": 3.072387132813696e-05,
+      "loss": 1.0026,
+      "step": 7131
+    },
+    {
+      "epoch": 0.6372551209596354,
+      "grad_norm": 0.45716649293899536,
+      "learning_rate": 3.071052002815866e-05,
+      "loss": 0.9762,
+      "step": 7132
+    },
+    {
+      "epoch": 0.6373444724908973,
+      "grad_norm": 0.4412686824798584,
+      "learning_rate": 3.069717034387991e-05,
+      "loss": 0.9536,
+      "step": 7133
+    },
+    {
+      "epoch": 0.6374338240221592,
+      "grad_norm": 0.5291409492492676,
+      "learning_rate": 3.0683822276418895e-05,
+      "loss": 0.9105,
+      "step": 7134
+    },
+    {
+      "epoch": 0.6375231755534211,
+      "grad_norm": 0.41035082936286926,
+      "learning_rate": 3.0670475826893664e-05,
+      "loss": 0.9441,
+      "step": 7135
+    },
+    {
+      "epoch": 0.637612527084683,
+      "grad_norm": 0.44016411900520325,
+      "learning_rate": 3.065713099642211e-05,
+      "loss": 0.9229,
+      "step": 7136
+    },
+    {
+      "epoch": 0.6377018786159447,
+      "grad_norm": 0.5165585875511169,
+      "learning_rate": 3.0643787786122026e-05,
+      "loss": 0.9092,
+      "step": 7137
+    },
+    {
+      "epoch": 0.6377912301472066,
+      "grad_norm": 0.46023446321487427,
+      "learning_rate": 3.063044619711104e-05,
+      "loss": 0.9572,
+      "step": 7138
+    },
+    {
+      "epoch": 0.6378805816784685,
+      "grad_norm": 0.4530174136161804,
+      "learning_rate": 3.0617106230506645e-05,
+      "loss": 0.9166,
+      "step": 7139
+    },
+    {
+      "epoch": 0.6379699332097304,
+      "grad_norm": 0.45306262373924255,
+      "learning_rate": 3.0603767887426224e-05,
+      "loss": 1.0331,
+      "step": 7140
+    },
+    {
+      "epoch": 0.6380592847409923,
+      "grad_norm": 0.440663605928421,
+      "learning_rate": 3.059043116898698e-05,
+      "loss": 1.0083,
+      "step": 7141
+    },
+    {
+      "epoch": 0.6381486362722542,
+      "grad_norm": 0.48396119475364685,
+      "learning_rate": 3.057709607630601e-05,
+      "loss": 0.9481,
+      "step": 7142
+    },
+    {
+      "epoch": 0.638237987803516,
+      "grad_norm": 0.4982928931713104,
+      "learning_rate": 3.056376261050028e-05,
+      "loss": 1.0025,
+      "step": 7143
+    },
+    {
+      "epoch": 0.6383273393347778,
+      "grad_norm": 0.4071272313594818,
+      "learning_rate": 3.05504307726866e-05,
+      "loss": 1.0177,
+      "step": 7144
+    },
+    {
+      "epoch": 0.6384166908660397,
+      "grad_norm": 0.48170456290245056,
+      "learning_rate": 3.053710056398167e-05,
+      "loss": 0.8774,
+      "step": 7145
+    },
+    {
+      "epoch": 0.6385060423973016,
+      "grad_norm": 0.5535345077514648,
+      "learning_rate": 3.052377198550204e-05,
+      "loss": 0.8821,
+      "step": 7146
+    },
+    {
+      "epoch": 0.6385953939285635,
+      "grad_norm": 0.5003737807273865,
+      "learning_rate": 3.051044503836409e-05,
+      "loss": 0.9331,
+      "step": 7147
+    },
+    {
+      "epoch": 0.6386847454598253,
+      "grad_norm": 0.4324195086956024,
+      "learning_rate": 3.0497119723684108e-05,
+      "loss": 1.0236,
+      "step": 7148
+    },
+    {
+      "epoch": 0.6387740969910872,
+      "grad_norm": 0.5235957503318787,
+      "learning_rate": 3.0483796042578246e-05,
+      "loss": 0.9891,
+      "step": 7149
+    },
+    {
+      "epoch": 0.638863448522349,
+      "grad_norm": 0.5359724760055542,
+      "learning_rate": 3.047047399616251e-05,
+      "loss": 0.896,
+      "step": 7150
+    },
+    {
+      "epoch": 0.6389528000536109,
+      "grad_norm": 0.44273999333381653,
+      "learning_rate": 3.0457153585552723e-05,
+      "loss": 0.9874,
+      "step": 7151
+    },
+    {
+      "epoch": 0.6390421515848728,
+      "grad_norm": 0.41611048579216003,
+      "learning_rate": 3.0443834811864635e-05,
+      "loss": 0.9371,
+      "step": 7152
+    },
+    {
+      "epoch": 0.6391315031161346,
+      "grad_norm": 0.5038164854049683,
+      "learning_rate": 3.043051767621383e-05,
+      "loss": 0.9263,
+      "step": 7153
+    },
+    {
+      "epoch": 0.6392208546473965,
+      "grad_norm": 0.49609890580177307,
+      "learning_rate": 3.0417202179715776e-05,
+      "loss": 0.9053,
+      "step": 7154
+    },
+    {
+      "epoch": 0.6393102061786584,
+      "grad_norm": 0.4573724567890167,
+      "learning_rate": 3.0403888323485775e-05,
+      "loss": 0.9369,
+      "step": 7155
+    },
+    {
+      "epoch": 0.6393995577099203,
+      "grad_norm": 0.4710655212402344,
+      "learning_rate": 3.0390576108639e-05,
+      "loss": 0.9565,
+      "step": 7156
+    },
+    {
+      "epoch": 0.6394889092411821,
+      "grad_norm": 0.5273383259773254,
+      "learning_rate": 3.03772655362905e-05,
+      "loss": 1.0381,
+      "step": 7157
+    },
+    {
+      "epoch": 0.639578260772444,
+      "grad_norm": 0.5512682795524597,
+      "learning_rate": 3.0363956607555177e-05,
+      "loss": 0.9397,
+      "step": 7158
+    },
+    {
+      "epoch": 0.6396676123037058,
+      "grad_norm": 0.45379549264907837,
+      "learning_rate": 3.0350649323547796e-05,
+      "loss": 0.9626,
+      "step": 7159
+    },
+    {
+      "epoch": 0.6397569638349677,
+      "grad_norm": 0.4535743296146393,
+      "learning_rate": 3.0337343685383e-05,
+      "loss": 0.948,
+      "step": 7160
+    },
+    {
+      "epoch": 0.6398463153662296,
+      "grad_norm": 0.45462849736213684,
+      "learning_rate": 3.0324039694175233e-05,
+      "loss": 1.0285,
+      "step": 7161
+    },
+    {
+      "epoch": 0.6399356668974915,
+      "grad_norm": 0.5407901406288147,
+      "learning_rate": 3.0310737351038875e-05,
+      "loss": 0.9403,
+      "step": 7162
+    },
+    {
+      "epoch": 0.6400250184287534,
+      "grad_norm": 0.410166472196579,
+      "learning_rate": 3.029743665708814e-05,
+      "loss": 1.0002,
+      "step": 7163
+    },
+    {
+      "epoch": 0.6401143699600151,
+      "grad_norm": 0.501658022403717,
+      "learning_rate": 3.0284137613437098e-05,
+      "loss": 0.9004,
+      "step": 7164
+    },
+    {
+      "epoch": 0.640203721491277,
+      "grad_norm": 0.4671003818511963,
+      "learning_rate": 3.027084022119969e-05,
+      "loss": 0.9121,
+      "step": 7165
+    },
+    {
+      "epoch": 0.6402930730225389,
+      "grad_norm": 0.4434639811515808,
+      "learning_rate": 3.0257544481489712e-05,
+      "loss": 0.9303,
+      "step": 7166
+    },
+    {
+      "epoch": 0.6403824245538008,
+      "grad_norm": 0.46848881244659424,
+      "learning_rate": 3.024425039542082e-05,
+      "loss": 0.8991,
+      "step": 7167
+    },
+    {
+      "epoch": 0.6404717760850627,
+      "grad_norm": 0.4049372971057892,
+      "learning_rate": 3.0230957964106532e-05,
+      "loss": 0.9814,
+      "step": 7168
+    },
+    {
+      "epoch": 0.6405611276163246,
+      "grad_norm": 0.5770232677459717,
+      "learning_rate": 3.0217667188660248e-05,
+      "loss": 0.9142,
+      "step": 7169
+    },
+    {
+      "epoch": 0.6406504791475864,
+      "grad_norm": 0.5074619650840759,
+      "learning_rate": 3.0204378070195218e-05,
+      "loss": 0.9127,
+      "step": 7170
+    },
+    {
+      "epoch": 0.6407398306788482,
+      "grad_norm": 0.539682924747467,
+      "learning_rate": 3.01910906098245e-05,
+      "loss": 0.9482,
+      "step": 7171
+    },
+    {
+      "epoch": 0.6408291822101101,
+      "grad_norm": 0.5305405259132385,
+      "learning_rate": 3.0177804808661103e-05,
+      "loss": 0.9165,
+      "step": 7172
+    },
+    {
+      "epoch": 0.640918533741372,
+      "grad_norm": 0.39787405729293823,
+      "learning_rate": 3.0164520667817842e-05,
+      "loss": 0.9578,
+      "step": 7173
+    },
+    {
+      "epoch": 0.6410078852726339,
+      "grad_norm": 0.40553003549575806,
+      "learning_rate": 3.01512381884074e-05,
+      "loss": 1.0191,
+      "step": 7174
+    },
+    {
+      "epoch": 0.6410972368038957,
+      "grad_norm": 0.5072142481803894,
+      "learning_rate": 3.0137957371542336e-05,
+      "loss": 0.9339,
+      "step": 7175
+    },
+    {
+      "epoch": 0.6411865883351576,
+      "grad_norm": 0.402472585439682,
+      "learning_rate": 3.0124678218335058e-05,
+      "loss": 0.919,
+      "step": 7176
+    },
+    {
+      "epoch": 0.6412759398664195,
+      "grad_norm": 0.5144462585449219,
+      "learning_rate": 3.0111400729897833e-05,
+      "loss": 0.9673,
+      "step": 7177
+    },
+    {
+      "epoch": 0.6413652913976813,
+      "grad_norm": 0.5433812141418457,
+      "learning_rate": 3.009812490734279e-05,
+      "loss": 0.9127,
+      "step": 7178
+    },
+    {
+      "epoch": 0.6414546429289432,
+      "grad_norm": 0.522948145866394,
+      "learning_rate": 3.008485075178194e-05,
+      "loss": 0.8578,
+      "step": 7179
+    },
+    {
+      "epoch": 0.641543994460205,
+      "grad_norm": 0.4444359242916107,
+      "learning_rate": 3.0071578264327116e-05,
+      "loss": 0.9434,
+      "step": 7180
+    },
+    {
+      "epoch": 0.6416333459914669,
+      "grad_norm": 0.5288926959037781,
+      "learning_rate": 3.005830744609003e-05,
+      "loss": 0.8699,
+      "step": 7181
+    },
+    {
+      "epoch": 0.6417226975227288,
+      "grad_norm": 0.5076068639755249,
+      "learning_rate": 3.004503829818225e-05,
+      "loss": 0.905,
+      "step": 7182
+    },
+    {
+      "epoch": 0.6418120490539907,
+      "grad_norm": 0.49402233958244324,
+      "learning_rate": 3.003177082171523e-05,
+      "loss": 0.9201,
+      "step": 7183
+    },
+    {
+      "epoch": 0.6419014005852526,
+      "grad_norm": 0.5896358489990234,
+      "learning_rate": 3.0018505017800246e-05,
+      "loss": 0.8503,
+      "step": 7184
+    },
+    {
+      "epoch": 0.6419907521165144,
+      "grad_norm": 0.5615974068641663,
+      "learning_rate": 3.0005240887548445e-05,
+      "loss": 1.0653,
+      "step": 7185
+    },
+    {
+      "epoch": 0.6420801036477762,
+      "grad_norm": 0.42065030336380005,
+      "learning_rate": 2.9991978432070856e-05,
+      "loss": 0.9577,
+      "step": 7186
+    },
+    {
+      "epoch": 0.6421694551790381,
+      "grad_norm": 0.44661208987236023,
+      "learning_rate": 2.9978717652478344e-05,
+      "loss": 0.999,
+      "step": 7187
+    },
+    {
+      "epoch": 0.6422588067103,
+      "grad_norm": 0.489004909992218,
+      "learning_rate": 2.9965458549881638e-05,
+      "loss": 0.872,
+      "step": 7188
+    },
+    {
+      "epoch": 0.6423481582415619,
+      "grad_norm": 0.43301481008529663,
+      "learning_rate": 2.9952201125391332e-05,
+      "loss": 0.9497,
+      "step": 7189
+    },
+    {
+      "epoch": 0.6424375097728238,
+      "grad_norm": 0.46835726499557495,
+      "learning_rate": 2.993894538011789e-05,
+      "loss": 1.0229,
+      "step": 7190
+    },
+    {
+      "epoch": 0.6425268613040857,
+      "grad_norm": 0.4566435217857361,
+      "learning_rate": 2.9925691315171594e-05,
+      "loss": 0.9432,
+      "step": 7191
+    },
+    {
+      "epoch": 0.6426162128353474,
+      "grad_norm": 0.490376353263855,
+      "learning_rate": 2.9912438931662624e-05,
+      "loss": 0.9158,
+      "step": 7192
+    },
+    {
+      "epoch": 0.6427055643666093,
+      "grad_norm": 0.45299986004829407,
+      "learning_rate": 2.9899188230701014e-05,
+      "loss": 0.942,
+      "step": 7193
+    },
+    {
+      "epoch": 0.6427949158978712,
+      "grad_norm": 0.40514639019966125,
+      "learning_rate": 2.9885939213396647e-05,
+      "loss": 0.987,
+      "step": 7194
+    },
+    {
+      "epoch": 0.6428842674291331,
+      "grad_norm": 0.46090513467788696,
+      "learning_rate": 2.987269188085927e-05,
+      "loss": 1.0108,
+      "step": 7195
+    },
+    {
+      "epoch": 0.642973618960395,
+      "grad_norm": 0.5533647537231445,
+      "learning_rate": 2.9859446234198494e-05,
+      "loss": 0.8314,
+      "step": 7196
+    },
+    {
+      "epoch": 0.6430629704916568,
+      "grad_norm": 0.5012393593788147,
+      "learning_rate": 2.9846202274523776e-05,
+      "loss": 0.8779,
+      "step": 7197
+    },
+    {
+      "epoch": 0.6431523220229187,
+      "grad_norm": 0.4513954818248749,
+      "learning_rate": 2.9832960002944454e-05,
+      "loss": 0.855,
+      "step": 7198
+    },
+    {
+      "epoch": 0.6432416735541805,
+      "grad_norm": 0.49379584193229675,
+      "learning_rate": 2.981971942056968e-05,
+      "loss": 0.9278,
+      "step": 7199
+    },
+    {
+      "epoch": 0.6433310250854424,
+      "grad_norm": 0.43649551272392273,
+      "learning_rate": 2.9806480528508517e-05,
+      "loss": 0.9354,
+      "step": 7200
+    },
+    {
+      "epoch": 0.6434203766167043,
+      "grad_norm": 0.35923606157302856,
+      "learning_rate": 2.9793243327869868e-05,
+      "loss": 0.9558,
+      "step": 7201
+    },
+    {
+      "epoch": 0.6435097281479661,
+      "grad_norm": 0.4612623453140259,
+      "learning_rate": 2.978000781976248e-05,
+      "loss": 0.9682,
+      "step": 7202
+    },
+    {
+      "epoch": 0.643599079679228,
+      "grad_norm": 0.4402258098125458,
+      "learning_rate": 2.9766774005294952e-05,
+      "loss": 1.0167,
+      "step": 7203
+    },
+    {
+      "epoch": 0.6436884312104899,
+      "grad_norm": 0.6181815266609192,
+      "learning_rate": 2.9753541885575777e-05,
+      "loss": 0.877,
+      "step": 7204
+    },
+    {
+      "epoch": 0.6437777827417518,
+      "grad_norm": 0.5466234683990479,
+      "learning_rate": 2.9740311461713273e-05,
+      "loss": 0.936,
+      "step": 7205
+    },
+    {
+      "epoch": 0.6438671342730136,
+      "grad_norm": 0.4415399432182312,
+      "learning_rate": 2.9727082734815637e-05,
+      "loss": 0.9895,
+      "step": 7206
+    },
+    {
+      "epoch": 0.6439564858042754,
+      "grad_norm": 0.47382205724716187,
+      "learning_rate": 2.9713855705990923e-05,
+      "loss": 0.9684,
+      "step": 7207
+    },
+    {
+      "epoch": 0.6440458373355373,
+      "grad_norm": 0.5399396419525146,
+      "learning_rate": 2.970063037634703e-05,
+      "loss": 0.8178,
+      "step": 7208
+    },
+    {
+      "epoch": 0.6441351888667992,
+      "grad_norm": 0.5536864995956421,
+      "learning_rate": 2.9687406746991708e-05,
+      "loss": 0.9333,
+      "step": 7209
+    },
+    {
+      "epoch": 0.6442245403980611,
+      "grad_norm": 0.4546963572502136,
+      "learning_rate": 2.967418481903259e-05,
+      "loss": 0.993,
+      "step": 7210
+    },
+    {
+      "epoch": 0.644313891929323,
+      "grad_norm": 0.42409881949424744,
+      "learning_rate": 2.966096459357718e-05,
+      "loss": 0.9383,
+      "step": 7211
+    },
+    {
+      "epoch": 0.6444032434605848,
+      "grad_norm": 0.4200609624385834,
+      "learning_rate": 2.9647746071732757e-05,
+      "loss": 0.8847,
+      "step": 7212
+    },
+    {
+      "epoch": 0.6444925949918466,
+      "grad_norm": 0.5027011632919312,
+      "learning_rate": 2.9634529254606542e-05,
+      "loss": 0.8904,
+      "step": 7213
+    },
+    {
+      "epoch": 0.6445819465231085,
+      "grad_norm": 0.4371449053287506,
+      "learning_rate": 2.962131414330558e-05,
+      "loss": 0.9849,
+      "step": 7214
+    },
+    {
+      "epoch": 0.6446712980543704,
+      "grad_norm": 0.42081284523010254,
+      "learning_rate": 2.9608100738936783e-05,
+      "loss": 0.9994,
+      "step": 7215
+    },
+    {
+      "epoch": 0.6447606495856323,
+      "grad_norm": 0.454153448343277,
+      "learning_rate": 2.9594889042606923e-05,
+      "loss": 0.9647,
+      "step": 7216
+    },
+    {
+      "epoch": 0.6448500011168942,
+      "grad_norm": 0.4745718836784363,
+      "learning_rate": 2.958167905542259e-05,
+      "loss": 0.959,
+      "step": 7217
+    },
+    {
+      "epoch": 0.644939352648156,
+      "grad_norm": 0.5167459845542908,
+      "learning_rate": 2.9568470778490287e-05,
+      "loss": 0.9293,
+      "step": 7218
+    },
+    {
+      "epoch": 0.6450287041794178,
+      "grad_norm": 0.5096081495285034,
+      "learning_rate": 2.9555264212916334e-05,
+      "loss": 0.9004,
+      "step": 7219
+    },
+    {
+      "epoch": 0.6451180557106797,
+      "grad_norm": 0.5786259770393372,
+      "learning_rate": 2.9542059359806935e-05,
+      "loss": 1.0463,
+      "step": 7220
+    },
+    {
+      "epoch": 0.6452074072419416,
+      "grad_norm": 0.3960155248641968,
+      "learning_rate": 2.9528856220268147e-05,
+      "loss": 0.9449,
+      "step": 7221
+    },
+    {
+      "epoch": 0.6452967587732035,
+      "grad_norm": 0.48387083411216736,
+      "learning_rate": 2.951565479540584e-05,
+      "loss": 0.9286,
+      "step": 7222
+    },
+    {
+      "epoch": 0.6453861103044654,
+      "grad_norm": 0.46584317088127136,
+      "learning_rate": 2.9502455086325787e-05,
+      "loss": 0.922,
+      "step": 7223
+    },
+    {
+      "epoch": 0.6454754618357272,
+      "grad_norm": 0.42123234272003174,
+      "learning_rate": 2.9489257094133616e-05,
+      "loss": 0.9701,
+      "step": 7224
+    },
+    {
+      "epoch": 0.6455648133669891,
+      "grad_norm": 0.5903416275978088,
+      "learning_rate": 2.9476060819934786e-05,
+      "loss": 0.896,
+      "step": 7225
+    },
+    {
+      "epoch": 0.6456541648982509,
+      "grad_norm": 0.4230227768421173,
+      "learning_rate": 2.946286626483463e-05,
+      "loss": 0.9556,
+      "step": 7226
+    },
+    {
+      "epoch": 0.6457435164295128,
+      "grad_norm": 0.5032885670661926,
+      "learning_rate": 2.9449673429938342e-05,
+      "loss": 0.934,
+      "step": 7227
+    },
+    {
+      "epoch": 0.6458328679607747,
+      "grad_norm": 0.5191749334335327,
+      "learning_rate": 2.943648231635095e-05,
+      "loss": 0.8587,
+      "step": 7228
+    },
+    {
+      "epoch": 0.6459222194920365,
+      "grad_norm": 0.49752548336982727,
+      "learning_rate": 2.942329292517736e-05,
+      "loss": 0.9805,
+      "step": 7229
+    },
+    {
+      "epoch": 0.6460115710232984,
+      "grad_norm": 0.5034040212631226,
+      "learning_rate": 2.9410105257522314e-05,
+      "loss": 0.9645,
+      "step": 7230
+    },
+    {
+      "epoch": 0.6461009225545603,
+      "grad_norm": 0.5222768783569336,
+      "learning_rate": 2.9396919314490447e-05,
+      "loss": 0.9295,
+      "step": 7231
+    },
+    {
+      "epoch": 0.6461902740858222,
+      "grad_norm": 0.4623466730117798,
+      "learning_rate": 2.9383735097186175e-05,
+      "loss": 0.9802,
+      "step": 7232
+    },
+    {
+      "epoch": 0.646279625617084,
+      "grad_norm": 0.49886736273765564,
+      "learning_rate": 2.9370552606713852e-05,
+      "loss": 0.9273,
+      "step": 7233
+    },
+    {
+      "epoch": 0.6463689771483458,
+      "grad_norm": 0.547577440738678,
+      "learning_rate": 2.935737184417764e-05,
+      "loss": 0.9833,
+      "step": 7234
+    },
+    {
+      "epoch": 0.6464583286796077,
+      "grad_norm": 0.4106947183609009,
+      "learning_rate": 2.9344192810681577e-05,
+      "loss": 1.0004,
+      "step": 7235
+    },
+    {
+      "epoch": 0.6465476802108696,
+      "grad_norm": 0.41953131556510925,
+      "learning_rate": 2.933101550732953e-05,
+      "loss": 0.9349,
+      "step": 7236
+    },
+    {
+      "epoch": 0.6466370317421315,
+      "grad_norm": 0.4484841823577881,
+      "learning_rate": 2.9317839935225254e-05,
+      "loss": 0.9114,
+      "step": 7237
+    },
+    {
+      "epoch": 0.6467263832733934,
+      "grad_norm": 0.4939296245574951,
+      "learning_rate": 2.9304666095472334e-05,
+      "loss": 0.9474,
+      "step": 7238
+    },
+    {
+      "epoch": 0.6468157348046553,
+      "grad_norm": 0.409227579832077,
+      "learning_rate": 2.9291493989174234e-05,
+      "loss": 1.0057,
+      "step": 7239
+    },
+    {
+      "epoch": 0.646905086335917,
+      "grad_norm": 0.44199416041374207,
+      "learning_rate": 2.9278323617434245e-05,
+      "loss": 0.899,
+      "step": 7240
+    },
+    {
+      "epoch": 0.6469944378671789,
+      "grad_norm": 0.4873824417591095,
+      "learning_rate": 2.9265154981355547e-05,
+      "loss": 0.9753,
+      "step": 7241
+    },
+    {
+      "epoch": 0.6470837893984408,
+      "grad_norm": 0.4727482497692108,
+      "learning_rate": 2.9251988082041115e-05,
+      "loss": 0.9691,
+      "step": 7242
+    },
+    {
+      "epoch": 0.6471731409297027,
+      "grad_norm": 0.4748595952987671,
+      "learning_rate": 2.9238822920593844e-05,
+      "loss": 0.993,
+      "step": 7243
+    },
+    {
+      "epoch": 0.6472624924609646,
+      "grad_norm": 0.490640252828598,
+      "learning_rate": 2.9225659498116452e-05,
+      "loss": 0.9925,
+      "step": 7244
+    },
+    {
+      "epoch": 0.6473518439922265,
+      "grad_norm": 0.5214620232582092,
+      "learning_rate": 2.9212497815711516e-05,
+      "loss": 0.9996,
+      "step": 7245
+    },
+    {
+      "epoch": 0.6474411955234883,
+      "grad_norm": 0.5297476649284363,
+      "learning_rate": 2.9199337874481465e-05,
+      "loss": 0.9351,
+      "step": 7246
+    },
+    {
+      "epoch": 0.6475305470547501,
+      "grad_norm": 0.4094926416873932,
+      "learning_rate": 2.9186179675528597e-05,
+      "loss": 0.9445,
+      "step": 7247
+    },
+    {
+      "epoch": 0.647619898586012,
+      "grad_norm": 0.5132883787155151,
+      "learning_rate": 2.9173023219955032e-05,
+      "loss": 0.8968,
+      "step": 7248
+    },
+    {
+      "epoch": 0.6477092501172739,
+      "grad_norm": 0.5564563870429993,
+      "learning_rate": 2.9159868508862766e-05,
+      "loss": 0.8778,
+      "step": 7249
+    },
+    {
+      "epoch": 0.6477986016485358,
+      "grad_norm": 0.5214046835899353,
+      "learning_rate": 2.9146715543353652e-05,
+      "loss": 0.9337,
+      "step": 7250
+    },
+    {
+      "epoch": 0.6478879531797976,
+      "grad_norm": 0.38234221935272217,
+      "learning_rate": 2.9133564324529415e-05,
+      "loss": 0.9849,
+      "step": 7251
+    },
+    {
+      "epoch": 0.6479773047110595,
+      "grad_norm": 0.4253350794315338,
+      "learning_rate": 2.912041485349157e-05,
+      "loss": 0.9478,
+      "step": 7252
+    },
+    {
+      "epoch": 0.6480666562423214,
+      "grad_norm": 0.436994731426239,
+      "learning_rate": 2.9107267131341537e-05,
+      "loss": 0.9162,
+      "step": 7253
+    },
+    {
+      "epoch": 0.6481560077735832,
+      "grad_norm": 0.4581502676010132,
+      "learning_rate": 2.9094121159180588e-05,
+      "loss": 0.9918,
+      "step": 7254
+    },
+    {
+      "epoch": 0.6482453593048451,
+      "grad_norm": 0.45387864112854004,
+      "learning_rate": 2.908097693810983e-05,
+      "loss": 0.8974,
+      "step": 7255
+    },
+    {
+      "epoch": 0.6483347108361069,
+      "grad_norm": 0.4383721947669983,
+      "learning_rate": 2.9067834469230225e-05,
+      "loss": 0.9876,
+      "step": 7256
+    },
+    {
+      "epoch": 0.6484240623673688,
+      "grad_norm": 0.4127906858921051,
+      "learning_rate": 2.9054693753642614e-05,
+      "loss": 0.9452,
+      "step": 7257
+    },
+    {
+      "epoch": 0.6485134138986307,
+      "grad_norm": 0.530967116355896,
+      "learning_rate": 2.9041554792447655e-05,
+      "loss": 0.918,
+      "step": 7258
+    },
+    {
+      "epoch": 0.6486027654298926,
+      "grad_norm": 0.48077839612960815,
+      "learning_rate": 2.9028417586745887e-05,
+      "loss": 0.9524,
+      "step": 7259
+    },
+    {
+      "epoch": 0.6486921169611545,
+      "grad_norm": 0.47966933250427246,
+      "learning_rate": 2.9015282137637688e-05,
+      "loss": 0.9938,
+      "step": 7260
+    },
+    {
+      "epoch": 0.6487814684924162,
+      "grad_norm": 0.5792236328125,
+      "learning_rate": 2.900214844622331e-05,
+      "loss": 0.9617,
+      "step": 7261
+    },
+    {
+      "epoch": 0.6488708200236781,
+      "grad_norm": 0.5350083708763123,
+      "learning_rate": 2.8989016513602802e-05,
+      "loss": 0.9807,
+      "step": 7262
+    },
+    {
+      "epoch": 0.64896017155494,
+      "grad_norm": 0.4303712844848633,
+      "learning_rate": 2.8975886340876117e-05,
+      "loss": 0.9293,
+      "step": 7263
+    },
+    {
+      "epoch": 0.6490495230862019,
+      "grad_norm": 0.451147198677063,
+      "learning_rate": 2.896275792914306e-05,
+      "loss": 1.0082,
+      "step": 7264
+    },
+    {
+      "epoch": 0.6491388746174638,
+      "grad_norm": 0.37288275361061096,
+      "learning_rate": 2.8949631279503264e-05,
+      "loss": 1.0025,
+      "step": 7265
+    },
+    {
+      "epoch": 0.6492282261487257,
+      "grad_norm": 0.480672687292099,
+      "learning_rate": 2.8936506393056223e-05,
+      "loss": 0.9353,
+      "step": 7266
+    },
+    {
+      "epoch": 0.6493175776799875,
+      "grad_norm": 0.547631561756134,
+      "learning_rate": 2.89233832709013e-05,
+      "loss": 0.925,
+      "step": 7267
+    },
+    {
+      "epoch": 0.6494069292112493,
+      "grad_norm": 0.43532001972198486,
+      "learning_rate": 2.8910261914137682e-05,
+      "loss": 0.9405,
+      "step": 7268
+    },
+    {
+      "epoch": 0.6494962807425112,
+      "grad_norm": 0.45452019572257996,
+      "learning_rate": 2.8897142323864433e-05,
+      "loss": 0.9886,
+      "step": 7269
+    },
+    {
+      "epoch": 0.6495856322737731,
+      "grad_norm": 0.4508405923843384,
+      "learning_rate": 2.8884024501180456e-05,
+      "loss": 0.9328,
+      "step": 7270
+    },
+    {
+      "epoch": 0.649674983805035,
+      "grad_norm": 0.47930291295051575,
+      "learning_rate": 2.887090844718453e-05,
+      "loss": 0.9157,
+      "step": 7271
+    },
+    {
+      "epoch": 0.6497643353362968,
+      "grad_norm": 0.47216659784317017,
+      "learning_rate": 2.8857794162975214e-05,
+      "loss": 0.9874,
+      "step": 7272
+    },
+    {
+      "epoch": 0.6498536868675587,
+      "grad_norm": 0.4564785063266754,
+      "learning_rate": 2.8844681649651e-05,
+      "loss": 0.9879,
+      "step": 7273
+    },
+    {
+      "epoch": 0.6499430383988205,
+      "grad_norm": 0.5167818069458008,
+      "learning_rate": 2.8831570908310202e-05,
+      "loss": 0.9293,
+      "step": 7274
+    },
+    {
+      "epoch": 0.6500323899300824,
+      "grad_norm": 0.4396851658821106,
+      "learning_rate": 2.8818461940050967e-05,
+      "loss": 0.8964,
+      "step": 7275
+    },
+    {
+      "epoch": 0.6501217414613443,
+      "grad_norm": 0.5146901607513428,
+      "learning_rate": 2.8805354745971337e-05,
+      "loss": 0.9672,
+      "step": 7276
+    },
+    {
+      "epoch": 0.6502110929926062,
+      "grad_norm": 0.5108723044395447,
+      "learning_rate": 2.879224932716918e-05,
+      "loss": 0.9948,
+      "step": 7277
+    },
+    {
+      "epoch": 0.650300444523868,
+      "grad_norm": 0.469088613986969,
+      "learning_rate": 2.877914568474218e-05,
+      "loss": 0.9109,
+      "step": 7278
+    },
+    {
+      "epoch": 0.6503897960551299,
+      "grad_norm": 0.4965616464614868,
+      "learning_rate": 2.8766043819787925e-05,
+      "loss": 0.8499,
+      "step": 7279
+    },
+    {
+      "epoch": 0.6504791475863918,
+      "grad_norm": 0.47035840153694153,
+      "learning_rate": 2.875294373340384e-05,
+      "loss": 0.9244,
+      "step": 7280
+    },
+    {
+      "epoch": 0.6505684991176536,
+      "grad_norm": 0.5100567936897278,
+      "learning_rate": 2.8739845426687218e-05,
+      "loss": 0.9717,
+      "step": 7281
+    },
+    {
+      "epoch": 0.6506578506489155,
+      "grad_norm": 0.40733376145362854,
+      "learning_rate": 2.8726748900735133e-05,
+      "loss": 0.9693,
+      "step": 7282
+    },
+    {
+      "epoch": 0.6507472021801773,
+      "grad_norm": 0.4507538676261902,
+      "learning_rate": 2.8713654156644588e-05,
+      "loss": 0.9329,
+      "step": 7283
+    },
+    {
+      "epoch": 0.6508365537114392,
+      "grad_norm": 0.44749560952186584,
+      "learning_rate": 2.87005611955124e-05,
+      "loss": 0.9209,
+      "step": 7284
+    },
+    {
+      "epoch": 0.6509259052427011,
+      "grad_norm": 0.3972071409225464,
+      "learning_rate": 2.8687470018435246e-05,
+      "loss": 1.0053,
+      "step": 7285
+    },
+    {
+      "epoch": 0.651015256773963,
+      "grad_norm": 0.4977121353149414,
+      "learning_rate": 2.867438062650966e-05,
+      "loss": 0.9951,
+      "step": 7286
+    },
+    {
+      "epoch": 0.6511046083052249,
+      "grad_norm": 0.42729929089546204,
+      "learning_rate": 2.866129302083201e-05,
+      "loss": 0.9505,
+      "step": 7287
+    },
+    {
+      "epoch": 0.6511939598364866,
+      "grad_norm": 0.3862477242946625,
+      "learning_rate": 2.8648207202498524e-05,
+      "loss": 0.9529,
+      "step": 7288
+    },
+    {
+      "epoch": 0.6512833113677485,
+      "grad_norm": 0.5908159017562866,
+      "learning_rate": 2.8635123172605273e-05,
+      "loss": 0.9252,
+      "step": 7289
+    },
+    {
+      "epoch": 0.6513726628990104,
+      "grad_norm": 0.5458805561065674,
+      "learning_rate": 2.8622040932248196e-05,
+      "loss": 0.9316,
+      "step": 7290
+    },
+    {
+      "epoch": 0.6514620144302723,
+      "grad_norm": 0.5677804350852966,
+      "learning_rate": 2.8608960482523056e-05,
+      "loss": 0.9166,
+      "step": 7291
+    },
+    {
+      "epoch": 0.6515513659615342,
+      "grad_norm": 0.57185298204422,
+      "learning_rate": 2.859588182452551e-05,
+      "loss": 0.9909,
+      "step": 7292
+    },
+    {
+      "epoch": 0.6516407174927961,
+      "grad_norm": 0.4821870028972626,
+      "learning_rate": 2.8582804959350994e-05,
+      "loss": 0.9059,
+      "step": 7293
+    },
+    {
+      "epoch": 0.651730069024058,
+      "grad_norm": 0.39487797021865845,
+      "learning_rate": 2.8569729888094853e-05,
+      "loss": 0.9615,
+      "step": 7294
+    },
+    {
+      "epoch": 0.6518194205553197,
+      "grad_norm": 0.5088570713996887,
+      "learning_rate": 2.8556656611852274e-05,
+      "loss": 0.9741,
+      "step": 7295
+    },
+    {
+      "epoch": 0.6519087720865816,
+      "grad_norm": 0.4211069941520691,
+      "learning_rate": 2.8543585131718263e-05,
+      "loss": 1.0125,
+      "step": 7296
+    },
+    {
+      "epoch": 0.6519981236178435,
+      "grad_norm": 0.46207818388938904,
+      "learning_rate": 2.853051544878771e-05,
+      "loss": 0.9296,
+      "step": 7297
+    },
+    {
+      "epoch": 0.6520874751491054,
+      "grad_norm": 0.454182893037796,
+      "learning_rate": 2.851744756415533e-05,
+      "loss": 0.8923,
+      "step": 7298
+    },
+    {
+      "epoch": 0.6521768266803672,
+      "grad_norm": 0.4857379198074341,
+      "learning_rate": 2.850438147891571e-05,
+      "loss": 0.9274,
+      "step": 7299
+    },
+    {
+      "epoch": 0.6522661782116291,
+      "grad_norm": 0.42418479919433594,
+      "learning_rate": 2.8491317194163265e-05,
+      "loss": 1.0274,
+      "step": 7300
+    },
+    {
+      "epoch": 0.652355529742891,
+      "grad_norm": 0.49535322189331055,
+      "learning_rate": 2.847825471099227e-05,
+      "loss": 0.9725,
+      "step": 7301
+    },
+    {
+      "epoch": 0.6524448812741528,
+      "grad_norm": 0.46349287033081055,
+      "learning_rate": 2.8465194030496872e-05,
+      "loss": 0.9706,
+      "step": 7302
+    },
+    {
+      "epoch": 0.6525342328054147,
+      "grad_norm": 0.5424239635467529,
+      "learning_rate": 2.8452135153771e-05,
+      "loss": 0.9002,
+      "step": 7303
+    },
+    {
+      "epoch": 0.6526235843366766,
+      "grad_norm": 0.5158531069755554,
+      "learning_rate": 2.843907808190849e-05,
+      "loss": 0.8912,
+      "step": 7304
+    },
+    {
+      "epoch": 0.6527129358679384,
+      "grad_norm": 0.44023188948631287,
+      "learning_rate": 2.8426022816003012e-05,
+      "loss": 1.0164,
+      "step": 7305
+    },
+    {
+      "epoch": 0.6528022873992003,
+      "grad_norm": 0.6754381060600281,
+      "learning_rate": 2.841296935714809e-05,
+      "loss": 0.9589,
+      "step": 7306
+    },
+    {
+      "epoch": 0.6528916389304622,
+      "grad_norm": 0.65952068567276,
+      "learning_rate": 2.8399917706437074e-05,
+      "loss": 0.8979,
+      "step": 7307
+    },
+    {
+      "epoch": 0.6529809904617241,
+      "grad_norm": 0.431598037481308,
+      "learning_rate": 2.8386867864963202e-05,
+      "loss": 0.8975,
+      "step": 7308
+    },
+    {
+      "epoch": 0.6530703419929859,
+      "grad_norm": 0.45453155040740967,
+      "learning_rate": 2.8373819833819526e-05,
+      "loss": 0.9543,
+      "step": 7309
+    },
+    {
+      "epoch": 0.6531596935242477,
+      "grad_norm": 0.4382008910179138,
+      "learning_rate": 2.836077361409897e-05,
+      "loss": 1.0249,
+      "step": 7310
+    },
+    {
+      "epoch": 0.6532490450555096,
+      "grad_norm": 0.4890137314796448,
+      "learning_rate": 2.8347729206894268e-05,
+      "loss": 0.9056,
+      "step": 7311
+    },
+    {
+      "epoch": 0.6533383965867715,
+      "grad_norm": 0.4039430022239685,
+      "learning_rate": 2.8334686613298034e-05,
+      "loss": 0.995,
+      "step": 7312
+    },
+    {
+      "epoch": 0.6534277481180334,
+      "grad_norm": 0.5166674256324768,
+      "learning_rate": 2.8321645834402737e-05,
+      "loss": 0.9533,
+      "step": 7313
+    },
+    {
+      "epoch": 0.6535170996492953,
+      "grad_norm": 0.5235624313354492,
+      "learning_rate": 2.8308606871300697e-05,
+      "loss": 0.9465,
+      "step": 7314
+    },
+    {
+      "epoch": 0.6536064511805572,
+      "grad_norm": 0.46239298582077026,
+      "learning_rate": 2.8295569725084027e-05,
+      "loss": 0.9373,
+      "step": 7315
+    },
+    {
+      "epoch": 0.6536958027118189,
+      "grad_norm": 0.46263396739959717,
+      "learning_rate": 2.828253439684474e-05,
+      "loss": 0.9634,
+      "step": 7316
+    },
+    {
+      "epoch": 0.6537851542430808,
+      "grad_norm": 0.6211736798286438,
+      "learning_rate": 2.8269500887674687e-05,
+      "loss": 0.9254,
+      "step": 7317
+    },
+    {
+      "epoch": 0.6538745057743427,
+      "grad_norm": 0.5086511969566345,
+      "learning_rate": 2.825646919866557e-05,
+      "loss": 0.8853,
+      "step": 7318
+    },
+    {
+      "epoch": 0.6539638573056046,
+      "grad_norm": 0.4071234464645386,
+      "learning_rate": 2.8243439330908926e-05,
+      "loss": 0.9643,
+      "step": 7319
+    },
+    {
+      "epoch": 0.6540532088368665,
+      "grad_norm": 0.4825928807258606,
+      "learning_rate": 2.8230411285496145e-05,
+      "loss": 0.9184,
+      "step": 7320
+    },
+    {
+      "epoch": 0.6541425603681283,
+      "grad_norm": 0.5398180484771729,
+      "learning_rate": 2.8217385063518463e-05,
+      "loss": 0.921,
+      "step": 7321
+    },
+    {
+      "epoch": 0.6542319118993902,
+      "grad_norm": 0.5476343631744385,
+      "learning_rate": 2.8204360666067e-05,
+      "loss": 0.9528,
+      "step": 7322
+    },
+    {
+      "epoch": 0.654321263430652,
+      "grad_norm": 0.4591327905654907,
+      "learning_rate": 2.819133809423262e-05,
+      "loss": 0.9868,
+      "step": 7323
+    },
+    {
+      "epoch": 0.6544106149619139,
+      "grad_norm": 0.5206472277641296,
+      "learning_rate": 2.8178317349106155e-05,
+      "loss": 0.9039,
+      "step": 7324
+    },
+    {
+      "epoch": 0.6544999664931758,
+      "grad_norm": 0.4743161201477051,
+      "learning_rate": 2.8165298431778197e-05,
+      "loss": 0.9459,
+      "step": 7325
+    },
+    {
+      "epoch": 0.6545893180244376,
+      "grad_norm": 0.48798489570617676,
+      "learning_rate": 2.8152281343339248e-05,
+      "loss": 0.9241,
+      "step": 7326
+    },
+    {
+      "epoch": 0.6546786695556995,
+      "grad_norm": 0.4876413643360138,
+      "learning_rate": 2.8139266084879614e-05,
+      "loss": 0.9567,
+      "step": 7327
+    },
+    {
+      "epoch": 0.6547680210869614,
+      "grad_norm": 0.5455504059791565,
+      "learning_rate": 2.812625265748946e-05,
+      "loss": 1.0707,
+      "step": 7328
+    },
+    {
+      "epoch": 0.6548573726182233,
+      "grad_norm": 0.4286678731441498,
+      "learning_rate": 2.811324106225881e-05,
+      "loss": 0.9682,
+      "step": 7329
+    },
+    {
+      "epoch": 0.6549467241494851,
+      "grad_norm": 0.4500292241573334,
+      "learning_rate": 2.8100231300277514e-05,
+      "loss": 0.8929,
+      "step": 7330
+    },
+    {
+      "epoch": 0.655036075680747,
+      "grad_norm": 0.4849509596824646,
+      "learning_rate": 2.8087223372635286e-05,
+      "loss": 1.0102,
+      "step": 7331
+    },
+    {
+      "epoch": 0.6551254272120088,
+      "grad_norm": 0.4313583970069885,
+      "learning_rate": 2.8074217280421688e-05,
+      "loss": 0.9825,
+      "step": 7332
+    },
+    {
+      "epoch": 0.6552147787432707,
+      "grad_norm": 0.485607773065567,
+      "learning_rate": 2.8061213024726085e-05,
+      "loss": 0.8919,
+      "step": 7333
+    },
+    {
+      "epoch": 0.6553041302745326,
+      "grad_norm": 0.5385865569114685,
+      "learning_rate": 2.8048210606637744e-05,
+      "loss": 0.9059,
+      "step": 7334
+    },
+    {
+      "epoch": 0.6553934818057945,
+      "grad_norm": 0.4551725387573242,
+      "learning_rate": 2.803521002724575e-05,
+      "loss": 0.9603,
+      "step": 7335
+    },
+    {
+      "epoch": 0.6554828333370563,
+      "grad_norm": 0.5102852582931519,
+      "learning_rate": 2.8022211287639044e-05,
+      "loss": 0.8227,
+      "step": 7336
+    },
+    {
+      "epoch": 0.6555721848683181,
+      "grad_norm": 0.5619865655899048,
+      "learning_rate": 2.8009214388906414e-05,
+      "loss": 0.9732,
+      "step": 7337
+    },
+    {
+      "epoch": 0.65566153639958,
+      "grad_norm": 0.4476434886455536,
+      "learning_rate": 2.7996219332136486e-05,
+      "loss": 1.0222,
+      "step": 7338
+    },
+    {
+      "epoch": 0.6557508879308419,
+      "grad_norm": 0.4480132758617401,
+      "learning_rate": 2.7983226118417728e-05,
+      "loss": 0.8978,
+      "step": 7339
+    },
+    {
+      "epoch": 0.6558402394621038,
+      "grad_norm": 0.5554249286651611,
+      "learning_rate": 2.7970234748838466e-05,
+      "loss": 0.8385,
+      "step": 7340
+    },
+    {
+      "epoch": 0.6559295909933657,
+      "grad_norm": 0.4294489622116089,
+      "learning_rate": 2.7957245224486862e-05,
+      "loss": 0.9374,
+      "step": 7341
+    },
+    {
+      "epoch": 0.6560189425246276,
+      "grad_norm": 0.5135436058044434,
+      "learning_rate": 2.7944257546450948e-05,
+      "loss": 0.9616,
+      "step": 7342
+    },
+    {
+      "epoch": 0.6561082940558893,
+      "grad_norm": 0.47493478655815125,
+      "learning_rate": 2.793127171581854e-05,
+      "loss": 0.867,
+      "step": 7343
+    },
+    {
+      "epoch": 0.6561976455871512,
+      "grad_norm": 0.4486094117164612,
+      "learning_rate": 2.7918287733677372e-05,
+      "loss": 0.9386,
+      "step": 7344
+    },
+    {
+      "epoch": 0.6562869971184131,
+      "grad_norm": 0.42966026067733765,
+      "learning_rate": 2.7905305601114972e-05,
+      "loss": 1.0198,
+      "step": 7345
+    },
+    {
+      "epoch": 0.656376348649675,
+      "grad_norm": 0.4275778830051422,
+      "learning_rate": 2.7892325319218744e-05,
+      "loss": 1.0089,
+      "step": 7346
+    },
+    {
+      "epoch": 0.6564657001809369,
+      "grad_norm": 0.4956458508968353,
+      "learning_rate": 2.787934688907594e-05,
+      "loss": 0.9616,
+      "step": 7347
+    },
+    {
+      "epoch": 0.6565550517121987,
+      "grad_norm": 0.5669902563095093,
+      "learning_rate": 2.7866370311773603e-05,
+      "loss": 0.8564,
+      "step": 7348
+    },
+    {
+      "epoch": 0.6566444032434606,
+      "grad_norm": 0.5185860991477966,
+      "learning_rate": 2.7853395588398677e-05,
+      "loss": 0.9535,
+      "step": 7349
+    },
+    {
+      "epoch": 0.6567337547747224,
+      "grad_norm": 0.4602813124656677,
+      "learning_rate": 2.784042272003794e-05,
+      "loss": 0.9282,
+      "step": 7350
+    },
+    {
+      "epoch": 0.6568231063059843,
+      "grad_norm": 0.42367443442344666,
+      "learning_rate": 2.7827451707778007e-05,
+      "loss": 0.9314,
+      "step": 7351
+    },
+    {
+      "epoch": 0.6569124578372462,
+      "grad_norm": 0.40423354506492615,
+      "learning_rate": 2.7814482552705346e-05,
+      "loss": 1.036,
+      "step": 7352
+    },
+    {
+      "epoch": 0.657001809368508,
+      "grad_norm": 0.4704086184501648,
+      "learning_rate": 2.780151525590624e-05,
+      "loss": 0.9161,
+      "step": 7353
+    },
+    {
+      "epoch": 0.6570911608997699,
+      "grad_norm": 0.4368595778942108,
+      "learning_rate": 2.7788549818466847e-05,
+      "loss": 1.052,
+      "step": 7354
+    },
+    {
+      "epoch": 0.6571805124310318,
+      "grad_norm": 0.5148953199386597,
+      "learning_rate": 2.7775586241473173e-05,
+      "loss": 0.9886,
+      "step": 7355
+    },
+    {
+      "epoch": 0.6572698639622937,
+      "grad_norm": 0.43235522508621216,
+      "learning_rate": 2.7762624526011038e-05,
+      "loss": 0.8993,
+      "step": 7356
+    },
+    {
+      "epoch": 0.6573592154935555,
+      "grad_norm": 0.5775102376937866,
+      "learning_rate": 2.774966467316613e-05,
+      "loss": 0.9632,
+      "step": 7357
+    },
+    {
+      "epoch": 0.6574485670248174,
+      "grad_norm": 0.4877246022224426,
+      "learning_rate": 2.7736706684023982e-05,
+      "loss": 0.8856,
+      "step": 7358
+    },
+    {
+      "epoch": 0.6575379185560792,
+      "grad_norm": 0.4451915919780731,
+      "learning_rate": 2.772375055966996e-05,
+      "loss": 0.9679,
+      "step": 7359
+    },
+    {
+      "epoch": 0.6576272700873411,
+      "grad_norm": 0.4021294414997101,
+      "learning_rate": 2.7710796301189268e-05,
+      "loss": 0.9819,
+      "step": 7360
+    },
+    {
+      "epoch": 0.657716621618603,
+      "grad_norm": 0.4669017791748047,
+      "learning_rate": 2.7697843909666977e-05,
+      "loss": 0.9477,
+      "step": 7361
+    },
+    {
+      "epoch": 0.6578059731498649,
+      "grad_norm": 0.48270532488822937,
+      "learning_rate": 2.7684893386188003e-05,
+      "loss": 0.993,
+      "step": 7362
+    },
+    {
+      "epoch": 0.6578953246811268,
+      "grad_norm": 0.4589235186576843,
+      "learning_rate": 2.767194473183705e-05,
+      "loss": 0.9158,
+      "step": 7363
+    },
+    {
+      "epoch": 0.6579846762123885,
+      "grad_norm": 0.4602046012878418,
+      "learning_rate": 2.765899794769873e-05,
+      "loss": 0.9123,
+      "step": 7364
+    },
+    {
+      "epoch": 0.6580740277436504,
+      "grad_norm": 0.4203275144100189,
+      "learning_rate": 2.7646053034857457e-05,
+      "loss": 0.9179,
+      "step": 7365
+    },
+    {
+      "epoch": 0.6581633792749123,
+      "grad_norm": 0.464804083108902,
+      "learning_rate": 2.7633109994397533e-05,
+      "loss": 0.9164,
+      "step": 7366
+    },
+    {
+      "epoch": 0.6582527308061742,
+      "grad_norm": 0.4711710810661316,
+      "learning_rate": 2.762016882740305e-05,
+      "loss": 0.9645,
+      "step": 7367
+    },
+    {
+      "epoch": 0.6583420823374361,
+      "grad_norm": 0.5707644820213318,
+      "learning_rate": 2.7607229534957984e-05,
+      "loss": 0.9153,
+      "step": 7368
+    },
+    {
+      "epoch": 0.658431433868698,
+      "grad_norm": 0.39793646335601807,
+      "learning_rate": 2.7594292118146137e-05,
+      "loss": 0.9331,
+      "step": 7369
+    },
+    {
+      "epoch": 0.6585207853999598,
+      "grad_norm": 0.4051116108894348,
+      "learning_rate": 2.7581356578051143e-05,
+      "loss": 0.9709,
+      "step": 7370
+    },
+    {
+      "epoch": 0.6586101369312216,
+      "grad_norm": 0.5276316404342651,
+      "learning_rate": 2.756842291575651e-05,
+      "loss": 0.9835,
+      "step": 7371
+    },
+    {
+      "epoch": 0.6586994884624835,
+      "grad_norm": 0.4137079417705536,
+      "learning_rate": 2.7555491132345557e-05,
+      "loss": 0.9525,
+      "step": 7372
+    },
+    {
+      "epoch": 0.6587888399937454,
+      "grad_norm": 0.5179648995399475,
+      "learning_rate": 2.7542561228901485e-05,
+      "loss": 1.0175,
+      "step": 7373
+    },
+    {
+      "epoch": 0.6588781915250073,
+      "grad_norm": 0.5228958129882812,
+      "learning_rate": 2.752963320650727e-05,
+      "loss": 0.8682,
+      "step": 7374
+    },
+    {
+      "epoch": 0.6589675430562691,
+      "grad_norm": 0.49394968152046204,
+      "learning_rate": 2.7516707066245796e-05,
+      "loss": 0.9222,
+      "step": 7375
+    },
+    {
+      "epoch": 0.659056894587531,
+      "grad_norm": 0.5419045686721802,
+      "learning_rate": 2.7503782809199753e-05,
+      "loss": 0.8827,
+      "step": 7376
+    },
+    {
+      "epoch": 0.6591462461187929,
+      "grad_norm": 0.4782697260379791,
+      "learning_rate": 2.7490860436451692e-05,
+      "loss": 0.9742,
+      "step": 7377
+    },
+    {
+      "epoch": 0.6592355976500547,
+      "grad_norm": 0.521144688129425,
+      "learning_rate": 2.7477939949084e-05,
+      "loss": 0.8683,
+      "step": 7378
+    },
+    {
+      "epoch": 0.6593249491813166,
+      "grad_norm": 0.4898735284805298,
+      "learning_rate": 2.7465021348178903e-05,
+      "loss": 0.9738,
+      "step": 7379
+    },
+    {
+      "epoch": 0.6594143007125784,
+      "grad_norm": 0.5147672891616821,
+      "learning_rate": 2.7452104634818497e-05,
+      "loss": 0.8741,
+      "step": 7380
+    },
+    {
+      "epoch": 0.6595036522438403,
+      "grad_norm": 0.5325078368186951,
+      "learning_rate": 2.7439189810084655e-05,
+      "loss": 0.9312,
+      "step": 7381
+    },
+    {
+      "epoch": 0.6595930037751022,
+      "grad_norm": 0.45045092701911926,
+      "learning_rate": 2.7426276875059143e-05,
+      "loss": 0.9429,
+      "step": 7382
+    },
+    {
+      "epoch": 0.6596823553063641,
+      "grad_norm": 0.45888474583625793,
+      "learning_rate": 2.7413365830823557e-05,
+      "loss": 0.9197,
+      "step": 7383
+    },
+    {
+      "epoch": 0.659771706837626,
+      "grad_norm": 0.4505671262741089,
+      "learning_rate": 2.7400456678459363e-05,
+      "loss": 0.8995,
+      "step": 7384
+    },
+    {
+      "epoch": 0.6598610583688878,
+      "grad_norm": 0.6128329634666443,
+      "learning_rate": 2.7387549419047798e-05,
+      "loss": 0.895,
+      "step": 7385
+    },
+    {
+      "epoch": 0.6599504099001496,
+      "grad_norm": 0.44439369440078735,
+      "learning_rate": 2.7374644053669997e-05,
+      "loss": 0.9927,
+      "step": 7386
+    },
+    {
+      "epoch": 0.6600397614314115,
+      "grad_norm": 0.520553469657898,
+      "learning_rate": 2.7361740583406924e-05,
+      "loss": 0.93,
+      "step": 7387
+    },
+    {
+      "epoch": 0.6601291129626734,
+      "grad_norm": 0.5277306437492371,
+      "learning_rate": 2.734883900933939e-05,
+      "loss": 0.9059,
+      "step": 7388
+    },
+    {
+      "epoch": 0.6602184644939353,
+      "grad_norm": 0.5034501552581787,
+      "learning_rate": 2.7335939332548032e-05,
+      "loss": 0.9507,
+      "step": 7389
+    },
+    {
+      "epoch": 0.6603078160251972,
+      "grad_norm": 0.5344181656837463,
+      "learning_rate": 2.7323041554113333e-05,
+      "loss": 0.834,
+      "step": 7390
+    },
+    {
+      "epoch": 0.660397167556459,
+      "grad_norm": 0.5650686621665955,
+      "learning_rate": 2.731014567511562e-05,
+      "loss": 0.9379,
+      "step": 7391
+    },
+    {
+      "epoch": 0.6604865190877208,
+      "grad_norm": 0.5536824464797974,
+      "learning_rate": 2.7297251696635074e-05,
+      "loss": 0.9655,
+      "step": 7392
+    },
+    {
+      "epoch": 0.6605758706189827,
+      "grad_norm": 0.6149418950080872,
+      "learning_rate": 2.7284359619751704e-05,
+      "loss": 0.9708,
+      "step": 7393
+    },
+    {
+      "epoch": 0.6606652221502446,
+      "grad_norm": 0.5840936303138733,
+      "learning_rate": 2.7271469445545327e-05,
+      "loss": 0.9482,
+      "step": 7394
+    },
+    {
+      "epoch": 0.6607545736815065,
+      "grad_norm": 0.42912179231643677,
+      "learning_rate": 2.7258581175095654e-05,
+      "loss": 0.9896,
+      "step": 7395
+    },
+    {
+      "epoch": 0.6608439252127684,
+      "grad_norm": 0.44300276041030884,
+      "learning_rate": 2.7245694809482214e-05,
+      "loss": 0.9567,
+      "step": 7396
+    },
+    {
+      "epoch": 0.6609332767440302,
+      "grad_norm": 0.482937753200531,
+      "learning_rate": 2.7232810349784375e-05,
+      "loss": 0.945,
+      "step": 7397
+    },
+    {
+      "epoch": 0.661022628275292,
+      "grad_norm": 0.40860891342163086,
+      "learning_rate": 2.721992779708136e-05,
+      "loss": 0.914,
+      "step": 7398
+    },
+    {
+      "epoch": 0.6611119798065539,
+      "grad_norm": 0.4908876121044159,
+      "learning_rate": 2.7207047152452196e-05,
+      "loss": 0.9459,
+      "step": 7399
+    },
+    {
+      "epoch": 0.6612013313378158,
+      "grad_norm": 0.475569486618042,
+      "learning_rate": 2.7194168416975797e-05,
+      "loss": 0.9524,
+      "step": 7400
+    },
+    {
+      "epoch": 0.6612906828690777,
+      "grad_norm": 0.44693323969841003,
+      "learning_rate": 2.7181291591730883e-05,
+      "loss": 0.886,
+      "step": 7401
+    },
+    {
+      "epoch": 0.6613800344003395,
+      "grad_norm": 0.613734781742096,
+      "learning_rate": 2.7168416677796028e-05,
+      "loss": 0.9082,
+      "step": 7402
+    },
+    {
+      "epoch": 0.6614693859316014,
+      "grad_norm": 0.43988481163978577,
+      "learning_rate": 2.715554367624966e-05,
+      "loss": 0.9724,
+      "step": 7403
+    },
+    {
+      "epoch": 0.6615587374628633,
+      "grad_norm": 0.37471234798431396,
+      "learning_rate": 2.7142672588170002e-05,
+      "loss": 0.9339,
+      "step": 7404
+    },
+    {
+      "epoch": 0.6616480889941251,
+      "grad_norm": 0.45363959670066833,
+      "learning_rate": 2.712980341463515e-05,
+      "loss": 0.9341,
+      "step": 7405
+    },
+    {
+      "epoch": 0.661737440525387,
+      "grad_norm": 0.5071055293083191,
+      "learning_rate": 2.711693615672305e-05,
+      "loss": 0.9066,
+      "step": 7406
+    },
+    {
+      "epoch": 0.6618267920566488,
+      "grad_norm": 0.4539679288864136,
+      "learning_rate": 2.710407081551145e-05,
+      "loss": 0.975,
+      "step": 7407
+    },
+    {
+      "epoch": 0.6619161435879107,
+      "grad_norm": 0.5037643909454346,
+      "learning_rate": 2.7091207392077977e-05,
+      "loss": 0.918,
+      "step": 7408
+    },
+    {
+      "epoch": 0.6620054951191726,
+      "grad_norm": 0.47487950325012207,
+      "learning_rate": 2.707834588750008e-05,
+      "loss": 0.8837,
+      "step": 7409
+    },
+    {
+      "epoch": 0.6620948466504345,
+      "grad_norm": 0.40642839670181274,
+      "learning_rate": 2.7065486302855037e-05,
+      "loss": 0.9693,
+      "step": 7410
+    },
+    {
+      "epoch": 0.6621841981816964,
+      "grad_norm": 0.4429019093513489,
+      "learning_rate": 2.705262863921998e-05,
+      "loss": 1.0066,
+      "step": 7411
+    },
+    {
+      "epoch": 0.6622735497129582,
+      "grad_norm": 0.5285657644271851,
+      "learning_rate": 2.703977289767188e-05,
+      "loss": 0.925,
+      "step": 7412
+    },
+    {
+      "epoch": 0.66236290124422,
+      "grad_norm": 0.533004105091095,
+      "learning_rate": 2.7026919079287555e-05,
+      "loss": 0.951,
+      "step": 7413
+    },
+    {
+      "epoch": 0.6624522527754819,
+      "grad_norm": 0.5454312562942505,
+      "learning_rate": 2.701406718514361e-05,
+      "loss": 0.96,
+      "step": 7414
+    },
+    {
+      "epoch": 0.6625416043067438,
+      "grad_norm": 0.4193102717399597,
+      "learning_rate": 2.7001217216316553e-05,
+      "loss": 0.9904,
+      "step": 7415
+    },
+    {
+      "epoch": 0.6626309558380057,
+      "grad_norm": 0.43523266911506653,
+      "learning_rate": 2.698836917388271e-05,
+      "loss": 0.9572,
+      "step": 7416
+    },
+    {
+      "epoch": 0.6627203073692676,
+      "grad_norm": 0.4869917035102844,
+      "learning_rate": 2.6975523058918252e-05,
+      "loss": 1.0183,
+      "step": 7417
+    },
+    {
+      "epoch": 0.6628096589005295,
+      "grad_norm": 0.4953290820121765,
+      "learning_rate": 2.6962678872499137e-05,
+      "loss": 0.8993,
+      "step": 7418
+    },
+    {
+      "epoch": 0.6628990104317912,
+      "grad_norm": 0.5153881907463074,
+      "learning_rate": 2.6949836615701225e-05,
+      "loss": 0.9066,
+      "step": 7419
+    },
+    {
+      "epoch": 0.6629883619630531,
+      "grad_norm": 0.42058584094047546,
+      "learning_rate": 2.6936996289600198e-05,
+      "loss": 0.989,
+      "step": 7420
+    },
+    {
+      "epoch": 0.663077713494315,
+      "grad_norm": 0.4563521146774292,
+      "learning_rate": 2.6924157895271563e-05,
+      "loss": 0.9595,
+      "step": 7421
+    },
+    {
+      "epoch": 0.6631670650255769,
+      "grad_norm": 0.5149219036102295,
+      "learning_rate": 2.6911321433790677e-05,
+      "loss": 0.9232,
+      "step": 7422
+    },
+    {
+      "epoch": 0.6632564165568388,
+      "grad_norm": 0.478681743144989,
+      "learning_rate": 2.6898486906232746e-05,
+      "loss": 0.9574,
+      "step": 7423
+    },
+    {
+      "epoch": 0.6633457680881006,
+      "grad_norm": 0.4329462945461273,
+      "learning_rate": 2.6885654313672763e-05,
+      "loss": 0.9712,
+      "step": 7424
+    },
+    {
+      "epoch": 0.6634351196193625,
+      "grad_norm": 0.5026130676269531,
+      "learning_rate": 2.6872823657185614e-05,
+      "loss": 0.9287,
+      "step": 7425
+    },
+    {
+      "epoch": 0.6635244711506243,
+      "grad_norm": 0.49470841884613037,
+      "learning_rate": 2.6859994937846e-05,
+      "loss": 0.876,
+      "step": 7426
+    },
+    {
+      "epoch": 0.6636138226818862,
+      "grad_norm": 0.49729084968566895,
+      "learning_rate": 2.6847168156728463e-05,
+      "loss": 0.8896,
+      "step": 7427
+    },
+    {
+      "epoch": 0.6637031742131481,
+      "grad_norm": 0.4407115876674652,
+      "learning_rate": 2.6834343314907384e-05,
+      "loss": 0.9779,
+      "step": 7428
+    },
+    {
+      "epoch": 0.66379252574441,
+      "grad_norm": 0.3972383737564087,
+      "learning_rate": 2.682152041345699e-05,
+      "loss": 0.9496,
+      "step": 7429
+    },
+    {
+      "epoch": 0.6638818772756718,
+      "grad_norm": 0.42927953600883484,
+      "learning_rate": 2.6808699453451313e-05,
+      "loss": 0.9761,
+      "step": 7430
+    },
+    {
+      "epoch": 0.6639712288069337,
+      "grad_norm": 0.37218815088272095,
+      "learning_rate": 2.679588043596427e-05,
+      "loss": 0.9326,
+      "step": 7431
+    },
+    {
+      "epoch": 0.6640605803381956,
+      "grad_norm": 0.513047456741333,
+      "learning_rate": 2.6783063362069573e-05,
+      "loss": 0.9675,
+      "step": 7432
+    },
+    {
+      "epoch": 0.6641499318694574,
+      "grad_norm": 0.4514136016368866,
+      "learning_rate": 2.677024823284081e-05,
+      "loss": 0.9793,
+      "step": 7433
+    },
+    {
+      "epoch": 0.6642392834007192,
+      "grad_norm": 0.5317848324775696,
+      "learning_rate": 2.6757435049351353e-05,
+      "loss": 0.8597,
+      "step": 7434
+    },
+    {
+      "epoch": 0.6643286349319811,
+      "grad_norm": 0.6172546148300171,
+      "learning_rate": 2.6744623812674463e-05,
+      "loss": 0.8547,
+      "step": 7435
+    },
+    {
+      "epoch": 0.664417986463243,
+      "grad_norm": 0.45025524497032166,
+      "learning_rate": 2.6731814523883202e-05,
+      "loss": 0.9209,
+      "step": 7436
+    },
+    {
+      "epoch": 0.6645073379945049,
+      "grad_norm": 0.4900471866130829,
+      "learning_rate": 2.6719007184050504e-05,
+      "loss": 0.8974,
+      "step": 7437
+    },
+    {
+      "epoch": 0.6645966895257668,
+      "grad_norm": 0.4560643434524536,
+      "learning_rate": 2.6706201794249108e-05,
+      "loss": 0.9878,
+      "step": 7438
+    },
+    {
+      "epoch": 0.6646860410570287,
+      "grad_norm": 0.41704943776130676,
+      "learning_rate": 2.6693398355551613e-05,
+      "loss": 1.011,
+      "step": 7439
+    },
+    {
+      "epoch": 0.6647753925882904,
+      "grad_norm": 0.6909659504890442,
+      "learning_rate": 2.668059686903043e-05,
+      "loss": 0.905,
+      "step": 7440
+    },
+    {
+      "epoch": 0.6648647441195523,
+      "grad_norm": 0.4692802429199219,
+      "learning_rate": 2.6667797335757827e-05,
+      "loss": 0.957,
+      "step": 7441
+    },
+    {
+      "epoch": 0.6649540956508142,
+      "grad_norm": 0.47277992963790894,
+      "learning_rate": 2.66549997568059e-05,
+      "loss": 0.9593,
+      "step": 7442
+    },
+    {
+      "epoch": 0.6650434471820761,
+      "grad_norm": 0.4334234595298767,
+      "learning_rate": 2.6642204133246605e-05,
+      "loss": 0.9594,
+      "step": 7443
+    },
+    {
+      "epoch": 0.665132798713338,
+      "grad_norm": 0.43913981318473816,
+      "learning_rate": 2.662941046615167e-05,
+      "loss": 0.9695,
+      "step": 7444
+    },
+    {
+      "epoch": 0.6652221502445999,
+      "grad_norm": 0.43409544229507446,
+      "learning_rate": 2.661661875659272e-05,
+      "loss": 0.9474,
+      "step": 7445
+    },
+    {
+      "epoch": 0.6653115017758617,
+      "grad_norm": 0.5566879510879517,
+      "learning_rate": 2.6603829005641202e-05,
+      "loss": 0.9446,
+      "step": 7446
+    },
+    {
+      "epoch": 0.6654008533071235,
+      "grad_norm": 0.503091037273407,
+      "learning_rate": 2.6591041214368385e-05,
+      "loss": 0.8959,
+      "step": 7447
+    },
+    {
+      "epoch": 0.6654902048383854,
+      "grad_norm": 0.4617300033569336,
+      "learning_rate": 2.6578255383845384e-05,
+      "loss": 1.0043,
+      "step": 7448
+    },
+    {
+      "epoch": 0.6655795563696473,
+      "grad_norm": 0.5025232434272766,
+      "learning_rate": 2.6565471515143157e-05,
+      "loss": 0.9259,
+      "step": 7449
+    },
+    {
+      "epoch": 0.6656689079009092,
+      "grad_norm": 0.5556004047393799,
+      "learning_rate": 2.6552689609332504e-05,
+      "loss": 0.9494,
+      "step": 7450
+    },
+    {
+      "epoch": 0.665758259432171,
+      "grad_norm": 0.5287407040596008,
+      "learning_rate": 2.653990966748401e-05,
+      "loss": 0.9567,
+      "step": 7451
+    },
+    {
+      "epoch": 0.6658476109634329,
+      "grad_norm": 0.5063120126724243,
+      "learning_rate": 2.652713169066815e-05,
+      "loss": 0.8983,
+      "step": 7452
+    },
+    {
+      "epoch": 0.6659369624946948,
+      "grad_norm": 0.4847102463245392,
+      "learning_rate": 2.6514355679955205e-05,
+      "loss": 0.9043,
+      "step": 7453
+    },
+    {
+      "epoch": 0.6660263140259566,
+      "grad_norm": 0.4578264355659485,
+      "learning_rate": 2.650158163641534e-05,
+      "loss": 0.974,
+      "step": 7454
+    },
+    {
+      "epoch": 0.6661156655572185,
+      "grad_norm": 0.481656014919281,
+      "learning_rate": 2.648880956111846e-05,
+      "loss": 0.8898,
+      "step": 7455
+    },
+    {
+      "epoch": 0.6662050170884803,
+      "grad_norm": 0.47429630160331726,
+      "learning_rate": 2.6476039455134393e-05,
+      "loss": 0.972,
+      "step": 7456
+    },
+    {
+      "epoch": 0.6662943686197422,
+      "grad_norm": 0.48434558510780334,
+      "learning_rate": 2.6463271319532766e-05,
+      "loss": 1.0256,
+      "step": 7457
+    },
+    {
+      "epoch": 0.6663837201510041,
+      "grad_norm": 0.5073295831680298,
+      "learning_rate": 2.645050515538306e-05,
+      "loss": 0.8717,
+      "step": 7458
+    },
+    {
+      "epoch": 0.666473071682266,
+      "grad_norm": 0.5684570670127869,
+      "learning_rate": 2.643774096375456e-05,
+      "loss": 0.8968,
+      "step": 7459
+    },
+    {
+      "epoch": 0.6665624232135279,
+      "grad_norm": 0.44450321793556213,
+      "learning_rate": 2.642497874571641e-05,
+      "loss": 0.9466,
+      "step": 7460
+    },
+    {
+      "epoch": 0.6666517747447896,
+      "grad_norm": 0.40622055530548096,
+      "learning_rate": 2.6412218502337582e-05,
+      "loss": 0.9863,
+      "step": 7461
+    },
+    {
+      "epoch": 0.6667411262760515,
+      "grad_norm": 0.42449328303337097,
+      "learning_rate": 2.6399460234686877e-05,
+      "loss": 0.935,
+      "step": 7462
+    },
+    {
+      "epoch": 0.6668304778073134,
+      "grad_norm": 0.5839447975158691,
+      "learning_rate": 2.6386703943832947e-05,
+      "loss": 0.9843,
+      "step": 7463
+    },
+    {
+      "epoch": 0.6669198293385753,
+      "grad_norm": 0.5379756093025208,
+      "learning_rate": 2.6373949630844287e-05,
+      "loss": 0.88,
+      "step": 7464
+    },
+    {
+      "epoch": 0.6670091808698372,
+      "grad_norm": 0.4397304058074951,
+      "learning_rate": 2.6361197296789153e-05,
+      "loss": 0.934,
+      "step": 7465
+    },
+    {
+      "epoch": 0.6670985324010991,
+      "grad_norm": 0.5913897156715393,
+      "learning_rate": 2.6348446942735716e-05,
+      "loss": 0.9605,
+      "step": 7466
+    },
+    {
+      "epoch": 0.6671878839323608,
+      "grad_norm": 0.4340857267379761,
+      "learning_rate": 2.6335698569751956e-05,
+      "loss": 0.9753,
+      "step": 7467
+    },
+    {
+      "epoch": 0.6672772354636227,
+      "grad_norm": 0.41628775000572205,
+      "learning_rate": 2.6322952178905692e-05,
+      "loss": 0.9943,
+      "step": 7468
+    },
+    {
+      "epoch": 0.6673665869948846,
+      "grad_norm": 0.5939970016479492,
+      "learning_rate": 2.631020777126455e-05,
+      "loss": 0.8947,
+      "step": 7469
+    },
+    {
+      "epoch": 0.6674559385261465,
+      "grad_norm": 0.413511723279953,
+      "learning_rate": 2.6297465347896026e-05,
+      "loss": 0.9991,
+      "step": 7470
+    },
+    {
+      "epoch": 0.6675452900574084,
+      "grad_norm": 0.48279207944869995,
+      "learning_rate": 2.6284724909867432e-05,
+      "loss": 0.9453,
+      "step": 7471
+    },
+    {
+      "epoch": 0.6676346415886703,
+      "grad_norm": 0.4602436125278473,
+      "learning_rate": 2.6271986458245912e-05,
+      "loss": 0.9822,
+      "step": 7472
+    },
+    {
+      "epoch": 0.6677239931199321,
+      "grad_norm": 0.4432784616947174,
+      "learning_rate": 2.6259249994098455e-05,
+      "loss": 0.9506,
+      "step": 7473
+    },
+    {
+      "epoch": 0.6678133446511939,
+      "grad_norm": 0.5207743644714355,
+      "learning_rate": 2.624651551849188e-05,
+      "loss": 0.8715,
+      "step": 7474
+    },
+    {
+      "epoch": 0.6679026961824558,
+      "grad_norm": 0.49934712052345276,
+      "learning_rate": 2.623378303249281e-05,
+      "loss": 0.9471,
+      "step": 7475
+    },
+    {
+      "epoch": 0.6679920477137177,
+      "grad_norm": 0.5130594968795776,
+      "learning_rate": 2.622105253716774e-05,
+      "loss": 0.9037,
+      "step": 7476
+    },
+    {
+      "epoch": 0.6680813992449796,
+      "grad_norm": 0.538914680480957,
+      "learning_rate": 2.6208324033582986e-05,
+      "loss": 0.9393,
+      "step": 7477
+    },
+    {
+      "epoch": 0.6681707507762414,
+      "grad_norm": 0.5530039668083191,
+      "learning_rate": 2.6195597522804692e-05,
+      "loss": 0.8773,
+      "step": 7478
+    },
+    {
+      "epoch": 0.6682601023075033,
+      "grad_norm": 0.4622054696083069,
+      "learning_rate": 2.6182873005898845e-05,
+      "loss": 0.9016,
+      "step": 7479
+    },
+    {
+      "epoch": 0.6683494538387652,
+      "grad_norm": 0.43117639422416687,
+      "learning_rate": 2.6170150483931257e-05,
+      "loss": 0.9483,
+      "step": 7480
+    },
+    {
+      "epoch": 0.668438805370027,
+      "grad_norm": 0.45902514457702637,
+      "learning_rate": 2.6157429957967566e-05,
+      "loss": 0.9158,
+      "step": 7481
+    },
+    {
+      "epoch": 0.6685281569012889,
+      "grad_norm": 0.6419569849967957,
+      "learning_rate": 2.6144711429073265e-05,
+      "loss": 0.913,
+      "step": 7482
+    },
+    {
+      "epoch": 0.6686175084325507,
+      "grad_norm": 0.38565734028816223,
+      "learning_rate": 2.6131994898313684e-05,
+      "loss": 0.9589,
+      "step": 7483
+    },
+    {
+      "epoch": 0.6687068599638126,
+      "grad_norm": 0.43412715196609497,
+      "learning_rate": 2.6119280366753917e-05,
+      "loss": 0.9686,
+      "step": 7484
+    },
+    {
+      "epoch": 0.6687962114950745,
+      "grad_norm": 0.4611121714115143,
+      "learning_rate": 2.610656783545898e-05,
+      "loss": 1.0003,
+      "step": 7485
+    },
+    {
+      "epoch": 0.6688855630263364,
+      "grad_norm": 0.4799754321575165,
+      "learning_rate": 2.6093857305493664e-05,
+      "loss": 0.969,
+      "step": 7486
+    },
+    {
+      "epoch": 0.6689749145575983,
+      "grad_norm": 0.556797444820404,
+      "learning_rate": 2.6081148777922643e-05,
+      "loss": 0.9443,
+      "step": 7487
+    },
+    {
+      "epoch": 0.66906426608886,
+      "grad_norm": 0.4632977247238159,
+      "learning_rate": 2.606844225381035e-05,
+      "loss": 0.9519,
+      "step": 7488
+    },
+    {
+      "epoch": 0.6691536176201219,
+      "grad_norm": 0.43375352025032043,
+      "learning_rate": 2.6055737734221108e-05,
+      "loss": 0.9369,
+      "step": 7489
+    },
+    {
+      "epoch": 0.6692429691513838,
+      "grad_norm": 0.48179739713668823,
+      "learning_rate": 2.604303522021906e-05,
+      "loss": 0.8513,
+      "step": 7490
+    },
+    {
+      "epoch": 0.6693323206826457,
+      "grad_norm": 0.45359307527542114,
+      "learning_rate": 2.6030334712868177e-05,
+      "loss": 1.0067,
+      "step": 7491
+    },
+    {
+      "epoch": 0.6694216722139076,
+      "grad_norm": 0.4388836622238159,
+      "learning_rate": 2.6017636213232255e-05,
+      "loss": 0.9204,
+      "step": 7492
+    },
+    {
+      "epoch": 0.6695110237451695,
+      "grad_norm": 0.44517624378204346,
+      "learning_rate": 2.600493972237493e-05,
+      "loss": 0.955,
+      "step": 7493
+    },
+    {
+      "epoch": 0.6696003752764313,
+      "grad_norm": 0.5145146250724792,
+      "learning_rate": 2.5992245241359702e-05,
+      "loss": 0.8947,
+      "step": 7494
+    },
+    {
+      "epoch": 0.6696897268076931,
+      "grad_norm": 0.6348103880882263,
+      "learning_rate": 2.5979552771249814e-05,
+      "loss": 0.975,
+      "step": 7495
+    },
+    {
+      "epoch": 0.669779078338955,
+      "grad_norm": 0.4274086654186249,
+      "learning_rate": 2.596686231310842e-05,
+      "loss": 0.9465,
+      "step": 7496
+    },
+    {
+      "epoch": 0.6698684298702169,
+      "grad_norm": 0.6270675659179688,
+      "learning_rate": 2.595417386799849e-05,
+      "loss": 0.8768,
+      "step": 7497
+    },
+    {
+      "epoch": 0.6699577814014788,
+      "grad_norm": 0.5795266032218933,
+      "learning_rate": 2.5941487436982803e-05,
+      "loss": 0.8254,
+      "step": 7498
+    },
+    {
+      "epoch": 0.6700471329327407,
+      "grad_norm": 0.4406552016735077,
+      "learning_rate": 2.592880302112399e-05,
+      "loss": 0.9277,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6701364844640025,
+      "grad_norm": 0.4358079135417938,
+      "learning_rate": 2.5916120621484498e-05,
+      "loss": 0.9509,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6702258359952644,
+      "grad_norm": 0.5085236430168152,
+      "learning_rate": 2.590344023912663e-05,
+      "loss": 0.8566,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6703151875265262,
+      "grad_norm": 0.5449510812759399,
+      "learning_rate": 2.5890761875112485e-05,
+      "loss": 0.9698,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6704045390577881,
+      "grad_norm": 0.49688923358917236,
+      "learning_rate": 2.587808553050402e-05,
+      "loss": 0.9613,
+      "step": 7503
+    },
+    {
+      "epoch": 0.67049389058905,
+      "grad_norm": 0.43229207396507263,
+      "learning_rate": 2.586541120636303e-05,
+      "loss": 0.9708,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6705832421203118,
+      "grad_norm": 0.4369189739227295,
+      "learning_rate": 2.5852738903751095e-05,
+      "loss": 0.9745,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6706725936515737,
+      "grad_norm": 0.42944756150245667,
+      "learning_rate": 2.5840068623729668e-05,
+      "loss": 0.9261,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6707619451828356,
+      "grad_norm": 0.4990726113319397,
+      "learning_rate": 2.5827400367360015e-05,
+      "loss": 0.975,
+      "step": 7507
+    },
+    {
+      "epoch": 0.6708512967140975,
+      "grad_norm": 0.4963349401950836,
+      "learning_rate": 2.5814734135703245e-05,
+      "loss": 0.9149,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6709406482453593,
+      "grad_norm": 0.4562934339046478,
+      "learning_rate": 2.5802069929820294e-05,
+      "loss": 1.0564,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6710299997766211,
+      "grad_norm": 0.4936973750591278,
+      "learning_rate": 2.578940775077191e-05,
+      "loss": 0.921,
+      "step": 7510
+    },
+    {
+      "epoch": 0.671119351307883,
+      "grad_norm": 0.4688092768192291,
+      "learning_rate": 2.5776747599618688e-05,
+      "loss": 0.8868,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6712087028391449,
+      "grad_norm": 0.525898277759552,
+      "learning_rate": 2.5764089477421067e-05,
+      "loss": 0.8827,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6712980543704068,
+      "grad_norm": 0.5553557872772217,
+      "learning_rate": 2.5751433385239288e-05,
+      "loss": 0.9034,
+      "step": 7513
+    },
+    {
+      "epoch": 0.6713874059016687,
+      "grad_norm": 0.5192905068397522,
+      "learning_rate": 2.5738779324133445e-05,
+      "loss": 0.9762,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6714767574329306,
+      "grad_norm": 0.45627352595329285,
+      "learning_rate": 2.5726127295163428e-05,
+      "loss": 0.9173,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6715661089641923,
+      "grad_norm": 0.5443771481513977,
+      "learning_rate": 2.5713477299388987e-05,
+      "loss": 0.9714,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6716554604954542,
+      "grad_norm": 0.4559047520160675,
+      "learning_rate": 2.57008293378697e-05,
+      "loss": 0.9483,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6717448120267161,
+      "grad_norm": 0.47910940647125244,
+      "learning_rate": 2.568818341166496e-05,
+      "loss": 0.9451,
+      "step": 7518
+    },
+    {
+      "epoch": 0.671834163557978,
+      "grad_norm": 0.4128173589706421,
+      "learning_rate": 2.5675539521834012e-05,
+      "loss": 0.9631,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6719235150892399,
+      "grad_norm": 0.424180805683136,
+      "learning_rate": 2.5662897669435925e-05,
+      "loss": 0.9432,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6720128666205017,
+      "grad_norm": 0.4650351405143738,
+      "learning_rate": 2.5650257855529558e-05,
+      "loss": 0.9263,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6721022181517636,
+      "grad_norm": 0.5273159742355347,
+      "learning_rate": 2.5637620081173642e-05,
+      "loss": 0.8891,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6721915696830254,
+      "grad_norm": 0.43151742219924927,
+      "learning_rate": 2.5624984347426727e-05,
+      "loss": 0.9896,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6722809212142873,
+      "grad_norm": 0.4380139112472534,
+      "learning_rate": 2.5612350655347195e-05,
+      "loss": 1.0411,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6723702727455492,
+      "grad_norm": 0.49319833517074585,
+      "learning_rate": 2.559971900599326e-05,
+      "loss": 0.8419,
+      "step": 7525
+    },
+    {
+      "epoch": 0.672459624276811,
+      "grad_norm": 0.44189539551734924,
+      "learning_rate": 2.5587089400422938e-05,
+      "loss": 0.9321,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6725489758080729,
+      "grad_norm": 0.4883861243724823,
+      "learning_rate": 2.55744618396941e-05,
+      "loss": 0.9207,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6726383273393348,
+      "grad_norm": 0.46611472964286804,
+      "learning_rate": 2.5561836324864442e-05,
+      "loss": 0.8813,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6727276788705966,
+      "grad_norm": 0.49315086007118225,
+      "learning_rate": 2.554921285699148e-05,
+      "loss": 0.8707,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6728170304018585,
+      "grad_norm": 0.47379156947135925,
+      "learning_rate": 2.5536591437132563e-05,
+      "loss": 0.9026,
+      "step": 7530
+    },
+    {
+      "epoch": 0.6729063819331204,
+      "grad_norm": 0.4609242081642151,
+      "learning_rate": 2.552397206634488e-05,
+      "loss": 0.9346,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6729957334643822,
+      "grad_norm": 0.4267633855342865,
+      "learning_rate": 2.5511354745685433e-05,
+      "loss": 0.9378,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6730850849956441,
+      "grad_norm": 0.4892594516277313,
+      "learning_rate": 2.5498739476211054e-05,
+      "loss": 0.9893,
+      "step": 7533
+    },
+    {
+      "epoch": 0.673174436526906,
+      "grad_norm": 0.6081541180610657,
+      "learning_rate": 2.5486126258978427e-05,
+      "loss": 0.9632,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6732637880581679,
+      "grad_norm": 0.49728408455848694,
+      "learning_rate": 2.547351509504401e-05,
+      "loss": 0.9476,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6733531395894297,
+      "grad_norm": 0.44689908623695374,
+      "learning_rate": 2.5460905985464134e-05,
+      "loss": 1.0134,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6734424911206915,
+      "grad_norm": 0.5165348052978516,
+      "learning_rate": 2.544829893129495e-05,
+      "loss": 0.9183,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6735318426519534,
+      "grad_norm": 0.45430701971054077,
+      "learning_rate": 2.5435693933592432e-05,
+      "loss": 0.9704,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6736211941832153,
+      "grad_norm": 0.4814325273036957,
+      "learning_rate": 2.5423090993412383e-05,
+      "loss": 0.9752,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6737105457144772,
+      "grad_norm": 0.5146889090538025,
+      "learning_rate": 2.5410490111810435e-05,
+      "loss": 0.9822,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6737998972457391,
+      "grad_norm": 0.47414878010749817,
+      "learning_rate": 2.5397891289842052e-05,
+      "loss": 1.0163,
+      "step": 7541
+    },
+    {
+      "epoch": 0.673889248777001,
+      "grad_norm": 0.4566519558429718,
+      "learning_rate": 2.5385294528562507e-05,
+      "loss": 0.9218,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6739786003082627,
+      "grad_norm": 0.7057138681411743,
+      "learning_rate": 2.537269982902692e-05,
+      "loss": 0.8424,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6740679518395246,
+      "grad_norm": 0.5519537329673767,
+      "learning_rate": 2.536010719229023e-05,
+      "loss": 0.8996,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6741573033707865,
+      "grad_norm": 0.46758851408958435,
+      "learning_rate": 2.5347516619407223e-05,
+      "loss": 0.8998,
+      "step": 7545
+    },
+    {
+      "epoch": 0.6742466549020484,
+      "grad_norm": 0.452288955450058,
+      "learning_rate": 2.533492811143246e-05,
+      "loss": 0.8949,
+      "step": 7546
+    },
+    {
+      "epoch": 0.6743360064333103,
+      "grad_norm": 0.46302369236946106,
+      "learning_rate": 2.532234166942038e-05,
+      "loss": 0.9162,
+      "step": 7547
+    },
+    {
+      "epoch": 0.6744253579645721,
+      "grad_norm": 0.46767961978912354,
+      "learning_rate": 2.5309757294425222e-05,
+      "loss": 0.9122,
+      "step": 7548
+    },
+    {
+      "epoch": 0.674514709495834,
+      "grad_norm": 0.42991432547569275,
+      "learning_rate": 2.5297174987501077e-05,
+      "loss": 1.0339,
+      "step": 7549
+    },
+    {
+      "epoch": 0.6746040610270958,
+      "grad_norm": 0.40119388699531555,
+      "learning_rate": 2.528459474970184e-05,
+      "loss": 1.0173,
+      "step": 7550
+    },
+    {
+      "epoch": 0.6746934125583577,
+      "grad_norm": 0.4925364553928375,
+      "learning_rate": 2.5272016582081236e-05,
+      "loss": 0.9063,
+      "step": 7551
+    },
+    {
+      "epoch": 0.6747827640896196,
+      "grad_norm": 0.4693504273891449,
+      "learning_rate": 2.525944048569282e-05,
+      "loss": 0.9513,
+      "step": 7552
+    },
+    {
+      "epoch": 0.6748721156208815,
+      "grad_norm": 0.4908875524997711,
+      "learning_rate": 2.524686646159001e-05,
+      "loss": 0.9932,
+      "step": 7553
+    },
+    {
+      "epoch": 0.6749614671521433,
+      "grad_norm": 0.5330108404159546,
+      "learning_rate": 2.5234294510825957e-05,
+      "loss": 0.925,
+      "step": 7554
+    },
+    {
+      "epoch": 0.6750508186834052,
+      "grad_norm": 0.41894859075546265,
+      "learning_rate": 2.5221724634453724e-05,
+      "loss": 0.9464,
+      "step": 7555
+    },
+    {
+      "epoch": 0.6751401702146671,
+      "grad_norm": 0.4790763556957245,
+      "learning_rate": 2.5209156833526172e-05,
+      "loss": 0.9183,
+      "step": 7556
+    },
+    {
+      "epoch": 0.6752295217459289,
+      "grad_norm": 0.6181892156600952,
+      "learning_rate": 2.5196591109096e-05,
+      "loss": 0.8195,
+      "step": 7557
+    },
+    {
+      "epoch": 0.6753188732771908,
+      "grad_norm": 0.4377688765525818,
+      "learning_rate": 2.5184027462215686e-05,
+      "loss": 0.9403,
+      "step": 7558
+    },
+    {
+      "epoch": 0.6754082248084526,
+      "grad_norm": 0.4455429017543793,
+      "learning_rate": 2.5171465893937602e-05,
+      "loss": 0.9458,
+      "step": 7559
+    },
+    {
+      "epoch": 0.6754975763397145,
+      "grad_norm": 0.451734334230423,
+      "learning_rate": 2.51589064053139e-05,
+      "loss": 0.9318,
+      "step": 7560
+    },
+    {
+      "epoch": 0.6755869278709764,
+      "grad_norm": 0.39565378427505493,
+      "learning_rate": 2.5146348997396567e-05,
+      "loss": 1.0195,
+      "step": 7561
+    },
+    {
+      "epoch": 0.6756762794022383,
+      "grad_norm": 0.5276391506195068,
+      "learning_rate": 2.5133793671237433e-05,
+      "loss": 0.9303,
+      "step": 7562
+    },
+    {
+      "epoch": 0.6757656309335002,
+      "grad_norm": 0.5773701667785645,
+      "learning_rate": 2.512124042788813e-05,
+      "loss": 0.9137,
+      "step": 7563
+    },
+    {
+      "epoch": 0.6758549824647619,
+      "grad_norm": 0.45947515964508057,
+      "learning_rate": 2.5108689268400132e-05,
+      "loss": 0.933,
+      "step": 7564
+    },
+    {
+      "epoch": 0.6759443339960238,
+      "grad_norm": 0.4164651930332184,
+      "learning_rate": 2.5096140193824748e-05,
+      "loss": 0.9255,
+      "step": 7565
+    },
+    {
+      "epoch": 0.6760336855272857,
+      "grad_norm": 0.5537623763084412,
+      "learning_rate": 2.5083593205213063e-05,
+      "loss": 0.8421,
+      "step": 7566
+    },
+    {
+      "epoch": 0.6761230370585476,
+      "grad_norm": 0.412296861410141,
+      "learning_rate": 2.5071048303616028e-05,
+      "loss": 0.9733,
+      "step": 7567
+    },
+    {
+      "epoch": 0.6762123885898095,
+      "grad_norm": 0.6003457307815552,
+      "learning_rate": 2.5058505490084428e-05,
+      "loss": 0.8458,
+      "step": 7568
+    },
+    {
+      "epoch": 0.6763017401210714,
+      "grad_norm": 0.4704027771949768,
+      "learning_rate": 2.504596476566885e-05,
+      "loss": 0.9236,
+      "step": 7569
+    },
+    {
+      "epoch": 0.6763910916523332,
+      "grad_norm": 0.40603572130203247,
+      "learning_rate": 2.5033426131419714e-05,
+      "loss": 0.9337,
+      "step": 7570
+    },
+    {
+      "epoch": 0.676480443183595,
+      "grad_norm": 0.433475136756897,
+      "learning_rate": 2.5020889588387266e-05,
+      "loss": 0.9907,
+      "step": 7571
+    },
+    {
+      "epoch": 0.6765697947148569,
+      "grad_norm": 0.44331127405166626,
+      "learning_rate": 2.500835513762157e-05,
+      "loss": 0.9091,
+      "step": 7572
+    },
+    {
+      "epoch": 0.6766591462461188,
+      "grad_norm": 0.43762436509132385,
+      "learning_rate": 2.4995822780172522e-05,
+      "loss": 0.9437,
+      "step": 7573
+    },
+    {
+      "epoch": 0.6767484977773807,
+      "grad_norm": 0.4718449115753174,
+      "learning_rate": 2.4983292517089846e-05,
+      "loss": 0.9747,
+      "step": 7574
+    },
+    {
+      "epoch": 0.6768378493086425,
+      "grad_norm": 0.48513391613960266,
+      "learning_rate": 2.4970764349423093e-05,
+      "loss": 0.9207,
+      "step": 7575
+    },
+    {
+      "epoch": 0.6769272008399044,
+      "grad_norm": 0.5163725018501282,
+      "learning_rate": 2.49582382782216e-05,
+      "loss": 0.915,
+      "step": 7576
+    },
+    {
+      "epoch": 0.6770165523711663,
+      "grad_norm": 0.592780351638794,
+      "learning_rate": 2.4945714304534585e-05,
+      "loss": 0.7994,
+      "step": 7577
+    },
+    {
+      "epoch": 0.6771059039024281,
+      "grad_norm": 0.44290387630462646,
+      "learning_rate": 2.4933192429411052e-05,
+      "loss": 0.9619,
+      "step": 7578
+    },
+    {
+      "epoch": 0.67719525543369,
+      "grad_norm": 0.5108716487884521,
+      "learning_rate": 2.4920672653899847e-05,
+      "loss": 0.9351,
+      "step": 7579
+    },
+    {
+      "epoch": 0.6772846069649519,
+      "grad_norm": 0.6224034428596497,
+      "learning_rate": 2.490815497904963e-05,
+      "loss": 0.9648,
+      "step": 7580
+    },
+    {
+      "epoch": 0.6773739584962137,
+      "grad_norm": 0.4487568736076355,
+      "learning_rate": 2.4895639405908894e-05,
+      "loss": 0.9194,
+      "step": 7581
+    },
+    {
+      "epoch": 0.6774633100274756,
+      "grad_norm": 0.5798953771591187,
+      "learning_rate": 2.4883125935525953e-05,
+      "loss": 0.8891,
+      "step": 7582
+    },
+    {
+      "epoch": 0.6775526615587375,
+      "grad_norm": 0.5566365122795105,
+      "learning_rate": 2.487061456894894e-05,
+      "loss": 0.9089,
+      "step": 7583
+    },
+    {
+      "epoch": 0.6776420130899994,
+      "grad_norm": 0.5041350722312927,
+      "learning_rate": 2.485810530722582e-05,
+      "loss": 1.0092,
+      "step": 7584
+    },
+    {
+      "epoch": 0.6777313646212612,
+      "grad_norm": 0.47851383686065674,
+      "learning_rate": 2.484559815140439e-05,
+      "loss": 0.8815,
+      "step": 7585
+    },
+    {
+      "epoch": 0.677820716152523,
+      "grad_norm": 0.5732426047325134,
+      "learning_rate": 2.4833093102532222e-05,
+      "loss": 0.9326,
+      "step": 7586
+    },
+    {
+      "epoch": 0.6779100676837849,
+      "grad_norm": 0.5327209234237671,
+      "learning_rate": 2.482059016165677e-05,
+      "loss": 0.8568,
+      "step": 7587
+    },
+    {
+      "epoch": 0.6779994192150468,
+      "grad_norm": 0.5125129222869873,
+      "learning_rate": 2.4808089329825286e-05,
+      "loss": 0.897,
+      "step": 7588
+    },
+    {
+      "epoch": 0.6780887707463087,
+      "grad_norm": 0.4826566278934479,
+      "learning_rate": 2.479559060808484e-05,
+      "loss": 0.9773,
+      "step": 7589
+    },
+    {
+      "epoch": 0.6781781222775706,
+      "grad_norm": 0.5305161476135254,
+      "learning_rate": 2.4783093997482364e-05,
+      "loss": 0.8638,
+      "step": 7590
+    },
+    {
+      "epoch": 0.6782674738088323,
+      "grad_norm": 0.4348098635673523,
+      "learning_rate": 2.477059949906454e-05,
+      "loss": 0.9555,
+      "step": 7591
+    },
+    {
+      "epoch": 0.6783568253400942,
+      "grad_norm": 0.4758349061012268,
+      "learning_rate": 2.4758107113877934e-05,
+      "loss": 1.0283,
+      "step": 7592
+    },
+    {
+      "epoch": 0.6784461768713561,
+      "grad_norm": 0.5508620142936707,
+      "learning_rate": 2.474561684296891e-05,
+      "loss": 0.8579,
+      "step": 7593
+    },
+    {
+      "epoch": 0.678535528402618,
+      "grad_norm": 0.5436617732048035,
+      "learning_rate": 2.4733128687383678e-05,
+      "loss": 0.9722,
+      "step": 7594
+    },
+    {
+      "epoch": 0.6786248799338799,
+      "grad_norm": 0.42772892117500305,
+      "learning_rate": 2.4720642648168256e-05,
+      "loss": 0.9366,
+      "step": 7595
+    },
+    {
+      "epoch": 0.6787142314651418,
+      "grad_norm": 0.6332234740257263,
+      "learning_rate": 2.4708158726368452e-05,
+      "loss": 0.8871,
+      "step": 7596
+    },
+    {
+      "epoch": 0.6788035829964036,
+      "grad_norm": 0.43669572472572327,
+      "learning_rate": 2.4695676923029952e-05,
+      "loss": 0.9658,
+      "step": 7597
+    },
+    {
+      "epoch": 0.6788929345276654,
+      "grad_norm": 0.492891401052475,
+      "learning_rate": 2.468319723919823e-05,
+      "loss": 0.9739,
+      "step": 7598
+    },
+    {
+      "epoch": 0.6789822860589273,
+      "grad_norm": 0.4240998923778534,
+      "learning_rate": 2.4670719675918597e-05,
+      "loss": 0.9061,
+      "step": 7599
+    },
+    {
+      "epoch": 0.6790716375901892,
+      "grad_norm": 0.4478083848953247,
+      "learning_rate": 2.465824423423618e-05,
+      "loss": 1.009,
+      "step": 7600
+    },
+    {
+      "epoch": 0.6791609891214511,
+      "grad_norm": 0.4915962815284729,
+      "learning_rate": 2.4645770915195937e-05,
+      "loss": 0.9018,
+      "step": 7601
+    },
+    {
+      "epoch": 0.679250340652713,
+      "grad_norm": 0.42449894547462463,
+      "learning_rate": 2.4633299719842633e-05,
+      "loss": 0.985,
+      "step": 7602
+    },
+    {
+      "epoch": 0.6793396921839748,
+      "grad_norm": 0.44549936056137085,
+      "learning_rate": 2.4620830649220873e-05,
+      "loss": 0.9358,
+      "step": 7603
+    },
+    {
+      "epoch": 0.6794290437152367,
+      "grad_norm": 0.47829920053482056,
+      "learning_rate": 2.460836370437506e-05,
+      "loss": 0.9446,
+      "step": 7604
+    },
+    {
+      "epoch": 0.6795183952464985,
+      "grad_norm": 0.5621015429496765,
+      "learning_rate": 2.4595898886349466e-05,
+      "loss": 0.9252,
+      "step": 7605
+    },
+    {
+      "epoch": 0.6796077467777604,
+      "grad_norm": 0.49867716431617737,
+      "learning_rate": 2.458343619618811e-05,
+      "loss": 0.9293,
+      "step": 7606
+    },
+    {
+      "epoch": 0.6796970983090223,
+      "grad_norm": 0.4213707447052002,
+      "learning_rate": 2.4570975634934888e-05,
+      "loss": 0.9489,
+      "step": 7607
+    },
+    {
+      "epoch": 0.6797864498402841,
+      "grad_norm": 0.48737624287605286,
+      "learning_rate": 2.455851720363352e-05,
+      "loss": 0.9583,
+      "step": 7608
+    },
+    {
+      "epoch": 0.679875801371546,
+      "grad_norm": 0.45338955521583557,
+      "learning_rate": 2.4546060903327512e-05,
+      "loss": 0.9993,
+      "step": 7609
+    },
+    {
+      "epoch": 0.6799651529028079,
+      "grad_norm": 0.509468138217926,
+      "learning_rate": 2.453360673506023e-05,
+      "loss": 0.8476,
+      "step": 7610
+    },
+    {
+      "epoch": 0.6800545044340698,
+      "grad_norm": 0.4940629303455353,
+      "learning_rate": 2.4521154699874833e-05,
+      "loss": 0.9198,
+      "step": 7611
+    },
+    {
+      "epoch": 0.6801438559653316,
+      "grad_norm": 0.41363492608070374,
+      "learning_rate": 2.450870479881432e-05,
+      "loss": 0.9574,
+      "step": 7612
+    },
+    {
+      "epoch": 0.6802332074965934,
+      "grad_norm": 0.4342886805534363,
+      "learning_rate": 2.4496257032921494e-05,
+      "loss": 0.9672,
+      "step": 7613
+    },
+    {
+      "epoch": 0.6803225590278553,
+      "grad_norm": 0.5025731921195984,
+      "learning_rate": 2.4483811403238987e-05,
+      "loss": 0.9102,
+      "step": 7614
+    },
+    {
+      "epoch": 0.6804119105591172,
+      "grad_norm": 0.5143190622329712,
+      "learning_rate": 2.4471367910809284e-05,
+      "loss": 0.9767,
+      "step": 7615
+    },
+    {
+      "epoch": 0.6805012620903791,
+      "grad_norm": 0.442281574010849,
+      "learning_rate": 2.4458926556674615e-05,
+      "loss": 0.9579,
+      "step": 7616
+    },
+    {
+      "epoch": 0.680590613621641,
+      "grad_norm": 0.4672459065914154,
+      "learning_rate": 2.4446487341877095e-05,
+      "loss": 0.9006,
+      "step": 7617
+    },
+    {
+      "epoch": 0.6806799651529029,
+      "grad_norm": 0.4161359667778015,
+      "learning_rate": 2.4434050267458636e-05,
+      "loss": 0.9494,
+      "step": 7618
+    },
+    {
+      "epoch": 0.6807693166841646,
+      "grad_norm": 0.4646640717983246,
+      "learning_rate": 2.4421615334460986e-05,
+      "loss": 0.9457,
+      "step": 7619
+    },
+    {
+      "epoch": 0.6808586682154265,
+      "grad_norm": 0.45113250613212585,
+      "learning_rate": 2.4409182543925698e-05,
+      "loss": 0.9247,
+      "step": 7620
+    },
+    {
+      "epoch": 0.6809480197466884,
+      "grad_norm": 0.4302740693092346,
+      "learning_rate": 2.4396751896894144e-05,
+      "loss": 0.9624,
+      "step": 7621
+    },
+    {
+      "epoch": 0.6810373712779503,
+      "grad_norm": 0.42872077226638794,
+      "learning_rate": 2.438432339440753e-05,
+      "loss": 0.9573,
+      "step": 7622
+    },
+    {
+      "epoch": 0.6811267228092122,
+      "grad_norm": 0.3714028596878052,
+      "learning_rate": 2.43718970375069e-05,
+      "loss": 0.9537,
+      "step": 7623
+    },
+    {
+      "epoch": 0.681216074340474,
+      "grad_norm": 0.42748263478279114,
+      "learning_rate": 2.435947282723305e-05,
+      "loss": 0.8772,
+      "step": 7624
+    },
+    {
+      "epoch": 0.6813054258717359,
+      "grad_norm": 0.42798498272895813,
+      "learning_rate": 2.4347050764626656e-05,
+      "loss": 0.9627,
+      "step": 7625
+    },
+    {
+      "epoch": 0.6813947774029977,
+      "grad_norm": 0.5037773251533508,
+      "learning_rate": 2.43346308507282e-05,
+      "loss": 0.9368,
+      "step": 7626
+    },
+    {
+      "epoch": 0.6814841289342596,
+      "grad_norm": 0.49071839451789856,
+      "learning_rate": 2.432221308657799e-05,
+      "loss": 0.9451,
+      "step": 7627
+    },
+    {
+      "epoch": 0.6815734804655215,
+      "grad_norm": 0.519668459892273,
+      "learning_rate": 2.430979747321615e-05,
+      "loss": 0.8967,
+      "step": 7628
+    },
+    {
+      "epoch": 0.6816628319967833,
+      "grad_norm": 0.389871746301651,
+      "learning_rate": 2.4297384011682595e-05,
+      "loss": 0.9651,
+      "step": 7629
+    },
+    {
+      "epoch": 0.6817521835280452,
+      "grad_norm": 0.5641822814941406,
+      "learning_rate": 2.42849727030171e-05,
+      "loss": 0.8365,
+      "step": 7630
+    },
+    {
+      "epoch": 0.6818415350593071,
+      "grad_norm": 0.4539962410926819,
+      "learning_rate": 2.427256354825924e-05,
+      "loss": 0.9822,
+      "step": 7631
+    },
+    {
+      "epoch": 0.681930886590569,
+      "grad_norm": 0.4748673439025879,
+      "learning_rate": 2.4260156548448427e-05,
+      "loss": 0.9252,
+      "step": 7632
+    },
+    {
+      "epoch": 0.6820202381218308,
+      "grad_norm": 0.4344809353351593,
+      "learning_rate": 2.424775170462386e-05,
+      "loss": 0.9967,
+      "step": 7633
+    },
+    {
+      "epoch": 0.6821095896530927,
+      "grad_norm": 0.4680819511413574,
+      "learning_rate": 2.4235349017824588e-05,
+      "loss": 0.9189,
+      "step": 7634
+    },
+    {
+      "epoch": 0.6821989411843545,
+      "grad_norm": 0.47244375944137573,
+      "learning_rate": 2.422294848908947e-05,
+      "loss": 0.9564,
+      "step": 7635
+    },
+    {
+      "epoch": 0.6822882927156164,
+      "grad_norm": 0.47211724519729614,
+      "learning_rate": 2.4210550119457197e-05,
+      "loss": 1.0198,
+      "step": 7636
+    },
+    {
+      "epoch": 0.6823776442468783,
+      "grad_norm": 0.4619308114051819,
+      "learning_rate": 2.419815390996623e-05,
+      "loss": 0.9311,
+      "step": 7637
+    },
+    {
+      "epoch": 0.6824669957781402,
+      "grad_norm": 0.5194936394691467,
+      "learning_rate": 2.41857598616549e-05,
+      "loss": 0.9223,
+      "step": 7638
+    },
+    {
+      "epoch": 0.6825563473094021,
+      "grad_norm": 0.4341084063053131,
+      "learning_rate": 2.4173367975561345e-05,
+      "loss": 0.9385,
+      "step": 7639
+    },
+    {
+      "epoch": 0.6826456988406638,
+      "grad_norm": 0.5083869695663452,
+      "learning_rate": 2.416097825272351e-05,
+      "loss": 0.9852,
+      "step": 7640
+    },
+    {
+      "epoch": 0.6827350503719257,
+      "grad_norm": 0.6154478192329407,
+      "learning_rate": 2.4148590694179168e-05,
+      "loss": 0.8429,
+      "step": 7641
+    },
+    {
+      "epoch": 0.6828244019031876,
+      "grad_norm": 0.49730563163757324,
+      "learning_rate": 2.413620530096592e-05,
+      "loss": 0.9575,
+      "step": 7642
+    },
+    {
+      "epoch": 0.6829137534344495,
+      "grad_norm": 0.5614471435546875,
+      "learning_rate": 2.412382207412116e-05,
+      "loss": 0.8966,
+      "step": 7643
+    },
+    {
+      "epoch": 0.6830031049657114,
+      "grad_norm": 0.4349358081817627,
+      "learning_rate": 2.4111441014682123e-05,
+      "loss": 0.9805,
+      "step": 7644
+    },
+    {
+      "epoch": 0.6830924564969733,
+      "grad_norm": 0.40343138575553894,
+      "learning_rate": 2.4099062123685852e-05,
+      "loss": 0.9149,
+      "step": 7645
+    },
+    {
+      "epoch": 0.6831818080282351,
+      "grad_norm": 0.4427891969680786,
+      "learning_rate": 2.4086685402169234e-05,
+      "loss": 0.9948,
+      "step": 7646
+    },
+    {
+      "epoch": 0.6832711595594969,
+      "grad_norm": 0.5709788203239441,
+      "learning_rate": 2.407431085116891e-05,
+      "loss": 0.946,
+      "step": 7647
+    },
+    {
+      "epoch": 0.6833605110907588,
+      "grad_norm": 0.4347441494464874,
+      "learning_rate": 2.4061938471721395e-05,
+      "loss": 1.0256,
+      "step": 7648
+    },
+    {
+      "epoch": 0.6834498626220207,
+      "grad_norm": 0.509070098400116,
+      "learning_rate": 2.4049568264863022e-05,
+      "loss": 0.885,
+      "step": 7649
+    },
+    {
+      "epoch": 0.6835392141532826,
+      "grad_norm": 0.5047274231910706,
+      "learning_rate": 2.403720023162991e-05,
+      "loss": 0.9298,
+      "step": 7650
+    },
+    {
+      "epoch": 0.6836285656845444,
+      "grad_norm": 0.4579404890537262,
+      "learning_rate": 2.4024834373058023e-05,
+      "loss": 0.9002,
+      "step": 7651
+    },
+    {
+      "epoch": 0.6837179172158063,
+      "grad_norm": 0.4215395748615265,
+      "learning_rate": 2.4012470690183136e-05,
+      "loss": 1.0175,
+      "step": 7652
+    },
+    {
+      "epoch": 0.6838072687470681,
+      "grad_norm": 0.4573095738887787,
+      "learning_rate": 2.4000109184040837e-05,
+      "loss": 0.9326,
+      "step": 7653
+    },
+    {
+      "epoch": 0.68389662027833,
+      "grad_norm": 0.4098723530769348,
+      "learning_rate": 2.3987749855666532e-05,
+      "loss": 1.0006,
+      "step": 7654
+    },
+    {
+      "epoch": 0.6839859718095919,
+      "grad_norm": 0.4912080764770508,
+      "learning_rate": 2.3975392706095446e-05,
+      "loss": 0.8999,
+      "step": 7655
+    },
+    {
+      "epoch": 0.6840753233408537,
+      "grad_norm": 0.4591172933578491,
+      "learning_rate": 2.3963037736362643e-05,
+      "loss": 0.9086,
+      "step": 7656
+    },
+    {
+      "epoch": 0.6841646748721156,
+      "grad_norm": 0.4539848864078522,
+      "learning_rate": 2.3950684947502944e-05,
+      "loss": 0.9633,
+      "step": 7657
+    },
+    {
+      "epoch": 0.6842540264033775,
+      "grad_norm": 0.462184876203537,
+      "learning_rate": 2.3938334340551044e-05,
+      "loss": 0.9912,
+      "step": 7658
+    },
+    {
+      "epoch": 0.6843433779346394,
+      "grad_norm": 0.46925488114356995,
+      "learning_rate": 2.3925985916541443e-05,
+      "loss": 0.9211,
+      "step": 7659
+    },
+    {
+      "epoch": 0.6844327294659012,
+      "grad_norm": 0.5577223300933838,
+      "learning_rate": 2.3913639676508472e-05,
+      "loss": 0.8373,
+      "step": 7660
+    },
+    {
+      "epoch": 0.684522080997163,
+      "grad_norm": 0.42695876955986023,
+      "learning_rate": 2.390129562148622e-05,
+      "loss": 0.8683,
+      "step": 7661
+    },
+    {
+      "epoch": 0.6846114325284249,
+      "grad_norm": 0.5081654191017151,
+      "learning_rate": 2.3888953752508647e-05,
+      "loss": 0.8314,
+      "step": 7662
+    },
+    {
+      "epoch": 0.6847007840596868,
+      "grad_norm": 0.6528906226158142,
+      "learning_rate": 2.387661407060952e-05,
+      "loss": 0.9019,
+      "step": 7663
+    },
+    {
+      "epoch": 0.6847901355909487,
+      "grad_norm": 0.3996724784374237,
+      "learning_rate": 2.3864276576822426e-05,
+      "loss": 0.9953,
+      "step": 7664
+    },
+    {
+      "epoch": 0.6848794871222106,
+      "grad_norm": 0.4197225868701935,
+      "learning_rate": 2.385194127218075e-05,
+      "loss": 1.0005,
+      "step": 7665
+    },
+    {
+      "epoch": 0.6849688386534725,
+      "grad_norm": 0.51161789894104,
+      "learning_rate": 2.3839608157717734e-05,
+      "loss": 0.9327,
+      "step": 7666
+    },
+    {
+      "epoch": 0.6850581901847342,
+      "grad_norm": 0.4613009989261627,
+      "learning_rate": 2.3827277234466362e-05,
+      "loss": 1.0265,
+      "step": 7667
+    },
+    {
+      "epoch": 0.6851475417159961,
+      "grad_norm": 0.4957435429096222,
+      "learning_rate": 2.3814948503459507e-05,
+      "loss": 0.9579,
+      "step": 7668
+    },
+    {
+      "epoch": 0.685236893247258,
+      "grad_norm": 0.5043631792068481,
+      "learning_rate": 2.380262196572982e-05,
+      "loss": 0.9527,
+      "step": 7669
+    },
+    {
+      "epoch": 0.6853262447785199,
+      "grad_norm": 0.5093596577644348,
+      "learning_rate": 2.3790297622309794e-05,
+      "loss": 0.9432,
+      "step": 7670
+    },
+    {
+      "epoch": 0.6854155963097818,
+      "grad_norm": 0.433877557516098,
+      "learning_rate": 2.3777975474231718e-05,
+      "loss": 0.8894,
+      "step": 7671
+    },
+    {
+      "epoch": 0.6855049478410437,
+      "grad_norm": 0.558192253112793,
+      "learning_rate": 2.3765655522527695e-05,
+      "loss": 0.9565,
+      "step": 7672
+    },
+    {
+      "epoch": 0.6855942993723055,
+      "grad_norm": 0.4382516145706177,
+      "learning_rate": 2.3753337768229667e-05,
+      "loss": 0.9967,
+      "step": 7673
+    },
+    {
+      "epoch": 0.6856836509035673,
+      "grad_norm": 0.5013807415962219,
+      "learning_rate": 2.374102221236937e-05,
+      "loss": 0.9572,
+      "step": 7674
+    },
+    {
+      "epoch": 0.6857730024348292,
+      "grad_norm": 0.43265098333358765,
+      "learning_rate": 2.372870885597836e-05,
+      "loss": 0.961,
+      "step": 7675
+    },
+    {
+      "epoch": 0.6858623539660911,
+      "grad_norm": 0.4570571780204773,
+      "learning_rate": 2.371639770008804e-05,
+      "loss": 0.9556,
+      "step": 7676
+    },
+    {
+      "epoch": 0.685951705497353,
+      "grad_norm": 0.4653104841709137,
+      "learning_rate": 2.370408874572955e-05,
+      "loss": 0.9336,
+      "step": 7677
+    },
+    {
+      "epoch": 0.6860410570286148,
+      "grad_norm": 0.5548616051673889,
+      "learning_rate": 2.3691781993933926e-05,
+      "loss": 1.0193,
+      "step": 7678
+    },
+    {
+      "epoch": 0.6861304085598767,
+      "grad_norm": 0.41967645287513733,
+      "learning_rate": 2.3679477445731986e-05,
+      "loss": 0.9393,
+      "step": 7679
+    },
+    {
+      "epoch": 0.6862197600911386,
+      "grad_norm": 0.4959544241428375,
+      "learning_rate": 2.366717510215436e-05,
+      "loss": 0.8876,
+      "step": 7680
+    },
+    {
+      "epoch": 0.6863091116224004,
+      "grad_norm": 0.45173823833465576,
+      "learning_rate": 2.3654874964231518e-05,
+      "loss": 0.9034,
+      "step": 7681
+    },
+    {
+      "epoch": 0.6863984631536623,
+      "grad_norm": 0.4388938248157501,
+      "learning_rate": 2.3642577032993705e-05,
+      "loss": 0.8964,
+      "step": 7682
+    },
+    {
+      "epoch": 0.6864878146849241,
+      "grad_norm": 0.42833590507507324,
+      "learning_rate": 2.363028130947102e-05,
+      "loss": 0.9542,
+      "step": 7683
+    },
+    {
+      "epoch": 0.686577166216186,
+      "grad_norm": 0.556098997592926,
+      "learning_rate": 2.361798779469336e-05,
+      "loss": 0.9357,
+      "step": 7684
+    },
+    {
+      "epoch": 0.6866665177474479,
+      "grad_norm": 0.5818011164665222,
+      "learning_rate": 2.3605696489690427e-05,
+      "loss": 0.8631,
+      "step": 7685
+    },
+    {
+      "epoch": 0.6867558692787098,
+      "grad_norm": 0.40382498502731323,
+      "learning_rate": 2.3593407395491778e-05,
+      "loss": 0.9778,
+      "step": 7686
+    },
+    {
+      "epoch": 0.6868452208099717,
+      "grad_norm": 0.44755667448043823,
+      "learning_rate": 2.358112051312672e-05,
+      "loss": 0.9621,
+      "step": 7687
+    },
+    {
+      "epoch": 0.6869345723412335,
+      "grad_norm": 0.5690574049949646,
+      "learning_rate": 2.3568835843624422e-05,
+      "loss": 0.897,
+      "step": 7688
+    },
+    {
+      "epoch": 0.6870239238724953,
+      "grad_norm": 0.5574625134468079,
+      "learning_rate": 2.3556553388013852e-05,
+      "loss": 0.9749,
+      "step": 7689
+    },
+    {
+      "epoch": 0.6871132754037572,
+      "grad_norm": 0.4903115928173065,
+      "learning_rate": 2.3544273147323807e-05,
+      "loss": 0.9872,
+      "step": 7690
+    },
+    {
+      "epoch": 0.6872026269350191,
+      "grad_norm": 0.4215288460254669,
+      "learning_rate": 2.3531995122582883e-05,
+      "loss": 0.9412,
+      "step": 7691
+    },
+    {
+      "epoch": 0.687291978466281,
+      "grad_norm": 0.5319769978523254,
+      "learning_rate": 2.3519719314819493e-05,
+      "loss": 0.9143,
+      "step": 7692
+    },
+    {
+      "epoch": 0.6873813299975429,
+      "grad_norm": 0.5102341175079346,
+      "learning_rate": 2.3507445725061895e-05,
+      "loss": 0.8844,
+      "step": 7693
+    },
+    {
+      "epoch": 0.6874706815288048,
+      "grad_norm": 0.6138080954551697,
+      "learning_rate": 2.3495174354338084e-05,
+      "loss": 0.9267,
+      "step": 7694
+    },
+    {
+      "epoch": 0.6875600330600665,
+      "grad_norm": 0.4891884922981262,
+      "learning_rate": 2.348290520367595e-05,
+      "loss": 0.9271,
+      "step": 7695
+    },
+    {
+      "epoch": 0.6876493845913284,
+      "grad_norm": 0.44753071665763855,
+      "learning_rate": 2.3470638274103147e-05,
+      "loss": 0.9276,
+      "step": 7696
+    },
+    {
+      "epoch": 0.6877387361225903,
+      "grad_norm": 0.44476214051246643,
+      "learning_rate": 2.3458373566647174e-05,
+      "loss": 0.9433,
+      "step": 7697
+    },
+    {
+      "epoch": 0.6878280876538522,
+      "grad_norm": 0.39918237924575806,
+      "learning_rate": 2.344611108233535e-05,
+      "loss": 0.9188,
+      "step": 7698
+    },
+    {
+      "epoch": 0.687917439185114,
+      "grad_norm": 0.4474645256996155,
+      "learning_rate": 2.343385082219475e-05,
+      "loss": 0.95,
+      "step": 7699
+    },
+    {
+      "epoch": 0.6880067907163759,
+      "grad_norm": 0.455310195684433,
+      "learning_rate": 2.342159278725231e-05,
+      "loss": 0.9567,
+      "step": 7700
+    },
+    {
+      "epoch": 0.6880961422476378,
+      "grad_norm": 0.5603063702583313,
+      "learning_rate": 2.3409336978534783e-05,
+      "loss": 0.8424,
+      "step": 7701
+    },
+    {
+      "epoch": 0.6881854937788996,
+      "grad_norm": 0.4089902937412262,
+      "learning_rate": 2.3397083397068724e-05,
+      "loss": 0.9721,
+      "step": 7702
+    },
+    {
+      "epoch": 0.6882748453101615,
+      "grad_norm": 0.563758373260498,
+      "learning_rate": 2.3384832043880495e-05,
+      "loss": 0.9647,
+      "step": 7703
+    },
+    {
+      "epoch": 0.6883641968414234,
+      "grad_norm": 0.4580053389072418,
+      "learning_rate": 2.337258291999628e-05,
+      "loss": 1.037,
+      "step": 7704
+    },
+    {
+      "epoch": 0.6884535483726852,
+      "grad_norm": 0.5256210565567017,
+      "learning_rate": 2.336033602644207e-05,
+      "loss": 0.9317,
+      "step": 7705
+    },
+    {
+      "epoch": 0.6885428999039471,
+      "grad_norm": 0.5540886521339417,
+      "learning_rate": 2.3348091364243703e-05,
+      "loss": 0.9263,
+      "step": 7706
+    },
+    {
+      "epoch": 0.688632251435209,
+      "grad_norm": 0.4322797954082489,
+      "learning_rate": 2.3335848934426746e-05,
+      "loss": 0.9876,
+      "step": 7707
+    },
+    {
+      "epoch": 0.6887216029664709,
+      "grad_norm": 0.49090680480003357,
+      "learning_rate": 2.3323608738016663e-05,
+      "loss": 0.9329,
+      "step": 7708
+    },
+    {
+      "epoch": 0.6888109544977327,
+      "grad_norm": 0.5516027808189392,
+      "learning_rate": 2.3311370776038698e-05,
+      "loss": 1.0222,
+      "step": 7709
+    },
+    {
+      "epoch": 0.6889003060289945,
+      "grad_norm": 0.4046795666217804,
+      "learning_rate": 2.3299135049517913e-05,
+      "loss": 0.9255,
+      "step": 7710
+    },
+    {
+      "epoch": 0.6889896575602564,
+      "grad_norm": 0.5095672607421875,
+      "learning_rate": 2.3286901559479175e-05,
+      "loss": 0.9252,
+      "step": 7711
+    },
+    {
+      "epoch": 0.6890790090915183,
+      "grad_norm": 0.5826500058174133,
+      "learning_rate": 2.3274670306947173e-05,
+      "loss": 0.7873,
+      "step": 7712
+    },
+    {
+      "epoch": 0.6891683606227802,
+      "grad_norm": 0.5110095143318176,
+      "learning_rate": 2.3262441292946407e-05,
+      "loss": 0.9253,
+      "step": 7713
+    },
+    {
+      "epoch": 0.6892577121540421,
+      "grad_norm": 0.4405474364757538,
+      "learning_rate": 2.3250214518501184e-05,
+      "loss": 0.9763,
+      "step": 7714
+    },
+    {
+      "epoch": 0.689347063685304,
+      "grad_norm": 0.42986834049224854,
+      "learning_rate": 2.3237989984635628e-05,
+      "loss": 0.9512,
+      "step": 7715
+    },
+    {
+      "epoch": 0.6894364152165657,
+      "grad_norm": 0.42974916100502014,
+      "learning_rate": 2.3225767692373686e-05,
+      "loss": 0.9491,
+      "step": 7716
+    },
+    {
+      "epoch": 0.6895257667478276,
+      "grad_norm": 0.41702863574028015,
+      "learning_rate": 2.3213547642739082e-05,
+      "loss": 0.9853,
+      "step": 7717
+    },
+    {
+      "epoch": 0.6896151182790895,
+      "grad_norm": 0.45753413438796997,
+      "learning_rate": 2.3201329836755382e-05,
+      "loss": 0.9201,
+      "step": 7718
+    },
+    {
+      "epoch": 0.6897044698103514,
+      "grad_norm": 0.6328713297843933,
+      "learning_rate": 2.3189114275445963e-05,
+      "loss": 0.8626,
+      "step": 7719
+    },
+    {
+      "epoch": 0.6897938213416133,
+      "grad_norm": 0.4760781228542328,
+      "learning_rate": 2.3176900959834004e-05,
+      "loss": 0.9065,
+      "step": 7720
+    },
+    {
+      "epoch": 0.6898831728728752,
+      "grad_norm": 0.5676954388618469,
+      "learning_rate": 2.3164689890942504e-05,
+      "loss": 0.9782,
+      "step": 7721
+    },
+    {
+      "epoch": 0.6899725244041369,
+      "grad_norm": 0.4092563986778259,
+      "learning_rate": 2.315248106979427e-05,
+      "loss": 1.0402,
+      "step": 7722
+    },
+    {
+      "epoch": 0.6900618759353988,
+      "grad_norm": 0.4689728915691376,
+      "learning_rate": 2.3140274497411918e-05,
+      "loss": 0.9736,
+      "step": 7723
+    },
+    {
+      "epoch": 0.6901512274666607,
+      "grad_norm": 0.5073100924491882,
+      "learning_rate": 2.3128070174817884e-05,
+      "loss": 0.9342,
+      "step": 7724
+    },
+    {
+      "epoch": 0.6902405789979226,
+      "grad_norm": 0.48991286754608154,
+      "learning_rate": 2.311586810303441e-05,
+      "loss": 0.9554,
+      "step": 7725
+    },
+    {
+      "epoch": 0.6903299305291845,
+      "grad_norm": 0.45478180050849915,
+      "learning_rate": 2.3103668283083564e-05,
+      "loss": 0.9772,
+      "step": 7726
+    },
+    {
+      "epoch": 0.6904192820604463,
+      "grad_norm": 0.5075895190238953,
+      "learning_rate": 2.3091470715987167e-05,
+      "loss": 0.8529,
+      "step": 7727
+    },
+    {
+      "epoch": 0.6905086335917082,
+      "grad_norm": 0.5047464966773987,
+      "learning_rate": 2.307927540276693e-05,
+      "loss": 0.9655,
+      "step": 7728
+    },
+    {
+      "epoch": 0.69059798512297,
+      "grad_norm": 0.47214803099632263,
+      "learning_rate": 2.306708234444433e-05,
+      "loss": 0.9132,
+      "step": 7729
+    },
+    {
+      "epoch": 0.6906873366542319,
+      "grad_norm": 0.4735872745513916,
+      "learning_rate": 2.305489154204067e-05,
+      "loss": 0.8661,
+      "step": 7730
+    },
+    {
+      "epoch": 0.6907766881854938,
+      "grad_norm": 0.4646959900856018,
+      "learning_rate": 2.304270299657707e-05,
+      "loss": 0.9096,
+      "step": 7731
+    },
+    {
+      "epoch": 0.6908660397167556,
+      "grad_norm": 0.45085829496383667,
+      "learning_rate": 2.3030516709074424e-05,
+      "loss": 0.9554,
+      "step": 7732
+    },
+    {
+      "epoch": 0.6909553912480175,
+      "grad_norm": 0.47358912229537964,
+      "learning_rate": 2.3018332680553477e-05,
+      "loss": 0.9515,
+      "step": 7733
+    },
+    {
+      "epoch": 0.6910447427792794,
+      "grad_norm": 0.4590679407119751,
+      "learning_rate": 2.3006150912034774e-05,
+      "loss": 0.9225,
+      "step": 7734
+    },
+    {
+      "epoch": 0.6911340943105413,
+      "grad_norm": 0.5170496702194214,
+      "learning_rate": 2.2993971404538668e-05,
+      "loss": 0.8962,
+      "step": 7735
+    },
+    {
+      "epoch": 0.6912234458418031,
+      "grad_norm": 0.4720090627670288,
+      "learning_rate": 2.298179415908531e-05,
+      "loss": 0.9482,
+      "step": 7736
+    },
+    {
+      "epoch": 0.691312797373065,
+      "grad_norm": 0.572025716304779,
+      "learning_rate": 2.296961917669471e-05,
+      "loss": 0.8859,
+      "step": 7737
+    },
+    {
+      "epoch": 0.6914021489043268,
+      "grad_norm": 0.4607776403427124,
+      "learning_rate": 2.295744645838661e-05,
+      "loss": 0.9991,
+      "step": 7738
+    },
+    {
+      "epoch": 0.6914915004355887,
+      "grad_norm": 0.5466684699058533,
+      "learning_rate": 2.2945276005180623e-05,
+      "loss": 0.896,
+      "step": 7739
+    },
+    {
+      "epoch": 0.6915808519668506,
+      "grad_norm": 0.48179617524147034,
+      "learning_rate": 2.293310781809615e-05,
+      "loss": 0.9066,
+      "step": 7740
+    },
+    {
+      "epoch": 0.6916702034981125,
+      "grad_norm": 0.4328363239765167,
+      "learning_rate": 2.292094189815241e-05,
+      "loss": 0.9941,
+      "step": 7741
+    },
+    {
+      "epoch": 0.6917595550293744,
+      "grad_norm": 0.4751478135585785,
+      "learning_rate": 2.290877824636843e-05,
+      "loss": 0.9366,
+      "step": 7742
+    },
+    {
+      "epoch": 0.6918489065606361,
+      "grad_norm": 0.47883662581443787,
+      "learning_rate": 2.2896616863763038e-05,
+      "loss": 1.0028,
+      "step": 7743
+    },
+    {
+      "epoch": 0.691938258091898,
+      "grad_norm": 0.5021872520446777,
+      "learning_rate": 2.2884457751354887e-05,
+      "loss": 0.9387,
+      "step": 7744
+    },
+    {
+      "epoch": 0.6920276096231599,
+      "grad_norm": 0.45752233266830444,
+      "learning_rate": 2.2872300910162436e-05,
+      "loss": 0.9381,
+      "step": 7745
+    },
+    {
+      "epoch": 0.6921169611544218,
+      "grad_norm": 0.465744286775589,
+      "learning_rate": 2.2860146341203937e-05,
+      "loss": 0.921,
+      "step": 7746
+    },
+    {
+      "epoch": 0.6922063126856837,
+      "grad_norm": 0.47951745986938477,
+      "learning_rate": 2.2847994045497496e-05,
+      "loss": 0.9759,
+      "step": 7747
+    },
+    {
+      "epoch": 0.6922956642169455,
+      "grad_norm": 0.4690631628036499,
+      "learning_rate": 2.2835844024060953e-05,
+      "loss": 0.9177,
+      "step": 7748
+    },
+    {
+      "epoch": 0.6923850157482074,
+      "grad_norm": 0.5409393310546875,
+      "learning_rate": 2.282369627791202e-05,
+      "loss": 0.905,
+      "step": 7749
+    },
+    {
+      "epoch": 0.6924743672794692,
+      "grad_norm": 0.5218047499656677,
+      "learning_rate": 2.2811550808068205e-05,
+      "loss": 0.9922,
+      "step": 7750
+    },
+    {
+      "epoch": 0.6925637188107311,
+      "grad_norm": 0.5113648176193237,
+      "learning_rate": 2.2799407615546815e-05,
+      "loss": 0.9426,
+      "step": 7751
+    },
+    {
+      "epoch": 0.692653070341993,
+      "grad_norm": 0.4514075517654419,
+      "learning_rate": 2.278726670136498e-05,
+      "loss": 0.9237,
+      "step": 7752
+    },
+    {
+      "epoch": 0.6927424218732549,
+      "grad_norm": 0.4475104510784149,
+      "learning_rate": 2.277512806653962e-05,
+      "loss": 0.9978,
+      "step": 7753
+    },
+    {
+      "epoch": 0.6928317734045167,
+      "grad_norm": 0.4956379532814026,
+      "learning_rate": 2.2762991712087484e-05,
+      "loss": 0.8725,
+      "step": 7754
+    },
+    {
+      "epoch": 0.6929211249357786,
+      "grad_norm": 0.52278733253479,
+      "learning_rate": 2.2750857639025113e-05,
+      "loss": 0.8942,
+      "step": 7755
+    },
+    {
+      "epoch": 0.6930104764670405,
+      "grad_norm": 0.5214927792549133,
+      "learning_rate": 2.2738725848368875e-05,
+      "loss": 0.9485,
+      "step": 7756
+    },
+    {
+      "epoch": 0.6930998279983023,
+      "grad_norm": 0.44215962290763855,
+      "learning_rate": 2.272659634113495e-05,
+      "loss": 0.9248,
+      "step": 7757
+    },
+    {
+      "epoch": 0.6931891795295642,
+      "grad_norm": 0.4372926652431488,
+      "learning_rate": 2.271446911833927e-05,
+      "loss": 0.9661,
+      "step": 7758
+    },
+    {
+      "epoch": 0.693278531060826,
+      "grad_norm": 0.4903610050678253,
+      "learning_rate": 2.2702344180997647e-05,
+      "loss": 0.945,
+      "step": 7759
+    },
+    {
+      "epoch": 0.6933678825920879,
+      "grad_norm": 0.5382173657417297,
+      "learning_rate": 2.2690221530125676e-05,
+      "loss": 0.9443,
+      "step": 7760
+    },
+    {
+      "epoch": 0.6934572341233498,
+      "grad_norm": 0.5113039016723633,
+      "learning_rate": 2.2678101166738746e-05,
+      "loss": 0.9368,
+      "step": 7761
+    },
+    {
+      "epoch": 0.6935465856546117,
+      "grad_norm": 0.6735799312591553,
+      "learning_rate": 2.2665983091852083e-05,
+      "loss": 0.7989,
+      "step": 7762
+    },
+    {
+      "epoch": 0.6936359371858736,
+      "grad_norm": 0.4851955771446228,
+      "learning_rate": 2.2653867306480708e-05,
+      "loss": 0.9514,
+      "step": 7763
+    },
+    {
+      "epoch": 0.6937252887171353,
+      "grad_norm": 0.4901287853717804,
+      "learning_rate": 2.2641753811639417e-05,
+      "loss": 0.9322,
+      "step": 7764
+    },
+    {
+      "epoch": 0.6938146402483972,
+      "grad_norm": 0.5687100291252136,
+      "learning_rate": 2.262964260834286e-05,
+      "loss": 0.8931,
+      "step": 7765
+    },
+    {
+      "epoch": 0.6939039917796591,
+      "grad_norm": 0.45068779587745667,
+      "learning_rate": 2.2617533697605485e-05,
+      "loss": 0.9799,
+      "step": 7766
+    },
+    {
+      "epoch": 0.693993343310921,
+      "grad_norm": 0.4677877724170685,
+      "learning_rate": 2.260542708044154e-05,
+      "loss": 0.9831,
+      "step": 7767
+    },
+    {
+      "epoch": 0.6940826948421829,
+      "grad_norm": 0.4960726201534271,
+      "learning_rate": 2.2593322757865097e-05,
+      "loss": 0.8693,
+      "step": 7768
+    },
+    {
+      "epoch": 0.6941720463734448,
+      "grad_norm": 0.5460396409034729,
+      "learning_rate": 2.258122073088999e-05,
+      "loss": 0.8695,
+      "step": 7769
+    },
+    {
+      "epoch": 0.6942613979047066,
+      "grad_norm": 0.44287580251693726,
+      "learning_rate": 2.2569121000529915e-05,
+      "loss": 0.9818,
+      "step": 7770
+    },
+    {
+      "epoch": 0.6943507494359684,
+      "grad_norm": 0.5541728734970093,
+      "learning_rate": 2.2557023567798342e-05,
+      "loss": 0.974,
+      "step": 7771
+    },
+    {
+      "epoch": 0.6944401009672303,
+      "grad_norm": 0.5429545044898987,
+      "learning_rate": 2.254492843370857e-05,
+      "loss": 0.9813,
+      "step": 7772
+    },
+    {
+      "epoch": 0.6945294524984922,
+      "grad_norm": 0.42696717381477356,
+      "learning_rate": 2.2532835599273687e-05,
+      "loss": 0.8715,
+      "step": 7773
+    },
+    {
+      "epoch": 0.6946188040297541,
+      "grad_norm": 0.49874842166900635,
+      "learning_rate": 2.2520745065506603e-05,
+      "loss": 0.8765,
+      "step": 7774
+    },
+    {
+      "epoch": 0.694708155561016,
+      "grad_norm": 0.49501675367355347,
+      "learning_rate": 2.2508656833420026e-05,
+      "loss": 0.9732,
+      "step": 7775
+    },
+    {
+      "epoch": 0.6947975070922778,
+      "grad_norm": 0.48265352845191956,
+      "learning_rate": 2.2496570904026483e-05,
+      "loss": 0.8776,
+      "step": 7776
+    },
+    {
+      "epoch": 0.6948868586235397,
+      "grad_norm": 0.46912387013435364,
+      "learning_rate": 2.2484487278338305e-05,
+      "loss": 0.9019,
+      "step": 7777
+    },
+    {
+      "epoch": 0.6949762101548015,
+      "grad_norm": 0.4739210605621338,
+      "learning_rate": 2.2472405957367593e-05,
+      "loss": 0.899,
+      "step": 7778
+    },
+    {
+      "epoch": 0.6950655616860634,
+      "grad_norm": 0.4149891436100006,
+      "learning_rate": 2.2460326942126307e-05,
+      "loss": 0.9424,
+      "step": 7779
+    },
+    {
+      "epoch": 0.6951549132173253,
+      "grad_norm": 0.43324029445648193,
+      "learning_rate": 2.2448250233626195e-05,
+      "loss": 0.927,
+      "step": 7780
+    },
+    {
+      "epoch": 0.6952442647485871,
+      "grad_norm": 0.7626486420631409,
+      "learning_rate": 2.2436175832878802e-05,
+      "loss": 0.9243,
+      "step": 7781
+    },
+    {
+      "epoch": 0.695333616279849,
+      "grad_norm": 0.4425656795501709,
+      "learning_rate": 2.242410374089549e-05,
+      "loss": 0.9539,
+      "step": 7782
+    },
+    {
+      "epoch": 0.6954229678111109,
+      "grad_norm": 0.4871158003807068,
+      "learning_rate": 2.2412033958687433e-05,
+      "loss": 0.8842,
+      "step": 7783
+    },
+    {
+      "epoch": 0.6955123193423727,
+      "grad_norm": 0.5576094388961792,
+      "learning_rate": 2.2399966487265596e-05,
+      "loss": 0.9452,
+      "step": 7784
+    },
+    {
+      "epoch": 0.6956016708736346,
+      "grad_norm": 0.41526561975479126,
+      "learning_rate": 2.238790132764076e-05,
+      "loss": 0.988,
+      "step": 7785
+    },
+    {
+      "epoch": 0.6956910224048964,
+      "grad_norm": 0.4997550845146179,
+      "learning_rate": 2.237583848082351e-05,
+      "loss": 0.942,
+      "step": 7786
+    },
+    {
+      "epoch": 0.6957803739361583,
+      "grad_norm": 0.44697871804237366,
+      "learning_rate": 2.2363777947824265e-05,
+      "loss": 0.9652,
+      "step": 7787
+    },
+    {
+      "epoch": 0.6958697254674202,
+      "grad_norm": 0.4410659372806549,
+      "learning_rate": 2.2351719729653175e-05,
+      "loss": 0.9358,
+      "step": 7788
+    },
+    {
+      "epoch": 0.6959590769986821,
+      "grad_norm": 0.39497098326683044,
+      "learning_rate": 2.233966382732027e-05,
+      "loss": 0.9459,
+      "step": 7789
+    },
+    {
+      "epoch": 0.696048428529944,
+      "grad_norm": 0.45145925879478455,
+      "learning_rate": 2.232761024183535e-05,
+      "loss": 0.9192,
+      "step": 7790
+    },
+    {
+      "epoch": 0.6961377800612057,
+      "grad_norm": 0.46766191720962524,
+      "learning_rate": 2.2315558974208045e-05,
+      "loss": 0.9029,
+      "step": 7791
+    },
+    {
+      "epoch": 0.6962271315924676,
+      "grad_norm": 0.4594910740852356,
+      "learning_rate": 2.2303510025447765e-05,
+      "loss": 0.8901,
+      "step": 7792
+    },
+    {
+      "epoch": 0.6963164831237295,
+      "grad_norm": 0.48282474279403687,
+      "learning_rate": 2.229146339656375e-05,
+      "loss": 0.9668,
+      "step": 7793
+    },
+    {
+      "epoch": 0.6964058346549914,
+      "grad_norm": 0.5255979895591736,
+      "learning_rate": 2.227941908856503e-05,
+      "loss": 0.932,
+      "step": 7794
+    },
+    {
+      "epoch": 0.6964951861862533,
+      "grad_norm": 0.4911576509475708,
+      "learning_rate": 2.2267377102460436e-05,
+      "loss": 0.9966,
+      "step": 7795
+    },
+    {
+      "epoch": 0.6965845377175152,
+      "grad_norm": 0.5194364190101624,
+      "learning_rate": 2.2255337439258633e-05,
+      "loss": 0.9707,
+      "step": 7796
+    },
+    {
+      "epoch": 0.696673889248777,
+      "grad_norm": 0.5468015074729919,
+      "learning_rate": 2.2243300099968046e-05,
+      "loss": 0.8358,
+      "step": 7797
+    },
+    {
+      "epoch": 0.6967632407800388,
+      "grad_norm": 0.47449544072151184,
+      "learning_rate": 2.2231265085596938e-05,
+      "loss": 0.9037,
+      "step": 7798
+    },
+    {
+      "epoch": 0.6968525923113007,
+      "grad_norm": 0.48344165086746216,
+      "learning_rate": 2.2219232397153366e-05,
+      "loss": 0.8762,
+      "step": 7799
+    },
+    {
+      "epoch": 0.6969419438425626,
+      "grad_norm": 0.433371365070343,
+      "learning_rate": 2.220720203564521e-05,
+      "loss": 0.9288,
+      "step": 7800
+    },
+    {
+      "epoch": 0.6970312953738245,
+      "grad_norm": 0.49428948760032654,
+      "learning_rate": 2.219517400208015e-05,
+      "loss": 0.9456,
+      "step": 7801
+    },
+    {
+      "epoch": 0.6971206469050863,
+      "grad_norm": 0.4290115535259247,
+      "learning_rate": 2.2183148297465627e-05,
+      "loss": 0.9207,
+      "step": 7802
+    },
+    {
+      "epoch": 0.6972099984363482,
+      "grad_norm": 0.4194071590900421,
+      "learning_rate": 2.217112492280894e-05,
+      "loss": 0.9429,
+      "step": 7803
+    },
+    {
+      "epoch": 0.6972993499676101,
+      "grad_norm": 0.4213373363018036,
+      "learning_rate": 2.2159103879117177e-05,
+      "loss": 0.9443,
+      "step": 7804
+    },
+    {
+      "epoch": 0.6973887014988719,
+      "grad_norm": 0.4857761263847351,
+      "learning_rate": 2.2147085167397223e-05,
+      "loss": 0.9155,
+      "step": 7805
+    },
+    {
+      "epoch": 0.6974780530301338,
+      "grad_norm": 0.4734998941421509,
+      "learning_rate": 2.2135068788655782e-05,
+      "loss": 0.9608,
+      "step": 7806
+    },
+    {
+      "epoch": 0.6975674045613957,
+      "grad_norm": 0.47898635268211365,
+      "learning_rate": 2.2123054743899346e-05,
+      "loss": 0.9165,
+      "step": 7807
+    },
+    {
+      "epoch": 0.6976567560926575,
+      "grad_norm": 0.556919515132904,
+      "learning_rate": 2.211104303413424e-05,
+      "loss": 0.9658,
+      "step": 7808
+    },
+    {
+      "epoch": 0.6977461076239194,
+      "grad_norm": 0.3948970139026642,
+      "learning_rate": 2.2099033660366537e-05,
+      "loss": 1.032,
+      "step": 7809
+    },
+    {
+      "epoch": 0.6978354591551813,
+      "grad_norm": 0.4545915126800537,
+      "learning_rate": 2.2087026623602164e-05,
+      "loss": 0.9332,
+      "step": 7810
+    },
+    {
+      "epoch": 0.6979248106864432,
+      "grad_norm": 0.4447174668312073,
+      "learning_rate": 2.207502192484685e-05,
+      "loss": 0.9648,
+      "step": 7811
+    },
+    {
+      "epoch": 0.698014162217705,
+      "grad_norm": 0.5227632522583008,
+      "learning_rate": 2.2063019565106102e-05,
+      "loss": 0.9064,
+      "step": 7812
+    },
+    {
+      "epoch": 0.6981035137489668,
+      "grad_norm": 0.41880759596824646,
+      "learning_rate": 2.2051019545385255e-05,
+      "loss": 0.9345,
+      "step": 7813
+    },
+    {
+      "epoch": 0.6981928652802287,
+      "grad_norm": 0.5738639831542969,
+      "learning_rate": 2.2039021866689435e-05,
+      "loss": 0.8841,
+      "step": 7814
+    },
+    {
+      "epoch": 0.6982822168114906,
+      "grad_norm": 0.38003218173980713,
+      "learning_rate": 2.202702653002358e-05,
+      "loss": 0.9731,
+      "step": 7815
+    },
+    {
+      "epoch": 0.6983715683427525,
+      "grad_norm": 0.46742263436317444,
+      "learning_rate": 2.2015033536392422e-05,
+      "loss": 0.9019,
+      "step": 7816
+    },
+    {
+      "epoch": 0.6984609198740144,
+      "grad_norm": 0.57480788230896,
+      "learning_rate": 2.20030428868005e-05,
+      "loss": 0.9423,
+      "step": 7817
+    },
+    {
+      "epoch": 0.6985502714052763,
+      "grad_norm": 0.4974348545074463,
+      "learning_rate": 2.199105458225218e-05,
+      "loss": 0.9786,
+      "step": 7818
+    },
+    {
+      "epoch": 0.698639622936538,
+      "grad_norm": 0.5195631980895996,
+      "learning_rate": 2.197906862375158e-05,
+      "loss": 0.9327,
+      "step": 7819
+    },
+    {
+      "epoch": 0.6987289744677999,
+      "grad_norm": 0.4291267991065979,
+      "learning_rate": 2.1967085012302663e-05,
+      "loss": 0.9707,
+      "step": 7820
+    },
+    {
+      "epoch": 0.6988183259990618,
+      "grad_norm": 0.5776417255401611,
+      "learning_rate": 2.1955103748909185e-05,
+      "loss": 0.836,
+      "step": 7821
+    },
+    {
+      "epoch": 0.6989076775303237,
+      "grad_norm": 0.45474281907081604,
+      "learning_rate": 2.19431248345747e-05,
+      "loss": 0.9669,
+      "step": 7822
+    },
+    {
+      "epoch": 0.6989970290615856,
+      "grad_norm": 0.41181471943855286,
+      "learning_rate": 2.193114827030258e-05,
+      "loss": 0.9725,
+      "step": 7823
+    },
+    {
+      "epoch": 0.6990863805928474,
+      "grad_norm": 0.5399705767631531,
+      "learning_rate": 2.191917405709598e-05,
+      "loss": 0.8856,
+      "step": 7824
+    },
+    {
+      "epoch": 0.6991757321241093,
+      "grad_norm": 0.37962204217910767,
+      "learning_rate": 2.1907202195957882e-05,
+      "loss": 0.9573,
+      "step": 7825
+    },
+    {
+      "epoch": 0.6992650836553711,
+      "grad_norm": 0.5898367762565613,
+      "learning_rate": 2.1895232687891043e-05,
+      "loss": 0.9001,
+      "step": 7826
+    },
+    {
+      "epoch": 0.699354435186633,
+      "grad_norm": 0.4448642432689667,
+      "learning_rate": 2.1883265533898038e-05,
+      "loss": 0.9668,
+      "step": 7827
+    },
+    {
+      "epoch": 0.6994437867178949,
+      "grad_norm": 0.42721062898635864,
+      "learning_rate": 2.187130073498127e-05,
+      "loss": 0.96,
+      "step": 7828
+    },
+    {
+      "epoch": 0.6995331382491567,
+      "grad_norm": 0.6047458648681641,
+      "learning_rate": 2.1859338292142876e-05,
+      "loss": 0.9264,
+      "step": 7829
+    },
+    {
+      "epoch": 0.6996224897804186,
+      "grad_norm": 0.4223198890686035,
+      "learning_rate": 2.184737820638486e-05,
+      "loss": 0.9497,
+      "step": 7830
+    },
+    {
+      "epoch": 0.6997118413116805,
+      "grad_norm": 0.5067864060401917,
+      "learning_rate": 2.1835420478709e-05,
+      "loss": 0.8954,
+      "step": 7831
+    },
+    {
+      "epoch": 0.6998011928429424,
+      "grad_norm": 0.4639846682548523,
+      "learning_rate": 2.182346511011689e-05,
+      "loss": 0.9765,
+      "step": 7832
+    },
+    {
+      "epoch": 0.6998905443742042,
+      "grad_norm": 0.44720324873924255,
+      "learning_rate": 2.1811512101609922e-05,
+      "loss": 0.9591,
+      "step": 7833
+    },
+    {
+      "epoch": 0.699979895905466,
+      "grad_norm": 0.4672408699989319,
+      "learning_rate": 2.17995614541893e-05,
+      "loss": 0.9872,
+      "step": 7834
+    },
+    {
+      "epoch": 0.7000692474367279,
+      "grad_norm": 0.5736651420593262,
+      "learning_rate": 2.1787613168855974e-05,
+      "loss": 0.9987,
+      "step": 7835
+    },
+    {
+      "epoch": 0.7001585989679898,
+      "grad_norm": 0.49598804116249084,
+      "learning_rate": 2.1775667246610775e-05,
+      "loss": 0.905,
+      "step": 7836
+    },
+    {
+      "epoch": 0.7002479504992517,
+      "grad_norm": 0.41947072744369507,
+      "learning_rate": 2.1763723688454298e-05,
+      "loss": 0.9417,
+      "step": 7837
+    },
+    {
+      "epoch": 0.7003373020305136,
+      "grad_norm": 0.4067612588405609,
+      "learning_rate": 2.1751782495386946e-05,
+      "loss": 0.9841,
+      "step": 7838
+    },
+    {
+      "epoch": 0.7004266535617755,
+      "grad_norm": 0.4044066071510315,
+      "learning_rate": 2.1739843668408904e-05,
+      "loss": 0.9668,
+      "step": 7839
+    },
+    {
+      "epoch": 0.7005160050930372,
+      "grad_norm": 0.712028980255127,
+      "learning_rate": 2.1727907208520187e-05,
+      "loss": 0.8248,
+      "step": 7840
+    },
+    {
+      "epoch": 0.7006053566242991,
+      "grad_norm": 0.4395425617694855,
+      "learning_rate": 2.1715973116720594e-05,
+      "loss": 0.9757,
+      "step": 7841
+    },
+    {
+      "epoch": 0.700694708155561,
+      "grad_norm": 0.4635796546936035,
+      "learning_rate": 2.1704041394009745e-05,
+      "loss": 0.9057,
+      "step": 7842
+    },
+    {
+      "epoch": 0.7007840596868229,
+      "grad_norm": 0.46956396102905273,
+      "learning_rate": 2.169211204138704e-05,
+      "loss": 0.9642,
+      "step": 7843
+    },
+    {
+      "epoch": 0.7008734112180848,
+      "grad_norm": 0.4673489034175873,
+      "learning_rate": 2.1680185059851692e-05,
+      "loss": 0.9328,
+      "step": 7844
+    },
+    {
+      "epoch": 0.7009627627493467,
+      "grad_norm": 0.4535006880760193,
+      "learning_rate": 2.166826045040271e-05,
+      "loss": 0.9427,
+      "step": 7845
+    },
+    {
+      "epoch": 0.7010521142806084,
+      "grad_norm": 0.5039295554161072,
+      "learning_rate": 2.165633821403892e-05,
+      "loss": 0.894,
+      "step": 7846
+    },
+    {
+      "epoch": 0.7011414658118703,
+      "grad_norm": 0.6144269704818726,
+      "learning_rate": 2.1644418351758917e-05,
+      "loss": 0.9041,
+      "step": 7847
+    },
+    {
+      "epoch": 0.7012308173431322,
+      "grad_norm": 0.5136603116989136,
+      "learning_rate": 2.1632500864561146e-05,
+      "loss": 0.8892,
+      "step": 7848
+    },
+    {
+      "epoch": 0.7013201688743941,
+      "grad_norm": 0.46805840730667114,
+      "learning_rate": 2.1620585753443786e-05,
+      "loss": 1.002,
+      "step": 7849
+    },
+    {
+      "epoch": 0.701409520405656,
+      "grad_norm": 0.41075465083122253,
+      "learning_rate": 2.160867301940487e-05,
+      "loss": 0.9253,
+      "step": 7850
+    },
+    {
+      "epoch": 0.7014988719369178,
+      "grad_norm": 0.46872520446777344,
+      "learning_rate": 2.1596762663442218e-05,
+      "loss": 0.909,
+      "step": 7851
+    },
+    {
+      "epoch": 0.7015882234681797,
+      "grad_norm": 0.5055715441703796,
+      "learning_rate": 2.1584854686553453e-05,
+      "loss": 0.9786,
+      "step": 7852
+    },
+    {
+      "epoch": 0.7016775749994415,
+      "grad_norm": 0.48132845759391785,
+      "learning_rate": 2.1572949089735987e-05,
+      "loss": 0.9081,
+      "step": 7853
+    },
+    {
+      "epoch": 0.7017669265307034,
+      "grad_norm": 0.5568394660949707,
+      "learning_rate": 2.1561045873987046e-05,
+      "loss": 0.8675,
+      "step": 7854
+    },
+    {
+      "epoch": 0.7018562780619653,
+      "grad_norm": 0.5073636770248413,
+      "learning_rate": 2.1549145040303654e-05,
+      "loss": 0.9566,
+      "step": 7855
+    },
+    {
+      "epoch": 0.7019456295932271,
+      "grad_norm": 0.42798787355422974,
+      "learning_rate": 2.153724658968263e-05,
+      "loss": 0.9682,
+      "step": 7856
+    },
+    {
+      "epoch": 0.702034981124489,
+      "grad_norm": 0.5035516023635864,
+      "learning_rate": 2.152535052312059e-05,
+      "loss": 0.9313,
+      "step": 7857
+    },
+    {
+      "epoch": 0.7021243326557509,
+      "grad_norm": 0.43961191177368164,
+      "learning_rate": 2.1513456841613982e-05,
+      "loss": 0.9328,
+      "step": 7858
+    },
+    {
+      "epoch": 0.7022136841870128,
+      "grad_norm": 0.4986516833305359,
+      "learning_rate": 2.1501565546158993e-05,
+      "loss": 0.9385,
+      "step": 7859
+    },
+    {
+      "epoch": 0.7023030357182746,
+      "grad_norm": 0.5121946930885315,
+      "learning_rate": 2.148967663775166e-05,
+      "loss": 0.8721,
+      "step": 7860
+    },
+    {
+      "epoch": 0.7023923872495365,
+      "grad_norm": 0.6112991571426392,
+      "learning_rate": 2.1477790117387808e-05,
+      "loss": 0.9234,
+      "step": 7861
+    },
+    {
+      "epoch": 0.7024817387807983,
+      "grad_norm": 0.4206686317920685,
+      "learning_rate": 2.1465905986063056e-05,
+      "loss": 0.9169,
+      "step": 7862
+    },
+    {
+      "epoch": 0.7025710903120602,
+      "grad_norm": 0.5358665585517883,
+      "learning_rate": 2.145402424477283e-05,
+      "loss": 0.9301,
+      "step": 7863
+    },
+    {
+      "epoch": 0.7026604418433221,
+      "grad_norm": 0.5316134095191956,
+      "learning_rate": 2.1442144894512352e-05,
+      "loss": 0.9495,
+      "step": 7864
+    },
+    {
+      "epoch": 0.702749793374584,
+      "grad_norm": 0.5568259954452515,
+      "learning_rate": 2.1430267936276637e-05,
+      "loss": 0.9321,
+      "step": 7865
+    },
+    {
+      "epoch": 0.7028391449058459,
+      "grad_norm": 0.39786767959594727,
+      "learning_rate": 2.1418393371060542e-05,
+      "loss": 0.9312,
+      "step": 7866
+    },
+    {
+      "epoch": 0.7029284964371076,
+      "grad_norm": 0.6305972933769226,
+      "learning_rate": 2.1406521199858637e-05,
+      "loss": 0.8965,
+      "step": 7867
+    },
+    {
+      "epoch": 0.7030178479683695,
+      "grad_norm": 0.5135489106178284,
+      "learning_rate": 2.1394651423665368e-05,
+      "loss": 1.0525,
+      "step": 7868
+    },
+    {
+      "epoch": 0.7031071994996314,
+      "grad_norm": 0.5106921792030334,
+      "learning_rate": 2.1382784043474953e-05,
+      "loss": 0.9549,
+      "step": 7869
+    },
+    {
+      "epoch": 0.7031965510308933,
+      "grad_norm": 0.462363064289093,
+      "learning_rate": 2.1370919060281415e-05,
+      "loss": 0.9733,
+      "step": 7870
+    },
+    {
+      "epoch": 0.7032859025621552,
+      "grad_norm": 0.4888690710067749,
+      "learning_rate": 2.135905647507858e-05,
+      "loss": 0.8726,
+      "step": 7871
+    },
+    {
+      "epoch": 0.7033752540934171,
+      "grad_norm": 0.5045296549797058,
+      "learning_rate": 2.1347196288860045e-05,
+      "loss": 0.9865,
+      "step": 7872
+    },
+    {
+      "epoch": 0.7034646056246789,
+      "grad_norm": 0.47808846831321716,
+      "learning_rate": 2.1335338502619233e-05,
+      "loss": 0.9196,
+      "step": 7873
+    },
+    {
+      "epoch": 0.7035539571559407,
+      "grad_norm": 0.42234256863594055,
+      "learning_rate": 2.1323483117349368e-05,
+      "loss": 0.9596,
+      "step": 7874
+    },
+    {
+      "epoch": 0.7036433086872026,
+      "grad_norm": 0.432258278131485,
+      "learning_rate": 2.131163013404346e-05,
+      "loss": 1.0006,
+      "step": 7875
+    },
+    {
+      "epoch": 0.7037326602184645,
+      "grad_norm": 0.45173177123069763,
+      "learning_rate": 2.1299779553694323e-05,
+      "loss": 0.9875,
+      "step": 7876
+    },
+    {
+      "epoch": 0.7038220117497264,
+      "grad_norm": 0.5817151069641113,
+      "learning_rate": 2.128793137729457e-05,
+      "loss": 0.8725,
+      "step": 7877
+    },
+    {
+      "epoch": 0.7039113632809882,
+      "grad_norm": 0.4585109055042267,
+      "learning_rate": 2.1276085605836636e-05,
+      "loss": 0.891,
+      "step": 7878
+    },
+    {
+      "epoch": 0.7040007148122501,
+      "grad_norm": 0.547097384929657,
+      "learning_rate": 2.1264242240312687e-05,
+      "loss": 0.9042,
+      "step": 7879
+    },
+    {
+      "epoch": 0.704090066343512,
+      "grad_norm": 0.6232752799987793,
+      "learning_rate": 2.1252401281714752e-05,
+      "loss": 0.8304,
+      "step": 7880
+    },
+    {
+      "epoch": 0.7041794178747738,
+      "grad_norm": 0.4578670859336853,
+      "learning_rate": 2.1240562731034635e-05,
+      "loss": 0.8873,
+      "step": 7881
+    },
+    {
+      "epoch": 0.7042687694060357,
+      "grad_norm": 0.47056087851524353,
+      "learning_rate": 2.1228726589263942e-05,
+      "loss": 0.9029,
+      "step": 7882
+    },
+    {
+      "epoch": 0.7043581209372975,
+      "grad_norm": 0.48560357093811035,
+      "learning_rate": 2.1216892857394083e-05,
+      "loss": 0.8779,
+      "step": 7883
+    },
+    {
+      "epoch": 0.7044474724685594,
+      "grad_norm": 0.536952555179596,
+      "learning_rate": 2.120506153641625e-05,
+      "loss": 0.8473,
+      "step": 7884
+    },
+    {
+      "epoch": 0.7045368239998213,
+      "grad_norm": 0.41716116666793823,
+      "learning_rate": 2.119323262732144e-05,
+      "loss": 0.9998,
+      "step": 7885
+    },
+    {
+      "epoch": 0.7046261755310832,
+      "grad_norm": 0.5672418475151062,
+      "learning_rate": 2.1181406131100463e-05,
+      "loss": 0.9055,
+      "step": 7886
+    },
+    {
+      "epoch": 0.7047155270623451,
+      "grad_norm": 0.5225016474723816,
+      "learning_rate": 2.1169582048743902e-05,
+      "loss": 0.9609,
+      "step": 7887
+    },
+    {
+      "epoch": 0.7048048785936069,
+      "grad_norm": 0.4605422019958496,
+      "learning_rate": 2.1157760381242174e-05,
+      "loss": 0.9329,
+      "step": 7888
+    },
+    {
+      "epoch": 0.7048942301248687,
+      "grad_norm": 0.4933791756629944,
+      "learning_rate": 2.1145941129585435e-05,
+      "loss": 0.9709,
+      "step": 7889
+    },
+    {
+      "epoch": 0.7049835816561306,
+      "grad_norm": 0.42105743288993835,
+      "learning_rate": 2.1134124294763678e-05,
+      "loss": 0.9333,
+      "step": 7890
+    },
+    {
+      "epoch": 0.7050729331873925,
+      "grad_norm": 0.5470524430274963,
+      "learning_rate": 2.112230987776671e-05,
+      "loss": 1.0609,
+      "step": 7891
+    },
+    {
+      "epoch": 0.7051622847186544,
+      "grad_norm": 0.48633435368537903,
+      "learning_rate": 2.11104978795841e-05,
+      "loss": 0.9119,
+      "step": 7892
+    },
+    {
+      "epoch": 0.7052516362499163,
+      "grad_norm": 0.6098655462265015,
+      "learning_rate": 2.1098688301205237e-05,
+      "loss": 0.8961,
+      "step": 7893
+    },
+    {
+      "epoch": 0.7053409877811782,
+      "grad_norm": 0.5170839428901672,
+      "learning_rate": 2.1086881143619292e-05,
+      "loss": 0.923,
+      "step": 7894
+    },
+    {
+      "epoch": 0.7054303393124399,
+      "grad_norm": 0.4735257923603058,
+      "learning_rate": 2.1075076407815243e-05,
+      "loss": 0.9574,
+      "step": 7895
+    },
+    {
+      "epoch": 0.7055196908437018,
+      "grad_norm": 0.6144537329673767,
+      "learning_rate": 2.106327409478186e-05,
+      "loss": 0.8456,
+      "step": 7896
+    },
+    {
+      "epoch": 0.7056090423749637,
+      "grad_norm": 0.5229153037071228,
+      "learning_rate": 2.1051474205507715e-05,
+      "loss": 0.9355,
+      "step": 7897
+    },
+    {
+      "epoch": 0.7056983939062256,
+      "grad_norm": 0.429524689912796,
+      "learning_rate": 2.1039676740981172e-05,
+      "loss": 0.954,
+      "step": 7898
+    },
+    {
+      "epoch": 0.7057877454374875,
+      "grad_norm": 0.5060709714889526,
+      "learning_rate": 2.1027881702190422e-05,
+      "loss": 0.8428,
+      "step": 7899
+    },
+    {
+      "epoch": 0.7058770969687493,
+      "grad_norm": 0.5516795516014099,
+      "learning_rate": 2.101608909012338e-05,
+      "loss": 0.9072,
+      "step": 7900
+    },
+    {
+      "epoch": 0.7059664485000112,
+      "grad_norm": 0.5318334102630615,
+      "learning_rate": 2.100429890576782e-05,
+      "loss": 0.9315,
+      "step": 7901
+    },
+    {
+      "epoch": 0.706055800031273,
+      "grad_norm": 0.54982990026474,
+      "learning_rate": 2.09925111501113e-05,
+      "loss": 0.969,
+      "step": 7902
+    },
+    {
+      "epoch": 0.7061451515625349,
+      "grad_norm": 0.4394816756248474,
+      "learning_rate": 2.0980725824141166e-05,
+      "loss": 1.042,
+      "step": 7903
+    },
+    {
+      "epoch": 0.7062345030937968,
+      "grad_norm": 0.579893171787262,
+      "learning_rate": 2.0968942928844593e-05,
+      "loss": 0.8299,
+      "step": 7904
+    },
+    {
+      "epoch": 0.7063238546250586,
+      "grad_norm": 0.4310462176799774,
+      "learning_rate": 2.0957162465208475e-05,
+      "loss": 0.9358,
+      "step": 7905
+    },
+    {
+      "epoch": 0.7064132061563205,
+      "grad_norm": 0.46636763215065,
+      "learning_rate": 2.094538443421958e-05,
+      "loss": 0.9262,
+      "step": 7906
+    },
+    {
+      "epoch": 0.7065025576875824,
+      "grad_norm": 0.5085285902023315,
+      "learning_rate": 2.0933608836864433e-05,
+      "loss": 0.9616,
+      "step": 7907
+    },
+    {
+      "epoch": 0.7065919092188442,
+      "grad_norm": 0.43514180183410645,
+      "learning_rate": 2.092183567412937e-05,
+      "loss": 1.0016,
+      "step": 7908
+    },
+    {
+      "epoch": 0.7066812607501061,
+      "grad_norm": 0.45315706729888916,
+      "learning_rate": 2.091006494700054e-05,
+      "loss": 0.9255,
+      "step": 7909
+    },
+    {
+      "epoch": 0.706770612281368,
+      "grad_norm": 0.46915316581726074,
+      "learning_rate": 2.0898296656463834e-05,
+      "loss": 0.9525,
+      "step": 7910
+    },
+    {
+      "epoch": 0.7068599638126298,
+      "grad_norm": 0.572519838809967,
+      "learning_rate": 2.0886530803504977e-05,
+      "loss": 0.9686,
+      "step": 7911
+    },
+    {
+      "epoch": 0.7069493153438917,
+      "grad_norm": 0.5691891312599182,
+      "learning_rate": 2.08747673891095e-05,
+      "loss": 0.9626,
+      "step": 7912
+    },
+    {
+      "epoch": 0.7070386668751536,
+      "grad_norm": 0.4699474275112152,
+      "learning_rate": 2.0863006414262703e-05,
+      "loss": 0.9125,
+      "step": 7913
+    },
+    {
+      "epoch": 0.7071280184064155,
+      "grad_norm": 0.5651034116744995,
+      "learning_rate": 2.0851247879949698e-05,
+      "loss": 0.8713,
+      "step": 7914
+    },
+    {
+      "epoch": 0.7072173699376773,
+      "grad_norm": 0.4276520609855652,
+      "learning_rate": 2.0839491787155387e-05,
+      "loss": 0.8954,
+      "step": 7915
+    },
+    {
+      "epoch": 0.7073067214689391,
+      "grad_norm": 0.5207197070121765,
+      "learning_rate": 2.0827738136864462e-05,
+      "loss": 0.9387,
+      "step": 7916
+    },
+    {
+      "epoch": 0.707396073000201,
+      "grad_norm": 0.43144193291664124,
+      "learning_rate": 2.0815986930061428e-05,
+      "loss": 0.8953,
+      "step": 7917
+    },
+    {
+      "epoch": 0.7074854245314629,
+      "grad_norm": 0.48623722791671753,
+      "learning_rate": 2.0804238167730566e-05,
+      "loss": 0.8455,
+      "step": 7918
+    },
+    {
+      "epoch": 0.7075747760627248,
+      "grad_norm": 0.4490886926651001,
+      "learning_rate": 2.0792491850855976e-05,
+      "loss": 0.9528,
+      "step": 7919
+    },
+    {
+      "epoch": 0.7076641275939867,
+      "grad_norm": 0.44778475165367126,
+      "learning_rate": 2.078074798042151e-05,
+      "loss": 0.9674,
+      "step": 7920
+    },
+    {
+      "epoch": 0.7077534791252486,
+      "grad_norm": 0.4196859300136566,
+      "learning_rate": 2.0769006557410858e-05,
+      "loss": 0.9325,
+      "step": 7921
+    },
+    {
+      "epoch": 0.7078428306565103,
+      "grad_norm": 0.4247509837150574,
+      "learning_rate": 2.0757267582807482e-05,
+      "loss": 0.9765,
+      "step": 7922
+    },
+    {
+      "epoch": 0.7079321821877722,
+      "grad_norm": 0.44996559619903564,
+      "learning_rate": 2.0745531057594654e-05,
+      "loss": 0.9291,
+      "step": 7923
+    },
+    {
+      "epoch": 0.7080215337190341,
+      "grad_norm": 0.4982374906539917,
+      "learning_rate": 2.0733796982755425e-05,
+      "loss": 0.8631,
+      "step": 7924
+    },
+    {
+      "epoch": 0.708110885250296,
+      "grad_norm": 0.515921413898468,
+      "learning_rate": 2.0722065359272657e-05,
+      "loss": 0.8803,
+      "step": 7925
+    },
+    {
+      "epoch": 0.7082002367815579,
+      "grad_norm": 0.4899452030658722,
+      "learning_rate": 2.0710336188128998e-05,
+      "loss": 0.8997,
+      "step": 7926
+    },
+    {
+      "epoch": 0.7082895883128197,
+      "grad_norm": 0.5061736106872559,
+      "learning_rate": 2.0698609470306885e-05,
+      "loss": 0.8947,
+      "step": 7927
+    },
+    {
+      "epoch": 0.7083789398440816,
+      "grad_norm": 0.45645084977149963,
+      "learning_rate": 2.0686885206788565e-05,
+      "loss": 0.9493,
+      "step": 7928
+    },
+    {
+      "epoch": 0.7084682913753434,
+      "grad_norm": 0.4615998864173889,
+      "learning_rate": 2.0675163398556073e-05,
+      "loss": 0.9419,
+      "step": 7929
+    },
+    {
+      "epoch": 0.7085576429066053,
+      "grad_norm": 0.5820156335830688,
+      "learning_rate": 2.066344404659122e-05,
+      "loss": 0.8803,
+      "step": 7930
+    },
+    {
+      "epoch": 0.7086469944378672,
+      "grad_norm": 0.48079875111579895,
+      "learning_rate": 2.065172715187562e-05,
+      "loss": 0.9185,
+      "step": 7931
+    },
+    {
+      "epoch": 0.708736345969129,
+      "grad_norm": 0.48415642976760864,
+      "learning_rate": 2.064001271539071e-05,
+      "loss": 0.9558,
+      "step": 7932
+    },
+    {
+      "epoch": 0.7088256975003909,
+      "grad_norm": 0.5177109241485596,
+      "learning_rate": 2.062830073811769e-05,
+      "loss": 0.9591,
+      "step": 7933
+    },
+    {
+      "epoch": 0.7089150490316528,
+      "grad_norm": 0.4747374653816223,
+      "learning_rate": 2.061659122103756e-05,
+      "loss": 1.048,
+      "step": 7934
+    },
+    {
+      "epoch": 0.7090044005629147,
+      "grad_norm": 0.5308428406715393,
+      "learning_rate": 2.0604884165131122e-05,
+      "loss": 0.8741,
+      "step": 7935
+    },
+    {
+      "epoch": 0.7090937520941765,
+      "grad_norm": 0.45492875576019287,
+      "learning_rate": 2.0593179571378964e-05,
+      "loss": 0.9447,
+      "step": 7936
+    },
+    {
+      "epoch": 0.7091831036254383,
+      "grad_norm": 0.534234881401062,
+      "learning_rate": 2.0581477440761488e-05,
+      "loss": 0.8724,
+      "step": 7937
+    },
+    {
+      "epoch": 0.7092724551567002,
+      "grad_norm": 0.4276201128959656,
+      "learning_rate": 2.0569777774258842e-05,
+      "loss": 0.9492,
+      "step": 7938
+    },
+    {
+      "epoch": 0.7093618066879621,
+      "grad_norm": 0.4283556044101715,
+      "learning_rate": 2.0558080572851002e-05,
+      "loss": 1.0475,
+      "step": 7939
+    },
+    {
+      "epoch": 0.709451158219224,
+      "grad_norm": 0.458793580532074,
+      "learning_rate": 2.054638583751775e-05,
+      "loss": 0.9828,
+      "step": 7940
+    },
+    {
+      "epoch": 0.7095405097504859,
+      "grad_norm": 0.41058874130249023,
+      "learning_rate": 2.053469356923865e-05,
+      "loss": 0.9937,
+      "step": 7941
+    },
+    {
+      "epoch": 0.7096298612817478,
+      "grad_norm": 0.4860825538635254,
+      "learning_rate": 2.0523003768993025e-05,
+      "loss": 0.9317,
+      "step": 7942
+    },
+    {
+      "epoch": 0.7097192128130095,
+      "grad_norm": 0.46478596329689026,
+      "learning_rate": 2.0511316437760042e-05,
+      "loss": 0.9276,
+      "step": 7943
+    },
+    {
+      "epoch": 0.7098085643442714,
+      "grad_norm": 0.504979133605957,
+      "learning_rate": 2.049963157651863e-05,
+      "loss": 0.9422,
+      "step": 7944
+    },
+    {
+      "epoch": 0.7098979158755333,
+      "grad_norm": 0.5935678482055664,
+      "learning_rate": 2.0487949186247524e-05,
+      "loss": 0.8924,
+      "step": 7945
+    },
+    {
+      "epoch": 0.7099872674067952,
+      "grad_norm": 0.4765070080757141,
+      "learning_rate": 2.0476269267925247e-05,
+      "loss": 0.9904,
+      "step": 7946
+    },
+    {
+      "epoch": 0.7100766189380571,
+      "grad_norm": 0.4645882844924927,
+      "learning_rate": 2.0464591822530123e-05,
+      "loss": 0.9594,
+      "step": 7947
+    },
+    {
+      "epoch": 0.710165970469319,
+      "grad_norm": 0.4871085584163666,
+      "learning_rate": 2.0452916851040256e-05,
+      "loss": 0.8937,
+      "step": 7948
+    },
+    {
+      "epoch": 0.7102553220005808,
+      "grad_norm": 0.44038885831832886,
+      "learning_rate": 2.0441244354433568e-05,
+      "loss": 0.951,
+      "step": 7949
+    },
+    {
+      "epoch": 0.7103446735318426,
+      "grad_norm": 0.5087218880653381,
+      "learning_rate": 2.042957433368773e-05,
+      "loss": 0.9512,
+      "step": 7950
+    },
+    {
+      "epoch": 0.7104340250631045,
+      "grad_norm": 0.5866579413414001,
+      "learning_rate": 2.0417906789780235e-05,
+      "loss": 0.8355,
+      "step": 7951
+    },
+    {
+      "epoch": 0.7105233765943664,
+      "grad_norm": 0.4508510231971741,
+      "learning_rate": 2.0406241723688362e-05,
+      "loss": 0.9801,
+      "step": 7952
+    },
+    {
+      "epoch": 0.7106127281256283,
+      "grad_norm": 0.4516523778438568,
+      "learning_rate": 2.0394579136389203e-05,
+      "loss": 0.9136,
+      "step": 7953
+    },
+    {
+      "epoch": 0.7107020796568901,
+      "grad_norm": 0.45436009764671326,
+      "learning_rate": 2.0382919028859605e-05,
+      "loss": 0.8881,
+      "step": 7954
+    },
+    {
+      "epoch": 0.710791431188152,
+      "grad_norm": 0.6993239521980286,
+      "learning_rate": 2.0371261402076236e-05,
+      "loss": 0.9434,
+      "step": 7955
+    },
+    {
+      "epoch": 0.7108807827194139,
+      "grad_norm": 0.47904080152511597,
+      "learning_rate": 2.0359606257015546e-05,
+      "loss": 1.0162,
+      "step": 7956
+    },
+    {
+      "epoch": 0.7109701342506757,
+      "grad_norm": 0.4850071370601654,
+      "learning_rate": 2.034795359465377e-05,
+      "loss": 0.9378,
+      "step": 7957
+    },
+    {
+      "epoch": 0.7110594857819376,
+      "grad_norm": 0.5238434672355652,
+      "learning_rate": 2.0336303415966952e-05,
+      "loss": 0.9051,
+      "step": 7958
+    },
+    {
+      "epoch": 0.7111488373131994,
+      "grad_norm": 0.5065889954566956,
+      "learning_rate": 2.0324655721930937e-05,
+      "loss": 0.9456,
+      "step": 7959
+    },
+    {
+      "epoch": 0.7112381888444613,
+      "grad_norm": 0.5274893045425415,
+      "learning_rate": 2.0313010513521298e-05,
+      "loss": 0.9309,
+      "step": 7960
+    },
+    {
+      "epoch": 0.7113275403757232,
+      "grad_norm": 0.47935160994529724,
+      "learning_rate": 2.030136779171347e-05,
+      "loss": 0.9542,
+      "step": 7961
+    },
+    {
+      "epoch": 0.7114168919069851,
+      "grad_norm": 0.4876161515712738,
+      "learning_rate": 2.0289727557482656e-05,
+      "loss": 0.881,
+      "step": 7962
+    },
+    {
+      "epoch": 0.711506243438247,
+      "grad_norm": 0.49741584062576294,
+      "learning_rate": 2.0278089811803846e-05,
+      "loss": 0.9275,
+      "step": 7963
+    },
+    {
+      "epoch": 0.7115955949695087,
+      "grad_norm": 0.5018129944801331,
+      "learning_rate": 2.0266454555651825e-05,
+      "loss": 0.9503,
+      "step": 7964
+    },
+    {
+      "epoch": 0.7116849465007706,
+      "grad_norm": 0.46502235531806946,
+      "learning_rate": 2.0254821790001177e-05,
+      "loss": 0.96,
+      "step": 7965
+    },
+    {
+      "epoch": 0.7117742980320325,
+      "grad_norm": 0.5179951786994934,
+      "learning_rate": 2.0243191515826265e-05,
+      "loss": 1.0107,
+      "step": 7966
+    },
+    {
+      "epoch": 0.7118636495632944,
+      "grad_norm": 0.49856871366500854,
+      "learning_rate": 2.0231563734101243e-05,
+      "loss": 0.9671,
+      "step": 7967
+    },
+    {
+      "epoch": 0.7119530010945563,
+      "grad_norm": 0.4230417013168335,
+      "learning_rate": 2.0219938445800074e-05,
+      "loss": 0.9217,
+      "step": 7968
+    },
+    {
+      "epoch": 0.7120423526258182,
+      "grad_norm": 0.5896261930465698,
+      "learning_rate": 2.0208315651896504e-05,
+      "loss": 0.9198,
+      "step": 7969
+    },
+    {
+      "epoch": 0.7121317041570799,
+      "grad_norm": 0.46407637000083923,
+      "learning_rate": 2.0196695353364042e-05,
+      "loss": 1.0145,
+      "step": 7970
+    },
+    {
+      "epoch": 0.7122210556883418,
+      "grad_norm": 0.39339298009872437,
+      "learning_rate": 2.018507755117602e-05,
+      "loss": 0.9654,
+      "step": 7971
+    },
+    {
+      "epoch": 0.7123104072196037,
+      "grad_norm": 0.5174713730812073,
+      "learning_rate": 2.017346224630556e-05,
+      "loss": 0.9395,
+      "step": 7972
+    },
+    {
+      "epoch": 0.7123997587508656,
+      "grad_norm": 0.5631799697875977,
+      "learning_rate": 2.0161849439725565e-05,
+      "loss": 0.8938,
+      "step": 7973
+    },
+    {
+      "epoch": 0.7124891102821275,
+      "grad_norm": 0.4893137812614441,
+      "learning_rate": 2.015023913240875e-05,
+      "loss": 0.9324,
+      "step": 7974
+    },
+    {
+      "epoch": 0.7125784618133894,
+      "grad_norm": 0.495174765586853,
+      "learning_rate": 2.0138631325327563e-05,
+      "loss": 0.9131,
+      "step": 7975
+    },
+    {
+      "epoch": 0.7126678133446512,
+      "grad_norm": 0.4327716827392578,
+      "learning_rate": 2.0127026019454304e-05,
+      "loss": 0.9635,
+      "step": 7976
+    },
+    {
+      "epoch": 0.712757164875913,
+      "grad_norm": 0.3611029386520386,
+      "learning_rate": 2.0115423215761037e-05,
+      "loss": 0.9878,
+      "step": 7977
+    },
+    {
+      "epoch": 0.7128465164071749,
+      "grad_norm": 0.4893472492694855,
+      "learning_rate": 2.0103822915219624e-05,
+      "loss": 0.9732,
+      "step": 7978
+    },
+    {
+      "epoch": 0.7129358679384368,
+      "grad_norm": 0.4027042090892792,
+      "learning_rate": 2.0092225118801706e-05,
+      "loss": 0.9217,
+      "step": 7979
+    },
+    {
+      "epoch": 0.7130252194696987,
+      "grad_norm": 0.4319385290145874,
+      "learning_rate": 2.0080629827478753e-05,
+      "loss": 0.92,
+      "step": 7980
+    },
+    {
+      "epoch": 0.7131145710009605,
+      "grad_norm": 0.48171430826187134,
+      "learning_rate": 2.006903704222195e-05,
+      "loss": 0.8967,
+      "step": 7981
+    },
+    {
+      "epoch": 0.7132039225322224,
+      "grad_norm": 0.5436350703239441,
+      "learning_rate": 2.0057446764002337e-05,
+      "loss": 0.9484,
+      "step": 7982
+    },
+    {
+      "epoch": 0.7132932740634843,
+      "grad_norm": 0.42786580324172974,
+      "learning_rate": 2.004585899379072e-05,
+      "loss": 1.0898,
+      "step": 7983
+    },
+    {
+      "epoch": 0.7133826255947461,
+      "grad_norm": 0.4779389798641205,
+      "learning_rate": 2.00342737325577e-05,
+      "loss": 0.9715,
+      "step": 7984
+    },
+    {
+      "epoch": 0.713471977126008,
+      "grad_norm": 0.47203120589256287,
+      "learning_rate": 2.0022690981273666e-05,
+      "loss": 0.9221,
+      "step": 7985
+    },
+    {
+      "epoch": 0.7135613286572698,
+      "grad_norm": 0.43961775302886963,
+      "learning_rate": 2.0011110740908802e-05,
+      "loss": 0.9608,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7136506801885317,
+      "grad_norm": 0.5077892541885376,
+      "learning_rate": 1.999953301243307e-05,
+      "loss": 0.9414,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7137400317197936,
+      "grad_norm": 0.5579219460487366,
+      "learning_rate": 1.998795779681623e-05,
+      "loss": 0.9396,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7138293832510555,
+      "grad_norm": 0.5244463682174683,
+      "learning_rate": 1.9976385095027826e-05,
+      "loss": 0.938,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7139187347823174,
+      "grad_norm": 0.5091515779495239,
+      "learning_rate": 1.9964814908037223e-05,
+      "loss": 0.9086,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7140080863135791,
+      "grad_norm": 0.5169358253479004,
+      "learning_rate": 1.9953247236813504e-05,
+      "loss": 0.9151,
+      "step": 7991
+    },
+    {
+      "epoch": 0.714097437844841,
+      "grad_norm": 0.509032666683197,
+      "learning_rate": 1.9941682082325602e-05,
+      "loss": 0.8878,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7141867893761029,
+      "grad_norm": 0.3874991238117218,
+      "learning_rate": 1.9930119445542227e-05,
+      "loss": 0.9988,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7142761409073648,
+      "grad_norm": 0.48893627524375916,
+      "learning_rate": 1.991855932743187e-05,
+      "loss": 0.9737,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7143654924386267,
+      "grad_norm": 0.5337826013565063,
+      "learning_rate": 1.990700172896281e-05,
+      "loss": 0.9399,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7144548439698886,
+      "grad_norm": 0.5026716589927673,
+      "learning_rate": 1.989544665110313e-05,
+      "loss": 0.9656,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7145441955011504,
+      "grad_norm": 0.5005045533180237,
+      "learning_rate": 1.988389409482068e-05,
+      "loss": 1.0192,
+      "step": 7997
+    },
+    {
+      "epoch": 0.7146335470324122,
+      "grad_norm": 0.45082736015319824,
+      "learning_rate": 1.9872344061083113e-05,
+      "loss": 0.9364,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7147228985636741,
+      "grad_norm": 0.4575754702091217,
+      "learning_rate": 1.9860796550857872e-05,
+      "loss": 0.9433,
+      "step": 7999
+    },
+    {
+      "epoch": 0.714812250094936,
+      "grad_norm": 0.5190432071685791,
+      "learning_rate": 1.98492515651122e-05,
+      "loss": 0.9311,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7149016016261979,
+      "grad_norm": 0.4815937578678131,
+      "learning_rate": 1.9837709104813073e-05,
+      "loss": 0.9925,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7149909531574598,
+      "grad_norm": 0.4584693908691406,
+      "learning_rate": 1.9826169170927317e-05,
+      "loss": 0.9623,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7150803046887216,
+      "grad_norm": 0.4569786489009857,
+      "learning_rate": 1.9814631764421522e-05,
+      "loss": 0.9571,
+      "step": 8003
+    },
+    {
+      "epoch": 0.7151696562199835,
+      "grad_norm": 0.47242796421051025,
+      "learning_rate": 1.9803096886262068e-05,
+      "loss": 0.864,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7152590077512453,
+      "grad_norm": 0.4647274613380432,
+      "learning_rate": 1.9791564537415124e-05,
+      "loss": 0.9579,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7153483592825072,
+      "grad_norm": 0.4390351474285126,
+      "learning_rate": 1.978003471884665e-05,
+      "loss": 0.9682,
+      "step": 8006
+    },
+    {
+      "epoch": 0.7154377108137691,
+      "grad_norm": 0.4374440908432007,
+      "learning_rate": 1.976850743152241e-05,
+      "loss": 0.9661,
+      "step": 8007
+    },
+    {
+      "epoch": 0.7155270623450309,
+      "grad_norm": 0.48356321454048157,
+      "learning_rate": 1.97569826764079e-05,
+      "loss": 0.9314,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7156164138762928,
+      "grad_norm": 0.37716495990753174,
+      "learning_rate": 1.9745460454468457e-05,
+      "loss": 0.9704,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7157057654075547,
+      "grad_norm": 0.3916030526161194,
+      "learning_rate": 1.973394076666919e-05,
+      "loss": 0.9635,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7157951169388166,
+      "grad_norm": 0.5709711909294128,
+      "learning_rate": 1.9722423613975016e-05,
+      "loss": 0.9523,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7158844684700784,
+      "grad_norm": 0.4928154945373535,
+      "learning_rate": 1.971090899735058e-05,
+      "loss": 0.8978,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7159738200013402,
+      "grad_norm": 0.4418443739414215,
+      "learning_rate": 1.9699396917760377e-05,
+      "loss": 0.8998,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7160631715326021,
+      "grad_norm": 0.5247768759727478,
+      "learning_rate": 1.9687887376168663e-05,
+      "loss": 1.0516,
+      "step": 8014
+    },
+    {
+      "epoch": 0.716152523063864,
+      "grad_norm": 0.5156433582305908,
+      "learning_rate": 1.9676380373539482e-05,
+      "loss": 0.9109,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7162418745951259,
+      "grad_norm": 0.4411882758140564,
+      "learning_rate": 1.9664875910836677e-05,
+      "loss": 0.9728,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7163312261263878,
+      "grad_norm": 0.5197750329971313,
+      "learning_rate": 1.965337398902386e-05,
+      "loss": 0.9223,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7164205776576497,
+      "grad_norm": 0.45580387115478516,
+      "learning_rate": 1.9641874609064443e-05,
+      "loss": 0.9332,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7165099291889114,
+      "grad_norm": 0.4132586419582367,
+      "learning_rate": 1.963037777192162e-05,
+      "loss": 0.9757,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7165992807201733,
+      "grad_norm": 0.5496894121170044,
+      "learning_rate": 1.96188834785584e-05,
+      "loss": 1.0128,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7166886322514352,
+      "grad_norm": 0.4949209988117218,
+      "learning_rate": 1.9607391729937503e-05,
+      "loss": 0.9594,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7167779837826971,
+      "grad_norm": 0.4753008782863617,
+      "learning_rate": 1.9595902527021513e-05,
+      "loss": 0.9873,
+      "step": 8022
+    },
+    {
+      "epoch": 0.716867335313959,
+      "grad_norm": 0.4486302435398102,
+      "learning_rate": 1.958441587077277e-05,
+      "loss": 0.8871,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7169566868452208,
+      "grad_norm": 0.42516544461250305,
+      "learning_rate": 1.9572931762153407e-05,
+      "loss": 1.0009,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7170460383764827,
+      "grad_norm": 0.4175739288330078,
+      "learning_rate": 1.9561450202125337e-05,
+      "loss": 0.9372,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7171353899077445,
+      "grad_norm": 0.5125389695167542,
+      "learning_rate": 1.9549971191650262e-05,
+      "loss": 0.9469,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7172247414390064,
+      "grad_norm": 0.5798017382621765,
+      "learning_rate": 1.953849473168968e-05,
+      "loss": 0.8865,
+      "step": 8027
+    },
+    {
+      "epoch": 0.7173140929702683,
+      "grad_norm": 0.4292055666446686,
+      "learning_rate": 1.9527020823204856e-05,
+      "loss": 0.991,
+      "step": 8028
+    },
+    {
+      "epoch": 0.7174034445015302,
+      "grad_norm": 0.41361159086227417,
+      "learning_rate": 1.951554946715686e-05,
+      "loss": 0.9773,
+      "step": 8029
+    },
+    {
+      "epoch": 0.717492796032792,
+      "grad_norm": 0.5009101629257202,
+      "learning_rate": 1.9504080664506546e-05,
+      "loss": 0.9358,
+      "step": 8030
+    },
+    {
+      "epoch": 0.7175821475640539,
+      "grad_norm": 0.6405025124549866,
+      "learning_rate": 1.9492614416214526e-05,
+      "loss": 0.9336,
+      "step": 8031
+    },
+    {
+      "epoch": 0.7176714990953158,
+      "grad_norm": 0.440360426902771,
+      "learning_rate": 1.9481150723241236e-05,
+      "loss": 0.923,
+      "step": 8032
+    },
+    {
+      "epoch": 0.7177608506265776,
+      "grad_norm": 0.4327067732810974,
+      "learning_rate": 1.946968958654688e-05,
+      "loss": 0.9564,
+      "step": 8033
+    },
+    {
+      "epoch": 0.7178502021578395,
+      "grad_norm": 0.6656114459037781,
+      "learning_rate": 1.9458231007091456e-05,
+      "loss": 0.9387,
+      "step": 8034
+    },
+    {
+      "epoch": 0.7179395536891013,
+      "grad_norm": 0.5067597031593323,
+      "learning_rate": 1.9446774985834726e-05,
+      "loss": 0.905,
+      "step": 8035
+    },
+    {
+      "epoch": 0.7180289052203632,
+      "grad_norm": 0.48535487055778503,
+      "learning_rate": 1.943532152373627e-05,
+      "loss": 0.8893,
+      "step": 8036
+    },
+    {
+      "epoch": 0.7181182567516251,
+      "grad_norm": 0.6070266962051392,
+      "learning_rate": 1.9423870621755434e-05,
+      "loss": 0.8946,
+      "step": 8037
+    },
+    {
+      "epoch": 0.718207608282887,
+      "grad_norm": 0.5179934501647949,
+      "learning_rate": 1.941242228085135e-05,
+      "loss": 0.8965,
+      "step": 8038
+    },
+    {
+      "epoch": 0.7182969598141488,
+      "grad_norm": 0.49246200919151306,
+      "learning_rate": 1.9400976501982943e-05,
+      "loss": 0.8797,
+      "step": 8039
+    },
+    {
+      "epoch": 0.7183863113454106,
+      "grad_norm": 0.47545865178108215,
+      "learning_rate": 1.938953328610893e-05,
+      "loss": 0.9602,
+      "step": 8040
+    },
+    {
+      "epoch": 0.7184756628766725,
+      "grad_norm": 0.5661696791648865,
+      "learning_rate": 1.9378092634187772e-05,
+      "loss": 0.8881,
+      "step": 8041
+    },
+    {
+      "epoch": 0.7185650144079344,
+      "grad_norm": 0.5181058645248413,
+      "learning_rate": 1.9366654547177764e-05,
+      "loss": 0.9065,
+      "step": 8042
+    },
+    {
+      "epoch": 0.7186543659391963,
+      "grad_norm": 0.42876681685447693,
+      "learning_rate": 1.9355219026036965e-05,
+      "loss": 1.0135,
+      "step": 8043
+    },
+    {
+      "epoch": 0.7187437174704582,
+      "grad_norm": 0.5706227421760559,
+      "learning_rate": 1.934378607172324e-05,
+      "loss": 0.9067,
+      "step": 8044
+    },
+    {
+      "epoch": 0.7188330690017201,
+      "grad_norm": 0.4501083791255951,
+      "learning_rate": 1.9332355685194182e-05,
+      "loss": 0.9817,
+      "step": 8045
+    },
+    {
+      "epoch": 0.7189224205329818,
+      "grad_norm": 0.5079956650733948,
+      "learning_rate": 1.9320927867407234e-05,
+      "loss": 0.9354,
+      "step": 8046
+    },
+    {
+      "epoch": 0.7190117720642437,
+      "grad_norm": 0.43281540274620056,
+      "learning_rate": 1.9309502619319592e-05,
+      "loss": 0.9358,
+      "step": 8047
+    },
+    {
+      "epoch": 0.7191011235955056,
+      "grad_norm": 0.45937132835388184,
+      "learning_rate": 1.9298079941888237e-05,
+      "loss": 0.9387,
+      "step": 8048
+    },
+    {
+      "epoch": 0.7191904751267675,
+      "grad_norm": 0.572298526763916,
+      "learning_rate": 1.9286659836069953e-05,
+      "loss": 0.8596,
+      "step": 8049
+    },
+    {
+      "epoch": 0.7192798266580294,
+      "grad_norm": 0.4632246494293213,
+      "learning_rate": 1.9275242302821302e-05,
+      "loss": 0.9057,
+      "step": 8050
+    },
+    {
+      "epoch": 0.7193691781892912,
+      "grad_norm": 0.46226373314857483,
+      "learning_rate": 1.9263827343098594e-05,
+      "loss": 0.9242,
+      "step": 8051
+    },
+    {
+      "epoch": 0.7194585297205531,
+      "grad_norm": 0.4486196041107178,
+      "learning_rate": 1.9252414957857966e-05,
+      "loss": 0.9282,
+      "step": 8052
+    },
+    {
+      "epoch": 0.7195478812518149,
+      "grad_norm": 0.5384345054626465,
+      "learning_rate": 1.9241005148055336e-05,
+      "loss": 0.904,
+      "step": 8053
+    },
+    {
+      "epoch": 0.7196372327830768,
+      "grad_norm": 0.5714906454086304,
+      "learning_rate": 1.922959791464639e-05,
+      "loss": 0.8597,
+      "step": 8054
+    },
+    {
+      "epoch": 0.7197265843143387,
+      "grad_norm": 0.5070673227310181,
+      "learning_rate": 1.921819325858661e-05,
+      "loss": 0.9542,
+      "step": 8055
+    },
+    {
+      "epoch": 0.7198159358456006,
+      "grad_norm": 0.5547393560409546,
+      "learning_rate": 1.920679118083125e-05,
+      "loss": 0.892,
+      "step": 8056
+    },
+    {
+      "epoch": 0.7199052873768624,
+      "grad_norm": 0.4647422134876251,
+      "learning_rate": 1.9195391682335368e-05,
+      "loss": 0.8927,
+      "step": 8057
+    },
+    {
+      "epoch": 0.7199946389081243,
+      "grad_norm": 0.4780432879924774,
+      "learning_rate": 1.918399476405378e-05,
+      "loss": 0.964,
+      "step": 8058
+    },
+    {
+      "epoch": 0.7200839904393862,
+      "grad_norm": 0.5166242122650146,
+      "learning_rate": 1.9172600426941113e-05,
+      "loss": 0.9151,
+      "step": 8059
+    },
+    {
+      "epoch": 0.720173341970648,
+      "grad_norm": 0.4829009771347046,
+      "learning_rate": 1.9161208671951763e-05,
+      "loss": 0.9347,
+      "step": 8060
+    },
+    {
+      "epoch": 0.7202626935019099,
+      "grad_norm": 0.5364348292350769,
+      "learning_rate": 1.9149819500039896e-05,
+      "loss": 0.9733,
+      "step": 8061
+    },
+    {
+      "epoch": 0.7203520450331717,
+      "grad_norm": 0.43714621663093567,
+      "learning_rate": 1.913843291215948e-05,
+      "loss": 0.9185,
+      "step": 8062
+    },
+    {
+      "epoch": 0.7204413965644336,
+      "grad_norm": 0.5605136752128601,
+      "learning_rate": 1.912704890926427e-05,
+      "loss": 0.8812,
+      "step": 8063
+    },
+    {
+      "epoch": 0.7205307480956955,
+      "grad_norm": 0.5223738551139832,
+      "learning_rate": 1.9115667492307797e-05,
+      "loss": 0.8438,
+      "step": 8064
+    },
+    {
+      "epoch": 0.7206200996269574,
+      "grad_norm": 0.46035128831863403,
+      "learning_rate": 1.9104288662243375e-05,
+      "loss": 0.9511,
+      "step": 8065
+    },
+    {
+      "epoch": 0.7207094511582193,
+      "grad_norm": 0.3924838602542877,
+      "learning_rate": 1.90929124200241e-05,
+      "loss": 0.9798,
+      "step": 8066
+    },
+    {
+      "epoch": 0.720798802689481,
+      "grad_norm": 0.47366228699684143,
+      "learning_rate": 1.9081538766602857e-05,
+      "loss": 1.0173,
+      "step": 8067
+    },
+    {
+      "epoch": 0.7208881542207429,
+      "grad_norm": 0.4460136294364929,
+      "learning_rate": 1.9070167702932313e-05,
+      "loss": 0.9475,
+      "step": 8068
+    },
+    {
+      "epoch": 0.7209775057520048,
+      "grad_norm": 0.4173930883407593,
+      "learning_rate": 1.9058799229964907e-05,
+      "loss": 0.9413,
+      "step": 8069
+    },
+    {
+      "epoch": 0.7210668572832667,
+      "grad_norm": 0.5706358551979065,
+      "learning_rate": 1.9047433348652876e-05,
+      "loss": 0.8534,
+      "step": 8070
+    },
+    {
+      "epoch": 0.7211562088145286,
+      "grad_norm": 0.47395849227905273,
+      "learning_rate": 1.9036070059948252e-05,
+      "loss": 0.9308,
+      "step": 8071
+    },
+    {
+      "epoch": 0.7212455603457905,
+      "grad_norm": 0.461373507976532,
+      "learning_rate": 1.9024709364802795e-05,
+      "loss": 0.8572,
+      "step": 8072
+    },
+    {
+      "epoch": 0.7213349118770523,
+      "grad_norm": 0.40039876103401184,
+      "learning_rate": 1.90133512641681e-05,
+      "loss": 0.9251,
+      "step": 8073
+    },
+    {
+      "epoch": 0.7214242634083141,
+      "grad_norm": 0.4980182945728302,
+      "learning_rate": 1.9001995758995533e-05,
+      "loss": 0.9338,
+      "step": 8074
+    },
+    {
+      "epoch": 0.721513614939576,
+      "grad_norm": 0.4447602927684784,
+      "learning_rate": 1.899064285023623e-05,
+      "loss": 0.9201,
+      "step": 8075
+    },
+    {
+      "epoch": 0.7216029664708379,
+      "grad_norm": 0.404520183801651,
+      "learning_rate": 1.8979292538841132e-05,
+      "loss": 0.9922,
+      "step": 8076
+    },
+    {
+      "epoch": 0.7216923180020998,
+      "grad_norm": 0.4557284116744995,
+      "learning_rate": 1.8967944825760954e-05,
+      "loss": 0.9549,
+      "step": 8077
+    },
+    {
+      "epoch": 0.7217816695333616,
+      "grad_norm": 0.45047318935394287,
+      "learning_rate": 1.8956599711946156e-05,
+      "loss": 0.9628,
+      "step": 8078
+    },
+    {
+      "epoch": 0.7218710210646235,
+      "grad_norm": 0.4499797224998474,
+      "learning_rate": 1.8945257198347034e-05,
+      "loss": 0.9302,
+      "step": 8079
+    },
+    {
+      "epoch": 0.7219603725958854,
+      "grad_norm": 0.4355414807796478,
+      "learning_rate": 1.893391728591364e-05,
+      "loss": 0.942,
+      "step": 8080
+    },
+    {
+      "epoch": 0.7220497241271472,
+      "grad_norm": 0.5795384049415588,
+      "learning_rate": 1.8922579975595835e-05,
+      "loss": 0.8788,
+      "step": 8081
+    },
+    {
+      "epoch": 0.7221390756584091,
+      "grad_norm": 0.5405740737915039,
+      "learning_rate": 1.891124526834319e-05,
+      "loss": 0.9337,
+      "step": 8082
+    },
+    {
+      "epoch": 0.722228427189671,
+      "grad_norm": 0.4552251696586609,
+      "learning_rate": 1.889991316510515e-05,
+      "loss": 1.0248,
+      "step": 8083
+    },
+    {
+      "epoch": 0.7223177787209328,
+      "grad_norm": 0.3918149471282959,
+      "learning_rate": 1.8888583666830876e-05,
+      "loss": 1.0259,
+      "step": 8084
+    },
+    {
+      "epoch": 0.7224071302521947,
+      "grad_norm": 0.5067173838615417,
+      "learning_rate": 1.887725677446935e-05,
+      "loss": 0.9839,
+      "step": 8085
+    },
+    {
+      "epoch": 0.7224964817834566,
+      "grad_norm": 0.5742607116699219,
+      "learning_rate": 1.8865932488969307e-05,
+      "loss": 0.8916,
+      "step": 8086
+    },
+    {
+      "epoch": 0.7225858333147185,
+      "grad_norm": 0.5914955735206604,
+      "learning_rate": 1.8854610811279288e-05,
+      "loss": 0.9132,
+      "step": 8087
+    },
+    {
+      "epoch": 0.7226751848459803,
+      "grad_norm": 0.479939341545105,
+      "learning_rate": 1.88432917423476e-05,
+      "loss": 1.0004,
+      "step": 8088
+    },
+    {
+      "epoch": 0.7227645363772421,
+      "grad_norm": 0.5861053466796875,
+      "learning_rate": 1.883197528312233e-05,
+      "loss": 0.9021,
+      "step": 8089
+    },
+    {
+      "epoch": 0.722853887908504,
+      "grad_norm": 0.607997715473175,
+      "learning_rate": 1.8820661434551362e-05,
+      "loss": 0.8781,
+      "step": 8090
+    },
+    {
+      "epoch": 0.7229432394397659,
+      "grad_norm": 0.4828149378299713,
+      "learning_rate": 1.8809350197582364e-05,
+      "loss": 0.865,
+      "step": 8091
+    },
+    {
+      "epoch": 0.7230325909710278,
+      "grad_norm": 0.6091709136962891,
+      "learning_rate": 1.879804157316274e-05,
+      "loss": 0.9282,
+      "step": 8092
+    },
+    {
+      "epoch": 0.7231219425022897,
+      "grad_norm": 0.5312778353691101,
+      "learning_rate": 1.878673556223972e-05,
+      "loss": 0.9184,
+      "step": 8093
+    },
+    {
+      "epoch": 0.7232112940335516,
+      "grad_norm": 0.41062840819358826,
+      "learning_rate": 1.8775432165760303e-05,
+      "loss": 0.9453,
+      "step": 8094
+    },
+    {
+      "epoch": 0.7233006455648133,
+      "grad_norm": 0.6178465485572815,
+      "learning_rate": 1.876413138467128e-05,
+      "loss": 0.9003,
+      "step": 8095
+    },
+    {
+      "epoch": 0.7233899970960752,
+      "grad_norm": 0.4309876561164856,
+      "learning_rate": 1.8752833219919197e-05,
+      "loss": 0.9821,
+      "step": 8096
+    },
+    {
+      "epoch": 0.7234793486273371,
+      "grad_norm": 0.5962274670600891,
+      "learning_rate": 1.8741537672450405e-05,
+      "loss": 0.9284,
+      "step": 8097
+    },
+    {
+      "epoch": 0.723568700158599,
+      "grad_norm": 0.4582142233848572,
+      "learning_rate": 1.8730244743211027e-05,
+      "loss": 0.938,
+      "step": 8098
+    },
+    {
+      "epoch": 0.7236580516898609,
+      "grad_norm": 0.4365943670272827,
+      "learning_rate": 1.8718954433146963e-05,
+      "loss": 0.9746,
+      "step": 8099
+    },
+    {
+      "epoch": 0.7237474032211227,
+      "grad_norm": 0.4775666892528534,
+      "learning_rate": 1.8707666743203893e-05,
+      "loss": 0.9665,
+      "step": 8100
+    },
+    {
+      "epoch": 0.7238367547523845,
+      "grad_norm": 0.43144118785858154,
+      "learning_rate": 1.8696381674327308e-05,
+      "loss": 0.9532,
+      "step": 8101
+    },
+    {
+      "epoch": 0.7239261062836464,
+      "grad_norm": 0.5456085205078125,
+      "learning_rate": 1.8685099227462406e-05,
+      "loss": 0.9138,
+      "step": 8102
+    },
+    {
+      "epoch": 0.7240154578149083,
+      "grad_norm": 0.44839584827423096,
+      "learning_rate": 1.8673819403554244e-05,
+      "loss": 0.9408,
+      "step": 8103
+    },
+    {
+      "epoch": 0.7241048093461702,
+      "grad_norm": 0.4968213737010956,
+      "learning_rate": 1.8662542203547616e-05,
+      "loss": 0.9101,
+      "step": 8104
+    },
+    {
+      "epoch": 0.724194160877432,
+      "grad_norm": 0.5285754203796387,
+      "learning_rate": 1.8651267628387108e-05,
+      "loss": 0.8168,
+      "step": 8105
+    },
+    {
+      "epoch": 0.7242835124086939,
+      "grad_norm": 0.4092321991920471,
+      "learning_rate": 1.86399956790171e-05,
+      "loss": 0.9397,
+      "step": 8106
+    },
+    {
+      "epoch": 0.7243728639399558,
+      "grad_norm": 0.6239436864852905,
+      "learning_rate": 1.862872635638172e-05,
+      "loss": 0.8964,
+      "step": 8107
+    },
+    {
+      "epoch": 0.7244622154712176,
+      "grad_norm": 0.5783222913742065,
+      "learning_rate": 1.86174596614249e-05,
+      "loss": 0.8807,
+      "step": 8108
+    },
+    {
+      "epoch": 0.7245515670024795,
+      "grad_norm": 0.43302804231643677,
+      "learning_rate": 1.8606195595090354e-05,
+      "loss": 0.9818,
+      "step": 8109
+    },
+    {
+      "epoch": 0.7246409185337414,
+      "grad_norm": 0.4686475694179535,
+      "learning_rate": 1.859493415832157e-05,
+      "loss": 0.9568,
+      "step": 8110
+    },
+    {
+      "epoch": 0.7247302700650032,
+      "grad_norm": 0.6813352704048157,
+      "learning_rate": 1.8583675352061792e-05,
+      "loss": 0.878,
+      "step": 8111
+    },
+    {
+      "epoch": 0.7248196215962651,
+      "grad_norm": 0.5830407738685608,
+      "learning_rate": 1.857241917725408e-05,
+      "loss": 0.8521,
+      "step": 8112
+    },
+    {
+      "epoch": 0.724908973127527,
+      "grad_norm": 0.5068725347518921,
+      "learning_rate": 1.856116563484125e-05,
+      "loss": 0.8299,
+      "step": 8113
+    },
+    {
+      "epoch": 0.7249983246587889,
+      "grad_norm": 0.4167884886264801,
+      "learning_rate": 1.8549914725765932e-05,
+      "loss": 1.0106,
+      "step": 8114
+    },
+    {
+      "epoch": 0.7250876761900507,
+      "grad_norm": 0.5126107335090637,
+      "learning_rate": 1.853866645097047e-05,
+      "loss": 0.893,
+      "step": 8115
+    },
+    {
+      "epoch": 0.7251770277213125,
+      "grad_norm": 0.46296408772468567,
+      "learning_rate": 1.852742081139705e-05,
+      "loss": 0.8905,
+      "step": 8116
+    },
+    {
+      "epoch": 0.7252663792525744,
+      "grad_norm": 0.5143711566925049,
+      "learning_rate": 1.8516177807987606e-05,
+      "loss": 0.9687,
+      "step": 8117
+    },
+    {
+      "epoch": 0.7253557307838363,
+      "grad_norm": 0.5294638276100159,
+      "learning_rate": 1.8504937441683868e-05,
+      "loss": 0.931,
+      "step": 8118
+    },
+    {
+      "epoch": 0.7254450823150982,
+      "grad_norm": 0.4719218909740448,
+      "learning_rate": 1.8493699713427333e-05,
+      "loss": 0.9807,
+      "step": 8119
+    },
+    {
+      "epoch": 0.7255344338463601,
+      "grad_norm": 0.4698675870895386,
+      "learning_rate": 1.848246462415928e-05,
+      "loss": 1.0442,
+      "step": 8120
+    },
+    {
+      "epoch": 0.725623785377622,
+      "grad_norm": 0.6009805798530579,
+      "learning_rate": 1.847123217482078e-05,
+      "loss": 0.9352,
+      "step": 8121
+    },
+    {
+      "epoch": 0.7257131369088837,
+      "grad_norm": 0.4387056827545166,
+      "learning_rate": 1.846000236635264e-05,
+      "loss": 0.9247,
+      "step": 8122
+    },
+    {
+      "epoch": 0.7258024884401456,
+      "grad_norm": 0.5181190967559814,
+      "learning_rate": 1.8448775199695502e-05,
+      "loss": 0.8485,
+      "step": 8123
+    },
+    {
+      "epoch": 0.7258918399714075,
+      "grad_norm": 0.4301597774028778,
+      "learning_rate": 1.843755067578975e-05,
+      "loss": 0.9481,
+      "step": 8124
+    },
+    {
+      "epoch": 0.7259811915026694,
+      "grad_norm": 0.49605125188827515,
+      "learning_rate": 1.842632879557556e-05,
+      "loss": 0.9434,
+      "step": 8125
+    },
+    {
+      "epoch": 0.7260705430339313,
+      "grad_norm": 0.43525439500808716,
+      "learning_rate": 1.8415109559992882e-05,
+      "loss": 0.9441,
+      "step": 8126
+    },
+    {
+      "epoch": 0.7261598945651931,
+      "grad_norm": 0.5193489789962769,
+      "learning_rate": 1.840389296998145e-05,
+      "loss": 0.9361,
+      "step": 8127
+    },
+    {
+      "epoch": 0.726249246096455,
+      "grad_norm": 0.45760974287986755,
+      "learning_rate": 1.8392679026480774e-05,
+      "loss": 0.9345,
+      "step": 8128
+    },
+    {
+      "epoch": 0.7263385976277168,
+      "grad_norm": 0.6482862234115601,
+      "learning_rate": 1.8381467730430134e-05,
+      "loss": 0.8445,
+      "step": 8129
+    },
+    {
+      "epoch": 0.7264279491589787,
+      "grad_norm": 0.45973724126815796,
+      "learning_rate": 1.837025908276861e-05,
+      "loss": 0.896,
+      "step": 8130
+    },
+    {
+      "epoch": 0.7265173006902406,
+      "grad_norm": 0.4498787522315979,
+      "learning_rate": 1.8359053084435046e-05,
+      "loss": 0.9654,
+      "step": 8131
+    },
+    {
+      "epoch": 0.7266066522215024,
+      "grad_norm": 0.4632214903831482,
+      "learning_rate": 1.834784973636804e-05,
+      "loss": 0.9408,
+      "step": 8132
+    },
+    {
+      "epoch": 0.7266960037527643,
+      "grad_norm": 0.4368852972984314,
+      "learning_rate": 1.8336649039506004e-05,
+      "loss": 0.9919,
+      "step": 8133
+    },
+    {
+      "epoch": 0.7267853552840262,
+      "grad_norm": 0.4225790798664093,
+      "learning_rate": 1.8325450994787124e-05,
+      "loss": 0.9568,
+      "step": 8134
+    },
+    {
+      "epoch": 0.7268747068152881,
+      "grad_norm": 0.40643781423568726,
+      "learning_rate": 1.8314255603149345e-05,
+      "loss": 0.9723,
+      "step": 8135
+    },
+    {
+      "epoch": 0.7269640583465499,
+      "grad_norm": 0.47153598070144653,
+      "learning_rate": 1.8303062865530406e-05,
+      "loss": 0.8922,
+      "step": 8136
+    },
+    {
+      "epoch": 0.7270534098778118,
+      "grad_norm": 0.45243874192237854,
+      "learning_rate": 1.8291872782867813e-05,
+      "loss": 1.0081,
+      "step": 8137
+    },
+    {
+      "epoch": 0.7271427614090736,
+      "grad_norm": 0.4166507124900818,
+      "learning_rate": 1.8280685356098863e-05,
+      "loss": 0.8909,
+      "step": 8138
+    },
+    {
+      "epoch": 0.7272321129403355,
+      "grad_norm": 0.5738488435745239,
+      "learning_rate": 1.826950058616062e-05,
+      "loss": 0.9549,
+      "step": 8139
+    },
+    {
+      "epoch": 0.7273214644715974,
+      "grad_norm": 0.5272175073623657,
+      "learning_rate": 1.825831847398992e-05,
+      "loss": 0.9165,
+      "step": 8140
+    },
+    {
+      "epoch": 0.7274108160028593,
+      "grad_norm": 0.38618120551109314,
+      "learning_rate": 1.8247139020523412e-05,
+      "loss": 0.9462,
+      "step": 8141
+    },
+    {
+      "epoch": 0.7275001675341212,
+      "grad_norm": 0.459420382976532,
+      "learning_rate": 1.8235962226697457e-05,
+      "loss": 0.9248,
+      "step": 8142
+    },
+    {
+      "epoch": 0.7275895190653829,
+      "grad_norm": 0.5098559260368347,
+      "learning_rate": 1.822478809344824e-05,
+      "loss": 0.9542,
+      "step": 8143
+    },
+    {
+      "epoch": 0.7276788705966448,
+      "grad_norm": 0.5070344805717468,
+      "learning_rate": 1.8213616621711722e-05,
+      "loss": 1.0229,
+      "step": 8144
+    },
+    {
+      "epoch": 0.7277682221279067,
+      "grad_norm": 0.434377521276474,
+      "learning_rate": 1.8202447812423634e-05,
+      "loss": 0.9772,
+      "step": 8145
+    },
+    {
+      "epoch": 0.7278575736591686,
+      "grad_norm": 0.5484136939048767,
+      "learning_rate": 1.8191281666519473e-05,
+      "loss": 0.9173,
+      "step": 8146
+    },
+    {
+      "epoch": 0.7279469251904305,
+      "grad_norm": 0.4595695734024048,
+      "learning_rate": 1.8180118184934548e-05,
+      "loss": 0.9066,
+      "step": 8147
+    },
+    {
+      "epoch": 0.7280362767216924,
+      "grad_norm": 0.5228673219680786,
+      "learning_rate": 1.8168957368603884e-05,
+      "loss": 0.8998,
+      "step": 8148
+    },
+    {
+      "epoch": 0.7281256282529542,
+      "grad_norm": 0.5042482018470764,
+      "learning_rate": 1.8157799218462335e-05,
+      "loss": 0.9067,
+      "step": 8149
+    },
+    {
+      "epoch": 0.728214979784216,
+      "grad_norm": 0.5893993973731995,
+      "learning_rate": 1.814664373544452e-05,
+      "loss": 0.8643,
+      "step": 8150
+    },
+    {
+      "epoch": 0.7283043313154779,
+      "grad_norm": 0.4281430244445801,
+      "learning_rate": 1.813549092048483e-05,
+      "loss": 0.8613,
+      "step": 8151
+    },
+    {
+      "epoch": 0.7283936828467398,
+      "grad_norm": 0.5029839873313904,
+      "learning_rate": 1.8124340774517418e-05,
+      "loss": 0.9724,
+      "step": 8152
+    },
+    {
+      "epoch": 0.7284830343780017,
+      "grad_norm": 0.49639925360679626,
+      "learning_rate": 1.8113193298476232e-05,
+      "loss": 0.9772,
+      "step": 8153
+    },
+    {
+      "epoch": 0.7285723859092635,
+      "grad_norm": 0.44179990887641907,
+      "learning_rate": 1.8102048493295006e-05,
+      "loss": 0.9583,
+      "step": 8154
+    },
+    {
+      "epoch": 0.7286617374405254,
+      "grad_norm": 0.45038869976997375,
+      "learning_rate": 1.8090906359907216e-05,
+      "loss": 0.993,
+      "step": 8155
+    },
+    {
+      "epoch": 0.7287510889717873,
+      "grad_norm": 0.5203961133956909,
+      "learning_rate": 1.8079766899246148e-05,
+      "loss": 0.8979,
+      "step": 8156
+    },
+    {
+      "epoch": 0.7288404405030491,
+      "grad_norm": 0.5539426803588867,
+      "learning_rate": 1.8068630112244843e-05,
+      "loss": 0.999,
+      "step": 8157
+    },
+    {
+      "epoch": 0.728929792034311,
+      "grad_norm": 0.48159733414649963,
+      "learning_rate": 1.8057495999836137e-05,
+      "loss": 1.0195,
+      "step": 8158
+    },
+    {
+      "epoch": 0.7290191435655728,
+      "grad_norm": 0.45767495036125183,
+      "learning_rate": 1.8046364562952624e-05,
+      "loss": 0.9883,
+      "step": 8159
+    },
+    {
+      "epoch": 0.7291084950968347,
+      "grad_norm": 0.5231384038925171,
+      "learning_rate": 1.8035235802526674e-05,
+      "loss": 0.8853,
+      "step": 8160
+    },
+    {
+      "epoch": 0.7291978466280966,
+      "grad_norm": 0.46192824840545654,
+      "learning_rate": 1.802410971949045e-05,
+      "loss": 0.9542,
+      "step": 8161
+    },
+    {
+      "epoch": 0.7292871981593585,
+      "grad_norm": 0.40954524278640747,
+      "learning_rate": 1.801298631477589e-05,
+      "loss": 0.9083,
+      "step": 8162
+    },
+    {
+      "epoch": 0.7293765496906203,
+      "grad_norm": 0.5644052624702454,
+      "learning_rate": 1.8001865589314664e-05,
+      "loss": 0.8587,
+      "step": 8163
+    },
+    {
+      "epoch": 0.7294659012218822,
+      "grad_norm": 0.4511362612247467,
+      "learning_rate": 1.7990747544038277e-05,
+      "loss": 0.9583,
+      "step": 8164
+    },
+    {
+      "epoch": 0.729555252753144,
+      "grad_norm": 0.4605756103992462,
+      "learning_rate": 1.7979632179877974e-05,
+      "loss": 0.942,
+      "step": 8165
+    },
+    {
+      "epoch": 0.7296446042844059,
+      "grad_norm": 0.49011021852493286,
+      "learning_rate": 1.7968519497764784e-05,
+      "loss": 0.9492,
+      "step": 8166
+    },
+    {
+      "epoch": 0.7297339558156678,
+      "grad_norm": 0.48554643988609314,
+      "learning_rate": 1.7957409498629522e-05,
+      "loss": 0.9447,
+      "step": 8167
+    },
+    {
+      "epoch": 0.7298233073469297,
+      "grad_norm": 0.43161845207214355,
+      "learning_rate": 1.794630218340277e-05,
+      "loss": 0.9717,
+      "step": 8168
+    },
+    {
+      "epoch": 0.7299126588781916,
+      "grad_norm": 0.47446951270103455,
+      "learning_rate": 1.793519755301487e-05,
+      "loss": 0.8648,
+      "step": 8169
+    },
+    {
+      "epoch": 0.7300020104094533,
+      "grad_norm": 0.41278836131095886,
+      "learning_rate": 1.7924095608395963e-05,
+      "loss": 0.9728,
+      "step": 8170
+    },
+    {
+      "epoch": 0.7300913619407152,
+      "grad_norm": 0.4663233160972595,
+      "learning_rate": 1.7912996350475954e-05,
+      "loss": 0.9122,
+      "step": 8171
+    },
+    {
+      "epoch": 0.7301807134719771,
+      "grad_norm": 0.6070008873939514,
+      "learning_rate": 1.7901899780184537e-05,
+      "loss": 0.8931,
+      "step": 8172
+    },
+    {
+      "epoch": 0.730270065003239,
+      "grad_norm": 0.49870383739471436,
+      "learning_rate": 1.789080589845114e-05,
+      "loss": 0.8633,
+      "step": 8173
+    },
+    {
+      "epoch": 0.7303594165345009,
+      "grad_norm": 0.4976552426815033,
+      "learning_rate": 1.7879714706205008e-05,
+      "loss": 0.9654,
+      "step": 8174
+    },
+    {
+      "epoch": 0.7304487680657628,
+      "grad_norm": 0.43188008666038513,
+      "learning_rate": 1.786862620437515e-05,
+      "loss": 0.9809,
+      "step": 8175
+    },
+    {
+      "epoch": 0.7305381195970246,
+      "grad_norm": 0.4651055335998535,
+      "learning_rate": 1.7857540393890337e-05,
+      "loss": 0.9153,
+      "step": 8176
+    },
+    {
+      "epoch": 0.7306274711282864,
+      "grad_norm": 0.42811328172683716,
+      "learning_rate": 1.7846457275679136e-05,
+      "loss": 0.8451,
+      "step": 8177
+    },
+    {
+      "epoch": 0.7307168226595483,
+      "grad_norm": 0.4533650875091553,
+      "learning_rate": 1.7835376850669866e-05,
+      "loss": 0.9642,
+      "step": 8178
+    },
+    {
+      "epoch": 0.7308061741908102,
+      "grad_norm": 0.4616091549396515,
+      "learning_rate": 1.7824299119790637e-05,
+      "loss": 0.9361,
+      "step": 8179
+    },
+    {
+      "epoch": 0.7308955257220721,
+      "grad_norm": 0.4673561155796051,
+      "learning_rate": 1.7813224083969344e-05,
+      "loss": 0.8883,
+      "step": 8180
+    },
+    {
+      "epoch": 0.7309848772533339,
+      "grad_norm": 0.4838724136352539,
+      "learning_rate": 1.78021517441336e-05,
+      "loss": 0.8978,
+      "step": 8181
+    },
+    {
+      "epoch": 0.7310742287845958,
+      "grad_norm": 0.5045418739318848,
+      "learning_rate": 1.7791082101210853e-05,
+      "loss": 0.9195,
+      "step": 8182
+    },
+    {
+      "epoch": 0.7311635803158577,
+      "grad_norm": 0.40996941924095154,
+      "learning_rate": 1.7780015156128305e-05,
+      "loss": 0.9633,
+      "step": 8183
+    },
+    {
+      "epoch": 0.7312529318471195,
+      "grad_norm": 0.577971875667572,
+      "learning_rate": 1.776895090981294e-05,
+      "loss": 1.0122,
+      "step": 8184
+    },
+    {
+      "epoch": 0.7313422833783814,
+      "grad_norm": 0.4948059916496277,
+      "learning_rate": 1.7757889363191483e-05,
+      "loss": 0.8988,
+      "step": 8185
+    },
+    {
+      "epoch": 0.7314316349096432,
+      "grad_norm": 0.4757169187068939,
+      "learning_rate": 1.7746830517190467e-05,
+      "loss": 0.9198,
+      "step": 8186
+    },
+    {
+      "epoch": 0.7315209864409051,
+      "grad_norm": 0.5007458925247192,
+      "learning_rate": 1.7735774372736187e-05,
+      "loss": 0.9536,
+      "step": 8187
+    },
+    {
+      "epoch": 0.731610337972167,
+      "grad_norm": 0.5878704190254211,
+      "learning_rate": 1.772472093075471e-05,
+      "loss": 0.9794,
+      "step": 8188
+    },
+    {
+      "epoch": 0.7316996895034289,
+      "grad_norm": 0.4866276979446411,
+      "learning_rate": 1.7713670192171895e-05,
+      "loss": 0.9257,
+      "step": 8189
+    },
+    {
+      "epoch": 0.7317890410346908,
+      "grad_norm": 0.4084816575050354,
+      "learning_rate": 1.7702622157913344e-05,
+      "loss": 0.9616,
+      "step": 8190
+    },
+    {
+      "epoch": 0.7318783925659526,
+      "grad_norm": 0.5077846050262451,
+      "learning_rate": 1.7691576828904456e-05,
+      "loss": 0.9011,
+      "step": 8191
+    },
+    {
+      "epoch": 0.7319677440972144,
+      "grad_norm": 0.46192851662635803,
+      "learning_rate": 1.7680534206070405e-05,
+      "loss": 0.9729,
+      "step": 8192
+    },
+    {
+      "epoch": 0.7320570956284763,
+      "grad_norm": 0.42796364426612854,
+      "learning_rate": 1.76694942903361e-05,
+      "loss": 0.9862,
+      "step": 8193
+    },
+    {
+      "epoch": 0.7321464471597382,
+      "grad_norm": 0.555330753326416,
+      "learning_rate": 1.765845708262626e-05,
+      "loss": 0.9163,
+      "step": 8194
+    },
+    {
+      "epoch": 0.7322357986910001,
+      "grad_norm": 0.456575870513916,
+      "learning_rate": 1.764742258386538e-05,
+      "loss": 0.9215,
+      "step": 8195
+    },
+    {
+      "epoch": 0.732325150222262,
+      "grad_norm": 0.4735051989555359,
+      "learning_rate": 1.7636390794977713e-05,
+      "loss": 0.9277,
+      "step": 8196
+    },
+    {
+      "epoch": 0.7324145017535238,
+      "grad_norm": 0.4335395097732544,
+      "learning_rate": 1.762536171688729e-05,
+      "loss": 0.9389,
+      "step": 8197
+    },
+    {
+      "epoch": 0.7325038532847856,
+      "grad_norm": 0.5177332758903503,
+      "learning_rate": 1.7614335350517915e-05,
+      "loss": 0.9101,
+      "step": 8198
+    },
+    {
+      "epoch": 0.7325932048160475,
+      "grad_norm": 0.5252842903137207,
+      "learning_rate": 1.760331169679315e-05,
+      "loss": 0.885,
+      "step": 8199
+    },
+    {
+      "epoch": 0.7326825563473094,
+      "grad_norm": 0.48446813225746155,
+      "learning_rate": 1.7592290756636365e-05,
+      "loss": 0.9211,
+      "step": 8200
+    },
+    {
+      "epoch": 0.7327719078785713,
+      "grad_norm": 0.5630969405174255,
+      "learning_rate": 1.7581272530970667e-05,
+      "loss": 0.8899,
+      "step": 8201
+    },
+    {
+      "epoch": 0.7328612594098332,
+      "grad_norm": 0.5070595741271973,
+      "learning_rate": 1.7570257020718967e-05,
+      "loss": 0.9564,
+      "step": 8202
+    },
+    {
+      "epoch": 0.732950610941095,
+      "grad_norm": 0.4257377088069916,
+      "learning_rate": 1.75592442268039e-05,
+      "loss": 1.001,
+      "step": 8203
+    },
+    {
+      "epoch": 0.7330399624723569,
+      "grad_norm": 0.40094104409217834,
+      "learning_rate": 1.7548234150147925e-05,
+      "loss": 0.9957,
+      "step": 8204
+    },
+    {
+      "epoch": 0.7331293140036187,
+      "grad_norm": 0.44128090143203735,
+      "learning_rate": 1.7537226791673256e-05,
+      "loss": 0.9949,
+      "step": 8205
+    },
+    {
+      "epoch": 0.7332186655348806,
+      "grad_norm": 0.46732866764068604,
+      "learning_rate": 1.7526222152301862e-05,
+      "loss": 0.9562,
+      "step": 8206
+    },
+    {
+      "epoch": 0.7333080170661425,
+      "grad_norm": 0.41671401262283325,
+      "learning_rate": 1.7515220232955513e-05,
+      "loss": 0.9315,
+      "step": 8207
+    },
+    {
+      "epoch": 0.7333973685974043,
+      "grad_norm": 0.5439326167106628,
+      "learning_rate": 1.7504221034555734e-05,
+      "loss": 0.9628,
+      "step": 8208
+    },
+    {
+      "epoch": 0.7334867201286662,
+      "grad_norm": 0.4819839894771576,
+      "learning_rate": 1.7493224558023825e-05,
+      "loss": 0.9094,
+      "step": 8209
+    },
+    {
+      "epoch": 0.7335760716599281,
+      "grad_norm": 0.4984486699104309,
+      "learning_rate": 1.7482230804280852e-05,
+      "loss": 0.9014,
+      "step": 8210
+    },
+    {
+      "epoch": 0.73366542319119,
+      "grad_norm": 0.5585265159606934,
+      "learning_rate": 1.7471239774247667e-05,
+      "loss": 0.9843,
+      "step": 8211
+    },
+    {
+      "epoch": 0.7337547747224518,
+      "grad_norm": 0.4539826214313507,
+      "learning_rate": 1.7460251468844895e-05,
+      "loss": 0.9753,
+      "step": 8212
+    },
+    {
+      "epoch": 0.7338441262537136,
+      "grad_norm": 0.44110792875289917,
+      "learning_rate": 1.74492658889929e-05,
+      "loss": 0.9038,
+      "step": 8213
+    },
+    {
+      "epoch": 0.7339334777849755,
+      "grad_norm": 0.44374075531959534,
+      "learning_rate": 1.7438283035611846e-05,
+      "loss": 0.9506,
+      "step": 8214
+    },
+    {
+      "epoch": 0.7340228293162374,
+      "grad_norm": 0.43943607807159424,
+      "learning_rate": 1.7427302909621672e-05,
+      "loss": 0.9645,
+      "step": 8215
+    },
+    {
+      "epoch": 0.7341121808474993,
+      "grad_norm": 0.5023413300514221,
+      "learning_rate": 1.7416325511942085e-05,
+      "loss": 0.8913,
+      "step": 8216
+    },
+    {
+      "epoch": 0.7342015323787612,
+      "grad_norm": 0.43534350395202637,
+      "learning_rate": 1.7405350843492567e-05,
+      "loss": 0.907,
+      "step": 8217
+    },
+    {
+      "epoch": 0.7342908839100231,
+      "grad_norm": 0.46004897356033325,
+      "learning_rate": 1.7394378905192334e-05,
+      "loss": 0.9859,
+      "step": 8218
+    },
+    {
+      "epoch": 0.7343802354412848,
+      "grad_norm": 0.49676865339279175,
+      "learning_rate": 1.738340969796042e-05,
+      "loss": 0.9193,
+      "step": 8219
+    },
+    {
+      "epoch": 0.7344695869725467,
+      "grad_norm": 0.47538769245147705,
+      "learning_rate": 1.7372443222715605e-05,
+      "loss": 0.9194,
+      "step": 8220
+    },
+    {
+      "epoch": 0.7345589385038086,
+      "grad_norm": 0.477346271276474,
+      "learning_rate": 1.736147948037646e-05,
+      "loss": 0.9213,
+      "step": 8221
+    },
+    {
+      "epoch": 0.7346482900350705,
+      "grad_norm": 0.3966813087463379,
+      "learning_rate": 1.7350518471861328e-05,
+      "loss": 0.9935,
+      "step": 8222
+    },
+    {
+      "epoch": 0.7347376415663324,
+      "grad_norm": 0.46678826212882996,
+      "learning_rate": 1.7339560198088273e-05,
+      "loss": 0.923,
+      "step": 8223
+    },
+    {
+      "epoch": 0.7348269930975942,
+      "grad_norm": 0.42243048548698425,
+      "learning_rate": 1.7328604659975184e-05,
+      "loss": 1.0018,
+      "step": 8224
+    },
+    {
+      "epoch": 0.734916344628856,
+      "grad_norm": 0.4616301655769348,
+      "learning_rate": 1.7317651858439714e-05,
+      "loss": 0.9319,
+      "step": 8225
+    },
+    {
+      "epoch": 0.7350056961601179,
+      "grad_norm": 0.644029438495636,
+      "learning_rate": 1.7306701794399266e-05,
+      "loss": 0.9191,
+      "step": 8226
+    },
+    {
+      "epoch": 0.7350950476913798,
+      "grad_norm": 0.4394361078739166,
+      "learning_rate": 1.7295754468771024e-05,
+      "loss": 0.9644,
+      "step": 8227
+    },
+    {
+      "epoch": 0.7351843992226417,
+      "grad_norm": 0.4633764922618866,
+      "learning_rate": 1.7284809882471954e-05,
+      "loss": 0.9198,
+      "step": 8228
+    },
+    {
+      "epoch": 0.7352737507539036,
+      "grad_norm": 0.49035075306892395,
+      "learning_rate": 1.727386803641877e-05,
+      "loss": 0.9324,
+      "step": 8229
+    },
+    {
+      "epoch": 0.7353631022851654,
+      "grad_norm": 0.45196330547332764,
+      "learning_rate": 1.7262928931527977e-05,
+      "loss": 0.9823,
+      "step": 8230
+    },
+    {
+      "epoch": 0.7354524538164273,
+      "grad_norm": 0.459506630897522,
+      "learning_rate": 1.7251992568715842e-05,
+      "loss": 0.9723,
+      "step": 8231
+    },
+    {
+      "epoch": 0.7355418053476891,
+      "grad_norm": 0.5117012858390808,
+      "learning_rate": 1.724105894889841e-05,
+      "loss": 0.9468,
+      "step": 8232
+    },
+    {
+      "epoch": 0.735631156878951,
+      "grad_norm": 0.43561848998069763,
+      "learning_rate": 1.7230128072991458e-05,
+      "loss": 0.933,
+      "step": 8233
+    },
+    {
+      "epoch": 0.7357205084102129,
+      "grad_norm": 0.5399268865585327,
+      "learning_rate": 1.721919994191058e-05,
+      "loss": 0.9901,
+      "step": 8234
+    },
+    {
+      "epoch": 0.7358098599414747,
+      "grad_norm": 0.4303232729434967,
+      "learning_rate": 1.720827455657113e-05,
+      "loss": 1.0334,
+      "step": 8235
+    },
+    {
+      "epoch": 0.7358992114727366,
+      "grad_norm": 0.5467885732650757,
+      "learning_rate": 1.719735191788822e-05,
+      "loss": 0.9286,
+      "step": 8236
+    },
+    {
+      "epoch": 0.7359885630039985,
+      "grad_norm": 0.43079063296318054,
+      "learning_rate": 1.7186432026776734e-05,
+      "loss": 0.8944,
+      "step": 8237
+    },
+    {
+      "epoch": 0.7360779145352604,
+      "grad_norm": 0.5494388341903687,
+      "learning_rate": 1.717551488415134e-05,
+      "loss": 0.8537,
+      "step": 8238
+    },
+    {
+      "epoch": 0.7361672660665222,
+      "grad_norm": 0.5178488492965698,
+      "learning_rate": 1.7164600490926454e-05,
+      "loss": 0.8705,
+      "step": 8239
+    },
+    {
+      "epoch": 0.736256617597784,
+      "grad_norm": 0.4486089050769806,
+      "learning_rate": 1.7153688848016277e-05,
+      "loss": 1.0209,
+      "step": 8240
+    },
+    {
+      "epoch": 0.7363459691290459,
+      "grad_norm": 0.4773981273174286,
+      "learning_rate": 1.7142779956334777e-05,
+      "loss": 0.9607,
+      "step": 8241
+    },
+    {
+      "epoch": 0.7364353206603078,
+      "grad_norm": 0.5104852318763733,
+      "learning_rate": 1.7131873816795684e-05,
+      "loss": 0.958,
+      "step": 8242
+    },
+    {
+      "epoch": 0.7365246721915697,
+      "grad_norm": 0.461008220911026,
+      "learning_rate": 1.7120970430312526e-05,
+      "loss": 0.9717,
+      "step": 8243
+    },
+    {
+      "epoch": 0.7366140237228316,
+      "grad_norm": 0.44321146607398987,
+      "learning_rate": 1.711006979779855e-05,
+      "loss": 0.9139,
+      "step": 8244
+    },
+    {
+      "epoch": 0.7367033752540935,
+      "grad_norm": 0.49801915884017944,
+      "learning_rate": 1.70991719201668e-05,
+      "loss": 0.8756,
+      "step": 8245
+    },
+    {
+      "epoch": 0.7367927267853552,
+      "grad_norm": 0.5032793283462524,
+      "learning_rate": 1.7088276798330106e-05,
+      "loss": 1.0369,
+      "step": 8246
+    },
+    {
+      "epoch": 0.7368820783166171,
+      "grad_norm": 0.5026294589042664,
+      "learning_rate": 1.7077384433201045e-05,
+      "loss": 0.9135,
+      "step": 8247
+    },
+    {
+      "epoch": 0.736971429847879,
+      "grad_norm": 0.4539667069911957,
+      "learning_rate": 1.7066494825691966e-05,
+      "loss": 0.9861,
+      "step": 8248
+    },
+    {
+      "epoch": 0.7370607813791409,
+      "grad_norm": 0.4462776184082031,
+      "learning_rate": 1.7055607976714988e-05,
+      "loss": 0.9653,
+      "step": 8249
+    },
+    {
+      "epoch": 0.7371501329104028,
+      "grad_norm": 0.4230305552482605,
+      "learning_rate": 1.7044723887182017e-05,
+      "loss": 0.9694,
+      "step": 8250
+    },
+    {
+      "epoch": 0.7372394844416646,
+      "grad_norm": 0.4417788088321686,
+      "learning_rate": 1.7033842558004693e-05,
+      "loss": 0.9562,
+      "step": 8251
+    },
+    {
+      "epoch": 0.7373288359729265,
+      "grad_norm": 0.4916762113571167,
+      "learning_rate": 1.7022963990094442e-05,
+      "loss": 0.8921,
+      "step": 8252
+    },
+    {
+      "epoch": 0.7374181875041883,
+      "grad_norm": 0.44411367177963257,
+      "learning_rate": 1.7012088184362467e-05,
+      "loss": 0.961,
+      "step": 8253
+    },
+    {
+      "epoch": 0.7375075390354502,
+      "grad_norm": 0.4935406446456909,
+      "learning_rate": 1.700121514171975e-05,
+      "loss": 0.9968,
+      "step": 8254
+    },
+    {
+      "epoch": 0.7375968905667121,
+      "grad_norm": 0.5243939757347107,
+      "learning_rate": 1.699034486307699e-05,
+      "loss": 0.9191,
+      "step": 8255
+    },
+    {
+      "epoch": 0.737686242097974,
+      "grad_norm": 0.5002910494804382,
+      "learning_rate": 1.69794773493447e-05,
+      "loss": 0.9762,
+      "step": 8256
+    },
+    {
+      "epoch": 0.7377755936292358,
+      "grad_norm": 0.42303359508514404,
+      "learning_rate": 1.6968612601433164e-05,
+      "loss": 0.9327,
+      "step": 8257
+    },
+    {
+      "epoch": 0.7378649451604977,
+      "grad_norm": 0.47965705394744873,
+      "learning_rate": 1.695775062025241e-05,
+      "loss": 0.906,
+      "step": 8258
+    },
+    {
+      "epoch": 0.7379542966917596,
+      "grad_norm": 0.44880691170692444,
+      "learning_rate": 1.6946891406712245e-05,
+      "loss": 0.9013,
+      "step": 8259
+    },
+    {
+      "epoch": 0.7380436482230214,
+      "grad_norm": 0.48894819617271423,
+      "learning_rate": 1.6936034961722247e-05,
+      "loss": 0.9125,
+      "step": 8260
+    },
+    {
+      "epoch": 0.7381329997542833,
+      "grad_norm": 0.4461333751678467,
+      "learning_rate": 1.6925181286191755e-05,
+      "loss": 0.9559,
+      "step": 8261
+    },
+    {
+      "epoch": 0.7382223512855451,
+      "grad_norm": 0.4503779411315918,
+      "learning_rate": 1.6914330381029888e-05,
+      "loss": 0.9575,
+      "step": 8262
+    },
+    {
+      "epoch": 0.738311702816807,
+      "grad_norm": 0.4670964181423187,
+      "learning_rate": 1.690348224714553e-05,
+      "loss": 0.9779,
+      "step": 8263
+    },
+    {
+      "epoch": 0.7384010543480689,
+      "grad_norm": 0.6177948713302612,
+      "learning_rate": 1.68926368854473e-05,
+      "loss": 0.994,
+      "step": 8264
+    },
+    {
+      "epoch": 0.7384904058793308,
+      "grad_norm": 0.4183749854564667,
+      "learning_rate": 1.6881794296843633e-05,
+      "loss": 0.9373,
+      "step": 8265
+    },
+    {
+      "epoch": 0.7385797574105927,
+      "grad_norm": 0.42233920097351074,
+      "learning_rate": 1.6870954482242707e-05,
+      "loss": 0.979,
+      "step": 8266
+    },
+    {
+      "epoch": 0.7386691089418544,
+      "grad_norm": 0.45563188195228577,
+      "learning_rate": 1.6860117442552477e-05,
+      "loss": 0.9276,
+      "step": 8267
+    },
+    {
+      "epoch": 0.7387584604731163,
+      "grad_norm": 0.43523088097572327,
+      "learning_rate": 1.6849283178680653e-05,
+      "loss": 1.0035,
+      "step": 8268
+    },
+    {
+      "epoch": 0.7388478120043782,
+      "grad_norm": 0.6393901109695435,
+      "learning_rate": 1.6838451691534724e-05,
+      "loss": 0.9238,
+      "step": 8269
+    },
+    {
+      "epoch": 0.7389371635356401,
+      "grad_norm": 0.5803185105323792,
+      "learning_rate": 1.6827622982021947e-05,
+      "loss": 0.8818,
+      "step": 8270
+    },
+    {
+      "epoch": 0.739026515066902,
+      "grad_norm": 0.3935569226741791,
+      "learning_rate": 1.6816797051049334e-05,
+      "loss": 0.914,
+      "step": 8271
+    },
+    {
+      "epoch": 0.7391158665981639,
+      "grad_norm": 0.43459293246269226,
+      "learning_rate": 1.6805973899523675e-05,
+      "loss": 0.9829,
+      "step": 8272
+    },
+    {
+      "epoch": 0.7392052181294257,
+      "grad_norm": 0.46992024779319763,
+      "learning_rate": 1.679515352835154e-05,
+      "loss": 0.9613,
+      "step": 8273
+    },
+    {
+      "epoch": 0.7392945696606875,
+      "grad_norm": 0.4815920889377594,
+      "learning_rate": 1.678433593843922e-05,
+      "loss": 0.9408,
+      "step": 8274
+    },
+    {
+      "epoch": 0.7393839211919494,
+      "grad_norm": 0.5394062399864197,
+      "learning_rate": 1.6773521130692822e-05,
+      "loss": 0.8696,
+      "step": 8275
+    },
+    {
+      "epoch": 0.7394732727232113,
+      "grad_norm": 0.4539199769496918,
+      "learning_rate": 1.6762709106018194e-05,
+      "loss": 0.9607,
+      "step": 8276
+    },
+    {
+      "epoch": 0.7395626242544732,
+      "grad_norm": 0.3883989751338959,
+      "learning_rate": 1.6751899865320963e-05,
+      "loss": 0.957,
+      "step": 8277
+    },
+    {
+      "epoch": 0.739651975785735,
+      "grad_norm": 0.4848140776157379,
+      "learning_rate": 1.674109340950652e-05,
+      "loss": 0.9505,
+      "step": 8278
+    },
+    {
+      "epoch": 0.7397413273169969,
+      "grad_norm": 0.42425286769866943,
+      "learning_rate": 1.6730289739480015e-05,
+      "loss": 0.9207,
+      "step": 8279
+    },
+    {
+      "epoch": 0.7398306788482588,
+      "grad_norm": 0.5808948278427124,
+      "learning_rate": 1.6719488856146377e-05,
+      "loss": 0.9163,
+      "step": 8280
+    },
+    {
+      "epoch": 0.7399200303795206,
+      "grad_norm": 0.481012761592865,
+      "learning_rate": 1.670869076041029e-05,
+      "loss": 0.9699,
+      "step": 8281
+    },
+    {
+      "epoch": 0.7400093819107825,
+      "grad_norm": 0.5220696926116943,
+      "learning_rate": 1.669789545317621e-05,
+      "loss": 0.912,
+      "step": 8282
+    },
+    {
+      "epoch": 0.7400987334420444,
+      "grad_norm": 0.46939072012901306,
+      "learning_rate": 1.668710293534838e-05,
+      "loss": 1.0163,
+      "step": 8283
+    },
+    {
+      "epoch": 0.7401880849733062,
+      "grad_norm": 0.4594574570655823,
+      "learning_rate": 1.6676313207830752e-05,
+      "loss": 0.9666,
+      "step": 8284
+    },
+    {
+      "epoch": 0.7402774365045681,
+      "grad_norm": 0.36939460039138794,
+      "learning_rate": 1.66655262715271e-05,
+      "loss": 0.9669,
+      "step": 8285
+    },
+    {
+      "epoch": 0.74036678803583,
+      "grad_norm": 0.45765185356140137,
+      "learning_rate": 1.6654742127340938e-05,
+      "loss": 0.8751,
+      "step": 8286
+    },
+    {
+      "epoch": 0.7404561395670918,
+      "grad_norm": 0.5784175992012024,
+      "learning_rate": 1.6643960776175577e-05,
+      "loss": 0.8804,
+      "step": 8287
+    },
+    {
+      "epoch": 0.7405454910983537,
+      "grad_norm": 0.4556187689304352,
+      "learning_rate": 1.6633182218934035e-05,
+      "loss": 0.9594,
+      "step": 8288
+    },
+    {
+      "epoch": 0.7406348426296155,
+      "grad_norm": 0.6318187117576599,
+      "learning_rate": 1.6622406456519146e-05,
+      "loss": 0.9063,
+      "step": 8289
+    },
+    {
+      "epoch": 0.7407241941608774,
+      "grad_norm": 0.5511077046394348,
+      "learning_rate": 1.66116334898335e-05,
+      "loss": 0.8661,
+      "step": 8290
+    },
+    {
+      "epoch": 0.7408135456921393,
+      "grad_norm": 0.5508573651313782,
+      "learning_rate": 1.6600863319779435e-05,
+      "loss": 0.8953,
+      "step": 8291
+    },
+    {
+      "epoch": 0.7409028972234012,
+      "grad_norm": 0.5418042540550232,
+      "learning_rate": 1.659009594725908e-05,
+      "loss": 0.8662,
+      "step": 8292
+    },
+    {
+      "epoch": 0.7409922487546631,
+      "grad_norm": 0.45847129821777344,
+      "learning_rate": 1.6579331373174335e-05,
+      "loss": 0.9173,
+      "step": 8293
+    },
+    {
+      "epoch": 0.7410816002859248,
+      "grad_norm": 0.45850539207458496,
+      "learning_rate": 1.6568569598426803e-05,
+      "loss": 0.8913,
+      "step": 8294
+    },
+    {
+      "epoch": 0.7411709518171867,
+      "grad_norm": 0.46910685300827026,
+      "learning_rate": 1.6557810623917923e-05,
+      "loss": 0.9585,
+      "step": 8295
+    },
+    {
+      "epoch": 0.7412603033484486,
+      "grad_norm": 0.5492746829986572,
+      "learning_rate": 1.654705445054887e-05,
+      "loss": 0.9454,
+      "step": 8296
+    },
+    {
+      "epoch": 0.7413496548797105,
+      "grad_norm": 0.42353761196136475,
+      "learning_rate": 1.6536301079220595e-05,
+      "loss": 0.9887,
+      "step": 8297
+    },
+    {
+      "epoch": 0.7414390064109724,
+      "grad_norm": 0.38787001371383667,
+      "learning_rate": 1.65255505108338e-05,
+      "loss": 0.9821,
+      "step": 8298
+    },
+    {
+      "epoch": 0.7415283579422343,
+      "grad_norm": 0.45528730750083923,
+      "learning_rate": 1.6514802746288955e-05,
+      "loss": 0.9411,
+      "step": 8299
+    },
+    {
+      "epoch": 0.7416177094734961,
+      "grad_norm": 0.4377180337905884,
+      "learning_rate": 1.6504057786486316e-05,
+      "loss": 0.9134,
+      "step": 8300
+    },
+    {
+      "epoch": 0.7417070610047579,
+      "grad_norm": 0.4238099753856659,
+      "learning_rate": 1.649331563232587e-05,
+      "loss": 0.9492,
+      "step": 8301
+    },
+    {
+      "epoch": 0.7417964125360198,
+      "grad_norm": 0.6428069472312927,
+      "learning_rate": 1.6482576284707402e-05,
+      "loss": 0.8924,
+      "step": 8302
+    },
+    {
+      "epoch": 0.7418857640672817,
+      "grad_norm": 0.5138502717018127,
+      "learning_rate": 1.6471839744530455e-05,
+      "loss": 1.0355,
+      "step": 8303
+    },
+    {
+      "epoch": 0.7419751155985436,
+      "grad_norm": 0.4536551833152771,
+      "learning_rate": 1.6461106012694293e-05,
+      "loss": 0.9953,
+      "step": 8304
+    },
+    {
+      "epoch": 0.7420644671298054,
+      "grad_norm": 0.45389628410339355,
+      "learning_rate": 1.6450375090098003e-05,
+      "loss": 0.9809,
+      "step": 8305
+    },
+    {
+      "epoch": 0.7421538186610673,
+      "grad_norm": 0.42163825035095215,
+      "learning_rate": 1.6439646977640417e-05,
+      "loss": 0.9368,
+      "step": 8306
+    },
+    {
+      "epoch": 0.7422431701923292,
+      "grad_norm": 0.48852425813674927,
+      "learning_rate": 1.6428921676220122e-05,
+      "loss": 0.9138,
+      "step": 8307
+    },
+    {
+      "epoch": 0.742332521723591,
+      "grad_norm": 0.4827854633331299,
+      "learning_rate": 1.6418199186735476e-05,
+      "loss": 0.9123,
+      "step": 8308
+    },
+    {
+      "epoch": 0.7424218732548529,
+      "grad_norm": 0.4420393705368042,
+      "learning_rate": 1.6407479510084607e-05,
+      "loss": 0.9481,
+      "step": 8309
+    },
+    {
+      "epoch": 0.7425112247861148,
+      "grad_norm": 0.47318053245544434,
+      "learning_rate": 1.6396762647165398e-05,
+      "loss": 0.9103,
+      "step": 8310
+    },
+    {
+      "epoch": 0.7426005763173766,
+      "grad_norm": 0.5290870070457458,
+      "learning_rate": 1.6386048598875502e-05,
+      "loss": 0.9708,
+      "step": 8311
+    },
+    {
+      "epoch": 0.7426899278486385,
+      "grad_norm": 0.52885901927948,
+      "learning_rate": 1.6375337366112336e-05,
+      "loss": 0.9811,
+      "step": 8312
+    },
+    {
+      "epoch": 0.7427792793799004,
+      "grad_norm": 0.49882233142852783,
+      "learning_rate": 1.6364628949773096e-05,
+      "loss": 0.9632,
+      "step": 8313
+    },
+    {
+      "epoch": 0.7428686309111623,
+      "grad_norm": 0.46031755208969116,
+      "learning_rate": 1.6353923350754692e-05,
+      "loss": 1.0104,
+      "step": 8314
+    },
+    {
+      "epoch": 0.7429579824424241,
+      "grad_norm": 0.49714195728302,
+      "learning_rate": 1.6343220569953848e-05,
+      "loss": 0.9287,
+      "step": 8315
+    },
+    {
+      "epoch": 0.7430473339736859,
+      "grad_norm": 0.5016975998878479,
+      "learning_rate": 1.6332520608267032e-05,
+      "loss": 0.9117,
+      "step": 8316
+    },
+    {
+      "epoch": 0.7431366855049478,
+      "grad_norm": 0.5067397952079773,
+      "learning_rate": 1.632182346659049e-05,
+      "loss": 0.9105,
+      "step": 8317
+    },
+    {
+      "epoch": 0.7432260370362097,
+      "grad_norm": 0.4985998272895813,
+      "learning_rate": 1.6311129145820218e-05,
+      "loss": 0.8446,
+      "step": 8318
+    },
+    {
+      "epoch": 0.7433153885674716,
+      "grad_norm": 0.4276972711086273,
+      "learning_rate": 1.6300437646851972e-05,
+      "loss": 0.9542,
+      "step": 8319
+    },
+    {
+      "epoch": 0.7434047400987335,
+      "grad_norm": 0.4463093876838684,
+      "learning_rate": 1.6289748970581307e-05,
+      "loss": 0.9895,
+      "step": 8320
+    },
+    {
+      "epoch": 0.7434940916299954,
+      "grad_norm": 0.4791722297668457,
+      "learning_rate": 1.627906311790347e-05,
+      "loss": 0.9321,
+      "step": 8321
+    },
+    {
+      "epoch": 0.7435834431612571,
+      "grad_norm": 0.4845844805240631,
+      "learning_rate": 1.6268380089713542e-05,
+      "loss": 0.9468,
+      "step": 8322
+    },
+    {
+      "epoch": 0.743672794692519,
+      "grad_norm": 0.49456390738487244,
+      "learning_rate": 1.625769988690633e-05,
+      "loss": 0.9305,
+      "step": 8323
+    },
+    {
+      "epoch": 0.7437621462237809,
+      "grad_norm": 0.5137072205543518,
+      "learning_rate": 1.6247022510376435e-05,
+      "loss": 0.9164,
+      "step": 8324
+    },
+    {
+      "epoch": 0.7438514977550428,
+      "grad_norm": 0.5762333869934082,
+      "learning_rate": 1.6236347961018172e-05,
+      "loss": 0.9437,
+      "step": 8325
+    },
+    {
+      "epoch": 0.7439408492863047,
+      "grad_norm": 0.43072283267974854,
+      "learning_rate": 1.6225676239725663e-05,
+      "loss": 0.927,
+      "step": 8326
+    },
+    {
+      "epoch": 0.7440302008175665,
+      "grad_norm": 0.42828169465065,
+      "learning_rate": 1.6215007347392775e-05,
+      "loss": 0.9505,
+      "step": 8327
+    },
+    {
+      "epoch": 0.7441195523488284,
+      "grad_norm": 0.44435444474220276,
+      "learning_rate": 1.6204341284913144e-05,
+      "loss": 0.9176,
+      "step": 8328
+    },
+    {
+      "epoch": 0.7442089038800902,
+      "grad_norm": 0.38234448432922363,
+      "learning_rate": 1.6193678053180168e-05,
+      "loss": 1.0191,
+      "step": 8329
+    },
+    {
+      "epoch": 0.7442982554113521,
+      "grad_norm": 0.5488905310630798,
+      "learning_rate": 1.6183017653087e-05,
+      "loss": 0.9291,
+      "step": 8330
+    },
+    {
+      "epoch": 0.744387606942614,
+      "grad_norm": 0.4858629107475281,
+      "learning_rate": 1.6172360085526565e-05,
+      "loss": 0.9887,
+      "step": 8331
+    },
+    {
+      "epoch": 0.7444769584738758,
+      "grad_norm": 0.517011821269989,
+      "learning_rate": 1.616170535139156e-05,
+      "loss": 0.9611,
+      "step": 8332
+    },
+    {
+      "epoch": 0.7445663100051377,
+      "grad_norm": 0.5602929592132568,
+      "learning_rate": 1.6151053451574416e-05,
+      "loss": 0.9426,
+      "step": 8333
+    },
+    {
+      "epoch": 0.7446556615363996,
+      "grad_norm": 0.4301270842552185,
+      "learning_rate": 1.614040438696736e-05,
+      "loss": 1.0005,
+      "step": 8334
+    },
+    {
+      "epoch": 0.7447450130676615,
+      "grad_norm": 0.42185020446777344,
+      "learning_rate": 1.612975815846235e-05,
+      "loss": 0.9301,
+      "step": 8335
+    },
+    {
+      "epoch": 0.7448343645989233,
+      "grad_norm": 0.5149445533752441,
+      "learning_rate": 1.6119114766951116e-05,
+      "loss": 0.9543,
+      "step": 8336
+    },
+    {
+      "epoch": 0.7449237161301852,
+      "grad_norm": 0.48916956782341003,
+      "learning_rate": 1.6108474213325165e-05,
+      "loss": 0.8676,
+      "step": 8337
+    },
+    {
+      "epoch": 0.745013067661447,
+      "grad_norm": 0.4371699392795563,
+      "learning_rate": 1.609783649847576e-05,
+      "loss": 0.9421,
+      "step": 8338
+    },
+    {
+      "epoch": 0.7451024191927089,
+      "grad_norm": 0.6675881147384644,
+      "learning_rate": 1.6087201623293917e-05,
+      "loss": 0.8675,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7451917707239708,
+      "grad_norm": 0.5161626935005188,
+      "learning_rate": 1.6076569588670425e-05,
+      "loss": 0.9024,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7452811222552327,
+      "grad_norm": 0.4626888334751129,
+      "learning_rate": 1.6065940395495825e-05,
+      "loss": 0.9111,
+      "step": 8341
+    },
+    {
+      "epoch": 0.7453704737864946,
+      "grad_norm": 0.45520034432411194,
+      "learning_rate": 1.605531404466043e-05,
+      "loss": 0.9532,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7454598253177563,
+      "grad_norm": 0.5057812929153442,
+      "learning_rate": 1.6044690537054306e-05,
+      "loss": 0.8732,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7455491768490182,
+      "grad_norm": 0.48415088653564453,
+      "learning_rate": 1.6034069873567303e-05,
+      "loss": 0.8654,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7456385283802801,
+      "grad_norm": 0.520176351070404,
+      "learning_rate": 1.6023452055088982e-05,
+      "loss": 0.9233,
+      "step": 8345
+    },
+    {
+      "epoch": 0.745727879911542,
+      "grad_norm": 0.4946928024291992,
+      "learning_rate": 1.601283708250872e-05,
+      "loss": 0.9015,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7458172314428039,
+      "grad_norm": 0.4103570580482483,
+      "learning_rate": 1.600222495671563e-05,
+      "loss": 0.9471,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7459065829740658,
+      "grad_norm": 0.41291913390159607,
+      "learning_rate": 1.599161567859858e-05,
+      "loss": 0.9129,
+      "step": 8348
+    },
+    {
+      "epoch": 0.7459959345053276,
+      "grad_norm": 0.4548855721950531,
+      "learning_rate": 1.598100924904623e-05,
+      "loss": 0.932,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7460852860365894,
+      "grad_norm": 0.43719008564949036,
+      "learning_rate": 1.597040566894697e-05,
+      "loss": 0.9768,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7461746375678513,
+      "grad_norm": 0.4432275593280792,
+      "learning_rate": 1.595980493918896e-05,
+      "loss": 0.9684,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7462639890991132,
+      "grad_norm": 0.5939651727676392,
+      "learning_rate": 1.5949207060660138e-05,
+      "loss": 0.9425,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7463533406303751,
+      "grad_norm": 0.45562806725502014,
+      "learning_rate": 1.5938612034248184e-05,
+      "loss": 0.9775,
+      "step": 8353
+    },
+    {
+      "epoch": 0.746442692161637,
+      "grad_norm": 0.4708113670349121,
+      "learning_rate": 1.5928019860840532e-05,
+      "loss": 0.975,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7465320436928988,
+      "grad_norm": 0.5317729115486145,
+      "learning_rate": 1.5917430541324398e-05,
+      "loss": 0.9226,
+      "step": 8355
+    },
+    {
+      "epoch": 0.7466213952241606,
+      "grad_norm": 0.5121520757675171,
+      "learning_rate": 1.5906844076586746e-05,
+      "loss": 0.9363,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7467107467554225,
+      "grad_norm": 0.629950225353241,
+      "learning_rate": 1.5896260467514336e-05,
+      "loss": 0.8352,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7468000982866844,
+      "grad_norm": 0.3894347846508026,
+      "learning_rate": 1.58856797149936e-05,
+      "loss": 0.9527,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7468894498179462,
+      "grad_norm": 0.5037431716918945,
+      "learning_rate": 1.5875101819910833e-05,
+      "loss": 0.9559,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7469788013492081,
+      "grad_norm": 0.46914321184158325,
+      "learning_rate": 1.5864526783152028e-05,
+      "loss": 0.9593,
+      "step": 8360
+    },
+    {
+      "epoch": 0.74706815288047,
+      "grad_norm": 0.42743295431137085,
+      "learning_rate": 1.5853954605602965e-05,
+      "loss": 0.9583,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7471575044117319,
+      "grad_norm": 0.5981160402297974,
+      "learning_rate": 1.584338528814917e-05,
+      "loss": 0.831,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7472468559429937,
+      "grad_norm": 0.45480284094810486,
+      "learning_rate": 1.5832818831675943e-05,
+      "loss": 0.9134,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7473362074742556,
+      "grad_norm": 0.4186408817768097,
+      "learning_rate": 1.5822255237068357e-05,
+      "loss": 0.9619,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7474255590055174,
+      "grad_norm": 0.5457665920257568,
+      "learning_rate": 1.5811694505211182e-05,
+      "loss": 0.9438,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7475149105367793,
+      "grad_norm": 0.485774427652359,
+      "learning_rate": 1.5801136636989012e-05,
+      "loss": 0.9763,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7476042620680412,
+      "grad_norm": 0.42799893021583557,
+      "learning_rate": 1.5790581633286184e-05,
+      "loss": 0.9938,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7476936135993031,
+      "grad_norm": 0.45776692032814026,
+      "learning_rate": 1.5780029494986794e-05,
+      "loss": 0.9387,
+      "step": 8368
+    },
+    {
+      "epoch": 0.747782965130565,
+      "grad_norm": 0.5079134702682495,
+      "learning_rate": 1.5769480222974685e-05,
+      "loss": 0.9021,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7478723166618267,
+      "grad_norm": 0.4531920850276947,
+      "learning_rate": 1.575893381813348e-05,
+      "loss": 0.9242,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7479616681930886,
+      "grad_norm": 0.5917310118675232,
+      "learning_rate": 1.5748390281346553e-05,
+      "loss": 0.8605,
+      "step": 8371
+    },
+    {
+      "epoch": 0.7480510197243505,
+      "grad_norm": 0.4818684756755829,
+      "learning_rate": 1.573784961349704e-05,
+      "loss": 0.9327,
+      "step": 8372
+    },
+    {
+      "epoch": 0.7481403712556124,
+      "grad_norm": 0.4459517300128937,
+      "learning_rate": 1.5727311815467825e-05,
+      "loss": 0.9717,
+      "step": 8373
+    },
+    {
+      "epoch": 0.7482297227868743,
+      "grad_norm": 0.4536815583705902,
+      "learning_rate": 1.5716776888141583e-05,
+      "loss": 0.9551,
+      "step": 8374
+    },
+    {
+      "epoch": 0.7483190743181362,
+      "grad_norm": 0.7309669256210327,
+      "learning_rate": 1.5706244832400696e-05,
+      "loss": 0.8589,
+      "step": 8375
+    },
+    {
+      "epoch": 0.748408425849398,
+      "grad_norm": 0.45956024527549744,
+      "learning_rate": 1.5695715649127345e-05,
+      "loss": 0.9848,
+      "step": 8376
+    },
+    {
+      "epoch": 0.7484977773806598,
+      "grad_norm": 0.450041264295578,
+      "learning_rate": 1.568518933920347e-05,
+      "loss": 0.938,
+      "step": 8377
+    },
+    {
+      "epoch": 0.7485871289119217,
+      "grad_norm": 0.45489928126335144,
+      "learning_rate": 1.5674665903510755e-05,
+      "loss": 1.0276,
+      "step": 8378
+    },
+    {
+      "epoch": 0.7486764804431836,
+      "grad_norm": 0.42867419123649597,
+      "learning_rate": 1.566414534293065e-05,
+      "loss": 0.9341,
+      "step": 8379
+    },
+    {
+      "epoch": 0.7487658319744455,
+      "grad_norm": 0.44365933537483215,
+      "learning_rate": 1.5653627658344374e-05,
+      "loss": 0.943,
+      "step": 8380
+    },
+    {
+      "epoch": 0.7488551835057073,
+      "grad_norm": 0.4565197229385376,
+      "learning_rate": 1.5643112850632884e-05,
+      "loss": 0.8627,
+      "step": 8381
+    },
+    {
+      "epoch": 0.7489445350369692,
+      "grad_norm": 0.4930249750614166,
+      "learning_rate": 1.563260092067691e-05,
+      "loss": 0.9596,
+      "step": 8382
+    },
+    {
+      "epoch": 0.7490338865682311,
+      "grad_norm": 0.5552300214767456,
+      "learning_rate": 1.5622091869356937e-05,
+      "loss": 0.9906,
+      "step": 8383
+    },
+    {
+      "epoch": 0.7491232380994929,
+      "grad_norm": 0.44053712487220764,
+      "learning_rate": 1.5611585697553232e-05,
+      "loss": 0.9295,
+      "step": 8384
+    },
+    {
+      "epoch": 0.7492125896307548,
+      "grad_norm": 0.4973897635936737,
+      "learning_rate": 1.5601082406145762e-05,
+      "loss": 0.9121,
+      "step": 8385
+    },
+    {
+      "epoch": 0.7493019411620166,
+      "grad_norm": 0.4836944043636322,
+      "learning_rate": 1.5590581996014304e-05,
+      "loss": 0.8984,
+      "step": 8386
+    },
+    {
+      "epoch": 0.7493912926932785,
+      "grad_norm": 0.4505099654197693,
+      "learning_rate": 1.5580084468038382e-05,
+      "loss": 0.9337,
+      "step": 8387
+    },
+    {
+      "epoch": 0.7494806442245404,
+      "grad_norm": 0.42860540747642517,
+      "learning_rate": 1.556958982309728e-05,
+      "loss": 0.9484,
+      "step": 8388
+    },
+    {
+      "epoch": 0.7495699957558023,
+      "grad_norm": 0.5451123714447021,
+      "learning_rate": 1.5559098062070028e-05,
+      "loss": 0.9509,
+      "step": 8389
+    },
+    {
+      "epoch": 0.7496593472870642,
+      "grad_norm": 0.5574720501899719,
+      "learning_rate": 1.5548609185835444e-05,
+      "loss": 0.8854,
+      "step": 8390
+    },
+    {
+      "epoch": 0.749748698818326,
+      "grad_norm": 0.42226511240005493,
+      "learning_rate": 1.5538123195272054e-05,
+      "loss": 0.9356,
+      "step": 8391
+    },
+    {
+      "epoch": 0.7498380503495878,
+      "grad_norm": 0.4949922561645508,
+      "learning_rate": 1.5527640091258177e-05,
+      "loss": 0.9313,
+      "step": 8392
+    },
+    {
+      "epoch": 0.7499274018808497,
+      "grad_norm": 0.49576741456985474,
+      "learning_rate": 1.5517159874671892e-05,
+      "loss": 0.8921,
+      "step": 8393
+    },
+    {
+      "epoch": 0.7500167534121116,
+      "grad_norm": 0.47223761677742004,
+      "learning_rate": 1.550668254639105e-05,
+      "loss": 0.9675,
+      "step": 8394
+    },
+    {
+      "epoch": 0.7501061049433735,
+      "grad_norm": 0.4839581847190857,
+      "learning_rate": 1.5496208107293197e-05,
+      "loss": 0.982,
+      "step": 8395
+    },
+    {
+      "epoch": 0.7501954564746354,
+      "grad_norm": 0.539761483669281,
+      "learning_rate": 1.5485736558255697e-05,
+      "loss": 0.9059,
+      "step": 8396
+    },
+    {
+      "epoch": 0.7502848080058973,
+      "grad_norm": 0.4613410532474518,
+      "learning_rate": 1.547526790015566e-05,
+      "loss": 0.9384,
+      "step": 8397
+    },
+    {
+      "epoch": 0.750374159537159,
+      "grad_norm": 0.47453439235687256,
+      "learning_rate": 1.5464802133869942e-05,
+      "loss": 0.9226,
+      "step": 8398
+    },
+    {
+      "epoch": 0.7504635110684209,
+      "grad_norm": 0.5073251128196716,
+      "learning_rate": 1.5454339260275165e-05,
+      "loss": 0.9383,
+      "step": 8399
+    },
+    {
+      "epoch": 0.7505528625996828,
+      "grad_norm": 0.6021450757980347,
+      "learning_rate": 1.5443879280247704e-05,
+      "loss": 0.9338,
+      "step": 8400
+    },
+    {
+      "epoch": 0.7506422141309447,
+      "grad_norm": 0.4408600330352783,
+      "learning_rate": 1.5433422194663693e-05,
+      "loss": 0.965,
+      "step": 8401
+    },
+    {
+      "epoch": 0.7507315656622066,
+      "grad_norm": 0.5570371747016907,
+      "learning_rate": 1.542296800439903e-05,
+      "loss": 0.9716,
+      "step": 8402
+    },
+    {
+      "epoch": 0.7508209171934684,
+      "grad_norm": 0.48043861985206604,
+      "learning_rate": 1.541251671032936e-05,
+      "loss": 0.9822,
+      "step": 8403
+    },
+    {
+      "epoch": 0.7509102687247303,
+      "grad_norm": 0.5600388646125793,
+      "learning_rate": 1.540206831333011e-05,
+      "loss": 0.9544,
+      "step": 8404
+    },
+    {
+      "epoch": 0.7509996202559921,
+      "grad_norm": 0.48118168115615845,
+      "learning_rate": 1.5391622814276408e-05,
+      "loss": 0.9213,
+      "step": 8405
+    },
+    {
+      "epoch": 0.751088971787254,
+      "grad_norm": 0.5037118196487427,
+      "learning_rate": 1.538118021404319e-05,
+      "loss": 0.8725,
+      "step": 8406
+    },
+    {
+      "epoch": 0.7511783233185159,
+      "grad_norm": 0.5096782445907593,
+      "learning_rate": 1.5370740513505143e-05,
+      "loss": 0.9062,
+      "step": 8407
+    },
+    {
+      "epoch": 0.7512676748497777,
+      "grad_norm": 0.450232595205307,
+      "learning_rate": 1.53603037135367e-05,
+      "loss": 0.9839,
+      "step": 8408
+    },
+    {
+      "epoch": 0.7513570263810396,
+      "grad_norm": 0.4577409625053406,
+      "learning_rate": 1.534986981501205e-05,
+      "loss": 0.946,
+      "step": 8409
+    },
+    {
+      "epoch": 0.7514463779123015,
+      "grad_norm": 0.4218578338623047,
+      "learning_rate": 1.5339438818805152e-05,
+      "loss": 0.9404,
+      "step": 8410
+    },
+    {
+      "epoch": 0.7515357294435634,
+      "grad_norm": 0.5897673964500427,
+      "learning_rate": 1.5329010725789704e-05,
+      "loss": 0.9004,
+      "step": 8411
+    },
+    {
+      "epoch": 0.7516250809748252,
+      "grad_norm": 0.4546933174133301,
+      "learning_rate": 1.531858553683918e-05,
+      "loss": 0.9354,
+      "step": 8412
+    },
+    {
+      "epoch": 0.751714432506087,
+      "grad_norm": 0.48495376110076904,
+      "learning_rate": 1.530816325282679e-05,
+      "loss": 1.0102,
+      "step": 8413
+    },
+    {
+      "epoch": 0.7518037840373489,
+      "grad_norm": 0.5461007952690125,
+      "learning_rate": 1.5297743874625515e-05,
+      "loss": 0.9229,
+      "step": 8414
+    },
+    {
+      "epoch": 0.7518931355686108,
+      "grad_norm": 0.42798981070518494,
+      "learning_rate": 1.5287327403108108e-05,
+      "loss": 0.9756,
+      "step": 8415
+    },
+    {
+      "epoch": 0.7519824870998727,
+      "grad_norm": 0.4769188463687897,
+      "learning_rate": 1.527691383914702e-05,
+      "loss": 0.9252,
+      "step": 8416
+    },
+    {
+      "epoch": 0.7520718386311346,
+      "grad_norm": 0.5491362810134888,
+      "learning_rate": 1.526650318361453e-05,
+      "loss": 0.912,
+      "step": 8417
+    },
+    {
+      "epoch": 0.7521611901623964,
+      "grad_norm": 0.4550434648990631,
+      "learning_rate": 1.5256095437382622e-05,
+      "loss": 0.9414,
+      "step": 8418
+    },
+    {
+      "epoch": 0.7522505416936582,
+      "grad_norm": 0.5997094511985779,
+      "learning_rate": 1.524569060132307e-05,
+      "loss": 0.9087,
+      "step": 8419
+    },
+    {
+      "epoch": 0.7523398932249201,
+      "grad_norm": 0.4672020971775055,
+      "learning_rate": 1.523528867630738e-05,
+      "loss": 0.8856,
+      "step": 8420
+    },
+    {
+      "epoch": 0.752429244756182,
+      "grad_norm": 0.5284818410873413,
+      "learning_rate": 1.5224889663206832e-05,
+      "loss": 0.7954,
+      "step": 8421
+    },
+    {
+      "epoch": 0.7525185962874439,
+      "grad_norm": 0.6233134269714355,
+      "learning_rate": 1.521449356289245e-05,
+      "loss": 0.9528,
+      "step": 8422
+    },
+    {
+      "epoch": 0.7526079478187058,
+      "grad_norm": 0.5407315492630005,
+      "learning_rate": 1.5204100376235036e-05,
+      "loss": 0.9614,
+      "step": 8423
+    },
+    {
+      "epoch": 0.7526972993499677,
+      "grad_norm": 0.5749227404594421,
+      "learning_rate": 1.5193710104105092e-05,
+      "loss": 0.8443,
+      "step": 8424
+    },
+    {
+      "epoch": 0.7527866508812294,
+      "grad_norm": 0.4009944796562195,
+      "learning_rate": 1.518332274737294e-05,
+      "loss": 0.9648,
+      "step": 8425
+    },
+    {
+      "epoch": 0.7528760024124913,
+      "grad_norm": 0.44395560026168823,
+      "learning_rate": 1.5172938306908623e-05,
+      "loss": 0.972,
+      "step": 8426
+    },
+    {
+      "epoch": 0.7529653539437532,
+      "grad_norm": 0.5940225720405579,
+      "learning_rate": 1.5162556783581971e-05,
+      "loss": 0.9004,
+      "step": 8427
+    },
+    {
+      "epoch": 0.7530547054750151,
+      "grad_norm": 0.4562346339225769,
+      "learning_rate": 1.5152178178262516e-05,
+      "loss": 0.9244,
+      "step": 8428
+    },
+    {
+      "epoch": 0.753144057006277,
+      "grad_norm": 0.483516663312912,
+      "learning_rate": 1.5141802491819584e-05,
+      "loss": 0.9087,
+      "step": 8429
+    },
+    {
+      "epoch": 0.7532334085375388,
+      "grad_norm": 0.668719470500946,
+      "learning_rate": 1.513142972512226e-05,
+      "loss": 0.7838,
+      "step": 8430
+    },
+    {
+      "epoch": 0.7533227600688007,
+      "grad_norm": 0.5353193879127502,
+      "learning_rate": 1.5121059879039367e-05,
+      "loss": 0.8938,
+      "step": 8431
+    },
+    {
+      "epoch": 0.7534121116000625,
+      "grad_norm": 0.5128493309020996,
+      "learning_rate": 1.5110692954439492e-05,
+      "loss": 0.9597,
+      "step": 8432
+    },
+    {
+      "epoch": 0.7535014631313244,
+      "grad_norm": 0.4552728831768036,
+      "learning_rate": 1.5100328952190973e-05,
+      "loss": 0.9204,
+      "step": 8433
+    },
+    {
+      "epoch": 0.7535908146625863,
+      "grad_norm": 0.5088723301887512,
+      "learning_rate": 1.5089967873161909e-05,
+      "loss": 0.9603,
+      "step": 8434
+    },
+    {
+      "epoch": 0.7536801661938481,
+      "grad_norm": 0.6546921133995056,
+      "learning_rate": 1.5079609718220166e-05,
+      "loss": 0.9088,
+      "step": 8435
+    },
+    {
+      "epoch": 0.75376951772511,
+      "grad_norm": 0.41693705320358276,
+      "learning_rate": 1.506925448823332e-05,
+      "loss": 0.9853,
+      "step": 8436
+    },
+    {
+      "epoch": 0.7538588692563719,
+      "grad_norm": 0.5113435387611389,
+      "learning_rate": 1.5058902184068741e-05,
+      "loss": 0.8789,
+      "step": 8437
+    },
+    {
+      "epoch": 0.7539482207876338,
+      "grad_norm": 0.4677006006240845,
+      "learning_rate": 1.5048552806593552e-05,
+      "loss": 0.9137,
+      "step": 8438
+    },
+    {
+      "epoch": 0.7540375723188956,
+      "grad_norm": 0.45884403586387634,
+      "learning_rate": 1.5038206356674623e-05,
+      "loss": 0.9903,
+      "step": 8439
+    },
+    {
+      "epoch": 0.7541269238501574,
+      "grad_norm": 0.49759143590927124,
+      "learning_rate": 1.5027862835178574e-05,
+      "loss": 0.905,
+      "step": 8440
+    },
+    {
+      "epoch": 0.7542162753814193,
+      "grad_norm": 0.4578782916069031,
+      "learning_rate": 1.5017522242971794e-05,
+      "loss": 0.9329,
+      "step": 8441
+    },
+    {
+      "epoch": 0.7543056269126812,
+      "grad_norm": 0.40164539217948914,
+      "learning_rate": 1.5007184580920408e-05,
+      "loss": 1.0047,
+      "step": 8442
+    },
+    {
+      "epoch": 0.7543949784439431,
+      "grad_norm": 0.4076503813266754,
+      "learning_rate": 1.499684984989031e-05,
+      "loss": 0.9041,
+      "step": 8443
+    },
+    {
+      "epoch": 0.754484329975205,
+      "grad_norm": 0.5915509462356567,
+      "learning_rate": 1.4986518050747145e-05,
+      "loss": 0.9639,
+      "step": 8444
+    },
+    {
+      "epoch": 0.7545736815064669,
+      "grad_norm": 0.3953361511230469,
+      "learning_rate": 1.4976189184356327e-05,
+      "loss": 0.9684,
+      "step": 8445
+    },
+    {
+      "epoch": 0.7546630330377286,
+      "grad_norm": 0.5043690800666809,
+      "learning_rate": 1.4965863251582974e-05,
+      "loss": 0.9368,
+      "step": 8446
+    },
+    {
+      "epoch": 0.7547523845689905,
+      "grad_norm": 0.4337480962276459,
+      "learning_rate": 1.495554025329201e-05,
+      "loss": 0.907,
+      "step": 8447
+    },
+    {
+      "epoch": 0.7548417361002524,
+      "grad_norm": 0.5515368580818176,
+      "learning_rate": 1.4945220190348102e-05,
+      "loss": 0.9356,
+      "step": 8448
+    },
+    {
+      "epoch": 0.7549310876315143,
+      "grad_norm": 0.5478143692016602,
+      "learning_rate": 1.4934903063615657e-05,
+      "loss": 0.967,
+      "step": 8449
+    },
+    {
+      "epoch": 0.7550204391627762,
+      "grad_norm": 0.5199536085128784,
+      "learning_rate": 1.4924588873958844e-05,
+      "loss": 0.8784,
+      "step": 8450
+    },
+    {
+      "epoch": 0.755109790694038,
+      "grad_norm": 0.667113184928894,
+      "learning_rate": 1.4914277622241597e-05,
+      "loss": 0.9662,
+      "step": 8451
+    },
+    {
+      "epoch": 0.7551991422252999,
+      "grad_norm": 0.4980166256427765,
+      "learning_rate": 1.4903969309327581e-05,
+      "loss": 0.9127,
+      "step": 8452
+    },
+    {
+      "epoch": 0.7552884937565617,
+      "grad_norm": 0.48468539118766785,
+      "learning_rate": 1.4893663936080232e-05,
+      "loss": 0.9079,
+      "step": 8453
+    },
+    {
+      "epoch": 0.7553778452878236,
+      "grad_norm": 0.43085187673568726,
+      "learning_rate": 1.4883361503362736e-05,
+      "loss": 1.0368,
+      "step": 8454
+    },
+    {
+      "epoch": 0.7554671968190855,
+      "grad_norm": 0.5685251355171204,
+      "learning_rate": 1.4873062012038047e-05,
+      "loss": 0.8219,
+      "step": 8455
+    },
+    {
+      "epoch": 0.7555565483503474,
+      "grad_norm": 0.5110030174255371,
+      "learning_rate": 1.4862765462968826e-05,
+      "loss": 0.9895,
+      "step": 8456
+    },
+    {
+      "epoch": 0.7556458998816092,
+      "grad_norm": 0.5514484643936157,
+      "learning_rate": 1.485247185701753e-05,
+      "loss": 0.9602,
+      "step": 8457
+    },
+    {
+      "epoch": 0.7557352514128711,
+      "grad_norm": 0.47170817852020264,
+      "learning_rate": 1.4842181195046361e-05,
+      "loss": 0.9424,
+      "step": 8458
+    },
+    {
+      "epoch": 0.755824602944133,
+      "grad_norm": 0.42950955033302307,
+      "learning_rate": 1.483189347791728e-05,
+      "loss": 0.9495,
+      "step": 8459
+    },
+    {
+      "epoch": 0.7559139544753948,
+      "grad_norm": 0.4732508659362793,
+      "learning_rate": 1.4821608706491996e-05,
+      "loss": 0.9515,
+      "step": 8460
+    },
+    {
+      "epoch": 0.7560033060066567,
+      "grad_norm": 0.47231292724609375,
+      "learning_rate": 1.4811326881631937e-05,
+      "loss": 0.9472,
+      "step": 8461
+    },
+    {
+      "epoch": 0.7560926575379185,
+      "grad_norm": 0.4015584886074066,
+      "learning_rate": 1.4801048004198342e-05,
+      "loss": 0.947,
+      "step": 8462
+    },
+    {
+      "epoch": 0.7561820090691804,
+      "grad_norm": 0.530529797077179,
+      "learning_rate": 1.4790772075052173e-05,
+      "loss": 0.9262,
+      "step": 8463
+    },
+    {
+      "epoch": 0.7562713606004423,
+      "grad_norm": 0.5107260942459106,
+      "learning_rate": 1.478049909505414e-05,
+      "loss": 0.9101,
+      "step": 8464
+    },
+    {
+      "epoch": 0.7563607121317042,
+      "grad_norm": 0.4164203703403473,
+      "learning_rate": 1.4770229065064738e-05,
+      "loss": 1.0175,
+      "step": 8465
+    },
+    {
+      "epoch": 0.7564500636629661,
+      "grad_norm": 0.4908786714076996,
+      "learning_rate": 1.475996198594416e-05,
+      "loss": 0.9977,
+      "step": 8466
+    },
+    {
+      "epoch": 0.7565394151942278,
+      "grad_norm": 0.5216073393821716,
+      "learning_rate": 1.4749697858552398e-05,
+      "loss": 0.9263,
+      "step": 8467
+    },
+    {
+      "epoch": 0.7566287667254897,
+      "grad_norm": 0.48185113072395325,
+      "learning_rate": 1.4739436683749181e-05,
+      "loss": 0.8621,
+      "step": 8468
+    },
+    {
+      "epoch": 0.7567181182567516,
+      "grad_norm": 0.49778178334236145,
+      "learning_rate": 1.472917846239399e-05,
+      "loss": 0.8806,
+      "step": 8469
+    },
+    {
+      "epoch": 0.7568074697880135,
+      "grad_norm": 0.5073883533477783,
+      "learning_rate": 1.4718923195346062e-05,
+      "loss": 0.9601,
+      "step": 8470
+    },
+    {
+      "epoch": 0.7568968213192754,
+      "grad_norm": 0.4837455153465271,
+      "learning_rate": 1.4708670883464393e-05,
+      "loss": 0.9467,
+      "step": 8471
+    },
+    {
+      "epoch": 0.7569861728505373,
+      "grad_norm": 0.38693663477897644,
+      "learning_rate": 1.469842152760771e-05,
+      "loss": 0.9904,
+      "step": 8472
+    },
+    {
+      "epoch": 0.7570755243817991,
+      "grad_norm": 0.46666041016578674,
+      "learning_rate": 1.4688175128634512e-05,
+      "loss": 0.9895,
+      "step": 8473
+    },
+    {
+      "epoch": 0.7571648759130609,
+      "grad_norm": 0.42904338240623474,
+      "learning_rate": 1.4677931687403046e-05,
+      "loss": 0.999,
+      "step": 8474
+    },
+    {
+      "epoch": 0.7572542274443228,
+      "grad_norm": 0.4167872369289398,
+      "learning_rate": 1.4667691204771322e-05,
+      "loss": 1.0134,
+      "step": 8475
+    },
+    {
+      "epoch": 0.7573435789755847,
+      "grad_norm": 0.43804195523262024,
+      "learning_rate": 1.4657453681597055e-05,
+      "loss": 0.9359,
+      "step": 8476
+    },
+    {
+      "epoch": 0.7574329305068466,
+      "grad_norm": 0.5330228805541992,
+      "learning_rate": 1.4647219118737771e-05,
+      "loss": 0.842,
+      "step": 8477
+    },
+    {
+      "epoch": 0.7575222820381085,
+      "grad_norm": 0.4981614649295807,
+      "learning_rate": 1.463698751705072e-05,
+      "loss": 0.9647,
+      "step": 8478
+    },
+    {
+      "epoch": 0.7576116335693703,
+      "grad_norm": 0.4153805077075958,
+      "learning_rate": 1.4626758877392904e-05,
+      "loss": 0.9969,
+      "step": 8479
+    },
+    {
+      "epoch": 0.7577009851006321,
+      "grad_norm": 0.4333755671977997,
+      "learning_rate": 1.461653320062109e-05,
+      "loss": 0.9647,
+      "step": 8480
+    },
+    {
+      "epoch": 0.757790336631894,
+      "grad_norm": 0.5431153178215027,
+      "learning_rate": 1.4606310487591773e-05,
+      "loss": 0.9157,
+      "step": 8481
+    },
+    {
+      "epoch": 0.7578796881631559,
+      "grad_norm": 0.4706348180770874,
+      "learning_rate": 1.4596090739161228e-05,
+      "loss": 0.8865,
+      "step": 8482
+    },
+    {
+      "epoch": 0.7579690396944178,
+      "grad_norm": 0.5356218814849854,
+      "learning_rate": 1.4585873956185458e-05,
+      "loss": 0.9127,
+      "step": 8483
+    },
+    {
+      "epoch": 0.7580583912256796,
+      "grad_norm": 0.45213526487350464,
+      "learning_rate": 1.4575660139520237e-05,
+      "loss": 0.9477,
+      "step": 8484
+    },
+    {
+      "epoch": 0.7581477427569415,
+      "grad_norm": 0.5066707134246826,
+      "learning_rate": 1.4565449290021088e-05,
+      "loss": 0.9715,
+      "step": 8485
+    },
+    {
+      "epoch": 0.7582370942882034,
+      "grad_norm": 0.4998951554298401,
+      "learning_rate": 1.4555241408543252e-05,
+      "loss": 0.9402,
+      "step": 8486
+    },
+    {
+      "epoch": 0.7583264458194652,
+      "grad_norm": 0.46853405237197876,
+      "learning_rate": 1.454503649594176e-05,
+      "loss": 0.9371,
+      "step": 8487
+    },
+    {
+      "epoch": 0.7584157973507271,
+      "grad_norm": 0.5197616219520569,
+      "learning_rate": 1.4534834553071386e-05,
+      "loss": 0.9598,
+      "step": 8488
+    },
+    {
+      "epoch": 0.7585051488819889,
+      "grad_norm": 0.4408440589904785,
+      "learning_rate": 1.452463558078665e-05,
+      "loss": 0.9645,
+      "step": 8489
+    },
+    {
+      "epoch": 0.7585945004132508,
+      "grad_norm": 0.6004998087882996,
+      "learning_rate": 1.4514439579941818e-05,
+      "loss": 0.9181,
+      "step": 8490
+    },
+    {
+      "epoch": 0.7586838519445127,
+      "grad_norm": 0.4586280286312103,
+      "learning_rate": 1.4504246551390927e-05,
+      "loss": 0.9704,
+      "step": 8491
+    },
+    {
+      "epoch": 0.7587732034757746,
+      "grad_norm": 0.4923453629016876,
+      "learning_rate": 1.4494056495987746e-05,
+      "loss": 0.9155,
+      "step": 8492
+    },
+    {
+      "epoch": 0.7588625550070365,
+      "grad_norm": 0.5735955238342285,
+      "learning_rate": 1.448386941458581e-05,
+      "loss": 0.9033,
+      "step": 8493
+    },
+    {
+      "epoch": 0.7589519065382982,
+      "grad_norm": 0.40873488783836365,
+      "learning_rate": 1.447368530803837e-05,
+      "loss": 0.9818,
+      "step": 8494
+    },
+    {
+      "epoch": 0.7590412580695601,
+      "grad_norm": 0.5386642217636108,
+      "learning_rate": 1.4463504177198473e-05,
+      "loss": 0.8663,
+      "step": 8495
+    },
+    {
+      "epoch": 0.759130609600822,
+      "grad_norm": 0.43717023730278015,
+      "learning_rate": 1.445332602291889e-05,
+      "loss": 0.9881,
+      "step": 8496
+    },
+    {
+      "epoch": 0.7592199611320839,
+      "grad_norm": 0.4573476314544678,
+      "learning_rate": 1.4443150846052167e-05,
+      "loss": 0.9784,
+      "step": 8497
+    },
+    {
+      "epoch": 0.7593093126633458,
+      "grad_norm": 0.46079909801483154,
+      "learning_rate": 1.4432978647450557e-05,
+      "loss": 0.9627,
+      "step": 8498
+    },
+    {
+      "epoch": 0.7593986641946077,
+      "grad_norm": 0.4422478973865509,
+      "learning_rate": 1.44228094279661e-05,
+      "loss": 0.9779,
+      "step": 8499
+    },
+    {
+      "epoch": 0.7594880157258695,
+      "grad_norm": 0.42531049251556396,
+      "learning_rate": 1.4412643188450581e-05,
+      "loss": 0.9694,
+      "step": 8500
+    },
+    {
+      "epoch": 0.7595773672571313,
+      "grad_norm": 0.5261896848678589,
+      "learning_rate": 1.440247992975553e-05,
+      "loss": 0.8767,
+      "step": 8501
+    },
+    {
+      "epoch": 0.7596667187883932,
+      "grad_norm": 0.4981927275657654,
+      "learning_rate": 1.4392319652732222e-05,
+      "loss": 0.9391,
+      "step": 8502
+    },
+    {
+      "epoch": 0.7597560703196551,
+      "grad_norm": 0.42882490158081055,
+      "learning_rate": 1.43821623582317e-05,
+      "loss": 0.9519,
+      "step": 8503
+    },
+    {
+      "epoch": 0.759845421850917,
+      "grad_norm": 0.558502197265625,
+      "learning_rate": 1.4372008047104735e-05,
+      "loss": 0.9239,
+      "step": 8504
+    },
+    {
+      "epoch": 0.7599347733821789,
+      "grad_norm": 0.4893379509449005,
+      "learning_rate": 1.4361856720201866e-05,
+      "loss": 0.952,
+      "step": 8505
+    },
+    {
+      "epoch": 0.7600241249134407,
+      "grad_norm": 0.5038098096847534,
+      "learning_rate": 1.4351708378373386e-05,
+      "loss": 0.9585,
+      "step": 8506
+    },
+    {
+      "epoch": 0.7601134764447026,
+      "grad_norm": 0.46212783455848694,
+      "learning_rate": 1.4341563022469296e-05,
+      "loss": 0.9344,
+      "step": 8507
+    },
+    {
+      "epoch": 0.7602028279759644,
+      "grad_norm": 0.43376246094703674,
+      "learning_rate": 1.4331420653339395e-05,
+      "loss": 0.9815,
+      "step": 8508
+    },
+    {
+      "epoch": 0.7602921795072263,
+      "grad_norm": 0.5722975730895996,
+      "learning_rate": 1.4321281271833214e-05,
+      "loss": 0.9366,
+      "step": 8509
+    },
+    {
+      "epoch": 0.7603815310384882,
+      "grad_norm": 0.5890293717384338,
+      "learning_rate": 1.4311144878800037e-05,
+      "loss": 0.842,
+      "step": 8510
+    },
+    {
+      "epoch": 0.76047088256975,
+      "grad_norm": 0.5008702278137207,
+      "learning_rate": 1.4301011475088889e-05,
+      "loss": 0.9603,
+      "step": 8511
+    },
+    {
+      "epoch": 0.7605602341010119,
+      "grad_norm": 0.4864904284477234,
+      "learning_rate": 1.4290881061548555e-05,
+      "loss": 0.9232,
+      "step": 8512
+    },
+    {
+      "epoch": 0.7606495856322738,
+      "grad_norm": 0.4470934271812439,
+      "learning_rate": 1.4280753639027566e-05,
+      "loss": 0.949,
+      "step": 8513
+    },
+    {
+      "epoch": 0.7607389371635357,
+      "grad_norm": 0.5684083700180054,
+      "learning_rate": 1.4270629208374197e-05,
+      "loss": 0.8836,
+      "step": 8514
+    },
+    {
+      "epoch": 0.7608282886947975,
+      "grad_norm": 0.4778299033641815,
+      "learning_rate": 1.4260507770436482e-05,
+      "loss": 0.9575,
+      "step": 8515
+    },
+    {
+      "epoch": 0.7609176402260593,
+      "grad_norm": 0.4669052064418793,
+      "learning_rate": 1.4250389326062207e-05,
+      "loss": 0.9437,
+      "step": 8516
+    },
+    {
+      "epoch": 0.7610069917573212,
+      "grad_norm": 0.4639929533004761,
+      "learning_rate": 1.424027387609888e-05,
+      "loss": 0.9677,
+      "step": 8517
+    },
+    {
+      "epoch": 0.7610963432885831,
+      "grad_norm": 0.40526002645492554,
+      "learning_rate": 1.4230161421393783e-05,
+      "loss": 0.9132,
+      "step": 8518
+    },
+    {
+      "epoch": 0.761185694819845,
+      "grad_norm": 0.45086291432380676,
+      "learning_rate": 1.422005196279395e-05,
+      "loss": 0.9638,
+      "step": 8519
+    },
+    {
+      "epoch": 0.7612750463511069,
+      "grad_norm": 0.48102667927742004,
+      "learning_rate": 1.420994550114615e-05,
+      "loss": 0.94,
+      "step": 8520
+    },
+    {
+      "epoch": 0.7613643978823688,
+      "grad_norm": 0.48567044734954834,
+      "learning_rate": 1.4199842037296906e-05,
+      "loss": 0.958,
+      "step": 8521
+    },
+    {
+      "epoch": 0.7614537494136305,
+      "grad_norm": 0.398173063993454,
+      "learning_rate": 1.4189741572092496e-05,
+      "loss": 1.0515,
+      "step": 8522
+    },
+    {
+      "epoch": 0.7615431009448924,
+      "grad_norm": 0.4758659601211548,
+      "learning_rate": 1.4179644106378942e-05,
+      "loss": 0.9731,
+      "step": 8523
+    },
+    {
+      "epoch": 0.7616324524761543,
+      "grad_norm": 0.5288426280021667,
+      "learning_rate": 1.4169549641002006e-05,
+      "loss": 0.9146,
+      "step": 8524
+    },
+    {
+      "epoch": 0.7617218040074162,
+      "grad_norm": 0.46394291520118713,
+      "learning_rate": 1.4159458176807217e-05,
+      "loss": 0.9605,
+      "step": 8525
+    },
+    {
+      "epoch": 0.7618111555386781,
+      "grad_norm": 0.4080126881599426,
+      "learning_rate": 1.4149369714639853e-05,
+      "loss": 0.9506,
+      "step": 8526
+    },
+    {
+      "epoch": 0.76190050706994,
+      "grad_norm": 0.5326384902000427,
+      "learning_rate": 1.4139284255344897e-05,
+      "loss": 0.8812,
+      "step": 8527
+    },
+    {
+      "epoch": 0.7619898586012018,
+      "grad_norm": 0.3904532194137573,
+      "learning_rate": 1.412920179976714e-05,
+      "loss": 0.9381,
+      "step": 8528
+    },
+    {
+      "epoch": 0.7620792101324636,
+      "grad_norm": 0.4524264931678772,
+      "learning_rate": 1.411912234875108e-05,
+      "loss": 1.029,
+      "step": 8529
+    },
+    {
+      "epoch": 0.7621685616637255,
+      "grad_norm": 0.4898856580257416,
+      "learning_rate": 1.4109045903141006e-05,
+      "loss": 0.9016,
+      "step": 8530
+    },
+    {
+      "epoch": 0.7622579131949874,
+      "grad_norm": 0.5062960982322693,
+      "learning_rate": 1.4098972463780885e-05,
+      "loss": 0.9609,
+      "step": 8531
+    },
+    {
+      "epoch": 0.7623472647262493,
+      "grad_norm": 0.5530799627304077,
+      "learning_rate": 1.4088902031514507e-05,
+      "loss": 0.9659,
+      "step": 8532
+    },
+    {
+      "epoch": 0.7624366162575111,
+      "grad_norm": 0.4216810464859009,
+      "learning_rate": 1.4078834607185364e-05,
+      "loss": 0.9497,
+      "step": 8533
+    },
+    {
+      "epoch": 0.762525967788773,
+      "grad_norm": 0.6185227036476135,
+      "learning_rate": 1.4068770191636716e-05,
+      "loss": 0.8765,
+      "step": 8534
+    },
+    {
+      "epoch": 0.7626153193200349,
+      "grad_norm": 0.5121302008628845,
+      "learning_rate": 1.405870878571156e-05,
+      "loss": 0.914,
+      "step": 8535
+    },
+    {
+      "epoch": 0.7627046708512967,
+      "grad_norm": 0.42362403869628906,
+      "learning_rate": 1.4048650390252671e-05,
+      "loss": 0.9526,
+      "step": 8536
+    },
+    {
+      "epoch": 0.7627940223825586,
+      "grad_norm": 0.4262167513370514,
+      "learning_rate": 1.4038595006102506e-05,
+      "loss": 0.9338,
+      "step": 8537
+    },
+    {
+      "epoch": 0.7628833739138204,
+      "grad_norm": 0.655556857585907,
+      "learning_rate": 1.4028542634103331e-05,
+      "loss": 0.9118,
+      "step": 8538
+    },
+    {
+      "epoch": 0.7629727254450823,
+      "grad_norm": 0.5181341767311096,
+      "learning_rate": 1.4018493275097139e-05,
+      "loss": 0.8456,
+      "step": 8539
+    },
+    {
+      "epoch": 0.7630620769763442,
+      "grad_norm": 0.46238502860069275,
+      "learning_rate": 1.4008446929925672e-05,
+      "loss": 0.9437,
+      "step": 8540
+    },
+    {
+      "epoch": 0.7631514285076061,
+      "grad_norm": 0.5332851409912109,
+      "learning_rate": 1.3998403599430415e-05,
+      "loss": 0.9058,
+      "step": 8541
+    },
+    {
+      "epoch": 0.7632407800388679,
+      "grad_norm": 0.47178998589515686,
+      "learning_rate": 1.398836328445261e-05,
+      "loss": 0.9194,
+      "step": 8542
+    },
+    {
+      "epoch": 0.7633301315701297,
+      "grad_norm": 0.5274223685264587,
+      "learning_rate": 1.3978325985833229e-05,
+      "loss": 0.9471,
+      "step": 8543
+    },
+    {
+      "epoch": 0.7634194831013916,
+      "grad_norm": 0.5024911165237427,
+      "learning_rate": 1.3968291704413018e-05,
+      "loss": 1.0011,
+      "step": 8544
+    },
+    {
+      "epoch": 0.7635088346326535,
+      "grad_norm": 0.5870459675788879,
+      "learning_rate": 1.3958260441032445e-05,
+      "loss": 0.8727,
+      "step": 8545
+    },
+    {
+      "epoch": 0.7635981861639154,
+      "grad_norm": 0.592512309551239,
+      "learning_rate": 1.3948232196531746e-05,
+      "loss": 0.8981,
+      "step": 8546
+    },
+    {
+      "epoch": 0.7636875376951773,
+      "grad_norm": 0.511677086353302,
+      "learning_rate": 1.3938206971750878e-05,
+      "loss": 0.9192,
+      "step": 8547
+    },
+    {
+      "epoch": 0.7637768892264392,
+      "grad_norm": 0.4113241732120514,
+      "learning_rate": 1.3928184767529562e-05,
+      "loss": 0.912,
+      "step": 8548
+    },
+    {
+      "epoch": 0.7638662407577009,
+      "grad_norm": 0.4762282967567444,
+      "learning_rate": 1.3918165584707271e-05,
+      "loss": 0.9661,
+      "step": 8549
+    },
+    {
+      "epoch": 0.7639555922889628,
+      "grad_norm": 0.5495911836624146,
+      "learning_rate": 1.3908149424123217e-05,
+      "loss": 1.0346,
+      "step": 8550
+    },
+    {
+      "epoch": 0.7640449438202247,
+      "grad_norm": 0.4995447099208832,
+      "learning_rate": 1.3898136286616364e-05,
+      "loss": 0.9112,
+      "step": 8551
+    },
+    {
+      "epoch": 0.7641342953514866,
+      "grad_norm": 0.48613351583480835,
+      "learning_rate": 1.3888126173025412e-05,
+      "loss": 0.9372,
+      "step": 8552
+    },
+    {
+      "epoch": 0.7642236468827485,
+      "grad_norm": 0.502858579158783,
+      "learning_rate": 1.3878119084188818e-05,
+      "loss": 0.9525,
+      "step": 8553
+    },
+    {
+      "epoch": 0.7643129984140103,
+      "grad_norm": 0.4795863628387451,
+      "learning_rate": 1.3868115020944783e-05,
+      "loss": 0.9097,
+      "step": 8554
+    },
+    {
+      "epoch": 0.7644023499452722,
+      "grad_norm": 0.49957504868507385,
+      "learning_rate": 1.385811398413125e-05,
+      "loss": 0.8775,
+      "step": 8555
+    },
+    {
+      "epoch": 0.764491701476534,
+      "grad_norm": 0.5270972847938538,
+      "learning_rate": 1.3848115974585934e-05,
+      "loss": 0.9626,
+      "step": 8556
+    },
+    {
+      "epoch": 0.7645810530077959,
+      "grad_norm": 0.4184163510799408,
+      "learning_rate": 1.3838120993146243e-05,
+      "loss": 0.8799,
+      "step": 8557
+    },
+    {
+      "epoch": 0.7646704045390578,
+      "grad_norm": 0.5301498770713806,
+      "learning_rate": 1.3828129040649374e-05,
+      "loss": 0.8824,
+      "step": 8558
+    },
+    {
+      "epoch": 0.7647597560703197,
+      "grad_norm": 0.4511061906814575,
+      "learning_rate": 1.381814011793226e-05,
+      "loss": 0.916,
+      "step": 8559
+    },
+    {
+      "epoch": 0.7648491076015815,
+      "grad_norm": 0.4901033043861389,
+      "learning_rate": 1.3808154225831583e-05,
+      "loss": 0.8872,
+      "step": 8560
+    },
+    {
+      "epoch": 0.7649384591328434,
+      "grad_norm": 0.45118212699890137,
+      "learning_rate": 1.3798171365183771e-05,
+      "loss": 0.9832,
+      "step": 8561
+    },
+    {
+      "epoch": 0.7650278106641053,
+      "grad_norm": 0.42388466000556946,
+      "learning_rate": 1.3788191536824984e-05,
+      "loss": 0.9311,
+      "step": 8562
+    },
+    {
+      "epoch": 0.7651171621953671,
+      "grad_norm": 0.4067128598690033,
+      "learning_rate": 1.3778214741591167e-05,
+      "loss": 0.9677,
+      "step": 8563
+    },
+    {
+      "epoch": 0.765206513726629,
+      "grad_norm": 0.5313117504119873,
+      "learning_rate": 1.3768240980317948e-05,
+      "loss": 0.8713,
+      "step": 8564
+    },
+    {
+      "epoch": 0.7652958652578908,
+      "grad_norm": 0.40129441022872925,
+      "learning_rate": 1.3758270253840744e-05,
+      "loss": 0.9213,
+      "step": 8565
+    },
+    {
+      "epoch": 0.7653852167891527,
+      "grad_norm": 0.5199376940727234,
+      "learning_rate": 1.374830256299472e-05,
+      "loss": 0.9241,
+      "step": 8566
+    },
+    {
+      "epoch": 0.7654745683204146,
+      "grad_norm": 0.4660845696926117,
+      "learning_rate": 1.3738337908614768e-05,
+      "loss": 0.9697,
+      "step": 8567
+    },
+    {
+      "epoch": 0.7655639198516765,
+      "grad_norm": 0.5757974982261658,
+      "learning_rate": 1.3728376291535555e-05,
+      "loss": 0.8874,
+      "step": 8568
+    },
+    {
+      "epoch": 0.7656532713829384,
+      "grad_norm": 0.45287203788757324,
+      "learning_rate": 1.3718417712591441e-05,
+      "loss": 0.9749,
+      "step": 8569
+    },
+    {
+      "epoch": 0.7657426229142001,
+      "grad_norm": 0.4740961492061615,
+      "learning_rate": 1.3708462172616577e-05,
+      "loss": 0.9843,
+      "step": 8570
+    },
+    {
+      "epoch": 0.765831974445462,
+      "grad_norm": 0.48708444833755493,
+      "learning_rate": 1.3698509672444843e-05,
+      "loss": 0.9231,
+      "step": 8571
+    },
+    {
+      "epoch": 0.7659213259767239,
+      "grad_norm": 0.4899809956550598,
+      "learning_rate": 1.3688560212909873e-05,
+      "loss": 0.9621,
+      "step": 8572
+    },
+    {
+      "epoch": 0.7660106775079858,
+      "grad_norm": 0.4719412922859192,
+      "learning_rate": 1.3678613794845035e-05,
+      "loss": 0.9742,
+      "step": 8573
+    },
+    {
+      "epoch": 0.7661000290392477,
+      "grad_norm": 0.44776198267936707,
+      "learning_rate": 1.3668670419083457e-05,
+      "loss": 0.8969,
+      "step": 8574
+    },
+    {
+      "epoch": 0.7661893805705096,
+      "grad_norm": 0.46503016352653503,
+      "learning_rate": 1.3658730086457988e-05,
+      "loss": 0.9827,
+      "step": 8575
+    },
+    {
+      "epoch": 0.7662787321017714,
+      "grad_norm": 0.5194110870361328,
+      "learning_rate": 1.3648792797801263e-05,
+      "loss": 0.959,
+      "step": 8576
+    },
+    {
+      "epoch": 0.7663680836330332,
+      "grad_norm": 0.46184584498405457,
+      "learning_rate": 1.36388585539456e-05,
+      "loss": 0.9375,
+      "step": 8577
+    },
+    {
+      "epoch": 0.7664574351642951,
+      "grad_norm": 0.416530579328537,
+      "learning_rate": 1.3628927355723114e-05,
+      "loss": 0.988,
+      "step": 8578
+    },
+    {
+      "epoch": 0.766546786695557,
+      "grad_norm": 0.4688471257686615,
+      "learning_rate": 1.3618999203965654e-05,
+      "loss": 0.9492,
+      "step": 8579
+    },
+    {
+      "epoch": 0.7666361382268189,
+      "grad_norm": 0.635256826877594,
+      "learning_rate": 1.3609074099504798e-05,
+      "loss": 0.8742,
+      "step": 8580
+    },
+    {
+      "epoch": 0.7667254897580807,
+      "grad_norm": 0.446486234664917,
+      "learning_rate": 1.3599152043171893e-05,
+      "loss": 0.9117,
+      "step": 8581
+    },
+    {
+      "epoch": 0.7668148412893426,
+      "grad_norm": 0.5363548994064331,
+      "learning_rate": 1.3589233035798005e-05,
+      "loss": 0.9194,
+      "step": 8582
+    },
+    {
+      "epoch": 0.7669041928206045,
+      "grad_norm": 0.5018768906593323,
+      "learning_rate": 1.357931707821396e-05,
+      "loss": 0.9498,
+      "step": 8583
+    },
+    {
+      "epoch": 0.7669935443518663,
+      "grad_norm": 0.43810418248176575,
+      "learning_rate": 1.3569404171250328e-05,
+      "loss": 0.9512,
+      "step": 8584
+    },
+    {
+      "epoch": 0.7670828958831282,
+      "grad_norm": 0.45606479048728943,
+      "learning_rate": 1.355949431573742e-05,
+      "loss": 0.9449,
+      "step": 8585
+    },
+    {
+      "epoch": 0.76717224741439,
+      "grad_norm": 0.5644077062606812,
+      "learning_rate": 1.354958751250529e-05,
+      "loss": 0.9936,
+      "step": 8586
+    },
+    {
+      "epoch": 0.7672615989456519,
+      "grad_norm": 0.483189195394516,
+      "learning_rate": 1.3539683762383753e-05,
+      "loss": 0.8959,
+      "step": 8587
+    },
+    {
+      "epoch": 0.7673509504769138,
+      "grad_norm": 0.497502863407135,
+      "learning_rate": 1.3529783066202329e-05,
+      "loss": 0.9242,
+      "step": 8588
+    },
+    {
+      "epoch": 0.7674403020081757,
+      "grad_norm": 0.5695610046386719,
+      "learning_rate": 1.3519885424790313e-05,
+      "loss": 0.8977,
+      "step": 8589
+    },
+    {
+      "epoch": 0.7675296535394376,
+      "grad_norm": 0.5756702423095703,
+      "learning_rate": 1.3509990838976744e-05,
+      "loss": 0.8939,
+      "step": 8590
+    },
+    {
+      "epoch": 0.7676190050706994,
+      "grad_norm": 0.5338248014450073,
+      "learning_rate": 1.3500099309590397e-05,
+      "loss": 0.9609,
+      "step": 8591
+    },
+    {
+      "epoch": 0.7677083566019612,
+      "grad_norm": 0.46391648054122925,
+      "learning_rate": 1.3490210837459799e-05,
+      "loss": 0.9912,
+      "step": 8592
+    },
+    {
+      "epoch": 0.7677977081332231,
+      "grad_norm": 0.5242441892623901,
+      "learning_rate": 1.3480325423413204e-05,
+      "loss": 0.8846,
+      "step": 8593
+    },
+    {
+      "epoch": 0.767887059664485,
+      "grad_norm": 0.45362526178359985,
+      "learning_rate": 1.3470443068278626e-05,
+      "loss": 0.9602,
+      "step": 8594
+    },
+    {
+      "epoch": 0.7679764111957469,
+      "grad_norm": 0.40796563029289246,
+      "learning_rate": 1.3460563772883822e-05,
+      "loss": 0.9368,
+      "step": 8595
+    },
+    {
+      "epoch": 0.7680657627270088,
+      "grad_norm": 0.481330931186676,
+      "learning_rate": 1.34506875380563e-05,
+      "loss": 0.9077,
+      "step": 8596
+    },
+    {
+      "epoch": 0.7681551142582707,
+      "grad_norm": 0.40942972898483276,
+      "learning_rate": 1.3440814364623267e-05,
+      "loss": 1.0025,
+      "step": 8597
+    },
+    {
+      "epoch": 0.7682444657895324,
+      "grad_norm": 0.533697247505188,
+      "learning_rate": 1.3430944253411727e-05,
+      "loss": 0.8807,
+      "step": 8598
+    },
+    {
+      "epoch": 0.7683338173207943,
+      "grad_norm": 0.4914191961288452,
+      "learning_rate": 1.34210772052484e-05,
+      "loss": 0.909,
+      "step": 8599
+    },
+    {
+      "epoch": 0.7684231688520562,
+      "grad_norm": 0.4288891851902008,
+      "learning_rate": 1.3411213220959773e-05,
+      "loss": 0.9436,
+      "step": 8600
+    },
+    {
+      "epoch": 0.7685125203833181,
+      "grad_norm": 0.4495335519313812,
+      "learning_rate": 1.340135230137204e-05,
+      "loss": 0.9656,
+      "step": 8601
+    },
+    {
+      "epoch": 0.76860187191458,
+      "grad_norm": 0.46838313341140747,
+      "learning_rate": 1.339149444731116e-05,
+      "loss": 0.9274,
+      "step": 8602
+    },
+    {
+      "epoch": 0.7686912234458418,
+      "grad_norm": 0.5495761632919312,
+      "learning_rate": 1.3381639659602841e-05,
+      "loss": 0.9106,
+      "step": 8603
+    },
+    {
+      "epoch": 0.7687805749771037,
+      "grad_norm": 0.4932359755039215,
+      "learning_rate": 1.3371787939072522e-05,
+      "loss": 0.9227,
+      "step": 8604
+    },
+    {
+      "epoch": 0.7688699265083655,
+      "grad_norm": 0.5239490270614624,
+      "learning_rate": 1.3361939286545388e-05,
+      "loss": 0.9407,
+      "step": 8605
+    },
+    {
+      "epoch": 0.7689592780396274,
+      "grad_norm": 0.46050575375556946,
+      "learning_rate": 1.335209370284638e-05,
+      "loss": 0.966,
+      "step": 8606
+    },
+    {
+      "epoch": 0.7690486295708893,
+      "grad_norm": 0.49059468507766724,
+      "learning_rate": 1.3342251188800175e-05,
+      "loss": 0.9366,
+      "step": 8607
+    },
+    {
+      "epoch": 0.7691379811021511,
+      "grad_norm": 0.46661099791526794,
+      "learning_rate": 1.333241174523116e-05,
+      "loss": 0.9926,
+      "step": 8608
+    },
+    {
+      "epoch": 0.769227332633413,
+      "grad_norm": 0.4436168372631073,
+      "learning_rate": 1.3322575372963515e-05,
+      "loss": 0.969,
+      "step": 8609
+    },
+    {
+      "epoch": 0.7693166841646749,
+      "grad_norm": 0.4760691225528717,
+      "learning_rate": 1.3312742072821127e-05,
+      "loss": 0.943,
+      "step": 8610
+    },
+    {
+      "epoch": 0.7694060356959367,
+      "grad_norm": 0.4735669195652008,
+      "learning_rate": 1.3302911845627658e-05,
+      "loss": 0.9113,
+      "step": 8611
+    },
+    {
+      "epoch": 0.7694953872271986,
+      "grad_norm": 0.515501081943512,
+      "learning_rate": 1.3293084692206476e-05,
+      "loss": 0.8752,
+      "step": 8612
+    },
+    {
+      "epoch": 0.7695847387584605,
+      "grad_norm": 0.46182894706726074,
+      "learning_rate": 1.3283260613380727e-05,
+      "loss": 0.9165,
+      "step": 8613
+    },
+    {
+      "epoch": 0.7696740902897223,
+      "grad_norm": 0.5648226737976074,
+      "learning_rate": 1.3273439609973271e-05,
+      "loss": 0.9357,
+      "step": 8614
+    },
+    {
+      "epoch": 0.7697634418209842,
+      "grad_norm": 0.5595985054969788,
+      "learning_rate": 1.326362168280672e-05,
+      "loss": 0.8938,
+      "step": 8615
+    },
+    {
+      "epoch": 0.7698527933522461,
+      "grad_norm": 0.5155600309371948,
+      "learning_rate": 1.3253806832703437e-05,
+      "loss": 0.9146,
+      "step": 8616
+    },
+    {
+      "epoch": 0.769942144883508,
+      "grad_norm": 0.4387071430683136,
+      "learning_rate": 1.3243995060485537e-05,
+      "loss": 0.9607,
+      "step": 8617
+    },
+    {
+      "epoch": 0.7700314964147698,
+      "grad_norm": 0.45591017603874207,
+      "learning_rate": 1.3234186366974822e-05,
+      "loss": 0.9637,
+      "step": 8618
+    },
+    {
+      "epoch": 0.7701208479460316,
+      "grad_norm": 0.47090432047843933,
+      "learning_rate": 1.3224380752992898e-05,
+      "loss": 0.9757,
+      "step": 8619
+    },
+    {
+      "epoch": 0.7702101994772935,
+      "grad_norm": 0.45392322540283203,
+      "learning_rate": 1.3214578219361085e-05,
+      "loss": 0.961,
+      "step": 8620
+    },
+    {
+      "epoch": 0.7702995510085554,
+      "grad_norm": 0.4720221161842346,
+      "learning_rate": 1.3204778766900445e-05,
+      "loss": 0.9789,
+      "step": 8621
+    },
+    {
+      "epoch": 0.7703889025398173,
+      "grad_norm": 0.674929141998291,
+      "learning_rate": 1.3194982396431798e-05,
+      "loss": 0.8242,
+      "step": 8622
+    },
+    {
+      "epoch": 0.7704782540710792,
+      "grad_norm": 0.5851225256919861,
+      "learning_rate": 1.3185189108775687e-05,
+      "loss": 0.87,
+      "step": 8623
+    },
+    {
+      "epoch": 0.770567605602341,
+      "grad_norm": 0.5186543464660645,
+      "learning_rate": 1.3175398904752407e-05,
+      "loss": 0.8479,
+      "step": 8624
+    },
+    {
+      "epoch": 0.7706569571336028,
+      "grad_norm": 0.5100897550582886,
+      "learning_rate": 1.3165611785181986e-05,
+      "loss": 0.9498,
+      "step": 8625
+    },
+    {
+      "epoch": 0.7707463086648647,
+      "grad_norm": 0.43524983525276184,
+      "learning_rate": 1.3155827750884209e-05,
+      "loss": 1.0058,
+      "step": 8626
+    },
+    {
+      "epoch": 0.7708356601961266,
+      "grad_norm": 0.6012685298919678,
+      "learning_rate": 1.3146046802678602e-05,
+      "loss": 0.8363,
+      "step": 8627
+    },
+    {
+      "epoch": 0.7709250117273885,
+      "grad_norm": 0.5322076082229614,
+      "learning_rate": 1.31362689413844e-05,
+      "loss": 0.9146,
+      "step": 8628
+    },
+    {
+      "epoch": 0.7710143632586504,
+      "grad_norm": 0.5446600317955017,
+      "learning_rate": 1.3126494167820607e-05,
+      "loss": 0.8362,
+      "step": 8629
+    },
+    {
+      "epoch": 0.7711037147899122,
+      "grad_norm": 0.4060341417789459,
+      "learning_rate": 1.311672248280597e-05,
+      "loss": 1.0057,
+      "step": 8630
+    },
+    {
+      "epoch": 0.7711930663211741,
+      "grad_norm": 0.45972713828086853,
+      "learning_rate": 1.310695388715898e-05,
+      "loss": 0.9607,
+      "step": 8631
+    },
+    {
+      "epoch": 0.7712824178524359,
+      "grad_norm": 0.4327675402164459,
+      "learning_rate": 1.3097188381697845e-05,
+      "loss": 0.9479,
+      "step": 8632
+    },
+    {
+      "epoch": 0.7713717693836978,
+      "grad_norm": 0.514522135257721,
+      "learning_rate": 1.3087425967240557e-05,
+      "loss": 0.9165,
+      "step": 8633
+    },
+    {
+      "epoch": 0.7714611209149597,
+      "grad_norm": 0.531057596206665,
+      "learning_rate": 1.307766664460479e-05,
+      "loss": 0.8906,
+      "step": 8634
+    },
+    {
+      "epoch": 0.7715504724462215,
+      "grad_norm": 0.5103472471237183,
+      "learning_rate": 1.3067910414608003e-05,
+      "loss": 1.0145,
+      "step": 8635
+    },
+    {
+      "epoch": 0.7716398239774834,
+      "grad_norm": 0.519164502620697,
+      "learning_rate": 1.305815727806739e-05,
+      "loss": 0.9374,
+      "step": 8636
+    },
+    {
+      "epoch": 0.7717291755087453,
+      "grad_norm": 0.5133655071258545,
+      "learning_rate": 1.3048407235799876e-05,
+      "loss": 0.9235,
+      "step": 8637
+    },
+    {
+      "epoch": 0.7718185270400072,
+      "grad_norm": 0.5075699687004089,
+      "learning_rate": 1.3038660288622145e-05,
+      "loss": 0.9772,
+      "step": 8638
+    },
+    {
+      "epoch": 0.771907878571269,
+      "grad_norm": 0.545677900314331,
+      "learning_rate": 1.302891643735058e-05,
+      "loss": 0.8661,
+      "step": 8639
+    },
+    {
+      "epoch": 0.7719972301025309,
+      "grad_norm": 0.5071326494216919,
+      "learning_rate": 1.3019175682801349e-05,
+      "loss": 0.8474,
+      "step": 8640
+    },
+    {
+      "epoch": 0.7720865816337927,
+      "grad_norm": 0.6671762466430664,
+      "learning_rate": 1.3009438025790337e-05,
+      "loss": 0.8427,
+      "step": 8641
+    },
+    {
+      "epoch": 0.7721759331650546,
+      "grad_norm": 0.4839807450771332,
+      "learning_rate": 1.2999703467133183e-05,
+      "loss": 0.9872,
+      "step": 8642
+    },
+    {
+      "epoch": 0.7722652846963165,
+      "grad_norm": 0.5609176754951477,
+      "learning_rate": 1.2989972007645263e-05,
+      "loss": 0.9293,
+      "step": 8643
+    },
+    {
+      "epoch": 0.7723546362275784,
+      "grad_norm": 0.5578341484069824,
+      "learning_rate": 1.2980243648141682e-05,
+      "loss": 0.9176,
+      "step": 8644
+    },
+    {
+      "epoch": 0.7724439877588403,
+      "grad_norm": 0.5111382007598877,
+      "learning_rate": 1.2970518389437297e-05,
+      "loss": 0.8884,
+      "step": 8645
+    },
+    {
+      "epoch": 0.772533339290102,
+      "grad_norm": 0.6080954670906067,
+      "learning_rate": 1.2960796232346706e-05,
+      "loss": 0.9019,
+      "step": 8646
+    },
+    {
+      "epoch": 0.7726226908213639,
+      "grad_norm": 0.4767414927482605,
+      "learning_rate": 1.295107717768425e-05,
+      "loss": 0.9232,
+      "step": 8647
+    },
+    {
+      "epoch": 0.7727120423526258,
+      "grad_norm": 0.4408529996871948,
+      "learning_rate": 1.2941361226263982e-05,
+      "loss": 0.9483,
+      "step": 8648
+    },
+    {
+      "epoch": 0.7728013938838877,
+      "grad_norm": 0.493099182844162,
+      "learning_rate": 1.293164837889973e-05,
+      "loss": 1.0169,
+      "step": 8649
+    },
+    {
+      "epoch": 0.7728907454151496,
+      "grad_norm": 0.554325520992279,
+      "learning_rate": 1.2921938636405045e-05,
+      "loss": 0.8812,
+      "step": 8650
+    },
+    {
+      "epoch": 0.7729800969464115,
+      "grad_norm": 0.513780951499939,
+      "learning_rate": 1.2912231999593222e-05,
+      "loss": 0.8577,
+      "step": 8651
+    },
+    {
+      "epoch": 0.7730694484776733,
+      "grad_norm": 0.4900115132331848,
+      "learning_rate": 1.2902528469277297e-05,
+      "loss": 0.9635,
+      "step": 8652
+    },
+    {
+      "epoch": 0.7731588000089351,
+      "grad_norm": 0.47862446308135986,
+      "learning_rate": 1.2892828046270038e-05,
+      "loss": 0.9151,
+      "step": 8653
+    },
+    {
+      "epoch": 0.773248151540197,
+      "grad_norm": 0.4748653173446655,
+      "learning_rate": 1.2883130731383969e-05,
+      "loss": 0.9588,
+      "step": 8654
+    },
+    {
+      "epoch": 0.7733375030714589,
+      "grad_norm": 0.4709342122077942,
+      "learning_rate": 1.2873436525431342e-05,
+      "loss": 0.9934,
+      "step": 8655
+    },
+    {
+      "epoch": 0.7734268546027208,
+      "grad_norm": 0.4540732800960541,
+      "learning_rate": 1.2863745429224144e-05,
+      "loss": 0.9338,
+      "step": 8656
+    },
+    {
+      "epoch": 0.7735162061339826,
+      "grad_norm": 0.5187547206878662,
+      "learning_rate": 1.2854057443574124e-05,
+      "loss": 0.8713,
+      "step": 8657
+    },
+    {
+      "epoch": 0.7736055576652445,
+      "grad_norm": 0.522544801235199,
+      "learning_rate": 1.2844372569292723e-05,
+      "loss": 0.8775,
+      "step": 8658
+    },
+    {
+      "epoch": 0.7736949091965064,
+      "grad_norm": 0.6091160774230957,
+      "learning_rate": 1.2834690807191174e-05,
+      "loss": 0.9253,
+      "step": 8659
+    },
+    {
+      "epoch": 0.7737842607277682,
+      "grad_norm": 0.5314405560493469,
+      "learning_rate": 1.2825012158080424e-05,
+      "loss": 0.9408,
+      "step": 8660
+    },
+    {
+      "epoch": 0.7738736122590301,
+      "grad_norm": 0.507613480091095,
+      "learning_rate": 1.2815336622771157e-05,
+      "loss": 0.9501,
+      "step": 8661
+    },
+    {
+      "epoch": 0.773962963790292,
+      "grad_norm": 0.4508688151836395,
+      "learning_rate": 1.2805664202073814e-05,
+      "loss": 1.0068,
+      "step": 8662
+    },
+    {
+      "epoch": 0.7740523153215538,
+      "grad_norm": 0.4795742332935333,
+      "learning_rate": 1.2795994896798551e-05,
+      "loss": 0.958,
+      "step": 8663
+    },
+    {
+      "epoch": 0.7741416668528157,
+      "grad_norm": 0.6187158226966858,
+      "learning_rate": 1.278632870775529e-05,
+      "loss": 0.8805,
+      "step": 8664
+    },
+    {
+      "epoch": 0.7742310183840776,
+      "grad_norm": 0.6057202816009521,
+      "learning_rate": 1.2776665635753665e-05,
+      "loss": 0.8507,
+      "step": 8665
+    },
+    {
+      "epoch": 0.7743203699153395,
+      "grad_norm": 0.43509769439697266,
+      "learning_rate": 1.2767005681603078e-05,
+      "loss": 0.9527,
+      "step": 8666
+    },
+    {
+      "epoch": 0.7744097214466013,
+      "grad_norm": 0.5171564817428589,
+      "learning_rate": 1.2757348846112626e-05,
+      "loss": 1.0171,
+      "step": 8667
+    },
+    {
+      "epoch": 0.7744990729778631,
+      "grad_norm": 0.5768764615058899,
+      "learning_rate": 1.2747695130091185e-05,
+      "loss": 0.912,
+      "step": 8668
+    },
+    {
+      "epoch": 0.774588424509125,
+      "grad_norm": 0.6042607426643372,
+      "learning_rate": 1.2738044534347365e-05,
+      "loss": 0.943,
+      "step": 8669
+    },
+    {
+      "epoch": 0.7746777760403869,
+      "grad_norm": 0.4685840606689453,
+      "learning_rate": 1.2728397059689495e-05,
+      "loss": 0.942,
+      "step": 8670
+    },
+    {
+      "epoch": 0.7747671275716488,
+      "grad_norm": 0.626491367816925,
+      "learning_rate": 1.271875270692567e-05,
+      "loss": 0.8608,
+      "step": 8671
+    },
+    {
+      "epoch": 0.7748564791029107,
+      "grad_norm": 0.5503062605857849,
+      "learning_rate": 1.2709111476863683e-05,
+      "loss": 0.8855,
+      "step": 8672
+    },
+    {
+      "epoch": 0.7749458306341724,
+      "grad_norm": 0.41604703664779663,
+      "learning_rate": 1.2699473370311099e-05,
+      "loss": 0.9734,
+      "step": 8673
+    },
+    {
+      "epoch": 0.7750351821654343,
+      "grad_norm": 0.48552244901657104,
+      "learning_rate": 1.268983838807522e-05,
+      "loss": 0.9216,
+      "step": 8674
+    },
+    {
+      "epoch": 0.7751245336966962,
+      "grad_norm": 0.45172011852264404,
+      "learning_rate": 1.2680206530963073e-05,
+      "loss": 0.9134,
+      "step": 8675
+    },
+    {
+      "epoch": 0.7752138852279581,
+      "grad_norm": 0.4474288523197174,
+      "learning_rate": 1.267057779978143e-05,
+      "loss": 1.0127,
+      "step": 8676
+    },
+    {
+      "epoch": 0.77530323675922,
+      "grad_norm": 0.5544458031654358,
+      "learning_rate": 1.2660952195336795e-05,
+      "loss": 0.8255,
+      "step": 8677
+    },
+    {
+      "epoch": 0.7753925882904819,
+      "grad_norm": 0.4368109703063965,
+      "learning_rate": 1.265132971843544e-05,
+      "loss": 0.9037,
+      "step": 8678
+    },
+    {
+      "epoch": 0.7754819398217437,
+      "grad_norm": 0.5210118889808655,
+      "learning_rate": 1.2641710369883308e-05,
+      "loss": 0.915,
+      "step": 8679
+    },
+    {
+      "epoch": 0.7755712913530055,
+      "grad_norm": 0.5763956308364868,
+      "learning_rate": 1.2632094150486146e-05,
+      "loss": 1.0095,
+      "step": 8680
+    },
+    {
+      "epoch": 0.7756606428842674,
+      "grad_norm": 0.5526300072669983,
+      "learning_rate": 1.2622481061049413e-05,
+      "loss": 0.9442,
+      "step": 8681
+    },
+    {
+      "epoch": 0.7757499944155293,
+      "grad_norm": 0.44834303855895996,
+      "learning_rate": 1.2612871102378304e-05,
+      "loss": 0.898,
+      "step": 8682
+    },
+    {
+      "epoch": 0.7758393459467912,
+      "grad_norm": 0.5319566130638123,
+      "learning_rate": 1.2603264275277766e-05,
+      "loss": 0.8959,
+      "step": 8683
+    },
+    {
+      "epoch": 0.775928697478053,
+      "grad_norm": 0.45926499366760254,
+      "learning_rate": 1.2593660580552457e-05,
+      "loss": 0.9541,
+      "step": 8684
+    },
+    {
+      "epoch": 0.7760180490093149,
+      "grad_norm": 0.4919742941856384,
+      "learning_rate": 1.2584060019006799e-05,
+      "loss": 0.8658,
+      "step": 8685
+    },
+    {
+      "epoch": 0.7761074005405768,
+      "grad_norm": 0.42770183086395264,
+      "learning_rate": 1.257446259144494e-05,
+      "loss": 0.9669,
+      "step": 8686
+    },
+    {
+      "epoch": 0.7761967520718386,
+      "grad_norm": 0.626804530620575,
+      "learning_rate": 1.2564868298670773e-05,
+      "loss": 0.8515,
+      "step": 8687
+    },
+    {
+      "epoch": 0.7762861036031005,
+      "grad_norm": 0.5289289355278015,
+      "learning_rate": 1.2555277141487925e-05,
+      "loss": 0.8717,
+      "step": 8688
+    },
+    {
+      "epoch": 0.7763754551343623,
+      "grad_norm": 0.3943426311016083,
+      "learning_rate": 1.2545689120699733e-05,
+      "loss": 0.9667,
+      "step": 8689
+    },
+    {
+      "epoch": 0.7764648066656242,
+      "grad_norm": 0.44126003980636597,
+      "learning_rate": 1.2536104237109314e-05,
+      "loss": 0.9189,
+      "step": 8690
+    },
+    {
+      "epoch": 0.7765541581968861,
+      "grad_norm": 0.5304616689682007,
+      "learning_rate": 1.2526522491519499e-05,
+      "loss": 0.891,
+      "step": 8691
+    },
+    {
+      "epoch": 0.776643509728148,
+      "grad_norm": 0.5561332106590271,
+      "learning_rate": 1.2516943884732862e-05,
+      "loss": 0.9357,
+      "step": 8692
+    },
+    {
+      "epoch": 0.7767328612594099,
+      "grad_norm": 0.48594895005226135,
+      "learning_rate": 1.2507368417551717e-05,
+      "loss": 0.8873,
+      "step": 8693
+    },
+    {
+      "epoch": 0.7768222127906717,
+      "grad_norm": 0.48896369338035583,
+      "learning_rate": 1.2497796090778113e-05,
+      "loss": 0.9618,
+      "step": 8694
+    },
+    {
+      "epoch": 0.7769115643219335,
+      "grad_norm": 0.5779229998588562,
+      "learning_rate": 1.2488226905213829e-05,
+      "loss": 0.9589,
+      "step": 8695
+    },
+    {
+      "epoch": 0.7770009158531954,
+      "grad_norm": 0.4492420256137848,
+      "learning_rate": 1.247866086166039e-05,
+      "loss": 0.9654,
+      "step": 8696
+    },
+    {
+      "epoch": 0.7770902673844573,
+      "grad_norm": 0.45466622710227966,
+      "learning_rate": 1.2469097960919052e-05,
+      "loss": 0.9783,
+      "step": 8697
+    },
+    {
+      "epoch": 0.7771796189157192,
+      "grad_norm": 0.4977153241634369,
+      "learning_rate": 1.2459538203790822e-05,
+      "loss": 0.9013,
+      "step": 8698
+    },
+    {
+      "epoch": 0.7772689704469811,
+      "grad_norm": 0.5779181122779846,
+      "learning_rate": 1.244998159107641e-05,
+      "loss": 0.8386,
+      "step": 8699
+    },
+    {
+      "epoch": 0.777358321978243,
+      "grad_norm": 0.4691259264945984,
+      "learning_rate": 1.2440428123576286e-05,
+      "loss": 1.0294,
+      "step": 8700
+    },
+    {
+      "epoch": 0.7774476735095047,
+      "grad_norm": 0.4812089502811432,
+      "learning_rate": 1.2430877802090674e-05,
+      "loss": 0.9459,
+      "step": 8701
+    },
+    {
+      "epoch": 0.7775370250407666,
+      "grad_norm": 0.5185266137123108,
+      "learning_rate": 1.2421330627419498e-05,
+      "loss": 0.8558,
+      "step": 8702
+    },
+    {
+      "epoch": 0.7776263765720285,
+      "grad_norm": 0.4775238037109375,
+      "learning_rate": 1.2411786600362457e-05,
+      "loss": 0.8715,
+      "step": 8703
+    },
+    {
+      "epoch": 0.7777157281032904,
+      "grad_norm": 0.5433192253112793,
+      "learning_rate": 1.2402245721718935e-05,
+      "loss": 0.9181,
+      "step": 8704
+    },
+    {
+      "epoch": 0.7778050796345523,
+      "grad_norm": 0.47637078166007996,
+      "learning_rate": 1.2392707992288095e-05,
+      "loss": 0.9268,
+      "step": 8705
+    },
+    {
+      "epoch": 0.7778944311658141,
+      "grad_norm": 0.45217010378837585,
+      "learning_rate": 1.2383173412868832e-05,
+      "loss": 0.9737,
+      "step": 8706
+    },
+    {
+      "epoch": 0.777983782697076,
+      "grad_norm": 0.40835416316986084,
+      "learning_rate": 1.2373641984259754e-05,
+      "loss": 1.0122,
+      "step": 8707
+    },
+    {
+      "epoch": 0.7780731342283378,
+      "grad_norm": 0.43482106924057007,
+      "learning_rate": 1.2364113707259251e-05,
+      "loss": 1.0017,
+      "step": 8708
+    },
+    {
+      "epoch": 0.7781624857595997,
+      "grad_norm": 0.4076036214828491,
+      "learning_rate": 1.2354588582665371e-05,
+      "loss": 0.9542,
+      "step": 8709
+    },
+    {
+      "epoch": 0.7782518372908616,
+      "grad_norm": 0.5018858313560486,
+      "learning_rate": 1.2345066611275973e-05,
+      "loss": 0.9303,
+      "step": 8710
+    },
+    {
+      "epoch": 0.7783411888221234,
+      "grad_norm": 0.45527756214141846,
+      "learning_rate": 1.2335547793888619e-05,
+      "loss": 1.0528,
+      "step": 8711
+    },
+    {
+      "epoch": 0.7784305403533853,
+      "grad_norm": 0.46973609924316406,
+      "learning_rate": 1.2326032131300613e-05,
+      "loss": 0.9795,
+      "step": 8712
+    },
+    {
+      "epoch": 0.7785198918846472,
+      "grad_norm": 0.46930214762687683,
+      "learning_rate": 1.2316519624308991e-05,
+      "loss": 0.8755,
+      "step": 8713
+    },
+    {
+      "epoch": 0.7786092434159091,
+      "grad_norm": 0.4111507534980774,
+      "learning_rate": 1.2307010273710528e-05,
+      "loss": 0.9699,
+      "step": 8714
+    },
+    {
+      "epoch": 0.7786985949471709,
+      "grad_norm": 0.5797132849693298,
+      "learning_rate": 1.229750408030173e-05,
+      "loss": 0.9161,
+      "step": 8715
+    },
+    {
+      "epoch": 0.7787879464784327,
+      "grad_norm": 0.5236135721206665,
+      "learning_rate": 1.2288001044878849e-05,
+      "loss": 0.9461,
+      "step": 8716
+    },
+    {
+      "epoch": 0.7788772980096946,
+      "grad_norm": 0.42989420890808105,
+      "learning_rate": 1.227850116823786e-05,
+      "loss": 0.973,
+      "step": 8717
+    },
+    {
+      "epoch": 0.7789666495409565,
+      "grad_norm": 0.5021857619285583,
+      "learning_rate": 1.2269004451174493e-05,
+      "loss": 0.9027,
+      "step": 8718
+    },
+    {
+      "epoch": 0.7790560010722184,
+      "grad_norm": 0.5190892219543457,
+      "learning_rate": 1.2259510894484173e-05,
+      "loss": 0.8612,
+      "step": 8719
+    },
+    {
+      "epoch": 0.7791453526034803,
+      "grad_norm": 0.4393077492713928,
+      "learning_rate": 1.22500204989621e-05,
+      "loss": 0.8991,
+      "step": 8720
+    },
+    {
+      "epoch": 0.7792347041347422,
+      "grad_norm": 0.5465923547744751,
+      "learning_rate": 1.2240533265403198e-05,
+      "loss": 0.8785,
+      "step": 8721
+    },
+    {
+      "epoch": 0.7793240556660039,
+      "grad_norm": 0.5483127236366272,
+      "learning_rate": 1.2231049194602122e-05,
+      "loss": 0.8618,
+      "step": 8722
+    },
+    {
+      "epoch": 0.7794134071972658,
+      "grad_norm": 0.4444792866706848,
+      "learning_rate": 1.2221568287353263e-05,
+      "loss": 0.9536,
+      "step": 8723
+    },
+    {
+      "epoch": 0.7795027587285277,
+      "grad_norm": 0.5139243006706238,
+      "learning_rate": 1.221209054445075e-05,
+      "loss": 0.8762,
+      "step": 8724
+    },
+    {
+      "epoch": 0.7795921102597896,
+      "grad_norm": 0.49245816469192505,
+      "learning_rate": 1.2202615966688442e-05,
+      "loss": 0.9376,
+      "step": 8725
+    },
+    {
+      "epoch": 0.7796814617910515,
+      "grad_norm": 0.5652595162391663,
+      "learning_rate": 1.2193144554859937e-05,
+      "loss": 0.8839,
+      "step": 8726
+    },
+    {
+      "epoch": 0.7797708133223133,
+      "grad_norm": 0.679634153842926,
+      "learning_rate": 1.2183676309758574e-05,
+      "loss": 0.8982,
+      "step": 8727
+    },
+    {
+      "epoch": 0.7798601648535752,
+      "grad_norm": 0.4609330892562866,
+      "learning_rate": 1.2174211232177419e-05,
+      "loss": 0.9649,
+      "step": 8728
+    },
+    {
+      "epoch": 0.779949516384837,
+      "grad_norm": 0.5099778771400452,
+      "learning_rate": 1.2164749322909257e-05,
+      "loss": 0.8654,
+      "step": 8729
+    },
+    {
+      "epoch": 0.7800388679160989,
+      "grad_norm": 0.5093337297439575,
+      "learning_rate": 1.2155290582746636e-05,
+      "loss": 0.9262,
+      "step": 8730
+    },
+    {
+      "epoch": 0.7801282194473608,
+      "grad_norm": 0.5493488907814026,
+      "learning_rate": 1.214583501248182e-05,
+      "loss": 0.9284,
+      "step": 8731
+    },
+    {
+      "epoch": 0.7802175709786227,
+      "grad_norm": 0.4473147392272949,
+      "learning_rate": 1.2136382612906822e-05,
+      "loss": 1.0044,
+      "step": 8732
+    },
+    {
+      "epoch": 0.7803069225098845,
+      "grad_norm": 0.4741600751876831,
+      "learning_rate": 1.2126933384813378e-05,
+      "loss": 0.913,
+      "step": 8733
+    },
+    {
+      "epoch": 0.7803962740411464,
+      "grad_norm": 0.42708808183670044,
+      "learning_rate": 1.2117487328992955e-05,
+      "loss": 0.9394,
+      "step": 8734
+    },
+    {
+      "epoch": 0.7804856255724082,
+      "grad_norm": 0.5554439425468445,
+      "learning_rate": 1.210804444623677e-05,
+      "loss": 0.9425,
+      "step": 8735
+    },
+    {
+      "epoch": 0.7805749771036701,
+      "grad_norm": 0.5091964602470398,
+      "learning_rate": 1.2098604737335778e-05,
+      "loss": 0.9622,
+      "step": 8736
+    },
+    {
+      "epoch": 0.780664328634932,
+      "grad_norm": 0.5323701500892639,
+      "learning_rate": 1.2089168203080625e-05,
+      "loss": 0.9593,
+      "step": 8737
+    },
+    {
+      "epoch": 0.7807536801661938,
+      "grad_norm": 0.5752742290496826,
+      "learning_rate": 1.207973484426173e-05,
+      "loss": 0.8755,
+      "step": 8738
+    },
+    {
+      "epoch": 0.7808430316974557,
+      "grad_norm": 0.4725722074508667,
+      "learning_rate": 1.2070304661669251e-05,
+      "loss": 1.0083,
+      "step": 8739
+    },
+    {
+      "epoch": 0.7809323832287176,
+      "grad_norm": 0.5769376754760742,
+      "learning_rate": 1.2060877656093051e-05,
+      "loss": 0.9272,
+      "step": 8740
+    },
+    {
+      "epoch": 0.7810217347599795,
+      "grad_norm": 0.5163832902908325,
+      "learning_rate": 1.2051453828322768e-05,
+      "loss": 0.9612,
+      "step": 8741
+    },
+    {
+      "epoch": 0.7811110862912413,
+      "grad_norm": 0.5350446105003357,
+      "learning_rate": 1.204203317914771e-05,
+      "loss": 0.9202,
+      "step": 8742
+    },
+    {
+      "epoch": 0.7812004378225031,
+      "grad_norm": 0.5735968351364136,
+      "learning_rate": 1.2032615709356981e-05,
+      "loss": 0.9389,
+      "step": 8743
+    },
+    {
+      "epoch": 0.781289789353765,
+      "grad_norm": 0.5735962390899658,
+      "learning_rate": 1.2023201419739389e-05,
+      "loss": 0.9112,
+      "step": 8744
+    },
+    {
+      "epoch": 0.7813791408850269,
+      "grad_norm": 0.5810607671737671,
+      "learning_rate": 1.2013790311083478e-05,
+      "loss": 0.9489,
+      "step": 8745
+    },
+    {
+      "epoch": 0.7814684924162888,
+      "grad_norm": 0.554716944694519,
+      "learning_rate": 1.2004382384177537e-05,
+      "loss": 0.9159,
+      "step": 8746
+    },
+    {
+      "epoch": 0.7815578439475507,
+      "grad_norm": 0.5520256757736206,
+      "learning_rate": 1.1994977639809574e-05,
+      "loss": 0.8593,
+      "step": 8747
+    },
+    {
+      "epoch": 0.7816471954788126,
+      "grad_norm": 0.4506448805332184,
+      "learning_rate": 1.1985576078767352e-05,
+      "loss": 0.9468,
+      "step": 8748
+    },
+    {
+      "epoch": 0.7817365470100743,
+      "grad_norm": 0.40401577949523926,
+      "learning_rate": 1.197617770183832e-05,
+      "loss": 0.9659,
+      "step": 8749
+    },
+    {
+      "epoch": 0.7818258985413362,
+      "grad_norm": 0.4281572699546814,
+      "learning_rate": 1.1966782509809715e-05,
+      "loss": 0.9421,
+      "step": 8750
+    },
+    {
+      "epoch": 0.7819152500725981,
+      "grad_norm": 0.4709114134311676,
+      "learning_rate": 1.195739050346848e-05,
+      "loss": 0.975,
+      "step": 8751
+    },
+    {
+      "epoch": 0.78200460160386,
+      "grad_norm": 0.4489923119544983,
+      "learning_rate": 1.1948001683601295e-05,
+      "loss": 1.0054,
+      "step": 8752
+    },
+    {
+      "epoch": 0.7820939531351219,
+      "grad_norm": 0.4930829107761383,
+      "learning_rate": 1.1938616050994572e-05,
+      "loss": 0.9004,
+      "step": 8753
+    },
+    {
+      "epoch": 0.7821833046663837,
+      "grad_norm": 0.5570724606513977,
+      "learning_rate": 1.192923360643446e-05,
+      "loss": 0.8782,
+      "step": 8754
+    },
+    {
+      "epoch": 0.7822726561976456,
+      "grad_norm": 0.44626376032829285,
+      "learning_rate": 1.1919854350706838e-05,
+      "loss": 0.9346,
+      "step": 8755
+    },
+    {
+      "epoch": 0.7823620077289074,
+      "grad_norm": 0.39036697149276733,
+      "learning_rate": 1.1910478284597321e-05,
+      "loss": 0.9688,
+      "step": 8756
+    },
+    {
+      "epoch": 0.7824513592601693,
+      "grad_norm": 0.5225417017936707,
+      "learning_rate": 1.1901105408891256e-05,
+      "loss": 0.8374,
+      "step": 8757
+    },
+    {
+      "epoch": 0.7825407107914312,
+      "grad_norm": 0.43829452991485596,
+      "learning_rate": 1.1891735724373726e-05,
+      "loss": 0.9661,
+      "step": 8758
+    },
+    {
+      "epoch": 0.782630062322693,
+      "grad_norm": 0.5867989659309387,
+      "learning_rate": 1.1882369231829526e-05,
+      "loss": 0.9218,
+      "step": 8759
+    },
+    {
+      "epoch": 0.7827194138539549,
+      "grad_norm": 0.46479299664497375,
+      "learning_rate": 1.1873005932043202e-05,
+      "loss": 0.94,
+      "step": 8760
+    },
+    {
+      "epoch": 0.7828087653852168,
+      "grad_norm": 0.543161153793335,
+      "learning_rate": 1.1863645825799042e-05,
+      "loss": 0.9706,
+      "step": 8761
+    },
+    {
+      "epoch": 0.7828981169164787,
+      "grad_norm": 0.4378313720226288,
+      "learning_rate": 1.185428891388104e-05,
+      "loss": 0.9379,
+      "step": 8762
+    },
+    {
+      "epoch": 0.7829874684477405,
+      "grad_norm": 0.44382184743881226,
+      "learning_rate": 1.1844935197072954e-05,
+      "loss": 0.9562,
+      "step": 8763
+    },
+    {
+      "epoch": 0.7830768199790024,
+      "grad_norm": 0.5598951578140259,
+      "learning_rate": 1.183558467615824e-05,
+      "loss": 0.8841,
+      "step": 8764
+    },
+    {
+      "epoch": 0.7831661715102642,
+      "grad_norm": 0.4546869695186615,
+      "learning_rate": 1.1826237351920123e-05,
+      "loss": 0.8926,
+      "step": 8765
+    },
+    {
+      "epoch": 0.7832555230415261,
+      "grad_norm": 0.3868406414985657,
+      "learning_rate": 1.1816893225141523e-05,
+      "loss": 1.0242,
+      "step": 8766
+    },
+    {
+      "epoch": 0.783344874572788,
+      "grad_norm": 0.4522659480571747,
+      "learning_rate": 1.1807552296605118e-05,
+      "loss": 0.9198,
+      "step": 8767
+    },
+    {
+      "epoch": 0.7834342261040499,
+      "grad_norm": 0.47773608565330505,
+      "learning_rate": 1.1798214567093313e-05,
+      "loss": 0.96,
+      "step": 8768
+    },
+    {
+      "epoch": 0.7835235776353118,
+      "grad_norm": 0.42124301195144653,
+      "learning_rate": 1.1788880037388256e-05,
+      "loss": 1.0142,
+      "step": 8769
+    },
+    {
+      "epoch": 0.7836129291665735,
+      "grad_norm": 0.4496397376060486,
+      "learning_rate": 1.1779548708271782e-05,
+      "loss": 0.964,
+      "step": 8770
+    },
+    {
+      "epoch": 0.7837022806978354,
+      "grad_norm": 0.4569949209690094,
+      "learning_rate": 1.1770220580525504e-05,
+      "loss": 0.9092,
+      "step": 8771
+    },
+    {
+      "epoch": 0.7837916322290973,
+      "grad_norm": 0.6700246334075928,
+      "learning_rate": 1.1760895654930748e-05,
+      "loss": 0.8237,
+      "step": 8772
+    },
+    {
+      "epoch": 0.7838809837603592,
+      "grad_norm": 0.4912089407444,
+      "learning_rate": 1.175157393226859e-05,
+      "loss": 0.8981,
+      "step": 8773
+    },
+    {
+      "epoch": 0.7839703352916211,
+      "grad_norm": 0.5014065504074097,
+      "learning_rate": 1.174225541331982e-05,
+      "loss": 0.9595,
+      "step": 8774
+    },
+    {
+      "epoch": 0.784059686822883,
+      "grad_norm": 0.5120930671691895,
+      "learning_rate": 1.1732940098864947e-05,
+      "loss": 0.9345,
+      "step": 8775
+    },
+    {
+      "epoch": 0.7841490383541448,
+      "grad_norm": 0.45890259742736816,
+      "learning_rate": 1.1723627989684239e-05,
+      "loss": 1.0086,
+      "step": 8776
+    },
+    {
+      "epoch": 0.7842383898854066,
+      "grad_norm": 0.43002086877822876,
+      "learning_rate": 1.171431908655768e-05,
+      "loss": 1.0457,
+      "step": 8777
+    },
+    {
+      "epoch": 0.7843277414166685,
+      "grad_norm": 0.4677255153656006,
+      "learning_rate": 1.1705013390264995e-05,
+      "loss": 0.8648,
+      "step": 8778
+    },
+    {
+      "epoch": 0.7844170929479304,
+      "grad_norm": 0.5233368277549744,
+      "learning_rate": 1.169571090158565e-05,
+      "loss": 0.8994,
+      "step": 8779
+    },
+    {
+      "epoch": 0.7845064444791923,
+      "grad_norm": 0.4638941287994385,
+      "learning_rate": 1.1686411621298793e-05,
+      "loss": 0.9106,
+      "step": 8780
+    },
+    {
+      "epoch": 0.7845957960104541,
+      "grad_norm": 0.4682579040527344,
+      "learning_rate": 1.167711555018336e-05,
+      "loss": 0.9533,
+      "step": 8781
+    },
+    {
+      "epoch": 0.784685147541716,
+      "grad_norm": 0.47554928064346313,
+      "learning_rate": 1.1667822689017988e-05,
+      "loss": 0.9157,
+      "step": 8782
+    },
+    {
+      "epoch": 0.7847744990729779,
+      "grad_norm": 0.5877280235290527,
+      "learning_rate": 1.165853303858106e-05,
+      "loss": 0.926,
+      "step": 8783
+    },
+    {
+      "epoch": 0.7848638506042397,
+      "grad_norm": 0.47542765736579895,
+      "learning_rate": 1.1649246599650681e-05,
+      "loss": 0.952,
+      "step": 8784
+    },
+    {
+      "epoch": 0.7849532021355016,
+      "grad_norm": 0.5808351635932922,
+      "learning_rate": 1.1639963373004691e-05,
+      "loss": 0.8716,
+      "step": 8785
+    },
+    {
+      "epoch": 0.7850425536667635,
+      "grad_norm": 0.4650845527648926,
+      "learning_rate": 1.1630683359420652e-05,
+      "loss": 0.9614,
+      "step": 8786
+    },
+    {
+      "epoch": 0.7851319051980253,
+      "grad_norm": 0.43621885776519775,
+      "learning_rate": 1.1621406559675873e-05,
+      "loss": 0.8881,
+      "step": 8787
+    },
+    {
+      "epoch": 0.7852212567292872,
+      "grad_norm": 0.47639355063438416,
+      "learning_rate": 1.1612132974547379e-05,
+      "loss": 0.9544,
+      "step": 8788
+    },
+    {
+      "epoch": 0.7853106082605491,
+      "grad_norm": 0.6018533706665039,
+      "learning_rate": 1.1602862604811955e-05,
+      "loss": 0.8597,
+      "step": 8789
+    },
+    {
+      "epoch": 0.785399959791811,
+      "grad_norm": 0.44231539964675903,
+      "learning_rate": 1.1593595451246047e-05,
+      "loss": 0.9587,
+      "step": 8790
+    },
+    {
+      "epoch": 0.7854893113230728,
+      "grad_norm": 0.5324105024337769,
+      "learning_rate": 1.1584331514625912e-05,
+      "loss": 0.9463,
+      "step": 8791
+    },
+    {
+      "epoch": 0.7855786628543346,
+      "grad_norm": 0.5201995372772217,
+      "learning_rate": 1.1575070795727489e-05,
+      "loss": 0.9776,
+      "step": 8792
+    },
+    {
+      "epoch": 0.7856680143855965,
+      "grad_norm": 0.45011529326438904,
+      "learning_rate": 1.1565813295326466e-05,
+      "loss": 0.9372,
+      "step": 8793
+    },
+    {
+      "epoch": 0.7857573659168584,
+      "grad_norm": 0.5564170479774475,
+      "learning_rate": 1.155655901419826e-05,
+      "loss": 0.9709,
+      "step": 8794
+    },
+    {
+      "epoch": 0.7858467174481203,
+      "grad_norm": 0.4589083790779114,
+      "learning_rate": 1.1547307953118014e-05,
+      "loss": 0.9465,
+      "step": 8795
+    },
+    {
+      "epoch": 0.7859360689793822,
+      "grad_norm": 0.5712124705314636,
+      "learning_rate": 1.1538060112860604e-05,
+      "loss": 0.8866,
+      "step": 8796
+    },
+    {
+      "epoch": 0.786025420510644,
+      "grad_norm": 0.45000290870666504,
+      "learning_rate": 1.1528815494200634e-05,
+      "loss": 0.9641,
+      "step": 8797
+    },
+    {
+      "epoch": 0.7861147720419058,
+      "grad_norm": 0.5889281630516052,
+      "learning_rate": 1.1519574097912444e-05,
+      "loss": 0.8468,
+      "step": 8798
+    },
+    {
+      "epoch": 0.7862041235731677,
+      "grad_norm": 0.5337108373641968,
+      "learning_rate": 1.1510335924770105e-05,
+      "loss": 0.9437,
+      "step": 8799
+    },
+    {
+      "epoch": 0.7862934751044296,
+      "grad_norm": 0.4237595498561859,
+      "learning_rate": 1.1501100975547385e-05,
+      "loss": 0.9497,
+      "step": 8800
+    },
+    {
+      "epoch": 0.7863828266356915,
+      "grad_norm": 0.4925912916660309,
+      "learning_rate": 1.1491869251017834e-05,
+      "loss": 0.945,
+      "step": 8801
+    },
+    {
+      "epoch": 0.7864721781669534,
+      "grad_norm": 0.6141911745071411,
+      "learning_rate": 1.1482640751954699e-05,
+      "loss": 0.8556,
+      "step": 8802
+    },
+    {
+      "epoch": 0.7865615296982152,
+      "grad_norm": 0.43323883414268494,
+      "learning_rate": 1.1473415479130962e-05,
+      "loss": 0.9278,
+      "step": 8803
+    },
+    {
+      "epoch": 0.786650881229477,
+      "grad_norm": 0.44862401485443115,
+      "learning_rate": 1.1464193433319347e-05,
+      "loss": 0.9978,
+      "step": 8804
+    },
+    {
+      "epoch": 0.7867402327607389,
+      "grad_norm": 0.46563851833343506,
+      "learning_rate": 1.145497461529229e-05,
+      "loss": 0.9032,
+      "step": 8805
+    },
+    {
+      "epoch": 0.7868295842920008,
+      "grad_norm": 0.6122609972953796,
+      "learning_rate": 1.144575902582199e-05,
+      "loss": 0.953,
+      "step": 8806
+    },
+    {
+      "epoch": 0.7869189358232627,
+      "grad_norm": 0.629047691822052,
+      "learning_rate": 1.1436546665680309e-05,
+      "loss": 0.945,
+      "step": 8807
+    },
+    {
+      "epoch": 0.7870082873545245,
+      "grad_norm": 0.47389715909957886,
+      "learning_rate": 1.1427337535638905e-05,
+      "loss": 0.9537,
+      "step": 8808
+    },
+    {
+      "epoch": 0.7870976388857864,
+      "grad_norm": 0.4402306079864502,
+      "learning_rate": 1.1418131636469137e-05,
+      "loss": 1.0046,
+      "step": 8809
+    },
+    {
+      "epoch": 0.7871869904170483,
+      "grad_norm": 0.4482784867286682,
+      "learning_rate": 1.1408928968942095e-05,
+      "loss": 0.8711,
+      "step": 8810
+    },
+    {
+      "epoch": 0.7872763419483101,
+      "grad_norm": 0.4512721598148346,
+      "learning_rate": 1.1399729533828623e-05,
+      "loss": 0.9093,
+      "step": 8811
+    },
+    {
+      "epoch": 0.787365693479572,
+      "grad_norm": 0.5375701189041138,
+      "learning_rate": 1.1390533331899234e-05,
+      "loss": 0.892,
+      "step": 8812
+    },
+    {
+      "epoch": 0.7874550450108339,
+      "grad_norm": 0.44732967019081116,
+      "learning_rate": 1.1381340363924226e-05,
+      "loss": 0.9631,
+      "step": 8813
+    },
+    {
+      "epoch": 0.7875443965420957,
+      "grad_norm": 0.5341693758964539,
+      "learning_rate": 1.1372150630673606e-05,
+      "loss": 0.8947,
+      "step": 8814
+    },
+    {
+      "epoch": 0.7876337480733576,
+      "grad_norm": 0.5837507247924805,
+      "learning_rate": 1.136296413291712e-05,
+      "loss": 0.9357,
+      "step": 8815
+    },
+    {
+      "epoch": 0.7877230996046195,
+      "grad_norm": 0.49844542145729065,
+      "learning_rate": 1.1353780871424225e-05,
+      "loss": 0.9402,
+      "step": 8816
+    },
+    {
+      "epoch": 0.7878124511358814,
+      "grad_norm": 0.5053272247314453,
+      "learning_rate": 1.134460084696412e-05,
+      "loss": 0.916,
+      "step": 8817
+    },
+    {
+      "epoch": 0.7879018026671432,
+      "grad_norm": 0.47650039196014404,
+      "learning_rate": 1.1335424060305733e-05,
+      "loss": 0.9513,
+      "step": 8818
+    },
+    {
+      "epoch": 0.787991154198405,
+      "grad_norm": 0.5099322199821472,
+      "learning_rate": 1.1326250512217728e-05,
+      "loss": 0.9191,
+      "step": 8819
+    },
+    {
+      "epoch": 0.7880805057296669,
+      "grad_norm": 0.6279272437095642,
+      "learning_rate": 1.1317080203468462e-05,
+      "loss": 0.935,
+      "step": 8820
+    },
+    {
+      "epoch": 0.7881698572609288,
+      "grad_norm": 0.5648776888847351,
+      "learning_rate": 1.1307913134826059e-05,
+      "loss": 0.8982,
+      "step": 8821
+    },
+    {
+      "epoch": 0.7882592087921907,
+      "grad_norm": 0.6467058062553406,
+      "learning_rate": 1.1298749307058359e-05,
+      "loss": 0.9648,
+      "step": 8822
+    },
+    {
+      "epoch": 0.7883485603234526,
+      "grad_norm": 0.5014479756355286,
+      "learning_rate": 1.1289588720932931e-05,
+      "loss": 0.928,
+      "step": 8823
+    },
+    {
+      "epoch": 0.7884379118547145,
+      "grad_norm": 0.5449494123458862,
+      "learning_rate": 1.1280431377217071e-05,
+      "loss": 1.0022,
+      "step": 8824
+    },
+    {
+      "epoch": 0.7885272633859762,
+      "grad_norm": 0.46424680948257446,
+      "learning_rate": 1.1271277276677805e-05,
+      "loss": 0.8855,
+      "step": 8825
+    },
+    {
+      "epoch": 0.7886166149172381,
+      "grad_norm": 0.548555314540863,
+      "learning_rate": 1.1262126420081887e-05,
+      "loss": 0.8627,
+      "step": 8826
+    },
+    {
+      "epoch": 0.7887059664485,
+      "grad_norm": 0.44086822867393494,
+      "learning_rate": 1.125297880819579e-05,
+      "loss": 0.9702,
+      "step": 8827
+    },
+    {
+      "epoch": 0.7887953179797619,
+      "grad_norm": 0.4535292088985443,
+      "learning_rate": 1.124383444178574e-05,
+      "loss": 0.8851,
+      "step": 8828
+    },
+    {
+      "epoch": 0.7888846695110238,
+      "grad_norm": 0.5612303614616394,
+      "learning_rate": 1.1234693321617673e-05,
+      "loss": 0.9546,
+      "step": 8829
+    },
+    {
+      "epoch": 0.7889740210422856,
+      "grad_norm": 0.467670738697052,
+      "learning_rate": 1.1225555448457242e-05,
+      "loss": 0.9503,
+      "step": 8830
+    },
+    {
+      "epoch": 0.7890633725735475,
+      "grad_norm": 0.4849298298358917,
+      "learning_rate": 1.1216420823069846e-05,
+      "loss": 0.9047,
+      "step": 8831
+    },
+    {
+      "epoch": 0.7891527241048093,
+      "grad_norm": 0.41470497846603394,
+      "learning_rate": 1.1207289446220604e-05,
+      "loss": 0.9613,
+      "step": 8832
+    },
+    {
+      "epoch": 0.7892420756360712,
+      "grad_norm": 0.4530922472476959,
+      "learning_rate": 1.1198161318674377e-05,
+      "loss": 0.9708,
+      "step": 8833
+    },
+    {
+      "epoch": 0.7893314271673331,
+      "grad_norm": 0.422997385263443,
+      "learning_rate": 1.1189036441195733e-05,
+      "loss": 0.9678,
+      "step": 8834
+    },
+    {
+      "epoch": 0.789420778698595,
+      "grad_norm": 0.5136018395423889,
+      "learning_rate": 1.1179914814548986e-05,
+      "loss": 0.9182,
+      "step": 8835
+    },
+    {
+      "epoch": 0.7895101302298568,
+      "grad_norm": 0.529691755771637,
+      "learning_rate": 1.1170796439498166e-05,
+      "loss": 0.8935,
+      "step": 8836
+    },
+    {
+      "epoch": 0.7895994817611187,
+      "grad_norm": 0.47432875633239746,
+      "learning_rate": 1.1161681316807032e-05,
+      "loss": 0.9239,
+      "step": 8837
+    },
+    {
+      "epoch": 0.7896888332923806,
+      "grad_norm": 0.563213586807251,
+      "learning_rate": 1.1152569447239075e-05,
+      "loss": 0.9184,
+      "step": 8838
+    },
+    {
+      "epoch": 0.7897781848236424,
+      "grad_norm": 0.4394233524799347,
+      "learning_rate": 1.1143460831557523e-05,
+      "loss": 0.9732,
+      "step": 8839
+    },
+    {
+      "epoch": 0.7898675363549043,
+      "grad_norm": 0.46250519156455994,
+      "learning_rate": 1.1134355470525293e-05,
+      "loss": 0.9073,
+      "step": 8840
+    },
+    {
+      "epoch": 0.7899568878861661,
+      "grad_norm": 0.43355268239974976,
+      "learning_rate": 1.1125253364905075e-05,
+      "loss": 0.9525,
+      "step": 8841
+    },
+    {
+      "epoch": 0.790046239417428,
+      "grad_norm": 0.4754622280597687,
+      "learning_rate": 1.111615451545926e-05,
+      "loss": 0.946,
+      "step": 8842
+    },
+    {
+      "epoch": 0.7901355909486899,
+      "grad_norm": 0.47392573952674866,
+      "learning_rate": 1.1107058922949975e-05,
+      "loss": 0.9907,
+      "step": 8843
+    },
+    {
+      "epoch": 0.7902249424799518,
+      "grad_norm": 0.4125828742980957,
+      "learning_rate": 1.1097966588139092e-05,
+      "loss": 0.9411,
+      "step": 8844
+    },
+    {
+      "epoch": 0.7903142940112137,
+      "grad_norm": 0.5055714249610901,
+      "learning_rate": 1.1088877511788154e-05,
+      "loss": 0.9096,
+      "step": 8845
+    },
+    {
+      "epoch": 0.7904036455424754,
+      "grad_norm": 0.5957999229431152,
+      "learning_rate": 1.1079791694658487e-05,
+      "loss": 0.922,
+      "step": 8846
+    },
+    {
+      "epoch": 0.7904929970737373,
+      "grad_norm": 0.48999646306037903,
+      "learning_rate": 1.1070709137511132e-05,
+      "loss": 0.9453,
+      "step": 8847
+    },
+    {
+      "epoch": 0.7905823486049992,
+      "grad_norm": 0.40712419152259827,
+      "learning_rate": 1.1061629841106836e-05,
+      "loss": 0.9806,
+      "step": 8848
+    },
+    {
+      "epoch": 0.7906717001362611,
+      "grad_norm": 0.5603836178779602,
+      "learning_rate": 1.1052553806206094e-05,
+      "loss": 1.014,
+      "step": 8849
+    },
+    {
+      "epoch": 0.790761051667523,
+      "grad_norm": 0.467892587184906,
+      "learning_rate": 1.1043481033569142e-05,
+      "loss": 0.9254,
+      "step": 8850
+    },
+    {
+      "epoch": 0.7908504031987849,
+      "grad_norm": 0.5536115169525146,
+      "learning_rate": 1.103441152395588e-05,
+      "loss": 0.9557,
+      "step": 8851
+    },
+    {
+      "epoch": 0.7909397547300467,
+      "grad_norm": 0.4258303940296173,
+      "learning_rate": 1.1025345278125998e-05,
+      "loss": 0.9526,
+      "step": 8852
+    },
+    {
+      "epoch": 0.7910291062613085,
+      "grad_norm": 0.4268374443054199,
+      "learning_rate": 1.1016282296838887e-05,
+      "loss": 0.9622,
+      "step": 8853
+    },
+    {
+      "epoch": 0.7911184577925704,
+      "grad_norm": 0.525230348110199,
+      "learning_rate": 1.100722258085367e-05,
+      "loss": 0.9293,
+      "step": 8854
+    },
+    {
+      "epoch": 0.7912078093238323,
+      "grad_norm": 0.49180522561073303,
+      "learning_rate": 1.0998166130929199e-05,
+      "loss": 1.0217,
+      "step": 8855
+    },
+    {
+      "epoch": 0.7912971608550942,
+      "grad_norm": 0.4953739643096924,
+      "learning_rate": 1.0989112947824043e-05,
+      "loss": 0.9168,
+      "step": 8856
+    },
+    {
+      "epoch": 0.791386512386356,
+      "grad_norm": 0.4349702298641205,
+      "learning_rate": 1.0980063032296501e-05,
+      "loss": 0.9029,
+      "step": 8857
+    },
+    {
+      "epoch": 0.7914758639176179,
+      "grad_norm": 0.4538852870464325,
+      "learning_rate": 1.0971016385104604e-05,
+      "loss": 0.9189,
+      "step": 8858
+    },
+    {
+      "epoch": 0.7915652154488797,
+      "grad_norm": 0.466751366853714,
+      "learning_rate": 1.0961973007006105e-05,
+      "loss": 0.941,
+      "step": 8859
+    },
+    {
+      "epoch": 0.7916545669801416,
+      "grad_norm": 0.5299864411354065,
+      "learning_rate": 1.0952932898758494e-05,
+      "loss": 0.9471,
+      "step": 8860
+    },
+    {
+      "epoch": 0.7917439185114035,
+      "grad_norm": 0.4934852719306946,
+      "learning_rate": 1.0943896061118953e-05,
+      "loss": 0.8567,
+      "step": 8861
+    },
+    {
+      "epoch": 0.7918332700426653,
+      "grad_norm": 0.442080557346344,
+      "learning_rate": 1.0934862494844427e-05,
+      "loss": 0.9867,
+      "step": 8862
+    },
+    {
+      "epoch": 0.7919226215739272,
+      "grad_norm": 0.5051554441452026,
+      "learning_rate": 1.0925832200691566e-05,
+      "loss": 0.9596,
+      "step": 8863
+    },
+    {
+      "epoch": 0.7920119731051891,
+      "grad_norm": 0.4397631883621216,
+      "learning_rate": 1.0916805179416761e-05,
+      "loss": 0.9463,
+      "step": 8864
+    },
+    {
+      "epoch": 0.792101324636451,
+      "grad_norm": 0.4761650264263153,
+      "learning_rate": 1.0907781431776121e-05,
+      "loss": 0.9427,
+      "step": 8865
+    },
+    {
+      "epoch": 0.7921906761677128,
+      "grad_norm": 0.5251712203025818,
+      "learning_rate": 1.0898760958525479e-05,
+      "loss": 0.9143,
+      "step": 8866
+    },
+    {
+      "epoch": 0.7922800276989747,
+      "grad_norm": 0.4629380404949188,
+      "learning_rate": 1.0889743760420396e-05,
+      "loss": 0.9407,
+      "step": 8867
+    },
+    {
+      "epoch": 0.7923693792302365,
+      "grad_norm": 0.3857233226299286,
+      "learning_rate": 1.0880729838216153e-05,
+      "loss": 0.9492,
+      "step": 8868
+    },
+    {
+      "epoch": 0.7924587307614984,
+      "grad_norm": 0.5180653929710388,
+      "learning_rate": 1.0871719192667773e-05,
+      "loss": 0.8831,
+      "step": 8869
+    },
+    {
+      "epoch": 0.7925480822927603,
+      "grad_norm": 0.49814748764038086,
+      "learning_rate": 1.0862711824530003e-05,
+      "loss": 0.9292,
+      "step": 8870
+    },
+    {
+      "epoch": 0.7926374338240222,
+      "grad_norm": 0.4596767723560333,
+      "learning_rate": 1.0853707734557272e-05,
+      "loss": 0.9822,
+      "step": 8871
+    },
+    {
+      "epoch": 0.7927267853552841,
+      "grad_norm": 0.4747507870197296,
+      "learning_rate": 1.084470692350379e-05,
+      "loss": 0.9528,
+      "step": 8872
+    },
+    {
+      "epoch": 0.7928161368865458,
+      "grad_norm": 0.5321727395057678,
+      "learning_rate": 1.0835709392123472e-05,
+      "loss": 0.9908,
+      "step": 8873
+    },
+    {
+      "epoch": 0.7929054884178077,
+      "grad_norm": 0.529675304889679,
+      "learning_rate": 1.0826715141169947e-05,
+      "loss": 0.9526,
+      "step": 8874
+    },
+    {
+      "epoch": 0.7929948399490696,
+      "grad_norm": 0.4251299798488617,
+      "learning_rate": 1.0817724171396592e-05,
+      "loss": 0.9465,
+      "step": 8875
+    },
+    {
+      "epoch": 0.7930841914803315,
+      "grad_norm": 0.44601455330848694,
+      "learning_rate": 1.0808736483556487e-05,
+      "loss": 0.9579,
+      "step": 8876
+    },
+    {
+      "epoch": 0.7931735430115934,
+      "grad_norm": 0.5946657061576843,
+      "learning_rate": 1.0799752078402469e-05,
+      "loss": 0.9116,
+      "step": 8877
+    },
+    {
+      "epoch": 0.7932628945428553,
+      "grad_norm": 0.460638165473938,
+      "learning_rate": 1.079077095668704e-05,
+      "loss": 0.9722,
+      "step": 8878
+    },
+    {
+      "epoch": 0.7933522460741171,
+      "grad_norm": 0.4943525493144989,
+      "learning_rate": 1.0781793119162487e-05,
+      "loss": 0.8539,
+      "step": 8879
+    },
+    {
+      "epoch": 0.7934415976053789,
+      "grad_norm": 0.45845288038253784,
+      "learning_rate": 1.0772818566580795e-05,
+      "loss": 0.9393,
+      "step": 8880
+    },
+    {
+      "epoch": 0.7935309491366408,
+      "grad_norm": 0.515739917755127,
+      "learning_rate": 1.0763847299693691e-05,
+      "loss": 0.8696,
+      "step": 8881
+    },
+    {
+      "epoch": 0.7936203006679027,
+      "grad_norm": 0.5935304164886475,
+      "learning_rate": 1.0754879319252591e-05,
+      "loss": 1.0198,
+      "step": 8882
+    },
+    {
+      "epoch": 0.7937096521991646,
+      "grad_norm": 0.4660301208496094,
+      "learning_rate": 1.0745914626008675e-05,
+      "loss": 0.8689,
+      "step": 8883
+    },
+    {
+      "epoch": 0.7937990037304264,
+      "grad_norm": 0.47837138175964355,
+      "learning_rate": 1.0736953220712826e-05,
+      "loss": 0.8852,
+      "step": 8884
+    },
+    {
+      "epoch": 0.7938883552616883,
+      "grad_norm": 0.4837789535522461,
+      "learning_rate": 1.0727995104115656e-05,
+      "loss": 0.8776,
+      "step": 8885
+    },
+    {
+      "epoch": 0.7939777067929502,
+      "grad_norm": 0.6152971386909485,
+      "learning_rate": 1.0719040276967507e-05,
+      "loss": 0.8065,
+      "step": 8886
+    },
+    {
+      "epoch": 0.794067058324212,
+      "grad_norm": 0.532853364944458,
+      "learning_rate": 1.0710088740018437e-05,
+      "loss": 0.8949,
+      "step": 8887
+    },
+    {
+      "epoch": 0.7941564098554739,
+      "grad_norm": 0.4451427161693573,
+      "learning_rate": 1.0701140494018242e-05,
+      "loss": 0.9532,
+      "step": 8888
+    },
+    {
+      "epoch": 0.7942457613867357,
+      "grad_norm": 0.4042229652404785,
+      "learning_rate": 1.0692195539716421e-05,
+      "loss": 0.9972,
+      "step": 8889
+    },
+    {
+      "epoch": 0.7943351129179976,
+      "grad_norm": 0.47671109437942505,
+      "learning_rate": 1.0683253877862225e-05,
+      "loss": 0.9228,
+      "step": 8890
+    },
+    {
+      "epoch": 0.7944244644492595,
+      "grad_norm": 0.4329127073287964,
+      "learning_rate": 1.0674315509204596e-05,
+      "loss": 0.93,
+      "step": 8891
+    },
+    {
+      "epoch": 0.7945138159805214,
+      "grad_norm": 0.47998568415641785,
+      "learning_rate": 1.0665380434492223e-05,
+      "loss": 0.9536,
+      "step": 8892
+    },
+    {
+      "epoch": 0.7946031675117833,
+      "grad_norm": 0.5138132572174072,
+      "learning_rate": 1.0656448654473517e-05,
+      "loss": 0.9208,
+      "step": 8893
+    },
+    {
+      "epoch": 0.794692519043045,
+      "grad_norm": 0.48235902190208435,
+      "learning_rate": 1.0647520169896607e-05,
+      "loss": 0.8611,
+      "step": 8894
+    },
+    {
+      "epoch": 0.7947818705743069,
+      "grad_norm": 0.394228458404541,
+      "learning_rate": 1.0638594981509348e-05,
+      "loss": 0.974,
+      "step": 8895
+    },
+    {
+      "epoch": 0.7948712221055688,
+      "grad_norm": 0.4859960675239563,
+      "learning_rate": 1.0629673090059322e-05,
+      "loss": 0.9407,
+      "step": 8896
+    },
+    {
+      "epoch": 0.7949605736368307,
+      "grad_norm": 0.5173121690750122,
+      "learning_rate": 1.0620754496293833e-05,
+      "loss": 1.0492,
+      "step": 8897
+    },
+    {
+      "epoch": 0.7950499251680926,
+      "grad_norm": 0.453870564699173,
+      "learning_rate": 1.061183920095991e-05,
+      "loss": 0.9226,
+      "step": 8898
+    },
+    {
+      "epoch": 0.7951392766993545,
+      "grad_norm": 0.44933760166168213,
+      "learning_rate": 1.0602927204804296e-05,
+      "loss": 0.9376,
+      "step": 8899
+    },
+    {
+      "epoch": 0.7952286282306164,
+      "grad_norm": 0.48174723982810974,
+      "learning_rate": 1.0594018508573489e-05,
+      "loss": 0.9127,
+      "step": 8900
+    },
+    {
+      "epoch": 0.7953179797618781,
+      "grad_norm": 0.5078961253166199,
+      "learning_rate": 1.0585113113013657e-05,
+      "loss": 0.8392,
+      "step": 8901
+    },
+    {
+      "epoch": 0.79540733129314,
+      "grad_norm": 0.4766399562358856,
+      "learning_rate": 1.0576211018870736e-05,
+      "loss": 0.9298,
+      "step": 8902
+    },
+    {
+      "epoch": 0.7954966828244019,
+      "grad_norm": 0.5817358493804932,
+      "learning_rate": 1.0567312226890364e-05,
+      "loss": 0.9257,
+      "step": 8903
+    },
+    {
+      "epoch": 0.7955860343556638,
+      "grad_norm": 0.49117544293403625,
+      "learning_rate": 1.0558416737817916e-05,
+      "loss": 0.9214,
+      "step": 8904
+    },
+    {
+      "epoch": 0.7956753858869257,
+      "grad_norm": 0.46288153529167175,
+      "learning_rate": 1.0549524552398488e-05,
+      "loss": 0.9006,
+      "step": 8905
+    },
+    {
+      "epoch": 0.7957647374181875,
+      "grad_norm": 0.4867618978023529,
+      "learning_rate": 1.0540635671376885e-05,
+      "loss": 0.9344,
+      "step": 8906
+    },
+    {
+      "epoch": 0.7958540889494494,
+      "grad_norm": 0.5538333058357239,
+      "learning_rate": 1.053175009549765e-05,
+      "loss": 0.7893,
+      "step": 8907
+    },
+    {
+      "epoch": 0.7959434404807112,
+      "grad_norm": 0.4785383939743042,
+      "learning_rate": 1.0522867825505051e-05,
+      "loss": 0.9209,
+      "step": 8908
+    },
+    {
+      "epoch": 0.7960327920119731,
+      "grad_norm": 0.5016849637031555,
+      "learning_rate": 1.0513988862143082e-05,
+      "loss": 0.9328,
+      "step": 8909
+    },
+    {
+      "epoch": 0.796122143543235,
+      "grad_norm": 0.4911724030971527,
+      "learning_rate": 1.050511320615542e-05,
+      "loss": 0.9874,
+      "step": 8910
+    },
+    {
+      "epoch": 0.7962114950744968,
+      "grad_norm": 0.5558311939239502,
+      "learning_rate": 1.0496240858285511e-05,
+      "loss": 0.9098,
+      "step": 8911
+    },
+    {
+      "epoch": 0.7963008466057587,
+      "grad_norm": 0.6585413217544556,
+      "learning_rate": 1.0487371819276509e-05,
+      "loss": 0.9099,
+      "step": 8912
+    },
+    {
+      "epoch": 0.7963901981370206,
+      "grad_norm": 0.5155404806137085,
+      "learning_rate": 1.0478506089871293e-05,
+      "loss": 0.9004,
+      "step": 8913
+    },
+    {
+      "epoch": 0.7964795496682825,
+      "grad_norm": 0.4853888154029846,
+      "learning_rate": 1.0469643670812479e-05,
+      "loss": 0.9659,
+      "step": 8914
+    },
+    {
+      "epoch": 0.7965689011995443,
+      "grad_norm": 0.5303243398666382,
+      "learning_rate": 1.0460784562842351e-05,
+      "loss": 0.8745,
+      "step": 8915
+    },
+    {
+      "epoch": 0.7966582527308061,
+      "grad_norm": 0.40654778480529785,
+      "learning_rate": 1.0451928766702979e-05,
+      "loss": 0.9293,
+      "step": 8916
+    },
+    {
+      "epoch": 0.796747604262068,
+      "grad_norm": 0.47125494480133057,
+      "learning_rate": 1.0443076283136122e-05,
+      "loss": 0.9285,
+      "step": 8917
+    },
+    {
+      "epoch": 0.7968369557933299,
+      "grad_norm": 0.5250342488288879,
+      "learning_rate": 1.043422711288327e-05,
+      "loss": 0.9306,
+      "step": 8918
+    },
+    {
+      "epoch": 0.7969263073245918,
+      "grad_norm": 0.495063841342926,
+      "learning_rate": 1.042538125668564e-05,
+      "loss": 0.9515,
+      "step": 8919
+    },
+    {
+      "epoch": 0.7970156588558537,
+      "grad_norm": 0.5038367509841919,
+      "learning_rate": 1.041653871528418e-05,
+      "loss": 0.9353,
+      "step": 8920
+    },
+    {
+      "epoch": 0.7971050103871156,
+      "grad_norm": 0.40401169657707214,
+      "learning_rate": 1.040769948941952e-05,
+      "loss": 0.9591,
+      "step": 8921
+    },
+    {
+      "epoch": 0.7971943619183773,
+      "grad_norm": 0.5238271951675415,
+      "learning_rate": 1.0398863579832047e-05,
+      "loss": 0.9321,
+      "step": 8922
+    },
+    {
+      "epoch": 0.7972837134496392,
+      "grad_norm": 0.5171613097190857,
+      "learning_rate": 1.0390030987261866e-05,
+      "loss": 0.957,
+      "step": 8923
+    },
+    {
+      "epoch": 0.7973730649809011,
+      "grad_norm": 0.49921807646751404,
+      "learning_rate": 1.0381201712448806e-05,
+      "loss": 0.94,
+      "step": 8924
+    },
+    {
+      "epoch": 0.797462416512163,
+      "grad_norm": 0.6377806067466736,
+      "learning_rate": 1.0372375756132408e-05,
+      "loss": 0.8859,
+      "step": 8925
+    },
+    {
+      "epoch": 0.7975517680434249,
+      "grad_norm": 0.4824712872505188,
+      "learning_rate": 1.036355311905194e-05,
+      "loss": 0.9774,
+      "step": 8926
+    },
+    {
+      "epoch": 0.7976411195746868,
+      "grad_norm": 0.5878000855445862,
+      "learning_rate": 1.0354733801946393e-05,
+      "loss": 0.91,
+      "step": 8927
+    },
+    {
+      "epoch": 0.7977304711059485,
+      "grad_norm": 0.5659178495407104,
+      "learning_rate": 1.0345917805554478e-05,
+      "loss": 0.9632,
+      "step": 8928
+    },
+    {
+      "epoch": 0.7978198226372104,
+      "grad_norm": 0.5241888165473938,
+      "learning_rate": 1.0337105130614627e-05,
+      "loss": 0.9609,
+      "step": 8929
+    },
+    {
+      "epoch": 0.7979091741684723,
+      "grad_norm": 0.4879510998725891,
+      "learning_rate": 1.0328295777865016e-05,
+      "loss": 0.8878,
+      "step": 8930
+    },
+    {
+      "epoch": 0.7979985256997342,
+      "grad_norm": 0.5221325159072876,
+      "learning_rate": 1.0319489748043486e-05,
+      "loss": 0.8727,
+      "step": 8931
+    },
+    {
+      "epoch": 0.7980878772309961,
+      "grad_norm": 0.4458564519882202,
+      "learning_rate": 1.0310687041887652e-05,
+      "loss": 0.8967,
+      "step": 8932
+    },
+    {
+      "epoch": 0.7981772287622579,
+      "grad_norm": 0.6000443696975708,
+      "learning_rate": 1.0301887660134841e-05,
+      "loss": 0.91,
+      "step": 8933
+    },
+    {
+      "epoch": 0.7982665802935198,
+      "grad_norm": 0.4376446008682251,
+      "learning_rate": 1.0293091603522081e-05,
+      "loss": 0.9492,
+      "step": 8934
+    },
+    {
+      "epoch": 0.7983559318247816,
+      "grad_norm": 0.4281294643878937,
+      "learning_rate": 1.0284298872786153e-05,
+      "loss": 0.9444,
+      "step": 8935
+    },
+    {
+      "epoch": 0.7984452833560435,
+      "grad_norm": 0.5830131769180298,
+      "learning_rate": 1.027550946866353e-05,
+      "loss": 0.8794,
+      "step": 8936
+    },
+    {
+      "epoch": 0.7985346348873054,
+      "grad_norm": 0.45756039023399353,
+      "learning_rate": 1.026672339189042e-05,
+      "loss": 0.9598,
+      "step": 8937
+    },
+    {
+      "epoch": 0.7986239864185672,
+      "grad_norm": 0.42528459429740906,
+      "learning_rate": 1.0257940643202757e-05,
+      "loss": 1.0198,
+      "step": 8938
+    },
+    {
+      "epoch": 0.7987133379498291,
+      "grad_norm": 0.4692322015762329,
+      "learning_rate": 1.0249161223336185e-05,
+      "loss": 1.0496,
+      "step": 8939
+    },
+    {
+      "epoch": 0.798802689481091,
+      "grad_norm": 0.5667924284934998,
+      "learning_rate": 1.024038513302607e-05,
+      "loss": 0.913,
+      "step": 8940
+    },
+    {
+      "epoch": 0.7988920410123529,
+      "grad_norm": 0.47331857681274414,
+      "learning_rate": 1.0231612373007521e-05,
+      "loss": 0.9722,
+      "step": 8941
+    },
+    {
+      "epoch": 0.7989813925436147,
+      "grad_norm": 0.4151710867881775,
+      "learning_rate": 1.0222842944015326e-05,
+      "loss": 0.9374,
+      "step": 8942
+    },
+    {
+      "epoch": 0.7990707440748765,
+      "grad_norm": 0.582118570804596,
+      "learning_rate": 1.0214076846784027e-05,
+      "loss": 0.9451,
+      "step": 8943
+    },
+    {
+      "epoch": 0.7991600956061384,
+      "grad_norm": 0.46437790989875793,
+      "learning_rate": 1.0205314082047879e-05,
+      "loss": 0.9203,
+      "step": 8944
+    },
+    {
+      "epoch": 0.7992494471374003,
+      "grad_norm": 0.4480501711368561,
+      "learning_rate": 1.0196554650540857e-05,
+      "loss": 0.9466,
+      "step": 8945
+    },
+    {
+      "epoch": 0.7993387986686622,
+      "grad_norm": 0.5746645331382751,
+      "learning_rate": 1.0187798552996653e-05,
+      "loss": 0.8555,
+      "step": 8946
+    },
+    {
+      "epoch": 0.7994281501999241,
+      "grad_norm": 0.4533185064792633,
+      "learning_rate": 1.0179045790148707e-05,
+      "loss": 0.9209,
+      "step": 8947
+    },
+    {
+      "epoch": 0.799517501731186,
+      "grad_norm": 0.4475330412387848,
+      "learning_rate": 1.0170296362730125e-05,
+      "loss": 0.9418,
+      "step": 8948
+    },
+    {
+      "epoch": 0.7996068532624477,
+      "grad_norm": 0.4576277434825897,
+      "learning_rate": 1.0161550271473774e-05,
+      "loss": 0.9764,
+      "step": 8949
+    },
+    {
+      "epoch": 0.7996962047937096,
+      "grad_norm": 0.5483040809631348,
+      "learning_rate": 1.0152807517112233e-05,
+      "loss": 0.9551,
+      "step": 8950
+    },
+    {
+      "epoch": 0.7997855563249715,
+      "grad_norm": 0.49113574624061584,
+      "learning_rate": 1.0144068100377818e-05,
+      "loss": 0.9638,
+      "step": 8951
+    },
+    {
+      "epoch": 0.7998749078562334,
+      "grad_norm": 0.4609823226928711,
+      "learning_rate": 1.013533202200252e-05,
+      "loss": 0.949,
+      "step": 8952
+    },
+    {
+      "epoch": 0.7999642593874953,
+      "grad_norm": 0.3998883068561554,
+      "learning_rate": 1.0126599282718096e-05,
+      "loss": 1.0431,
+      "step": 8953
+    },
+    {
+      "epoch": 0.8000536109187572,
+      "grad_norm": 0.6044648289680481,
+      "learning_rate": 1.0117869883255998e-05,
+      "loss": 0.8911,
+      "step": 8954
+    },
+    {
+      "epoch": 0.800142962450019,
+      "grad_norm": 0.4625280499458313,
+      "learning_rate": 1.0109143824347411e-05,
+      "loss": 0.9561,
+      "step": 8955
+    },
+    {
+      "epoch": 0.8002323139812808,
+      "grad_norm": 0.45291417837142944,
+      "learning_rate": 1.0100421106723234e-05,
+      "loss": 0.9568,
+      "step": 8956
+    },
+    {
+      "epoch": 0.8003216655125427,
+      "grad_norm": 0.4366665780544281,
+      "learning_rate": 1.009170173111409e-05,
+      "loss": 1.0054,
+      "step": 8957
+    },
+    {
+      "epoch": 0.8004110170438046,
+      "grad_norm": 0.5513303875923157,
+      "learning_rate": 1.0082985698250313e-05,
+      "loss": 0.9255,
+      "step": 8958
+    },
+    {
+      "epoch": 0.8005003685750665,
+      "grad_norm": 0.42455315589904785,
+      "learning_rate": 1.0074273008861973e-05,
+      "loss": 0.9084,
+      "step": 8959
+    },
+    {
+      "epoch": 0.8005897201063283,
+      "grad_norm": 0.4331103563308716,
+      "learning_rate": 1.0065563663678845e-05,
+      "loss": 0.9705,
+      "step": 8960
+    },
+    {
+      "epoch": 0.8006790716375902,
+      "grad_norm": 0.4532991349697113,
+      "learning_rate": 1.005685766343044e-05,
+      "loss": 0.8854,
+      "step": 8961
+    },
+    {
+      "epoch": 0.8007684231688521,
+      "grad_norm": 0.4681682288646698,
+      "learning_rate": 1.0048155008845962e-05,
+      "loss": 0.9396,
+      "step": 8962
+    },
+    {
+      "epoch": 0.8008577747001139,
+      "grad_norm": 0.4511670172214508,
+      "learning_rate": 1.003945570065436e-05,
+      "loss": 0.8475,
+      "step": 8963
+    },
+    {
+      "epoch": 0.8009471262313758,
+      "grad_norm": 0.475967675447464,
+      "learning_rate": 1.0030759739584284e-05,
+      "loss": 0.9374,
+      "step": 8964
+    },
+    {
+      "epoch": 0.8010364777626376,
+      "grad_norm": 0.547709047794342,
+      "learning_rate": 1.0022067126364126e-05,
+      "loss": 1.0038,
+      "step": 8965
+    },
+    {
+      "epoch": 0.8011258292938995,
+      "grad_norm": 0.44402167201042175,
+      "learning_rate": 1.001337786172198e-05,
+      "loss": 0.9718,
+      "step": 8966
+    },
+    {
+      "epoch": 0.8012151808251614,
+      "grad_norm": 0.40516430139541626,
+      "learning_rate": 1.0004691946385663e-05,
+      "loss": 1.0515,
+      "step": 8967
+    },
+    {
+      "epoch": 0.8013045323564233,
+      "grad_norm": 0.4301491975784302,
+      "learning_rate": 9.996009381082717e-06,
+      "loss": 0.9238,
+      "step": 8968
+    },
+    {
+      "epoch": 0.8013938838876852,
+      "grad_norm": 0.5374286770820618,
+      "learning_rate": 9.987330166540398e-06,
+      "loss": 0.9712,
+      "step": 8969
+    },
+    {
+      "epoch": 0.801483235418947,
+      "grad_norm": 0.453789085149765,
+      "learning_rate": 9.97865430348568e-06,
+      "loss": 0.9511,
+      "step": 8970
+    },
+    {
+      "epoch": 0.8015725869502088,
+      "grad_norm": 0.5180380344390869,
+      "learning_rate": 9.969981792645273e-06,
+      "loss": 0.8728,
+      "step": 8971
+    },
+    {
+      "epoch": 0.8016619384814707,
+      "grad_norm": 0.46428194642066956,
+      "learning_rate": 9.961312634745567e-06,
+      "loss": 0.95,
+      "step": 8972
+    },
+    {
+      "epoch": 0.8017512900127326,
+      "grad_norm": 0.5181485414505005,
+      "learning_rate": 9.952646830512712e-06,
+      "loss": 0.9228,
+      "step": 8973
+    },
+    {
+      "epoch": 0.8018406415439945,
+      "grad_norm": 0.44874072074890137,
+      "learning_rate": 9.94398438067256e-06,
+      "loss": 0.9679,
+      "step": 8974
+    },
+    {
+      "epoch": 0.8019299930752564,
+      "grad_norm": 0.47210460901260376,
+      "learning_rate": 9.93532528595068e-06,
+      "loss": 0.9655,
+      "step": 8975
+    },
+    {
+      "epoch": 0.8020193446065182,
+      "grad_norm": 0.5403384566307068,
+      "learning_rate": 9.926669547072364e-06,
+      "loss": 0.9232,
+      "step": 8976
+    },
+    {
+      "epoch": 0.80210869613778,
+      "grad_norm": 0.4292760193347931,
+      "learning_rate": 9.918017164762627e-06,
+      "loss": 0.9782,
+      "step": 8977
+    },
+    {
+      "epoch": 0.8021980476690419,
+      "grad_norm": 0.4274854362010956,
+      "learning_rate": 9.909368139746194e-06,
+      "loss": 0.9455,
+      "step": 8978
+    },
+    {
+      "epoch": 0.8022873992003038,
+      "grad_norm": 0.4722135663032532,
+      "learning_rate": 9.90072247274752e-06,
+      "loss": 1.0058,
+      "step": 8979
+    },
+    {
+      "epoch": 0.8023767507315657,
+      "grad_norm": 0.4381501376628876,
+      "learning_rate": 9.892080164490774e-06,
+      "loss": 0.9518,
+      "step": 8980
+    },
+    {
+      "epoch": 0.8024661022628276,
+      "grad_norm": 0.488213449716568,
+      "learning_rate": 9.883441215699823e-06,
+      "loss": 0.9261,
+      "step": 8981
+    },
+    {
+      "epoch": 0.8025554537940894,
+      "grad_norm": 0.5511339902877808,
+      "learning_rate": 9.874805627098282e-06,
+      "loss": 0.9793,
+      "step": 8982
+    },
+    {
+      "epoch": 0.8026448053253513,
+      "grad_norm": 0.5285826921463013,
+      "learning_rate": 9.866173399409474e-06,
+      "loss": 0.9814,
+      "step": 8983
+    },
+    {
+      "epoch": 0.8027341568566131,
+      "grad_norm": 0.5300377607345581,
+      "learning_rate": 9.857544533356456e-06,
+      "loss": 1.0165,
+      "step": 8984
+    },
+    {
+      "epoch": 0.802823508387875,
+      "grad_norm": 0.4703865051269531,
+      "learning_rate": 9.848919029661952e-06,
+      "loss": 0.8869,
+      "step": 8985
+    },
+    {
+      "epoch": 0.8029128599191369,
+      "grad_norm": 0.5108640193939209,
+      "learning_rate": 9.840296889048462e-06,
+      "loss": 0.9011,
+      "step": 8986
+    },
+    {
+      "epoch": 0.8030022114503987,
+      "grad_norm": 0.49105626344680786,
+      "learning_rate": 9.831678112238173e-06,
+      "loss": 0.8887,
+      "step": 8987
+    },
+    {
+      "epoch": 0.8030915629816606,
+      "grad_norm": 0.5277919769287109,
+      "learning_rate": 9.823062699953012e-06,
+      "loss": 0.944,
+      "step": 8988
+    },
+    {
+      "epoch": 0.8031809145129225,
+      "grad_norm": 0.4541589021682739,
+      "learning_rate": 9.81445065291461e-06,
+      "loss": 0.9229,
+      "step": 8989
+    },
+    {
+      "epoch": 0.8032702660441843,
+      "grad_norm": 0.4479905366897583,
+      "learning_rate": 9.805841971844305e-06,
+      "loss": 0.9027,
+      "step": 8990
+    },
+    {
+      "epoch": 0.8033596175754462,
+      "grad_norm": 0.4373217225074768,
+      "learning_rate": 9.797236657463188e-06,
+      "loss": 0.9373,
+      "step": 8991
+    },
+    {
+      "epoch": 0.803448969106708,
+      "grad_norm": 0.407236248254776,
+      "learning_rate": 9.788634710492017e-06,
+      "loss": 0.9314,
+      "step": 8992
+    },
+    {
+      "epoch": 0.8035383206379699,
+      "grad_norm": 0.5327048301696777,
+      "learning_rate": 9.780036131651315e-06,
+      "loss": 0.9155,
+      "step": 8993
+    },
+    {
+      "epoch": 0.8036276721692318,
+      "grad_norm": 0.5354421138763428,
+      "learning_rate": 9.7714409216613e-06,
+      "loss": 0.8684,
+      "step": 8994
+    },
+    {
+      "epoch": 0.8037170237004937,
+      "grad_norm": 0.5037555694580078,
+      "learning_rate": 9.762849081241915e-06,
+      "loss": 0.9443,
+      "step": 8995
+    },
+    {
+      "epoch": 0.8038063752317556,
+      "grad_norm": 0.5367538928985596,
+      "learning_rate": 9.754260611112814e-06,
+      "loss": 0.9278,
+      "step": 8996
+    },
+    {
+      "epoch": 0.8038957267630173,
+      "grad_norm": 0.4558742046356201,
+      "learning_rate": 9.74567551199338e-06,
+      "loss": 0.848,
+      "step": 8997
+    },
+    {
+      "epoch": 0.8039850782942792,
+      "grad_norm": 0.45363011956214905,
+      "learning_rate": 9.737093784602697e-06,
+      "loss": 0.9489,
+      "step": 8998
+    },
+    {
+      "epoch": 0.8040744298255411,
+      "grad_norm": 0.4492855668067932,
+      "learning_rate": 9.728515429659585e-06,
+      "loss": 0.9211,
+      "step": 8999
+    },
+    {
+      "epoch": 0.804163781356803,
+      "grad_norm": 0.44268351793289185,
+      "learning_rate": 9.719940447882563e-06,
+      "loss": 0.9763,
+      "step": 9000
+    },
+    {
+      "epoch": 0.8042531328880649,
+      "grad_norm": 0.4664279520511627,
+      "learning_rate": 9.711368839989905e-06,
+      "loss": 0.9177,
+      "step": 9001
+    },
+    {
+      "epoch": 0.8043424844193268,
+      "grad_norm": 0.4780477285385132,
+      "learning_rate": 9.70280060669953e-06,
+      "loss": 0.8817,
+      "step": 9002
+    },
+    {
+      "epoch": 0.8044318359505886,
+      "grad_norm": 0.5579407811164856,
+      "learning_rate": 9.694235748729146e-06,
+      "loss": 0.8823,
+      "step": 9003
+    },
+    {
+      "epoch": 0.8045211874818504,
+      "grad_norm": 0.47675004601478577,
+      "learning_rate": 9.685674266796146e-06,
+      "loss": 0.9631,
+      "step": 9004
+    },
+    {
+      "epoch": 0.8046105390131123,
+      "grad_norm": 0.3818524479866028,
+      "learning_rate": 9.67711616161765e-06,
+      "loss": 0.937,
+      "step": 9005
+    },
+    {
+      "epoch": 0.8046998905443742,
+      "grad_norm": 0.4102923572063446,
+      "learning_rate": 9.668561433910484e-06,
+      "loss": 0.9733,
+      "step": 9006
+    },
+    {
+      "epoch": 0.8047892420756361,
+      "grad_norm": 0.41497570276260376,
+      "learning_rate": 9.660010084391197e-06,
+      "loss": 1.0069,
+      "step": 9007
+    },
+    {
+      "epoch": 0.804878593606898,
+      "grad_norm": 0.6005346775054932,
+      "learning_rate": 9.65146211377606e-06,
+      "loss": 0.925,
+      "step": 9008
+    },
+    {
+      "epoch": 0.8049679451381598,
+      "grad_norm": 0.5480002164840698,
+      "learning_rate": 9.642917522781058e-06,
+      "loss": 0.9914,
+      "step": 9009
+    },
+    {
+      "epoch": 0.8050572966694217,
+      "grad_norm": 0.5394856333732605,
+      "learning_rate": 9.63437631212189e-06,
+      "loss": 0.8734,
+      "step": 9010
+    },
+    {
+      "epoch": 0.8051466482006835,
+      "grad_norm": 0.48310962319374084,
+      "learning_rate": 9.625838482513983e-06,
+      "loss": 1.0201,
+      "step": 9011
+    },
+    {
+      "epoch": 0.8052359997319454,
+      "grad_norm": 0.485075443983078,
+      "learning_rate": 9.617304034672448e-06,
+      "loss": 0.916,
+      "step": 9012
+    },
+    {
+      "epoch": 0.8053253512632073,
+      "grad_norm": 0.5205706357955933,
+      "learning_rate": 9.608772969312147e-06,
+      "loss": 0.9511,
+      "step": 9013
+    },
+    {
+      "epoch": 0.8054147027944691,
+      "grad_norm": 0.48902761936187744,
+      "learning_rate": 9.600245287147652e-06,
+      "loss": 0.896,
+      "step": 9014
+    },
+    {
+      "epoch": 0.805504054325731,
+      "grad_norm": 0.5032730102539062,
+      "learning_rate": 9.591720988893244e-06,
+      "loss": 0.9751,
+      "step": 9015
+    },
+    {
+      "epoch": 0.8055934058569929,
+      "grad_norm": 0.537691593170166,
+      "learning_rate": 9.583200075262921e-06,
+      "loss": 0.983,
+      "step": 9016
+    },
+    {
+      "epoch": 0.8056827573882548,
+      "grad_norm": 0.5454861521720886,
+      "learning_rate": 9.57468254697042e-06,
+      "loss": 0.9127,
+      "step": 9017
+    },
+    {
+      "epoch": 0.8057721089195166,
+      "grad_norm": 0.4477491080760956,
+      "learning_rate": 9.566168404729148e-06,
+      "loss": 0.995,
+      "step": 9018
+    },
+    {
+      "epoch": 0.8058614604507784,
+      "grad_norm": 0.5274454951286316,
+      "learning_rate": 9.55765764925226e-06,
+      "loss": 0.8299,
+      "step": 9019
+    },
+    {
+      "epoch": 0.8059508119820403,
+      "grad_norm": 0.5348902940750122,
+      "learning_rate": 9.549150281252633e-06,
+      "loss": 0.9079,
+      "step": 9020
+    },
+    {
+      "epoch": 0.8060401635133022,
+      "grad_norm": 0.5387945771217346,
+      "learning_rate": 9.54064630144284e-06,
+      "loss": 0.9053,
+      "step": 9021
+    },
+    {
+      "epoch": 0.8061295150445641,
+      "grad_norm": 0.5068916082382202,
+      "learning_rate": 9.532145710535207e-06,
+      "loss": 0.926,
+      "step": 9022
+    },
+    {
+      "epoch": 0.806218866575826,
+      "grad_norm": 0.39943739771842957,
+      "learning_rate": 9.523648509241706e-06,
+      "loss": 0.9773,
+      "step": 9023
+    },
+    {
+      "epoch": 0.8063082181070879,
+      "grad_norm": 0.676214873790741,
+      "learning_rate": 9.515154698274093e-06,
+      "loss": 0.8005,
+      "step": 9024
+    },
+    {
+      "epoch": 0.8063975696383496,
+      "grad_norm": 0.45418787002563477,
+      "learning_rate": 9.506664278343808e-06,
+      "loss": 0.9662,
+      "step": 9025
+    },
+    {
+      "epoch": 0.8064869211696115,
+      "grad_norm": 0.5260841846466064,
+      "learning_rate": 9.498177250162022e-06,
+      "loss": 1.0117,
+      "step": 9026
+    },
+    {
+      "epoch": 0.8065762727008734,
+      "grad_norm": 0.502400815486908,
+      "learning_rate": 9.489693614439605e-06,
+      "loss": 0.8963,
+      "step": 9027
+    },
+    {
+      "epoch": 0.8066656242321353,
+      "grad_norm": 0.5157961249351501,
+      "learning_rate": 9.48121337188716e-06,
+      "loss": 0.9942,
+      "step": 9028
+    },
+    {
+      "epoch": 0.8067549757633972,
+      "grad_norm": 0.6219414472579956,
+      "learning_rate": 9.472736523214993e-06,
+      "loss": 0.9617,
+      "step": 9029
+    },
+    {
+      "epoch": 0.806844327294659,
+      "grad_norm": 0.41551846265792847,
+      "learning_rate": 9.464263069133134e-06,
+      "loss": 0.9622,
+      "step": 9030
+    },
+    {
+      "epoch": 0.8069336788259209,
+      "grad_norm": 0.454323947429657,
+      "learning_rate": 9.455793010351321e-06,
+      "loss": 0.9332,
+      "step": 9031
+    },
+    {
+      "epoch": 0.8070230303571827,
+      "grad_norm": 0.4970146119594574,
+      "learning_rate": 9.447326347579028e-06,
+      "loss": 0.9591,
+      "step": 9032
+    },
+    {
+      "epoch": 0.8071123818884446,
+      "grad_norm": 0.44244319200515747,
+      "learning_rate": 9.438863081525396e-06,
+      "loss": 0.9451,
+      "step": 9033
+    },
+    {
+      "epoch": 0.8072017334197065,
+      "grad_norm": 0.47928622364997864,
+      "learning_rate": 9.430403212899336e-06,
+      "loss": 0.8969,
+      "step": 9034
+    },
+    {
+      "epoch": 0.8072910849509684,
+      "grad_norm": 0.45787954330444336,
+      "learning_rate": 9.421946742409448e-06,
+      "loss": 0.9851,
+      "step": 9035
+    },
+    {
+      "epoch": 0.8073804364822302,
+      "grad_norm": 0.603981077671051,
+      "learning_rate": 9.41349367076405e-06,
+      "loss": 0.9052,
+      "step": 9036
+    },
+    {
+      "epoch": 0.8074697880134921,
+      "grad_norm": 0.47599250078201294,
+      "learning_rate": 9.405043998671176e-06,
+      "loss": 0.9525,
+      "step": 9037
+    },
+    {
+      "epoch": 0.807559139544754,
+      "grad_norm": 0.4036829471588135,
+      "learning_rate": 9.39659772683858e-06,
+      "loss": 0.9176,
+      "step": 9038
+    },
+    {
+      "epoch": 0.8076484910760158,
+      "grad_norm": 0.49439674615859985,
+      "learning_rate": 9.38815485597373e-06,
+      "loss": 0.9366,
+      "step": 9039
+    },
+    {
+      "epoch": 0.8077378426072777,
+      "grad_norm": 0.5236937403678894,
+      "learning_rate": 9.379715386783794e-06,
+      "loss": 0.9086,
+      "step": 9040
+    },
+    {
+      "epoch": 0.8078271941385395,
+      "grad_norm": 0.4477667808532715,
+      "learning_rate": 9.371279319975678e-06,
+      "loss": 0.9332,
+      "step": 9041
+    },
+    {
+      "epoch": 0.8079165456698014,
+      "grad_norm": 0.509158730506897,
+      "learning_rate": 9.362846656256008e-06,
+      "loss": 0.9012,
+      "step": 9042
+    },
+    {
+      "epoch": 0.8080058972010633,
+      "grad_norm": 0.43936219811439514,
+      "learning_rate": 9.354417396331073e-06,
+      "loss": 0.9957,
+      "step": 9043
+    },
+    {
+      "epoch": 0.8080952487323252,
+      "grad_norm": 0.5267393589019775,
+      "learning_rate": 9.345991540906934e-06,
+      "loss": 0.8568,
+      "step": 9044
+    },
+    {
+      "epoch": 0.8081846002635871,
+      "grad_norm": 0.5685756802558899,
+      "learning_rate": 9.337569090689346e-06,
+      "loss": 0.9036,
+      "step": 9045
+    },
+    {
+      "epoch": 0.8082739517948488,
+      "grad_norm": 0.5792932510375977,
+      "learning_rate": 9.329150046383772e-06,
+      "loss": 0.8976,
+      "step": 9046
+    },
+    {
+      "epoch": 0.8083633033261107,
+      "grad_norm": 0.4724472761154175,
+      "learning_rate": 9.320734408695403e-06,
+      "loss": 0.8995,
+      "step": 9047
+    },
+    {
+      "epoch": 0.8084526548573726,
+      "grad_norm": 0.4375053346157074,
+      "learning_rate": 9.312322178329142e-06,
+      "loss": 1.06,
+      "step": 9048
+    },
+    {
+      "epoch": 0.8085420063886345,
+      "grad_norm": 0.6214897036552429,
+      "learning_rate": 9.303913355989596e-06,
+      "loss": 0.9019,
+      "step": 9049
+    },
+    {
+      "epoch": 0.8086313579198964,
+      "grad_norm": 0.49101778864860535,
+      "learning_rate": 9.295507942381103e-06,
+      "loss": 0.9121,
+      "step": 9050
+    },
+    {
+      "epoch": 0.8087207094511583,
+      "grad_norm": 0.4864928424358368,
+      "learning_rate": 9.28710593820769e-06,
+      "loss": 0.8843,
+      "step": 9051
+    },
+    {
+      "epoch": 0.80881006098242,
+      "grad_norm": 0.4251103401184082,
+      "learning_rate": 9.278707344173126e-06,
+      "loss": 0.9909,
+      "step": 9052
+    },
+    {
+      "epoch": 0.8088994125136819,
+      "grad_norm": 0.5644087195396423,
+      "learning_rate": 9.270312160980876e-06,
+      "loss": 0.9755,
+      "step": 9053
+    },
+    {
+      "epoch": 0.8089887640449438,
+      "grad_norm": 0.42497533559799194,
+      "learning_rate": 9.261920389334139e-06,
+      "loss": 0.9219,
+      "step": 9054
+    },
+    {
+      "epoch": 0.8090781155762057,
+      "grad_norm": 0.37208980321884155,
+      "learning_rate": 9.253532029935796e-06,
+      "loss": 0.9487,
+      "step": 9055
+    },
+    {
+      "epoch": 0.8091674671074676,
+      "grad_norm": 0.5022962689399719,
+      "learning_rate": 9.245147083488476e-06,
+      "loss": 0.9539,
+      "step": 9056
+    },
+    {
+      "epoch": 0.8092568186387294,
+      "grad_norm": 0.4653746485710144,
+      "learning_rate": 9.236765550694498e-06,
+      "loss": 1.0132,
+      "step": 9057
+    },
+    {
+      "epoch": 0.8093461701699913,
+      "grad_norm": 0.4346064329147339,
+      "learning_rate": 9.228387432255915e-06,
+      "loss": 0.9706,
+      "step": 9058
+    },
+    {
+      "epoch": 0.8094355217012531,
+      "grad_norm": 0.6596688032150269,
+      "learning_rate": 9.220012728874473e-06,
+      "loss": 0.8905,
+      "step": 9059
+    },
+    {
+      "epoch": 0.809524873232515,
+      "grad_norm": 0.5736172795295715,
+      "learning_rate": 9.21164144125165e-06,
+      "loss": 0.9162,
+      "step": 9060
+    },
+    {
+      "epoch": 0.8096142247637769,
+      "grad_norm": 0.43389225006103516,
+      "learning_rate": 9.203273570088634e-06,
+      "loss": 0.9322,
+      "step": 9061
+    },
+    {
+      "epoch": 0.8097035762950388,
+      "grad_norm": 0.516688346862793,
+      "learning_rate": 9.194909116086325e-06,
+      "loss": 0.9036,
+      "step": 9062
+    },
+    {
+      "epoch": 0.8097929278263006,
+      "grad_norm": 0.4250909984111786,
+      "learning_rate": 9.186548079945318e-06,
+      "loss": 0.9199,
+      "step": 9063
+    },
+    {
+      "epoch": 0.8098822793575625,
+      "grad_norm": 0.5514670014381409,
+      "learning_rate": 9.178190462365949e-06,
+      "loss": 0.9008,
+      "step": 9064
+    },
+    {
+      "epoch": 0.8099716308888244,
+      "grad_norm": 0.529180109500885,
+      "learning_rate": 9.169836264048259e-06,
+      "loss": 0.8523,
+      "step": 9065
+    },
+    {
+      "epoch": 0.8100609824200862,
+      "grad_norm": 0.4616588056087494,
+      "learning_rate": 9.161485485692001e-06,
+      "loss": 0.9791,
+      "step": 9066
+    },
+    {
+      "epoch": 0.8101503339513481,
+      "grad_norm": 0.5188908576965332,
+      "learning_rate": 9.15313812799664e-06,
+      "loss": 0.9484,
+      "step": 9067
+    },
+    {
+      "epoch": 0.8102396854826099,
+      "grad_norm": 0.43526217341423035,
+      "learning_rate": 9.144794191661355e-06,
+      "loss": 0.9121,
+      "step": 9068
+    },
+    {
+      "epoch": 0.8103290370138718,
+      "grad_norm": 0.44872915744781494,
+      "learning_rate": 9.136453677385043e-06,
+      "loss": 1.0135,
+      "step": 9069
+    },
+    {
+      "epoch": 0.8104183885451337,
+      "grad_norm": 0.4532172679901123,
+      "learning_rate": 9.128116585866308e-06,
+      "loss": 0.9565,
+      "step": 9070
+    },
+    {
+      "epoch": 0.8105077400763956,
+      "grad_norm": 0.5452584028244019,
+      "learning_rate": 9.119782917803476e-06,
+      "loss": 0.9311,
+      "step": 9071
+    },
+    {
+      "epoch": 0.8105970916076575,
+      "grad_norm": 0.5876386165618896,
+      "learning_rate": 9.111452673894588e-06,
+      "loss": 0.8881,
+      "step": 9072
+    },
+    {
+      "epoch": 0.8106864431389192,
+      "grad_norm": 0.4356098175048828,
+      "learning_rate": 9.103125854837362e-06,
+      "loss": 0.9527,
+      "step": 9073
+    },
+    {
+      "epoch": 0.8107757946701811,
+      "grad_norm": 0.47350767254829407,
+      "learning_rate": 9.094802461329277e-06,
+      "loss": 0.9068,
+      "step": 9074
+    },
+    {
+      "epoch": 0.810865146201443,
+      "grad_norm": 0.4826947748661041,
+      "learning_rate": 9.086482494067506e-06,
+      "loss": 0.8969,
+      "step": 9075
+    },
+    {
+      "epoch": 0.8109544977327049,
+      "grad_norm": 0.5265054702758789,
+      "learning_rate": 9.078165953748936e-06,
+      "loss": 0.9365,
+      "step": 9076
+    },
+    {
+      "epoch": 0.8110438492639668,
+      "grad_norm": 0.41897475719451904,
+      "learning_rate": 9.069852841070159e-06,
+      "loss": 0.9803,
+      "step": 9077
+    },
+    {
+      "epoch": 0.8111332007952287,
+      "grad_norm": 0.4707808196544647,
+      "learning_rate": 9.061543156727486e-06,
+      "loss": 0.8801,
+      "step": 9078
+    },
+    {
+      "epoch": 0.8112225523264905,
+      "grad_norm": 0.5158072710037231,
+      "learning_rate": 9.05323690141695e-06,
+      "loss": 0.9694,
+      "step": 9079
+    },
+    {
+      "epoch": 0.8113119038577523,
+      "grad_norm": 0.47056683897972107,
+      "learning_rate": 9.044934075834289e-06,
+      "loss": 0.9452,
+      "step": 9080
+    },
+    {
+      "epoch": 0.8114012553890142,
+      "grad_norm": 0.5674601793289185,
+      "learning_rate": 9.036634680674943e-06,
+      "loss": 0.8907,
+      "step": 9081
+    },
+    {
+      "epoch": 0.8114906069202761,
+      "grad_norm": 0.44375982880592346,
+      "learning_rate": 9.028338716634094e-06,
+      "loss": 1.0265,
+      "step": 9082
+    },
+    {
+      "epoch": 0.811579958451538,
+      "grad_norm": 0.4866836965084076,
+      "learning_rate": 9.020046184406593e-06,
+      "loss": 0.9754,
+      "step": 9083
+    },
+    {
+      "epoch": 0.8116693099827998,
+      "grad_norm": 0.42839664220809937,
+      "learning_rate": 9.01175708468704e-06,
+      "loss": 0.9104,
+      "step": 9084
+    },
+    {
+      "epoch": 0.8117586615140617,
+      "grad_norm": 0.4285668432712555,
+      "learning_rate": 9.003471418169734e-06,
+      "loss": 0.9743,
+      "step": 9085
+    },
+    {
+      "epoch": 0.8118480130453236,
+      "grad_norm": 0.6068586111068726,
+      "learning_rate": 8.995189185548686e-06,
+      "loss": 0.978,
+      "step": 9086
+    },
+    {
+      "epoch": 0.8119373645765854,
+      "grad_norm": 0.4698917865753174,
+      "learning_rate": 8.986910387517639e-06,
+      "loss": 0.9434,
+      "step": 9087
+    },
+    {
+      "epoch": 0.8120267161078473,
+      "grad_norm": 0.4826454222202301,
+      "learning_rate": 8.97863502477e-06,
+      "loss": 0.9581,
+      "step": 9088
+    },
+    {
+      "epoch": 0.8121160676391092,
+      "grad_norm": 0.5964739322662354,
+      "learning_rate": 8.970363097998936e-06,
+      "loss": 0.8737,
+      "step": 9089
+    },
+    {
+      "epoch": 0.812205419170371,
+      "grad_norm": 0.48209187388420105,
+      "learning_rate": 8.962094607897303e-06,
+      "loss": 0.8962,
+      "step": 9090
+    },
+    {
+      "epoch": 0.8122947707016329,
+      "grad_norm": 0.5253683924674988,
+      "learning_rate": 8.953829555157684e-06,
+      "loss": 0.874,
+      "step": 9091
+    },
+    {
+      "epoch": 0.8123841222328948,
+      "grad_norm": 0.3659655451774597,
+      "learning_rate": 8.945567940472371e-06,
+      "loss": 1.0097,
+      "step": 9092
+    },
+    {
+      "epoch": 0.8124734737641567,
+      "grad_norm": 0.5746763944625854,
+      "learning_rate": 8.937309764533335e-06,
+      "loss": 0.8948,
+      "step": 9093
+    },
+    {
+      "epoch": 0.8125628252954185,
+      "grad_norm": 0.5120271444320679,
+      "learning_rate": 8.929055028032312e-06,
+      "loss": 0.9083,
+      "step": 9094
+    },
+    {
+      "epoch": 0.8126521768266803,
+      "grad_norm": 0.4482232928276062,
+      "learning_rate": 8.92080373166071e-06,
+      "loss": 0.8775,
+      "step": 9095
+    },
+    {
+      "epoch": 0.8127415283579422,
+      "grad_norm": 0.7028616070747375,
+      "learning_rate": 8.912555876109668e-06,
+      "loss": 0.8673,
+      "step": 9096
+    },
+    {
+      "epoch": 0.8128308798892041,
+      "grad_norm": 0.5362510681152344,
+      "learning_rate": 8.904311462070031e-06,
+      "loss": 0.8818,
+      "step": 9097
+    },
+    {
+      "epoch": 0.812920231420466,
+      "grad_norm": 0.42716673016548157,
+      "learning_rate": 8.89607049023236e-06,
+      "loss": 0.9894,
+      "step": 9098
+    },
+    {
+      "epoch": 0.8130095829517279,
+      "grad_norm": 0.6123215556144714,
+      "learning_rate": 8.887832961286923e-06,
+      "loss": 0.9649,
+      "step": 9099
+    },
+    {
+      "epoch": 0.8130989344829898,
+      "grad_norm": 0.4452812671661377,
+      "learning_rate": 8.8795988759237e-06,
+      "loss": 0.9172,
+      "step": 9100
+    },
+    {
+      "epoch": 0.8131882860142515,
+      "grad_norm": 0.49280375242233276,
+      "learning_rate": 8.871368234832378e-06,
+      "loss": 0.9344,
+      "step": 9101
+    },
+    {
+      "epoch": 0.8132776375455134,
+      "grad_norm": 0.4815100431442261,
+      "learning_rate": 8.863141038702382e-06,
+      "loss": 0.8467,
+      "step": 9102
+    },
+    {
+      "epoch": 0.8133669890767753,
+      "grad_norm": 0.48964032530784607,
+      "learning_rate": 8.854917288222803e-06,
+      "loss": 0.9917,
+      "step": 9103
+    },
+    {
+      "epoch": 0.8134563406080372,
+      "grad_norm": 0.45770344138145447,
+      "learning_rate": 8.84669698408247e-06,
+      "loss": 0.9124,
+      "step": 9104
+    },
+    {
+      "epoch": 0.8135456921392991,
+      "grad_norm": 0.49135828018188477,
+      "learning_rate": 8.838480126969928e-06,
+      "loss": 0.8979,
+      "step": 9105
+    },
+    {
+      "epoch": 0.8136350436705609,
+      "grad_norm": 0.44308167695999146,
+      "learning_rate": 8.830266717573427e-06,
+      "loss": 1.0013,
+      "step": 9106
+    },
+    {
+      "epoch": 0.8137243952018228,
+      "grad_norm": 0.4538305699825287,
+      "learning_rate": 8.822056756580926e-06,
+      "loss": 0.9082,
+      "step": 9107
+    },
+    {
+      "epoch": 0.8138137467330846,
+      "grad_norm": 0.5253844857215881,
+      "learning_rate": 8.8138502446801e-06,
+      "loss": 0.8479,
+      "step": 9108
+    },
+    {
+      "epoch": 0.8139030982643465,
+      "grad_norm": 0.47331321239471436,
+      "learning_rate": 8.805647182558324e-06,
+      "loss": 0.8898,
+      "step": 9109
+    },
+    {
+      "epoch": 0.8139924497956084,
+      "grad_norm": 0.5414149761199951,
+      "learning_rate": 8.797447570902695e-06,
+      "loss": 0.9196,
+      "step": 9110
+    },
+    {
+      "epoch": 0.8140818013268702,
+      "grad_norm": 0.45004722476005554,
+      "learning_rate": 8.789251410400023e-06,
+      "loss": 0.94,
+      "step": 9111
+    },
+    {
+      "epoch": 0.8141711528581321,
+      "grad_norm": 0.4916594624519348,
+      "learning_rate": 8.781058701736822e-06,
+      "loss": 0.9332,
+      "step": 9112
+    },
+    {
+      "epoch": 0.814260504389394,
+      "grad_norm": 0.4902633726596832,
+      "learning_rate": 8.77286944559933e-06,
+      "loss": 0.8747,
+      "step": 9113
+    },
+    {
+      "epoch": 0.8143498559206558,
+      "grad_norm": 0.5728904008865356,
+      "learning_rate": 8.764683642673455e-06,
+      "loss": 0.8908,
+      "step": 9114
+    },
+    {
+      "epoch": 0.8144392074519177,
+      "grad_norm": 0.5360121726989746,
+      "learning_rate": 8.756501293644865e-06,
+      "loss": 0.8705,
+      "step": 9115
+    },
+    {
+      "epoch": 0.8145285589831796,
+      "grad_norm": 0.5026247501373291,
+      "learning_rate": 8.748322399198916e-06,
+      "loss": 1.0092,
+      "step": 9116
+    },
+    {
+      "epoch": 0.8146179105144414,
+      "grad_norm": 0.46945619583129883,
+      "learning_rate": 8.740146960020673e-06,
+      "loss": 0.9574,
+      "step": 9117
+    },
+    {
+      "epoch": 0.8147072620457033,
+      "grad_norm": 0.4218684434890747,
+      "learning_rate": 8.731974976794926e-06,
+      "loss": 0.9714,
+      "step": 9118
+    },
+    {
+      "epoch": 0.8147966135769652,
+      "grad_norm": 0.5485289692878723,
+      "learning_rate": 8.723806450206157e-06,
+      "loss": 0.9748,
+      "step": 9119
+    },
+    {
+      "epoch": 0.8148859651082271,
+      "grad_norm": 0.5682856440544128,
+      "learning_rate": 8.715641380938583e-06,
+      "loss": 0.9386,
+      "step": 9120
+    },
+    {
+      "epoch": 0.8149753166394889,
+      "grad_norm": 0.49733036756515503,
+      "learning_rate": 8.707479769676091e-06,
+      "loss": 0.9663,
+      "step": 9121
+    },
+    {
+      "epoch": 0.8150646681707507,
+      "grad_norm": 0.42391541600227356,
+      "learning_rate": 8.699321617102312e-06,
+      "loss": 0.92,
+      "step": 9122
+    },
+    {
+      "epoch": 0.8151540197020126,
+      "grad_norm": 0.48684948682785034,
+      "learning_rate": 8.691166923900585e-06,
+      "loss": 0.96,
+      "step": 9123
+    },
+    {
+      "epoch": 0.8152433712332745,
+      "grad_norm": 0.4087164103984833,
+      "learning_rate": 8.68301569075396e-06,
+      "loss": 0.9743,
+      "step": 9124
+    },
+    {
+      "epoch": 0.8153327227645364,
+      "grad_norm": 0.5374601483345032,
+      "learning_rate": 8.67486791834517e-06,
+      "loss": 0.9377,
+      "step": 9125
+    },
+    {
+      "epoch": 0.8154220742957983,
+      "grad_norm": 0.4423445463180542,
+      "learning_rate": 8.66672360735668e-06,
+      "loss": 0.9455,
+      "step": 9126
+    },
+    {
+      "epoch": 0.8155114258270602,
+      "grad_norm": 0.47707852721214294,
+      "learning_rate": 8.658582758470673e-06,
+      "loss": 0.926,
+      "step": 9127
+    },
+    {
+      "epoch": 0.8156007773583219,
+      "grad_norm": 0.4866050183773041,
+      "learning_rate": 8.650445372369025e-06,
+      "loss": 0.9738,
+      "step": 9128
+    },
+    {
+      "epoch": 0.8156901288895838,
+      "grad_norm": 0.5553078651428223,
+      "learning_rate": 8.642311449733331e-06,
+      "loss": 0.8595,
+      "step": 9129
+    },
+    {
+      "epoch": 0.8157794804208457,
+      "grad_norm": 0.4209044277667999,
+      "learning_rate": 8.634180991244894e-06,
+      "loss": 0.9365,
+      "step": 9130
+    },
+    {
+      "epoch": 0.8158688319521076,
+      "grad_norm": 0.4722152352333069,
+      "learning_rate": 8.626053997584732e-06,
+      "loss": 0.9841,
+      "step": 9131
+    },
+    {
+      "epoch": 0.8159581834833695,
+      "grad_norm": 0.5327540040016174,
+      "learning_rate": 8.61793046943356e-06,
+      "loss": 0.9321,
+      "step": 9132
+    },
+    {
+      "epoch": 0.8160475350146313,
+      "grad_norm": 0.4482969045639038,
+      "learning_rate": 8.609810407471824e-06,
+      "loss": 0.9341,
+      "step": 9133
+    },
+    {
+      "epoch": 0.8161368865458932,
+      "grad_norm": 0.46904319524765015,
+      "learning_rate": 8.60169381237964e-06,
+      "loss": 1.0379,
+      "step": 9134
+    },
+    {
+      "epoch": 0.816226238077155,
+      "grad_norm": 0.5137619972229004,
+      "learning_rate": 8.593580684836877e-06,
+      "loss": 0.8599,
+      "step": 9135
+    },
+    {
+      "epoch": 0.8163155896084169,
+      "grad_norm": 0.5669674873352051,
+      "learning_rate": 8.58547102552309e-06,
+      "loss": 0.857,
+      "step": 9136
+    },
+    {
+      "epoch": 0.8164049411396788,
+      "grad_norm": 0.45517855882644653,
+      "learning_rate": 8.577364835117552e-06,
+      "loss": 1.0162,
+      "step": 9137
+    },
+    {
+      "epoch": 0.8164942926709406,
+      "grad_norm": 0.5764082074165344,
+      "learning_rate": 8.569262114299243e-06,
+      "loss": 0.9356,
+      "step": 9138
+    },
+    {
+      "epoch": 0.8165836442022025,
+      "grad_norm": 0.69561767578125,
+      "learning_rate": 8.561162863746847e-06,
+      "loss": 0.8668,
+      "step": 9139
+    },
+    {
+      "epoch": 0.8166729957334644,
+      "grad_norm": 0.6161789298057556,
+      "learning_rate": 8.553067084138772e-06,
+      "loss": 0.905,
+      "step": 9140
+    },
+    {
+      "epoch": 0.8167623472647263,
+      "grad_norm": 0.5391383171081543,
+      "learning_rate": 8.544974776153124e-06,
+      "loss": 0.9331,
+      "step": 9141
+    },
+    {
+      "epoch": 0.8168516987959881,
+      "grad_norm": 0.504325270652771,
+      "learning_rate": 8.536885940467715e-06,
+      "loss": 0.9957,
+      "step": 9142
+    },
+    {
+      "epoch": 0.81694105032725,
+      "grad_norm": 0.4798524081707001,
+      "learning_rate": 8.528800577760083e-06,
+      "loss": 0.9162,
+      "step": 9143
+    },
+    {
+      "epoch": 0.8170304018585118,
+      "grad_norm": 0.48512718081474304,
+      "learning_rate": 8.520718688707447e-06,
+      "loss": 0.9431,
+      "step": 9144
+    },
+    {
+      "epoch": 0.8171197533897737,
+      "grad_norm": 0.5134716033935547,
+      "learning_rate": 8.512640273986756e-06,
+      "loss": 0.9639,
+      "step": 9145
+    },
+    {
+      "epoch": 0.8172091049210356,
+      "grad_norm": 0.3981763422489166,
+      "learning_rate": 8.504565334274662e-06,
+      "loss": 0.9315,
+      "step": 9146
+    },
+    {
+      "epoch": 0.8172984564522975,
+      "grad_norm": 0.43053102493286133,
+      "learning_rate": 8.496493870247535e-06,
+      "loss": 0.9206,
+      "step": 9147
+    },
+    {
+      "epoch": 0.8173878079835594,
+      "grad_norm": 0.5616499185562134,
+      "learning_rate": 8.488425882581436e-06,
+      "loss": 0.9045,
+      "step": 9148
+    },
+    {
+      "epoch": 0.8174771595148211,
+      "grad_norm": 0.4507736265659332,
+      "learning_rate": 8.480361371952156e-06,
+      "loss": 0.9098,
+      "step": 9149
+    },
+    {
+      "epoch": 0.817566511046083,
+      "grad_norm": 0.4418569505214691,
+      "learning_rate": 8.472300339035178e-06,
+      "loss": 0.9385,
+      "step": 9150
+    },
+    {
+      "epoch": 0.8176558625773449,
+      "grad_norm": 0.39274096488952637,
+      "learning_rate": 8.4642427845057e-06,
+      "loss": 0.9688,
+      "step": 9151
+    },
+    {
+      "epoch": 0.8177452141086068,
+      "grad_norm": 0.48565006256103516,
+      "learning_rate": 8.456188709038632e-06,
+      "loss": 0.9807,
+      "step": 9152
+    },
+    {
+      "epoch": 0.8178345656398687,
+      "grad_norm": 0.5166711211204529,
+      "learning_rate": 8.44813811330859e-06,
+      "loss": 1.0159,
+      "step": 9153
+    },
+    {
+      "epoch": 0.8179239171711306,
+      "grad_norm": 0.5695293545722961,
+      "learning_rate": 8.440090997989885e-06,
+      "loss": 0.936,
+      "step": 9154
+    },
+    {
+      "epoch": 0.8180132687023924,
+      "grad_norm": 0.4986543655395508,
+      "learning_rate": 8.432047363756551e-06,
+      "loss": 0.9075,
+      "step": 9155
+    },
+    {
+      "epoch": 0.8181026202336542,
+      "grad_norm": 0.5056511163711548,
+      "learning_rate": 8.424007211282337e-06,
+      "loss": 0.9455,
+      "step": 9156
+    },
+    {
+      "epoch": 0.8181919717649161,
+      "grad_norm": 0.5469470620155334,
+      "learning_rate": 8.415970541240697e-06,
+      "loss": 0.8948,
+      "step": 9157
+    },
+    {
+      "epoch": 0.818281323296178,
+      "grad_norm": 0.5168966054916382,
+      "learning_rate": 8.407937354304769e-06,
+      "loss": 0.8955,
+      "step": 9158
+    },
+    {
+      "epoch": 0.8183706748274399,
+      "grad_norm": 0.4793681800365448,
+      "learning_rate": 8.39990765114742e-06,
+      "loss": 0.9188,
+      "step": 9159
+    },
+    {
+      "epoch": 0.8184600263587017,
+      "grad_norm": 0.5143187046051025,
+      "learning_rate": 8.39188143244124e-06,
+      "loss": 0.9672,
+      "step": 9160
+    },
+    {
+      "epoch": 0.8185493778899636,
+      "grad_norm": 0.5519688725471497,
+      "learning_rate": 8.383858698858494e-06,
+      "loss": 0.9495,
+      "step": 9161
+    },
+    {
+      "epoch": 0.8186387294212255,
+      "grad_norm": 0.45014795660972595,
+      "learning_rate": 8.375839451071183e-06,
+      "loss": 0.8814,
+      "step": 9162
+    },
+    {
+      "epoch": 0.8187280809524873,
+      "grad_norm": 0.4443376064300537,
+      "learning_rate": 8.367823689751009e-06,
+      "loss": 0.9321,
+      "step": 9163
+    },
+    {
+      "epoch": 0.8188174324837492,
+      "grad_norm": 0.5371510982513428,
+      "learning_rate": 8.359811415569352e-06,
+      "loss": 0.9393,
+      "step": 9164
+    },
+    {
+      "epoch": 0.818906784015011,
+      "grad_norm": 0.5210739374160767,
+      "learning_rate": 8.351802629197348e-06,
+      "loss": 0.949,
+      "step": 9165
+    },
+    {
+      "epoch": 0.8189961355462729,
+      "grad_norm": 0.4387337565422058,
+      "learning_rate": 8.343797331305809e-06,
+      "loss": 0.9241,
+      "step": 9166
+    },
+    {
+      "epoch": 0.8190854870775348,
+      "grad_norm": 0.4759118854999542,
+      "learning_rate": 8.335795522565264e-06,
+      "loss": 0.9132,
+      "step": 9167
+    },
+    {
+      "epoch": 0.8191748386087967,
+      "grad_norm": 0.5519284605979919,
+      "learning_rate": 8.327797203645954e-06,
+      "loss": 0.9745,
+      "step": 9168
+    },
+    {
+      "epoch": 0.8192641901400586,
+      "grad_norm": 0.4415756165981293,
+      "learning_rate": 8.319802375217821e-06,
+      "loss": 1.0214,
+      "step": 9169
+    },
+    {
+      "epoch": 0.8193535416713204,
+      "grad_norm": 0.5409457683563232,
+      "learning_rate": 8.311811037950522e-06,
+      "loss": 0.979,
+      "step": 9170
+    },
+    {
+      "epoch": 0.8194428932025822,
+      "grad_norm": 0.49359166622161865,
+      "learning_rate": 8.30382319251341e-06,
+      "loss": 0.8628,
+      "step": 9171
+    },
+    {
+      "epoch": 0.8195322447338441,
+      "grad_norm": 0.45984289050102234,
+      "learning_rate": 8.295838839575559e-06,
+      "loss": 0.9556,
+      "step": 9172
+    },
+    {
+      "epoch": 0.819621596265106,
+      "grad_norm": 0.520918607711792,
+      "learning_rate": 8.287857979805746e-06,
+      "loss": 0.9423,
+      "step": 9173
+    },
+    {
+      "epoch": 0.8197109477963679,
+      "grad_norm": 0.4687787890434265,
+      "learning_rate": 8.279880613872438e-06,
+      "loss": 0.9551,
+      "step": 9174
+    },
+    {
+      "epoch": 0.8198002993276298,
+      "grad_norm": 0.4171432852745056,
+      "learning_rate": 8.271906742443836e-06,
+      "loss": 0.9329,
+      "step": 9175
+    },
+    {
+      "epoch": 0.8198896508588915,
+      "grad_norm": 0.5635870099067688,
+      "learning_rate": 8.263936366187824e-06,
+      "loss": 0.8315,
+      "step": 9176
+    },
+    {
+      "epoch": 0.8199790023901534,
+      "grad_norm": 0.5225967764854431,
+      "learning_rate": 8.255969485772025e-06,
+      "loss": 0.9817,
+      "step": 9177
+    },
+    {
+      "epoch": 0.8200683539214153,
+      "grad_norm": 0.47580423951148987,
+      "learning_rate": 8.248006101863736e-06,
+      "loss": 0.9601,
+      "step": 9178
+    },
+    {
+      "epoch": 0.8201577054526772,
+      "grad_norm": 0.5381955504417419,
+      "learning_rate": 8.240046215129982e-06,
+      "loss": 0.9547,
+      "step": 9179
+    },
+    {
+      "epoch": 0.8202470569839391,
+      "grad_norm": 0.5118776559829712,
+      "learning_rate": 8.232089826237487e-06,
+      "loss": 0.9135,
+      "step": 9180
+    },
+    {
+      "epoch": 0.820336408515201,
+      "grad_norm": 0.6581920981407166,
+      "learning_rate": 8.224136935852683e-06,
+      "loss": 0.8308,
+      "step": 9181
+    },
+    {
+      "epoch": 0.8204257600464628,
+      "grad_norm": 0.45607322454452515,
+      "learning_rate": 8.216187544641706e-06,
+      "loss": 0.881,
+      "step": 9182
+    },
+    {
+      "epoch": 0.8205151115777246,
+      "grad_norm": 0.4387778639793396,
+      "learning_rate": 8.208241653270421e-06,
+      "loss": 0.9486,
+      "step": 9183
+    },
+    {
+      "epoch": 0.8206044631089865,
+      "grad_norm": 0.5159300565719604,
+      "learning_rate": 8.200299262404353e-06,
+      "loss": 1.005,
+      "step": 9184
+    },
+    {
+      "epoch": 0.8206938146402484,
+      "grad_norm": 0.6031007766723633,
+      "learning_rate": 8.19236037270878e-06,
+      "loss": 0.8316,
+      "step": 9185
+    },
+    {
+      "epoch": 0.8207831661715103,
+      "grad_norm": 0.4924890995025635,
+      "learning_rate": 8.184424984848655e-06,
+      "loss": 0.9564,
+      "step": 9186
+    },
+    {
+      "epoch": 0.8208725177027721,
+      "grad_norm": 0.4664970934391022,
+      "learning_rate": 8.176493099488663e-06,
+      "loss": 0.9416,
+      "step": 9187
+    },
+    {
+      "epoch": 0.820961869234034,
+      "grad_norm": 0.46760138869285583,
+      "learning_rate": 8.168564717293182e-06,
+      "loss": 0.9566,
+      "step": 9188
+    },
+    {
+      "epoch": 0.8210512207652959,
+      "grad_norm": 0.5146795511245728,
+      "learning_rate": 8.160639838926293e-06,
+      "loss": 0.8951,
+      "step": 9189
+    },
+    {
+      "epoch": 0.8211405722965577,
+      "grad_norm": 0.4736398458480835,
+      "learning_rate": 8.152718465051806e-06,
+      "loss": 1.0171,
+      "step": 9190
+    },
+    {
+      "epoch": 0.8212299238278196,
+      "grad_norm": 0.6582403182983398,
+      "learning_rate": 8.144800596333196e-06,
+      "loss": 0.9182,
+      "step": 9191
+    },
+    {
+      "epoch": 0.8213192753590814,
+      "grad_norm": 0.49453866481781006,
+      "learning_rate": 8.136886233433683e-06,
+      "loss": 0.8465,
+      "step": 9192
+    },
+    {
+      "epoch": 0.8214086268903433,
+      "grad_norm": 0.4479573965072632,
+      "learning_rate": 8.128975377016174e-06,
+      "loss": 0.8996,
+      "step": 9193
+    },
+    {
+      "epoch": 0.8214979784216052,
+      "grad_norm": 0.5152574777603149,
+      "learning_rate": 8.1210680277433e-06,
+      "loss": 0.9579,
+      "step": 9194
+    },
+    {
+      "epoch": 0.8215873299528671,
+      "grad_norm": 0.5516067743301392,
+      "learning_rate": 8.113164186277367e-06,
+      "loss": 0.9078,
+      "step": 9195
+    },
+    {
+      "epoch": 0.821676681484129,
+      "grad_norm": 0.4801537096500397,
+      "learning_rate": 8.105263853280416e-06,
+      "loss": 1.0086,
+      "step": 9196
+    },
+    {
+      "epoch": 0.8217660330153908,
+      "grad_norm": 0.4999988377094269,
+      "learning_rate": 8.097367029414182e-06,
+      "loss": 0.8929,
+      "step": 9197
+    },
+    {
+      "epoch": 0.8218553845466526,
+      "grad_norm": 0.45241791009902954,
+      "learning_rate": 8.089473715340107e-06,
+      "loss": 0.9495,
+      "step": 9198
+    },
+    {
+      "epoch": 0.8219447360779145,
+      "grad_norm": 0.4731530249118805,
+      "learning_rate": 8.081583911719343e-06,
+      "loss": 0.9343,
+      "step": 9199
+    },
+    {
+      "epoch": 0.8220340876091764,
+      "grad_norm": 0.5134366154670715,
+      "learning_rate": 8.073697619212745e-06,
+      "loss": 0.9878,
+      "step": 9200
+    },
+    {
+      "epoch": 0.8221234391404383,
+      "grad_norm": 0.4594072103500366,
+      "learning_rate": 8.065814838480879e-06,
+      "loss": 0.9403,
+      "step": 9201
+    },
+    {
+      "epoch": 0.8222127906717002,
+      "grad_norm": 0.49172157049179077,
+      "learning_rate": 8.057935570184e-06,
+      "loss": 0.9553,
+      "step": 9202
+    },
+    {
+      "epoch": 0.822302142202962,
+      "grad_norm": 0.44806215167045593,
+      "learning_rate": 8.050059814982092e-06,
+      "loss": 0.9714,
+      "step": 9203
+    },
+    {
+      "epoch": 0.8223914937342238,
+      "grad_norm": 0.4985756576061249,
+      "learning_rate": 8.042187573534836e-06,
+      "loss": 0.9834,
+      "step": 9204
+    },
+    {
+      "epoch": 0.8224808452654857,
+      "grad_norm": 0.5466139912605286,
+      "learning_rate": 8.034318846501598e-06,
+      "loss": 0.9727,
+      "step": 9205
+    },
+    {
+      "epoch": 0.8225701967967476,
+      "grad_norm": 0.4964464008808136,
+      "learning_rate": 8.026453634541481e-06,
+      "loss": 0.9444,
+      "step": 9206
+    },
+    {
+      "epoch": 0.8226595483280095,
+      "grad_norm": 0.49595358967781067,
+      "learning_rate": 8.018591938313275e-06,
+      "loss": 0.973,
+      "step": 9207
+    },
+    {
+      "epoch": 0.8227488998592714,
+      "grad_norm": 0.5283821821212769,
+      "learning_rate": 8.010733758475486e-06,
+      "loss": 0.956,
+      "step": 9208
+    },
+    {
+      "epoch": 0.8228382513905332,
+      "grad_norm": 0.4949239492416382,
+      "learning_rate": 8.002879095686317e-06,
+      "loss": 0.9192,
+      "step": 9209
+    },
+    {
+      "epoch": 0.8229276029217951,
+      "grad_norm": 0.48823899030685425,
+      "learning_rate": 7.995027950603683e-06,
+      "loss": 0.8878,
+      "step": 9210
+    },
+    {
+      "epoch": 0.8230169544530569,
+      "grad_norm": 0.483600378036499,
+      "learning_rate": 7.987180323885196e-06,
+      "loss": 0.9239,
+      "step": 9211
+    },
+    {
+      "epoch": 0.8231063059843188,
+      "grad_norm": 0.4983593225479126,
+      "learning_rate": 7.979336216188182e-06,
+      "loss": 0.9776,
+      "step": 9212
+    },
+    {
+      "epoch": 0.8231956575155807,
+      "grad_norm": 0.4743238389492035,
+      "learning_rate": 7.971495628169667e-06,
+      "loss": 0.9436,
+      "step": 9213
+    },
+    {
+      "epoch": 0.8232850090468425,
+      "grad_norm": 0.4517209529876709,
+      "learning_rate": 7.963658560486399e-06,
+      "loss": 0.9313,
+      "step": 9214
+    },
+    {
+      "epoch": 0.8233743605781044,
+      "grad_norm": 0.5047935247421265,
+      "learning_rate": 7.955825013794793e-06,
+      "loss": 0.9093,
+      "step": 9215
+    },
+    {
+      "epoch": 0.8234637121093663,
+      "grad_norm": 0.603001594543457,
+      "learning_rate": 7.947994988750995e-06,
+      "loss": 0.9476,
+      "step": 9216
+    },
+    {
+      "epoch": 0.8235530636406282,
+      "grad_norm": 0.4813896119594574,
+      "learning_rate": 7.940168486010862e-06,
+      "loss": 0.8947,
+      "step": 9217
+    },
+    {
+      "epoch": 0.82364241517189,
+      "grad_norm": 0.5081338882446289,
+      "learning_rate": 7.932345506229944e-06,
+      "loss": 0.8794,
+      "step": 9218
+    },
+    {
+      "epoch": 0.8237317667031518,
+      "grad_norm": 0.4924575984477997,
+      "learning_rate": 7.924526050063497e-06,
+      "loss": 0.9315,
+      "step": 9219
+    },
+    {
+      "epoch": 0.8238211182344137,
+      "grad_norm": 0.4447862207889557,
+      "learning_rate": 7.916710118166482e-06,
+      "loss": 0.9681,
+      "step": 9220
+    },
+    {
+      "epoch": 0.8239104697656756,
+      "grad_norm": 0.5768123269081116,
+      "learning_rate": 7.908897711193575e-06,
+      "loss": 0.9272,
+      "step": 9221
+    },
+    {
+      "epoch": 0.8239998212969375,
+      "grad_norm": 0.6067156791687012,
+      "learning_rate": 7.901088829799142e-06,
+      "loss": 0.9376,
+      "step": 9222
+    },
+    {
+      "epoch": 0.8240891728281994,
+      "grad_norm": 0.6196531057357788,
+      "learning_rate": 7.893283474637275e-06,
+      "loss": 0.8734,
+      "step": 9223
+    },
+    {
+      "epoch": 0.8241785243594613,
+      "grad_norm": 0.49695178866386414,
+      "learning_rate": 7.88548164636173e-06,
+      "loss": 0.9231,
+      "step": 9224
+    },
+    {
+      "epoch": 0.824267875890723,
+      "grad_norm": 0.4435543417930603,
+      "learning_rate": 7.877683345626008e-06,
+      "loss": 0.9304,
+      "step": 9225
+    },
+    {
+      "epoch": 0.8243572274219849,
+      "grad_norm": 0.4490625560283661,
+      "learning_rate": 7.869888573083294e-06,
+      "loss": 0.9596,
+      "step": 9226
+    },
+    {
+      "epoch": 0.8244465789532468,
+      "grad_norm": 0.4157819151878357,
+      "learning_rate": 7.862097329386497e-06,
+      "loss": 0.9515,
+      "step": 9227
+    },
+    {
+      "epoch": 0.8245359304845087,
+      "grad_norm": 0.4776933491230011,
+      "learning_rate": 7.854309615188198e-06,
+      "loss": 0.9247,
+      "step": 9228
+    },
+    {
+      "epoch": 0.8246252820157706,
+      "grad_norm": 0.4814663529396057,
+      "learning_rate": 7.84652543114071e-06,
+      "loss": 0.9378,
+      "step": 9229
+    },
+    {
+      "epoch": 0.8247146335470324,
+      "grad_norm": 0.45337748527526855,
+      "learning_rate": 7.83874477789604e-06,
+      "loss": 0.9357,
+      "step": 9230
+    },
+    {
+      "epoch": 0.8248039850782943,
+      "grad_norm": 0.44966641068458557,
+      "learning_rate": 7.830967656105903e-06,
+      "loss": 0.9343,
+      "step": 9231
+    },
+    {
+      "epoch": 0.8248933366095561,
+      "grad_norm": 0.4764994978904724,
+      "learning_rate": 7.823194066421707e-06,
+      "loss": 0.8798,
+      "step": 9232
+    },
+    {
+      "epoch": 0.824982688140818,
+      "grad_norm": 0.44345101714134216,
+      "learning_rate": 7.815424009494588e-06,
+      "loss": 0.9255,
+      "step": 9233
+    },
+    {
+      "epoch": 0.8250720396720799,
+      "grad_norm": 0.44141438603401184,
+      "learning_rate": 7.807657485975372e-06,
+      "loss": 0.9607,
+      "step": 9234
+    },
+    {
+      "epoch": 0.8251613912033418,
+      "grad_norm": 0.4870518147945404,
+      "learning_rate": 7.79989449651457e-06,
+      "loss": 0.8497,
+      "step": 9235
+    },
+    {
+      "epoch": 0.8252507427346036,
+      "grad_norm": 0.6152021288871765,
+      "learning_rate": 7.792135041762421e-06,
+      "loss": 0.8489,
+      "step": 9236
+    },
+    {
+      "epoch": 0.8253400942658655,
+      "grad_norm": 0.6551914811134338,
+      "learning_rate": 7.784379122368873e-06,
+      "loss": 0.9771,
+      "step": 9237
+    },
+    {
+      "epoch": 0.8254294457971274,
+      "grad_norm": 0.4421270787715912,
+      "learning_rate": 7.776626738983556e-06,
+      "loss": 0.9964,
+      "step": 9238
+    },
+    {
+      "epoch": 0.8255187973283892,
+      "grad_norm": 0.55922931432724,
+      "learning_rate": 7.768877892255816e-06,
+      "loss": 0.968,
+      "step": 9239
+    },
+    {
+      "epoch": 0.8256081488596511,
+      "grad_norm": 0.4829465448856354,
+      "learning_rate": 7.761132582834708e-06,
+      "loss": 0.967,
+      "step": 9240
+    },
+    {
+      "epoch": 0.8256975003909129,
+      "grad_norm": 0.5545680522918701,
+      "learning_rate": 7.753390811368971e-06,
+      "loss": 0.9614,
+      "step": 9241
+    },
+    {
+      "epoch": 0.8257868519221748,
+      "grad_norm": 0.5379648804664612,
+      "learning_rate": 7.745652578507079e-06,
+      "loss": 0.846,
+      "step": 9242
+    },
+    {
+      "epoch": 0.8258762034534367,
+      "grad_norm": 0.6437764167785645,
+      "learning_rate": 7.737917884897177e-06,
+      "loss": 0.9404,
+      "step": 9243
+    },
+    {
+      "epoch": 0.8259655549846986,
+      "grad_norm": 0.4301976263523102,
+      "learning_rate": 7.73018673118715e-06,
+      "loss": 0.9681,
+      "step": 9244
+    },
+    {
+      "epoch": 0.8260549065159604,
+      "grad_norm": 0.46342891454696655,
+      "learning_rate": 7.722459118024538e-06,
+      "loss": 0.9234,
+      "step": 9245
+    },
+    {
+      "epoch": 0.8261442580472222,
+      "grad_norm": 0.45039889216423035,
+      "learning_rate": 7.714735046056615e-06,
+      "loss": 0.9904,
+      "step": 9246
+    },
+    {
+      "epoch": 0.8262336095784841,
+      "grad_norm": 0.5035346746444702,
+      "learning_rate": 7.707014515930366e-06,
+      "loss": 0.891,
+      "step": 9247
+    },
+    {
+      "epoch": 0.826322961109746,
+      "grad_norm": 0.41054171323776245,
+      "learning_rate": 7.69929752829246e-06,
+      "loss": 0.943,
+      "step": 9248
+    },
+    {
+      "epoch": 0.8264123126410079,
+      "grad_norm": 0.4819393754005432,
+      "learning_rate": 7.691584083789277e-06,
+      "loss": 0.9219,
+      "step": 9249
+    },
+    {
+      "epoch": 0.8265016641722698,
+      "grad_norm": 0.4592518210411072,
+      "learning_rate": 7.683874183066903e-06,
+      "loss": 0.9504,
+      "step": 9250
+    },
+    {
+      "epoch": 0.8265910157035317,
+      "grad_norm": 0.5692583322525024,
+      "learning_rate": 7.676167826771124e-06,
+      "loss": 0.8824,
+      "step": 9251
+    },
+    {
+      "epoch": 0.8266803672347934,
+      "grad_norm": 0.48362407088279724,
+      "learning_rate": 7.668465015547427e-06,
+      "loss": 0.8867,
+      "step": 9252
+    },
+    {
+      "epoch": 0.8267697187660553,
+      "grad_norm": 0.5431807041168213,
+      "learning_rate": 7.66076575004101e-06,
+      "loss": 0.8867,
+      "step": 9253
+    },
+    {
+      "epoch": 0.8268590702973172,
+      "grad_norm": 0.6217208504676819,
+      "learning_rate": 7.653070030896774e-06,
+      "loss": 0.9971,
+      "step": 9254
+    },
+    {
+      "epoch": 0.8269484218285791,
+      "grad_norm": 0.4189354181289673,
+      "learning_rate": 7.6453778587593e-06,
+      "loss": 0.9719,
+      "step": 9255
+    },
+    {
+      "epoch": 0.827037773359841,
+      "grad_norm": 0.50577312707901,
+      "learning_rate": 7.637689234272899e-06,
+      "loss": 0.9394,
+      "step": 9256
+    },
+    {
+      "epoch": 0.8271271248911028,
+      "grad_norm": 0.5943676829338074,
+      "learning_rate": 7.630004158081572e-06,
+      "loss": 0.8626,
+      "step": 9257
+    },
+    {
+      "epoch": 0.8272164764223647,
+      "grad_norm": 0.42551112174987793,
+      "learning_rate": 7.622322630829032e-06,
+      "loss": 0.9436,
+      "step": 9258
+    },
+    {
+      "epoch": 0.8273058279536265,
+      "grad_norm": 0.6228631138801575,
+      "learning_rate": 7.6146446531586815e-06,
+      "loss": 0.9074,
+      "step": 9259
+    },
+    {
+      "epoch": 0.8273951794848884,
+      "grad_norm": 0.4399883449077606,
+      "learning_rate": 7.60697022571365e-06,
+      "loss": 0.9957,
+      "step": 9260
+    },
+    {
+      "epoch": 0.8274845310161503,
+      "grad_norm": 0.5171003341674805,
+      "learning_rate": 7.599299349136723e-06,
+      "loss": 0.9123,
+      "step": 9261
+    },
+    {
+      "epoch": 0.8275738825474122,
+      "grad_norm": 0.4070376753807068,
+      "learning_rate": 7.591632024070439e-06,
+      "loss": 0.9231,
+      "step": 9262
+    },
+    {
+      "epoch": 0.827663234078674,
+      "grad_norm": 0.47427892684936523,
+      "learning_rate": 7.583968251157014e-06,
+      "loss": 0.9726,
+      "step": 9263
+    },
+    {
+      "epoch": 0.8277525856099359,
+      "grad_norm": 0.5160714983940125,
+      "learning_rate": 7.576308031038381e-06,
+      "loss": 0.9672,
+      "step": 9264
+    },
+    {
+      "epoch": 0.8278419371411978,
+      "grad_norm": 0.5352666974067688,
+      "learning_rate": 7.568651364356144e-06,
+      "loss": 0.8548,
+      "step": 9265
+    },
+    {
+      "epoch": 0.8279312886724596,
+      "grad_norm": 0.4556531310081482,
+      "learning_rate": 7.560998251751639e-06,
+      "loss": 0.9801,
+      "step": 9266
+    },
+    {
+      "epoch": 0.8280206402037215,
+      "grad_norm": 0.46089810132980347,
+      "learning_rate": 7.553348693865897e-06,
+      "loss": 0.9177,
+      "step": 9267
+    },
+    {
+      "epoch": 0.8281099917349833,
+      "grad_norm": 0.4947889447212219,
+      "learning_rate": 7.545702691339657e-06,
+      "loss": 0.9936,
+      "step": 9268
+    },
+    {
+      "epoch": 0.8281993432662452,
+      "grad_norm": 0.46438440680503845,
+      "learning_rate": 7.538060244813339e-06,
+      "loss": 0.9395,
+      "step": 9269
+    },
+    {
+      "epoch": 0.8282886947975071,
+      "grad_norm": 0.46110087633132935,
+      "learning_rate": 7.530421354927092e-06,
+      "loss": 0.9576,
+      "step": 9270
+    },
+    {
+      "epoch": 0.828378046328769,
+      "grad_norm": 0.4900936484336853,
+      "learning_rate": 7.522786022320744e-06,
+      "loss": 0.9473,
+      "step": 9271
+    },
+    {
+      "epoch": 0.8284673978600309,
+      "grad_norm": 0.48400017619132996,
+      "learning_rate": 7.5151542476338485e-06,
+      "loss": 0.9329,
+      "step": 9272
+    },
+    {
+      "epoch": 0.8285567493912926,
+      "grad_norm": 0.49266934394836426,
+      "learning_rate": 7.507526031505635e-06,
+      "loss": 0.8979,
+      "step": 9273
+    },
+    {
+      "epoch": 0.8286461009225545,
+      "grad_norm": 0.5182574987411499,
+      "learning_rate": 7.499901374575069e-06,
+      "loss": 0.8894,
+      "step": 9274
+    },
+    {
+      "epoch": 0.8287354524538164,
+      "grad_norm": 0.5964707732200623,
+      "learning_rate": 7.492280277480768e-06,
+      "loss": 0.9323,
+      "step": 9275
+    },
+    {
+      "epoch": 0.8288248039850783,
+      "grad_norm": 0.45841360092163086,
+      "learning_rate": 7.4846627408610925e-06,
+      "loss": 0.9406,
+      "step": 9276
+    },
+    {
+      "epoch": 0.8289141555163402,
+      "grad_norm": 0.4415697753429413,
+      "learning_rate": 7.477048765354094e-06,
+      "loss": 0.9698,
+      "step": 9277
+    },
+    {
+      "epoch": 0.8290035070476021,
+      "grad_norm": 0.49116966128349304,
+      "learning_rate": 7.469438351597524e-06,
+      "loss": 0.9453,
+      "step": 9278
+    },
+    {
+      "epoch": 0.829092858578864,
+      "grad_norm": 0.5854166746139526,
+      "learning_rate": 7.4618315002288384e-06,
+      "loss": 0.9059,
+      "step": 9279
+    },
+    {
+      "epoch": 0.8291822101101257,
+      "grad_norm": 0.46509405970573425,
+      "learning_rate": 7.454228211885184e-06,
+      "loss": 0.9222,
+      "step": 9280
+    },
+    {
+      "epoch": 0.8292715616413876,
+      "grad_norm": 0.442025363445282,
+      "learning_rate": 7.446628487203422e-06,
+      "loss": 0.9237,
+      "step": 9281
+    },
+    {
+      "epoch": 0.8293609131726495,
+      "grad_norm": 0.6018460392951965,
+      "learning_rate": 7.439032326820117e-06,
+      "loss": 0.9529,
+      "step": 9282
+    },
+    {
+      "epoch": 0.8294502647039114,
+      "grad_norm": 0.48638513684272766,
+      "learning_rate": 7.4314397313715145e-06,
+      "loss": 0.8751,
+      "step": 9283
+    },
+    {
+      "epoch": 0.8295396162351732,
+      "grad_norm": 0.4461117684841156,
+      "learning_rate": 7.423850701493585e-06,
+      "loss": 0.9923,
+      "step": 9284
+    },
+    {
+      "epoch": 0.8296289677664351,
+      "grad_norm": 0.5634244084358215,
+      "learning_rate": 7.416265237822001e-06,
+      "loss": 0.9615,
+      "step": 9285
+    },
+    {
+      "epoch": 0.829718319297697,
+      "grad_norm": 0.5193483233451843,
+      "learning_rate": 7.408683340992101e-06,
+      "loss": 0.8622,
+      "step": 9286
+    },
+    {
+      "epoch": 0.8298076708289588,
+      "grad_norm": 0.5628817081451416,
+      "learning_rate": 7.401105011638965e-06,
+      "loss": 0.8551,
+      "step": 9287
+    },
+    {
+      "epoch": 0.8298970223602207,
+      "grad_norm": 0.5427822470664978,
+      "learning_rate": 7.3935302503973515e-06,
+      "loss": 0.8488,
+      "step": 9288
+    },
+    {
+      "epoch": 0.8299863738914826,
+      "grad_norm": 0.4602343440055847,
+      "learning_rate": 7.385959057901737e-06,
+      "loss": 1.054,
+      "step": 9289
+    },
+    {
+      "epoch": 0.8300757254227444,
+      "grad_norm": 0.5840815305709839,
+      "learning_rate": 7.378391434786281e-06,
+      "loss": 0.911,
+      "step": 9290
+    },
+    {
+      "epoch": 0.8301650769540063,
+      "grad_norm": 0.514077365398407,
+      "learning_rate": 7.370827381684853e-06,
+      "loss": 0.923,
+      "step": 9291
+    },
+    {
+      "epoch": 0.8302544284852682,
+      "grad_norm": 0.5028949975967407,
+      "learning_rate": 7.3632668992310305e-06,
+      "loss": 0.8987,
+      "step": 9292
+    },
+    {
+      "epoch": 0.8303437800165301,
+      "grad_norm": 0.3772270083427429,
+      "learning_rate": 7.35570998805809e-06,
+      "loss": 0.9484,
+      "step": 9293
+    },
+    {
+      "epoch": 0.8304331315477919,
+      "grad_norm": 0.5475994944572449,
+      "learning_rate": 7.348156648798981e-06,
+      "loss": 0.9564,
+      "step": 9294
+    },
+    {
+      "epoch": 0.8305224830790537,
+      "grad_norm": 0.5647354125976562,
+      "learning_rate": 7.340606882086393e-06,
+      "loss": 0.894,
+      "step": 9295
+    },
+    {
+      "epoch": 0.8306118346103156,
+      "grad_norm": 0.5140451788902283,
+      "learning_rate": 7.33306068855269e-06,
+      "loss": 0.9726,
+      "step": 9296
+    },
+    {
+      "epoch": 0.8307011861415775,
+      "grad_norm": 0.45875853300094604,
+      "learning_rate": 7.325518068829967e-06,
+      "loss": 0.9158,
+      "step": 9297
+    },
+    {
+      "epoch": 0.8307905376728394,
+      "grad_norm": 0.38240689039230347,
+      "learning_rate": 7.31797902354997e-06,
+      "loss": 0.9517,
+      "step": 9298
+    },
+    {
+      "epoch": 0.8308798892041013,
+      "grad_norm": 0.5880802869796753,
+      "learning_rate": 7.310443553344182e-06,
+      "loss": 0.9576,
+      "step": 9299
+    },
+    {
+      "epoch": 0.8309692407353632,
+      "grad_norm": 0.5717623829841614,
+      "learning_rate": 7.302911658843792e-06,
+      "loss": 0.9193,
+      "step": 9300
+    },
+    {
+      "epoch": 0.8310585922666249,
+      "grad_norm": 0.4772949814796448,
+      "learning_rate": 7.2953833406796675e-06,
+      "loss": 0.8271,
+      "step": 9301
+    },
+    {
+      "epoch": 0.8311479437978868,
+      "grad_norm": 0.43257272243499756,
+      "learning_rate": 7.287858599482383e-06,
+      "loss": 0.946,
+      "step": 9302
+    },
+    {
+      "epoch": 0.8312372953291487,
+      "grad_norm": 0.5265113115310669,
+      "learning_rate": 7.280337435882223e-06,
+      "loss": 0.8861,
+      "step": 9303
+    },
+    {
+      "epoch": 0.8313266468604106,
+      "grad_norm": 0.4444237947463989,
+      "learning_rate": 7.272819850509161e-06,
+      "loss": 0.9427,
+      "step": 9304
+    },
+    {
+      "epoch": 0.8314159983916725,
+      "grad_norm": 0.53780198097229,
+      "learning_rate": 7.265305843992881e-06,
+      "loss": 0.8518,
+      "step": 9305
+    },
+    {
+      "epoch": 0.8315053499229343,
+      "grad_norm": 0.5039278268814087,
+      "learning_rate": 7.257795416962753e-06,
+      "loss": 0.9144,
+      "step": 9306
+    },
+    {
+      "epoch": 0.8315947014541961,
+      "grad_norm": 0.5485106110572815,
+      "learning_rate": 7.250288570047853e-06,
+      "loss": 0.9448,
+      "step": 9307
+    },
+    {
+      "epoch": 0.831684052985458,
+      "grad_norm": 0.592823326587677,
+      "learning_rate": 7.242785303876965e-06,
+      "loss": 0.9934,
+      "step": 9308
+    },
+    {
+      "epoch": 0.8317734045167199,
+      "grad_norm": 0.42520976066589355,
+      "learning_rate": 7.235285619078569e-06,
+      "loss": 0.9649,
+      "step": 9309
+    },
+    {
+      "epoch": 0.8318627560479818,
+      "grad_norm": 0.4947132170200348,
+      "learning_rate": 7.2277895162808416e-06,
+      "loss": 0.9205,
+      "step": 9310
+    },
+    {
+      "epoch": 0.8319521075792436,
+      "grad_norm": 0.42473235726356506,
+      "learning_rate": 7.220296996111658e-06,
+      "loss": 0.9293,
+      "step": 9311
+    },
+    {
+      "epoch": 0.8320414591105055,
+      "grad_norm": 0.5703517198562622,
+      "learning_rate": 7.2128080591986e-06,
+      "loss": 0.8687,
+      "step": 9312
+    },
+    {
+      "epoch": 0.8321308106417674,
+      "grad_norm": 0.5139065980911255,
+      "learning_rate": 7.205322706168949e-06,
+      "loss": 0.8515,
+      "step": 9313
+    },
+    {
+      "epoch": 0.8322201621730292,
+      "grad_norm": 0.49325382709503174,
+      "learning_rate": 7.19784093764968e-06,
+      "loss": 0.9548,
+      "step": 9314
+    },
+    {
+      "epoch": 0.8323095137042911,
+      "grad_norm": 0.4829562306404114,
+      "learning_rate": 7.190362754267477e-06,
+      "loss": 0.923,
+      "step": 9315
+    },
+    {
+      "epoch": 0.832398865235553,
+      "grad_norm": 0.4533390402793884,
+      "learning_rate": 7.182888156648704e-06,
+      "loss": 0.9178,
+      "step": 9316
+    },
+    {
+      "epoch": 0.8324882167668148,
+      "grad_norm": 0.42005079984664917,
+      "learning_rate": 7.175417145419444e-06,
+      "loss": 0.912,
+      "step": 9317
+    },
+    {
+      "epoch": 0.8325775682980767,
+      "grad_norm": 0.5088595747947693,
+      "learning_rate": 7.16794972120548e-06,
+      "loss": 0.9998,
+      "step": 9318
+    },
+    {
+      "epoch": 0.8326669198293386,
+      "grad_norm": 0.4803122580051422,
+      "learning_rate": 7.160485884632278e-06,
+      "loss": 0.8923,
+      "step": 9319
+    },
+    {
+      "epoch": 0.8327562713606005,
+      "grad_norm": 0.439439058303833,
+      "learning_rate": 7.153025636325023e-06,
+      "loss": 0.9342,
+      "step": 9320
+    },
+    {
+      "epoch": 0.8328456228918623,
+      "grad_norm": 0.4299473464488983,
+      "learning_rate": 7.145568976908584e-06,
+      "loss": 0.9986,
+      "step": 9321
+    },
+    {
+      "epoch": 0.8329349744231241,
+      "grad_norm": 0.46970871090888977,
+      "learning_rate": 7.138115907007537e-06,
+      "loss": 0.9392,
+      "step": 9322
+    },
+    {
+      "epoch": 0.833024325954386,
+      "grad_norm": 0.500106692314148,
+      "learning_rate": 7.1306664272461635e-06,
+      "loss": 0.9441,
+      "step": 9323
+    },
+    {
+      "epoch": 0.8331136774856479,
+      "grad_norm": 0.5206024050712585,
+      "learning_rate": 7.123220538248426e-06,
+      "loss": 0.8643,
+      "step": 9324
+    },
+    {
+      "epoch": 0.8332030290169098,
+      "grad_norm": 0.4776027798652649,
+      "learning_rate": 7.115778240638016e-06,
+      "loss": 0.9586,
+      "step": 9325
+    },
+    {
+      "epoch": 0.8332923805481717,
+      "grad_norm": 0.45335862040519714,
+      "learning_rate": 7.1083395350382776e-06,
+      "loss": 0.941,
+      "step": 9326
+    },
+    {
+      "epoch": 0.8333817320794336,
+      "grad_norm": 0.47026875615119934,
+      "learning_rate": 7.100904422072296e-06,
+      "loss": 0.9022,
+      "step": 9327
+    },
+    {
+      "epoch": 0.8334710836106953,
+      "grad_norm": 0.4703599810600281,
+      "learning_rate": 7.093472902362841e-06,
+      "loss": 1.0376,
+      "step": 9328
+    },
+    {
+      "epoch": 0.8335604351419572,
+      "grad_norm": 0.48730385303497314,
+      "learning_rate": 7.086044976532386e-06,
+      "loss": 0.983,
+      "step": 9329
+    },
+    {
+      "epoch": 0.8336497866732191,
+      "grad_norm": 0.44514140486717224,
+      "learning_rate": 7.078620645203099e-06,
+      "loss": 0.8899,
+      "step": 9330
+    },
+    {
+      "epoch": 0.833739138204481,
+      "grad_norm": 0.4243552088737488,
+      "learning_rate": 7.0711999089968335e-06,
+      "loss": 0.9852,
+      "step": 9331
+    },
+    {
+      "epoch": 0.8338284897357429,
+      "grad_norm": 0.49389246106147766,
+      "learning_rate": 7.06378276853516e-06,
+      "loss": 0.8915,
+      "step": 9332
+    },
+    {
+      "epoch": 0.8339178412670047,
+      "grad_norm": 0.4797744154930115,
+      "learning_rate": 7.056369224439352e-06,
+      "loss": 0.9607,
+      "step": 9333
+    },
+    {
+      "epoch": 0.8340071927982666,
+      "grad_norm": 0.5037931203842163,
+      "learning_rate": 7.048959277330363e-06,
+      "loss": 0.9384,
+      "step": 9334
+    },
+    {
+      "epoch": 0.8340965443295284,
+      "grad_norm": 0.474246084690094,
+      "learning_rate": 7.0415529278288765e-06,
+      "loss": 1.0259,
+      "step": 9335
+    },
+    {
+      "epoch": 0.8341858958607903,
+      "grad_norm": 0.46430274844169617,
+      "learning_rate": 7.034150176555221e-06,
+      "loss": 0.9752,
+      "step": 9336
+    },
+    {
+      "epoch": 0.8342752473920522,
+      "grad_norm": 0.5320345759391785,
+      "learning_rate": 7.0267510241294745e-06,
+      "loss": 0.917,
+      "step": 9337
+    },
+    {
+      "epoch": 0.834364598923314,
+      "grad_norm": 0.4332819879055023,
+      "learning_rate": 7.01935547117139e-06,
+      "loss": 0.9781,
+      "step": 9338
+    },
+    {
+      "epoch": 0.8344539504545759,
+      "grad_norm": 0.4649805724620819,
+      "learning_rate": 7.011963518300424e-06,
+      "loss": 0.9405,
+      "step": 9339
+    },
+    {
+      "epoch": 0.8345433019858378,
+      "grad_norm": 0.4542677700519562,
+      "learning_rate": 7.004575166135735e-06,
+      "loss": 0.9736,
+      "step": 9340
+    },
+    {
+      "epoch": 0.8346326535170997,
+      "grad_norm": 0.4982972741127014,
+      "learning_rate": 6.997190415296173e-06,
+      "loss": 0.8426,
+      "step": 9341
+    },
+    {
+      "epoch": 0.8347220050483615,
+      "grad_norm": 0.49553072452545166,
+      "learning_rate": 6.989809266400288e-06,
+      "loss": 0.92,
+      "step": 9342
+    },
+    {
+      "epoch": 0.8348113565796234,
+      "grad_norm": 0.4943731427192688,
+      "learning_rate": 6.982431720066335e-06,
+      "loss": 0.9163,
+      "step": 9343
+    },
+    {
+      "epoch": 0.8349007081108852,
+      "grad_norm": 0.4164564311504364,
+      "learning_rate": 6.975057776912258e-06,
+      "loss": 0.9058,
+      "step": 9344
+    },
+    {
+      "epoch": 0.8349900596421471,
+      "grad_norm": 0.4946102797985077,
+      "learning_rate": 6.96768743755572e-06,
+      "loss": 0.9323,
+      "step": 9345
+    },
+    {
+      "epoch": 0.835079411173409,
+      "grad_norm": 0.5377085208892822,
+      "learning_rate": 6.960320702614037e-06,
+      "loss": 0.8814,
+      "step": 9346
+    },
+    {
+      "epoch": 0.8351687627046709,
+      "grad_norm": 0.5740790367126465,
+      "learning_rate": 6.952957572704266e-06,
+      "loss": 0.9291,
+      "step": 9347
+    },
+    {
+      "epoch": 0.8352581142359328,
+      "grad_norm": 0.5461665987968445,
+      "learning_rate": 6.945598048443147e-06,
+      "loss": 0.8508,
+      "step": 9348
+    },
+    {
+      "epoch": 0.8353474657671945,
+      "grad_norm": 0.48720571398735046,
+      "learning_rate": 6.93824213044712e-06,
+      "loss": 0.9276,
+      "step": 9349
+    },
+    {
+      "epoch": 0.8354368172984564,
+      "grad_norm": 0.4371801018714905,
+      "learning_rate": 6.93088981933232e-06,
+      "loss": 0.9428,
+      "step": 9350
+    },
+    {
+      "epoch": 0.8355261688297183,
+      "grad_norm": 0.661524772644043,
+      "learning_rate": 6.9235411157145765e-06,
+      "loss": 0.8726,
+      "step": 9351
+    },
+    {
+      "epoch": 0.8356155203609802,
+      "grad_norm": 0.4543387293815613,
+      "learning_rate": 6.916196020209431e-06,
+      "loss": 0.9342,
+      "step": 9352
+    },
+    {
+      "epoch": 0.8357048718922421,
+      "grad_norm": 0.40771013498306274,
+      "learning_rate": 6.908854533432107e-06,
+      "loss": 0.9474,
+      "step": 9353
+    },
+    {
+      "epoch": 0.835794223423504,
+      "grad_norm": 0.4102661609649658,
+      "learning_rate": 6.901516655997536e-06,
+      "loss": 0.9604,
+      "step": 9354
+    },
+    {
+      "epoch": 0.8358835749547658,
+      "grad_norm": 0.47201770544052124,
+      "learning_rate": 6.8941823885203485e-06,
+      "loss": 0.9408,
+      "step": 9355
+    },
+    {
+      "epoch": 0.8359729264860276,
+      "grad_norm": 0.508849561214447,
+      "learning_rate": 6.886851731614857e-06,
+      "loss": 0.8948,
+      "step": 9356
+    },
+    {
+      "epoch": 0.8360622780172895,
+      "grad_norm": 0.48625683784484863,
+      "learning_rate": 6.879524685895084e-06,
+      "loss": 0.9328,
+      "step": 9357
+    },
+    {
+      "epoch": 0.8361516295485514,
+      "grad_norm": 0.4217790961265564,
+      "learning_rate": 6.872201251974747e-06,
+      "loss": 1.0213,
+      "step": 9358
+    },
+    {
+      "epoch": 0.8362409810798133,
+      "grad_norm": 0.4759739637374878,
+      "learning_rate": 6.864881430467262e-06,
+      "loss": 0.9459,
+      "step": 9359
+    },
+    {
+      "epoch": 0.8363303326110751,
+      "grad_norm": 0.4266846776008606,
+      "learning_rate": 6.857565221985751e-06,
+      "loss": 0.9863,
+      "step": 9360
+    },
+    {
+      "epoch": 0.836419684142337,
+      "grad_norm": 0.5146978497505188,
+      "learning_rate": 6.850252627143017e-06,
+      "loss": 0.963,
+      "step": 9361
+    },
+    {
+      "epoch": 0.8365090356735989,
+      "grad_norm": 0.4575260877609253,
+      "learning_rate": 6.842943646551564e-06,
+      "loss": 0.9757,
+      "step": 9362
+    },
+    {
+      "epoch": 0.8365983872048607,
+      "grad_norm": 0.4543364346027374,
+      "learning_rate": 6.835638280823614e-06,
+      "loss": 0.9393,
+      "step": 9363
+    },
+    {
+      "epoch": 0.8366877387361226,
+      "grad_norm": 0.47015100717544556,
+      "learning_rate": 6.82833653057105e-06,
+      "loss": 1.009,
+      "step": 9364
+    },
+    {
+      "epoch": 0.8367770902673844,
+      "grad_norm": 0.4650666415691376,
+      "learning_rate": 6.821038396405477e-06,
+      "loss": 0.9546,
+      "step": 9365
+    },
+    {
+      "epoch": 0.8368664417986463,
+      "grad_norm": 0.537955105304718,
+      "learning_rate": 6.813743878938189e-06,
+      "loss": 0.9242,
+      "step": 9366
+    },
+    {
+      "epoch": 0.8369557933299082,
+      "grad_norm": 0.5456058979034424,
+      "learning_rate": 6.806452978780198e-06,
+      "loss": 0.8273,
+      "step": 9367
+    },
+    {
+      "epoch": 0.8370451448611701,
+      "grad_norm": 0.48449915647506714,
+      "learning_rate": 6.79916569654217e-06,
+      "loss": 0.9456,
+      "step": 9368
+    },
+    {
+      "epoch": 0.8371344963924319,
+      "grad_norm": 0.5084838271141052,
+      "learning_rate": 6.791882032834501e-06,
+      "loss": 0.9431,
+      "step": 9369
+    },
+    {
+      "epoch": 0.8372238479236938,
+      "grad_norm": 0.5198284387588501,
+      "learning_rate": 6.784601988267281e-06,
+      "loss": 0.9717,
+      "step": 9370
+    },
+    {
+      "epoch": 0.8373131994549556,
+      "grad_norm": 0.5668700933456421,
+      "learning_rate": 6.777325563450282e-06,
+      "loss": 0.9418,
+      "step": 9371
+    },
+    {
+      "epoch": 0.8374025509862175,
+      "grad_norm": 0.6485138535499573,
+      "learning_rate": 6.77005275899299e-06,
+      "loss": 0.966,
+      "step": 9372
+    },
+    {
+      "epoch": 0.8374919025174794,
+      "grad_norm": 0.39778807759284973,
+      "learning_rate": 6.762783575504578e-06,
+      "loss": 0.9455,
+      "step": 9373
+    },
+    {
+      "epoch": 0.8375812540487413,
+      "grad_norm": 0.48835721611976624,
+      "learning_rate": 6.755518013593914e-06,
+      "loss": 0.978,
+      "step": 9374
+    },
+    {
+      "epoch": 0.8376706055800032,
+      "grad_norm": 0.6808341145515442,
+      "learning_rate": 6.748256073869575e-06,
+      "loss": 0.8574,
+      "step": 9375
+    },
+    {
+      "epoch": 0.8377599571112649,
+      "grad_norm": 0.43278858065605164,
+      "learning_rate": 6.7409977569398265e-06,
+      "loss": 0.964,
+      "step": 9376
+    },
+    {
+      "epoch": 0.8378493086425268,
+      "grad_norm": 0.4334334433078766,
+      "learning_rate": 6.733743063412612e-06,
+      "loss": 0.9898,
+      "step": 9377
+    },
+    {
+      "epoch": 0.8379386601737887,
+      "grad_norm": 0.4900834262371063,
+      "learning_rate": 6.726491993895606e-06,
+      "loss": 0.8638,
+      "step": 9378
+    },
+    {
+      "epoch": 0.8380280117050506,
+      "grad_norm": 0.6439689993858337,
+      "learning_rate": 6.719244548996157e-06,
+      "loss": 0.8321,
+      "step": 9379
+    },
+    {
+      "epoch": 0.8381173632363125,
+      "grad_norm": 0.4418420195579529,
+      "learning_rate": 6.712000729321311e-06,
+      "loss": 0.9594,
+      "step": 9380
+    },
+    {
+      "epoch": 0.8382067147675744,
+      "grad_norm": 0.5338155031204224,
+      "learning_rate": 6.704760535477828e-06,
+      "loss": 0.891,
+      "step": 9381
+    },
+    {
+      "epoch": 0.8382960662988362,
+      "grad_norm": 0.48710137605667114,
+      "learning_rate": 6.697523968072139e-06,
+      "loss": 0.9445,
+      "step": 9382
+    },
+    {
+      "epoch": 0.838385417830098,
+      "grad_norm": 0.42618831992149353,
+      "learning_rate": 6.690291027710394e-06,
+      "loss": 0.9553,
+      "step": 9383
+    },
+    {
+      "epoch": 0.8384747693613599,
+      "grad_norm": 0.4654203951358795,
+      "learning_rate": 6.6830617149984175e-06,
+      "loss": 0.9249,
+      "step": 9384
+    },
+    {
+      "epoch": 0.8385641208926218,
+      "grad_norm": 0.5858107805252075,
+      "learning_rate": 6.675836030541755e-06,
+      "loss": 0.8551,
+      "step": 9385
+    },
+    {
+      "epoch": 0.8386534724238837,
+      "grad_norm": 0.5533519387245178,
+      "learning_rate": 6.668613974945631e-06,
+      "loss": 0.9609,
+      "step": 9386
+    },
+    {
+      "epoch": 0.8387428239551455,
+      "grad_norm": 0.4850926399230957,
+      "learning_rate": 6.661395548814958e-06,
+      "loss": 0.8791,
+      "step": 9387
+    },
+    {
+      "epoch": 0.8388321754864074,
+      "grad_norm": 0.3984906077384949,
+      "learning_rate": 6.6541807527543635e-06,
+      "loss": 0.9768,
+      "step": 9388
+    },
+    {
+      "epoch": 0.8389215270176693,
+      "grad_norm": 0.5376686453819275,
+      "learning_rate": 6.646969587368168e-06,
+      "loss": 0.957,
+      "step": 9389
+    },
+    {
+      "epoch": 0.8390108785489311,
+      "grad_norm": 0.5016261339187622,
+      "learning_rate": 6.6397620532603735e-06,
+      "loss": 1.0018,
+      "step": 9390
+    },
+    {
+      "epoch": 0.839100230080193,
+      "grad_norm": 0.4670695960521698,
+      "learning_rate": 6.632558151034701e-06,
+      "loss": 0.9015,
+      "step": 9391
+    },
+    {
+      "epoch": 0.8391895816114548,
+      "grad_norm": 0.653988778591156,
+      "learning_rate": 6.6253578812945414e-06,
+      "loss": 0.8509,
+      "step": 9392
+    },
+    {
+      "epoch": 0.8392789331427167,
+      "grad_norm": 0.5167982578277588,
+      "learning_rate": 6.6181612446430005e-06,
+      "loss": 0.8648,
+      "step": 9393
+    },
+    {
+      "epoch": 0.8393682846739786,
+      "grad_norm": 0.5537070631980896,
+      "learning_rate": 6.610968241682875e-06,
+      "loss": 0.8879,
+      "step": 9394
+    },
+    {
+      "epoch": 0.8394576362052405,
+      "grad_norm": 0.489687979221344,
+      "learning_rate": 6.603778873016653e-06,
+      "loss": 0.92,
+      "step": 9395
+    },
+    {
+      "epoch": 0.8395469877365024,
+      "grad_norm": 0.4659658670425415,
+      "learning_rate": 6.596593139246532e-06,
+      "loss": 0.9537,
+      "step": 9396
+    },
+    {
+      "epoch": 0.8396363392677642,
+      "grad_norm": 0.45842671394348145,
+      "learning_rate": 6.589411040974369e-06,
+      "loss": 0.9171,
+      "step": 9397
+    },
+    {
+      "epoch": 0.839725690799026,
+      "grad_norm": 0.5154210329055786,
+      "learning_rate": 6.582232578801756e-06,
+      "loss": 0.8698,
+      "step": 9398
+    },
+    {
+      "epoch": 0.8398150423302879,
+      "grad_norm": 0.5210223197937012,
+      "learning_rate": 6.575057753329966e-06,
+      "loss": 0.8417,
+      "step": 9399
+    },
+    {
+      "epoch": 0.8399043938615498,
+      "grad_norm": 0.605532169342041,
+      "learning_rate": 6.567886565159975e-06,
+      "loss": 0.8945,
+      "step": 9400
+    },
+    {
+      "epoch": 0.8399937453928117,
+      "grad_norm": 0.4984580874443054,
+      "learning_rate": 6.560719014892425e-06,
+      "loss": 0.9281,
+      "step": 9401
+    },
+    {
+      "epoch": 0.8400830969240736,
+      "grad_norm": 0.4397487938404083,
+      "learning_rate": 6.55355510312769e-06,
+      "loss": 0.9569,
+      "step": 9402
+    },
+    {
+      "epoch": 0.8401724484553355,
+      "grad_norm": 0.5144175291061401,
+      "learning_rate": 6.546394830465819e-06,
+      "loss": 0.9275,
+      "step": 9403
+    },
+    {
+      "epoch": 0.8402617999865972,
+      "grad_norm": 0.4532061219215393,
+      "learning_rate": 6.539238197506564e-06,
+      "loss": 0.9112,
+      "step": 9404
+    },
+    {
+      "epoch": 0.8403511515178591,
+      "grad_norm": 0.5863621234893799,
+      "learning_rate": 6.53208520484937e-06,
+      "loss": 0.9107,
+      "step": 9405
+    },
+    {
+      "epoch": 0.840440503049121,
+      "grad_norm": 0.5393437147140503,
+      "learning_rate": 6.524935853093383e-06,
+      "loss": 0.8843,
+      "step": 9406
+    },
+    {
+      "epoch": 0.8405298545803829,
+      "grad_norm": 0.4814126193523407,
+      "learning_rate": 6.5177901428374154e-06,
+      "loss": 0.9806,
+      "step": 9407
+    },
+    {
+      "epoch": 0.8406192061116448,
+      "grad_norm": 0.38951998949050903,
+      "learning_rate": 6.510648074680015e-06,
+      "loss": 0.9301,
+      "step": 9408
+    },
+    {
+      "epoch": 0.8407085576429066,
+      "grad_norm": 0.487699955701828,
+      "learning_rate": 6.503509649219403e-06,
+      "loss": 0.9929,
+      "step": 9409
+    },
+    {
+      "epoch": 0.8407979091741685,
+      "grad_norm": 0.6564955711364746,
+      "learning_rate": 6.496374867053495e-06,
+      "loss": 0.8443,
+      "step": 9410
+    },
+    {
+      "epoch": 0.8408872607054303,
+      "grad_norm": 0.4607921838760376,
+      "learning_rate": 6.489243728779904e-06,
+      "loss": 0.9666,
+      "step": 9411
+    },
+    {
+      "epoch": 0.8409766122366922,
+      "grad_norm": 0.6501184105873108,
+      "learning_rate": 6.482116234995944e-06,
+      "loss": 0.8588,
+      "step": 9412
+    },
+    {
+      "epoch": 0.8410659637679541,
+      "grad_norm": 0.48878639936447144,
+      "learning_rate": 6.474992386298617e-06,
+      "loss": 0.964,
+      "step": 9413
+    },
+    {
+      "epoch": 0.8411553152992159,
+      "grad_norm": 0.5011094808578491,
+      "learning_rate": 6.467872183284623e-06,
+      "loss": 0.8937,
+      "step": 9414
+    },
+    {
+      "epoch": 0.8412446668304778,
+      "grad_norm": 0.429160475730896,
+      "learning_rate": 6.460755626550352e-06,
+      "loss": 0.9129,
+      "step": 9415
+    },
+    {
+      "epoch": 0.8413340183617397,
+      "grad_norm": 0.4783921539783478,
+      "learning_rate": 6.453642716691905e-06,
+      "loss": 0.9367,
+      "step": 9416
+    },
+    {
+      "epoch": 0.8414233698930016,
+      "grad_norm": 0.4726746380329132,
+      "learning_rate": 6.446533454305037e-06,
+      "loss": 0.8902,
+      "step": 9417
+    },
+    {
+      "epoch": 0.8415127214242634,
+      "grad_norm": 0.4634154438972473,
+      "learning_rate": 6.43942783998524e-06,
+      "loss": 0.9265,
+      "step": 9418
+    },
+    {
+      "epoch": 0.8416020729555252,
+      "grad_norm": 0.44272157549858093,
+      "learning_rate": 6.432325874327683e-06,
+      "loss": 0.9241,
+      "step": 9419
+    },
+    {
+      "epoch": 0.8416914244867871,
+      "grad_norm": 0.49922677874565125,
+      "learning_rate": 6.425227557927232e-06,
+      "loss": 0.9276,
+      "step": 9420
+    },
+    {
+      "epoch": 0.841780776018049,
+      "grad_norm": 0.4457327723503113,
+      "learning_rate": 6.418132891378448e-06,
+      "loss": 0.9738,
+      "step": 9421
+    },
+    {
+      "epoch": 0.8418701275493109,
+      "grad_norm": 0.48788073658943176,
+      "learning_rate": 6.411041875275581e-06,
+      "loss": 0.9399,
+      "step": 9422
+    },
+    {
+      "epoch": 0.8419594790805728,
+      "grad_norm": 0.4666908383369446,
+      "learning_rate": 6.4039545102125845e-06,
+      "loss": 0.8912,
+      "step": 9423
+    },
+    {
+      "epoch": 0.8420488306118347,
+      "grad_norm": 0.5896921753883362,
+      "learning_rate": 6.396870796783095e-06,
+      "loss": 0.8858,
+      "step": 9424
+    },
+    {
+      "epoch": 0.8421381821430964,
+      "grad_norm": 0.4257669150829315,
+      "learning_rate": 6.389790735580458e-06,
+      "loss": 0.9384,
+      "step": 9425
+    },
+    {
+      "epoch": 0.8422275336743583,
+      "grad_norm": 0.43601882457733154,
+      "learning_rate": 6.382714327197703e-06,
+      "loss": 1.0046,
+      "step": 9426
+    },
+    {
+      "epoch": 0.8423168852056202,
+      "grad_norm": 0.5371537208557129,
+      "learning_rate": 6.375641572227542e-06,
+      "loss": 0.9728,
+      "step": 9427
+    },
+    {
+      "epoch": 0.8424062367368821,
+      "grad_norm": 0.5477351546287537,
+      "learning_rate": 6.368572471262402e-06,
+      "loss": 0.9175,
+      "step": 9428
+    },
+    {
+      "epoch": 0.842495588268144,
+      "grad_norm": 0.4891539216041565,
+      "learning_rate": 6.361507024894392e-06,
+      "loss": 0.9221,
+      "step": 9429
+    },
+    {
+      "epoch": 0.8425849397994059,
+      "grad_norm": 0.5484804511070251,
+      "learning_rate": 6.354445233715328e-06,
+      "loss": 0.8749,
+      "step": 9430
+    },
+    {
+      "epoch": 0.8426742913306676,
+      "grad_norm": 0.5287595391273499,
+      "learning_rate": 6.347387098316698e-06,
+      "loss": 0.9356,
+      "step": 9431
+    },
+    {
+      "epoch": 0.8427636428619295,
+      "grad_norm": 0.4888148605823517,
+      "learning_rate": 6.340332619289701e-06,
+      "loss": 0.9794,
+      "step": 9432
+    },
+    {
+      "epoch": 0.8428529943931914,
+      "grad_norm": 0.5393836498260498,
+      "learning_rate": 6.333281797225244e-06,
+      "loss": 0.8226,
+      "step": 9433
+    },
+    {
+      "epoch": 0.8429423459244533,
+      "grad_norm": 0.5648978352546692,
+      "learning_rate": 6.326234632713873e-06,
+      "loss": 0.902,
+      "step": 9434
+    },
+    {
+      "epoch": 0.8430316974557152,
+      "grad_norm": 0.4472680687904358,
+      "learning_rate": 6.319191126345881e-06,
+      "loss": 0.9544,
+      "step": 9435
+    },
+    {
+      "epoch": 0.843121048986977,
+      "grad_norm": 0.4734359681606293,
+      "learning_rate": 6.312151278711237e-06,
+      "loss": 0.8552,
+      "step": 9436
+    },
+    {
+      "epoch": 0.8432104005182389,
+      "grad_norm": 0.492206335067749,
+      "learning_rate": 6.305115090399616e-06,
+      "loss": 0.9213,
+      "step": 9437
+    },
+    {
+      "epoch": 0.8432997520495007,
+      "grad_norm": 0.4859955608844757,
+      "learning_rate": 6.2980825620003455e-06,
+      "loss": 0.8452,
+      "step": 9438
+    },
+    {
+      "epoch": 0.8433891035807626,
+      "grad_norm": 0.4734523296356201,
+      "learning_rate": 6.291053694102489e-06,
+      "loss": 0.9237,
+      "step": 9439
+    },
+    {
+      "epoch": 0.8434784551120245,
+      "grad_norm": 0.5449217557907104,
+      "learning_rate": 6.284028487294796e-06,
+      "loss": 0.8981,
+      "step": 9440
+    },
+    {
+      "epoch": 0.8435678066432863,
+      "grad_norm": 0.4198647439479828,
+      "learning_rate": 6.277006942165692e-06,
+      "loss": 0.9808,
+      "step": 9441
+    },
+    {
+      "epoch": 0.8436571581745482,
+      "grad_norm": 0.4316485524177551,
+      "learning_rate": 6.269989059303305e-06,
+      "loss": 0.9737,
+      "step": 9442
+    },
+    {
+      "epoch": 0.8437465097058101,
+      "grad_norm": 0.46406179666519165,
+      "learning_rate": 6.262974839295471e-06,
+      "loss": 0.9257,
+      "step": 9443
+    },
+    {
+      "epoch": 0.843835861237072,
+      "grad_norm": 0.4460984170436859,
+      "learning_rate": 6.255964282729692e-06,
+      "loss": 0.9946,
+      "step": 9444
+    },
+    {
+      "epoch": 0.8439252127683338,
+      "grad_norm": 0.511184573173523,
+      "learning_rate": 6.248957390193183e-06,
+      "loss": 0.9457,
+      "step": 9445
+    },
+    {
+      "epoch": 0.8440145642995956,
+      "grad_norm": 0.5544686913490295,
+      "learning_rate": 6.2419541622728565e-06,
+      "loss": 0.9525,
+      "step": 9446
+    },
+    {
+      "epoch": 0.8441039158308575,
+      "grad_norm": 0.44529908895492554,
+      "learning_rate": 6.2349545995552875e-06,
+      "loss": 0.9527,
+      "step": 9447
+    },
+    {
+      "epoch": 0.8441932673621194,
+      "grad_norm": 0.5516801476478577,
+      "learning_rate": 6.227958702626769e-06,
+      "loss": 0.8854,
+      "step": 9448
+    },
+    {
+      "epoch": 0.8442826188933813,
+      "grad_norm": 0.4624561369419098,
+      "learning_rate": 6.220966472073287e-06,
+      "loss": 0.9874,
+      "step": 9449
+    },
+    {
+      "epoch": 0.8443719704246432,
+      "grad_norm": 0.4866270422935486,
+      "learning_rate": 6.213977908480518e-06,
+      "loss": 0.9601,
+      "step": 9450
+    },
+    {
+      "epoch": 0.8444613219559051,
+      "grad_norm": 0.4873912036418915,
+      "learning_rate": 6.206993012433815e-06,
+      "loss": 0.9031,
+      "step": 9451
+    },
+    {
+      "epoch": 0.8445506734871668,
+      "grad_norm": 0.4816325306892395,
+      "learning_rate": 6.200011784518256e-06,
+      "loss": 0.9123,
+      "step": 9452
+    },
+    {
+      "epoch": 0.8446400250184287,
+      "grad_norm": 0.5002326965332031,
+      "learning_rate": 6.193034225318578e-06,
+      "loss": 0.8816,
+      "step": 9453
+    },
+    {
+      "epoch": 0.8447293765496906,
+      "grad_norm": 0.4894864559173584,
+      "learning_rate": 6.186060335419236e-06,
+      "loss": 0.9751,
+      "step": 9454
+    },
+    {
+      "epoch": 0.8448187280809525,
+      "grad_norm": 0.5670939683914185,
+      "learning_rate": 6.179090115404368e-06,
+      "loss": 0.8294,
+      "step": 9455
+    },
+    {
+      "epoch": 0.8449080796122144,
+      "grad_norm": 0.553982138633728,
+      "learning_rate": 6.172123565857796e-06,
+      "loss": 0.8847,
+      "step": 9456
+    },
+    {
+      "epoch": 0.8449974311434763,
+      "grad_norm": 0.48611873388290405,
+      "learning_rate": 6.165160687363053e-06,
+      "loss": 0.9528,
+      "step": 9457
+    },
+    {
+      "epoch": 0.8450867826747381,
+      "grad_norm": 0.4134733974933624,
+      "learning_rate": 6.158201480503345e-06,
+      "loss": 1.0282,
+      "step": 9458
+    },
+    {
+      "epoch": 0.8451761342059999,
+      "grad_norm": 0.6506361961364746,
+      "learning_rate": 6.151245945861578e-06,
+      "loss": 0.8545,
+      "step": 9459
+    },
+    {
+      "epoch": 0.8452654857372618,
+      "grad_norm": 0.49033278226852417,
+      "learning_rate": 6.1442940840203575e-06,
+      "loss": 0.857,
+      "step": 9460
+    },
+    {
+      "epoch": 0.8453548372685237,
+      "grad_norm": 0.5886410474777222,
+      "learning_rate": 6.13734589556198e-06,
+      "loss": 0.8042,
+      "step": 9461
+    },
+    {
+      "epoch": 0.8454441887997856,
+      "grad_norm": 0.4791713058948517,
+      "learning_rate": 6.130401381068424e-06,
+      "loss": 0.9883,
+      "step": 9462
+    },
+    {
+      "epoch": 0.8455335403310474,
+      "grad_norm": 0.5414884686470032,
+      "learning_rate": 6.123460541121368e-06,
+      "loss": 0.9804,
+      "step": 9463
+    },
+    {
+      "epoch": 0.8456228918623093,
+      "grad_norm": 0.5500231981277466,
+      "learning_rate": 6.11652337630218e-06,
+      "loss": 0.9065,
+      "step": 9464
+    },
+    {
+      "epoch": 0.8457122433935712,
+      "grad_norm": 0.4416286051273346,
+      "learning_rate": 6.109589887191924e-06,
+      "loss": 0.9156,
+      "step": 9465
+    },
+    {
+      "epoch": 0.845801594924833,
+      "grad_norm": 0.4755689799785614,
+      "learning_rate": 6.102660074371369e-06,
+      "loss": 0.9018,
+      "step": 9466
+    },
+    {
+      "epoch": 0.8458909464560949,
+      "grad_norm": 0.5172240138053894,
+      "learning_rate": 6.095733938420928e-06,
+      "loss": 0.9731,
+      "step": 9467
+    },
+    {
+      "epoch": 0.8459802979873567,
+      "grad_norm": 0.47983941435813904,
+      "learning_rate": 6.0888114799207586e-06,
+      "loss": 0.99,
+      "step": 9468
+    },
+    {
+      "epoch": 0.8460696495186186,
+      "grad_norm": 0.4617604911327362,
+      "learning_rate": 6.081892699450687e-06,
+      "loss": 0.9207,
+      "step": 9469
+    },
+    {
+      "epoch": 0.8461590010498805,
+      "grad_norm": 0.43974384665489197,
+      "learning_rate": 6.0749775975902424e-06,
+      "loss": 0.919,
+      "step": 9470
+    },
+    {
+      "epoch": 0.8462483525811424,
+      "grad_norm": 0.5544477701187134,
+      "learning_rate": 6.068066174918624e-06,
+      "loss": 0.9164,
+      "step": 9471
+    },
+    {
+      "epoch": 0.8463377041124043,
+      "grad_norm": 0.5196595191955566,
+      "learning_rate": 6.0611584320147465e-06,
+      "loss": 0.9605,
+      "step": 9472
+    },
+    {
+      "epoch": 0.846427055643666,
+      "grad_norm": 0.47909513115882874,
+      "learning_rate": 6.054254369457202e-06,
+      "loss": 0.8867,
+      "step": 9473
+    },
+    {
+      "epoch": 0.8465164071749279,
+      "grad_norm": 0.5090166926383972,
+      "learning_rate": 6.047353987824278e-06,
+      "loss": 0.8925,
+      "step": 9474
+    },
+    {
+      "epoch": 0.8466057587061898,
+      "grad_norm": 0.4992344081401825,
+      "learning_rate": 6.040457287693963e-06,
+      "loss": 0.9032,
+      "step": 9475
+    },
+    {
+      "epoch": 0.8466951102374517,
+      "grad_norm": 0.5999143123626709,
+      "learning_rate": 6.033564269643926e-06,
+      "loss": 0.9008,
+      "step": 9476
+    },
+    {
+      "epoch": 0.8467844617687136,
+      "grad_norm": 0.5238880515098572,
+      "learning_rate": 6.026674934251542e-06,
+      "loss": 0.9067,
+      "step": 9477
+    },
+    {
+      "epoch": 0.8468738132999755,
+      "grad_norm": 0.5634439587593079,
+      "learning_rate": 6.0197892820938405e-06,
+      "loss": 0.8967,
+      "step": 9478
+    },
+    {
+      "epoch": 0.8469631648312373,
+      "grad_norm": 0.4486282467842102,
+      "learning_rate": 6.012907313747585e-06,
+      "loss": 0.9542,
+      "step": 9479
+    },
+    {
+      "epoch": 0.8470525163624991,
+      "grad_norm": 0.49240148067474365,
+      "learning_rate": 6.006029029789206e-06,
+      "loss": 0.9843,
+      "step": 9480
+    },
+    {
+      "epoch": 0.847141867893761,
+      "grad_norm": 0.5785838961601257,
+      "learning_rate": 5.99915443079484e-06,
+      "loss": 0.9319,
+      "step": 9481
+    },
+    {
+      "epoch": 0.8472312194250229,
+      "grad_norm": 0.6052419543266296,
+      "learning_rate": 5.992283517340308e-06,
+      "loss": 0.7929,
+      "step": 9482
+    },
+    {
+      "epoch": 0.8473205709562848,
+      "grad_norm": 0.452104777097702,
+      "learning_rate": 5.985416290001117e-06,
+      "loss": 0.9212,
+      "step": 9483
+    },
+    {
+      "epoch": 0.8474099224875467,
+      "grad_norm": 0.49323323369026184,
+      "learning_rate": 5.978552749352473e-06,
+      "loss": 0.8627,
+      "step": 9484
+    },
+    {
+      "epoch": 0.8474992740188085,
+      "grad_norm": 0.5414223670959473,
+      "learning_rate": 5.971692895969272e-06,
+      "loss": 0.9574,
+      "step": 9485
+    },
+    {
+      "epoch": 0.8475886255500704,
+      "grad_norm": 0.6366201043128967,
+      "learning_rate": 5.964836730426099e-06,
+      "loss": 0.9616,
+      "step": 9486
+    },
+    {
+      "epoch": 0.8476779770813322,
+      "grad_norm": 0.6016992926597595,
+      "learning_rate": 5.957984253297244e-06,
+      "loss": 0.9459,
+      "step": 9487
+    },
+    {
+      "epoch": 0.8477673286125941,
+      "grad_norm": 0.48284274339675903,
+      "learning_rate": 5.9511354651566485e-06,
+      "loss": 0.9174,
+      "step": 9488
+    },
+    {
+      "epoch": 0.847856680143856,
+      "grad_norm": 0.42610475420951843,
+      "learning_rate": 5.944290366577987e-06,
+      "loss": 0.965,
+      "step": 9489
+    },
+    {
+      "epoch": 0.8479460316751178,
+      "grad_norm": 0.4493615925312042,
+      "learning_rate": 5.937448958134606e-06,
+      "loss": 0.9472,
+      "step": 9490
+    },
+    {
+      "epoch": 0.8480353832063797,
+      "grad_norm": 0.46299251914024353,
+      "learning_rate": 5.930611240399548e-06,
+      "loss": 0.948,
+      "step": 9491
+    },
+    {
+      "epoch": 0.8481247347376416,
+      "grad_norm": 0.482902854681015,
+      "learning_rate": 5.923777213945541e-06,
+      "loss": 0.905,
+      "step": 9492
+    },
+    {
+      "epoch": 0.8482140862689035,
+      "grad_norm": 0.5270372033119202,
+      "learning_rate": 5.916946879345015e-06,
+      "loss": 0.9044,
+      "step": 9493
+    },
+    {
+      "epoch": 0.8483034378001653,
+      "grad_norm": 0.47987642884254456,
+      "learning_rate": 5.91012023717008e-06,
+      "loss": 0.9566,
+      "step": 9494
+    },
+    {
+      "epoch": 0.8483927893314271,
+      "grad_norm": 0.4810824990272522,
+      "learning_rate": 5.903297287992538e-06,
+      "loss": 0.8675,
+      "step": 9495
+    },
+    {
+      "epoch": 0.848482140862689,
+      "grad_norm": 0.5205212831497192,
+      "learning_rate": 5.89647803238389e-06,
+      "loss": 0.9361,
+      "step": 9496
+    },
+    {
+      "epoch": 0.8485714923939509,
+      "grad_norm": 0.45572689175605774,
+      "learning_rate": 5.889662470915319e-06,
+      "loss": 0.9834,
+      "step": 9497
+    },
+    {
+      "epoch": 0.8486608439252128,
+      "grad_norm": 0.4855954945087433,
+      "learning_rate": 5.882850604157691e-06,
+      "loss": 0.8503,
+      "step": 9498
+    },
+    {
+      "epoch": 0.8487501954564747,
+      "grad_norm": 0.4647740423679352,
+      "learning_rate": 5.876042432681584e-06,
+      "loss": 0.943,
+      "step": 9499
+    },
+    {
+      "epoch": 0.8488395469877364,
+      "grad_norm": 0.5627195239067078,
+      "learning_rate": 5.869237957057244e-06,
+      "loss": 0.8846,
+      "step": 9500
+    },
+    {
+      "epoch": 0.8489288985189983,
+      "grad_norm": 0.4476568102836609,
+      "learning_rate": 5.86243717785463e-06,
+      "loss": 0.9631,
+      "step": 9501
+    },
+    {
+      "epoch": 0.8490182500502602,
+      "grad_norm": 0.42511650919914246,
+      "learning_rate": 5.855640095643372e-06,
+      "loss": 0.8941,
+      "step": 9502
+    },
+    {
+      "epoch": 0.8491076015815221,
+      "grad_norm": 0.6032181978225708,
+      "learning_rate": 5.848846710992817e-06,
+      "loss": 0.9229,
+      "step": 9503
+    },
+    {
+      "epoch": 0.849196953112784,
+      "grad_norm": 0.49917519092559814,
+      "learning_rate": 5.842057024471953e-06,
+      "loss": 0.9726,
+      "step": 9504
+    },
+    {
+      "epoch": 0.8492863046440459,
+      "grad_norm": 0.4302211105823517,
+      "learning_rate": 5.835271036649503e-06,
+      "loss": 0.9793,
+      "step": 9505
+    },
+    {
+      "epoch": 0.8493756561753077,
+      "grad_norm": 0.6006591320037842,
+      "learning_rate": 5.8284887480938636e-06,
+      "loss": 0.9136,
+      "step": 9506
+    },
+    {
+      "epoch": 0.8494650077065695,
+      "grad_norm": 0.5111293792724609,
+      "learning_rate": 5.821710159373128e-06,
+      "loss": 0.8993,
+      "step": 9507
+    },
+    {
+      "epoch": 0.8495543592378314,
+      "grad_norm": 0.5666483640670776,
+      "learning_rate": 5.814935271055083e-06,
+      "loss": 0.921,
+      "step": 9508
+    },
+    {
+      "epoch": 0.8496437107690933,
+      "grad_norm": 0.522527813911438,
+      "learning_rate": 5.808164083707179e-06,
+      "loss": 0.9923,
+      "step": 9509
+    },
+    {
+      "epoch": 0.8497330623003552,
+      "grad_norm": 0.5020400285720825,
+      "learning_rate": 5.8013965978965855e-06,
+      "loss": 0.9426,
+      "step": 9510
+    },
+    {
+      "epoch": 0.849822413831617,
+      "grad_norm": 0.4437236487865448,
+      "learning_rate": 5.794632814190148e-06,
+      "loss": 0.9501,
+      "step": 9511
+    },
+    {
+      "epoch": 0.8499117653628789,
+      "grad_norm": 0.49814578890800476,
+      "learning_rate": 5.787872733154409e-06,
+      "loss": 0.936,
+      "step": 9512
+    },
+    {
+      "epoch": 0.8500011168941408,
+      "grad_norm": 0.550423264503479,
+      "learning_rate": 5.781116355355593e-06,
+      "loss": 1.0181,
+      "step": 9513
+    },
+    {
+      "epoch": 0.8500904684254026,
+      "grad_norm": 0.7392222285270691,
+      "learning_rate": 5.774363681359624e-06,
+      "loss": 0.9255,
+      "step": 9514
+    },
+    {
+      "epoch": 0.8501798199566645,
+      "grad_norm": 0.5093478560447693,
+      "learning_rate": 5.767614711732111e-06,
+      "loss": 0.8951,
+      "step": 9515
+    },
+    {
+      "epoch": 0.8502691714879264,
+      "grad_norm": 0.5823287963867188,
+      "learning_rate": 5.760869447038348e-06,
+      "loss": 0.861,
+      "step": 9516
+    },
+    {
+      "epoch": 0.8503585230191882,
+      "grad_norm": 0.42861655354499817,
+      "learning_rate": 5.75412788784333e-06,
+      "loss": 0.9809,
+      "step": 9517
+    },
+    {
+      "epoch": 0.8504478745504501,
+      "grad_norm": 0.44293922185897827,
+      "learning_rate": 5.74739003471172e-06,
+      "loss": 0.9667,
+      "step": 9518
+    },
+    {
+      "epoch": 0.850537226081712,
+      "grad_norm": 0.46408700942993164,
+      "learning_rate": 5.740655888207897e-06,
+      "loss": 0.9613,
+      "step": 9519
+    },
+    {
+      "epoch": 0.8506265776129739,
+      "grad_norm": 0.42871806025505066,
+      "learning_rate": 5.73392544889591e-06,
+      "loss": 0.987,
+      "step": 9520
+    },
+    {
+      "epoch": 0.8507159291442357,
+      "grad_norm": 0.5720949172973633,
+      "learning_rate": 5.727198717339511e-06,
+      "loss": 0.8512,
+      "step": 9521
+    },
+    {
+      "epoch": 0.8508052806754975,
+      "grad_norm": 0.46692514419555664,
+      "learning_rate": 5.720475694102128e-06,
+      "loss": 0.9157,
+      "step": 9522
+    },
+    {
+      "epoch": 0.8508946322067594,
+      "grad_norm": 0.47131991386413574,
+      "learning_rate": 5.713756379746898e-06,
+      "loss": 0.9097,
+      "step": 9523
+    },
+    {
+      "epoch": 0.8509839837380213,
+      "grad_norm": 0.5264059901237488,
+      "learning_rate": 5.707040774836625e-06,
+      "loss": 0.8181,
+      "step": 9524
+    },
+    {
+      "epoch": 0.8510733352692832,
+      "grad_norm": 0.5986820459365845,
+      "learning_rate": 5.70032887993382e-06,
+      "loss": 0.9024,
+      "step": 9525
+    },
+    {
+      "epoch": 0.8511626868005451,
+      "grad_norm": 0.5732765197753906,
+      "learning_rate": 5.693620695600671e-06,
+      "loss": 0.8888,
+      "step": 9526
+    },
+    {
+      "epoch": 0.851252038331807,
+      "grad_norm": 0.5805981755256653,
+      "learning_rate": 5.686916222399069e-06,
+      "loss": 0.8615,
+      "step": 9527
+    },
+    {
+      "epoch": 0.8513413898630687,
+      "grad_norm": 0.4108264744281769,
+      "learning_rate": 5.68021546089057e-06,
+      "loss": 1.0108,
+      "step": 9528
+    },
+    {
+      "epoch": 0.8514307413943306,
+      "grad_norm": 0.48710349202156067,
+      "learning_rate": 5.673518411636436e-06,
+      "loss": 0.9861,
+      "step": 9529
+    },
+    {
+      "epoch": 0.8515200929255925,
+      "grad_norm": 0.5818296670913696,
+      "learning_rate": 5.666825075197624e-06,
+      "loss": 0.882,
+      "step": 9530
+    },
+    {
+      "epoch": 0.8516094444568544,
+      "grad_norm": 0.433623343706131,
+      "learning_rate": 5.6601354521347685e-06,
+      "loss": 0.9407,
+      "step": 9531
+    },
+    {
+      "epoch": 0.8516987959881163,
+      "grad_norm": 0.5480315685272217,
+      "learning_rate": 5.653449543008199e-06,
+      "loss": 0.8762,
+      "step": 9532
+    },
+    {
+      "epoch": 0.8517881475193781,
+      "grad_norm": 0.45491036772727966,
+      "learning_rate": 5.646767348377935e-06,
+      "loss": 0.9376,
+      "step": 9533
+    },
+    {
+      "epoch": 0.85187749905064,
+      "grad_norm": 0.43504709005355835,
+      "learning_rate": 5.640088868803673e-06,
+      "loss": 0.9852,
+      "step": 9534
+    },
+    {
+      "epoch": 0.8519668505819018,
+      "grad_norm": 0.5115997791290283,
+      "learning_rate": 5.633414104844808e-06,
+      "loss": 0.9546,
+      "step": 9535
+    },
+    {
+      "epoch": 0.8520562021131637,
+      "grad_norm": 0.3994527757167816,
+      "learning_rate": 5.6267430570604405e-06,
+      "loss": 0.9405,
+      "step": 9536
+    },
+    {
+      "epoch": 0.8521455536444256,
+      "grad_norm": 0.5334661602973938,
+      "learning_rate": 5.620075726009311e-06,
+      "loss": 0.9164,
+      "step": 9537
+    },
+    {
+      "epoch": 0.8522349051756875,
+      "grad_norm": 0.41900575160980225,
+      "learning_rate": 5.6134121122498995e-06,
+      "loss": 0.9545,
+      "step": 9538
+    },
+    {
+      "epoch": 0.8523242567069493,
+      "grad_norm": 0.583827555179596,
+      "learning_rate": 5.606752216340349e-06,
+      "loss": 0.8339,
+      "step": 9539
+    },
+    {
+      "epoch": 0.8524136082382112,
+      "grad_norm": 0.6421988606452942,
+      "learning_rate": 5.6000960388385095e-06,
+      "loss": 0.8723,
+      "step": 9540
+    },
+    {
+      "epoch": 0.8525029597694731,
+      "grad_norm": 0.4667465090751648,
+      "learning_rate": 5.593443580301888e-06,
+      "loss": 0.934,
+      "step": 9541
+    },
+    {
+      "epoch": 0.8525923113007349,
+      "grad_norm": 0.48862892389297485,
+      "learning_rate": 5.586794841287701e-06,
+      "loss": 0.9056,
+      "step": 9542
+    },
+    {
+      "epoch": 0.8526816628319968,
+      "grad_norm": 0.4161957800388336,
+      "learning_rate": 5.5801498223528644e-06,
+      "loss": 0.9632,
+      "step": 9543
+    },
+    {
+      "epoch": 0.8527710143632586,
+      "grad_norm": 0.6144065260887146,
+      "learning_rate": 5.573508524053955e-06,
+      "loss": 0.8685,
+      "step": 9544
+    },
+    {
+      "epoch": 0.8528603658945205,
+      "grad_norm": 0.44452551007270813,
+      "learning_rate": 5.566870946947261e-06,
+      "loss": 0.9958,
+      "step": 9545
+    },
+    {
+      "epoch": 0.8529497174257824,
+      "grad_norm": 0.6683482527732849,
+      "learning_rate": 5.5602370915887494e-06,
+      "loss": 0.8042,
+      "step": 9546
+    },
+    {
+      "epoch": 0.8530390689570443,
+      "grad_norm": 0.4790489971637726,
+      "learning_rate": 5.553606958534074e-06,
+      "loss": 0.9908,
+      "step": 9547
+    },
+    {
+      "epoch": 0.8531284204883062,
+      "grad_norm": 0.4621983766555786,
+      "learning_rate": 5.546980548338593e-06,
+      "loss": 0.9601,
+      "step": 9548
+    },
+    {
+      "epoch": 0.8532177720195679,
+      "grad_norm": 0.42011430859565735,
+      "learning_rate": 5.54035786155731e-06,
+      "loss": 0.965,
+      "step": 9549
+    },
+    {
+      "epoch": 0.8533071235508298,
+      "grad_norm": 0.4877914488315582,
+      "learning_rate": 5.533738898744967e-06,
+      "loss": 0.8975,
+      "step": 9550
+    },
+    {
+      "epoch": 0.8533964750820917,
+      "grad_norm": 0.49201685190200806,
+      "learning_rate": 5.527123660455969e-06,
+      "loss": 0.9032,
+      "step": 9551
+    },
+    {
+      "epoch": 0.8534858266133536,
+      "grad_norm": 0.4355340898036957,
+      "learning_rate": 5.5205121472444055e-06,
+      "loss": 0.9941,
+      "step": 9552
+    },
+    {
+      "epoch": 0.8535751781446155,
+      "grad_norm": 0.43532299995422363,
+      "learning_rate": 5.513904359664074e-06,
+      "loss": 0.9603,
+      "step": 9553
+    },
+    {
+      "epoch": 0.8536645296758774,
+      "grad_norm": 0.5881833434104919,
+      "learning_rate": 5.507300298268436e-06,
+      "loss": 0.9591,
+      "step": 9554
+    },
+    {
+      "epoch": 0.8537538812071392,
+      "grad_norm": 0.4763388931751251,
+      "learning_rate": 5.500699963610656e-06,
+      "loss": 0.9825,
+      "step": 9555
+    },
+    {
+      "epoch": 0.853843232738401,
+      "grad_norm": 0.45608317852020264,
+      "learning_rate": 5.49410335624358e-06,
+      "loss": 0.9578,
+      "step": 9556
+    },
+    {
+      "epoch": 0.8539325842696629,
+      "grad_norm": 0.5364043712615967,
+      "learning_rate": 5.487510476719748e-06,
+      "loss": 1.0419,
+      "step": 9557
+    },
+    {
+      "epoch": 0.8540219358009248,
+      "grad_norm": 0.41415050625801086,
+      "learning_rate": 5.480921325591398e-06,
+      "loss": 0.9417,
+      "step": 9558
+    },
+    {
+      "epoch": 0.8541112873321867,
+      "grad_norm": 0.42311692237854004,
+      "learning_rate": 5.47433590341041e-06,
+      "loss": 0.9467,
+      "step": 9559
+    },
+    {
+      "epoch": 0.8542006388634485,
+      "grad_norm": 0.44652259349823,
+      "learning_rate": 5.467754210728404e-06,
+      "loss": 1.0331,
+      "step": 9560
+    },
+    {
+      "epoch": 0.8542899903947104,
+      "grad_norm": 0.4289079010486603,
+      "learning_rate": 5.4611762480966555e-06,
+      "loss": 0.9254,
+      "step": 9561
+    },
+    {
+      "epoch": 0.8543793419259722,
+      "grad_norm": 0.4332316219806671,
+      "learning_rate": 5.4546020160661515e-06,
+      "loss": 1.0019,
+      "step": 9562
+    },
+    {
+      "epoch": 0.8544686934572341,
+      "grad_norm": 0.5433120727539062,
+      "learning_rate": 5.448031515187552e-06,
+      "loss": 0.8531,
+      "step": 9563
+    },
+    {
+      "epoch": 0.854558044988496,
+      "grad_norm": 0.4383524954319,
+      "learning_rate": 5.441464746011194e-06,
+      "loss": 0.9369,
+      "step": 9564
+    },
+    {
+      "epoch": 0.8546473965197579,
+      "grad_norm": 0.4835048019886017,
+      "learning_rate": 5.434901709087131e-06,
+      "loss": 0.9205,
+      "step": 9565
+    },
+    {
+      "epoch": 0.8547367480510197,
+      "grad_norm": 0.44369155168533325,
+      "learning_rate": 5.428342404965076e-06,
+      "loss": 0.9472,
+      "step": 9566
+    },
+    {
+      "epoch": 0.8548260995822816,
+      "grad_norm": 0.4760458469390869,
+      "learning_rate": 5.421786834194442e-06,
+      "loss": 0.8901,
+      "step": 9567
+    },
+    {
+      "epoch": 0.8549154511135435,
+      "grad_norm": 0.4910159707069397,
+      "learning_rate": 5.415234997324348e-06,
+      "loss": 1.0034,
+      "step": 9568
+    },
+    {
+      "epoch": 0.8550048026448053,
+      "grad_norm": 0.600982129573822,
+      "learning_rate": 5.4086868949035475e-06,
+      "loss": 0.838,
+      "step": 9569
+    },
+    {
+      "epoch": 0.8550941541760672,
+      "grad_norm": 0.6115323901176453,
+      "learning_rate": 5.4021425274805245e-06,
+      "loss": 0.8826,
+      "step": 9570
+    },
+    {
+      "epoch": 0.855183505707329,
+      "grad_norm": 0.41822290420532227,
+      "learning_rate": 5.395601895603453e-06,
+      "loss": 0.9599,
+      "step": 9571
+    },
+    {
+      "epoch": 0.8552728572385909,
+      "grad_norm": 0.5147873163223267,
+      "learning_rate": 5.389064999820165e-06,
+      "loss": 0.9285,
+      "step": 9572
+    },
+    {
+      "epoch": 0.8553622087698528,
+      "grad_norm": 0.5221825838088989,
+      "learning_rate": 5.3825318406782125e-06,
+      "loss": 0.9447,
+      "step": 9573
+    },
+    {
+      "epoch": 0.8554515603011147,
+      "grad_norm": 0.4737570285797119,
+      "learning_rate": 5.376002418724796e-06,
+      "loss": 0.9483,
+      "step": 9574
+    },
+    {
+      "epoch": 0.8555409118323766,
+      "grad_norm": 0.5008856058120728,
+      "learning_rate": 5.3694767345068345e-06,
+      "loss": 0.8795,
+      "step": 9575
+    },
+    {
+      "epoch": 0.8556302633636383,
+      "grad_norm": 0.4690033197402954,
+      "learning_rate": 5.362954788570928e-06,
+      "loss": 0.9913,
+      "step": 9576
+    },
+    {
+      "epoch": 0.8557196148949002,
+      "grad_norm": 0.5061206221580505,
+      "learning_rate": 5.356436581463353e-06,
+      "loss": 0.8482,
+      "step": 9577
+    },
+    {
+      "epoch": 0.8558089664261621,
+      "grad_norm": 0.5794250965118408,
+      "learning_rate": 5.3499221137300885e-06,
+      "loss": 0.8922,
+      "step": 9578
+    },
+    {
+      "epoch": 0.855898317957424,
+      "grad_norm": 0.5473932027816772,
+      "learning_rate": 5.343411385916769e-06,
+      "loss": 0.9497,
+      "step": 9579
+    },
+    {
+      "epoch": 0.8559876694886859,
+      "grad_norm": 0.42786940932273865,
+      "learning_rate": 5.3369043985687595e-06,
+      "loss": 0.9562,
+      "step": 9580
+    },
+    {
+      "epoch": 0.8560770210199478,
+      "grad_norm": 0.5070851445198059,
+      "learning_rate": 5.330401152231074e-06,
+      "loss": 0.9229,
+      "step": 9581
+    },
+    {
+      "epoch": 0.8561663725512096,
+      "grad_norm": 0.45324572920799255,
+      "learning_rate": 5.323901647448437e-06,
+      "loss": 1.023,
+      "step": 9582
+    },
+    {
+      "epoch": 0.8562557240824714,
+      "grad_norm": 0.6096236109733582,
+      "learning_rate": 5.317405884765253e-06,
+      "loss": 0.9328,
+      "step": 9583
+    },
+    {
+      "epoch": 0.8563450756137333,
+      "grad_norm": 0.5863158106803894,
+      "learning_rate": 5.310913864725609e-06,
+      "loss": 0.915,
+      "step": 9584
+    },
+    {
+      "epoch": 0.8564344271449952,
+      "grad_norm": 0.4709130525588989,
+      "learning_rate": 5.304425587873274e-06,
+      "loss": 0.8696,
+      "step": 9585
+    },
+    {
+      "epoch": 0.8565237786762571,
+      "grad_norm": 0.607227623462677,
+      "learning_rate": 5.29794105475172e-06,
+      "loss": 0.924,
+      "step": 9586
+    },
+    {
+      "epoch": 0.856613130207519,
+      "grad_norm": 0.4545291066169739,
+      "learning_rate": 5.291460265904097e-06,
+      "loss": 0.9702,
+      "step": 9587
+    },
+    {
+      "epoch": 0.8567024817387808,
+      "grad_norm": 0.44768497347831726,
+      "learning_rate": 5.284983221873241e-06,
+      "loss": 1.0336,
+      "step": 9588
+    },
+    {
+      "epoch": 0.8567918332700427,
+      "grad_norm": 0.666070282459259,
+      "learning_rate": 5.278509923201657e-06,
+      "loss": 0.8109,
+      "step": 9589
+    },
+    {
+      "epoch": 0.8568811848013045,
+      "grad_norm": 0.5126855969429016,
+      "learning_rate": 5.2720403704315635e-06,
+      "loss": 0.904,
+      "step": 9590
+    },
+    {
+      "epoch": 0.8569705363325664,
+      "grad_norm": 0.47559547424316406,
+      "learning_rate": 5.265574564104858e-06,
+      "loss": 0.9819,
+      "step": 9591
+    },
+    {
+      "epoch": 0.8570598878638283,
+      "grad_norm": 0.43424925208091736,
+      "learning_rate": 5.259112504763114e-06,
+      "loss": 0.9388,
+      "step": 9592
+    },
+    {
+      "epoch": 0.8571492393950901,
+      "grad_norm": 0.3889505863189697,
+      "learning_rate": 5.252654192947603e-06,
+      "loss": 1.003,
+      "step": 9593
+    },
+    {
+      "epoch": 0.857238590926352,
+      "grad_norm": 0.47286322712898254,
+      "learning_rate": 5.246199629199277e-06,
+      "loss": 0.8823,
+      "step": 9594
+    },
+    {
+      "epoch": 0.8573279424576139,
+      "grad_norm": 0.46248650550842285,
+      "learning_rate": 5.23974881405877e-06,
+      "loss": 0.9612,
+      "step": 9595
+    },
+    {
+      "epoch": 0.8574172939888758,
+      "grad_norm": 0.5204552412033081,
+      "learning_rate": 5.233301748066416e-06,
+      "loss": 0.9113,
+      "step": 9596
+    },
+    {
+      "epoch": 0.8575066455201376,
+      "grad_norm": 0.42434239387512207,
+      "learning_rate": 5.226858431762216e-06,
+      "loss": 0.9449,
+      "step": 9597
+    },
+    {
+      "epoch": 0.8575959970513994,
+      "grad_norm": 0.6636669039726257,
+      "learning_rate": 5.2204188656858775e-06,
+      "loss": 0.9023,
+      "step": 9598
+    },
+    {
+      "epoch": 0.8576853485826613,
+      "grad_norm": 0.49965164065361023,
+      "learning_rate": 5.213983050376764e-06,
+      "loss": 0.9547,
+      "step": 9599
+    },
+    {
+      "epoch": 0.8577747001139232,
+      "grad_norm": 0.43195703625679016,
+      "learning_rate": 5.207550986373961e-06,
+      "loss": 0.9084,
+      "step": 9600
+    },
+    {
+      "epoch": 0.8578640516451851,
+      "grad_norm": 0.4506111145019531,
+      "learning_rate": 5.201122674216208e-06,
+      "loss": 0.9734,
+      "step": 9601
+    },
+    {
+      "epoch": 0.857953403176447,
+      "grad_norm": 0.4218018054962158,
+      "learning_rate": 5.194698114441959e-06,
+      "loss": 0.9299,
+      "step": 9602
+    },
+    {
+      "epoch": 0.8580427547077089,
+      "grad_norm": 0.5177663564682007,
+      "learning_rate": 5.188277307589334e-06,
+      "loss": 0.8914,
+      "step": 9603
+    },
+    {
+      "epoch": 0.8581321062389706,
+      "grad_norm": 0.5173326134681702,
+      "learning_rate": 5.181860254196136e-06,
+      "loss": 0.9674,
+      "step": 9604
+    },
+    {
+      "epoch": 0.8582214577702325,
+      "grad_norm": 0.49760016798973083,
+      "learning_rate": 5.175446954799873e-06,
+      "loss": 0.8798,
+      "step": 9605
+    },
+    {
+      "epoch": 0.8583108093014944,
+      "grad_norm": 0.48603054881095886,
+      "learning_rate": 5.169037409937733e-06,
+      "loss": 0.8737,
+      "step": 9606
+    },
+    {
+      "epoch": 0.8584001608327563,
+      "grad_norm": 0.4998380243778229,
+      "learning_rate": 5.162631620146563e-06,
+      "loss": 0.9471,
+      "step": 9607
+    },
+    {
+      "epoch": 0.8584895123640182,
+      "grad_norm": 0.5357629656791687,
+      "learning_rate": 5.15622958596293e-06,
+      "loss": 0.9189,
+      "step": 9608
+    },
+    {
+      "epoch": 0.85857886389528,
+      "grad_norm": 0.5003303289413452,
+      "learning_rate": 5.149831307923064e-06,
+      "loss": 0.9853,
+      "step": 9609
+    },
+    {
+      "epoch": 0.8586682154265419,
+      "grad_norm": 0.4467265009880066,
+      "learning_rate": 5.1434367865629e-06,
+      "loss": 0.9858,
+      "step": 9610
+    },
+    {
+      "epoch": 0.8587575669578037,
+      "grad_norm": 0.6121494770050049,
+      "learning_rate": 5.137046022418046e-06,
+      "loss": 0.859,
+      "step": 9611
+    },
+    {
+      "epoch": 0.8588469184890656,
+      "grad_norm": 0.5334439873695374,
+      "learning_rate": 5.130659016023787e-06,
+      "loss": 0.884,
+      "step": 9612
+    },
+    {
+      "epoch": 0.8589362700203275,
+      "grad_norm": 0.4914329946041107,
+      "learning_rate": 5.12427576791511e-06,
+      "loss": 0.9521,
+      "step": 9613
+    },
+    {
+      "epoch": 0.8590256215515893,
+      "grad_norm": 0.5462926626205444,
+      "learning_rate": 5.117896278626671e-06,
+      "loss": 0.9285,
+      "step": 9614
+    },
+    {
+      "epoch": 0.8591149730828512,
+      "grad_norm": 0.4273386001586914,
+      "learning_rate": 5.111520548692833e-06,
+      "loss": 0.9606,
+      "step": 9615
+    },
+    {
+      "epoch": 0.8592043246141131,
+      "grad_norm": 0.6681180596351624,
+      "learning_rate": 5.105148578647623e-06,
+      "loss": 0.773,
+      "step": 9616
+    },
+    {
+      "epoch": 0.859293676145375,
+      "grad_norm": 0.4087236225605011,
+      "learning_rate": 5.0987803690247635e-06,
+      "loss": 0.9703,
+      "step": 9617
+    },
+    {
+      "epoch": 0.8593830276766368,
+      "grad_norm": 0.499735027551651,
+      "learning_rate": 5.092415920357674e-06,
+      "loss": 0.9526,
+      "step": 9618
+    },
+    {
+      "epoch": 0.8594723792078987,
+      "grad_norm": 0.482085257768631,
+      "learning_rate": 5.086055233179421e-06,
+      "loss": 0.8669,
+      "step": 9619
+    },
+    {
+      "epoch": 0.8595617307391605,
+      "grad_norm": 0.448794960975647,
+      "learning_rate": 5.07969830802279e-06,
+      "loss": 0.9589,
+      "step": 9620
+    },
+    {
+      "epoch": 0.8596510822704224,
+      "grad_norm": 0.464096337556839,
+      "learning_rate": 5.073345145420238e-06,
+      "loss": 0.9648,
+      "step": 9621
+    },
+    {
+      "epoch": 0.8597404338016843,
+      "grad_norm": 0.5646736025810242,
+      "learning_rate": 5.066995745903919e-06,
+      "loss": 0.96,
+      "step": 9622
+    },
+    {
+      "epoch": 0.8598297853329462,
+      "grad_norm": 0.5770590901374817,
+      "learning_rate": 5.060650110005655e-06,
+      "loss": 0.9433,
+      "step": 9623
+    },
+    {
+      "epoch": 0.859919136864208,
+      "grad_norm": 0.5076577067375183,
+      "learning_rate": 5.0543082382569666e-06,
+      "loss": 0.8988,
+      "step": 9624
+    },
+    {
+      "epoch": 0.8600084883954698,
+      "grad_norm": 0.40074893832206726,
+      "learning_rate": 5.047970131189045e-06,
+      "loss": 0.9946,
+      "step": 9625
+    },
+    {
+      "epoch": 0.8600978399267317,
+      "grad_norm": 0.530560314655304,
+      "learning_rate": 5.0416357893327826e-06,
+      "loss": 1.0576,
+      "step": 9626
+    },
+    {
+      "epoch": 0.8601871914579936,
+      "grad_norm": 0.5239288806915283,
+      "learning_rate": 5.035305213218744e-06,
+      "loss": 0.9083,
+      "step": 9627
+    },
+    {
+      "epoch": 0.8602765429892555,
+      "grad_norm": 0.6042668223381042,
+      "learning_rate": 5.028978403377182e-06,
+      "loss": 0.8771,
+      "step": 9628
+    },
+    {
+      "epoch": 0.8603658945205174,
+      "grad_norm": 0.5551727414131165,
+      "learning_rate": 5.022655360338047e-06,
+      "loss": 0.9087,
+      "step": 9629
+    },
+    {
+      "epoch": 0.8604552460517793,
+      "grad_norm": 0.5318360328674316,
+      "learning_rate": 5.016336084630935e-06,
+      "loss": 0.8723,
+      "step": 9630
+    },
+    {
+      "epoch": 0.860544597583041,
+      "grad_norm": 0.42179903388023376,
+      "learning_rate": 5.010020576785174e-06,
+      "loss": 0.9642,
+      "step": 9631
+    },
+    {
+      "epoch": 0.8606339491143029,
+      "grad_norm": 0.5042697787284851,
+      "learning_rate": 5.0037088373297455e-06,
+      "loss": 0.8744,
+      "step": 9632
+    },
+    {
+      "epoch": 0.8607233006455648,
+      "grad_norm": 0.539291501045227,
+      "learning_rate": 4.997400866793328e-06,
+      "loss": 0.9179,
+      "step": 9633
+    },
+    {
+      "epoch": 0.8608126521768267,
+      "grad_norm": 0.5395140647888184,
+      "learning_rate": 4.991096665704281e-06,
+      "loss": 0.9229,
+      "step": 9634
+    },
+    {
+      "epoch": 0.8609020037080886,
+      "grad_norm": 0.46536731719970703,
+      "learning_rate": 4.984796234590644e-06,
+      "loss": 0.9679,
+      "step": 9635
+    },
+    {
+      "epoch": 0.8609913552393504,
+      "grad_norm": 0.4683972895145416,
+      "learning_rate": 4.978499573980155e-06,
+      "loss": 0.9381,
+      "step": 9636
+    },
+    {
+      "epoch": 0.8610807067706123,
+      "grad_norm": 0.49697425961494446,
+      "learning_rate": 4.972206684400221e-06,
+      "loss": 0.9342,
+      "step": 9637
+    },
+    {
+      "epoch": 0.8611700583018741,
+      "grad_norm": 0.5050958395004272,
+      "learning_rate": 4.9659175663779365e-06,
+      "loss": 0.9076,
+      "step": 9638
+    },
+    {
+      "epoch": 0.861259409833136,
+      "grad_norm": 0.5287109613418579,
+      "learning_rate": 4.959632220440097e-06,
+      "loss": 0.8461,
+      "step": 9639
+    },
+    {
+      "epoch": 0.8613487613643979,
+      "grad_norm": 0.7550068497657776,
+      "learning_rate": 4.953350647113148e-06,
+      "loss": 0.8207,
+      "step": 9640
+    },
+    {
+      "epoch": 0.8614381128956597,
+      "grad_norm": 0.4816301763057709,
+      "learning_rate": 4.94707284692324e-06,
+      "loss": 0.911,
+      "step": 9641
+    },
+    {
+      "epoch": 0.8615274644269216,
+      "grad_norm": 0.4800666570663452,
+      "learning_rate": 4.940798820396214e-06,
+      "loss": 0.8894,
+      "step": 9642
+    },
+    {
+      "epoch": 0.8616168159581835,
+      "grad_norm": 0.4487608075141907,
+      "learning_rate": 4.934528568057589e-06,
+      "loss": 0.9399,
+      "step": 9643
+    },
+    {
+      "epoch": 0.8617061674894454,
+      "grad_norm": 0.5139451026916504,
+      "learning_rate": 4.928262090432556e-06,
+      "loss": 0.9371,
+      "step": 9644
+    },
+    {
+      "epoch": 0.8617955190207072,
+      "grad_norm": 0.44007474184036255,
+      "learning_rate": 4.921999388045995e-06,
+      "loss": 0.9442,
+      "step": 9645
+    },
+    {
+      "epoch": 0.861884870551969,
+      "grad_norm": 0.4355262219905853,
+      "learning_rate": 4.915740461422491e-06,
+      "loss": 0.9202,
+      "step": 9646
+    },
+    {
+      "epoch": 0.8619742220832309,
+      "grad_norm": 0.425823450088501,
+      "learning_rate": 4.909485311086281e-06,
+      "loss": 1.0389,
+      "step": 9647
+    },
+    {
+      "epoch": 0.8620635736144928,
+      "grad_norm": 0.5076099038124084,
+      "learning_rate": 4.903233937561308e-06,
+      "loss": 0.9965,
+      "step": 9648
+    },
+    {
+      "epoch": 0.8621529251457547,
+      "grad_norm": 0.5130895972251892,
+      "learning_rate": 4.896986341371201e-06,
+      "loss": 0.8541,
+      "step": 9649
+    },
+    {
+      "epoch": 0.8622422766770166,
+      "grad_norm": 0.4927546977996826,
+      "learning_rate": 4.890742523039238e-06,
+      "loss": 0.8913,
+      "step": 9650
+    },
+    {
+      "epoch": 0.8623316282082785,
+      "grad_norm": 0.6410254240036011,
+      "learning_rate": 4.884502483088421e-06,
+      "loss": 0.8739,
+      "step": 9651
+    },
+    {
+      "epoch": 0.8624209797395402,
+      "grad_norm": 0.4251442849636078,
+      "learning_rate": 4.8782662220414225e-06,
+      "loss": 0.9444,
+      "step": 9652
+    },
+    {
+      "epoch": 0.8625103312708021,
+      "grad_norm": 0.46380364894866943,
+      "learning_rate": 4.872033740420584e-06,
+      "loss": 0.9366,
+      "step": 9653
+    },
+    {
+      "epoch": 0.862599682802064,
+      "grad_norm": 0.4519484043121338,
+      "learning_rate": 4.86580503874795e-06,
+      "loss": 0.8937,
+      "step": 9654
+    },
+    {
+      "epoch": 0.8626890343333259,
+      "grad_norm": 0.6642407178878784,
+      "learning_rate": 4.8595801175452425e-06,
+      "loss": 0.8715,
+      "step": 9655
+    },
+    {
+      "epoch": 0.8627783858645878,
+      "grad_norm": 0.6670805215835571,
+      "learning_rate": 4.853358977333866e-06,
+      "loss": 0.8974,
+      "step": 9656
+    },
+    {
+      "epoch": 0.8628677373958497,
+      "grad_norm": 0.4275446832180023,
+      "learning_rate": 4.847141618634898e-06,
+      "loss": 0.9749,
+      "step": 9657
+    },
+    {
+      "epoch": 0.8629570889271115,
+      "grad_norm": 0.4982154369354248,
+      "learning_rate": 4.8409280419691176e-06,
+      "loss": 0.9995,
+      "step": 9658
+    },
+    {
+      "epoch": 0.8630464404583733,
+      "grad_norm": 0.47910749912261963,
+      "learning_rate": 4.834718247856978e-06,
+      "loss": 0.944,
+      "step": 9659
+    },
+    {
+      "epoch": 0.8631357919896352,
+      "grad_norm": 0.633050799369812,
+      "learning_rate": 4.828512236818611e-06,
+      "loss": 0.863,
+      "step": 9660
+    },
+    {
+      "epoch": 0.8632251435208971,
+      "grad_norm": 0.6047154068946838,
+      "learning_rate": 4.822310009373832e-06,
+      "loss": 0.91,
+      "step": 9661
+    },
+    {
+      "epoch": 0.863314495052159,
+      "grad_norm": 0.49586907029151917,
+      "learning_rate": 4.816111566042153e-06,
+      "loss": 0.9588,
+      "step": 9662
+    },
+    {
+      "epoch": 0.8634038465834208,
+      "grad_norm": 0.5318160057067871,
+      "learning_rate": 4.809916907342754e-06,
+      "loss": 0.8139,
+      "step": 9663
+    },
+    {
+      "epoch": 0.8634931981146827,
+      "grad_norm": 0.5686520338058472,
+      "learning_rate": 4.803726033794509e-06,
+      "loss": 0.9418,
+      "step": 9664
+    },
+    {
+      "epoch": 0.8635825496459446,
+      "grad_norm": 0.5216206908226013,
+      "learning_rate": 4.797538945915958e-06,
+      "loss": 0.9086,
+      "step": 9665
+    },
+    {
+      "epoch": 0.8636719011772064,
+      "grad_norm": 0.4904700815677643,
+      "learning_rate": 4.791355644225354e-06,
+      "loss": 0.93,
+      "step": 9666
+    },
+    {
+      "epoch": 0.8637612527084683,
+      "grad_norm": 0.4535919725894928,
+      "learning_rate": 4.7851761292405985e-06,
+      "loss": 0.924,
+      "step": 9667
+    },
+    {
+      "epoch": 0.8638506042397301,
+      "grad_norm": 0.5180773138999939,
+      "learning_rate": 4.7790004014793e-06,
+      "loss": 0.9589,
+      "step": 9668
+    },
+    {
+      "epoch": 0.863939955770992,
+      "grad_norm": 0.4436119496822357,
+      "learning_rate": 4.772828461458745e-06,
+      "loss": 0.9582,
+      "step": 9669
+    },
+    {
+      "epoch": 0.8640293073022539,
+      "grad_norm": 0.41180846095085144,
+      "learning_rate": 4.766660309695886e-06,
+      "loss": 0.9333,
+      "step": 9670
+    },
+    {
+      "epoch": 0.8641186588335158,
+      "grad_norm": 0.4437258243560791,
+      "learning_rate": 4.7604959467073774e-06,
+      "loss": 0.9341,
+      "step": 9671
+    },
+    {
+      "epoch": 0.8642080103647777,
+      "grad_norm": 0.5144766569137573,
+      "learning_rate": 4.754335373009555e-06,
+      "loss": 0.9834,
+      "step": 9672
+    },
+    {
+      "epoch": 0.8642973618960395,
+      "grad_norm": 0.43276622891426086,
+      "learning_rate": 4.7481785891184295e-06,
+      "loss": 0.9718,
+      "step": 9673
+    },
+    {
+      "epoch": 0.8643867134273013,
+      "grad_norm": 0.5187088847160339,
+      "learning_rate": 4.7420255955496925e-06,
+      "loss": 0.9335,
+      "step": 9674
+    },
+    {
+      "epoch": 0.8644760649585632,
+      "grad_norm": 0.5164356827735901,
+      "learning_rate": 4.735876392818727e-06,
+      "loss": 0.9012,
+      "step": 9675
+    },
+    {
+      "epoch": 0.8645654164898251,
+      "grad_norm": 0.49884214997291565,
+      "learning_rate": 4.729730981440611e-06,
+      "loss": 0.9389,
+      "step": 9676
+    },
+    {
+      "epoch": 0.864654768021087,
+      "grad_norm": 0.5525695085525513,
+      "learning_rate": 4.723589361930053e-06,
+      "loss": 0.9312,
+      "step": 9677
+    },
+    {
+      "epoch": 0.8647441195523489,
+      "grad_norm": 0.4606095850467682,
+      "learning_rate": 4.717451534801504e-06,
+      "loss": 0.9609,
+      "step": 9678
+    },
+    {
+      "epoch": 0.8648334710836107,
+      "grad_norm": 0.513602077960968,
+      "learning_rate": 4.711317500569068e-06,
+      "loss": 0.9551,
+      "step": 9679
+    },
+    {
+      "epoch": 0.8649228226148725,
+      "grad_norm": 0.4917820990085602,
+      "learning_rate": 4.70518725974653e-06,
+      "loss": 0.878,
+      "step": 9680
+    },
+    {
+      "epoch": 0.8650121741461344,
+      "grad_norm": 0.49893245100975037,
+      "learning_rate": 4.699060812847378e-06,
+      "loss": 0.9257,
+      "step": 9681
+    },
+    {
+      "epoch": 0.8651015256773963,
+      "grad_norm": 0.5214920043945312,
+      "learning_rate": 4.692938160384747e-06,
+      "loss": 0.8962,
+      "step": 9682
+    },
+    {
+      "epoch": 0.8651908772086582,
+      "grad_norm": 0.5985978841781616,
+      "learning_rate": 4.686819302871481e-06,
+      "loss": 0.8617,
+      "step": 9683
+    },
+    {
+      "epoch": 0.86528022873992,
+      "grad_norm": 0.47699883580207825,
+      "learning_rate": 4.6807042408201e-06,
+      "loss": 0.9344,
+      "step": 9684
+    },
+    {
+      "epoch": 0.8653695802711819,
+      "grad_norm": 0.4804041087627411,
+      "learning_rate": 4.674592974742814e-06,
+      "loss": 0.9556,
+      "step": 9685
+    },
+    {
+      "epoch": 0.8654589318024437,
+      "grad_norm": 0.5328719019889832,
+      "learning_rate": 4.668485505151498e-06,
+      "loss": 0.9159,
+      "step": 9686
+    },
+    {
+      "epoch": 0.8655482833337056,
+      "grad_norm": 0.5171124935150146,
+      "learning_rate": 4.662381832557722e-06,
+      "loss": 0.9119,
+      "step": 9687
+    },
+    {
+      "epoch": 0.8656376348649675,
+      "grad_norm": 0.44296491146087646,
+      "learning_rate": 4.65628195747273e-06,
+      "loss": 0.9932,
+      "step": 9688
+    },
+    {
+      "epoch": 0.8657269863962294,
+      "grad_norm": 0.5319826602935791,
+      "learning_rate": 4.65018588040746e-06,
+      "loss": 0.9008,
+      "step": 9689
+    },
+    {
+      "epoch": 0.8658163379274912,
+      "grad_norm": 0.5790243744850159,
+      "learning_rate": 4.644093601872513e-06,
+      "loss": 0.8628,
+      "step": 9690
+    },
+    {
+      "epoch": 0.8659056894587531,
+      "grad_norm": 0.46617162227630615,
+      "learning_rate": 4.638005122378181e-06,
+      "loss": 0.9073,
+      "step": 9691
+    },
+    {
+      "epoch": 0.865995040990015,
+      "grad_norm": 0.4653133451938629,
+      "learning_rate": 4.631920442434446e-06,
+      "loss": 0.9665,
+      "step": 9692
+    },
+    {
+      "epoch": 0.8660843925212768,
+      "grad_norm": 0.4677208662033081,
+      "learning_rate": 4.625839562550965e-06,
+      "loss": 0.8954,
+      "step": 9693
+    },
+    {
+      "epoch": 0.8661737440525387,
+      "grad_norm": 0.5210303664207458,
+      "learning_rate": 4.619762483237072e-06,
+      "loss": 0.9758,
+      "step": 9694
+    },
+    {
+      "epoch": 0.8662630955838005,
+      "grad_norm": 0.5137758255004883,
+      "learning_rate": 4.613689205001792e-06,
+      "loss": 0.8698,
+      "step": 9695
+    },
+    {
+      "epoch": 0.8663524471150624,
+      "grad_norm": 0.45305004715919495,
+      "learning_rate": 4.607619728353818e-06,
+      "loss": 0.9656,
+      "step": 9696
+    },
+    {
+      "epoch": 0.8664417986463243,
+      "grad_norm": 0.5791176557540894,
+      "learning_rate": 4.601554053801549e-06,
+      "loss": 0.8502,
+      "step": 9697
+    },
+    {
+      "epoch": 0.8665311501775862,
+      "grad_norm": 0.40173932909965515,
+      "learning_rate": 4.595492181853034e-06,
+      "loss": 0.9574,
+      "step": 9698
+    },
+    {
+      "epoch": 0.8666205017088481,
+      "grad_norm": 0.4410465955734253,
+      "learning_rate": 4.589434113016039e-06,
+      "loss": 0.9465,
+      "step": 9699
+    },
+    {
+      "epoch": 0.8667098532401099,
+      "grad_norm": 0.49588507413864136,
+      "learning_rate": 4.583379847797964e-06,
+      "loss": 0.9517,
+      "step": 9700
+    },
+    {
+      "epoch": 0.8667992047713717,
+      "grad_norm": 0.46362167596817017,
+      "learning_rate": 4.577329386705942e-06,
+      "loss": 0.9281,
+      "step": 9701
+    },
+    {
+      "epoch": 0.8668885563026336,
+      "grad_norm": 0.5212817192077637,
+      "learning_rate": 4.571282730246745e-06,
+      "loss": 0.9243,
+      "step": 9702
+    },
+    {
+      "epoch": 0.8669779078338955,
+      "grad_norm": 0.45512816309928894,
+      "learning_rate": 4.565239878926863e-06,
+      "loss": 0.9509,
+      "step": 9703
+    },
+    {
+      "epoch": 0.8670672593651574,
+      "grad_norm": 0.5159767866134644,
+      "learning_rate": 4.5592008332524364e-06,
+      "loss": 0.8746,
+      "step": 9704
+    },
+    {
+      "epoch": 0.8671566108964193,
+      "grad_norm": 0.4418531656265259,
+      "learning_rate": 4.553165593729303e-06,
+      "loss": 1.0525,
+      "step": 9705
+    },
+    {
+      "epoch": 0.8672459624276811,
+      "grad_norm": 0.5434200763702393,
+      "learning_rate": 4.547134160862981e-06,
+      "loss": 0.8959,
+      "step": 9706
+    },
+    {
+      "epoch": 0.8673353139589429,
+      "grad_norm": 0.4653039276599884,
+      "learning_rate": 4.541106535158668e-06,
+      "loss": 0.9446,
+      "step": 9707
+    },
+    {
+      "epoch": 0.8674246654902048,
+      "grad_norm": 0.46755141019821167,
+      "learning_rate": 4.5350827171212366e-06,
+      "loss": 0.8746,
+      "step": 9708
+    },
+    {
+      "epoch": 0.8675140170214667,
+      "grad_norm": 0.5309289693832397,
+      "learning_rate": 4.529062707255261e-06,
+      "loss": 0.9246,
+      "step": 9709
+    },
+    {
+      "epoch": 0.8676033685527286,
+      "grad_norm": 0.5540649890899658,
+      "learning_rate": 4.5230465060649595e-06,
+      "loss": 0.8565,
+      "step": 9710
+    },
+    {
+      "epoch": 0.8676927200839905,
+      "grad_norm": 0.5240215063095093,
+      "learning_rate": 4.517034114054258e-06,
+      "loss": 0.9513,
+      "step": 9711
+    },
+    {
+      "epoch": 0.8677820716152523,
+      "grad_norm": 0.4945807158946991,
+      "learning_rate": 4.5110255317267704e-06,
+      "loss": 1.0085,
+      "step": 9712
+    },
+    {
+      "epoch": 0.8678714231465142,
+      "grad_norm": 0.446004182100296,
+      "learning_rate": 4.505020759585765e-06,
+      "loss": 0.9658,
+      "step": 9713
+    },
+    {
+      "epoch": 0.867960774677776,
+      "grad_norm": 0.4994924068450928,
+      "learning_rate": 4.499019798134224e-06,
+      "loss": 0.9725,
+      "step": 9714
+    },
+    {
+      "epoch": 0.8680501262090379,
+      "grad_norm": 0.42142254114151,
+      "learning_rate": 4.493022647874773e-06,
+      "loss": 0.9469,
+      "step": 9715
+    },
+    {
+      "epoch": 0.8681394777402998,
+      "grad_norm": 0.49049112200737,
+      "learning_rate": 4.487029309309748e-06,
+      "loss": 0.9365,
+      "step": 9716
+    },
+    {
+      "epoch": 0.8682288292715616,
+      "grad_norm": 0.5305113792419434,
+      "learning_rate": 4.481039782941143e-06,
+      "loss": 0.9549,
+      "step": 9717
+    },
+    {
+      "epoch": 0.8683181808028235,
+      "grad_norm": 0.5698569416999817,
+      "learning_rate": 4.4750540692706625e-06,
+      "loss": 0.9227,
+      "step": 9718
+    },
+    {
+      "epoch": 0.8684075323340854,
+      "grad_norm": 0.5504754185676575,
+      "learning_rate": 4.469072168799659e-06,
+      "loss": 1.002,
+      "step": 9719
+    },
+    {
+      "epoch": 0.8684968838653473,
+      "grad_norm": 0.4171115756034851,
+      "learning_rate": 4.4630940820291955e-06,
+      "loss": 0.9375,
+      "step": 9720
+    },
+    {
+      "epoch": 0.8685862353966091,
+      "grad_norm": 0.5085521340370178,
+      "learning_rate": 4.4571198094599866e-06,
+      "loss": 0.8873,
+      "step": 9721
+    },
+    {
+      "epoch": 0.868675586927871,
+      "grad_norm": 0.3779529929161072,
+      "learning_rate": 4.451149351592437e-06,
+      "loss": 1.0097,
+      "step": 9722
+    },
+    {
+      "epoch": 0.8687649384591328,
+      "grad_norm": 0.4312131702899933,
+      "learning_rate": 4.445182708926654e-06,
+      "loss": 0.9988,
+      "step": 9723
+    },
+    {
+      "epoch": 0.8688542899903947,
+      "grad_norm": 0.578957200050354,
+      "learning_rate": 4.439219881962392e-06,
+      "loss": 0.8544,
+      "step": 9724
+    },
+    {
+      "epoch": 0.8689436415216566,
+      "grad_norm": 0.47734835743904114,
+      "learning_rate": 4.433260871199113e-06,
+      "loss": 0.9883,
+      "step": 9725
+    },
+    {
+      "epoch": 0.8690329930529185,
+      "grad_norm": 0.5002614259719849,
+      "learning_rate": 4.42730567713594e-06,
+      "loss": 0.9303,
+      "step": 9726
+    },
+    {
+      "epoch": 0.8691223445841804,
+      "grad_norm": 0.4868052303791046,
+      "learning_rate": 4.4213543002716904e-06,
+      "loss": 0.9285,
+      "step": 9727
+    },
+    {
+      "epoch": 0.8692116961154421,
+      "grad_norm": 0.4923020005226135,
+      "learning_rate": 4.415406741104844e-06,
+      "loss": 0.971,
+      "step": 9728
+    },
+    {
+      "epoch": 0.869301047646704,
+      "grad_norm": 0.4633644223213196,
+      "learning_rate": 4.409463000133584e-06,
+      "loss": 0.9052,
+      "step": 9729
+    },
+    {
+      "epoch": 0.8693903991779659,
+      "grad_norm": 0.5671141147613525,
+      "learning_rate": 4.40352307785577e-06,
+      "loss": 0.9258,
+      "step": 9730
+    },
+    {
+      "epoch": 0.8694797507092278,
+      "grad_norm": 0.40942826867103577,
+      "learning_rate": 4.397586974768908e-06,
+      "loss": 0.9291,
+      "step": 9731
+    },
+    {
+      "epoch": 0.8695691022404897,
+      "grad_norm": 0.38688722252845764,
+      "learning_rate": 4.391654691370229e-06,
+      "loss": 0.9367,
+      "step": 9732
+    },
+    {
+      "epoch": 0.8696584537717515,
+      "grad_norm": 0.43114298582077026,
+      "learning_rate": 4.385726228156617e-06,
+      "loss": 0.9637,
+      "step": 9733
+    },
+    {
+      "epoch": 0.8697478053030134,
+      "grad_norm": 0.4530818462371826,
+      "learning_rate": 4.379801585624643e-06,
+      "loss": 0.9887,
+      "step": 9734
+    },
+    {
+      "epoch": 0.8698371568342752,
+      "grad_norm": 0.582784116268158,
+      "learning_rate": 4.373880764270566e-06,
+      "loss": 0.9341,
+      "step": 9735
+    },
+    {
+      "epoch": 0.8699265083655371,
+      "grad_norm": 0.5109930634498596,
+      "learning_rate": 4.367963764590311e-06,
+      "loss": 0.9377,
+      "step": 9736
+    },
+    {
+      "epoch": 0.870015859896799,
+      "grad_norm": 0.5649825930595398,
+      "learning_rate": 4.362050587079497e-06,
+      "loss": 0.9036,
+      "step": 9737
+    },
+    {
+      "epoch": 0.8701052114280609,
+      "grad_norm": 0.6133013963699341,
+      "learning_rate": 4.356141232233413e-06,
+      "loss": 0.8088,
+      "step": 9738
+    },
+    {
+      "epoch": 0.8701945629593227,
+      "grad_norm": 0.6256003379821777,
+      "learning_rate": 4.350235700547028e-06,
+      "loss": 0.8784,
+      "step": 9739
+    },
+    {
+      "epoch": 0.8702839144905846,
+      "grad_norm": 0.5290203094482422,
+      "learning_rate": 4.344333992515004e-06,
+      "loss": 0.8624,
+      "step": 9740
+    },
+    {
+      "epoch": 0.8703732660218465,
+      "grad_norm": 0.44871601462364197,
+      "learning_rate": 4.338436108631649e-06,
+      "loss": 0.9352,
+      "step": 9741
+    },
+    {
+      "epoch": 0.8704626175531083,
+      "grad_norm": 0.6028452515602112,
+      "learning_rate": 4.332542049390992e-06,
+      "loss": 0.8648,
+      "step": 9742
+    },
+    {
+      "epoch": 0.8705519690843702,
+      "grad_norm": 0.4009053409099579,
+      "learning_rate": 4.3266518152867145e-06,
+      "loss": 0.9586,
+      "step": 9743
+    },
+    {
+      "epoch": 0.870641320615632,
+      "grad_norm": 0.5013288259506226,
+      "learning_rate": 4.320765406812194e-06,
+      "loss": 0.8975,
+      "step": 9744
+    },
+    {
+      "epoch": 0.8707306721468939,
+      "grad_norm": 0.5538919568061829,
+      "learning_rate": 4.314882824460475e-06,
+      "loss": 1.09,
+      "step": 9745
+    },
+    {
+      "epoch": 0.8708200236781558,
+      "grad_norm": 0.4379560947418213,
+      "learning_rate": 4.309004068724298e-06,
+      "loss": 0.9737,
+      "step": 9746
+    },
+    {
+      "epoch": 0.8709093752094177,
+      "grad_norm": 0.47535446286201477,
+      "learning_rate": 4.303129140096052e-06,
+      "loss": 0.9569,
+      "step": 9747
+    },
+    {
+      "epoch": 0.8709987267406795,
+      "grad_norm": 0.5819507837295532,
+      "learning_rate": 4.297258039067831e-06,
+      "loss": 0.8848,
+      "step": 9748
+    },
+    {
+      "epoch": 0.8710880782719413,
+      "grad_norm": 0.4755156934261322,
+      "learning_rate": 4.291390766131409e-06,
+      "loss": 0.8717,
+      "step": 9749
+    },
+    {
+      "epoch": 0.8711774298032032,
+      "grad_norm": 0.605339527130127,
+      "learning_rate": 4.285527321778232e-06,
+      "loss": 0.9916,
+      "step": 9750
+    },
+    {
+      "epoch": 0.8712667813344651,
+      "grad_norm": 0.6067926287651062,
+      "learning_rate": 4.279667706499424e-06,
+      "loss": 0.9132,
+      "step": 9751
+    },
+    {
+      "epoch": 0.871356132865727,
+      "grad_norm": 0.5548110604286194,
+      "learning_rate": 4.273811920785786e-06,
+      "loss": 0.8592,
+      "step": 9752
+    },
+    {
+      "epoch": 0.8714454843969889,
+      "grad_norm": 0.533915638923645,
+      "learning_rate": 4.2679599651278045e-06,
+      "loss": 0.9079,
+      "step": 9753
+    },
+    {
+      "epoch": 0.8715348359282508,
+      "grad_norm": 0.5274013876914978,
+      "learning_rate": 4.262111840015642e-06,
+      "loss": 0.9541,
+      "step": 9754
+    },
+    {
+      "epoch": 0.8716241874595125,
+      "grad_norm": 0.4624837636947632,
+      "learning_rate": 4.256267545939147e-06,
+      "loss": 0.8876,
+      "step": 9755
+    },
+    {
+      "epoch": 0.8717135389907744,
+      "grad_norm": 0.5232068300247192,
+      "learning_rate": 4.250427083387837e-06,
+      "loss": 0.9834,
+      "step": 9756
+    },
+    {
+      "epoch": 0.8718028905220363,
+      "grad_norm": 0.666278600692749,
+      "learning_rate": 4.244590452850916e-06,
+      "loss": 0.8801,
+      "step": 9757
+    },
+    {
+      "epoch": 0.8718922420532982,
+      "grad_norm": 0.49066150188446045,
+      "learning_rate": 4.2387576548172605e-06,
+      "loss": 0.907,
+      "step": 9758
+    },
+    {
+      "epoch": 0.8719815935845601,
+      "grad_norm": 0.599941074848175,
+      "learning_rate": 4.232928689775428e-06,
+      "loss": 0.9306,
+      "step": 9759
+    },
+    {
+      "epoch": 0.872070945115822,
+      "grad_norm": 0.5583867430686951,
+      "learning_rate": 4.227103558213674e-06,
+      "loss": 0.8552,
+      "step": 9760
+    },
+    {
+      "epoch": 0.8721602966470838,
+      "grad_norm": 0.5444775819778442,
+      "learning_rate": 4.221282260619891e-06,
+      "loss": 0.9385,
+      "step": 9761
+    },
+    {
+      "epoch": 0.8722496481783456,
+      "grad_norm": 0.5456721782684326,
+      "learning_rate": 4.215464797481683e-06,
+      "loss": 0.8826,
+      "step": 9762
+    },
+    {
+      "epoch": 0.8723389997096075,
+      "grad_norm": 0.5117126107215881,
+      "learning_rate": 4.2096511692863275e-06,
+      "loss": 0.9481,
+      "step": 9763
+    },
+    {
+      "epoch": 0.8724283512408694,
+      "grad_norm": 0.5069064497947693,
+      "learning_rate": 4.203841376520773e-06,
+      "loss": 0.9362,
+      "step": 9764
+    },
+    {
+      "epoch": 0.8725177027721313,
+      "grad_norm": 0.5353370904922485,
+      "learning_rate": 4.198035419671658e-06,
+      "loss": 0.8717,
+      "step": 9765
+    },
+    {
+      "epoch": 0.8726070543033931,
+      "grad_norm": 0.4016374945640564,
+      "learning_rate": 4.19223329922529e-06,
+      "loss": 0.8964,
+      "step": 9766
+    },
+    {
+      "epoch": 0.872696405834655,
+      "grad_norm": 0.4020116627216339,
+      "learning_rate": 4.186435015667661e-06,
+      "loss": 1.0015,
+      "step": 9767
+    },
+    {
+      "epoch": 0.8727857573659169,
+      "grad_norm": 0.6577222347259521,
+      "learning_rate": 4.180640569484434e-06,
+      "loss": 0.9153,
+      "step": 9768
+    },
+    {
+      "epoch": 0.8728751088971787,
+      "grad_norm": 0.5189597010612488,
+      "learning_rate": 4.17484996116096e-06,
+      "loss": 0.903,
+      "step": 9769
+    },
+    {
+      "epoch": 0.8729644604284406,
+      "grad_norm": 0.44712257385253906,
+      "learning_rate": 4.169063191182271e-06,
+      "loss": 0.9051,
+      "step": 9770
+    },
+    {
+      "epoch": 0.8730538119597024,
+      "grad_norm": 0.497765451669693,
+      "learning_rate": 4.163280260033053e-06,
+      "loss": 0.9542,
+      "step": 9771
+    },
+    {
+      "epoch": 0.8731431634909643,
+      "grad_norm": 0.4439071714878082,
+      "learning_rate": 4.157501168197703e-06,
+      "loss": 0.9827,
+      "step": 9772
+    },
+    {
+      "epoch": 0.8732325150222262,
+      "grad_norm": 0.4466901421546936,
+      "learning_rate": 4.1517259161602705e-06,
+      "loss": 0.939,
+      "step": 9773
+    },
+    {
+      "epoch": 0.8733218665534881,
+      "grad_norm": 0.5371690988540649,
+      "learning_rate": 4.145954504404498e-06,
+      "loss": 0.9615,
+      "step": 9774
+    },
+    {
+      "epoch": 0.87341121808475,
+      "grad_norm": 0.4514279067516327,
+      "learning_rate": 4.140186933413809e-06,
+      "loss": 0.9192,
+      "step": 9775
+    },
+    {
+      "epoch": 0.8735005696160117,
+      "grad_norm": 0.47442543506622314,
+      "learning_rate": 4.134423203671295e-06,
+      "loss": 0.9659,
+      "step": 9776
+    },
+    {
+      "epoch": 0.8735899211472736,
+      "grad_norm": 0.532922625541687,
+      "learning_rate": 4.128663315659725e-06,
+      "loss": 0.9492,
+      "step": 9777
+    },
+    {
+      "epoch": 0.8736792726785355,
+      "grad_norm": 0.6389763355255127,
+      "learning_rate": 4.122907269861559e-06,
+      "loss": 0.8919,
+      "step": 9778
+    },
+    {
+      "epoch": 0.8737686242097974,
+      "grad_norm": 0.4954306483268738,
+      "learning_rate": 4.117155066758938e-06,
+      "loss": 1.0198,
+      "step": 9779
+    },
+    {
+      "epoch": 0.8738579757410593,
+      "grad_norm": 0.46348780393600464,
+      "learning_rate": 4.111406706833637e-06,
+      "loss": 0.9218,
+      "step": 9780
+    },
+    {
+      "epoch": 0.8739473272723212,
+      "grad_norm": 0.47658228874206543,
+      "learning_rate": 4.105662190567166e-06,
+      "loss": 0.8703,
+      "step": 9781
+    },
+    {
+      "epoch": 0.874036678803583,
+      "grad_norm": 0.5489630103111267,
+      "learning_rate": 4.099921518440686e-06,
+      "loss": 0.9327,
+      "step": 9782
+    },
+    {
+      "epoch": 0.8741260303348448,
+      "grad_norm": 0.5429866909980774,
+      "learning_rate": 4.094184690935038e-06,
+      "loss": 0.8676,
+      "step": 9783
+    },
+    {
+      "epoch": 0.8742153818661067,
+      "grad_norm": 0.5859069228172302,
+      "learning_rate": 4.088451708530755e-06,
+      "loss": 0.8897,
+      "step": 9784
+    },
+    {
+      "epoch": 0.8743047333973686,
+      "grad_norm": 0.5639720559120178,
+      "learning_rate": 4.082722571708008e-06,
+      "loss": 0.9152,
+      "step": 9785
+    },
+    {
+      "epoch": 0.8743940849286305,
+      "grad_norm": 0.4945251941680908,
+      "learning_rate": 4.076997280946693e-06,
+      "loss": 0.9027,
+      "step": 9786
+    },
+    {
+      "epoch": 0.8744834364598923,
+      "grad_norm": 0.49459564685821533,
+      "learning_rate": 4.071275836726357e-06,
+      "loss": 0.8952,
+      "step": 9787
+    },
+    {
+      "epoch": 0.8745727879911542,
+      "grad_norm": 0.5829160213470459,
+      "learning_rate": 4.065558239526241e-06,
+      "loss": 0.9146,
+      "step": 9788
+    },
+    {
+      "epoch": 0.8746621395224161,
+      "grad_norm": 0.472232848405838,
+      "learning_rate": 4.059844489825243e-06,
+      "loss": 0.9677,
+      "step": 9789
+    },
+    {
+      "epoch": 0.8747514910536779,
+      "grad_norm": 0.4521077275276184,
+      "learning_rate": 4.054134588101965e-06,
+      "loss": 0.9456,
+      "step": 9790
+    },
+    {
+      "epoch": 0.8748408425849398,
+      "grad_norm": 0.5067082047462463,
+      "learning_rate": 4.048428534834653e-06,
+      "loss": 0.9167,
+      "step": 9791
+    },
+    {
+      "epoch": 0.8749301941162017,
+      "grad_norm": 0.6378511786460876,
+      "learning_rate": 4.042726330501262e-06,
+      "loss": 0.8104,
+      "step": 9792
+    },
+    {
+      "epoch": 0.8750195456474635,
+      "grad_norm": 0.483568400144577,
+      "learning_rate": 4.03702797557941e-06,
+      "loss": 0.9183,
+      "step": 9793
+    },
+    {
+      "epoch": 0.8751088971787254,
+      "grad_norm": 0.47959351539611816,
+      "learning_rate": 4.031333470546394e-06,
+      "loss": 0.9048,
+      "step": 9794
+    },
+    {
+      "epoch": 0.8751982487099873,
+      "grad_norm": 0.5103621482849121,
+      "learning_rate": 4.025642815879188e-06,
+      "loss": 0.8854,
+      "step": 9795
+    },
+    {
+      "epoch": 0.8752876002412492,
+      "grad_norm": 0.4524891674518585,
+      "learning_rate": 4.019956012054455e-06,
+      "loss": 1.0298,
+      "step": 9796
+    },
+    {
+      "epoch": 0.875376951772511,
+      "grad_norm": 0.4642522633075714,
+      "learning_rate": 4.014273059548512e-06,
+      "loss": 0.9861,
+      "step": 9797
+    },
+    {
+      "epoch": 0.8754663033037728,
+      "grad_norm": 0.456893652677536,
+      "learning_rate": 4.0085939588373754e-06,
+      "loss": 0.9152,
+      "step": 9798
+    },
+    {
+      "epoch": 0.8755556548350347,
+      "grad_norm": 0.491441547870636,
+      "learning_rate": 4.0029187103967245e-06,
+      "loss": 0.9677,
+      "step": 9799
+    },
+    {
+      "epoch": 0.8756450063662966,
+      "grad_norm": 0.4507903754711151,
+      "learning_rate": 3.997247314701935e-06,
+      "loss": 0.9474,
+      "step": 9800
+    },
+    {
+      "epoch": 0.8757343578975585,
+      "grad_norm": 0.6348841190338135,
+      "learning_rate": 3.991579772228032e-06,
+      "loss": 0.9522,
+      "step": 9801
+    },
+    {
+      "epoch": 0.8758237094288204,
+      "grad_norm": 0.40709996223449707,
+      "learning_rate": 3.985916083449737e-06,
+      "loss": 0.9453,
+      "step": 9802
+    },
+    {
+      "epoch": 0.8759130609600823,
+      "grad_norm": 0.46370193362236023,
+      "learning_rate": 3.980256248841441e-06,
+      "loss": 0.9726,
+      "step": 9803
+    },
+    {
+      "epoch": 0.876002412491344,
+      "grad_norm": 0.3987545073032379,
+      "learning_rate": 3.974600268877221e-06,
+      "loss": 0.9591,
+      "step": 9804
+    },
+    {
+      "epoch": 0.8760917640226059,
+      "grad_norm": 0.5190705060958862,
+      "learning_rate": 3.9689481440308265e-06,
+      "loss": 0.9205,
+      "step": 9805
+    },
+    {
+      "epoch": 0.8761811155538678,
+      "grad_norm": 0.4716487228870392,
+      "learning_rate": 3.963299874775678e-06,
+      "loss": 0.9981,
+      "step": 9806
+    },
+    {
+      "epoch": 0.8762704670851297,
+      "grad_norm": 0.4962429702281952,
+      "learning_rate": 3.957655461584881e-06,
+      "loss": 0.897,
+      "step": 9807
+    },
+    {
+      "epoch": 0.8763598186163916,
+      "grad_norm": 0.6615248322486877,
+      "learning_rate": 3.952014904931217e-06,
+      "loss": 0.8475,
+      "step": 9808
+    },
+    {
+      "epoch": 0.8764491701476534,
+      "grad_norm": 0.47163599729537964,
+      "learning_rate": 3.946378205287138e-06,
+      "loss": 0.9074,
+      "step": 9809
+    },
+    {
+      "epoch": 0.8765385216789153,
+      "grad_norm": 0.6437599658966064,
+      "learning_rate": 3.940745363124787e-06,
+      "loss": 0.8462,
+      "step": 9810
+    },
+    {
+      "epoch": 0.8766278732101771,
+      "grad_norm": 0.5201234221458435,
+      "learning_rate": 3.935116378915971e-06,
+      "loss": 0.9933,
+      "step": 9811
+    },
+    {
+      "epoch": 0.876717224741439,
+      "grad_norm": 0.47189584374427795,
+      "learning_rate": 3.929491253132167e-06,
+      "loss": 0.9501,
+      "step": 9812
+    },
+    {
+      "epoch": 0.8768065762727009,
+      "grad_norm": 0.41938087344169617,
+      "learning_rate": 3.923869986244549e-06,
+      "loss": 0.9382,
+      "step": 9813
+    },
+    {
+      "epoch": 0.8768959278039627,
+      "grad_norm": 0.42869406938552856,
+      "learning_rate": 3.918252578723952e-06,
+      "loss": 0.9126,
+      "step": 9814
+    },
+    {
+      "epoch": 0.8769852793352246,
+      "grad_norm": 0.4569510221481323,
+      "learning_rate": 3.912639031040899e-06,
+      "loss": 0.9354,
+      "step": 9815
+    },
+    {
+      "epoch": 0.8770746308664865,
+      "grad_norm": 0.46077507734298706,
+      "learning_rate": 3.907029343665586e-06,
+      "loss": 0.9599,
+      "step": 9816
+    },
+    {
+      "epoch": 0.8771639823977483,
+      "grad_norm": 0.5191541314125061,
+      "learning_rate": 3.901423517067887e-06,
+      "loss": 0.8914,
+      "step": 9817
+    },
+    {
+      "epoch": 0.8772533339290102,
+      "grad_norm": 0.48782411217689514,
+      "learning_rate": 3.895821551717338e-06,
+      "loss": 0.8774,
+      "step": 9818
+    },
+    {
+      "epoch": 0.877342685460272,
+      "grad_norm": 0.42482325434684753,
+      "learning_rate": 3.890223448083163e-06,
+      "loss": 0.9598,
+      "step": 9819
+    },
+    {
+      "epoch": 0.8774320369915339,
+      "grad_norm": 0.3945348858833313,
+      "learning_rate": 3.884629206634277e-06,
+      "loss": 0.9229,
+      "step": 9820
+    },
+    {
+      "epoch": 0.8775213885227958,
+      "grad_norm": 0.41309240460395813,
+      "learning_rate": 3.879038827839255e-06,
+      "loss": 0.975,
+      "step": 9821
+    },
+    {
+      "epoch": 0.8776107400540577,
+      "grad_norm": 0.5235944390296936,
+      "learning_rate": 3.873452312166337e-06,
+      "loss": 0.8413,
+      "step": 9822
+    },
+    {
+      "epoch": 0.8777000915853196,
+      "grad_norm": 0.4890744388103485,
+      "learning_rate": 3.867869660083456e-06,
+      "loss": 0.8615,
+      "step": 9823
+    },
+    {
+      "epoch": 0.8777894431165814,
+      "grad_norm": 0.46650564670562744,
+      "learning_rate": 3.862290872058233e-06,
+      "loss": 0.9691,
+      "step": 9824
+    },
+    {
+      "epoch": 0.8778787946478432,
+      "grad_norm": 0.46951133012771606,
+      "learning_rate": 3.856715948557938e-06,
+      "loss": 0.9661,
+      "step": 9825
+    },
+    {
+      "epoch": 0.8779681461791051,
+      "grad_norm": 0.5398757457733154,
+      "learning_rate": 3.851144890049535e-06,
+      "loss": 0.9167,
+      "step": 9826
+    },
+    {
+      "epoch": 0.878057497710367,
+      "grad_norm": 0.49195584654808044,
+      "learning_rate": 3.845577696999659e-06,
+      "loss": 0.9914,
+      "step": 9827
+    },
+    {
+      "epoch": 0.8781468492416289,
+      "grad_norm": 0.5577439665794373,
+      "learning_rate": 3.840014369874617e-06,
+      "loss": 0.856,
+      "step": 9828
+    },
+    {
+      "epoch": 0.8782362007728908,
+      "grad_norm": 0.5505506992340088,
+      "learning_rate": 3.834454909140406e-06,
+      "loss": 0.8946,
+      "step": 9829
+    },
+    {
+      "epoch": 0.8783255523041527,
+      "grad_norm": 0.4931388795375824,
+      "learning_rate": 3.828899315262685e-06,
+      "loss": 0.9288,
+      "step": 9830
+    },
+    {
+      "epoch": 0.8784149038354144,
+      "grad_norm": 0.6808313131332397,
+      "learning_rate": 3.823347588706805e-06,
+      "loss": 0.9366,
+      "step": 9831
+    },
+    {
+      "epoch": 0.8785042553666763,
+      "grad_norm": 0.6877908706665039,
+      "learning_rate": 3.81779972993776e-06,
+      "loss": 0.8903,
+      "step": 9832
+    },
+    {
+      "epoch": 0.8785936068979382,
+      "grad_norm": 0.5493754744529724,
+      "learning_rate": 3.812255739420256e-06,
+      "loss": 0.9696,
+      "step": 9833
+    },
+    {
+      "epoch": 0.8786829584292001,
+      "grad_norm": 0.48021358251571655,
+      "learning_rate": 3.8067156176186616e-06,
+      "loss": 0.9312,
+      "step": 9834
+    },
+    {
+      "epoch": 0.878772309960462,
+      "grad_norm": 0.46737149357795715,
+      "learning_rate": 3.8011793649970207e-06,
+      "loss": 0.9433,
+      "step": 9835
+    },
+    {
+      "epoch": 0.8788616614917238,
+      "grad_norm": 0.4425142705440521,
+      "learning_rate": 3.7956469820190465e-06,
+      "loss": 0.9032,
+      "step": 9836
+    },
+    {
+      "epoch": 0.8789510130229857,
+      "grad_norm": 0.49997785687446594,
+      "learning_rate": 3.790118469148146e-06,
+      "loss": 0.8658,
+      "step": 9837
+    },
+    {
+      "epoch": 0.8790403645542475,
+      "grad_norm": 0.47397980093955994,
+      "learning_rate": 3.7845938268473823e-06,
+      "loss": 0.9049,
+      "step": 9838
+    },
+    {
+      "epoch": 0.8791297160855094,
+      "grad_norm": 0.5154423713684082,
+      "learning_rate": 3.7790730555795075e-06,
+      "loss": 0.956,
+      "step": 9839
+    },
+    {
+      "epoch": 0.8792190676167713,
+      "grad_norm": 0.5313198566436768,
+      "learning_rate": 3.7735561558069455e-06,
+      "loss": 0.915,
+      "step": 9840
+    },
+    {
+      "epoch": 0.8793084191480331,
+      "grad_norm": 0.499161034822464,
+      "learning_rate": 3.7680431279917994e-06,
+      "loss": 0.8453,
+      "step": 9841
+    },
+    {
+      "epoch": 0.879397770679295,
+      "grad_norm": 0.4321475028991699,
+      "learning_rate": 3.762533972595833e-06,
+      "loss": 0.9776,
+      "step": 9842
+    },
+    {
+      "epoch": 0.8794871222105569,
+      "grad_norm": 0.52239590883255,
+      "learning_rate": 3.7570286900804998e-06,
+      "loss": 0.9106,
+      "step": 9843
+    },
+    {
+      "epoch": 0.8795764737418188,
+      "grad_norm": 0.5335208177566528,
+      "learning_rate": 3.7515272809069303e-06,
+      "loss": 0.9377,
+      "step": 9844
+    },
+    {
+      "epoch": 0.8796658252730806,
+      "grad_norm": 0.5441358089447021,
+      "learning_rate": 3.746029745535923e-06,
+      "loss": 0.8779,
+      "step": 9845
+    },
+    {
+      "epoch": 0.8797551768043425,
+      "grad_norm": 0.5699974894523621,
+      "learning_rate": 3.7405360844279537e-06,
+      "loss": 0.8933,
+      "step": 9846
+    },
+    {
+      "epoch": 0.8798445283356043,
+      "grad_norm": 0.42555707693099976,
+      "learning_rate": 3.735046298043182e-06,
+      "loss": 0.9373,
+      "step": 9847
+    },
+    {
+      "epoch": 0.8799338798668662,
+      "grad_norm": 0.5089072585105896,
+      "learning_rate": 3.7295603868414297e-06,
+      "loss": 0.863,
+      "step": 9848
+    },
+    {
+      "epoch": 0.8800232313981281,
+      "grad_norm": 0.45135679841041565,
+      "learning_rate": 3.724078351282212e-06,
+      "loss": 0.9436,
+      "step": 9849
+    },
+    {
+      "epoch": 0.88011258292939,
+      "grad_norm": 0.5915647745132446,
+      "learning_rate": 3.7186001918246893e-06,
+      "loss": 0.8365,
+      "step": 9850
+    },
+    {
+      "epoch": 0.8802019344606519,
+      "grad_norm": 0.49184906482696533,
+      "learning_rate": 3.713125908927728e-06,
+      "loss": 0.8851,
+      "step": 9851
+    },
+    {
+      "epoch": 0.8802912859919136,
+      "grad_norm": 0.4521387815475464,
+      "learning_rate": 3.7076555030498506e-06,
+      "loss": 0.8955,
+      "step": 9852
+    },
+    {
+      "epoch": 0.8803806375231755,
+      "grad_norm": 0.43427813053131104,
+      "learning_rate": 3.7021889746492676e-06,
+      "loss": 0.9594,
+      "step": 9853
+    },
+    {
+      "epoch": 0.8804699890544374,
+      "grad_norm": 0.4693211019039154,
+      "learning_rate": 3.6967263241838636e-06,
+      "loss": 0.8923,
+      "step": 9854
+    },
+    {
+      "epoch": 0.8805593405856993,
+      "grad_norm": 0.4870004653930664,
+      "learning_rate": 3.691267552111183e-06,
+      "loss": 0.9082,
+      "step": 9855
+    },
+    {
+      "epoch": 0.8806486921169612,
+      "grad_norm": 0.4808718264102936,
+      "learning_rate": 3.6858126588884544e-06,
+      "loss": 0.9936,
+      "step": 9856
+    },
+    {
+      "epoch": 0.8807380436482231,
+      "grad_norm": 0.602287232875824,
+      "learning_rate": 3.6803616449725964e-06,
+      "loss": 0.8709,
+      "step": 9857
+    },
+    {
+      "epoch": 0.8808273951794849,
+      "grad_norm": 0.549919843673706,
+      "learning_rate": 3.6749145108201766e-06,
+      "loss": 0.9254,
+      "step": 9858
+    },
+    {
+      "epoch": 0.8809167467107467,
+      "grad_norm": 0.6031754612922668,
+      "learning_rate": 3.6694712568874577e-06,
+      "loss": 0.8894,
+      "step": 9859
+    },
+    {
+      "epoch": 0.8810060982420086,
+      "grad_norm": 0.6014052033424377,
+      "learning_rate": 3.66403188363037e-06,
+      "loss": 0.9024,
+      "step": 9860
+    },
+    {
+      "epoch": 0.8810954497732705,
+      "grad_norm": 0.4804718792438507,
+      "learning_rate": 3.6585963915045264e-06,
+      "loss": 0.9533,
+      "step": 9861
+    },
+    {
+      "epoch": 0.8811848013045324,
+      "grad_norm": 0.5219516754150391,
+      "learning_rate": 3.6531647809651904e-06,
+      "loss": 0.9346,
+      "step": 9862
+    },
+    {
+      "epoch": 0.8812741528357942,
+      "grad_norm": 0.4467054307460785,
+      "learning_rate": 3.647737052467326e-06,
+      "loss": 0.969,
+      "step": 9863
+    },
+    {
+      "epoch": 0.8813635043670561,
+      "grad_norm": 0.4762718081474304,
+      "learning_rate": 3.642313206465564e-06,
+      "loss": 0.9917,
+      "step": 9864
+    },
+    {
+      "epoch": 0.881452855898318,
+      "grad_norm": 0.5307314395904541,
+      "learning_rate": 3.6368932434142076e-06,
+      "loss": 0.9542,
+      "step": 9865
+    },
+    {
+      "epoch": 0.8815422074295798,
+      "grad_norm": 0.38814452290534973,
+      "learning_rate": 3.631477163767233e-06,
+      "loss": 0.9693,
+      "step": 9866
+    },
+    {
+      "epoch": 0.8816315589608417,
+      "grad_norm": 0.4484859108924866,
+      "learning_rate": 3.6260649679783044e-06,
+      "loss": 0.9018,
+      "step": 9867
+    },
+    {
+      "epoch": 0.8817209104921035,
+      "grad_norm": 0.6517726182937622,
+      "learning_rate": 3.620656656500743e-06,
+      "loss": 0.7952,
+      "step": 9868
+    },
+    {
+      "epoch": 0.8818102620233654,
+      "grad_norm": 0.5478501915931702,
+      "learning_rate": 3.615252229787558e-06,
+      "loss": 0.8884,
+      "step": 9869
+    },
+    {
+      "epoch": 0.8818996135546273,
+      "grad_norm": 0.5623330473899841,
+      "learning_rate": 3.6098516882914213e-06,
+      "loss": 0.9665,
+      "step": 9870
+    },
+    {
+      "epoch": 0.8819889650858892,
+      "grad_norm": 0.5018959641456604,
+      "learning_rate": 3.6044550324646987e-06,
+      "loss": 0.9083,
+      "step": 9871
+    },
+    {
+      "epoch": 0.8820783166171511,
+      "grad_norm": 0.3743535280227661,
+      "learning_rate": 3.599062262759395e-06,
+      "loss": 0.9624,
+      "step": 9872
+    },
+    {
+      "epoch": 0.8821676681484129,
+      "grad_norm": 0.6572086215019226,
+      "learning_rate": 3.5936733796272327e-06,
+      "loss": 0.8587,
+      "step": 9873
+    },
+    {
+      "epoch": 0.8822570196796747,
+      "grad_norm": 0.49752864241600037,
+      "learning_rate": 3.588288383519578e-06,
+      "loss": 0.8681,
+      "step": 9874
+    },
+    {
+      "epoch": 0.8823463712109366,
+      "grad_norm": 0.5613961219787598,
+      "learning_rate": 3.5829072748874813e-06,
+      "loss": 0.9468,
+      "step": 9875
+    },
+    {
+      "epoch": 0.8824357227421985,
+      "grad_norm": 0.5418183207511902,
+      "learning_rate": 3.577530054181677e-06,
+      "loss": 0.8794,
+      "step": 9876
+    },
+    {
+      "epoch": 0.8825250742734604,
+      "grad_norm": 0.5422852039337158,
+      "learning_rate": 3.5721567218525542e-06,
+      "loss": 0.9458,
+      "step": 9877
+    },
+    {
+      "epoch": 0.8826144258047223,
+      "grad_norm": 0.42068326473236084,
+      "learning_rate": 3.566787278350192e-06,
+      "loss": 0.9237,
+      "step": 9878
+    },
+    {
+      "epoch": 0.882703777335984,
+      "grad_norm": 0.4543195068836212,
+      "learning_rate": 3.5614217241243363e-06,
+      "loss": 0.9494,
+      "step": 9879
+    },
+    {
+      "epoch": 0.8827931288672459,
+      "grad_norm": 0.5260372757911682,
+      "learning_rate": 3.556060059624411e-06,
+      "loss": 0.9434,
+      "step": 9880
+    },
+    {
+      "epoch": 0.8828824803985078,
+      "grad_norm": 0.5226284861564636,
+      "learning_rate": 3.550702285299523e-06,
+      "loss": 0.9078,
+      "step": 9881
+    },
+    {
+      "epoch": 0.8829718319297697,
+      "grad_norm": 0.4836858808994293,
+      "learning_rate": 3.5453484015984253e-06,
+      "loss": 0.9321,
+      "step": 9882
+    },
+    {
+      "epoch": 0.8830611834610316,
+      "grad_norm": 0.47180086374282837,
+      "learning_rate": 3.539998408969569e-06,
+      "loss": 0.885,
+      "step": 9883
+    },
+    {
+      "epoch": 0.8831505349922935,
+      "grad_norm": 0.473090261220932,
+      "learning_rate": 3.5346523078610748e-06,
+      "loss": 0.9625,
+      "step": 9884
+    },
+    {
+      "epoch": 0.8832398865235553,
+      "grad_norm": 0.4888576567173004,
+      "learning_rate": 3.529310098720734e-06,
+      "loss": 0.8954,
+      "step": 9885
+    },
+    {
+      "epoch": 0.8833292380548171,
+      "grad_norm": 0.46975868940353394,
+      "learning_rate": 3.5239717819960104e-06,
+      "loss": 1.0308,
+      "step": 9886
+    },
+    {
+      "epoch": 0.883418589586079,
+      "grad_norm": 0.4399878680706024,
+      "learning_rate": 3.5186373581340636e-06,
+      "loss": 0.9445,
+      "step": 9887
+    },
+    {
+      "epoch": 0.8835079411173409,
+      "grad_norm": 0.4216008484363556,
+      "learning_rate": 3.5133068275816806e-06,
+      "loss": 0.9615,
+      "step": 9888
+    },
+    {
+      "epoch": 0.8835972926486028,
+      "grad_norm": 0.6720105409622192,
+      "learning_rate": 3.5079801907853648e-06,
+      "loss": 0.8721,
+      "step": 9889
+    },
+    {
+      "epoch": 0.8836866441798646,
+      "grad_norm": 0.4864836037158966,
+      "learning_rate": 3.5026574481912767e-06,
+      "loss": 0.9017,
+      "step": 9890
+    },
+    {
+      "epoch": 0.8837759957111265,
+      "grad_norm": 0.44451501965522766,
+      "learning_rate": 3.4973386002452535e-06,
+      "loss": 0.9123,
+      "step": 9891
+    },
+    {
+      "epoch": 0.8838653472423884,
+      "grad_norm": 0.4861214756965637,
+      "learning_rate": 3.492023647392817e-06,
+      "loss": 0.9691,
+      "step": 9892
+    },
+    {
+      "epoch": 0.8839546987736502,
+      "grad_norm": 0.5341703295707703,
+      "learning_rate": 3.4867125900791274e-06,
+      "loss": 0.8504,
+      "step": 9893
+    },
+    {
+      "epoch": 0.8840440503049121,
+      "grad_norm": 0.4494452178478241,
+      "learning_rate": 3.481405428749057e-06,
+      "loss": 0.962,
+      "step": 9894
+    },
+    {
+      "epoch": 0.884133401836174,
+      "grad_norm": 0.5674958229064941,
+      "learning_rate": 3.4761021638471337e-06,
+      "loss": 0.9775,
+      "step": 9895
+    },
+    {
+      "epoch": 0.8842227533674358,
+      "grad_norm": 0.480878621339798,
+      "learning_rate": 3.4708027958175625e-06,
+      "loss": 0.8908,
+      "step": 9896
+    },
+    {
+      "epoch": 0.8843121048986977,
+      "grad_norm": 0.6041078567504883,
+      "learning_rate": 3.4655073251042226e-06,
+      "loss": 0.8541,
+      "step": 9897
+    },
+    {
+      "epoch": 0.8844014564299596,
+      "grad_norm": 0.47140049934387207,
+      "learning_rate": 3.4602157521506638e-06,
+      "loss": 0.9046,
+      "step": 9898
+    },
+    {
+      "epoch": 0.8844908079612215,
+      "grad_norm": 0.5176533460617065,
+      "learning_rate": 3.4549280774001158e-06,
+      "loss": 0.9528,
+      "step": 9899
+    },
+    {
+      "epoch": 0.8845801594924833,
+      "grad_norm": 0.48900794982910156,
+      "learning_rate": 3.4496443012954795e-06,
+      "loss": 0.904,
+      "step": 9900
+    },
+    {
+      "epoch": 0.8846695110237451,
+      "grad_norm": 0.4905250668525696,
+      "learning_rate": 3.444364424279323e-06,
+      "loss": 0.8826,
+      "step": 9901
+    },
+    {
+      "epoch": 0.884758862555007,
+      "grad_norm": 0.5473768711090088,
+      "learning_rate": 3.4390884467938978e-06,
+      "loss": 0.9235,
+      "step": 9902
+    },
+    {
+      "epoch": 0.8848482140862689,
+      "grad_norm": 0.4095969498157501,
+      "learning_rate": 3.433816369281112e-06,
+      "loss": 0.9529,
+      "step": 9903
+    },
+    {
+      "epoch": 0.8849375656175308,
+      "grad_norm": 0.5141059756278992,
+      "learning_rate": 3.428548192182568e-06,
+      "loss": 0.9323,
+      "step": 9904
+    },
+    {
+      "epoch": 0.8850269171487927,
+      "grad_norm": 0.5582877993583679,
+      "learning_rate": 3.423283915939529e-06,
+      "loss": 0.9181,
+      "step": 9905
+    },
+    {
+      "epoch": 0.8851162686800546,
+      "grad_norm": 0.4510602653026581,
+      "learning_rate": 3.418023540992932e-06,
+      "loss": 1.0325,
+      "step": 9906
+    },
+    {
+      "epoch": 0.8852056202113163,
+      "grad_norm": 0.47843095660209656,
+      "learning_rate": 3.41276706778339e-06,
+      "loss": 1.007,
+      "step": 9907
+    },
+    {
+      "epoch": 0.8852949717425782,
+      "grad_norm": 0.5486652255058289,
+      "learning_rate": 3.4075144967511963e-06,
+      "loss": 0.9213,
+      "step": 9908
+    },
+    {
+      "epoch": 0.8853843232738401,
+      "grad_norm": 0.4576515257358551,
+      "learning_rate": 3.4022658283362985e-06,
+      "loss": 0.9397,
+      "step": 9909
+    },
+    {
+      "epoch": 0.885473674805102,
+      "grad_norm": 0.47343116998672485,
+      "learning_rate": 3.397021062978334e-06,
+      "loss": 0.9073,
+      "step": 9910
+    },
+    {
+      "epoch": 0.8855630263363639,
+      "grad_norm": 0.5906479358673096,
+      "learning_rate": 3.3917802011166067e-06,
+      "loss": 0.867,
+      "step": 9911
+    },
+    {
+      "epoch": 0.8856523778676257,
+      "grad_norm": 0.4365607798099518,
+      "learning_rate": 3.3865432431901046e-06,
+      "loss": 0.9728,
+      "step": 9912
+    },
+    {
+      "epoch": 0.8857417293988876,
+      "grad_norm": 0.47159871459007263,
+      "learning_rate": 3.3813101896374653e-06,
+      "loss": 0.9237,
+      "step": 9913
+    },
+    {
+      "epoch": 0.8858310809301494,
+      "grad_norm": 0.6291021704673767,
+      "learning_rate": 3.3760810408970113e-06,
+      "loss": 0.9258,
+      "step": 9914
+    },
+    {
+      "epoch": 0.8859204324614113,
+      "grad_norm": 0.4011688530445099,
+      "learning_rate": 3.3708557974067523e-06,
+      "loss": 0.9217,
+      "step": 9915
+    },
+    {
+      "epoch": 0.8860097839926732,
+      "grad_norm": 0.6337020397186279,
+      "learning_rate": 3.3656344596043442e-06,
+      "loss": 0.9362,
+      "step": 9916
+    },
+    {
+      "epoch": 0.886099135523935,
+      "grad_norm": 0.47050267457962036,
+      "learning_rate": 3.3604170279271374e-06,
+      "loss": 0.9049,
+      "step": 9917
+    },
+    {
+      "epoch": 0.8861884870551969,
+      "grad_norm": 0.5305570960044861,
+      "learning_rate": 3.3552035028121486e-06,
+      "loss": 0.8664,
+      "step": 9918
+    },
+    {
+      "epoch": 0.8862778385864588,
+      "grad_norm": 0.4407976269721985,
+      "learning_rate": 3.3499938846960675e-06,
+      "loss": 0.9272,
+      "step": 9919
+    },
+    {
+      "epoch": 0.8863671901177207,
+      "grad_norm": 0.4790654182434082,
+      "learning_rate": 3.3447881740152566e-06,
+      "loss": 0.93,
+      "step": 9920
+    },
+    {
+      "epoch": 0.8864565416489825,
+      "grad_norm": 0.5529206395149231,
+      "learning_rate": 3.3395863712057383e-06,
+      "loss": 0.8797,
+      "step": 9921
+    },
+    {
+      "epoch": 0.8865458931802443,
+      "grad_norm": 0.6508722901344299,
+      "learning_rate": 3.334388476703226e-06,
+      "loss": 0.8327,
+      "step": 9922
+    },
+    {
+      "epoch": 0.8866352447115062,
+      "grad_norm": 0.4920750558376312,
+      "learning_rate": 3.329194490943094e-06,
+      "loss": 0.9091,
+      "step": 9923
+    },
+    {
+      "epoch": 0.8867245962427681,
+      "grad_norm": 0.5537749528884888,
+      "learning_rate": 3.32400441436041e-06,
+      "loss": 0.9325,
+      "step": 9924
+    },
+    {
+      "epoch": 0.88681394777403,
+      "grad_norm": 0.5373173952102661,
+      "learning_rate": 3.3188182473898767e-06,
+      "loss": 0.8741,
+      "step": 9925
+    },
+    {
+      "epoch": 0.8869032993052919,
+      "grad_norm": 0.5053735375404358,
+      "learning_rate": 3.313635990465902e-06,
+      "loss": 0.8817,
+      "step": 9926
+    },
+    {
+      "epoch": 0.8869926508365538,
+      "grad_norm": 0.5587336421012878,
+      "learning_rate": 3.3084576440225555e-06,
+      "loss": 0.9233,
+      "step": 9927
+    },
+    {
+      "epoch": 0.8870820023678155,
+      "grad_norm": 0.45530253648757935,
+      "learning_rate": 3.3032832084935795e-06,
+      "loss": 0.9825,
+      "step": 9928
+    },
+    {
+      "epoch": 0.8871713538990774,
+      "grad_norm": 0.575091540813446,
+      "learning_rate": 3.298112684312382e-06,
+      "loss": 0.8839,
+      "step": 9929
+    },
+    {
+      "epoch": 0.8872607054303393,
+      "grad_norm": 0.5958606004714966,
+      "learning_rate": 3.292946071912051e-06,
+      "loss": 0.8694,
+      "step": 9930
+    },
+    {
+      "epoch": 0.8873500569616012,
+      "grad_norm": 0.5950548648834229,
+      "learning_rate": 3.2877833717253503e-06,
+      "loss": 0.9257,
+      "step": 9931
+    },
+    {
+      "epoch": 0.8874394084928631,
+      "grad_norm": 0.46099838614463806,
+      "learning_rate": 3.282624584184718e-06,
+      "loss": 0.9811,
+      "step": 9932
+    },
+    {
+      "epoch": 0.887528760024125,
+      "grad_norm": 0.5730208158493042,
+      "learning_rate": 3.277469709722242e-06,
+      "loss": 0.9349,
+      "step": 9933
+    },
+    {
+      "epoch": 0.8876181115553868,
+      "grad_norm": 0.4635668992996216,
+      "learning_rate": 3.2723187487696982e-06,
+      "loss": 0.9519,
+      "step": 9934
+    },
+    {
+      "epoch": 0.8877074630866486,
+      "grad_norm": 0.5630105137825012,
+      "learning_rate": 3.267171701758548e-06,
+      "loss": 0.9327,
+      "step": 9935
+    },
+    {
+      "epoch": 0.8877968146179105,
+      "grad_norm": 0.43999359011650085,
+      "learning_rate": 3.262028569119896e-06,
+      "loss": 0.9838,
+      "step": 9936
+    },
+    {
+      "epoch": 0.8878861661491724,
+      "grad_norm": 0.4146654009819031,
+      "learning_rate": 3.2568893512845477e-06,
+      "loss": 0.9695,
+      "step": 9937
+    },
+    {
+      "epoch": 0.8879755176804343,
+      "grad_norm": 0.48017647862434387,
+      "learning_rate": 3.251754048682959e-06,
+      "loss": 0.8955,
+      "step": 9938
+    },
+    {
+      "epoch": 0.8880648692116961,
+      "grad_norm": 0.48065900802612305,
+      "learning_rate": 3.2466226617452745e-06,
+      "loss": 0.9756,
+      "step": 9939
+    },
+    {
+      "epoch": 0.888154220742958,
+      "grad_norm": 0.5396913290023804,
+      "learning_rate": 3.2414951909012946e-06,
+      "loss": 0.96,
+      "step": 9940
+    },
+    {
+      "epoch": 0.8882435722742198,
+      "grad_norm": 0.665141224861145,
+      "learning_rate": 3.2363716365804984e-06,
+      "loss": 0.887,
+      "step": 9941
+    },
+    {
+      "epoch": 0.8883329238054817,
+      "grad_norm": 0.548674464225769,
+      "learning_rate": 3.2312519992120538e-06,
+      "loss": 0.8511,
+      "step": 9942
+    },
+    {
+      "epoch": 0.8884222753367436,
+      "grad_norm": 0.6099333763122559,
+      "learning_rate": 3.226136279224762e-06,
+      "loss": 0.9248,
+      "step": 9943
+    },
+    {
+      "epoch": 0.8885116268680054,
+      "grad_norm": 0.5216543078422546,
+      "learning_rate": 3.2210244770471356e-06,
+      "loss": 0.8861,
+      "step": 9944
+    },
+    {
+      "epoch": 0.8886009783992673,
+      "grad_norm": 0.4764314591884613,
+      "learning_rate": 3.215916593107332e-06,
+      "loss": 0.945,
+      "step": 9945
+    },
+    {
+      "epoch": 0.8886903299305292,
+      "grad_norm": 0.42511335015296936,
+      "learning_rate": 3.2108126278331983e-06,
+      "loss": 0.9197,
+      "step": 9946
+    },
+    {
+      "epoch": 0.8887796814617911,
+      "grad_norm": 0.5028632283210754,
+      "learning_rate": 3.2057125816522483e-06,
+      "loss": 0.9817,
+      "step": 9947
+    },
+    {
+      "epoch": 0.8888690329930529,
+      "grad_norm": 0.40669888257980347,
+      "learning_rate": 3.2006164549916563e-06,
+      "loss": 1.0339,
+      "step": 9948
+    },
+    {
+      "epoch": 0.8889583845243147,
+      "grad_norm": 0.4078708291053772,
+      "learning_rate": 3.195524248278281e-06,
+      "loss": 0.9188,
+      "step": 9949
+    },
+    {
+      "epoch": 0.8890477360555766,
+      "grad_norm": 0.45386314392089844,
+      "learning_rate": 3.190435961938654e-06,
+      "loss": 0.9065,
+      "step": 9950
+    },
+    {
+      "epoch": 0.8891370875868385,
+      "grad_norm": 0.39312687516212463,
+      "learning_rate": 3.185351596398961e-06,
+      "loss": 0.9465,
+      "step": 9951
+    },
+    {
+      "epoch": 0.8892264391181004,
+      "grad_norm": 0.4432995319366455,
+      "learning_rate": 3.1802711520850957e-06,
+      "loss": 0.9804,
+      "step": 9952
+    },
+    {
+      "epoch": 0.8893157906493623,
+      "grad_norm": 0.48735445737838745,
+      "learning_rate": 3.1751946294225733e-06,
+      "loss": 0.8804,
+      "step": 9953
+    },
+    {
+      "epoch": 0.8894051421806242,
+      "grad_norm": 0.4991060793399811,
+      "learning_rate": 3.1701220288366197e-06,
+      "loss": 0.9357,
+      "step": 9954
+    },
+    {
+      "epoch": 0.8894944937118859,
+      "grad_norm": 0.5495728254318237,
+      "learning_rate": 3.165053350752112e-06,
+      "loss": 0.8973,
+      "step": 9955
+    },
+    {
+      "epoch": 0.8895838452431478,
+      "grad_norm": 0.5644074082374573,
+      "learning_rate": 3.159988595593616e-06,
+      "loss": 0.8905,
+      "step": 9956
+    },
+    {
+      "epoch": 0.8896731967744097,
+      "grad_norm": 0.5330132842063904,
+      "learning_rate": 3.1549277637853593e-06,
+      "loss": 0.8993,
+      "step": 9957
+    },
+    {
+      "epoch": 0.8897625483056716,
+      "grad_norm": 0.4881329834461212,
+      "learning_rate": 3.1498708557512246e-06,
+      "loss": 0.9377,
+      "step": 9958
+    },
+    {
+      "epoch": 0.8898518998369335,
+      "grad_norm": 0.43445295095443726,
+      "learning_rate": 3.1448178719147957e-06,
+      "loss": 0.9774,
+      "step": 9959
+    },
+    {
+      "epoch": 0.8899412513681954,
+      "grad_norm": 0.5099343657493591,
+      "learning_rate": 3.1397688126993065e-06,
+      "loss": 0.8734,
+      "step": 9960
+    },
+    {
+      "epoch": 0.8900306028994572,
+      "grad_norm": 0.4378454387187958,
+      "learning_rate": 3.134723678527679e-06,
+      "loss": 0.945,
+      "step": 9961
+    },
+    {
+      "epoch": 0.890119954430719,
+      "grad_norm": 0.45414188504219055,
+      "learning_rate": 3.1296824698224924e-06,
+      "loss": 0.9504,
+      "step": 9962
+    },
+    {
+      "epoch": 0.8902093059619809,
+      "grad_norm": 0.40855854749679565,
+      "learning_rate": 3.1246451870059977e-06,
+      "loss": 0.9502,
+      "step": 9963
+    },
+    {
+      "epoch": 0.8902986574932428,
+      "grad_norm": 0.521061360836029,
+      "learning_rate": 3.1196118305001243e-06,
+      "loss": 0.8663,
+      "step": 9964
+    },
+    {
+      "epoch": 0.8903880090245047,
+      "grad_norm": 0.4307143986225128,
+      "learning_rate": 3.114582400726468e-06,
+      "loss": 0.9124,
+      "step": 9965
+    },
+    {
+      "epoch": 0.8904773605557665,
+      "grad_norm": 0.5008573532104492,
+      "learning_rate": 3.109556898106297e-06,
+      "loss": 0.9045,
+      "step": 9966
+    },
+    {
+      "epoch": 0.8905667120870284,
+      "grad_norm": 0.5884765982627869,
+      "learning_rate": 3.1045353230605535e-06,
+      "loss": 0.9376,
+      "step": 9967
+    },
+    {
+      "epoch": 0.8906560636182903,
+      "grad_norm": 0.49903714656829834,
+      "learning_rate": 3.0995176760098445e-06,
+      "loss": 0.9345,
+      "step": 9968
+    },
+    {
+      "epoch": 0.8907454151495521,
+      "grad_norm": 0.4943690299987793,
+      "learning_rate": 3.094503957374456e-06,
+      "loss": 0.9906,
+      "step": 9969
+    },
+    {
+      "epoch": 0.890834766680814,
+      "grad_norm": 0.5311302542686462,
+      "learning_rate": 3.089494167574336e-06,
+      "loss": 0.9613,
+      "step": 9970
+    },
+    {
+      "epoch": 0.8909241182120758,
+      "grad_norm": 0.5208662748336792,
+      "learning_rate": 3.08448830702911e-06,
+      "loss": 0.9179,
+      "step": 9971
+    },
+    {
+      "epoch": 0.8910134697433377,
+      "grad_norm": 0.49993789196014404,
+      "learning_rate": 3.0794863761580805e-06,
+      "loss": 0.8284,
+      "step": 9972
+    },
+    {
+      "epoch": 0.8911028212745996,
+      "grad_norm": 0.5702754259109497,
+      "learning_rate": 3.074488375380197e-06,
+      "loss": 0.9488,
+      "step": 9973
+    },
+    {
+      "epoch": 0.8911921728058615,
+      "grad_norm": 0.41398748755455017,
+      "learning_rate": 3.0694943051140958e-06,
+      "loss": 0.9195,
+      "step": 9974
+    },
+    {
+      "epoch": 0.8912815243371234,
+      "grad_norm": 0.46768733859062195,
+      "learning_rate": 3.0645041657780927e-06,
+      "loss": 0.9542,
+      "step": 9975
+    },
+    {
+      "epoch": 0.8913708758683851,
+      "grad_norm": 0.47688964009284973,
+      "learning_rate": 3.0595179577901643e-06,
+      "loss": 0.8927,
+      "step": 9976
+    },
+    {
+      "epoch": 0.891460227399647,
+      "grad_norm": 0.501356840133667,
+      "learning_rate": 3.054535681567955e-06,
+      "loss": 0.9192,
+      "step": 9977
+    },
+    {
+      "epoch": 0.8915495789309089,
+      "grad_norm": 0.44670113921165466,
+      "learning_rate": 3.0495573375287854e-06,
+      "loss": 0.9633,
+      "step": 9978
+    },
+    {
+      "epoch": 0.8916389304621708,
+      "grad_norm": 0.5801134705543518,
+      "learning_rate": 3.044582926089645e-06,
+      "loss": 0.8739,
+      "step": 9979
+    },
+    {
+      "epoch": 0.8917282819934327,
+      "grad_norm": 0.5927140712738037,
+      "learning_rate": 3.039612447667195e-06,
+      "loss": 0.9109,
+      "step": 9980
+    },
+    {
+      "epoch": 0.8918176335246946,
+      "grad_norm": 0.49671000242233276,
+      "learning_rate": 3.034645902677763e-06,
+      "loss": 0.9923,
+      "step": 9981
+    },
+    {
+      "epoch": 0.8919069850559564,
+      "grad_norm": 0.46051478385925293,
+      "learning_rate": 3.0296832915373497e-06,
+      "loss": 0.9259,
+      "step": 9982
+    },
+    {
+      "epoch": 0.8919963365872182,
+      "grad_norm": 0.5306558609008789,
+      "learning_rate": 3.024724614661639e-06,
+      "loss": 0.913,
+      "step": 9983
+    },
+    {
+      "epoch": 0.8920856881184801,
+      "grad_norm": 0.3877553939819336,
+      "learning_rate": 3.0197698724659497e-06,
+      "loss": 1.001,
+      "step": 9984
+    },
+    {
+      "epoch": 0.892175039649742,
+      "grad_norm": 0.5105606913566589,
+      "learning_rate": 3.0148190653653096e-06,
+      "loss": 0.9371,
+      "step": 9985
+    },
+    {
+      "epoch": 0.8922643911810039,
+      "grad_norm": 0.5007322430610657,
+      "learning_rate": 3.009872193774399e-06,
+      "loss": 0.9932,
+      "step": 9986
+    },
+    {
+      "epoch": 0.8923537427122658,
+      "grad_norm": 0.5739557147026062,
+      "learning_rate": 3.0049292581075692e-06,
+      "loss": 0.8694,
+      "step": 9987
+    },
+    {
+      "epoch": 0.8924430942435276,
+      "grad_norm": 0.4719444215297699,
+      "learning_rate": 2.9999902587788507e-06,
+      "loss": 1.0219,
+      "step": 9988
+    },
+    {
+      "epoch": 0.8925324457747895,
+      "grad_norm": 0.4949089586734772,
+      "learning_rate": 2.9950551962019293e-06,
+      "loss": 0.9235,
+      "step": 9989
+    },
+    {
+      "epoch": 0.8926217973060513,
+      "grad_norm": 0.45871758460998535,
+      "learning_rate": 2.990124070790179e-06,
+      "loss": 0.9429,
+      "step": 9990
+    },
+    {
+      "epoch": 0.8927111488373132,
+      "grad_norm": 0.5253780484199524,
+      "learning_rate": 2.98519688295662e-06,
+      "loss": 0.939,
+      "step": 9991
+    },
+    {
+      "epoch": 0.8928005003685751,
+      "grad_norm": 0.4314812123775482,
+      "learning_rate": 2.9802736331139615e-06,
+      "loss": 0.9584,
+      "step": 9992
+    },
+    {
+      "epoch": 0.8928898518998369,
+      "grad_norm": 0.473132848739624,
+      "learning_rate": 2.9753543216745784e-06,
+      "loss": 0.9659,
+      "step": 9993
+    },
+    {
+      "epoch": 0.8929792034310988,
+      "grad_norm": 0.4422551989555359,
+      "learning_rate": 2.9704389490505303e-06,
+      "loss": 0.9317,
+      "step": 9994
+    },
+    {
+      "epoch": 0.8930685549623607,
+      "grad_norm": 0.5181037783622742,
+      "learning_rate": 2.9655275156535103e-06,
+      "loss": 0.9623,
+      "step": 9995
+    },
+    {
+      "epoch": 0.8931579064936226,
+      "grad_norm": 0.5638878345489502,
+      "learning_rate": 2.960620021894911e-06,
+      "loss": 0.908,
+      "step": 9996
+    },
+    {
+      "epoch": 0.8932472580248844,
+      "grad_norm": 0.5014554858207703,
+      "learning_rate": 2.955716468185793e-06,
+      "loss": 0.8998,
+      "step": 9997
+    },
+    {
+      "epoch": 0.8933366095561462,
+      "grad_norm": 0.46494272351264954,
+      "learning_rate": 2.950816854936872e-06,
+      "loss": 0.9615,
+      "step": 9998
+    },
+    {
+      "epoch": 0.8934259610874081,
+      "grad_norm": 0.4722679555416107,
+      "learning_rate": 2.9459211825585475e-06,
+      "loss": 0.9962,
+      "step": 9999
+    },
+    {
+      "epoch": 0.89351531261867,
+      "grad_norm": 0.6044051051139832,
+      "learning_rate": 2.941029451460886e-06,
+      "loss": 0.9006,
+      "step": 10000
+    },
+    {
+      "epoch": 0.8936046641499319,
+      "grad_norm": 0.4716538190841675,
+      "learning_rate": 2.936141662053621e-06,
+      "loss": 0.9416,
+      "step": 10001
+    },
+    {
+      "epoch": 0.8936940156811938,
+      "grad_norm": 0.5190637111663818,
+      "learning_rate": 2.931257814746158e-06,
+      "loss": 0.8654,
+      "step": 10002
+    },
+    {
+      "epoch": 0.8937833672124555,
+      "grad_norm": 0.46541985869407654,
+      "learning_rate": 2.926377909947575e-06,
+      "loss": 0.9391,
+      "step": 10003
+    },
+    {
+      "epoch": 0.8938727187437174,
+      "grad_norm": 0.46355336904525757,
+      "learning_rate": 2.9215019480666015e-06,
+      "loss": 0.8925,
+      "step": 10004
+    },
+    {
+      "epoch": 0.8939620702749793,
+      "grad_norm": 0.5161470770835876,
+      "learning_rate": 2.916629929511666e-06,
+      "loss": 0.8671,
+      "step": 10005
+    },
+    {
+      "epoch": 0.8940514218062412,
+      "grad_norm": 0.544810950756073,
+      "learning_rate": 2.911761854690842e-06,
+      "loss": 0.911,
+      "step": 10006
+    },
+    {
+      "epoch": 0.8941407733375031,
+      "grad_norm": 0.4915909469127655,
+      "learning_rate": 2.9068977240118867e-06,
+      "loss": 0.8727,
+      "step": 10007
+    },
+    {
+      "epoch": 0.894230124868765,
+      "grad_norm": 0.658275306224823,
+      "learning_rate": 2.9020375378822297e-06,
+      "loss": 0.9053,
+      "step": 10008
+    },
+    {
+      "epoch": 0.8943194764000268,
+      "grad_norm": 0.48863136768341064,
+      "learning_rate": 2.897181296708951e-06,
+      "loss": 0.8799,
+      "step": 10009
+    },
+    {
+      "epoch": 0.8944088279312886,
+      "grad_norm": 0.46332526206970215,
+      "learning_rate": 2.8923290008988193e-06,
+      "loss": 0.9081,
+      "step": 10010
+    },
+    {
+      "epoch": 0.8944981794625505,
+      "grad_norm": 0.4719219505786896,
+      "learning_rate": 2.8874806508582652e-06,
+      "loss": 0.9298,
+      "step": 10011
+    },
+    {
+      "epoch": 0.8945875309938124,
+      "grad_norm": 0.4817638099193573,
+      "learning_rate": 2.882636246993392e-06,
+      "loss": 0.9128,
+      "step": 10012
+    },
+    {
+      "epoch": 0.8946768825250743,
+      "grad_norm": 0.49703267216682434,
+      "learning_rate": 2.877795789709975e-06,
+      "loss": 0.8182,
+      "step": 10013
+    },
+    {
+      "epoch": 0.8947662340563362,
+      "grad_norm": 0.4652135372161865,
+      "learning_rate": 2.87295927941344e-06,
+      "loss": 0.9703,
+      "step": 10014
+    },
+    {
+      "epoch": 0.894855585587598,
+      "grad_norm": 0.4329572916030884,
+      "learning_rate": 2.868126716508901e-06,
+      "loss": 0.939,
+      "step": 10015
+    },
+    {
+      "epoch": 0.8949449371188599,
+      "grad_norm": 0.4363684356212616,
+      "learning_rate": 2.8632981014011463e-06,
+      "loss": 0.8893,
+      "step": 10016
+    },
+    {
+      "epoch": 0.8950342886501217,
+      "grad_norm": 0.564899742603302,
+      "learning_rate": 2.8584734344946073e-06,
+      "loss": 0.9229,
+      "step": 10017
+    },
+    {
+      "epoch": 0.8951236401813836,
+      "grad_norm": 0.5711920261383057,
+      "learning_rate": 2.853652716193417e-06,
+      "loss": 0.8848,
+      "step": 10018
+    },
+    {
+      "epoch": 0.8952129917126455,
+      "grad_norm": 0.4611048996448517,
+      "learning_rate": 2.8488359469013514e-06,
+      "loss": 0.8717,
+      "step": 10019
+    },
+    {
+      "epoch": 0.8953023432439073,
+      "grad_norm": 0.7279905676841736,
+      "learning_rate": 2.844023127021872e-06,
+      "loss": 0.7405,
+      "step": 10020
+    },
+    {
+      "epoch": 0.8953916947751692,
+      "grad_norm": 0.5130376219749451,
+      "learning_rate": 2.839214256958106e-06,
+      "loss": 0.9462,
+      "step": 10021
+    },
+    {
+      "epoch": 0.8954810463064311,
+      "grad_norm": 0.43503376841545105,
+      "learning_rate": 2.8344093371128424e-06,
+      "loss": 0.8974,
+      "step": 10022
+    },
+    {
+      "epoch": 0.895570397837693,
+      "grad_norm": 0.5059356689453125,
+      "learning_rate": 2.8296083678885477e-06,
+      "loss": 0.9436,
+      "step": 10023
+    },
+    {
+      "epoch": 0.8956597493689548,
+      "grad_norm": 0.5730863213539124,
+      "learning_rate": 2.8248113496873507e-06,
+      "loss": 0.8933,
+      "step": 10024
+    },
+    {
+      "epoch": 0.8957491009002166,
+      "grad_norm": 0.4714071750640869,
+      "learning_rate": 2.8200182829110523e-06,
+      "loss": 0.9942,
+      "step": 10025
+    },
+    {
+      "epoch": 0.8958384524314785,
+      "grad_norm": 0.4940183758735657,
+      "learning_rate": 2.8152291679611255e-06,
+      "loss": 0.9125,
+      "step": 10026
+    },
+    {
+      "epoch": 0.8959278039627404,
+      "grad_norm": 0.6629910469055176,
+      "learning_rate": 2.810444005238716e-06,
+      "loss": 0.8461,
+      "step": 10027
+    },
+    {
+      "epoch": 0.8960171554940023,
+      "grad_norm": 0.4740060865879059,
+      "learning_rate": 2.80566279514462e-06,
+      "loss": 1.0388,
+      "step": 10028
+    },
+    {
+      "epoch": 0.8961065070252642,
+      "grad_norm": 0.4990270733833313,
+      "learning_rate": 2.800885538079323e-06,
+      "loss": 0.8857,
+      "step": 10029
+    },
+    {
+      "epoch": 0.8961958585565261,
+      "grad_norm": 0.4073428213596344,
+      "learning_rate": 2.796112234442966e-06,
+      "loss": 0.9675,
+      "step": 10030
+    },
+    {
+      "epoch": 0.8962852100877878,
+      "grad_norm": 0.541317343711853,
+      "learning_rate": 2.791342884635362e-06,
+      "loss": 0.9653,
+      "step": 10031
+    },
+    {
+      "epoch": 0.8963745616190497,
+      "grad_norm": 0.3983585834503174,
+      "learning_rate": 2.7865774890560025e-06,
+      "loss": 0.9158,
+      "step": 10032
+    },
+    {
+      "epoch": 0.8964639131503116,
+      "grad_norm": 0.4200194776058197,
+      "learning_rate": 2.7818160481040465e-06,
+      "loss": 0.9852,
+      "step": 10033
+    },
+    {
+      "epoch": 0.8965532646815735,
+      "grad_norm": 0.48257651925086975,
+      "learning_rate": 2.7770585621782973e-06,
+      "loss": 0.9329,
+      "step": 10034
+    },
+    {
+      "epoch": 0.8966426162128354,
+      "grad_norm": 0.4446716010570526,
+      "learning_rate": 2.772305031677258e-06,
+      "loss": 0.9151,
+      "step": 10035
+    },
+    {
+      "epoch": 0.8967319677440972,
+      "grad_norm": 0.4650631546974182,
+      "learning_rate": 2.767555456999077e-06,
+      "loss": 0.956,
+      "step": 10036
+    },
+    {
+      "epoch": 0.8968213192753591,
+      "grad_norm": 0.5073647499084473,
+      "learning_rate": 2.762809838541591e-06,
+      "loss": 0.9766,
+      "step": 10037
+    },
+    {
+      "epoch": 0.8969106708066209,
+      "grad_norm": 0.5146875977516174,
+      "learning_rate": 2.758068176702294e-06,
+      "loss": 0.9526,
+      "step": 10038
+    },
+    {
+      "epoch": 0.8970000223378828,
+      "grad_norm": 0.5599371790885925,
+      "learning_rate": 2.7533304718783516e-06,
+      "loss": 0.934,
+      "step": 10039
+    },
+    {
+      "epoch": 0.8970893738691447,
+      "grad_norm": 0.5773612856864929,
+      "learning_rate": 2.74859672446659e-06,
+      "loss": 0.9475,
+      "step": 10040
+    },
+    {
+      "epoch": 0.8971787254004066,
+      "grad_norm": 0.5281670093536377,
+      "learning_rate": 2.7438669348635202e-06,
+      "loss": 0.9335,
+      "step": 10041
+    },
+    {
+      "epoch": 0.8972680769316684,
+      "grad_norm": 0.48857590556144714,
+      "learning_rate": 2.7391411034653094e-06,
+      "loss": 0.8603,
+      "step": 10042
+    },
+    {
+      "epoch": 0.8973574284629303,
+      "grad_norm": 0.4282703995704651,
+      "learning_rate": 2.734419230667801e-06,
+      "loss": 0.9902,
+      "step": 10043
+    },
+    {
+      "epoch": 0.8974467799941922,
+      "grad_norm": 0.5266483426094055,
+      "learning_rate": 2.7297013168664897e-06,
+      "loss": 0.8483,
+      "step": 10044
+    },
+    {
+      "epoch": 0.897536131525454,
+      "grad_norm": 0.5313611626625061,
+      "learning_rate": 2.7249873624565604e-06,
+      "loss": 0.8934,
+      "step": 10045
+    },
+    {
+      "epoch": 0.8976254830567159,
+      "grad_norm": 0.4362044632434845,
+      "learning_rate": 2.720277367832852e-06,
+      "loss": 0.9271,
+      "step": 10046
+    },
+    {
+      "epoch": 0.8977148345879777,
+      "grad_norm": 0.5316069722175598,
+      "learning_rate": 2.7155713333898825e-06,
+      "loss": 0.8928,
+      "step": 10047
+    },
+    {
+      "epoch": 0.8978041861192396,
+      "grad_norm": 0.5463095307350159,
+      "learning_rate": 2.7108692595218254e-06,
+      "loss": 0.932,
+      "step": 10048
+    },
+    {
+      "epoch": 0.8978935376505015,
+      "grad_norm": 0.475721150636673,
+      "learning_rate": 2.706171146622538e-06,
+      "loss": 0.9658,
+      "step": 10049
+    },
+    {
+      "epoch": 0.8979828891817634,
+      "grad_norm": 0.4900808036327362,
+      "learning_rate": 2.7014769950855334e-06,
+      "loss": 0.963,
+      "step": 10050
+    },
+    {
+      "epoch": 0.8980722407130253,
+      "grad_norm": 0.411359578371048,
+      "learning_rate": 2.6967868053039913e-06,
+      "loss": 0.972,
+      "step": 10051
+    },
+    {
+      "epoch": 0.898161592244287,
+      "grad_norm": 0.5783373713493347,
+      "learning_rate": 2.6921005776707755e-06,
+      "loss": 0.9393,
+      "step": 10052
+    },
+    {
+      "epoch": 0.8982509437755489,
+      "grad_norm": 0.4613523483276367,
+      "learning_rate": 2.6874183125784047e-06,
+      "loss": 0.9651,
+      "step": 10053
+    },
+    {
+      "epoch": 0.8983402953068108,
+      "grad_norm": 0.5736072063446045,
+      "learning_rate": 2.682740010419066e-06,
+      "loss": 0.8834,
+      "step": 10054
+    },
+    {
+      "epoch": 0.8984296468380727,
+      "grad_norm": 0.4773879051208496,
+      "learning_rate": 2.678065671584612e-06,
+      "loss": 0.9388,
+      "step": 10055
+    },
+    {
+      "epoch": 0.8985189983693346,
+      "grad_norm": 0.4866504669189453,
+      "learning_rate": 2.673395296466574e-06,
+      "loss": 0.9726,
+      "step": 10056
+    },
+    {
+      "epoch": 0.8986083499005965,
+      "grad_norm": 0.5097846984863281,
+      "learning_rate": 2.6687288854561455e-06,
+      "loss": 0.8879,
+      "step": 10057
+    },
+    {
+      "epoch": 0.8986977014318583,
+      "grad_norm": 0.6318731307983398,
+      "learning_rate": 2.664066438944185e-06,
+      "loss": 0.9752,
+      "step": 10058
+    },
+    {
+      "epoch": 0.8987870529631201,
+      "grad_norm": 0.47188690304756165,
+      "learning_rate": 2.6594079573212303e-06,
+      "loss": 0.8645,
+      "step": 10059
+    },
+    {
+      "epoch": 0.898876404494382,
+      "grad_norm": 0.43872636556625366,
+      "learning_rate": 2.654753440977481e-06,
+      "loss": 0.9638,
+      "step": 10060
+    },
+    {
+      "epoch": 0.8989657560256439,
+      "grad_norm": 0.521487295627594,
+      "learning_rate": 2.650102890302786e-06,
+      "loss": 0.957,
+      "step": 10061
+    },
+    {
+      "epoch": 0.8990551075569058,
+      "grad_norm": 0.5581962466239929,
+      "learning_rate": 2.6454563056866834e-06,
+      "loss": 0.855,
+      "step": 10062
+    },
+    {
+      "epoch": 0.8991444590881676,
+      "grad_norm": 0.43780362606048584,
+      "learning_rate": 2.640813687518384e-06,
+      "loss": 0.9888,
+      "step": 10063
+    },
+    {
+      "epoch": 0.8992338106194295,
+      "grad_norm": 0.46411171555519104,
+      "learning_rate": 2.6361750361867554e-06,
+      "loss": 0.8972,
+      "step": 10064
+    },
+    {
+      "epoch": 0.8993231621506913,
+      "grad_norm": 0.5126439332962036,
+      "learning_rate": 2.631540352080325e-06,
+      "loss": 0.8815,
+      "step": 10065
+    },
+    {
+      "epoch": 0.8994125136819532,
+      "grad_norm": 0.449725866317749,
+      "learning_rate": 2.626909635587299e-06,
+      "loss": 0.9732,
+      "step": 10066
+    },
+    {
+      "epoch": 0.8995018652132151,
+      "grad_norm": 0.47947466373443604,
+      "learning_rate": 2.6222828870955505e-06,
+      "loss": 0.8826,
+      "step": 10067
+    },
+    {
+      "epoch": 0.899591216744477,
+      "grad_norm": 0.4623918831348419,
+      "learning_rate": 2.6176601069926255e-06,
+      "loss": 0.9146,
+      "step": 10068
+    },
+    {
+      "epoch": 0.8996805682757388,
+      "grad_norm": 0.5010434985160828,
+      "learning_rate": 2.613041295665719e-06,
+      "loss": 0.9345,
+      "step": 10069
+    },
+    {
+      "epoch": 0.8997699198070007,
+      "grad_norm": 0.5494645833969116,
+      "learning_rate": 2.608426453501722e-06,
+      "loss": 0.8777,
+      "step": 10070
+    },
+    {
+      "epoch": 0.8998592713382626,
+      "grad_norm": 0.4975086450576782,
+      "learning_rate": 2.6038155808871587e-06,
+      "loss": 0.8943,
+      "step": 10071
+    },
+    {
+      "epoch": 0.8999486228695244,
+      "grad_norm": 0.4562755823135376,
+      "learning_rate": 2.5992086782082536e-06,
+      "loss": 0.8846,
+      "step": 10072
+    },
+    {
+      "epoch": 0.9000379744007863,
+      "grad_norm": 0.5983653664588928,
+      "learning_rate": 2.5946057458508756e-06,
+      "loss": 0.8868,
+      "step": 10073
+    },
+    {
+      "epoch": 0.9001273259320481,
+      "grad_norm": 0.586626410484314,
+      "learning_rate": 2.5900067842005772e-06,
+      "loss": 0.8883,
+      "step": 10074
+    },
+    {
+      "epoch": 0.90021667746331,
+      "grad_norm": 0.5138862729072571,
+      "learning_rate": 2.585411793642556e-06,
+      "loss": 0.9107,
+      "step": 10075
+    },
+    {
+      "epoch": 0.9003060289945719,
+      "grad_norm": 0.45086416602134705,
+      "learning_rate": 2.580820774561704e-06,
+      "loss": 0.9627,
+      "step": 10076
+    },
+    {
+      "epoch": 0.9003953805258338,
+      "grad_norm": 0.5956899523735046,
+      "learning_rate": 2.576233727342564e-06,
+      "loss": 0.9463,
+      "step": 10077
+    },
+    {
+      "epoch": 0.9004847320570957,
+      "grad_norm": 0.47443798184394836,
+      "learning_rate": 2.571650652369351e-06,
+      "loss": 0.9167,
+      "step": 10078
+    },
+    {
+      "epoch": 0.9005740835883574,
+      "grad_norm": 0.5572949051856995,
+      "learning_rate": 2.5670715500259403e-06,
+      "loss": 0.8379,
+      "step": 10079
+    },
+    {
+      "epoch": 0.9006634351196193,
+      "grad_norm": 0.6084146499633789,
+      "learning_rate": 2.5624964206958924e-06,
+      "loss": 0.8469,
+      "step": 10080
+    },
+    {
+      "epoch": 0.9007527866508812,
+      "grad_norm": 0.5597148537635803,
+      "learning_rate": 2.557925264762412e-06,
+      "loss": 0.8754,
+      "step": 10081
+    },
+    {
+      "epoch": 0.9008421381821431,
+      "grad_norm": 0.5465840101242065,
+      "learning_rate": 2.5533580826083926e-06,
+      "loss": 0.8959,
+      "step": 10082
+    },
+    {
+      "epoch": 0.900931489713405,
+      "grad_norm": 0.39822328090667725,
+      "learning_rate": 2.5487948746163726e-06,
+      "loss": 0.9557,
+      "step": 10083
+    },
+    {
+      "epoch": 0.9010208412446669,
+      "grad_norm": 0.45787888765335083,
+      "learning_rate": 2.544235641168585e-06,
+      "loss": 0.8887,
+      "step": 10084
+    },
+    {
+      "epoch": 0.9011101927759287,
+      "grad_norm": 0.5872011184692383,
+      "learning_rate": 2.5396803826468975e-06,
+      "loss": 0.8258,
+      "step": 10085
+    },
+    {
+      "epoch": 0.9011995443071905,
+      "grad_norm": 0.5175109505653381,
+      "learning_rate": 2.53512909943287e-06,
+      "loss": 0.9667,
+      "step": 10086
+    },
+    {
+      "epoch": 0.9012888958384524,
+      "grad_norm": 0.4502319097518921,
+      "learning_rate": 2.5305817919077157e-06,
+      "loss": 0.9462,
+      "step": 10087
+    },
+    {
+      "epoch": 0.9013782473697143,
+      "grad_norm": 0.4700741171836853,
+      "learning_rate": 2.526038460452329e-06,
+      "loss": 0.9298,
+      "step": 10088
+    },
+    {
+      "epoch": 0.9014675989009762,
+      "grad_norm": 0.4288919270038605,
+      "learning_rate": 2.521499105447256e-06,
+      "loss": 0.9719,
+      "step": 10089
+    },
+    {
+      "epoch": 0.901556950432238,
+      "grad_norm": 0.38474833965301514,
+      "learning_rate": 2.516963727272714e-06,
+      "loss": 1.0002,
+      "step": 10090
+    },
+    {
+      "epoch": 0.9016463019634999,
+      "grad_norm": 0.5012978911399841,
+      "learning_rate": 2.512432326308595e-06,
+      "loss": 0.9018,
+      "step": 10091
+    },
+    {
+      "epoch": 0.9017356534947618,
+      "grad_norm": 0.5238143801689148,
+      "learning_rate": 2.5079049029344492e-06,
+      "loss": 0.8814,
+      "step": 10092
+    },
+    {
+      "epoch": 0.9018250050260236,
+      "grad_norm": 0.4294685423374176,
+      "learning_rate": 2.503381457529508e-06,
+      "loss": 0.9577,
+      "step": 10093
+    },
+    {
+      "epoch": 0.9019143565572855,
+      "grad_norm": 0.5884724259376526,
+      "learning_rate": 2.498861990472634e-06,
+      "loss": 0.8814,
+      "step": 10094
+    },
+    {
+      "epoch": 0.9020037080885474,
+      "grad_norm": 0.40673425793647766,
+      "learning_rate": 2.494346502142397e-06,
+      "loss": 0.952,
+      "step": 10095
+    },
+    {
+      "epoch": 0.9020930596198092,
+      "grad_norm": 0.4494728744029999,
+      "learning_rate": 2.4898349929170116e-06,
+      "loss": 0.937,
+      "step": 10096
+    },
+    {
+      "epoch": 0.9021824111510711,
+      "grad_norm": 0.5147069692611694,
+      "learning_rate": 2.4853274631743807e-06,
+      "loss": 0.8933,
+      "step": 10097
+    },
+    {
+      "epoch": 0.902271762682333,
+      "grad_norm": 0.42955800890922546,
+      "learning_rate": 2.4808239132920297e-06,
+      "loss": 0.9895,
+      "step": 10098
+    },
+    {
+      "epoch": 0.9023611142135949,
+      "grad_norm": 0.5820243954658508,
+      "learning_rate": 2.4763243436472016e-06,
+      "loss": 0.8731,
+      "step": 10099
+    },
+    {
+      "epoch": 0.9024504657448567,
+      "grad_norm": 0.46939095854759216,
+      "learning_rate": 2.471828754616773e-06,
+      "loss": 0.9266,
+      "step": 10100
+    },
+    {
+      "epoch": 0.9025398172761185,
+      "grad_norm": 0.5100577473640442,
+      "learning_rate": 2.4673371465772978e-06,
+      "loss": 0.9469,
+      "step": 10101
+    },
+    {
+      "epoch": 0.9026291688073804,
+      "grad_norm": 0.5236694812774658,
+      "learning_rate": 2.4628495199050027e-06,
+      "loss": 0.9135,
+      "step": 10102
+    },
+    {
+      "epoch": 0.9027185203386423,
+      "grad_norm": 0.5193259716033936,
+      "learning_rate": 2.4583658749757656e-06,
+      "loss": 0.9942,
+      "step": 10103
+    },
+    {
+      "epoch": 0.9028078718699042,
+      "grad_norm": 0.553813099861145,
+      "learning_rate": 2.453886212165152e-06,
+      "loss": 0.9319,
+      "step": 10104
+    },
+    {
+      "epoch": 0.9028972234011661,
+      "grad_norm": 0.5031151175498962,
+      "learning_rate": 2.4494105318483674e-06,
+      "loss": 0.8901,
+      "step": 10105
+    },
+    {
+      "epoch": 0.902986574932428,
+      "grad_norm": 0.5096839070320129,
+      "learning_rate": 2.444938834400301e-06,
+      "loss": 0.8819,
+      "step": 10106
+    },
+    {
+      "epoch": 0.9030759264636897,
+      "grad_norm": 0.48195287585258484,
+      "learning_rate": 2.4404711201955088e-06,
+      "loss": 0.969,
+      "step": 10107
+    },
+    {
+      "epoch": 0.9031652779949516,
+      "grad_norm": 0.46852657198905945,
+      "learning_rate": 2.4360073896082138e-06,
+      "loss": 0.8956,
+      "step": 10108
+    },
+    {
+      "epoch": 0.9032546295262135,
+      "grad_norm": 0.4824883043766022,
+      "learning_rate": 2.4315476430122884e-06,
+      "loss": 0.9332,
+      "step": 10109
+    },
+    {
+      "epoch": 0.9033439810574754,
+      "grad_norm": 0.4670718312263489,
+      "learning_rate": 2.4270918807812958e-06,
+      "loss": 0.9048,
+      "step": 10110
+    },
+    {
+      "epoch": 0.9034333325887373,
+      "grad_norm": 0.4740527868270874,
+      "learning_rate": 2.422640103288443e-06,
+      "loss": 0.9462,
+      "step": 10111
+    },
+    {
+      "epoch": 0.9035226841199991,
+      "grad_norm": 0.49257439374923706,
+      "learning_rate": 2.418192310906625e-06,
+      "loss": 0.9087,
+      "step": 10112
+    },
+    {
+      "epoch": 0.903612035651261,
+      "grad_norm": 0.5299413204193115,
+      "learning_rate": 2.413748504008384e-06,
+      "loss": 0.8926,
+      "step": 10113
+    },
+    {
+      "epoch": 0.9037013871825228,
+      "grad_norm": 0.46102991700172424,
+      "learning_rate": 2.4093086829659495e-06,
+      "loss": 0.9209,
+      "step": 10114
+    },
+    {
+      "epoch": 0.9037907387137847,
+      "grad_norm": 0.5814657211303711,
+      "learning_rate": 2.404872848151185e-06,
+      "loss": 1.0147,
+      "step": 10115
+    },
+    {
+      "epoch": 0.9038800902450466,
+      "grad_norm": 0.5453367829322815,
+      "learning_rate": 2.4004409999356437e-06,
+      "loss": 0.8692,
+      "step": 10116
+    },
+    {
+      "epoch": 0.9039694417763084,
+      "grad_norm": 0.5346057415008545,
+      "learning_rate": 2.396013138690545e-06,
+      "loss": 0.9964,
+      "step": 10117
+    },
+    {
+      "epoch": 0.9040587933075703,
+      "grad_norm": 0.541984498500824,
+      "learning_rate": 2.39158926478677e-06,
+      "loss": 0.9529,
+      "step": 10118
+    },
+    {
+      "epoch": 0.9041481448388322,
+      "grad_norm": 0.43673083186149597,
+      "learning_rate": 2.3871693785948614e-06,
+      "loss": 1.0007,
+      "step": 10119
+    },
+    {
+      "epoch": 0.9042374963700941,
+      "grad_norm": 0.42175018787384033,
+      "learning_rate": 2.3827534804850336e-06,
+      "loss": 0.9554,
+      "step": 10120
+    },
+    {
+      "epoch": 0.9043268479013559,
+      "grad_norm": 0.5283427238464355,
+      "learning_rate": 2.3783415708271696e-06,
+      "loss": 0.8791,
+      "step": 10121
+    },
+    {
+      "epoch": 0.9044161994326178,
+      "grad_norm": 0.41755416989326477,
+      "learning_rate": 2.3739336499908005e-06,
+      "loss": 0.9335,
+      "step": 10122
+    },
+    {
+      "epoch": 0.9045055509638796,
+      "grad_norm": 0.4738496243953705,
+      "learning_rate": 2.3695297183451536e-06,
+      "loss": 0.9587,
+      "step": 10123
+    },
+    {
+      "epoch": 0.9045949024951415,
+      "grad_norm": 0.4655263423919678,
+      "learning_rate": 2.3651297762591006e-06,
+      "loss": 0.8765,
+      "step": 10124
+    },
+    {
+      "epoch": 0.9046842540264034,
+      "grad_norm": 0.4394261837005615,
+      "learning_rate": 2.3607338241011747e-06,
+      "loss": 0.9538,
+      "step": 10125
+    },
+    {
+      "epoch": 0.9047736055576653,
+      "grad_norm": 0.4117361605167389,
+      "learning_rate": 2.356341862239586e-06,
+      "loss": 1.0001,
+      "step": 10126
+    },
+    {
+      "epoch": 0.9048629570889272,
+      "grad_norm": 0.5142612457275391,
+      "learning_rate": 2.3519538910422134e-06,
+      "loss": 0.9423,
+      "step": 10127
+    },
+    {
+      "epoch": 0.9049523086201889,
+      "grad_norm": 0.5198079347610474,
+      "learning_rate": 2.3475699108765958e-06,
+      "loss": 0.9456,
+      "step": 10128
+    },
+    {
+      "epoch": 0.9050416601514508,
+      "grad_norm": 0.521613359451294,
+      "learning_rate": 2.3431899221099342e-06,
+      "loss": 0.9063,
+      "step": 10129
+    },
+    {
+      "epoch": 0.9051310116827127,
+      "grad_norm": 0.4279034435749054,
+      "learning_rate": 2.3388139251091067e-06,
+      "loss": 1.0047,
+      "step": 10130
+    },
+    {
+      "epoch": 0.9052203632139746,
+      "grad_norm": 0.5167081356048584,
+      "learning_rate": 2.334441920240643e-06,
+      "loss": 0.9027,
+      "step": 10131
+    },
+    {
+      "epoch": 0.9053097147452365,
+      "grad_norm": 0.494266152381897,
+      "learning_rate": 2.3300739078707446e-06,
+      "loss": 0.929,
+      "step": 10132
+    },
+    {
+      "epoch": 0.9053990662764984,
+      "grad_norm": 0.4619883596897125,
+      "learning_rate": 2.3257098883652795e-06,
+      "loss": 0.944,
+      "step": 10133
+    },
+    {
+      "epoch": 0.9054884178077601,
+      "grad_norm": 0.46336403489112854,
+      "learning_rate": 2.321349862089789e-06,
+      "loss": 0.9754,
+      "step": 10134
+    },
+    {
+      "epoch": 0.905577769339022,
+      "grad_norm": 0.5507116913795471,
+      "learning_rate": 2.3169938294094582e-06,
+      "loss": 0.8841,
+      "step": 10135
+    },
+    {
+      "epoch": 0.9056671208702839,
+      "grad_norm": 0.49154022336006165,
+      "learning_rate": 2.312641790689163e-06,
+      "loss": 0.887,
+      "step": 10136
+    },
+    {
+      "epoch": 0.9057564724015458,
+      "grad_norm": 0.5543971061706543,
+      "learning_rate": 2.3082937462934274e-06,
+      "loss": 0.876,
+      "step": 10137
+    },
+    {
+      "epoch": 0.9058458239328077,
+      "grad_norm": 0.5523365139961243,
+      "learning_rate": 2.3039496965864436e-06,
+      "loss": 0.9618,
+      "step": 10138
+    },
+    {
+      "epoch": 0.9059351754640695,
+      "grad_norm": 0.5370588898658752,
+      "learning_rate": 2.2996096419320824e-06,
+      "loss": 0.9194,
+      "step": 10139
+    },
+    {
+      "epoch": 0.9060245269953314,
+      "grad_norm": 0.4241883158683777,
+      "learning_rate": 2.2952735826938576e-06,
+      "loss": 0.8904,
+      "step": 10140
+    },
+    {
+      "epoch": 0.9061138785265932,
+      "grad_norm": 0.45826974511146545,
+      "learning_rate": 2.290941519234968e-06,
+      "loss": 0.9604,
+      "step": 10141
+    },
+    {
+      "epoch": 0.9062032300578551,
+      "grad_norm": 0.4920784533023834,
+      "learning_rate": 2.286613451918268e-06,
+      "loss": 0.9322,
+      "step": 10142
+    },
+    {
+      "epoch": 0.906292581589117,
+      "grad_norm": 0.4059630334377289,
+      "learning_rate": 2.2822893811062786e-06,
+      "loss": 0.9157,
+      "step": 10143
+    },
+    {
+      "epoch": 0.9063819331203788,
+      "grad_norm": 0.4405556321144104,
+      "learning_rate": 2.2779693071611986e-06,
+      "loss": 0.9764,
+      "step": 10144
+    },
+    {
+      "epoch": 0.9064712846516407,
+      "grad_norm": 0.4672204852104187,
+      "learning_rate": 2.273653230444861e-06,
+      "loss": 0.9871,
+      "step": 10145
+    },
+    {
+      "epoch": 0.9065606361829026,
+      "grad_norm": 0.5110070109367371,
+      "learning_rate": 2.269341151318788e-06,
+      "loss": 0.9669,
+      "step": 10146
+    },
+    {
+      "epoch": 0.9066499877141645,
+      "grad_norm": 0.6213862299919128,
+      "learning_rate": 2.2650330701441678e-06,
+      "loss": 0.9349,
+      "step": 10147
+    },
+    {
+      "epoch": 0.9067393392454263,
+      "grad_norm": 0.5088279247283936,
+      "learning_rate": 2.260728987281846e-06,
+      "loss": 0.895,
+      "step": 10148
+    },
+    {
+      "epoch": 0.9068286907766882,
+      "grad_norm": 0.5393834710121155,
+      "learning_rate": 2.2564289030923393e-06,
+      "loss": 0.9513,
+      "step": 10149
+    },
+    {
+      "epoch": 0.90691804230795,
+      "grad_norm": 0.4829711318016052,
+      "learning_rate": 2.2521328179358146e-06,
+      "loss": 0.8979,
+      "step": 10150
+    },
+    {
+      "epoch": 0.9070073938392119,
+      "grad_norm": 0.5531469583511353,
+      "learning_rate": 2.2478407321721296e-06,
+      "loss": 0.8851,
+      "step": 10151
+    },
+    {
+      "epoch": 0.9070967453704738,
+      "grad_norm": 0.45315778255462646,
+      "learning_rate": 2.243552646160779e-06,
+      "loss": 0.9384,
+      "step": 10152
+    },
+    {
+      "epoch": 0.9071860969017357,
+      "grad_norm": 0.5223442912101746,
+      "learning_rate": 2.239268560260943e-06,
+      "loss": 0.8615,
+      "step": 10153
+    },
+    {
+      "epoch": 0.9072754484329976,
+      "grad_norm": 0.47916871309280396,
+      "learning_rate": 2.234988474831462e-06,
+      "loss": 0.9782,
+      "step": 10154
+    },
+    {
+      "epoch": 0.9073647999642593,
+      "grad_norm": 0.4964684247970581,
+      "learning_rate": 2.230712390230838e-06,
+      "loss": 0.8816,
+      "step": 10155
+    },
+    {
+      "epoch": 0.9074541514955212,
+      "grad_norm": 0.5099236965179443,
+      "learning_rate": 2.226440306817229e-06,
+      "loss": 0.9071,
+      "step": 10156
+    },
+    {
+      "epoch": 0.9075435030267831,
+      "grad_norm": 0.48252931237220764,
+      "learning_rate": 2.222172224948471e-06,
+      "loss": 0.9396,
+      "step": 10157
+    },
+    {
+      "epoch": 0.907632854558045,
+      "grad_norm": 0.5542041063308716,
+      "learning_rate": 2.2179081449820672e-06,
+      "loss": 0.901,
+      "step": 10158
+    },
+    {
+      "epoch": 0.9077222060893069,
+      "grad_norm": 0.5302069783210754,
+      "learning_rate": 2.213648067275176e-06,
+      "loss": 0.8758,
+      "step": 10159
+    },
+    {
+      "epoch": 0.9078115576205688,
+      "grad_norm": 0.6605767607688904,
+      "learning_rate": 2.2093919921846283e-06,
+      "loss": 0.8491,
+      "step": 10160
+    },
+    {
+      "epoch": 0.9079009091518306,
+      "grad_norm": 0.47050368785858154,
+      "learning_rate": 2.2051399200669065e-06,
+      "loss": 0.9554,
+      "step": 10161
+    },
+    {
+      "epoch": 0.9079902606830924,
+      "grad_norm": 0.49454933404922485,
+      "learning_rate": 2.200891851278175e-06,
+      "loss": 0.8927,
+      "step": 10162
+    },
+    {
+      "epoch": 0.9080796122143543,
+      "grad_norm": 0.46944358944892883,
+      "learning_rate": 2.1966477861742607e-06,
+      "loss": 0.9581,
+      "step": 10163
+    },
+    {
+      "epoch": 0.9081689637456162,
+      "grad_norm": 0.48867207765579224,
+      "learning_rate": 2.1924077251106347e-06,
+      "loss": 1.0078,
+      "step": 10164
+    },
+    {
+      "epoch": 0.9082583152768781,
+      "grad_norm": 0.46018025279045105,
+      "learning_rate": 2.1881716684424568e-06,
+      "loss": 0.9385,
+      "step": 10165
+    },
+    {
+      "epoch": 0.9083476668081399,
+      "grad_norm": 0.4881027340888977,
+      "learning_rate": 2.183939616524533e-06,
+      "loss": 0.9308,
+      "step": 10166
+    },
+    {
+      "epoch": 0.9084370183394018,
+      "grad_norm": 0.4833429157733917,
+      "learning_rate": 2.1797115697113624e-06,
+      "loss": 1.0144,
+      "step": 10167
+    },
+    {
+      "epoch": 0.9085263698706637,
+      "grad_norm": 0.42278674244880676,
+      "learning_rate": 2.175487528357062e-06,
+      "loss": 0.9192,
+      "step": 10168
+    },
+    {
+      "epoch": 0.9086157214019255,
+      "grad_norm": 0.632034957408905,
+      "learning_rate": 2.17126749281546e-06,
+      "loss": 0.9533,
+      "step": 10169
+    },
+    {
+      "epoch": 0.9087050729331874,
+      "grad_norm": 0.5221367478370667,
+      "learning_rate": 2.1670514634400173e-06,
+      "loss": 0.9187,
+      "step": 10170
+    },
+    {
+      "epoch": 0.9087944244644492,
+      "grad_norm": 0.4272196292877197,
+      "learning_rate": 2.1628394405838803e-06,
+      "loss": 1.0073,
+      "step": 10171
+    },
+    {
+      "epoch": 0.9088837759957111,
+      "grad_norm": 0.44251886010169983,
+      "learning_rate": 2.1586314245998497e-06,
+      "loss": 0.9993,
+      "step": 10172
+    },
+    {
+      "epoch": 0.908973127526973,
+      "grad_norm": 0.6176732778549194,
+      "learning_rate": 2.1544274158403877e-06,
+      "loss": 0.8595,
+      "step": 10173
+    },
+    {
+      "epoch": 0.9090624790582349,
+      "grad_norm": 0.4189254343509674,
+      "learning_rate": 2.150227414657624e-06,
+      "loss": 0.9154,
+      "step": 10174
+    },
+    {
+      "epoch": 0.9091518305894968,
+      "grad_norm": 0.47814953327178955,
+      "learning_rate": 2.1460314214033662e-06,
+      "loss": 0.94,
+      "step": 10175
+    },
+    {
+      "epoch": 0.9092411821207586,
+      "grad_norm": 0.4514091908931732,
+      "learning_rate": 2.141839436429055e-06,
+      "loss": 0.9883,
+      "step": 10176
+    },
+    {
+      "epoch": 0.9093305336520204,
+      "grad_norm": 0.38325291872024536,
+      "learning_rate": 2.137651460085821e-06,
+      "loss": 0.9605,
+      "step": 10177
+    },
+    {
+      "epoch": 0.9094198851832823,
+      "grad_norm": 0.4543987512588501,
+      "learning_rate": 2.1334674927244556e-06,
+      "loss": 0.9641,
+      "step": 10178
+    },
+    {
+      "epoch": 0.9095092367145442,
+      "grad_norm": 0.4916648268699646,
+      "learning_rate": 2.1292875346954123e-06,
+      "loss": 0.9521,
+      "step": 10179
+    },
+    {
+      "epoch": 0.9095985882458061,
+      "grad_norm": 0.49808797240257263,
+      "learning_rate": 2.1251115863487934e-06,
+      "loss": 0.9676,
+      "step": 10180
+    },
+    {
+      "epoch": 0.909687939777068,
+      "grad_norm": 0.4320428669452667,
+      "learning_rate": 2.1209396480343977e-06,
+      "loss": 0.9669,
+      "step": 10181
+    },
+    {
+      "epoch": 0.9097772913083298,
+      "grad_norm": 0.5682074427604675,
+      "learning_rate": 2.1167717201016568e-06,
+      "loss": 0.8171,
+      "step": 10182
+    },
+    {
+      "epoch": 0.9098666428395916,
+      "grad_norm": 0.5299460291862488,
+      "learning_rate": 2.11260780289968e-06,
+      "loss": 0.9178,
+      "step": 10183
+    },
+    {
+      "epoch": 0.9099559943708535,
+      "grad_norm": 0.4480501711368561,
+      "learning_rate": 2.1084478967772494e-06,
+      "loss": 0.9736,
+      "step": 10184
+    },
+    {
+      "epoch": 0.9100453459021154,
+      "grad_norm": 0.5181623697280884,
+      "learning_rate": 2.1042920020827974e-06,
+      "loss": 0.9647,
+      "step": 10185
+    },
+    {
+      "epoch": 0.9101346974333773,
+      "grad_norm": 0.6131975650787354,
+      "learning_rate": 2.100140119164412e-06,
+      "loss": 0.9173,
+      "step": 10186
+    },
+    {
+      "epoch": 0.9102240489646392,
+      "grad_norm": 0.5780476927757263,
+      "learning_rate": 2.095992248369871e-06,
+      "loss": 0.9846,
+      "step": 10187
+    },
+    {
+      "epoch": 0.910313400495901,
+      "grad_norm": 0.4459938108921051,
+      "learning_rate": 2.0918483900466025e-06,
+      "loss": 0.9798,
+      "step": 10188
+    },
+    {
+      "epoch": 0.9104027520271629,
+      "grad_norm": 0.47280803322792053,
+      "learning_rate": 2.087708544541689e-06,
+      "loss": 0.9052,
+      "step": 10189
+    },
+    {
+      "epoch": 0.9104921035584247,
+      "grad_norm": 0.5585520267486572,
+      "learning_rate": 2.083572712201898e-06,
+      "loss": 0.8532,
+      "step": 10190
+    },
+    {
+      "epoch": 0.9105814550896866,
+      "grad_norm": 0.5118190050125122,
+      "learning_rate": 2.079440893373641e-06,
+      "loss": 0.9004,
+      "step": 10191
+    },
+    {
+      "epoch": 0.9106708066209485,
+      "grad_norm": 0.503633975982666,
+      "learning_rate": 2.075313088403008e-06,
+      "loss": 0.8772,
+      "step": 10192
+    },
+    {
+      "epoch": 0.9107601581522103,
+      "grad_norm": 0.45112094283103943,
+      "learning_rate": 2.071189297635745e-06,
+      "loss": 0.9163,
+      "step": 10193
+    },
+    {
+      "epoch": 0.9108495096834722,
+      "grad_norm": 0.4446631669998169,
+      "learning_rate": 2.067069521417264e-06,
+      "loss": 0.9806,
+      "step": 10194
+    },
+    {
+      "epoch": 0.9109388612147341,
+      "grad_norm": 0.525492250919342,
+      "learning_rate": 2.0629537600926395e-06,
+      "loss": 0.9365,
+      "step": 10195
+    },
+    {
+      "epoch": 0.9110282127459959,
+      "grad_norm": 0.4737085998058319,
+      "learning_rate": 2.0588420140066067e-06,
+      "loss": 1.0117,
+      "step": 10196
+    },
+    {
+      "epoch": 0.9111175642772578,
+      "grad_norm": 0.4597386121749878,
+      "learning_rate": 2.0547342835035733e-06,
+      "loss": 0.896,
+      "step": 10197
+    },
+    {
+      "epoch": 0.9112069158085196,
+      "grad_norm": 0.5008763074874878,
+      "learning_rate": 2.050630568927603e-06,
+      "loss": 0.9568,
+      "step": 10198
+    },
+    {
+      "epoch": 0.9112962673397815,
+      "grad_norm": 0.5139524340629578,
+      "learning_rate": 2.0465308706224207e-06,
+      "loss": 0.8624,
+      "step": 10199
+    },
+    {
+      "epoch": 0.9113856188710434,
+      "grad_norm": 0.5094760060310364,
+      "learning_rate": 2.0424351889314354e-06,
+      "loss": 0.9532,
+      "step": 10200
+    },
+    {
+      "epoch": 0.9114749704023053,
+      "grad_norm": 0.5391070246696472,
+      "learning_rate": 2.038343524197689e-06,
+      "loss": 0.8551,
+      "step": 10201
+    },
+    {
+      "epoch": 0.9115643219335672,
+      "grad_norm": 0.5024666786193848,
+      "learning_rate": 2.0342558767639074e-06,
+      "loss": 0.8772,
+      "step": 10202
+    },
+    {
+      "epoch": 0.911653673464829,
+      "grad_norm": 0.48186391592025757,
+      "learning_rate": 2.0301722469724726e-06,
+      "loss": 0.875,
+      "step": 10203
+    },
+    {
+      "epoch": 0.9117430249960908,
+      "grad_norm": 0.44918540120124817,
+      "learning_rate": 2.026092635165433e-06,
+      "loss": 0.9593,
+      "step": 10204
+    },
+    {
+      "epoch": 0.9118323765273527,
+      "grad_norm": 0.48004093766212463,
+      "learning_rate": 2.022017041684504e-06,
+      "loss": 0.9952,
+      "step": 10205
+    },
+    {
+      "epoch": 0.9119217280586146,
+      "grad_norm": 0.4600532352924347,
+      "learning_rate": 2.0179454668710575e-06,
+      "loss": 0.9583,
+      "step": 10206
+    },
+    {
+      "epoch": 0.9120110795898765,
+      "grad_norm": 0.4377675950527191,
+      "learning_rate": 2.0138779110661252e-06,
+      "loss": 0.8938,
+      "step": 10207
+    },
+    {
+      "epoch": 0.9121004311211384,
+      "grad_norm": 0.6013002395629883,
+      "learning_rate": 2.0098143746104135e-06,
+      "loss": 0.7857,
+      "step": 10208
+    },
+    {
+      "epoch": 0.9121897826524002,
+      "grad_norm": 0.5890505313873291,
+      "learning_rate": 2.005754857844283e-06,
+      "loss": 0.9102,
+      "step": 10209
+    },
+    {
+      "epoch": 0.912279134183662,
+      "grad_norm": 0.48680880665779114,
+      "learning_rate": 2.0016993611077726e-06,
+      "loss": 0.944,
+      "step": 10210
+    },
+    {
+      "epoch": 0.9123684857149239,
+      "grad_norm": 0.5501675009727478,
+      "learning_rate": 1.997647884740561e-06,
+      "loss": 0.872,
+      "step": 10211
+    },
+    {
+      "epoch": 0.9124578372461858,
+      "grad_norm": 0.46747952699661255,
+      "learning_rate": 1.9936004290820098e-06,
+      "loss": 0.9101,
+      "step": 10212
+    },
+    {
+      "epoch": 0.9125471887774477,
+      "grad_norm": 0.47848084568977356,
+      "learning_rate": 1.989556994471131e-06,
+      "loss": 0.9689,
+      "step": 10213
+    },
+    {
+      "epoch": 0.9126365403087096,
+      "grad_norm": 0.42791107296943665,
+      "learning_rate": 1.98551758124661e-06,
+      "loss": 0.9226,
+      "step": 10214
+    },
+    {
+      "epoch": 0.9127258918399714,
+      "grad_norm": 0.6010297536849976,
+      "learning_rate": 1.9814821897467973e-06,
+      "loss": 0.8729,
+      "step": 10215
+    },
+    {
+      "epoch": 0.9128152433712333,
+      "grad_norm": 0.5505849719047546,
+      "learning_rate": 1.977450820309684e-06,
+      "loss": 0.9354,
+      "step": 10216
+    },
+    {
+      "epoch": 0.9129045949024951,
+      "grad_norm": 0.7786350250244141,
+      "learning_rate": 1.973423473272945e-06,
+      "loss": 0.8643,
+      "step": 10217
+    },
+    {
+      "epoch": 0.912993946433757,
+      "grad_norm": 0.596895694732666,
+      "learning_rate": 1.9694001489739213e-06,
+      "loss": 0.9079,
+      "step": 10218
+    },
+    {
+      "epoch": 0.9130832979650189,
+      "grad_norm": 0.4010644555091858,
+      "learning_rate": 1.9653808477496038e-06,
+      "loss": 0.9816,
+      "step": 10219
+    },
+    {
+      "epoch": 0.9131726494962807,
+      "grad_norm": 0.537714421749115,
+      "learning_rate": 1.9613655699366464e-06,
+      "loss": 0.9397,
+      "step": 10220
+    },
+    {
+      "epoch": 0.9132620010275426,
+      "grad_norm": 0.4720018804073334,
+      "learning_rate": 1.957354315871385e-06,
+      "loss": 0.9971,
+      "step": 10221
+    },
+    {
+      "epoch": 0.9133513525588045,
+      "grad_norm": 0.45322003960609436,
+      "learning_rate": 1.953347085889795e-06,
+      "loss": 0.9162,
+      "step": 10222
+    },
+    {
+      "epoch": 0.9134407040900664,
+      "grad_norm": 0.5041544437408447,
+      "learning_rate": 1.9493438803275257e-06,
+      "loss": 0.9266,
+      "step": 10223
+    },
+    {
+      "epoch": 0.9135300556213282,
+      "grad_norm": 0.5631766319274902,
+      "learning_rate": 1.945344699519891e-06,
+      "loss": 0.896,
+      "step": 10224
+    },
+    {
+      "epoch": 0.91361940715259,
+      "grad_norm": 0.5237100124359131,
+      "learning_rate": 1.9413495438018736e-06,
+      "loss": 0.9955,
+      "step": 10225
+    },
+    {
+      "epoch": 0.9137087586838519,
+      "grad_norm": 0.5129988193511963,
+      "learning_rate": 1.937358413508089e-06,
+      "loss": 0.9253,
+      "step": 10226
+    },
+    {
+      "epoch": 0.9137981102151138,
+      "grad_norm": 0.45968538522720337,
+      "learning_rate": 1.933371308972848e-06,
+      "loss": 0.9623,
+      "step": 10227
+    },
+    {
+      "epoch": 0.9138874617463757,
+      "grad_norm": 0.5253444314002991,
+      "learning_rate": 1.929388230530116e-06,
+      "loss": 1.0036,
+      "step": 10228
+    },
+    {
+      "epoch": 0.9139768132776376,
+      "grad_norm": 0.5315740704536438,
+      "learning_rate": 1.9254091785135153e-06,
+      "loss": 0.89,
+      "step": 10229
+    },
+    {
+      "epoch": 0.9140661648088995,
+      "grad_norm": 0.4819380044937134,
+      "learning_rate": 1.9214341532563296e-06,
+      "loss": 0.936,
+      "step": 10230
+    },
+    {
+      "epoch": 0.9141555163401612,
+      "grad_norm": 0.5596315264701843,
+      "learning_rate": 1.91746315509152e-06,
+      "loss": 0.8871,
+      "step": 10231
+    },
+    {
+      "epoch": 0.9142448678714231,
+      "grad_norm": 0.4711841344833374,
+      "learning_rate": 1.913496184351693e-06,
+      "loss": 0.9442,
+      "step": 10232
+    },
+    {
+      "epoch": 0.914334219402685,
+      "grad_norm": 0.4977385997772217,
+      "learning_rate": 1.9095332413691326e-06,
+      "loss": 0.9563,
+      "step": 10233
+    },
+    {
+      "epoch": 0.9144235709339469,
+      "grad_norm": 0.592835545539856,
+      "learning_rate": 1.905574326475762e-06,
+      "loss": 0.8969,
+      "step": 10234
+    },
+    {
+      "epoch": 0.9145129224652088,
+      "grad_norm": 0.47582802176475525,
+      "learning_rate": 1.9016194400031884e-06,
+      "loss": 0.9354,
+      "step": 10235
+    },
+    {
+      "epoch": 0.9146022739964706,
+      "grad_norm": 0.4953663647174835,
+      "learning_rate": 1.8976685822826856e-06,
+      "loss": 0.9497,
+      "step": 10236
+    },
+    {
+      "epoch": 0.9146916255277325,
+      "grad_norm": 0.49911701679229736,
+      "learning_rate": 1.8937217536451778e-06,
+      "loss": 0.9374,
+      "step": 10237
+    },
+    {
+      "epoch": 0.9147809770589943,
+      "grad_norm": 0.5435790419578552,
+      "learning_rate": 1.8897789544212396e-06,
+      "loss": 0.8251,
+      "step": 10238
+    },
+    {
+      "epoch": 0.9148703285902562,
+      "grad_norm": 0.5200591683387756,
+      "learning_rate": 1.8858401849411344e-06,
+      "loss": 0.9244,
+      "step": 10239
+    },
+    {
+      "epoch": 0.9149596801215181,
+      "grad_norm": 0.6190335750579834,
+      "learning_rate": 1.8819054455347707e-06,
+      "loss": 0.8485,
+      "step": 10240
+    },
+    {
+      "epoch": 0.91504903165278,
+      "grad_norm": 0.5141425132751465,
+      "learning_rate": 1.877974736531729e-06,
+      "loss": 0.946,
+      "step": 10241
+    },
+    {
+      "epoch": 0.9151383831840418,
+      "grad_norm": 0.5247118473052979,
+      "learning_rate": 1.8740480582612519e-06,
+      "loss": 0.942,
+      "step": 10242
+    },
+    {
+      "epoch": 0.9152277347153037,
+      "grad_norm": 0.5821022391319275,
+      "learning_rate": 1.8701254110522315e-06,
+      "loss": 0.9661,
+      "step": 10243
+    },
+    {
+      "epoch": 0.9153170862465656,
+      "grad_norm": 0.6091264486312866,
+      "learning_rate": 1.8662067952332386e-06,
+      "loss": 0.9005,
+      "step": 10244
+    },
+    {
+      "epoch": 0.9154064377778274,
+      "grad_norm": 0.44241493940353394,
+      "learning_rate": 1.8622922111324937e-06,
+      "loss": 0.9736,
+      "step": 10245
+    },
+    {
+      "epoch": 0.9154957893090893,
+      "grad_norm": 0.4991171360015869,
+      "learning_rate": 1.8583816590778901e-06,
+      "loss": 0.9084,
+      "step": 10246
+    },
+    {
+      "epoch": 0.9155851408403511,
+      "grad_norm": 0.6471422910690308,
+      "learning_rate": 1.8544751393969716e-06,
+      "loss": 0.8957,
+      "step": 10247
+    },
+    {
+      "epoch": 0.915674492371613,
+      "grad_norm": 0.5588064193725586,
+      "learning_rate": 1.8505726524169598e-06,
+      "loss": 0.8469,
+      "step": 10248
+    },
+    {
+      "epoch": 0.9157638439028749,
+      "grad_norm": 0.4519459307193756,
+      "learning_rate": 1.8466741984647151e-06,
+      "loss": 0.9619,
+      "step": 10249
+    },
+    {
+      "epoch": 0.9158531954341368,
+      "grad_norm": 0.5813220739364624,
+      "learning_rate": 1.8427797778667931e-06,
+      "loss": 0.9573,
+      "step": 10250
+    },
+    {
+      "epoch": 0.9159425469653987,
+      "grad_norm": 0.4593013823032379,
+      "learning_rate": 1.8388893909493775e-06,
+      "loss": 1.0341,
+      "step": 10251
+    },
+    {
+      "epoch": 0.9160318984966604,
+      "grad_norm": 0.45809662342071533,
+      "learning_rate": 1.8350030380383355e-06,
+      "loss": 0.957,
+      "step": 10252
+    },
+    {
+      "epoch": 0.9161212500279223,
+      "grad_norm": 0.58559250831604,
+      "learning_rate": 1.831120719459195e-06,
+      "loss": 0.8983,
+      "step": 10253
+    },
+    {
+      "epoch": 0.9162106015591842,
+      "grad_norm": 0.4311509132385254,
+      "learning_rate": 1.8272424355371353e-06,
+      "loss": 0.903,
+      "step": 10254
+    },
+    {
+      "epoch": 0.9162999530904461,
+      "grad_norm": 0.49322575330734253,
+      "learning_rate": 1.8233681865970077e-06,
+      "loss": 0.937,
+      "step": 10255
+    },
+    {
+      "epoch": 0.916389304621708,
+      "grad_norm": 0.4198033809661865,
+      "learning_rate": 1.8194979729633244e-06,
+      "loss": 1.0386,
+      "step": 10256
+    },
+    {
+      "epoch": 0.9164786561529699,
+      "grad_norm": 0.4912882149219513,
+      "learning_rate": 1.8156317949602486e-06,
+      "loss": 0.9599,
+      "step": 10257
+    },
+    {
+      "epoch": 0.9165680076842316,
+      "grad_norm": 0.42434144020080566,
+      "learning_rate": 1.8117696529116213e-06,
+      "loss": 0.943,
+      "step": 10258
+    },
+    {
+      "epoch": 0.9166573592154935,
+      "grad_norm": 0.5602464079856873,
+      "learning_rate": 1.807911547140928e-06,
+      "loss": 0.851,
+      "step": 10259
+    },
+    {
+      "epoch": 0.9167467107467554,
+      "grad_norm": 0.4524661898612976,
+      "learning_rate": 1.8040574779713382e-06,
+      "loss": 1.0004,
+      "step": 10260
+    },
+    {
+      "epoch": 0.9168360622780173,
+      "grad_norm": 0.5537035465240479,
+      "learning_rate": 1.8002074457256658e-06,
+      "loss": 0.9157,
+      "step": 10261
+    },
+    {
+      "epoch": 0.9169254138092792,
+      "grad_norm": 0.48981863260269165,
+      "learning_rate": 1.796361450726397e-06,
+      "loss": 0.9785,
+      "step": 10262
+    },
+    {
+      "epoch": 0.917014765340541,
+      "grad_norm": 0.6159743070602417,
+      "learning_rate": 1.7925194932956635e-06,
+      "loss": 0.9111,
+      "step": 10263
+    },
+    {
+      "epoch": 0.9171041168718029,
+      "grad_norm": 0.40217313170433044,
+      "learning_rate": 1.7886815737552797e-06,
+      "loss": 0.9608,
+      "step": 10264
+    },
+    {
+      "epoch": 0.9171934684030647,
+      "grad_norm": 0.4525063633918762,
+      "learning_rate": 1.784847692426711e-06,
+      "loss": 1.0066,
+      "step": 10265
+    },
+    {
+      "epoch": 0.9172828199343266,
+      "grad_norm": 0.6385212540626526,
+      "learning_rate": 1.7810178496310891e-06,
+      "loss": 0.8665,
+      "step": 10266
+    },
+    {
+      "epoch": 0.9173721714655885,
+      "grad_norm": 0.4889377951622009,
+      "learning_rate": 1.7771920456891966e-06,
+      "loss": 0.9201,
+      "step": 10267
+    },
+    {
+      "epoch": 0.9174615229968504,
+      "grad_norm": 0.5491172671318054,
+      "learning_rate": 1.7733702809214825e-06,
+      "loss": 0.8458,
+      "step": 10268
+    },
+    {
+      "epoch": 0.9175508745281122,
+      "grad_norm": 0.48966601490974426,
+      "learning_rate": 1.7695525556480686e-06,
+      "loss": 0.9615,
+      "step": 10269
+    },
+    {
+      "epoch": 0.9176402260593741,
+      "grad_norm": 0.5428771376609802,
+      "learning_rate": 1.7657388701887379e-06,
+      "loss": 0.9203,
+      "step": 10270
+    },
+    {
+      "epoch": 0.917729577590636,
+      "grad_norm": 0.45766574144363403,
+      "learning_rate": 1.7619292248629071e-06,
+      "loss": 0.9307,
+      "step": 10271
+    },
+    {
+      "epoch": 0.9178189291218978,
+      "grad_norm": 0.49701622128486633,
+      "learning_rate": 1.7581236199896879e-06,
+      "loss": 0.868,
+      "step": 10272
+    },
+    {
+      "epoch": 0.9179082806531597,
+      "grad_norm": 0.4536900520324707,
+      "learning_rate": 1.754322055887836e-06,
+      "loss": 0.9504,
+      "step": 10273
+    },
+    {
+      "epoch": 0.9179976321844215,
+      "grad_norm": 0.5021615028381348,
+      "learning_rate": 1.750524532875769e-06,
+      "loss": 0.9365,
+      "step": 10274
+    },
+    {
+      "epoch": 0.9180869837156834,
+      "grad_norm": 0.4921415448188782,
+      "learning_rate": 1.7467310512715774e-06,
+      "loss": 0.966,
+      "step": 10275
+    },
+    {
+      "epoch": 0.9181763352469453,
+      "grad_norm": 0.5077376961708069,
+      "learning_rate": 1.742941611393012e-06,
+      "loss": 0.9469,
+      "step": 10276
+    },
+    {
+      "epoch": 0.9182656867782072,
+      "grad_norm": 0.45509567856788635,
+      "learning_rate": 1.7391562135574634e-06,
+      "loss": 0.9654,
+      "step": 10277
+    },
+    {
+      "epoch": 0.9183550383094691,
+      "grad_norm": 0.4772971570491791,
+      "learning_rate": 1.7353748580820061e-06,
+      "loss": 0.9162,
+      "step": 10278
+    },
+    {
+      "epoch": 0.9184443898407308,
+      "grad_norm": 0.5425154566764832,
+      "learning_rate": 1.7315975452833645e-06,
+      "loss": 0.9155,
+      "step": 10279
+    },
+    {
+      "epoch": 0.9185337413719927,
+      "grad_norm": 0.5885468125343323,
+      "learning_rate": 1.727824275477935e-06,
+      "loss": 0.9321,
+      "step": 10280
+    },
+    {
+      "epoch": 0.9186230929032546,
+      "grad_norm": 0.41930609941482544,
+      "learning_rate": 1.7240550489817653e-06,
+      "loss": 0.9593,
+      "step": 10281
+    },
+    {
+      "epoch": 0.9187124444345165,
+      "grad_norm": 0.5219174027442932,
+      "learning_rate": 1.7202898661105748e-06,
+      "loss": 0.9323,
+      "step": 10282
+    },
+    {
+      "epoch": 0.9188017959657784,
+      "grad_norm": 0.5876883268356323,
+      "learning_rate": 1.716528727179728e-06,
+      "loss": 0.9279,
+      "step": 10283
+    },
+    {
+      "epoch": 0.9188911474970403,
+      "grad_norm": 0.5892945528030396,
+      "learning_rate": 1.712771632504262e-06,
+      "loss": 0.8964,
+      "step": 10284
+    },
+    {
+      "epoch": 0.9189804990283021,
+      "grad_norm": 0.4626266062259674,
+      "learning_rate": 1.7090185823988857e-06,
+      "loss": 0.907,
+      "step": 10285
+    },
+    {
+      "epoch": 0.9190698505595639,
+      "grad_norm": 0.6098506450653076,
+      "learning_rate": 1.7052695771779481e-06,
+      "loss": 0.8302,
+      "step": 10286
+    },
+    {
+      "epoch": 0.9191592020908258,
+      "grad_norm": 0.5348356366157532,
+      "learning_rate": 1.7015246171554644e-06,
+      "loss": 0.9675,
+      "step": 10287
+    },
+    {
+      "epoch": 0.9192485536220877,
+      "grad_norm": 0.46181049942970276,
+      "learning_rate": 1.697783702645117e-06,
+      "loss": 0.849,
+      "step": 10288
+    },
+    {
+      "epoch": 0.9193379051533496,
+      "grad_norm": 0.5558209419250488,
+      "learning_rate": 1.69404683396025e-06,
+      "loss": 0.9733,
+      "step": 10289
+    },
+    {
+      "epoch": 0.9194272566846114,
+      "grad_norm": 0.48256248235702515,
+      "learning_rate": 1.6903140114138627e-06,
+      "loss": 0.948,
+      "step": 10290
+    },
+    {
+      "epoch": 0.9195166082158733,
+      "grad_norm": 0.6082019805908203,
+      "learning_rate": 1.6865852353186218e-06,
+      "loss": 0.9981,
+      "step": 10291
+    },
+    {
+      "epoch": 0.9196059597471352,
+      "grad_norm": 0.4192001223564148,
+      "learning_rate": 1.6828605059868552e-06,
+      "loss": 1.004,
+      "step": 10292
+    },
+    {
+      "epoch": 0.919695311278397,
+      "grad_norm": 0.5222572684288025,
+      "learning_rate": 1.6791398237305412e-06,
+      "loss": 0.9626,
+      "step": 10293
+    },
+    {
+      "epoch": 0.9197846628096589,
+      "grad_norm": 0.48702099919319153,
+      "learning_rate": 1.6754231888613304e-06,
+      "loss": 0.9927,
+      "step": 10294
+    },
+    {
+      "epoch": 0.9198740143409208,
+      "grad_norm": 0.5022723078727722,
+      "learning_rate": 1.6717106016905348e-06,
+      "loss": 0.9374,
+      "step": 10295
+    },
+    {
+      "epoch": 0.9199633658721826,
+      "grad_norm": 0.3965294063091278,
+      "learning_rate": 1.6680020625291227e-06,
+      "loss": 0.9798,
+      "step": 10296
+    },
+    {
+      "epoch": 0.9200527174034445,
+      "grad_norm": 0.48680707812309265,
+      "learning_rate": 1.6642975716877118e-06,
+      "loss": 0.9358,
+      "step": 10297
+    },
+    {
+      "epoch": 0.9201420689347064,
+      "grad_norm": 0.5684248805046082,
+      "learning_rate": 1.6605971294766044e-06,
+      "loss": 0.848,
+      "step": 10298
+    },
+    {
+      "epoch": 0.9202314204659683,
+      "grad_norm": 0.5401350855827332,
+      "learning_rate": 1.6569007362057465e-06,
+      "loss": 0.9419,
+      "step": 10299
+    },
+    {
+      "epoch": 0.9203207719972301,
+      "grad_norm": 0.5058586001396179,
+      "learning_rate": 1.653208392184752e-06,
+      "loss": 0.877,
+      "step": 10300
+    },
+    {
+      "epoch": 0.9204101235284919,
+      "grad_norm": 0.4828777611255646,
+      "learning_rate": 1.6495200977228897e-06,
+      "loss": 0.9071,
+      "step": 10301
+    },
+    {
+      "epoch": 0.9204994750597538,
+      "grad_norm": 0.5211459994316101,
+      "learning_rate": 1.6458358531291074e-06,
+      "loss": 0.926,
+      "step": 10302
+    },
+    {
+      "epoch": 0.9205888265910157,
+      "grad_norm": 0.4530414044857025,
+      "learning_rate": 1.6421556587119913e-06,
+      "loss": 0.9432,
+      "step": 10303
+    },
+    {
+      "epoch": 0.9206781781222776,
+      "grad_norm": 0.4439500868320465,
+      "learning_rate": 1.6384795147797894e-06,
+      "loss": 0.9568,
+      "step": 10304
+    },
+    {
+      "epoch": 0.9207675296535395,
+      "grad_norm": 0.43074101209640503,
+      "learning_rate": 1.6348074216404273e-06,
+      "loss": 1.1121,
+      "step": 10305
+    },
+    {
+      "epoch": 0.9208568811848014,
+      "grad_norm": 0.5717933177947998,
+      "learning_rate": 1.6311393796014819e-06,
+      "loss": 0.8781,
+      "step": 10306
+    },
+    {
+      "epoch": 0.9209462327160631,
+      "grad_norm": 0.4109342396259308,
+      "learning_rate": 1.62747538897019e-06,
+      "loss": 0.9698,
+      "step": 10307
+    },
+    {
+      "epoch": 0.921035584247325,
+      "grad_norm": 0.4915206730365753,
+      "learning_rate": 1.6238154500534452e-06,
+      "loss": 0.9776,
+      "step": 10308
+    },
+    {
+      "epoch": 0.9211249357785869,
+      "grad_norm": 0.5269554853439331,
+      "learning_rate": 1.620159563157808e-06,
+      "loss": 0.8828,
+      "step": 10309
+    },
+    {
+      "epoch": 0.9212142873098488,
+      "grad_norm": 0.446888267993927,
+      "learning_rate": 1.6165077285895002e-06,
+      "loss": 0.9208,
+      "step": 10310
+    },
+    {
+      "epoch": 0.9213036388411107,
+      "grad_norm": 0.4172409772872925,
+      "learning_rate": 1.6128599466543993e-06,
+      "loss": 0.9789,
+      "step": 10311
+    },
+    {
+      "epoch": 0.9213929903723725,
+      "grad_norm": 0.5068113803863525,
+      "learning_rate": 1.6092162176580494e-06,
+      "loss": 0.8979,
+      "step": 10312
+    },
+    {
+      "epoch": 0.9214823419036344,
+      "grad_norm": 0.5210028290748596,
+      "learning_rate": 1.6055765419056456e-06,
+      "loss": 0.9481,
+      "step": 10313
+    },
+    {
+      "epoch": 0.9215716934348962,
+      "grad_norm": 0.48873379826545715,
+      "learning_rate": 1.6019409197020607e-06,
+      "loss": 0.9693,
+      "step": 10314
+    },
+    {
+      "epoch": 0.9216610449661581,
+      "grad_norm": 0.4486122727394104,
+      "learning_rate": 1.5983093513518066e-06,
+      "loss": 0.946,
+      "step": 10315
+    },
+    {
+      "epoch": 0.92175039649742,
+      "grad_norm": 0.38552728295326233,
+      "learning_rate": 1.5946818371590787e-06,
+      "loss": 0.9608,
+      "step": 10316
+    },
+    {
+      "epoch": 0.9218397480286818,
+      "grad_norm": 0.5378084182739258,
+      "learning_rate": 1.591058377427701e-06,
+      "loss": 0.9168,
+      "step": 10317
+    },
+    {
+      "epoch": 0.9219290995599437,
+      "grad_norm": 0.48840612173080444,
+      "learning_rate": 1.587438972461186e-06,
+      "loss": 0.9724,
+      "step": 10318
+    },
+    {
+      "epoch": 0.9220184510912056,
+      "grad_norm": 0.5533527135848999,
+      "learning_rate": 1.5838236225626968e-06,
+      "loss": 0.8688,
+      "step": 10319
+    },
+    {
+      "epoch": 0.9221078026224674,
+      "grad_norm": 0.6588174700737,
+      "learning_rate": 1.5802123280350633e-06,
+      "loss": 0.8956,
+      "step": 10320
+    },
+    {
+      "epoch": 0.9221971541537293,
+      "grad_norm": 0.4313110411167145,
+      "learning_rate": 1.5766050891807604e-06,
+      "loss": 0.9738,
+      "step": 10321
+    },
+    {
+      "epoch": 0.9222865056849912,
+      "grad_norm": 0.44930917024612427,
+      "learning_rate": 1.5730019063019407e-06,
+      "loss": 1.0021,
+      "step": 10322
+    },
+    {
+      "epoch": 0.922375857216253,
+      "grad_norm": 0.6756159067153931,
+      "learning_rate": 1.5694027797004073e-06,
+      "loss": 0.8297,
+      "step": 10323
+    },
+    {
+      "epoch": 0.9224652087475149,
+      "grad_norm": 0.5268403887748718,
+      "learning_rate": 1.5658077096776192e-06,
+      "loss": 0.9241,
+      "step": 10324
+    },
+    {
+      "epoch": 0.9225545602787768,
+      "grad_norm": 0.4503045082092285,
+      "learning_rate": 1.562216696534713e-06,
+      "loss": 0.8687,
+      "step": 10325
+    },
+    {
+      "epoch": 0.9226439118100387,
+      "grad_norm": 0.5141546130180359,
+      "learning_rate": 1.5586297405724648e-06,
+      "loss": 0.8619,
+      "step": 10326
+    },
+    {
+      "epoch": 0.9227332633413005,
+      "grad_norm": 0.5367921590805054,
+      "learning_rate": 1.5550468420913288e-06,
+      "loss": 0.9026,
+      "step": 10327
+    },
+    {
+      "epoch": 0.9228226148725623,
+      "grad_norm": 0.575573205947876,
+      "learning_rate": 1.5514680013913984e-06,
+      "loss": 0.897,
+      "step": 10328
+    },
+    {
+      "epoch": 0.9229119664038242,
+      "grad_norm": 0.44511479139328003,
+      "learning_rate": 1.5478932187724504e-06,
+      "loss": 0.9417,
+      "step": 10329
+    },
+    {
+      "epoch": 0.9230013179350861,
+      "grad_norm": 0.6052654981613159,
+      "learning_rate": 1.5443224945339063e-06,
+      "loss": 0.9232,
+      "step": 10330
+    },
+    {
+      "epoch": 0.923090669466348,
+      "grad_norm": 0.4598505198955536,
+      "learning_rate": 1.5407558289748547e-06,
+      "loss": 1.0025,
+      "step": 10331
+    },
+    {
+      "epoch": 0.9231800209976099,
+      "grad_norm": 0.5528203845024109,
+      "learning_rate": 1.53719322239404e-06,
+      "loss": 0.9608,
+      "step": 10332
+    },
+    {
+      "epoch": 0.9232693725288718,
+      "grad_norm": 0.47037777304649353,
+      "learning_rate": 1.5336346750898678e-06,
+      "loss": 0.9687,
+      "step": 10333
+    },
+    {
+      "epoch": 0.9233587240601335,
+      "grad_norm": 0.5850094556808472,
+      "learning_rate": 1.530080187360411e-06,
+      "loss": 0.9053,
+      "step": 10334
+    },
+    {
+      "epoch": 0.9234480755913954,
+      "grad_norm": 0.552905797958374,
+      "learning_rate": 1.5265297595033868e-06,
+      "loss": 0.9073,
+      "step": 10335
+    },
+    {
+      "epoch": 0.9235374271226573,
+      "grad_norm": 0.5310145020484924,
+      "learning_rate": 1.5229833918161906e-06,
+      "loss": 0.9366,
+      "step": 10336
+    },
+    {
+      "epoch": 0.9236267786539192,
+      "grad_norm": 0.4303600490093231,
+      "learning_rate": 1.5194410845958574e-06,
+      "loss": 0.9336,
+      "step": 10337
+    },
+    {
+      "epoch": 0.9237161301851811,
+      "grad_norm": 0.44541823863983154,
+      "learning_rate": 1.5159028381390994e-06,
+      "loss": 0.9732,
+      "step": 10338
+    },
+    {
+      "epoch": 0.9238054817164429,
+      "grad_norm": 0.6362563967704773,
+      "learning_rate": 1.5123686527422854e-06,
+      "loss": 0.9495,
+      "step": 10339
+    },
+    {
+      "epoch": 0.9238948332477048,
+      "grad_norm": 0.4896201491355896,
+      "learning_rate": 1.5088385287014395e-06,
+      "loss": 0.8939,
+      "step": 10340
+    },
+    {
+      "epoch": 0.9239841847789666,
+      "grad_norm": 0.5114855766296387,
+      "learning_rate": 1.5053124663122419e-06,
+      "loss": 0.8782,
+      "step": 10341
+    },
+    {
+      "epoch": 0.9240735363102285,
+      "grad_norm": 0.4664299190044403,
+      "learning_rate": 1.5017904658700398e-06,
+      "loss": 0.9548,
+      "step": 10342
+    },
+    {
+      "epoch": 0.9241628878414904,
+      "grad_norm": 0.5291177034378052,
+      "learning_rate": 1.4982725276698418e-06,
+      "loss": 0.8842,
+      "step": 10343
+    },
+    {
+      "epoch": 0.9242522393727522,
+      "grad_norm": 0.5422632694244385,
+      "learning_rate": 1.4947586520063062e-06,
+      "loss": 0.8602,
+      "step": 10344
+    },
+    {
+      "epoch": 0.9243415909040141,
+      "grad_norm": 0.6073907613754272,
+      "learning_rate": 1.49124883917377e-06,
+      "loss": 0.8669,
+      "step": 10345
+    },
+    {
+      "epoch": 0.924430942435276,
+      "grad_norm": 0.6418290734291077,
+      "learning_rate": 1.4877430894662036e-06,
+      "loss": 0.8869,
+      "step": 10346
+    },
+    {
+      "epoch": 0.9245202939665379,
+      "grad_norm": 0.634617030620575,
+      "learning_rate": 1.4842414031772612e-06,
+      "loss": 0.8845,
+      "step": 10347
+    },
+    {
+      "epoch": 0.9246096454977997,
+      "grad_norm": 0.5437284708023071,
+      "learning_rate": 1.480743780600241e-06,
+      "loss": 0.9179,
+      "step": 10348
+    },
+    {
+      "epoch": 0.9246989970290616,
+      "grad_norm": 0.44016095995903015,
+      "learning_rate": 1.4772502220281093e-06,
+      "loss": 0.9627,
+      "step": 10349
+    },
+    {
+      "epoch": 0.9247883485603234,
+      "grad_norm": 0.39443960785865784,
+      "learning_rate": 1.4737607277534815e-06,
+      "loss": 0.9157,
+      "step": 10350
+    },
+    {
+      "epoch": 0.9248777000915853,
+      "grad_norm": 0.4765508770942688,
+      "learning_rate": 1.4702752980686462e-06,
+      "loss": 0.973,
+      "step": 10351
+    },
+    {
+      "epoch": 0.9249670516228472,
+      "grad_norm": 0.5301365256309509,
+      "learning_rate": 1.4667939332655478e-06,
+      "loss": 0.8771,
+      "step": 10352
+    },
+    {
+      "epoch": 0.9250564031541091,
+      "grad_norm": 0.48352834582328796,
+      "learning_rate": 1.4633166336357807e-06,
+      "loss": 0.9077,
+      "step": 10353
+    },
+    {
+      "epoch": 0.925145754685371,
+      "grad_norm": 0.540468156337738,
+      "learning_rate": 1.4598433994706117e-06,
+      "loss": 0.9076,
+      "step": 10354
+    },
+    {
+      "epoch": 0.9252351062166327,
+      "grad_norm": 0.42415082454681396,
+      "learning_rate": 1.4563742310609529e-06,
+      "loss": 0.9269,
+      "step": 10355
+    },
+    {
+      "epoch": 0.9253244577478946,
+      "grad_norm": 0.4903179407119751,
+      "learning_rate": 1.4529091286973995e-06,
+      "loss": 0.933,
+      "step": 10356
+    },
+    {
+      "epoch": 0.9254138092791565,
+      "grad_norm": 0.5395703315734863,
+      "learning_rate": 1.4494480926701803e-06,
+      "loss": 0.8883,
+      "step": 10357
+    },
+    {
+      "epoch": 0.9255031608104184,
+      "grad_norm": 0.5688163042068481,
+      "learning_rate": 1.4459911232691914e-06,
+      "loss": 0.9469,
+      "step": 10358
+    },
+    {
+      "epoch": 0.9255925123416803,
+      "grad_norm": 0.41127413511276245,
+      "learning_rate": 1.4425382207839954e-06,
+      "loss": 1.0157,
+      "step": 10359
+    },
+    {
+      "epoch": 0.9256818638729422,
+      "grad_norm": 0.46226775646209717,
+      "learning_rate": 1.4390893855038057e-06,
+      "loss": 0.9535,
+      "step": 10360
+    },
+    {
+      "epoch": 0.925771215404204,
+      "grad_norm": 0.5173824429512024,
+      "learning_rate": 1.4356446177175077e-06,
+      "loss": 0.9242,
+      "step": 10361
+    },
+    {
+      "epoch": 0.9258605669354658,
+      "grad_norm": 0.518804669380188,
+      "learning_rate": 1.432203917713626e-06,
+      "loss": 0.9252,
+      "step": 10362
+    },
+    {
+      "epoch": 0.9259499184667277,
+      "grad_norm": 0.5390114188194275,
+      "learning_rate": 1.4287672857803636e-06,
+      "loss": 0.8615,
+      "step": 10363
+    },
+    {
+      "epoch": 0.9260392699979896,
+      "grad_norm": 0.48673346638679504,
+      "learning_rate": 1.4253347222055735e-06,
+      "loss": 0.9964,
+      "step": 10364
+    },
+    {
+      "epoch": 0.9261286215292515,
+      "grad_norm": 0.5194094181060791,
+      "learning_rate": 1.4219062272767703e-06,
+      "loss": 0.9008,
+      "step": 10365
+    },
+    {
+      "epoch": 0.9262179730605133,
+      "grad_norm": 0.5350120067596436,
+      "learning_rate": 1.418481801281124e-06,
+      "loss": 0.8815,
+      "step": 10366
+    },
+    {
+      "epoch": 0.9263073245917752,
+      "grad_norm": 0.4165932536125183,
+      "learning_rate": 1.4150614445054778e-06,
+      "loss": 0.9316,
+      "step": 10367
+    },
+    {
+      "epoch": 0.9263966761230371,
+      "grad_norm": 0.6019562482833862,
+      "learning_rate": 1.411645157236302e-06,
+      "loss": 0.9976,
+      "step": 10368
+    },
+    {
+      "epoch": 0.9264860276542989,
+      "grad_norm": 0.5063771605491638,
+      "learning_rate": 1.4082329397597626e-06,
+      "loss": 0.972,
+      "step": 10369
+    },
+    {
+      "epoch": 0.9265753791855608,
+      "grad_norm": 0.46804505586624146,
+      "learning_rate": 1.4048247923616642e-06,
+      "loss": 0.9119,
+      "step": 10370
+    },
+    {
+      "epoch": 0.9266647307168226,
+      "grad_norm": 0.4946081340312958,
+      "learning_rate": 1.4014207153274783e-06,
+      "loss": 0.9554,
+      "step": 10371
+    },
+    {
+      "epoch": 0.9267540822480845,
+      "grad_norm": 0.5020267963409424,
+      "learning_rate": 1.3980207089423326e-06,
+      "loss": 0.9053,
+      "step": 10372
+    },
+    {
+      "epoch": 0.9268434337793464,
+      "grad_norm": 0.5129795670509338,
+      "learning_rate": 1.3946247734910156e-06,
+      "loss": 0.9016,
+      "step": 10373
+    },
+    {
+      "epoch": 0.9269327853106083,
+      "grad_norm": 0.43974798917770386,
+      "learning_rate": 1.3912329092579668e-06,
+      "loss": 0.963,
+      "step": 10374
+    },
+    {
+      "epoch": 0.9270221368418702,
+      "grad_norm": 0.4542407691478729,
+      "learning_rate": 1.387845116527292e-06,
+      "loss": 0.9474,
+      "step": 10375
+    },
+    {
+      "epoch": 0.927111488373132,
+      "grad_norm": 0.418317973613739,
+      "learning_rate": 1.3844613955827535e-06,
+      "loss": 0.9627,
+      "step": 10376
+    },
+    {
+      "epoch": 0.9272008399043938,
+      "grad_norm": 0.5148499011993408,
+      "learning_rate": 1.3810817467077852e-06,
+      "loss": 0.9356,
+      "step": 10377
+    },
+    {
+      "epoch": 0.9272901914356557,
+      "grad_norm": 0.44719183444976807,
+      "learning_rate": 1.377706170185461e-06,
+      "loss": 0.9662,
+      "step": 10378
+    },
+    {
+      "epoch": 0.9273795429669176,
+      "grad_norm": 0.48192569613456726,
+      "learning_rate": 1.3743346662985157e-06,
+      "loss": 0.9405,
+      "step": 10379
+    },
+    {
+      "epoch": 0.9274688944981795,
+      "grad_norm": 0.6048439741134644,
+      "learning_rate": 1.3709672353293568e-06,
+      "loss": 0.9381,
+      "step": 10380
+    },
+    {
+      "epoch": 0.9275582460294414,
+      "grad_norm": 0.4958132207393646,
+      "learning_rate": 1.3676038775600364e-06,
+      "loss": 0.9396,
+      "step": 10381
+    },
+    {
+      "epoch": 0.9276475975607033,
+      "grad_norm": 0.4941558539867401,
+      "learning_rate": 1.3642445932722792e-06,
+      "loss": 0.9128,
+      "step": 10382
+    },
+    {
+      "epoch": 0.927736949091965,
+      "grad_norm": 0.5161193609237671,
+      "learning_rate": 1.36088938274746e-06,
+      "loss": 0.9273,
+      "step": 10383
+    },
+    {
+      "epoch": 0.9278263006232269,
+      "grad_norm": 0.47243648767471313,
+      "learning_rate": 1.3575382462666042e-06,
+      "loss": 0.9353,
+      "step": 10384
+    },
+    {
+      "epoch": 0.9279156521544888,
+      "grad_norm": 0.432420551776886,
+      "learning_rate": 1.3541911841104149e-06,
+      "loss": 0.9181,
+      "step": 10385
+    },
+    {
+      "epoch": 0.9280050036857507,
+      "grad_norm": 0.507323145866394,
+      "learning_rate": 1.3508481965592401e-06,
+      "loss": 0.8991,
+      "step": 10386
+    },
+    {
+      "epoch": 0.9280943552170126,
+      "grad_norm": 0.5412149429321289,
+      "learning_rate": 1.3475092838930947e-06,
+      "loss": 0.8649,
+      "step": 10387
+    },
+    {
+      "epoch": 0.9281837067482744,
+      "grad_norm": 0.42463141679763794,
+      "learning_rate": 1.344174446391644e-06,
+      "loss": 0.9389,
+      "step": 10388
+    },
+    {
+      "epoch": 0.9282730582795362,
+      "grad_norm": 0.4768441319465637,
+      "learning_rate": 1.3408436843342142e-06,
+      "loss": 0.9983,
+      "step": 10389
+    },
+    {
+      "epoch": 0.9283624098107981,
+      "grad_norm": 0.584004282951355,
+      "learning_rate": 1.3375169979997992e-06,
+      "loss": 0.9239,
+      "step": 10390
+    },
+    {
+      "epoch": 0.92845176134206,
+      "grad_norm": 0.4398382604122162,
+      "learning_rate": 1.3341943876670371e-06,
+      "loss": 0.9366,
+      "step": 10391
+    },
+    {
+      "epoch": 0.9285411128733219,
+      "grad_norm": 0.4435214102268219,
+      "learning_rate": 1.3308758536142384e-06,
+      "loss": 0.9742,
+      "step": 10392
+    },
+    {
+      "epoch": 0.9286304644045837,
+      "grad_norm": 0.5455980896949768,
+      "learning_rate": 1.327561396119359e-06,
+      "loss": 0.953,
+      "step": 10393
+    },
+    {
+      "epoch": 0.9287198159358456,
+      "grad_norm": 0.4934171438217163,
+      "learning_rate": 1.3242510154600207e-06,
+      "loss": 0.8277,
+      "step": 10394
+    },
+    {
+      "epoch": 0.9288091674671075,
+      "grad_norm": 0.4403308629989624,
+      "learning_rate": 1.3209447119135132e-06,
+      "loss": 0.9771,
+      "step": 10395
+    },
+    {
+      "epoch": 0.9288985189983693,
+      "grad_norm": 0.600942075252533,
+      "learning_rate": 1.3176424857567648e-06,
+      "loss": 0.8728,
+      "step": 10396
+    },
+    {
+      "epoch": 0.9289878705296312,
+      "grad_norm": 0.6132327318191528,
+      "learning_rate": 1.3143443372663767e-06,
+      "loss": 0.9608,
+      "step": 10397
+    },
+    {
+      "epoch": 0.929077222060893,
+      "grad_norm": 0.7262205481529236,
+      "learning_rate": 1.3110502667185997e-06,
+      "loss": 0.8809,
+      "step": 10398
+    },
+    {
+      "epoch": 0.9291665735921549,
+      "grad_norm": 0.49000728130340576,
+      "learning_rate": 1.3077602743893523e-06,
+      "loss": 0.8944,
+      "step": 10399
+    },
+    {
+      "epoch": 0.9292559251234168,
+      "grad_norm": 0.5410964488983154,
+      "learning_rate": 1.3044743605541975e-06,
+      "loss": 0.9214,
+      "step": 10400
+    },
+    {
+      "epoch": 0.9293452766546787,
+      "grad_norm": 0.46295785903930664,
+      "learning_rate": 1.3011925254883761e-06,
+      "loss": 0.9303,
+      "step": 10401
+    },
+    {
+      "epoch": 0.9294346281859406,
+      "grad_norm": 0.4503743052482605,
+      "learning_rate": 1.2979147694667738e-06,
+      "loss": 0.9973,
+      "step": 10402
+    },
+    {
+      "epoch": 0.9295239797172024,
+      "grad_norm": 0.4440877437591553,
+      "learning_rate": 1.2946410927639374e-06,
+      "loss": 0.9027,
+      "step": 10403
+    },
+    {
+      "epoch": 0.9296133312484642,
+      "grad_norm": 0.5882024765014648,
+      "learning_rate": 1.29137149565407e-06,
+      "loss": 0.8308,
+      "step": 10404
+    },
+    {
+      "epoch": 0.9297026827797261,
+      "grad_norm": 0.6086094975471497,
+      "learning_rate": 1.2881059784110362e-06,
+      "loss": 0.9265,
+      "step": 10405
+    },
+    {
+      "epoch": 0.929792034310988,
+      "grad_norm": 0.522862434387207,
+      "learning_rate": 1.284844541308361e-06,
+      "loss": 0.9177,
+      "step": 10406
+    },
+    {
+      "epoch": 0.9298813858422499,
+      "grad_norm": 0.42290636897087097,
+      "learning_rate": 1.2815871846192152e-06,
+      "loss": 0.9461,
+      "step": 10407
+    },
+    {
+      "epoch": 0.9299707373735118,
+      "grad_norm": 0.47796404361724854,
+      "learning_rate": 1.278333908616447e-06,
+      "loss": 0.9295,
+      "step": 10408
+    },
+    {
+      "epoch": 0.9300600889047737,
+      "grad_norm": 0.46141406893730164,
+      "learning_rate": 1.2750847135725496e-06,
+      "loss": 0.9564,
+      "step": 10409
+    },
+    {
+      "epoch": 0.9301494404360354,
+      "grad_norm": 0.5057287216186523,
+      "learning_rate": 1.2718395997596833e-06,
+      "loss": 0.9667,
+      "step": 10410
+    },
+    {
+      "epoch": 0.9302387919672973,
+      "grad_norm": 0.5116837620735168,
+      "learning_rate": 1.268598567449647e-06,
+      "loss": 0.9412,
+      "step": 10411
+    },
+    {
+      "epoch": 0.9303281434985592,
+      "grad_norm": 0.47364234924316406,
+      "learning_rate": 1.2653616169139237e-06,
+      "loss": 0.9878,
+      "step": 10412
+    },
+    {
+      "epoch": 0.9304174950298211,
+      "grad_norm": 0.5040470957756042,
+      "learning_rate": 1.262128748423641e-06,
+      "loss": 1.0276,
+      "step": 10413
+    },
+    {
+      "epoch": 0.930506846561083,
+      "grad_norm": 0.5426703095436096,
+      "learning_rate": 1.2588999622495768e-06,
+      "loss": 0.9327,
+      "step": 10414
+    },
+    {
+      "epoch": 0.9305961980923448,
+      "grad_norm": 0.5746845602989197,
+      "learning_rate": 1.2556752586621868e-06,
+      "loss": 0.9452,
+      "step": 10415
+    },
+    {
+      "epoch": 0.9306855496236067,
+      "grad_norm": 0.5168527364730835,
+      "learning_rate": 1.2524546379315717e-06,
+      "loss": 0.9003,
+      "step": 10416
+    },
+    {
+      "epoch": 0.9307749011548685,
+      "grad_norm": 0.40214788913726807,
+      "learning_rate": 1.2492381003274934e-06,
+      "loss": 0.9048,
+      "step": 10417
+    },
+    {
+      "epoch": 0.9308642526861304,
+      "grad_norm": 0.43776369094848633,
+      "learning_rate": 1.2460256461193754e-06,
+      "loss": 0.8901,
+      "step": 10418
+    },
+    {
+      "epoch": 0.9309536042173923,
+      "grad_norm": 0.4549417197704315,
+      "learning_rate": 1.2428172755762802e-06,
+      "loss": 0.9345,
+      "step": 10419
+    },
+    {
+      "epoch": 0.9310429557486541,
+      "grad_norm": 0.6516943573951721,
+      "learning_rate": 1.2396129889669595e-06,
+      "loss": 0.8175,
+      "step": 10420
+    },
+    {
+      "epoch": 0.931132307279916,
+      "grad_norm": 0.3830028772354126,
+      "learning_rate": 1.236412786559793e-06,
+      "loss": 0.9797,
+      "step": 10421
+    },
+    {
+      "epoch": 0.9312216588111779,
+      "grad_norm": 0.5979812741279602,
+      "learning_rate": 1.2332166686228386e-06,
+      "loss": 0.8904,
+      "step": 10422
+    },
+    {
+      "epoch": 0.9313110103424398,
+      "grad_norm": 0.44422391057014465,
+      "learning_rate": 1.2300246354238042e-06,
+      "loss": 0.9199,
+      "step": 10423
+    },
+    {
+      "epoch": 0.9314003618737016,
+      "grad_norm": 0.45250019431114197,
+      "learning_rate": 1.2268366872300597e-06,
+      "loss": 1.0105,
+      "step": 10424
+    },
+    {
+      "epoch": 0.9314897134049634,
+      "grad_norm": 0.5378450155258179,
+      "learning_rate": 1.2236528243086298e-06,
+      "loss": 0.9126,
+      "step": 10425
+    },
+    {
+      "epoch": 0.9315790649362253,
+      "grad_norm": 0.518375039100647,
+      "learning_rate": 1.2204730469261906e-06,
+      "loss": 0.835,
+      "step": 10426
+    },
+    {
+      "epoch": 0.9316684164674872,
+      "grad_norm": 0.6360508799552917,
+      "learning_rate": 1.217297355349084e-06,
+      "loss": 0.7857,
+      "step": 10427
+    },
+    {
+      "epoch": 0.9317577679987491,
+      "grad_norm": 0.43396756052970886,
+      "learning_rate": 1.21412574984332e-06,
+      "loss": 0.9583,
+      "step": 10428
+    },
+    {
+      "epoch": 0.931847119530011,
+      "grad_norm": 0.4555058777332306,
+      "learning_rate": 1.210958230674536e-06,
+      "loss": 0.939,
+      "step": 10429
+    },
+    {
+      "epoch": 0.9319364710612729,
+      "grad_norm": 0.38424840569496155,
+      "learning_rate": 1.2077947981080584e-06,
+      "loss": 1.0483,
+      "step": 10430
+    },
+    {
+      "epoch": 0.9320258225925346,
+      "grad_norm": 0.5196928977966309,
+      "learning_rate": 1.2046354524088477e-06,
+      "loss": 0.901,
+      "step": 10431
+    },
+    {
+      "epoch": 0.9321151741237965,
+      "grad_norm": 0.443092942237854,
+      "learning_rate": 1.2014801938415422e-06,
+      "loss": 0.9166,
+      "step": 10432
+    },
+    {
+      "epoch": 0.9322045256550584,
+      "grad_norm": 0.6846402883529663,
+      "learning_rate": 1.1983290226704247e-06,
+      "loss": 0.9144,
+      "step": 10433
+    },
+    {
+      "epoch": 0.9322938771863203,
+      "grad_norm": 0.5252601504325867,
+      "learning_rate": 1.1951819391594398e-06,
+      "loss": 0.837,
+      "step": 10434
+    },
+    {
+      "epoch": 0.9323832287175822,
+      "grad_norm": 0.48273909091949463,
+      "learning_rate": 1.1920389435721935e-06,
+      "loss": 0.9412,
+      "step": 10435
+    },
+    {
+      "epoch": 0.932472580248844,
+      "grad_norm": 0.4483881890773773,
+      "learning_rate": 1.188900036171936e-06,
+      "loss": 0.9079,
+      "step": 10436
+    },
+    {
+      "epoch": 0.9325619317801059,
+      "grad_norm": 0.5480309128761292,
+      "learning_rate": 1.1857652172215905e-06,
+      "loss": 0.8815,
+      "step": 10437
+    },
+    {
+      "epoch": 0.9326512833113677,
+      "grad_norm": 0.4683678150177002,
+      "learning_rate": 1.1826344869837359e-06,
+      "loss": 0.9372,
+      "step": 10438
+    },
+    {
+      "epoch": 0.9327406348426296,
+      "grad_norm": 0.43585410714149475,
+      "learning_rate": 1.1795078457205956e-06,
+      "loss": 0.9516,
+      "step": 10439
+    },
+    {
+      "epoch": 0.9328299863738915,
+      "grad_norm": 0.5120673775672913,
+      "learning_rate": 1.17638529369406e-06,
+      "loss": 0.8705,
+      "step": 10440
+    },
+    {
+      "epoch": 0.9329193379051534,
+      "grad_norm": 0.4925757944583893,
+      "learning_rate": 1.1732668311656815e-06,
+      "loss": 0.9006,
+      "step": 10441
+    },
+    {
+      "epoch": 0.9330086894364152,
+      "grad_norm": 0.5412935614585876,
+      "learning_rate": 1.1701524583966562e-06,
+      "loss": 0.9419,
+      "step": 10442
+    },
+    {
+      "epoch": 0.9330980409676771,
+      "grad_norm": 0.5587424039840698,
+      "learning_rate": 1.1670421756478589e-06,
+      "loss": 0.9397,
+      "step": 10443
+    },
+    {
+      "epoch": 0.933187392498939,
+      "grad_norm": 0.6443805694580078,
+      "learning_rate": 1.163935983179798e-06,
+      "loss": 0.9046,
+      "step": 10444
+    },
+    {
+      "epoch": 0.9332767440302008,
+      "grad_norm": 0.4743395149707794,
+      "learning_rate": 1.1608338812526487e-06,
+      "loss": 0.8918,
+      "step": 10445
+    },
+    {
+      "epoch": 0.9333660955614627,
+      "grad_norm": 0.44435209035873413,
+      "learning_rate": 1.157735870126253e-06,
+      "loss": 0.9825,
+      "step": 10446
+    },
+    {
+      "epoch": 0.9334554470927245,
+      "grad_norm": 0.46846985816955566,
+      "learning_rate": 1.1546419500601036e-06,
+      "loss": 0.9809,
+      "step": 10447
+    },
+    {
+      "epoch": 0.9335447986239864,
+      "grad_norm": 0.5460909008979797,
+      "learning_rate": 1.1515521213133429e-06,
+      "loss": 0.9302,
+      "step": 10448
+    },
+    {
+      "epoch": 0.9336341501552483,
+      "grad_norm": 0.5399797558784485,
+      "learning_rate": 1.1484663841447807e-06,
+      "loss": 0.9177,
+      "step": 10449
+    },
+    {
+      "epoch": 0.9337235016865102,
+      "grad_norm": 0.6021638512611389,
+      "learning_rate": 1.1453847388128712e-06,
+      "loss": 0.964,
+      "step": 10450
+    },
+    {
+      "epoch": 0.933812853217772,
+      "grad_norm": 0.4933503568172455,
+      "learning_rate": 1.1423071855757473e-06,
+      "loss": 0.8803,
+      "step": 10451
+    },
+    {
+      "epoch": 0.9339022047490338,
+      "grad_norm": 0.5711500644683838,
+      "learning_rate": 1.13923372469118e-06,
+      "loss": 0.9169,
+      "step": 10452
+    },
+    {
+      "epoch": 0.9339915562802957,
+      "grad_norm": 0.49056991934776306,
+      "learning_rate": 1.136164356416608e-06,
+      "loss": 0.8962,
+      "step": 10453
+    },
+    {
+      "epoch": 0.9340809078115576,
+      "grad_norm": 0.5031728148460388,
+      "learning_rate": 1.1330990810091257e-06,
+      "loss": 0.963,
+      "step": 10454
+    },
+    {
+      "epoch": 0.9341702593428195,
+      "grad_norm": 0.42288509011268616,
+      "learning_rate": 1.1300378987254723e-06,
+      "loss": 1.0058,
+      "step": 10455
+    },
+    {
+      "epoch": 0.9342596108740814,
+      "grad_norm": 0.4871813952922821,
+      "learning_rate": 1.1269808098220647e-06,
+      "loss": 0.9285,
+      "step": 10456
+    },
+    {
+      "epoch": 0.9343489624053433,
+      "grad_norm": 0.537005603313446,
+      "learning_rate": 1.1239278145549648e-06,
+      "loss": 0.9045,
+      "step": 10457
+    },
+    {
+      "epoch": 0.934438313936605,
+      "grad_norm": 0.4379669725894928,
+      "learning_rate": 1.1208789131798958e-06,
+      "loss": 0.9574,
+      "step": 10458
+    },
+    {
+      "epoch": 0.9345276654678669,
+      "grad_norm": 0.4513357877731323,
+      "learning_rate": 1.1178341059522256e-06,
+      "loss": 0.9283,
+      "step": 10459
+    },
+    {
+      "epoch": 0.9346170169991288,
+      "grad_norm": 0.653385579586029,
+      "learning_rate": 1.114793393126995e-06,
+      "loss": 0.8258,
+      "step": 10460
+    },
+    {
+      "epoch": 0.9347063685303907,
+      "grad_norm": 0.44483664631843567,
+      "learning_rate": 1.1117567749588997e-06,
+      "loss": 0.9546,
+      "step": 10461
+    },
+    {
+      "epoch": 0.9347957200616526,
+      "grad_norm": 0.4962851405143738,
+      "learning_rate": 1.1087242517022812e-06,
+      "loss": 0.9296,
+      "step": 10462
+    },
+    {
+      "epoch": 0.9348850715929145,
+      "grad_norm": 0.4276025891304016,
+      "learning_rate": 1.1056958236111525e-06,
+      "loss": 0.9155,
+      "step": 10463
+    },
+    {
+      "epoch": 0.9349744231241763,
+      "grad_norm": 0.5110339522361755,
+      "learning_rate": 1.1026714909391778e-06,
+      "loss": 0.9607,
+      "step": 10464
+    },
+    {
+      "epoch": 0.9350637746554381,
+      "grad_norm": 0.42139366269111633,
+      "learning_rate": 1.0996512539396707e-06,
+      "loss": 1.0041,
+      "step": 10465
+    },
+    {
+      "epoch": 0.9351531261867,
+      "grad_norm": 0.4701647162437439,
+      "learning_rate": 1.0966351128656072e-06,
+      "loss": 0.9622,
+      "step": 10466
+    },
+    {
+      "epoch": 0.9352424777179619,
+      "grad_norm": 0.47479528188705444,
+      "learning_rate": 1.093623067969629e-06,
+      "loss": 0.9374,
+      "step": 10467
+    },
+    {
+      "epoch": 0.9353318292492238,
+      "grad_norm": 0.5557815432548523,
+      "learning_rate": 1.0906151195040294e-06,
+      "loss": 0.92,
+      "step": 10468
+    },
+    {
+      "epoch": 0.9354211807804856,
+      "grad_norm": 0.46990975737571716,
+      "learning_rate": 1.087611267720745e-06,
+      "loss": 0.912,
+      "step": 10469
+    },
+    {
+      "epoch": 0.9355105323117475,
+      "grad_norm": 0.5090728402137756,
+      "learning_rate": 1.0846115128713862e-06,
+      "loss": 0.9633,
+      "step": 10470
+    },
+    {
+      "epoch": 0.9355998838430094,
+      "grad_norm": 0.45003437995910645,
+      "learning_rate": 1.081615855207213e-06,
+      "loss": 0.9622,
+      "step": 10471
+    },
+    {
+      "epoch": 0.9356892353742712,
+      "grad_norm": 0.5591026544570923,
+      "learning_rate": 1.0786242949791415e-06,
+      "loss": 0.9682,
+      "step": 10472
+    },
+    {
+      "epoch": 0.9357785869055331,
+      "grad_norm": 0.48961174488067627,
+      "learning_rate": 1.0756368324377542e-06,
+      "loss": 0.9561,
+      "step": 10473
+    },
+    {
+      "epoch": 0.9358679384367949,
+      "grad_norm": 0.6354184746742249,
+      "learning_rate": 1.0726534678332733e-06,
+      "loss": 0.8864,
+      "step": 10474
+    },
+    {
+      "epoch": 0.9359572899680568,
+      "grad_norm": 0.3932569921016693,
+      "learning_rate": 1.069674201415599e-06,
+      "loss": 0.9948,
+      "step": 10475
+    },
+    {
+      "epoch": 0.9360466414993187,
+      "grad_norm": 0.4889991283416748,
+      "learning_rate": 1.0666990334342707e-06,
+      "loss": 0.9149,
+      "step": 10476
+    },
+    {
+      "epoch": 0.9361359930305806,
+      "grad_norm": 0.46199727058410645,
+      "learning_rate": 1.0637279641384834e-06,
+      "loss": 0.9391,
+      "step": 10477
+    },
+    {
+      "epoch": 0.9362253445618425,
+      "grad_norm": 0.5159189701080322,
+      "learning_rate": 1.0607609937771046e-06,
+      "loss": 0.9332,
+      "step": 10478
+    },
+    {
+      "epoch": 0.9363146960931042,
+      "grad_norm": 0.5201767086982727,
+      "learning_rate": 1.0577981225986467e-06,
+      "loss": 0.8602,
+      "step": 10479
+    },
+    {
+      "epoch": 0.9364040476243661,
+      "grad_norm": 0.4806266725063324,
+      "learning_rate": 1.0548393508512887e-06,
+      "loss": 0.9179,
+      "step": 10480
+    },
+    {
+      "epoch": 0.936493399155628,
+      "grad_norm": 0.4578525722026825,
+      "learning_rate": 1.0518846787828496e-06,
+      "loss": 0.9347,
+      "step": 10481
+    },
+    {
+      "epoch": 0.9365827506868899,
+      "grad_norm": 0.4183438718318939,
+      "learning_rate": 1.0489341066408142e-06,
+      "loss": 0.9347,
+      "step": 10482
+    },
+    {
+      "epoch": 0.9366721022181518,
+      "grad_norm": 0.5696753859519958,
+      "learning_rate": 1.0459876346723297e-06,
+      "loss": 0.8983,
+      "step": 10483
+    },
+    {
+      "epoch": 0.9367614537494137,
+      "grad_norm": 0.46501290798187256,
+      "learning_rate": 1.0430452631241928e-06,
+      "loss": 0.9691,
+      "step": 10484
+    },
+    {
+      "epoch": 0.9368508052806755,
+      "grad_norm": 0.5632017254829407,
+      "learning_rate": 1.0401069922428619e-06,
+      "loss": 0.9061,
+      "step": 10485
+    },
+    {
+      "epoch": 0.9369401568119373,
+      "grad_norm": 0.4606979787349701,
+      "learning_rate": 1.0371728222744402e-06,
+      "loss": 0.9337,
+      "step": 10486
+    },
+    {
+      "epoch": 0.9370295083431992,
+      "grad_norm": 0.42771193385124207,
+      "learning_rate": 1.0342427534647036e-06,
+      "loss": 0.9453,
+      "step": 10487
+    },
+    {
+      "epoch": 0.9371188598744611,
+      "grad_norm": 0.49479082226753235,
+      "learning_rate": 1.0313167860590777e-06,
+      "loss": 0.9413,
+      "step": 10488
+    },
+    {
+      "epoch": 0.937208211405723,
+      "grad_norm": 0.5891922116279602,
+      "learning_rate": 1.0283949203026332e-06,
+      "loss": 0.934,
+      "step": 10489
+    },
+    {
+      "epoch": 0.9372975629369849,
+      "grad_norm": 0.4088734984397888,
+      "learning_rate": 1.0254771564401189e-06,
+      "loss": 0.9396,
+      "step": 10490
+    },
+    {
+      "epoch": 0.9373869144682467,
+      "grad_norm": 0.5687556862831116,
+      "learning_rate": 1.0225634947159223e-06,
+      "loss": 0.9473,
+      "step": 10491
+    },
+    {
+      "epoch": 0.9374762659995086,
+      "grad_norm": 0.46405094861984253,
+      "learning_rate": 1.019653935374093e-06,
+      "loss": 0.9307,
+      "step": 10492
+    },
+    {
+      "epoch": 0.9375656175307704,
+      "grad_norm": 0.451276570558548,
+      "learning_rate": 1.0167484786583359e-06,
+      "loss": 0.9777,
+      "step": 10493
+    },
+    {
+      "epoch": 0.9376549690620323,
+      "grad_norm": 0.46850451827049255,
+      "learning_rate": 1.013847124812023e-06,
+      "loss": 0.9389,
+      "step": 10494
+    },
+    {
+      "epoch": 0.9377443205932942,
+      "grad_norm": 0.454631507396698,
+      "learning_rate": 1.010949874078171e-06,
+      "loss": 0.8892,
+      "step": 10495
+    },
+    {
+      "epoch": 0.937833672124556,
+      "grad_norm": 0.6396118402481079,
+      "learning_rate": 1.0080567266994466e-06,
+      "loss": 0.8246,
+      "step": 10496
+    },
+    {
+      "epoch": 0.9379230236558179,
+      "grad_norm": 0.5484038591384888,
+      "learning_rate": 1.00516768291819e-06,
+      "loss": 0.8602,
+      "step": 10497
+    },
+    {
+      "epoch": 0.9380123751870798,
+      "grad_norm": 0.5306549072265625,
+      "learning_rate": 1.0022827429763903e-06,
+      "loss": 1.019,
+      "step": 10498
+    },
+    {
+      "epoch": 0.9381017267183417,
+      "grad_norm": 0.5081790089607239,
+      "learning_rate": 9.99401907115688e-07,
+      "loss": 0.8561,
+      "step": 10499
+    },
+    {
+      "epoch": 0.9381910782496035,
+      "grad_norm": 0.43956875801086426,
+      "learning_rate": 9.96525175577384e-07,
+      "loss": 0.9825,
+      "step": 10500
+    },
+    {
+      "epoch": 0.9382804297808653,
+      "grad_norm": 0.6813711524009705,
+      "learning_rate": 9.936525486024362e-07,
+      "loss": 0.7838,
+      "step": 10501
+    },
+    {
+      "epoch": 0.9383697813121272,
+      "grad_norm": 0.5399671792984009,
+      "learning_rate": 9.907840264314572e-07,
+      "loss": 0.943,
+      "step": 10502
+    },
+    {
+      "epoch": 0.9384591328433891,
+      "grad_norm": 0.4913422763347626,
+      "learning_rate": 9.87919609304716e-07,
+      "loss": 0.9302,
+      "step": 10503
+    },
+    {
+      "epoch": 0.938548484374651,
+      "grad_norm": 0.5128286480903625,
+      "learning_rate": 9.850592974621375e-07,
+      "loss": 0.9291,
+      "step": 10504
+    },
+    {
+      "epoch": 0.9386378359059129,
+      "grad_norm": 0.6238715052604675,
+      "learning_rate": 9.822030911433023e-07,
+      "loss": 0.9145,
+      "step": 10505
+    },
+    {
+      "epoch": 0.9387271874371748,
+      "grad_norm": 0.5400130748748779,
+      "learning_rate": 9.793509905874576e-07,
+      "loss": 0.9463,
+      "step": 10506
+    },
+    {
+      "epoch": 0.9388165389684365,
+      "grad_norm": 0.534290075302124,
+      "learning_rate": 9.765029960334849e-07,
+      "loss": 0.9139,
+      "step": 10507
+    },
+    {
+      "epoch": 0.9389058904996984,
+      "grad_norm": 0.48370060324668884,
+      "learning_rate": 9.736591077199374e-07,
+      "loss": 0.9766,
+      "step": 10508
+    },
+    {
+      "epoch": 0.9389952420309603,
+      "grad_norm": 0.44061046838760376,
+      "learning_rate": 9.708193258850306e-07,
+      "loss": 0.9452,
+      "step": 10509
+    },
+    {
+      "epoch": 0.9390845935622222,
+      "grad_norm": 0.4897972643375397,
+      "learning_rate": 9.679836507666185e-07,
+      "loss": 0.9245,
+      "step": 10510
+    },
+    {
+      "epoch": 0.9391739450934841,
+      "grad_norm": 0.5403768420219421,
+      "learning_rate": 9.65152082602211e-07,
+      "loss": 0.9311,
+      "step": 10511
+    },
+    {
+      "epoch": 0.939263296624746,
+      "grad_norm": 0.4956395626068115,
+      "learning_rate": 9.62324621628996e-07,
+      "loss": 0.9089,
+      "step": 10512
+    },
+    {
+      "epoch": 0.9393526481560077,
+      "grad_norm": 0.5809354782104492,
+      "learning_rate": 9.595012680838012e-07,
+      "loss": 0.8998,
+      "step": 10513
+    },
+    {
+      "epoch": 0.9394419996872696,
+      "grad_norm": 0.46071991324424744,
+      "learning_rate": 9.566820222031036e-07,
+      "loss": 0.8987,
+      "step": 10514
+    },
+    {
+      "epoch": 0.9395313512185315,
+      "grad_norm": 0.49813321232795715,
+      "learning_rate": 9.538668842230537e-07,
+      "loss": 0.9903,
+      "step": 10515
+    },
+    {
+      "epoch": 0.9396207027497934,
+      "grad_norm": 0.47896721959114075,
+      "learning_rate": 9.510558543794457e-07,
+      "loss": 1.0244,
+      "step": 10516
+    },
+    {
+      "epoch": 0.9397100542810553,
+      "grad_norm": 0.4987740218639374,
+      "learning_rate": 9.482489329077304e-07,
+      "loss": 0.9312,
+      "step": 10517
+    },
+    {
+      "epoch": 0.9397994058123171,
+      "grad_norm": 0.4500676989555359,
+      "learning_rate": 9.454461200430253e-07,
+      "loss": 0.9262,
+      "step": 10518
+    },
+    {
+      "epoch": 0.939888757343579,
+      "grad_norm": 0.4563513398170471,
+      "learning_rate": 9.426474160200926e-07,
+      "loss": 0.9153,
+      "step": 10519
+    },
+    {
+      "epoch": 0.9399781088748408,
+      "grad_norm": 0.4408057928085327,
+      "learning_rate": 9.398528210733504e-07,
+      "loss": 0.9599,
+      "step": 10520
+    },
+    {
+      "epoch": 0.9400674604061027,
+      "grad_norm": 0.5273951888084412,
+      "learning_rate": 9.370623354368779e-07,
+      "loss": 0.9386,
+      "step": 10521
+    },
+    {
+      "epoch": 0.9401568119373646,
+      "grad_norm": 0.6240958571434021,
+      "learning_rate": 9.342759593444106e-07,
+      "loss": 0.8736,
+      "step": 10522
+    },
+    {
+      "epoch": 0.9402461634686264,
+      "grad_norm": 0.48620718717575073,
+      "learning_rate": 9.314936930293283e-07,
+      "loss": 0.9117,
+      "step": 10523
+    },
+    {
+      "epoch": 0.9403355149998883,
+      "grad_norm": 0.45210275053977966,
+      "learning_rate": 9.287155367246891e-07,
+      "loss": 0.9074,
+      "step": 10524
+    },
+    {
+      "epoch": 0.9404248665311502,
+      "grad_norm": 0.5615553259849548,
+      "learning_rate": 9.259414906631791e-07,
+      "loss": 0.9937,
+      "step": 10525
+    },
+    {
+      "epoch": 0.9405142180624121,
+      "grad_norm": 0.46976426243782043,
+      "learning_rate": 9.23171555077168e-07,
+      "loss": 1.0001,
+      "step": 10526
+    },
+    {
+      "epoch": 0.9406035695936739,
+      "grad_norm": 0.5349940061569214,
+      "learning_rate": 9.204057301986535e-07,
+      "loss": 0.8587,
+      "step": 10527
+    },
+    {
+      "epoch": 0.9406929211249357,
+      "grad_norm": 0.48549193143844604,
+      "learning_rate": 9.176440162593169e-07,
+      "loss": 0.9925,
+      "step": 10528
+    },
+    {
+      "epoch": 0.9407822726561976,
+      "grad_norm": 0.48987144231796265,
+      "learning_rate": 9.148864134904733e-07,
+      "loss": 0.9652,
+      "step": 10529
+    },
+    {
+      "epoch": 0.9408716241874595,
+      "grad_norm": 0.4827273190021515,
+      "learning_rate": 9.12132922123099e-07,
+      "loss": 0.8644,
+      "step": 10530
+    },
+    {
+      "epoch": 0.9409609757187214,
+      "grad_norm": 0.5220755934715271,
+      "learning_rate": 9.093835423878317e-07,
+      "loss": 0.9001,
+      "step": 10531
+    },
+    {
+      "epoch": 0.9410503272499833,
+      "grad_norm": 0.4924890995025635,
+      "learning_rate": 9.066382745149649e-07,
+      "loss": 0.9662,
+      "step": 10532
+    },
+    {
+      "epoch": 0.9411396787812452,
+      "grad_norm": 0.5399982929229736,
+      "learning_rate": 9.038971187344369e-07,
+      "loss": 0.8605,
+      "step": 10533
+    },
+    {
+      "epoch": 0.9412290303125069,
+      "grad_norm": 0.5098146200180054,
+      "learning_rate": 9.011600752758531e-07,
+      "loss": 0.9553,
+      "step": 10534
+    },
+    {
+      "epoch": 0.9413183818437688,
+      "grad_norm": 0.42846280336380005,
+      "learning_rate": 8.984271443684633e-07,
+      "loss": 0.9544,
+      "step": 10535
+    },
+    {
+      "epoch": 0.9414077333750307,
+      "grad_norm": 0.5252984762191772,
+      "learning_rate": 8.9569832624119e-07,
+      "loss": 0.8778,
+      "step": 10536
+    },
+    {
+      "epoch": 0.9414970849062926,
+      "grad_norm": 0.5305891036987305,
+      "learning_rate": 8.929736211226003e-07,
+      "loss": 0.8921,
+      "step": 10537
+    },
+    {
+      "epoch": 0.9415864364375545,
+      "grad_norm": 0.4615854024887085,
+      "learning_rate": 8.902530292409062e-07,
+      "loss": 0.9217,
+      "step": 10538
+    },
+    {
+      "epoch": 0.9416757879688163,
+      "grad_norm": 0.45759010314941406,
+      "learning_rate": 8.87536550824003e-07,
+      "loss": 0.9712,
+      "step": 10539
+    },
+    {
+      "epoch": 0.9417651395000782,
+      "grad_norm": 0.48792535066604614,
+      "learning_rate": 8.848241860994089e-07,
+      "loss": 0.876,
+      "step": 10540
+    },
+    {
+      "epoch": 0.94185449103134,
+      "grad_norm": 0.4174993932247162,
+      "learning_rate": 8.821159352943143e-07,
+      "loss": 0.9367,
+      "step": 10541
+    },
+    {
+      "epoch": 0.9419438425626019,
+      "grad_norm": 0.4825289845466614,
+      "learning_rate": 8.794117986355766e-07,
+      "loss": 0.9969,
+      "step": 10542
+    },
+    {
+      "epoch": 0.9420331940938638,
+      "grad_norm": 0.44167813658714294,
+      "learning_rate": 8.767117763496813e-07,
+      "loss": 0.8798,
+      "step": 10543
+    },
+    {
+      "epoch": 0.9421225456251257,
+      "grad_norm": 0.5294803380966187,
+      "learning_rate": 8.740158686627975e-07,
+      "loss": 0.8875,
+      "step": 10544
+    },
+    {
+      "epoch": 0.9422118971563875,
+      "grad_norm": 0.4376986622810364,
+      "learning_rate": 8.71324075800728e-07,
+      "loss": 0.9939,
+      "step": 10545
+    },
+    {
+      "epoch": 0.9423012486876494,
+      "grad_norm": 0.5412419438362122,
+      "learning_rate": 8.686363979889478e-07,
+      "loss": 0.9387,
+      "step": 10546
+    },
+    {
+      "epoch": 0.9423906002189113,
+      "grad_norm": 0.5909743309020996,
+      "learning_rate": 8.659528354525603e-07,
+      "loss": 0.8535,
+      "step": 10547
+    },
+    {
+      "epoch": 0.9424799517501731,
+      "grad_norm": 0.430073082447052,
+      "learning_rate": 8.632733884163635e-07,
+      "loss": 0.9364,
+      "step": 10548
+    },
+    {
+      "epoch": 0.942569303281435,
+      "grad_norm": 0.5627753734588623,
+      "learning_rate": 8.605980571047723e-07,
+      "loss": 0.9279,
+      "step": 10549
+    },
+    {
+      "epoch": 0.9426586548126968,
+      "grad_norm": 0.526467502117157,
+      "learning_rate": 8.579268417418851e-07,
+      "loss": 0.9335,
+      "step": 10550
+    },
+    {
+      "epoch": 0.9427480063439587,
+      "grad_norm": 0.5456862449645996,
+      "learning_rate": 8.552597425514508e-07,
+      "loss": 0.8866,
+      "step": 10551
+    },
+    {
+      "epoch": 0.9428373578752206,
+      "grad_norm": 0.5069810152053833,
+      "learning_rate": 8.525967597568463e-07,
+      "loss": 0.9052,
+      "step": 10552
+    },
+    {
+      "epoch": 0.9429267094064825,
+      "grad_norm": 0.625331461429596,
+      "learning_rate": 8.499378935811431e-07,
+      "loss": 0.9148,
+      "step": 10553
+    },
+    {
+      "epoch": 0.9430160609377444,
+      "grad_norm": 0.5902383327484131,
+      "learning_rate": 8.472831442470408e-07,
+      "loss": 0.802,
+      "step": 10554
+    },
+    {
+      "epoch": 0.9431054124690061,
+      "grad_norm": 0.42086061835289,
+      "learning_rate": 8.446325119769061e-07,
+      "loss": 0.9733,
+      "step": 10555
+    },
+    {
+      "epoch": 0.943194764000268,
+      "grad_norm": 0.5746432542800903,
+      "learning_rate": 8.419859969927557e-07,
+      "loss": 0.8944,
+      "step": 10556
+    },
+    {
+      "epoch": 0.9432841155315299,
+      "grad_norm": 0.5022989511489868,
+      "learning_rate": 8.393435995162624e-07,
+      "loss": 0.9403,
+      "step": 10557
+    },
+    {
+      "epoch": 0.9433734670627918,
+      "grad_norm": 0.443338543176651,
+      "learning_rate": 8.367053197687602e-07,
+      "loss": 0.9092,
+      "step": 10558
+    },
+    {
+      "epoch": 0.9434628185940537,
+      "grad_norm": 0.592194676399231,
+      "learning_rate": 8.340711579712391e-07,
+      "loss": 0.8213,
+      "step": 10559
+    },
+    {
+      "epoch": 0.9435521701253156,
+      "grad_norm": 0.47716042399406433,
+      "learning_rate": 8.314411143443168e-07,
+      "loss": 0.9157,
+      "step": 10560
+    },
+    {
+      "epoch": 0.9436415216565774,
+      "grad_norm": 0.49664196372032166,
+      "learning_rate": 8.288151891083062e-07,
+      "loss": 0.8716,
+      "step": 10561
+    },
+    {
+      "epoch": 0.9437308731878392,
+      "grad_norm": 0.471483439207077,
+      "learning_rate": 8.261933824831481e-07,
+      "loss": 0.9732,
+      "step": 10562
+    },
+    {
+      "epoch": 0.9438202247191011,
+      "grad_norm": 0.4351029396057129,
+      "learning_rate": 8.235756946884554e-07,
+      "loss": 1.0083,
+      "step": 10563
+    },
+    {
+      "epoch": 0.943909576250363,
+      "grad_norm": 0.4930981993675232,
+      "learning_rate": 8.209621259434753e-07,
+      "loss": 0.9445,
+      "step": 10564
+    },
+    {
+      "epoch": 0.9439989277816249,
+      "grad_norm": 0.42323318123817444,
+      "learning_rate": 8.183526764671267e-07,
+      "loss": 0.9145,
+      "step": 10565
+    },
+    {
+      "epoch": 0.9440882793128867,
+      "grad_norm": 0.4111577570438385,
+      "learning_rate": 8.157473464779852e-07,
+      "loss": 0.9204,
+      "step": 10566
+    },
+    {
+      "epoch": 0.9441776308441486,
+      "grad_norm": 0.6320559978485107,
+      "learning_rate": 8.13146136194265e-07,
+      "loss": 0.9821,
+      "step": 10567
+    },
+    {
+      "epoch": 0.9442669823754105,
+      "grad_norm": 0.47050777077674866,
+      "learning_rate": 8.105490458338527e-07,
+      "loss": 0.9344,
+      "step": 10568
+    },
+    {
+      "epoch": 0.9443563339066723,
+      "grad_norm": 0.6935842037200928,
+      "learning_rate": 8.079560756142857e-07,
+      "loss": 0.9204,
+      "step": 10569
+    },
+    {
+      "epoch": 0.9444456854379342,
+      "grad_norm": 0.5569193363189697,
+      "learning_rate": 8.0536722575274e-07,
+      "loss": 0.8738,
+      "step": 10570
+    },
+    {
+      "epoch": 0.944535036969196,
+      "grad_norm": 0.4778384864330292,
+      "learning_rate": 8.027824964660646e-07,
+      "loss": 0.924,
+      "step": 10571
+    },
+    {
+      "epoch": 0.9446243885004579,
+      "grad_norm": 0.4366244375705719,
+      "learning_rate": 8.00201887970764e-07,
+      "loss": 0.9712,
+      "step": 10572
+    },
+    {
+      "epoch": 0.9447137400317198,
+      "grad_norm": 0.6287685036659241,
+      "learning_rate": 7.976254004829875e-07,
+      "loss": 0.9883,
+      "step": 10573
+    },
+    {
+      "epoch": 0.9448030915629817,
+      "grad_norm": 0.464234858751297,
+      "learning_rate": 7.950530342185402e-07,
+      "loss": 0.9353,
+      "step": 10574
+    },
+    {
+      "epoch": 0.9448924430942435,
+      "grad_norm": 0.5004445314407349,
+      "learning_rate": 7.924847893928888e-07,
+      "loss": 0.8881,
+      "step": 10575
+    },
+    {
+      "epoch": 0.9449817946255054,
+      "grad_norm": 0.5838055610656738,
+      "learning_rate": 7.8992066622115e-07,
+      "loss": 0.9222,
+      "step": 10576
+    },
+    {
+      "epoch": 0.9450711461567672,
+      "grad_norm": 0.4792827367782593,
+      "learning_rate": 7.873606649180965e-07,
+      "loss": 0.9294,
+      "step": 10577
+    },
+    {
+      "epoch": 0.9451604976880291,
+      "grad_norm": 0.4360794723033905,
+      "learning_rate": 7.848047856981622e-07,
+      "loss": 0.9488,
+      "step": 10578
+    },
+    {
+      "epoch": 0.945249849219291,
+      "grad_norm": 0.5515472888946533,
+      "learning_rate": 7.822530287754204e-07,
+      "loss": 0.8643,
+      "step": 10579
+    },
+    {
+      "epoch": 0.9453392007505529,
+      "grad_norm": 0.6533846855163574,
+      "learning_rate": 7.797053943636112e-07,
+      "loss": 0.9626,
+      "step": 10580
+    },
+    {
+      "epoch": 0.9454285522818148,
+      "grad_norm": 0.5607351660728455,
+      "learning_rate": 7.771618826761252e-07,
+      "loss": 0.9153,
+      "step": 10581
+    },
+    {
+      "epoch": 0.9455179038130765,
+      "grad_norm": 0.5081503987312317,
+      "learning_rate": 7.746224939260083e-07,
+      "loss": 0.8538,
+      "step": 10582
+    },
+    {
+      "epoch": 0.9456072553443384,
+      "grad_norm": 0.5000874996185303,
+      "learning_rate": 7.720872283259684e-07,
+      "loss": 0.9195,
+      "step": 10583
+    },
+    {
+      "epoch": 0.9456966068756003,
+      "grad_norm": 0.49381589889526367,
+      "learning_rate": 7.695560860883467e-07,
+      "loss": 0.9285,
+      "step": 10584
+    },
+    {
+      "epoch": 0.9457859584068622,
+      "grad_norm": 0.54331374168396,
+      "learning_rate": 7.670290674251679e-07,
+      "loss": 0.9005,
+      "step": 10585
+    },
+    {
+      "epoch": 0.9458753099381241,
+      "grad_norm": 0.5094651579856873,
+      "learning_rate": 7.64506172548085e-07,
+      "loss": 0.8645,
+      "step": 10586
+    },
+    {
+      "epoch": 0.945964661469386,
+      "grad_norm": 0.5358061194419861,
+      "learning_rate": 7.619874016684237e-07,
+      "loss": 0.8771,
+      "step": 10587
+    },
+    {
+      "epoch": 0.9460540130006478,
+      "grad_norm": 0.45969656109809875,
+      "learning_rate": 7.594727549971592e-07,
+      "loss": 0.9333,
+      "step": 10588
+    },
+    {
+      "epoch": 0.9461433645319096,
+      "grad_norm": 0.5184633135795593,
+      "learning_rate": 7.569622327449177e-07,
+      "loss": 0.8593,
+      "step": 10589
+    },
+    {
+      "epoch": 0.9462327160631715,
+      "grad_norm": 0.6350348591804504,
+      "learning_rate": 7.54455835121981e-07,
+      "loss": 0.8568,
+      "step": 10590
+    },
+    {
+      "epoch": 0.9463220675944334,
+      "grad_norm": 0.45934930443763733,
+      "learning_rate": 7.519535623382867e-07,
+      "loss": 0.8905,
+      "step": 10591
+    },
+    {
+      "epoch": 0.9464114191256953,
+      "grad_norm": 0.6196146607398987,
+      "learning_rate": 7.494554146034338e-07,
+      "loss": 0.9422,
+      "step": 10592
+    },
+    {
+      "epoch": 0.9465007706569571,
+      "grad_norm": 0.4321631193161011,
+      "learning_rate": 7.46961392126655e-07,
+      "loss": 0.9287,
+      "step": 10593
+    },
+    {
+      "epoch": 0.946590122188219,
+      "grad_norm": 0.494315505027771,
+      "learning_rate": 7.444714951168663e-07,
+      "loss": 0.8973,
+      "step": 10594
+    },
+    {
+      "epoch": 0.9466794737194809,
+      "grad_norm": 0.4586873948574066,
+      "learning_rate": 7.419857237826122e-07,
+      "loss": 0.8857,
+      "step": 10595
+    },
+    {
+      "epoch": 0.9467688252507427,
+      "grad_norm": 0.46470096707344055,
+      "learning_rate": 7.395040783321039e-07,
+      "loss": 0.9417,
+      "step": 10596
+    },
+    {
+      "epoch": 0.9468581767820046,
+      "grad_norm": 0.45211324095726013,
+      "learning_rate": 7.37026558973214e-07,
+      "loss": 0.8929,
+      "step": 10597
+    },
+    {
+      "epoch": 0.9469475283132665,
+      "grad_norm": 0.5554351210594177,
+      "learning_rate": 7.345531659134486e-07,
+      "loss": 0.9703,
+      "step": 10598
+    },
+    {
+      "epoch": 0.9470368798445283,
+      "grad_norm": 0.46897411346435547,
+      "learning_rate": 7.320838993599921e-07,
+      "loss": 0.9272,
+      "step": 10599
+    },
+    {
+      "epoch": 0.9471262313757902,
+      "grad_norm": 0.6147266030311584,
+      "learning_rate": 7.29618759519668e-07,
+      "loss": 0.98,
+      "step": 10600
+    },
+    {
+      "epoch": 0.9472155829070521,
+      "grad_norm": 0.5554401278495789,
+      "learning_rate": 7.271577465989555e-07,
+      "loss": 0.9312,
+      "step": 10601
+    },
+    {
+      "epoch": 0.947304934438314,
+      "grad_norm": 0.5745187401771545,
+      "learning_rate": 7.247008608039952e-07,
+      "loss": 0.8978,
+      "step": 10602
+    },
+    {
+      "epoch": 0.9473942859695758,
+      "grad_norm": 0.46855631470680237,
+      "learning_rate": 7.222481023405725e-07,
+      "loss": 0.949,
+      "step": 10603
+    },
+    {
+      "epoch": 0.9474836375008376,
+      "grad_norm": 0.5807746052742004,
+      "learning_rate": 7.197994714141343e-07,
+      "loss": 0.8789,
+      "step": 10604
+    },
+    {
+      "epoch": 0.9475729890320995,
+      "grad_norm": 0.539803147315979,
+      "learning_rate": 7.173549682297775e-07,
+      "loss": 1.0313,
+      "step": 10605
+    },
+    {
+      "epoch": 0.9476623405633614,
+      "grad_norm": 0.45081114768981934,
+      "learning_rate": 7.149145929922607e-07,
+      "loss": 0.9441,
+      "step": 10606
+    },
+    {
+      "epoch": 0.9477516920946233,
+      "grad_norm": 0.44997450709342957,
+      "learning_rate": 7.124783459059869e-07,
+      "loss": 0.9809,
+      "step": 10607
+    },
+    {
+      "epoch": 0.9478410436258852,
+      "grad_norm": 0.43726733326911926,
+      "learning_rate": 7.100462271750153e-07,
+      "loss": 0.9145,
+      "step": 10608
+    },
+    {
+      "epoch": 0.947930395157147,
+      "grad_norm": 0.5305030345916748,
+      "learning_rate": 7.076182370030715e-07,
+      "loss": 0.9287,
+      "step": 10609
+    },
+    {
+      "epoch": 0.9480197466884088,
+      "grad_norm": 0.3688342869281769,
+      "learning_rate": 7.051943755935208e-07,
+      "loss": 0.9817,
+      "step": 10610
+    },
+    {
+      "epoch": 0.9481090982196707,
+      "grad_norm": 0.40400731563568115,
+      "learning_rate": 7.027746431493787e-07,
+      "loss": 0.9076,
+      "step": 10611
+    },
+    {
+      "epoch": 0.9481984497509326,
+      "grad_norm": 0.5000659823417664,
+      "learning_rate": 7.003590398733329e-07,
+      "loss": 0.9145,
+      "step": 10612
+    },
+    {
+      "epoch": 0.9482878012821945,
+      "grad_norm": 0.4885746240615845,
+      "learning_rate": 6.97947565967716e-07,
+      "loss": 0.9152,
+      "step": 10613
+    },
+    {
+      "epoch": 0.9483771528134564,
+      "grad_norm": 0.48515942692756653,
+      "learning_rate": 6.955402216345108e-07,
+      "loss": 0.9315,
+      "step": 10614
+    },
+    {
+      "epoch": 0.9484665043447182,
+      "grad_norm": 0.43972206115722656,
+      "learning_rate": 6.931370070753618e-07,
+      "loss": 0.9436,
+      "step": 10615
+    },
+    {
+      "epoch": 0.9485558558759801,
+      "grad_norm": 0.5218010544776917,
+      "learning_rate": 6.907379224915633e-07,
+      "loss": 0.9323,
+      "step": 10616
+    },
+    {
+      "epoch": 0.9486452074072419,
+      "grad_norm": 0.5221048593521118,
+      "learning_rate": 6.883429680840602e-07,
+      "loss": 0.9423,
+      "step": 10617
+    },
+    {
+      "epoch": 0.9487345589385038,
+      "grad_norm": 0.5904656648635864,
+      "learning_rate": 6.859521440534533e-07,
+      "loss": 0.8302,
+      "step": 10618
+    },
+    {
+      "epoch": 0.9488239104697657,
+      "grad_norm": 0.4238061010837555,
+      "learning_rate": 6.8356545060001e-07,
+      "loss": 1.0147,
+      "step": 10619
+    },
+    {
+      "epoch": 0.9489132620010275,
+      "grad_norm": 0.47028911113739014,
+      "learning_rate": 6.81182887923637e-07,
+      "loss": 0.8621,
+      "step": 10620
+    },
+    {
+      "epoch": 0.9490026135322894,
+      "grad_norm": 0.5632848143577576,
+      "learning_rate": 6.788044562238971e-07,
+      "loss": 0.9579,
+      "step": 10621
+    },
+    {
+      "epoch": 0.9490919650635513,
+      "grad_norm": 0.6260432004928589,
+      "learning_rate": 6.764301557000086e-07,
+      "loss": 0.8357,
+      "step": 10622
+    },
+    {
+      "epoch": 0.9491813165948132,
+      "grad_norm": 0.6236041784286499,
+      "learning_rate": 6.740599865508457e-07,
+      "loss": 0.8724,
+      "step": 10623
+    },
+    {
+      "epoch": 0.949270668126075,
+      "grad_norm": 0.4447486400604248,
+      "learning_rate": 6.716939489749329e-07,
+      "loss": 0.9633,
+      "step": 10624
+    },
+    {
+      "epoch": 0.9493600196573369,
+      "grad_norm": 0.43622735142707825,
+      "learning_rate": 6.693320431704564e-07,
+      "loss": 0.93,
+      "step": 10625
+    },
+    {
+      "epoch": 0.9494493711885987,
+      "grad_norm": 0.5855457186698914,
+      "learning_rate": 6.669742693352521e-07,
+      "loss": 0.8923,
+      "step": 10626
+    },
+    {
+      "epoch": 0.9495387227198606,
+      "grad_norm": 0.4751369059085846,
+      "learning_rate": 6.64620627666801e-07,
+      "loss": 0.8902,
+      "step": 10627
+    },
+    {
+      "epoch": 0.9496280742511225,
+      "grad_norm": 0.5162234902381897,
+      "learning_rate": 6.622711183622454e-07,
+      "loss": 0.9346,
+      "step": 10628
+    },
+    {
+      "epoch": 0.9497174257823844,
+      "grad_norm": 0.3996429145336151,
+      "learning_rate": 6.599257416183946e-07,
+      "loss": 0.9754,
+      "step": 10629
+    },
+    {
+      "epoch": 0.9498067773136463,
+      "grad_norm": 0.4350222647190094,
+      "learning_rate": 6.575844976316859e-07,
+      "loss": 0.9303,
+      "step": 10630
+    },
+    {
+      "epoch": 0.949896128844908,
+      "grad_norm": 0.4600478708744049,
+      "learning_rate": 6.552473865982289e-07,
+      "loss": 0.8937,
+      "step": 10631
+    },
+    {
+      "epoch": 0.9499854803761699,
+      "grad_norm": 0.6685845851898193,
+      "learning_rate": 6.52914408713784e-07,
+      "loss": 0.9852,
+      "step": 10632
+    },
+    {
+      "epoch": 0.9500748319074318,
+      "grad_norm": 0.43094709515571594,
+      "learning_rate": 6.505855641737502e-07,
+      "loss": 1.0011,
+      "step": 10633
+    },
+    {
+      "epoch": 0.9501641834386937,
+      "grad_norm": 0.4459003806114197,
+      "learning_rate": 6.482608531732104e-07,
+      "loss": 0.9094,
+      "step": 10634
+    },
+    {
+      "epoch": 0.9502535349699556,
+      "grad_norm": 0.4444451630115509,
+      "learning_rate": 6.4594027590687e-07,
+      "loss": 0.993,
+      "step": 10635
+    },
+    {
+      "epoch": 0.9503428865012175,
+      "grad_norm": 0.5474777817726135,
+      "learning_rate": 6.436238325691125e-07,
+      "loss": 0.8881,
+      "step": 10636
+    },
+    {
+      "epoch": 0.9504322380324792,
+      "grad_norm": 0.5109720826148987,
+      "learning_rate": 6.413115233539601e-07,
+      "loss": 0.9903,
+      "step": 10637
+    },
+    {
+      "epoch": 0.9505215895637411,
+      "grad_norm": 0.4545588195323944,
+      "learning_rate": 6.390033484550917e-07,
+      "loss": 0.9167,
+      "step": 10638
+    },
+    {
+      "epoch": 0.950610941095003,
+      "grad_norm": 0.5279081463813782,
+      "learning_rate": 6.366993080658413e-07,
+      "loss": 0.9207,
+      "step": 10639
+    },
+    {
+      "epoch": 0.9507002926262649,
+      "grad_norm": 0.5103138089179993,
+      "learning_rate": 6.343994023792043e-07,
+      "loss": 0.9425,
+      "step": 10640
+    },
+    {
+      "epoch": 0.9507896441575268,
+      "grad_norm": 0.49605733156204224,
+      "learning_rate": 6.32103631587816e-07,
+      "loss": 0.898,
+      "step": 10641
+    },
+    {
+      "epoch": 0.9508789956887886,
+      "grad_norm": 0.5339604616165161,
+      "learning_rate": 6.298119958839721e-07,
+      "loss": 0.8929,
+      "step": 10642
+    },
+    {
+      "epoch": 0.9509683472200505,
+      "grad_norm": 0.48836469650268555,
+      "learning_rate": 6.275244954596193e-07,
+      "loss": 0.8902,
+      "step": 10643
+    },
+    {
+      "epoch": 0.9510576987513123,
+      "grad_norm": 0.4708765149116516,
+      "learning_rate": 6.252411305063599e-07,
+      "loss": 0.9173,
+      "step": 10644
+    },
+    {
+      "epoch": 0.9511470502825742,
+      "grad_norm": 0.47018665075302124,
+      "learning_rate": 6.229619012154575e-07,
+      "loss": 0.9882,
+      "step": 10645
+    },
+    {
+      "epoch": 0.9512364018138361,
+      "grad_norm": 0.5642026662826538,
+      "learning_rate": 6.206868077778149e-07,
+      "loss": 0.9281,
+      "step": 10646
+    },
+    {
+      "epoch": 0.951325753345098,
+      "grad_norm": 0.6329473853111267,
+      "learning_rate": 6.18415850384002e-07,
+      "loss": 0.887,
+      "step": 10647
+    },
+    {
+      "epoch": 0.9514151048763598,
+      "grad_norm": 0.47655338048934937,
+      "learning_rate": 6.161490292242278e-07,
+      "loss": 0.908,
+      "step": 10648
+    },
+    {
+      "epoch": 0.9515044564076217,
+      "grad_norm": 0.4613625109195709,
+      "learning_rate": 6.138863444883735e-07,
+      "loss": 0.9321,
+      "step": 10649
+    },
+    {
+      "epoch": 0.9515938079388836,
+      "grad_norm": 0.5077817440032959,
+      "learning_rate": 6.116277963659489e-07,
+      "loss": 0.986,
+      "step": 10650
+    },
+    {
+      "epoch": 0.9516831594701454,
+      "grad_norm": 0.4817694425582886,
+      "learning_rate": 6.093733850461358e-07,
+      "loss": 0.9819,
+      "step": 10651
+    },
+    {
+      "epoch": 0.9517725110014073,
+      "grad_norm": 0.6474151015281677,
+      "learning_rate": 6.071231107177722e-07,
+      "loss": 0.8317,
+      "step": 10652
+    },
+    {
+      "epoch": 0.9518618625326691,
+      "grad_norm": 0.529495894908905,
+      "learning_rate": 6.048769735693404e-07,
+      "loss": 0.872,
+      "step": 10653
+    },
+    {
+      "epoch": 0.951951214063931,
+      "grad_norm": 0.574099600315094,
+      "learning_rate": 6.026349737889736e-07,
+      "loss": 0.9059,
+      "step": 10654
+    },
+    {
+      "epoch": 0.9520405655951929,
+      "grad_norm": 0.5554847717285156,
+      "learning_rate": 6.003971115644657e-07,
+      "loss": 0.9797,
+      "step": 10655
+    },
+    {
+      "epoch": 0.9521299171264548,
+      "grad_norm": 0.5695211887359619,
+      "learning_rate": 5.981633870832614e-07,
+      "loss": 0.9082,
+      "step": 10656
+    },
+    {
+      "epoch": 0.9522192686577167,
+      "grad_norm": 0.6171822547912598,
+      "learning_rate": 5.959338005324611e-07,
+      "loss": 0.9138,
+      "step": 10657
+    },
+    {
+      "epoch": 0.9523086201889784,
+      "grad_norm": 0.4547090232372284,
+      "learning_rate": 5.937083520988151e-07,
+      "loss": 0.9651,
+      "step": 10658
+    },
+    {
+      "epoch": 0.9523979717202403,
+      "grad_norm": 0.5358104705810547,
+      "learning_rate": 5.914870419687247e-07,
+      "loss": 0.9548,
+      "step": 10659
+    },
+    {
+      "epoch": 0.9524873232515022,
+      "grad_norm": 0.5056635737419128,
+      "learning_rate": 5.892698703282517e-07,
+      "loss": 0.8873,
+      "step": 10660
+    },
+    {
+      "epoch": 0.9525766747827641,
+      "grad_norm": 0.5599133968353271,
+      "learning_rate": 5.870568373631091e-07,
+      "loss": 0.8852,
+      "step": 10661
+    },
+    {
+      "epoch": 0.952666026314026,
+      "grad_norm": 0.4533610939979553,
+      "learning_rate": 5.848479432586596e-07,
+      "loss": 0.9582,
+      "step": 10662
+    },
+    {
+      "epoch": 0.9527553778452879,
+      "grad_norm": 0.42300939559936523,
+      "learning_rate": 5.826431881999217e-07,
+      "loss": 0.938,
+      "step": 10663
+    },
+    {
+      "epoch": 0.9528447293765497,
+      "grad_norm": 0.525570273399353,
+      "learning_rate": 5.804425723715701e-07,
+      "loss": 0.9239,
+      "step": 10664
+    },
+    {
+      "epoch": 0.9529340809078115,
+      "grad_norm": 0.5064711570739746,
+      "learning_rate": 5.782460959579239e-07,
+      "loss": 0.8886,
+      "step": 10665
+    },
+    {
+      "epoch": 0.9530234324390734,
+      "grad_norm": 0.4518652558326721,
+      "learning_rate": 5.760537591429694e-07,
+      "loss": 0.9666,
+      "step": 10666
+    },
+    {
+      "epoch": 0.9531127839703353,
+      "grad_norm": 0.4667336344718933,
+      "learning_rate": 5.738655621103317e-07,
+      "loss": 0.9292,
+      "step": 10667
+    },
+    {
+      "epoch": 0.9532021355015972,
+      "grad_norm": 0.4580380916595459,
+      "learning_rate": 5.716815050432978e-07,
+      "loss": 0.934,
+      "step": 10668
+    },
+    {
+      "epoch": 0.953291487032859,
+      "grad_norm": 0.5494727492332458,
+      "learning_rate": 5.6950158812481e-07,
+      "loss": 0.9544,
+      "step": 10669
+    },
+    {
+      "epoch": 0.9533808385641209,
+      "grad_norm": 0.5223135948181152,
+      "learning_rate": 5.673258115374502e-07,
+      "loss": 0.8561,
+      "step": 10670
+    },
+    {
+      "epoch": 0.9534701900953828,
+      "grad_norm": 0.5879161953926086,
+      "learning_rate": 5.651541754634726e-07,
+      "loss": 0.8332,
+      "step": 10671
+    },
+    {
+      "epoch": 0.9535595416266446,
+      "grad_norm": 0.4340555965900421,
+      "learning_rate": 5.629866800847649e-07,
+      "loss": 0.9383,
+      "step": 10672
+    },
+    {
+      "epoch": 0.9536488931579065,
+      "grad_norm": 0.46510595083236694,
+      "learning_rate": 5.608233255828876e-07,
+      "loss": 0.8908,
+      "step": 10673
+    },
+    {
+      "epoch": 0.9537382446891683,
+      "grad_norm": 0.5728726983070374,
+      "learning_rate": 5.586641121390401e-07,
+      "loss": 0.8977,
+      "step": 10674
+    },
+    {
+      "epoch": 0.9538275962204302,
+      "grad_norm": 0.4889596402645111,
+      "learning_rate": 5.565090399340778e-07,
+      "loss": 0.9271,
+      "step": 10675
+    },
+    {
+      "epoch": 0.9539169477516921,
+      "grad_norm": 0.46741414070129395,
+      "learning_rate": 5.543581091485117e-07,
+      "loss": 0.9841,
+      "step": 10676
+    },
+    {
+      "epoch": 0.954006299282954,
+      "grad_norm": 0.48655492067337036,
+      "learning_rate": 5.522113199625145e-07,
+      "loss": 0.942,
+      "step": 10677
+    },
+    {
+      "epoch": 0.9540956508142159,
+      "grad_norm": 0.4470749497413635,
+      "learning_rate": 5.500686725558868e-07,
+      "loss": 0.9452,
+      "step": 10678
+    },
+    {
+      "epoch": 0.9541850023454777,
+      "grad_norm": 0.4536135494709015,
+      "learning_rate": 5.479301671081072e-07,
+      "loss": 0.9409,
+      "step": 10679
+    },
+    {
+      "epoch": 0.9542743538767395,
+      "grad_norm": 0.4575904905796051,
+      "learning_rate": 5.457958037982991e-07,
+      "loss": 0.9012,
+      "step": 10680
+    },
+    {
+      "epoch": 0.9543637054080014,
+      "grad_norm": 0.4889681935310364,
+      "learning_rate": 5.436655828052417e-07,
+      "loss": 0.971,
+      "step": 10681
+    },
+    {
+      "epoch": 0.9544530569392633,
+      "grad_norm": 0.443453311920166,
+      "learning_rate": 5.415395043073535e-07,
+      "loss": 0.9073,
+      "step": 10682
+    },
+    {
+      "epoch": 0.9545424084705252,
+      "grad_norm": 0.4644291400909424,
+      "learning_rate": 5.394175684827196e-07,
+      "loss": 0.9093,
+      "step": 10683
+    },
+    {
+      "epoch": 0.9546317600017871,
+      "grad_norm": 0.5638459920883179,
+      "learning_rate": 5.372997755090759e-07,
+      "loss": 0.8817,
+      "step": 10684
+    },
+    {
+      "epoch": 0.954721111533049,
+      "grad_norm": 0.42751070857048035,
+      "learning_rate": 5.351861255638135e-07,
+      "loss": 1.0156,
+      "step": 10685
+    },
+    {
+      "epoch": 0.9548104630643107,
+      "grad_norm": 0.5429770350456238,
+      "learning_rate": 5.330766188239689e-07,
+      "loss": 0.9907,
+      "step": 10686
+    },
+    {
+      "epoch": 0.9548998145955726,
+      "grad_norm": 0.5348584651947021,
+      "learning_rate": 5.309712554662338e-07,
+      "loss": 0.8703,
+      "step": 10687
+    },
+    {
+      "epoch": 0.9549891661268345,
+      "grad_norm": 0.5381792783737183,
+      "learning_rate": 5.288700356669618e-07,
+      "loss": 0.9238,
+      "step": 10688
+    },
+    {
+      "epoch": 0.9550785176580964,
+      "grad_norm": 0.5914027690887451,
+      "learning_rate": 5.267729596021509e-07,
+      "loss": 0.9369,
+      "step": 10689
+    },
+    {
+      "epoch": 0.9551678691893583,
+      "grad_norm": 0.573617696762085,
+      "learning_rate": 5.246800274474439e-07,
+      "loss": 0.8617,
+      "step": 10690
+    },
+    {
+      "epoch": 0.9552572207206201,
+      "grad_norm": 0.4635196030139923,
+      "learning_rate": 5.225912393781617e-07,
+      "loss": 0.9357,
+      "step": 10691
+    },
+    {
+      "epoch": 0.955346572251882,
+      "grad_norm": 0.4225331246852875,
+      "learning_rate": 5.205065955692534e-07,
+      "loss": 0.9849,
+      "step": 10692
+    },
+    {
+      "epoch": 0.9554359237831438,
+      "grad_norm": 0.5266596078872681,
+      "learning_rate": 5.184260961953236e-07,
+      "loss": 0.9057,
+      "step": 10693
+    },
+    {
+      "epoch": 0.9555252753144057,
+      "grad_norm": 0.48360317945480347,
+      "learning_rate": 5.163497414306495e-07,
+      "loss": 0.897,
+      "step": 10694
+    },
+    {
+      "epoch": 0.9556146268456676,
+      "grad_norm": 0.44401875138282776,
+      "learning_rate": 5.142775314491422e-07,
+      "loss": 0.9219,
+      "step": 10695
+    },
+    {
+      "epoch": 0.9557039783769294,
+      "grad_norm": 0.4794129431247711,
+      "learning_rate": 5.122094664243681e-07,
+      "loss": 0.9605,
+      "step": 10696
+    },
+    {
+      "epoch": 0.9557933299081913,
+      "grad_norm": 0.5471349358558655,
+      "learning_rate": 5.101455465295557e-07,
+      "loss": 0.9078,
+      "step": 10697
+    },
+    {
+      "epoch": 0.9558826814394532,
+      "grad_norm": 0.5558369755744934,
+      "learning_rate": 5.080857719375776e-07,
+      "loss": 0.9239,
+      "step": 10698
+    },
+    {
+      "epoch": 0.9559720329707151,
+      "grad_norm": 0.39528143405914307,
+      "learning_rate": 5.060301428209624e-07,
+      "loss": 0.9493,
+      "step": 10699
+    },
+    {
+      "epoch": 0.9560613845019769,
+      "grad_norm": 0.44044265151023865,
+      "learning_rate": 5.039786593518892e-07,
+      "loss": 0.9668,
+      "step": 10700
+    },
+    {
+      "epoch": 0.9561507360332387,
+      "grad_norm": 0.44840747117996216,
+      "learning_rate": 5.019313217021982e-07,
+      "loss": 0.943,
+      "step": 10701
+    },
+    {
+      "epoch": 0.9562400875645006,
+      "grad_norm": 0.4926467537879944,
+      "learning_rate": 4.998881300433688e-07,
+      "loss": 0.9684,
+      "step": 10702
+    },
+    {
+      "epoch": 0.9563294390957625,
+      "grad_norm": 0.44300875067710876,
+      "learning_rate": 4.978490845465367e-07,
+      "loss": 0.9856,
+      "step": 10703
+    },
+    {
+      "epoch": 0.9564187906270244,
+      "grad_norm": 0.4651431143283844,
+      "learning_rate": 4.958141853825038e-07,
+      "loss": 0.9457,
+      "step": 10704
+    },
+    {
+      "epoch": 0.9565081421582863,
+      "grad_norm": 0.4982956051826477,
+      "learning_rate": 4.937834327217061e-07,
+      "loss": 0.8796,
+      "step": 10705
+    },
+    {
+      "epoch": 0.956597493689548,
+      "grad_norm": 0.46135175228118896,
+      "learning_rate": 4.917568267342465e-07,
+      "loss": 0.8973,
+      "step": 10706
+    },
+    {
+      "epoch": 0.9566868452208099,
+      "grad_norm": 0.5590856075286865,
+      "learning_rate": 4.897343675898669e-07,
+      "loss": 0.9275,
+      "step": 10707
+    },
+    {
+      "epoch": 0.9567761967520718,
+      "grad_norm": 0.5813520550727844,
+      "learning_rate": 4.877160554579818e-07,
+      "loss": 0.9401,
+      "step": 10708
+    },
+    {
+      "epoch": 0.9568655482833337,
+      "grad_norm": 0.44653165340423584,
+      "learning_rate": 4.857018905076394e-07,
+      "loss": 1.0116,
+      "step": 10709
+    },
+    {
+      "epoch": 0.9569548998145956,
+      "grad_norm": 0.47590041160583496,
+      "learning_rate": 4.836918729075435e-07,
+      "loss": 0.9912,
+      "step": 10710
+    },
+    {
+      "epoch": 0.9570442513458575,
+      "grad_norm": 0.42629966139793396,
+      "learning_rate": 4.81686002826065e-07,
+      "loss": 0.9783,
+      "step": 10711
+    },
+    {
+      "epoch": 0.9571336028771193,
+      "grad_norm": 0.5415294766426086,
+      "learning_rate": 4.796842804312085e-07,
+      "loss": 0.9848,
+      "step": 10712
+    },
+    {
+      "epoch": 0.9572229544083811,
+      "grad_norm": 0.5828223824501038,
+      "learning_rate": 4.776867058906453e-07,
+      "loss": 0.8597,
+      "step": 10713
+    },
+    {
+      "epoch": 0.957312305939643,
+      "grad_norm": 0.5175209045410156,
+      "learning_rate": 4.756932793716862e-07,
+      "loss": 0.9513,
+      "step": 10714
+    },
+    {
+      "epoch": 0.9574016574709049,
+      "grad_norm": 0.5043903589248657,
+      "learning_rate": 4.737040010413085e-07,
+      "loss": 0.8341,
+      "step": 10715
+    },
+    {
+      "epoch": 0.9574910090021668,
+      "grad_norm": 0.47045809030532837,
+      "learning_rate": 4.717188710661291e-07,
+      "loss": 0.8929,
+      "step": 10716
+    },
+    {
+      "epoch": 0.9575803605334287,
+      "grad_norm": 0.44731298089027405,
+      "learning_rate": 4.697378896124316e-07,
+      "loss": 0.9371,
+      "step": 10717
+    },
+    {
+      "epoch": 0.9576697120646905,
+      "grad_norm": 0.47598275542259216,
+      "learning_rate": 4.677610568461388e-07,
+      "loss": 0.9035,
+      "step": 10718
+    },
+    {
+      "epoch": 0.9577590635959524,
+      "grad_norm": 0.49408817291259766,
+      "learning_rate": 4.657883729328405e-07,
+      "loss": 0.8944,
+      "step": 10719
+    },
+    {
+      "epoch": 0.9578484151272142,
+      "grad_norm": 0.41959306597709656,
+      "learning_rate": 4.638198380377545e-07,
+      "loss": 0.9322,
+      "step": 10720
+    },
+    {
+      "epoch": 0.9579377666584761,
+      "grad_norm": 0.43749403953552246,
+      "learning_rate": 4.6185545232577676e-07,
+      "loss": 0.9512,
+      "step": 10721
+    },
+    {
+      "epoch": 0.958027118189738,
+      "grad_norm": 0.426736980676651,
+      "learning_rate": 4.5989521596144226e-07,
+      "loss": 0.9449,
+      "step": 10722
+    },
+    {
+      "epoch": 0.9581164697209998,
+      "grad_norm": 0.5225173830986023,
+      "learning_rate": 4.579391291089419e-07,
+      "loss": 0.8694,
+      "step": 10723
+    },
+    {
+      "epoch": 0.9582058212522617,
+      "grad_norm": 0.6672084927558899,
+      "learning_rate": 4.559871919321279e-07,
+      "loss": 0.8843,
+      "step": 10724
+    },
+    {
+      "epoch": 0.9582951727835236,
+      "grad_norm": 0.4476062059402466,
+      "learning_rate": 4.5403940459448067e-07,
+      "loss": 0.898,
+      "step": 10725
+    },
+    {
+      "epoch": 0.9583845243147855,
+      "grad_norm": 0.4649253189563751,
+      "learning_rate": 4.5209576725915304e-07,
+      "loss": 0.8641,
+      "step": 10726
+    },
+    {
+      "epoch": 0.9584738758460473,
+      "grad_norm": 0.4393841028213501,
+      "learning_rate": 4.501562800889536e-07,
+      "loss": 0.9845,
+      "step": 10727
+    },
+    {
+      "epoch": 0.9585632273773091,
+      "grad_norm": 0.5080018639564514,
+      "learning_rate": 4.482209432463247e-07,
+      "loss": 0.8999,
+      "step": 10728
+    },
+    {
+      "epoch": 0.958652578908571,
+      "grad_norm": 0.5580213665962219,
+      "learning_rate": 4.46289756893381e-07,
+      "loss": 0.9335,
+      "step": 10729
+    },
+    {
+      "epoch": 0.9587419304398329,
+      "grad_norm": 0.6076408624649048,
+      "learning_rate": 4.443627211918711e-07,
+      "loss": 0.9558,
+      "step": 10730
+    },
+    {
+      "epoch": 0.9588312819710948,
+      "grad_norm": 0.4416392147541046,
+      "learning_rate": 4.424398363032101e-07,
+      "loss": 1.0056,
+      "step": 10731
+    },
+    {
+      "epoch": 0.9589206335023567,
+      "grad_norm": 0.4628840684890747,
+      "learning_rate": 4.405211023884581e-07,
+      "loss": 0.9612,
+      "step": 10732
+    },
+    {
+      "epoch": 0.9590099850336186,
+      "grad_norm": 0.5556047558784485,
+      "learning_rate": 4.3860651960832557e-07,
+      "loss": 1.0094,
+      "step": 10733
+    },
+    {
+      "epoch": 0.9590993365648803,
+      "grad_norm": 0.4615866541862488,
+      "learning_rate": 4.3669608812318965e-07,
+      "loss": 0.9931,
+      "step": 10734
+    },
+    {
+      "epoch": 0.9591886880961422,
+      "grad_norm": 0.4226515293121338,
+      "learning_rate": 4.347898080930557e-07,
+      "loss": 0.909,
+      "step": 10735
+    },
+    {
+      "epoch": 0.9592780396274041,
+      "grad_norm": 0.4294814169406891,
+      "learning_rate": 4.3288767967760715e-07,
+      "loss": 0.9615,
+      "step": 10736
+    },
+    {
+      "epoch": 0.959367391158666,
+      "grad_norm": 0.5078641772270203,
+      "learning_rate": 4.3098970303616646e-07,
+      "loss": 0.8305,
+      "step": 10737
+    },
+    {
+      "epoch": 0.9594567426899279,
+      "grad_norm": 0.5030995607376099,
+      "learning_rate": 4.2909587832770103e-07,
+      "loss": 1.0022,
+      "step": 10738
+    },
+    {
+      "epoch": 0.9595460942211897,
+      "grad_norm": 0.5276926159858704,
+      "learning_rate": 4.27206205710845e-07,
+      "loss": 0.8767,
+      "step": 10739
+    },
+    {
+      "epoch": 0.9596354457524516,
+      "grad_norm": 0.450339138507843,
+      "learning_rate": 4.2532068534387737e-07,
+      "loss": 0.9567,
+      "step": 10740
+    },
+    {
+      "epoch": 0.9597247972837134,
+      "grad_norm": 0.4343004822731018,
+      "learning_rate": 4.2343931738473284e-07,
+      "loss": 0.9784,
+      "step": 10741
+    },
+    {
+      "epoch": 0.9598141488149753,
+      "grad_norm": 0.45386016368865967,
+      "learning_rate": 4.215621019909854e-07,
+      "loss": 0.8928,
+      "step": 10742
+    },
+    {
+      "epoch": 0.9599035003462372,
+      "grad_norm": 0.6192027926445007,
+      "learning_rate": 4.196890393198871e-07,
+      "loss": 0.9318,
+      "step": 10743
+    },
+    {
+      "epoch": 0.959992851877499,
+      "grad_norm": 0.5403250455856323,
+      "learning_rate": 4.1782012952831796e-07,
+      "loss": 0.841,
+      "step": 10744
+    },
+    {
+      "epoch": 0.9600822034087609,
+      "grad_norm": 0.525173544883728,
+      "learning_rate": 4.159553727728194e-07,
+      "loss": 0.9311,
+      "step": 10745
+    },
+    {
+      "epoch": 0.9601715549400228,
+      "grad_norm": 0.5325625538825989,
+      "learning_rate": 4.140947692095887e-07,
+      "loss": 0.8706,
+      "step": 10746
+    },
+    {
+      "epoch": 0.9602609064712847,
+      "grad_norm": 0.49182307720184326,
+      "learning_rate": 4.1223831899446785e-07,
+      "loss": 0.9359,
+      "step": 10747
+    },
+    {
+      "epoch": 0.9603502580025465,
+      "grad_norm": 0.444618821144104,
+      "learning_rate": 4.103860222829603e-07,
+      "loss": 0.9907,
+      "step": 10748
+    },
+    {
+      "epoch": 0.9604396095338084,
+      "grad_norm": 0.4664936065673828,
+      "learning_rate": 4.0853787923020303e-07,
+      "loss": 1.0142,
+      "step": 10749
+    },
+    {
+      "epoch": 0.9605289610650702,
+      "grad_norm": 0.6530784964561462,
+      "learning_rate": 4.066938899910111e-07,
+      "loss": 0.8257,
+      "step": 10750
+    },
+    {
+      "epoch": 0.9606183125963321,
+      "grad_norm": 0.5376946330070496,
+      "learning_rate": 4.048540547198332e-07,
+      "loss": 0.8809,
+      "step": 10751
+    },
+    {
+      "epoch": 0.960707664127594,
+      "grad_norm": 0.4881064295768738,
+      "learning_rate": 4.030183735707682e-07,
+      "loss": 0.9446,
+      "step": 10752
+    },
+    {
+      "epoch": 0.9607970156588559,
+      "grad_norm": 0.46255356073379517,
+      "learning_rate": 4.011868466975821e-07,
+      "loss": 0.9665,
+      "step": 10753
+    },
+    {
+      "epoch": 0.9608863671901178,
+      "grad_norm": 0.5297724008560181,
+      "learning_rate": 3.9935947425368546e-07,
+      "loss": 0.9761,
+      "step": 10754
+    },
+    {
+      "epoch": 0.9609757187213795,
+      "grad_norm": 0.5168519616127014,
+      "learning_rate": 3.9753625639213366e-07,
+      "loss": 0.9226,
+      "step": 10755
+    },
+    {
+      "epoch": 0.9610650702526414,
+      "grad_norm": 0.5492185354232788,
+      "learning_rate": 3.9571719326564894e-07,
+      "loss": 0.8866,
+      "step": 10756
+    },
+    {
+      "epoch": 0.9611544217839033,
+      "grad_norm": 0.5541948676109314,
+      "learning_rate": 3.939022850265928e-07,
+      "loss": 0.9604,
+      "step": 10757
+    },
+    {
+      "epoch": 0.9612437733151652,
+      "grad_norm": 0.6160182952880859,
+      "learning_rate": 3.920915318269824e-07,
+      "loss": 0.8829,
+      "step": 10758
+    },
+    {
+      "epoch": 0.9613331248464271,
+      "grad_norm": 0.5799499154090881,
+      "learning_rate": 3.902849338184911e-07,
+      "loss": 0.8777,
+      "step": 10759
+    },
+    {
+      "epoch": 0.961422476377689,
+      "grad_norm": 0.5014825463294983,
+      "learning_rate": 3.8848249115243097e-07,
+      "loss": 0.9618,
+      "step": 10760
+    },
+    {
+      "epoch": 0.9615118279089508,
+      "grad_norm": 0.4629524052143097,
+      "learning_rate": 3.866842039797869e-07,
+      "loss": 0.949,
+      "step": 10761
+    },
+    {
+      "epoch": 0.9616011794402126,
+      "grad_norm": 0.46119171380996704,
+      "learning_rate": 3.848900724511828e-07,
+      "loss": 0.9926,
+      "step": 10762
+    },
+    {
+      "epoch": 0.9616905309714745,
+      "grad_norm": 0.5495138764381409,
+      "learning_rate": 3.8310009671689297e-07,
+      "loss": 0.9226,
+      "step": 10763
+    },
+    {
+      "epoch": 0.9617798825027364,
+      "grad_norm": 0.49449822306632996,
+      "learning_rate": 3.813142769268474e-07,
+      "loss": 0.9972,
+      "step": 10764
+    },
+    {
+      "epoch": 0.9618692340339983,
+      "grad_norm": 0.4808973968029022,
+      "learning_rate": 3.7953261323063205e-07,
+      "loss": 0.9218,
+      "step": 10765
+    },
+    {
+      "epoch": 0.9619585855652601,
+      "grad_norm": 0.45000120997428894,
+      "learning_rate": 3.7775510577747195e-07,
+      "loss": 0.9093,
+      "step": 10766
+    },
+    {
+      "epoch": 0.962047937096522,
+      "grad_norm": 0.42627647519111633,
+      "learning_rate": 3.759817547162536e-07,
+      "loss": 0.9849,
+      "step": 10767
+    },
+    {
+      "epoch": 0.9621372886277838,
+      "grad_norm": 0.4252340793609619,
+      "learning_rate": 3.742125601955249e-07,
+      "loss": 0.9683,
+      "step": 10768
+    },
+    {
+      "epoch": 0.9622266401590457,
+      "grad_norm": 0.45607298612594604,
+      "learning_rate": 3.724475223634616e-07,
+      "loss": 1.0085,
+      "step": 10769
+    },
+    {
+      "epoch": 0.9623159916903076,
+      "grad_norm": 0.5643563866615295,
+      "learning_rate": 3.706866413679122e-07,
+      "loss": 0.9372,
+      "step": 10770
+    },
+    {
+      "epoch": 0.9624053432215695,
+      "grad_norm": 0.5665273666381836,
+      "learning_rate": 3.689299173563643e-07,
+      "loss": 0.9523,
+      "step": 10771
+    },
+    {
+      "epoch": 0.9624946947528313,
+      "grad_norm": 0.5675973892211914,
+      "learning_rate": 3.6717735047597233e-07,
+      "loss": 0.8598,
+      "step": 10772
+    },
+    {
+      "epoch": 0.9625840462840932,
+      "grad_norm": 0.48373866081237793,
+      "learning_rate": 3.6542894087351896e-07,
+      "loss": 0.9075,
+      "step": 10773
+    },
+    {
+      "epoch": 0.9626733978153551,
+      "grad_norm": 0.5747187733650208,
+      "learning_rate": 3.6368468869545926e-07,
+      "loss": 0.94,
+      "step": 10774
+    },
+    {
+      "epoch": 0.9627627493466169,
+      "grad_norm": 0.4350042939186096,
+      "learning_rate": 3.619445940878929e-07,
+      "loss": 0.958,
+      "step": 10775
+    },
+    {
+      "epoch": 0.9628521008778788,
+      "grad_norm": 0.5669698119163513,
+      "learning_rate": 3.6020865719657016e-07,
+      "loss": 0.9641,
+      "step": 10776
+    },
+    {
+      "epoch": 0.9629414524091406,
+      "grad_norm": 0.5079529881477356,
+      "learning_rate": 3.584768781668968e-07,
+      "loss": 0.9328,
+      "step": 10777
+    },
+    {
+      "epoch": 0.9630308039404025,
+      "grad_norm": 0.6118546724319458,
+      "learning_rate": 3.5674925714391796e-07,
+      "loss": 0.8199,
+      "step": 10778
+    },
+    {
+      "epoch": 0.9631201554716644,
+      "grad_norm": 0.4477030038833618,
+      "learning_rate": 3.5502579427235673e-07,
+      "loss": 0.9591,
+      "step": 10779
+    },
+    {
+      "epoch": 0.9632095070029263,
+      "grad_norm": 0.454915314912796,
+      "learning_rate": 3.5330648969655876e-07,
+      "loss": 0.9534,
+      "step": 10780
+    },
+    {
+      "epoch": 0.9632988585341882,
+      "grad_norm": 0.5509199500083923,
+      "learning_rate": 3.515913435605367e-07,
+      "loss": 0.8643,
+      "step": 10781
+    },
+    {
+      "epoch": 0.96338821006545,
+      "grad_norm": 0.4031422734260559,
+      "learning_rate": 3.4988035600795886e-07,
+      "loss": 0.9174,
+      "step": 10782
+    },
+    {
+      "epoch": 0.9634775615967118,
+      "grad_norm": 0.636800229549408,
+      "learning_rate": 3.481735271821274e-07,
+      "loss": 1.0021,
+      "step": 10783
+    },
+    {
+      "epoch": 0.9635669131279737,
+      "grad_norm": 0.4733976721763611,
+      "learning_rate": 3.464708572260167e-07,
+      "loss": 0.9752,
+      "step": 10784
+    },
+    {
+      "epoch": 0.9636562646592356,
+      "grad_norm": 0.4569736123085022,
+      "learning_rate": 3.44772346282235e-07,
+      "loss": 0.933,
+      "step": 10785
+    },
+    {
+      "epoch": 0.9637456161904975,
+      "grad_norm": 0.49324387311935425,
+      "learning_rate": 3.4307799449306286e-07,
+      "loss": 0.9589,
+      "step": 10786
+    },
+    {
+      "epoch": 0.9638349677217594,
+      "grad_norm": 0.5697484016418457,
+      "learning_rate": 3.413878020004091e-07,
+      "loss": 0.9579,
+      "step": 10787
+    },
+    {
+      "epoch": 0.9639243192530212,
+      "grad_norm": 0.5456252694129944,
+      "learning_rate": 3.397017689458548e-07,
+      "loss": 0.8727,
+      "step": 10788
+    },
+    {
+      "epoch": 0.964013670784283,
+      "grad_norm": 0.5522463917732239,
+      "learning_rate": 3.3801989547061484e-07,
+      "loss": 0.9692,
+      "step": 10789
+    },
+    {
+      "epoch": 0.9641030223155449,
+      "grad_norm": 0.5635389685630798,
+      "learning_rate": 3.363421817155654e-07,
+      "loss": 0.8955,
+      "step": 10790
+    },
+    {
+      "epoch": 0.9641923738468068,
+      "grad_norm": 0.525848388671875,
+      "learning_rate": 3.3466862782123853e-07,
+      "loss": 0.8466,
+      "step": 10791
+    },
+    {
+      "epoch": 0.9642817253780687,
+      "grad_norm": 0.5157281160354614,
+      "learning_rate": 3.3299923392780543e-07,
+      "loss": 1.019,
+      "step": 10792
+    },
+    {
+      "epoch": 0.9643710769093305,
+      "grad_norm": 0.6579477787017822,
+      "learning_rate": 3.3133400017509865e-07,
+      "loss": 0.8579,
+      "step": 10793
+    },
+    {
+      "epoch": 0.9644604284405924,
+      "grad_norm": 0.5229929089546204,
+      "learning_rate": 3.2967292670260106e-07,
+      "loss": 0.889,
+      "step": 10794
+    },
+    {
+      "epoch": 0.9645497799718543,
+      "grad_norm": 0.5078433752059937,
+      "learning_rate": 3.280160136494459e-07,
+      "loss": 0.9574,
+      "step": 10795
+    },
+    {
+      "epoch": 0.9646391315031161,
+      "grad_norm": 0.5875293016433716,
+      "learning_rate": 3.2636326115441097e-07,
+      "loss": 0.8764,
+      "step": 10796
+    },
+    {
+      "epoch": 0.964728483034378,
+      "grad_norm": 0.47622236609458923,
+      "learning_rate": 3.247146693559355e-07,
+      "loss": 0.9398,
+      "step": 10797
+    },
+    {
+      "epoch": 0.9648178345656399,
+      "grad_norm": 0.5416803956031799,
+      "learning_rate": 3.2307023839210914e-07,
+      "loss": 0.9299,
+      "step": 10798
+    },
+    {
+      "epoch": 0.9649071860969017,
+      "grad_norm": 0.47890692949295044,
+      "learning_rate": 3.214299684006661e-07,
+      "loss": 0.9296,
+      "step": 10799
+    },
+    {
+      "epoch": 0.9649965376281636,
+      "grad_norm": 0.42355430126190186,
+      "learning_rate": 3.197938595189964e-07,
+      "loss": 0.9806,
+      "step": 10800
+    },
+    {
+      "epoch": 0.9650858891594255,
+      "grad_norm": 0.46457356214523315,
+      "learning_rate": 3.1816191188415166e-07,
+      "loss": 0.9908,
+      "step": 10801
+    },
+    {
+      "epoch": 0.9651752406906874,
+      "grad_norm": 0.5232931971549988,
+      "learning_rate": 3.1653412563281135e-07,
+      "loss": 0.8684,
+      "step": 10802
+    },
+    {
+      "epoch": 0.9652645922219492,
+      "grad_norm": 0.4481331408023834,
+      "learning_rate": 3.1491050090132757e-07,
+      "loss": 0.9863,
+      "step": 10803
+    },
+    {
+      "epoch": 0.965353943753211,
+      "grad_norm": 0.5186441540718079,
+      "learning_rate": 3.1329103782569145e-07,
+      "loss": 0.8764,
+      "step": 10804
+    },
+    {
+      "epoch": 0.9654432952844729,
+      "grad_norm": 0.5062096118927002,
+      "learning_rate": 3.116757365415557e-07,
+      "loss": 0.8986,
+      "step": 10805
+    },
+    {
+      "epoch": 0.9655326468157348,
+      "grad_norm": 0.42634356021881104,
+      "learning_rate": 3.1006459718421755e-07,
+      "loss": 0.9799,
+      "step": 10806
+    },
+    {
+      "epoch": 0.9656219983469967,
+      "grad_norm": 0.4971635937690735,
+      "learning_rate": 3.0845761988862464e-07,
+      "loss": 0.8796,
+      "step": 10807
+    },
+    {
+      "epoch": 0.9657113498782586,
+      "grad_norm": 0.5731960535049438,
+      "learning_rate": 3.068548047893804e-07,
+      "loss": 0.8347,
+      "step": 10808
+    },
+    {
+      "epoch": 0.9658007014095205,
+      "grad_norm": 0.6698244214057922,
+      "learning_rate": 3.0525615202073863e-07,
+      "loss": 0.8799,
+      "step": 10809
+    },
+    {
+      "epoch": 0.9658900529407822,
+      "grad_norm": 0.4804188907146454,
+      "learning_rate": 3.036616617165977e-07,
+      "loss": 0.959,
+      "step": 10810
+    },
+    {
+      "epoch": 0.9659794044720441,
+      "grad_norm": 0.45491814613342285,
+      "learning_rate": 3.020713340105175e-07,
+      "loss": 0.9305,
+      "step": 10811
+    },
+    {
+      "epoch": 0.966068756003306,
+      "grad_norm": 0.43385374546051025,
+      "learning_rate": 3.0048516903571357e-07,
+      "loss": 0.9343,
+      "step": 10812
+    },
+    {
+      "epoch": 0.9661581075345679,
+      "grad_norm": 0.42967671155929565,
+      "learning_rate": 2.989031669250297e-07,
+      "loss": 0.978,
+      "step": 10813
+    },
+    {
+      "epoch": 0.9662474590658298,
+      "grad_norm": 0.4216921031475067,
+      "learning_rate": 2.973253278109767e-07,
+      "loss": 1.0218,
+      "step": 10814
+    },
+    {
+      "epoch": 0.9663368105970916,
+      "grad_norm": 0.45912298560142517,
+      "learning_rate": 2.957516518257264e-07,
+      "loss": 0.9561,
+      "step": 10815
+    },
+    {
+      "epoch": 0.9664261621283535,
+      "grad_norm": 0.4412069022655487,
+      "learning_rate": 2.9418213910107907e-07,
+      "loss": 1.0146,
+      "step": 10816
+    },
+    {
+      "epoch": 0.9665155136596153,
+      "grad_norm": 0.4734534025192261,
+      "learning_rate": 2.9261678976850726e-07,
+      "loss": 1.0074,
+      "step": 10817
+    },
+    {
+      "epoch": 0.9666048651908772,
+      "grad_norm": 0.44548165798187256,
+      "learning_rate": 2.910556039591228e-07,
+      "loss": 0.9156,
+      "step": 10818
+    },
+    {
+      "epoch": 0.9666942167221391,
+      "grad_norm": 0.45493119955062866,
+      "learning_rate": 2.894985818036877e-07,
+      "loss": 0.9426,
+      "step": 10819
+    },
+    {
+      "epoch": 0.966783568253401,
+      "grad_norm": 0.4516558051109314,
+      "learning_rate": 2.879457234326255e-07,
+      "loss": 0.9776,
+      "step": 10820
+    },
+    {
+      "epoch": 0.9668729197846628,
+      "grad_norm": 0.4548396170139313,
+      "learning_rate": 2.8639702897599873e-07,
+      "loss": 0.8973,
+      "step": 10821
+    },
+    {
+      "epoch": 0.9669622713159247,
+      "grad_norm": 0.4782199263572693,
+      "learning_rate": 2.8485249856353147e-07,
+      "loss": 0.8619,
+      "step": 10822
+    },
+    {
+      "epoch": 0.9670516228471866,
+      "grad_norm": 0.6326056718826294,
+      "learning_rate": 2.833121323245924e-07,
+      "loss": 0.8101,
+      "step": 10823
+    },
+    {
+      "epoch": 0.9671409743784484,
+      "grad_norm": 0.4548356533050537,
+      "learning_rate": 2.817759303882006e-07,
+      "loss": 0.9549,
+      "step": 10824
+    },
+    {
+      "epoch": 0.9672303259097103,
+      "grad_norm": 0.5176000595092773,
+      "learning_rate": 2.802438928830364e-07,
+      "loss": 0.8751,
+      "step": 10825
+    },
+    {
+      "epoch": 0.9673196774409721,
+      "grad_norm": 0.4941912293434143,
+      "learning_rate": 2.7871601993741947e-07,
+      "loss": 0.9054,
+      "step": 10826
+    },
+    {
+      "epoch": 0.967409028972234,
+      "grad_norm": 0.5169671773910522,
+      "learning_rate": 2.7719231167933067e-07,
+      "loss": 0.9378,
+      "step": 10827
+    },
+    {
+      "epoch": 0.9674983805034959,
+      "grad_norm": 0.5568271279335022,
+      "learning_rate": 2.7567276823639023e-07,
+      "loss": 0.9216,
+      "step": 10828
+    },
+    {
+      "epoch": 0.9675877320347578,
+      "grad_norm": 0.6154254674911499,
+      "learning_rate": 2.741573897358796e-07,
+      "loss": 0.7888,
+      "step": 10829
+    },
+    {
+      "epoch": 0.9676770835660196,
+      "grad_norm": 0.4801642894744873,
+      "learning_rate": 2.72646176304725e-07,
+      "loss": 0.9863,
+      "step": 10830
+    },
+    {
+      "epoch": 0.9677664350972814,
+      "grad_norm": 0.5245612263679504,
+      "learning_rate": 2.711391280695086e-07,
+      "loss": 0.9012,
+      "step": 10831
+    },
+    {
+      "epoch": 0.9678557866285433,
+      "grad_norm": 0.5288980603218079,
+      "learning_rate": 2.6963624515646266e-07,
+      "loss": 0.9981,
+      "step": 10832
+    },
+    {
+      "epoch": 0.9679451381598052,
+      "grad_norm": 0.3883173167705536,
+      "learning_rate": 2.681375276914644e-07,
+      "loss": 0.9628,
+      "step": 10833
+    },
+    {
+      "epoch": 0.9680344896910671,
+      "grad_norm": 0.6051060557365417,
+      "learning_rate": 2.666429758000577e-07,
+      "loss": 0.8418,
+      "step": 10834
+    },
+    {
+      "epoch": 0.968123841222329,
+      "grad_norm": 0.4276251196861267,
+      "learning_rate": 2.651525896074203e-07,
+      "loss": 0.9092,
+      "step": 10835
+    },
+    {
+      "epoch": 0.9682131927535909,
+      "grad_norm": 0.48356980085372925,
+      "learning_rate": 2.636663692383856e-07,
+      "loss": 1.0608,
+      "step": 10836
+    },
+    {
+      "epoch": 0.9683025442848526,
+      "grad_norm": 0.5755243897438049,
+      "learning_rate": 2.62184314817443e-07,
+      "loss": 0.831,
+      "step": 10837
+    },
+    {
+      "epoch": 0.9683918958161145,
+      "grad_norm": 0.4929053783416748,
+      "learning_rate": 2.6070642646873757e-07,
+      "loss": 0.9273,
+      "step": 10838
+    },
+    {
+      "epoch": 0.9684812473473764,
+      "grad_norm": 0.48126858472824097,
+      "learning_rate": 2.5923270431604804e-07,
+      "loss": 0.9546,
+      "step": 10839
+    },
+    {
+      "epoch": 0.9685705988786383,
+      "grad_norm": 0.524333119392395,
+      "learning_rate": 2.577631484828147e-07,
+      "loss": 0.8945,
+      "step": 10840
+    },
+    {
+      "epoch": 0.9686599504099002,
+      "grad_norm": 0.5595930218696594,
+      "learning_rate": 2.5629775909213337e-07,
+      "loss": 0.9558,
+      "step": 10841
+    },
+    {
+      "epoch": 0.968749301941162,
+      "grad_norm": 0.507494330406189,
+      "learning_rate": 2.5483653626675043e-07,
+      "loss": 1.0194,
+      "step": 10842
+    },
+    {
+      "epoch": 0.9688386534724239,
+      "grad_norm": 0.43305492401123047,
+      "learning_rate": 2.5337948012904566e-07,
+      "loss": 0.9712,
+      "step": 10843
+    },
+    {
+      "epoch": 0.9689280050036857,
+      "grad_norm": 0.46509480476379395,
+      "learning_rate": 2.5192659080107704e-07,
+      "loss": 0.9215,
+      "step": 10844
+    },
+    {
+      "epoch": 0.9690173565349476,
+      "grad_norm": 0.44424694776535034,
+      "learning_rate": 2.5047786840452504e-07,
+      "loss": 0.9837,
+      "step": 10845
+    },
+    {
+      "epoch": 0.9691067080662095,
+      "grad_norm": 0.44861480593681335,
+      "learning_rate": 2.4903331306074804e-07,
+      "loss": 0.9725,
+      "step": 10846
+    },
+    {
+      "epoch": 0.9691960595974713,
+      "grad_norm": 0.4651372730731964,
+      "learning_rate": 2.475929248907383e-07,
+      "loss": 0.9143,
+      "step": 10847
+    },
+    {
+      "epoch": 0.9692854111287332,
+      "grad_norm": 0.5831353664398193,
+      "learning_rate": 2.4615670401514356e-07,
+      "loss": 0.9379,
+      "step": 10848
+    },
+    {
+      "epoch": 0.9693747626599951,
+      "grad_norm": 0.45759260654449463,
+      "learning_rate": 2.4472465055426217e-07,
+      "loss": 0.9993,
+      "step": 10849
+    },
+    {
+      "epoch": 0.969464114191257,
+      "grad_norm": 0.6619382500648499,
+      "learning_rate": 2.432967646280426e-07,
+      "loss": 0.8364,
+      "step": 10850
+    },
+    {
+      "epoch": 0.9695534657225188,
+      "grad_norm": 0.3795022964477539,
+      "learning_rate": 2.4187304635608923e-07,
+      "loss": 0.9585,
+      "step": 10851
+    },
+    {
+      "epoch": 0.9696428172537807,
+      "grad_norm": 0.5020806193351746,
+      "learning_rate": 2.4045349585765097e-07,
+      "loss": 0.9079,
+      "step": 10852
+    },
+    {
+      "epoch": 0.9697321687850425,
+      "grad_norm": 0.4140533208847046,
+      "learning_rate": 2.3903811325163283e-07,
+      "loss": 0.9358,
+      "step": 10853
+    },
+    {
+      "epoch": 0.9698215203163044,
+      "grad_norm": 0.6230595707893372,
+      "learning_rate": 2.3762689865658438e-07,
+      "loss": 0.8465,
+      "step": 10854
+    },
+    {
+      "epoch": 0.9699108718475663,
+      "grad_norm": 0.5708101391792297,
+      "learning_rate": 2.3621985219071108e-07,
+      "loss": 0.8622,
+      "step": 10855
+    },
+    {
+      "epoch": 0.9700002233788282,
+      "grad_norm": 0.5391978025436401,
+      "learning_rate": 2.3481697397187418e-07,
+      "loss": 0.905,
+      "step": 10856
+    },
+    {
+      "epoch": 0.9700895749100901,
+      "grad_norm": 0.4433526396751404,
+      "learning_rate": 2.334182641175686e-07,
+      "loss": 0.9798,
+      "step": 10857
+    },
+    {
+      "epoch": 0.9701789264413518,
+      "grad_norm": 0.669391393661499,
+      "learning_rate": 2.320237227449562e-07,
+      "loss": 0.8246,
+      "step": 10858
+    },
+    {
+      "epoch": 0.9702682779726137,
+      "grad_norm": 0.47016438841819763,
+      "learning_rate": 2.3063334997084907e-07,
+      "loss": 0.8787,
+      "step": 10859
+    },
+    {
+      "epoch": 0.9703576295038756,
+      "grad_norm": 0.5404382944107056,
+      "learning_rate": 2.2924714591170403e-07,
+      "loss": 0.9036,
+      "step": 10860
+    },
+    {
+      "epoch": 0.9704469810351375,
+      "grad_norm": 0.44831082224845886,
+      "learning_rate": 2.2786511068362826e-07,
+      "loss": 0.9216,
+      "step": 10861
+    },
+    {
+      "epoch": 0.9705363325663994,
+      "grad_norm": 0.44715404510498047,
+      "learning_rate": 2.2648724440237913e-07,
+      "loss": 0.9903,
+      "step": 10862
+    },
+    {
+      "epoch": 0.9706256840976613,
+      "grad_norm": 0.5612984895706177,
+      "learning_rate": 2.2511354718336986e-07,
+      "loss": 0.8568,
+      "step": 10863
+    },
+    {
+      "epoch": 0.9707150356289231,
+      "grad_norm": 0.45841625332832336,
+      "learning_rate": 2.2374401914166953e-07,
+      "loss": 0.9588,
+      "step": 10864
+    },
+    {
+      "epoch": 0.9708043871601849,
+      "grad_norm": 0.4593586325645447,
+      "learning_rate": 2.2237866039198085e-07,
+      "loss": 0.9923,
+      "step": 10865
+    },
+    {
+      "epoch": 0.9708937386914468,
+      "grad_norm": 0.46852678060531616,
+      "learning_rate": 2.2101747104866788e-07,
+      "loss": 0.9463,
+      "step": 10866
+    },
+    {
+      "epoch": 0.9709830902227087,
+      "grad_norm": 0.4527819752693176,
+      "learning_rate": 2.1966045122575052e-07,
+      "loss": 0.9759,
+      "step": 10867
+    },
+    {
+      "epoch": 0.9710724417539706,
+      "grad_norm": 0.42089274525642395,
+      "learning_rate": 2.1830760103688784e-07,
+      "loss": 0.9125,
+      "step": 10868
+    },
+    {
+      "epoch": 0.9711617932852324,
+      "grad_norm": 0.4603272080421448,
+      "learning_rate": 2.1695892059540035e-07,
+      "loss": 0.9218,
+      "step": 10869
+    },
+    {
+      "epoch": 0.9712511448164943,
+      "grad_norm": 0.444743275642395,
+      "learning_rate": 2.156144100142532e-07,
+      "loss": 0.9502,
+      "step": 10870
+    },
+    {
+      "epoch": 0.9713404963477562,
+      "grad_norm": 0.5526580810546875,
+      "learning_rate": 2.1427406940606187e-07,
+      "loss": 0.9189,
+      "step": 10871
+    },
+    {
+      "epoch": 0.971429847879018,
+      "grad_norm": 0.4762706160545349,
+      "learning_rate": 2.1293789888309212e-07,
+      "loss": 0.9151,
+      "step": 10872
+    },
+    {
+      "epoch": 0.9715191994102799,
+      "grad_norm": 0.5218154191970825,
+      "learning_rate": 2.116058985572711e-07,
+      "loss": 0.8963,
+      "step": 10873
+    },
+    {
+      "epoch": 0.9716085509415417,
+      "grad_norm": 0.5006052851676941,
+      "learning_rate": 2.1027806854015954e-07,
+      "loss": 0.9783,
+      "step": 10874
+    },
+    {
+      "epoch": 0.9716979024728036,
+      "grad_norm": 0.45393121242523193,
+      "learning_rate": 2.0895440894297402e-07,
+      "loss": 0.9236,
+      "step": 10875
+    },
+    {
+      "epoch": 0.9717872540040655,
+      "grad_norm": 0.5509636998176575,
+      "learning_rate": 2.0763491987659812e-07,
+      "loss": 0.9322,
+      "step": 10876
+    },
+    {
+      "epoch": 0.9718766055353274,
+      "grad_norm": 0.49217623472213745,
+      "learning_rate": 2.0631960145154338e-07,
+      "loss": 0.9342,
+      "step": 10877
+    },
+    {
+      "epoch": 0.9719659570665893,
+      "grad_norm": 0.48069292306900024,
+      "learning_rate": 2.050084537779884e-07,
+      "loss": 0.9461,
+      "step": 10878
+    },
+    {
+      "epoch": 0.972055308597851,
+      "grad_norm": 0.4449180066585541,
+      "learning_rate": 2.0370147696574526e-07,
+      "loss": 0.9321,
+      "step": 10879
+    },
+    {
+      "epoch": 0.9721446601291129,
+      "grad_norm": 0.5779398083686829,
+      "learning_rate": 2.0239867112429868e-07,
+      "loss": 0.9169,
+      "step": 10880
+    },
+    {
+      "epoch": 0.9722340116603748,
+      "grad_norm": 0.5261817574501038,
+      "learning_rate": 2.0110003636276687e-07,
+      "loss": 0.9289,
+      "step": 10881
+    },
+    {
+      "epoch": 0.9723233631916367,
+      "grad_norm": 0.5714827179908752,
+      "learning_rate": 1.9980557278992397e-07,
+      "loss": 0.8583,
+      "step": 10882
+    },
+    {
+      "epoch": 0.9724127147228986,
+      "grad_norm": 0.5161953568458557,
+      "learning_rate": 1.9851528051419988e-07,
+      "loss": 0.9715,
+      "step": 10883
+    },
+    {
+      "epoch": 0.9725020662541605,
+      "grad_norm": 0.5811159610748291,
+      "learning_rate": 1.9722915964366372e-07,
+      "loss": 0.8745,
+      "step": 10884
+    },
+    {
+      "epoch": 0.9725914177854224,
+      "grad_norm": 0.4967767000198364,
+      "learning_rate": 1.95947210286046e-07,
+      "loss": 0.9639,
+      "step": 10885
+    },
+    {
+      "epoch": 0.9726807693166841,
+      "grad_norm": 0.538653552532196,
+      "learning_rate": 1.9466943254872193e-07,
+      "loss": 0.9367,
+      "step": 10886
+    },
+    {
+      "epoch": 0.972770120847946,
+      "grad_norm": 0.5293512344360352,
+      "learning_rate": 1.9339582653871703e-07,
+      "loss": 0.9583,
+      "step": 10887
+    },
+    {
+      "epoch": 0.9728594723792079,
+      "grad_norm": 0.4885563254356384,
+      "learning_rate": 1.9212639236271256e-07,
+      "loss": 0.8537,
+      "step": 10888
+    },
+    {
+      "epoch": 0.9729488239104698,
+      "grad_norm": 0.5381324291229248,
+      "learning_rate": 1.908611301270402e-07,
+      "loss": 0.9268,
+      "step": 10889
+    },
+    {
+      "epoch": 0.9730381754417317,
+      "grad_norm": 0.550106406211853,
+      "learning_rate": 1.8960003993767073e-07,
+      "loss": 0.9345,
+      "step": 10890
+    },
+    {
+      "epoch": 0.9731275269729935,
+      "grad_norm": 0.534343421459198,
+      "learning_rate": 1.8834312190024183e-07,
+      "loss": 0.8775,
+      "step": 10891
+    },
+    {
+      "epoch": 0.9732168785042553,
+      "grad_norm": 0.5603974461555481,
+      "learning_rate": 1.8709037612003045e-07,
+      "loss": 0.9408,
+      "step": 10892
+    },
+    {
+      "epoch": 0.9733062300355172,
+      "grad_norm": 0.3925410509109497,
+      "learning_rate": 1.8584180270196926e-07,
+      "loss": 0.9445,
+      "step": 10893
+    },
+    {
+      "epoch": 0.9733955815667791,
+      "grad_norm": 0.44486695528030396,
+      "learning_rate": 1.8459740175063577e-07,
+      "loss": 0.9394,
+      "step": 10894
+    },
+    {
+      "epoch": 0.973484933098041,
+      "grad_norm": 0.5443863868713379,
+      "learning_rate": 1.8335717337026326e-07,
+      "loss": 0.9071,
+      "step": 10895
+    },
+    {
+      "epoch": 0.9735742846293028,
+      "grad_norm": 0.504984974861145,
+      "learning_rate": 1.8212111766473528e-07,
+      "loss": 0.8573,
+      "step": 10896
+    },
+    {
+      "epoch": 0.9736636361605647,
+      "grad_norm": 0.49120062589645386,
+      "learning_rate": 1.8088923473758568e-07,
+      "loss": 0.929,
+      "step": 10897
+    },
+    {
+      "epoch": 0.9737529876918266,
+      "grad_norm": 0.4665651321411133,
+      "learning_rate": 1.7966152469199305e-07,
+      "loss": 0.9986,
+      "step": 10898
+    },
+    {
+      "epoch": 0.9738423392230884,
+      "grad_norm": 0.46804356575012207,
+      "learning_rate": 1.7843798763079733e-07,
+      "loss": 0.9408,
+      "step": 10899
+    },
+    {
+      "epoch": 0.9739316907543503,
+      "grad_norm": 0.49248629808425903,
+      "learning_rate": 1.7721862365647767e-07,
+      "loss": 0.8993,
+      "step": 10900
+    },
+    {
+      "epoch": 0.9740210422856121,
+      "grad_norm": 0.5523046255111694,
+      "learning_rate": 1.7600343287116904e-07,
+      "loss": 0.9051,
+      "step": 10901
+    },
+    {
+      "epoch": 0.974110393816874,
+      "grad_norm": 0.46335333585739136,
+      "learning_rate": 1.7479241537666225e-07,
+      "loss": 0.9142,
+      "step": 10902
+    },
+    {
+      "epoch": 0.9741997453481359,
+      "grad_norm": 0.4312766194343567,
+      "learning_rate": 1.7358557127438723e-07,
+      "loss": 0.9891,
+      "step": 10903
+    },
+    {
+      "epoch": 0.9742890968793978,
+      "grad_norm": 0.4239676594734192,
+      "learning_rate": 1.7238290066543538e-07,
+      "loss": 0.9441,
+      "step": 10904
+    },
+    {
+      "epoch": 0.9743784484106597,
+      "grad_norm": 0.4999157190322876,
+      "learning_rate": 1.7118440365053722e-07,
+      "loss": 0.9123,
+      "step": 10905
+    },
+    {
+      "epoch": 0.9744677999419215,
+      "grad_norm": 0.5276448130607605,
+      "learning_rate": 1.6999008033007913e-07,
+      "loss": 0.8139,
+      "step": 10906
+    },
+    {
+      "epoch": 0.9745571514731833,
+      "grad_norm": 0.44836434721946716,
+      "learning_rate": 1.6879993080410327e-07,
+      "loss": 0.9032,
+      "step": 10907
+    },
+    {
+      "epoch": 0.9746465030044452,
+      "grad_norm": 0.49218374490737915,
+      "learning_rate": 1.6761395517230215e-07,
+      "loss": 1.0147,
+      "step": 10908
+    },
+    {
+      "epoch": 0.9747358545357071,
+      "grad_norm": 0.4299551248550415,
+      "learning_rate": 1.6643215353400188e-07,
+      "loss": 0.9801,
+      "step": 10909
+    },
+    {
+      "epoch": 0.974825206066969,
+      "grad_norm": 0.4688587784767151,
+      "learning_rate": 1.6525452598819547e-07,
+      "loss": 0.8839,
+      "step": 10910
+    },
+    {
+      "epoch": 0.9749145575982309,
+      "grad_norm": 0.509926438331604,
+      "learning_rate": 1.6408107263352069e-07,
+      "loss": 0.9189,
+      "step": 10911
+    },
+    {
+      "epoch": 0.9750039091294928,
+      "grad_norm": 0.6090724468231201,
+      "learning_rate": 1.6291179356827114e-07,
+      "loss": 0.9022,
+      "step": 10912
+    },
+    {
+      "epoch": 0.9750932606607545,
+      "grad_norm": 0.5029346942901611,
+      "learning_rate": 1.617466888903907e-07,
+      "loss": 0.9822,
+      "step": 10913
+    },
+    {
+      "epoch": 0.9751826121920164,
+      "grad_norm": 0.594288170337677,
+      "learning_rate": 1.6058575869745686e-07,
+      "loss": 0.8495,
+      "step": 10914
+    },
+    {
+      "epoch": 0.9752719637232783,
+      "grad_norm": 0.4893123507499695,
+      "learning_rate": 1.5942900308671405e-07,
+      "loss": 0.9206,
+      "step": 10915
+    },
+    {
+      "epoch": 0.9753613152545402,
+      "grad_norm": 0.4604051113128662,
+      "learning_rate": 1.58276422155057e-07,
+      "loss": 0.9293,
+      "step": 10916
+    },
+    {
+      "epoch": 0.9754506667858021,
+      "grad_norm": 0.4965328872203827,
+      "learning_rate": 1.5712801599902515e-07,
+      "loss": 0.8888,
+      "step": 10917
+    },
+    {
+      "epoch": 0.9755400183170639,
+      "grad_norm": 0.47335171699523926,
+      "learning_rate": 1.5598378471480267e-07,
+      "loss": 0.9663,
+      "step": 10918
+    },
+    {
+      "epoch": 0.9756293698483258,
+      "grad_norm": 0.4554191827774048,
+      "learning_rate": 1.548437283982407e-07,
+      "loss": 0.9362,
+      "step": 10919
+    },
+    {
+      "epoch": 0.9757187213795876,
+      "grad_norm": 0.4438236951828003,
+      "learning_rate": 1.5370784714482944e-07,
+      "loss": 0.9775,
+      "step": 10920
+    },
+    {
+      "epoch": 0.9758080729108495,
+      "grad_norm": 0.5050124526023865,
+      "learning_rate": 1.5257614104970952e-07,
+      "loss": 0.9673,
+      "step": 10921
+    },
+    {
+      "epoch": 0.9758974244421114,
+      "grad_norm": 0.4388897716999054,
+      "learning_rate": 1.5144861020767176e-07,
+      "loss": 0.9194,
+      "step": 10922
+    },
+    {
+      "epoch": 0.9759867759733732,
+      "grad_norm": 0.6075052618980408,
+      "learning_rate": 1.5032525471316284e-07,
+      "loss": 0.8963,
+      "step": 10923
+    },
+    {
+      "epoch": 0.9760761275046351,
+      "grad_norm": 0.4434504806995392,
+      "learning_rate": 1.4920607466026858e-07,
+      "loss": 0.9343,
+      "step": 10924
+    },
+    {
+      "epoch": 0.976165479035897,
+      "grad_norm": 0.4541495442390442,
+      "learning_rate": 1.4809107014274182e-07,
+      "loss": 0.9642,
+      "step": 10925
+    },
+    {
+      "epoch": 0.9762548305671589,
+      "grad_norm": 0.49039971828460693,
+      "learning_rate": 1.4698024125396892e-07,
+      "loss": 0.9018,
+      "step": 10926
+    },
+    {
+      "epoch": 0.9763441820984207,
+      "grad_norm": 0.46627819538116455,
+      "learning_rate": 1.4587358808699215e-07,
+      "loss": 1.0147,
+      "step": 10927
+    },
+    {
+      "epoch": 0.9764335336296825,
+      "grad_norm": 0.4772813618183136,
+      "learning_rate": 1.447711107345151e-07,
+      "loss": 0.9074,
+      "step": 10928
+    },
+    {
+      "epoch": 0.9765228851609444,
+      "grad_norm": 0.606759250164032,
+      "learning_rate": 1.436728092888695e-07,
+      "loss": 0.8712,
+      "step": 10929
+    },
+    {
+      "epoch": 0.9766122366922063,
+      "grad_norm": 0.4552794098854065,
+      "learning_rate": 1.4257868384206508e-07,
+      "loss": 0.9339,
+      "step": 10930
+    },
+    {
+      "epoch": 0.9767015882234682,
+      "grad_norm": 0.5520709753036499,
+      "learning_rate": 1.4148873448573408e-07,
+      "loss": 0.869,
+      "step": 10931
+    },
+    {
+      "epoch": 0.9767909397547301,
+      "grad_norm": 0.49193331599235535,
+      "learning_rate": 1.4040296131117013e-07,
+      "loss": 0.9923,
+      "step": 10932
+    },
+    {
+      "epoch": 0.976880291285992,
+      "grad_norm": 0.5167152881622314,
+      "learning_rate": 1.393213644093283e-07,
+      "loss": 0.8921,
+      "step": 10933
+    },
+    {
+      "epoch": 0.9769696428172537,
+      "grad_norm": 0.5435425639152527,
+      "learning_rate": 1.382439438707972e-07,
+      "loss": 0.9036,
+      "step": 10934
+    },
+    {
+      "epoch": 0.9770589943485156,
+      "grad_norm": 0.5024846196174622,
+      "learning_rate": 1.3717069978582687e-07,
+      "loss": 0.899,
+      "step": 10935
+    },
+    {
+      "epoch": 0.9771483458797775,
+      "grad_norm": 0.4377569854259491,
+      "learning_rate": 1.361016322443065e-07,
+      "loss": 0.9288,
+      "step": 10936
+    },
+    {
+      "epoch": 0.9772376974110394,
+      "grad_norm": 0.549490213394165,
+      "learning_rate": 1.350367413357867e-07,
+      "loss": 0.9464,
+      "step": 10937
+    },
+    {
+      "epoch": 0.9773270489423013,
+      "grad_norm": 0.5465390086174011,
+      "learning_rate": 1.3397602714946278e-07,
+      "loss": 0.7978,
+      "step": 10938
+    },
+    {
+      "epoch": 0.9774164004735632,
+      "grad_norm": 0.4476923942565918,
+      "learning_rate": 1.3291948977418033e-07,
+      "loss": 0.9909,
+      "step": 10939
+    },
+    {
+      "epoch": 0.977505752004825,
+      "grad_norm": 0.4725693464279175,
+      "learning_rate": 1.318671292984297e-07,
+      "loss": 0.9513,
+      "step": 10940
+    },
+    {
+      "epoch": 0.9775951035360868,
+      "grad_norm": 0.4664715528488159,
+      "learning_rate": 1.3081894581036813e-07,
+      "loss": 0.9367,
+      "step": 10941
+    },
+    {
+      "epoch": 0.9776844550673487,
+      "grad_norm": 0.5551668405532837,
+      "learning_rate": 1.297749393977865e-07,
+      "loss": 0.9184,
+      "step": 10942
+    },
+    {
+      "epoch": 0.9777738065986106,
+      "grad_norm": 0.44665080308914185,
+      "learning_rate": 1.2873511014813155e-07,
+      "loss": 0.9797,
+      "step": 10943
+    },
+    {
+      "epoch": 0.9778631581298725,
+      "grad_norm": 0.5233182907104492,
+      "learning_rate": 1.2769945814850582e-07,
+      "loss": 0.9347,
+      "step": 10944
+    },
+    {
+      "epoch": 0.9779525096611343,
+      "grad_norm": 0.48261916637420654,
+      "learning_rate": 1.2666798348564546e-07,
+      "loss": 0.9018,
+      "step": 10945
+    },
+    {
+      "epoch": 0.9780418611923962,
+      "grad_norm": 0.5972064733505249,
+      "learning_rate": 1.256406862459536e-07,
+      "loss": 0.9394,
+      "step": 10946
+    },
+    {
+      "epoch": 0.9781312127236581,
+      "grad_norm": 0.49780553579330444,
+      "learning_rate": 1.2461756651547807e-07,
+      "loss": 0.9478,
+      "step": 10947
+    },
+    {
+      "epoch": 0.9782205642549199,
+      "grad_norm": 0.48487037420272827,
+      "learning_rate": 1.2359862437991144e-07,
+      "loss": 0.8759,
+      "step": 10948
+    },
+    {
+      "epoch": 0.9783099157861818,
+      "grad_norm": 0.6491232514381409,
+      "learning_rate": 1.2258385992460764e-07,
+      "loss": 0.8842,
+      "step": 10949
+    },
+    {
+      "epoch": 0.9783992673174436,
+      "grad_norm": 0.4678862690925598,
+      "learning_rate": 1.215732732345598e-07,
+      "loss": 0.9332,
+      "step": 10950
+    },
+    {
+      "epoch": 0.9784886188487055,
+      "grad_norm": 0.4849478304386139,
+      "learning_rate": 1.205668643944169e-07,
+      "loss": 0.9802,
+      "step": 10951
+    },
+    {
+      "epoch": 0.9785779703799674,
+      "grad_norm": 0.5359340310096741,
+      "learning_rate": 1.195646334884726e-07,
+      "loss": 0.9359,
+      "step": 10952
+    },
+    {
+      "epoch": 0.9786673219112293,
+      "grad_norm": 0.49691978096961975,
+      "learning_rate": 1.1856658060068193e-07,
+      "loss": 0.9549,
+      "step": 10953
+    },
+    {
+      "epoch": 0.9787566734424911,
+      "grad_norm": 0.6014655232429504,
+      "learning_rate": 1.1757270581463364e-07,
+      "loss": 0.9857,
+      "step": 10954
+    },
+    {
+      "epoch": 0.978846024973753,
+      "grad_norm": 0.4421977400779724,
+      "learning_rate": 1.1658300921358334e-07,
+      "loss": 0.9145,
+      "step": 10955
+    },
+    {
+      "epoch": 0.9789353765050148,
+      "grad_norm": 0.552693247795105,
+      "learning_rate": 1.1559749088042026e-07,
+      "loss": 0.8431,
+      "step": 10956
+    },
+    {
+      "epoch": 0.9790247280362767,
+      "grad_norm": 0.46435338258743286,
+      "learning_rate": 1.1461615089770062e-07,
+      "loss": 0.9223,
+      "step": 10957
+    },
+    {
+      "epoch": 0.9791140795675386,
+      "grad_norm": 0.4661587178707123,
+      "learning_rate": 1.1363898934761974e-07,
+      "loss": 0.9771,
+      "step": 10958
+    },
+    {
+      "epoch": 0.9792034310988005,
+      "grad_norm": 0.5485957264900208,
+      "learning_rate": 1.1266600631202328e-07,
+      "loss": 0.9378,
+      "step": 10959
+    },
+    {
+      "epoch": 0.9792927826300624,
+      "grad_norm": 0.5378150939941406,
+      "learning_rate": 1.1169720187240718e-07,
+      "loss": 0.9114,
+      "step": 10960
+    },
+    {
+      "epoch": 0.9793821341613241,
+      "grad_norm": 0.5579792261123657,
+      "learning_rate": 1.1073257610991761e-07,
+      "loss": 0.9124,
+      "step": 10961
+    },
+    {
+      "epoch": 0.979471485692586,
+      "grad_norm": 0.3841468393802643,
+      "learning_rate": 1.0977212910536217e-07,
+      "loss": 1.0001,
+      "step": 10962
+    },
+    {
+      "epoch": 0.9795608372238479,
+      "grad_norm": 0.4615533649921417,
+      "learning_rate": 1.0881586093918205e-07,
+      "loss": 0.918,
+      "step": 10963
+    },
+    {
+      "epoch": 0.9796501887551098,
+      "grad_norm": 0.46879342198371887,
+      "learning_rate": 1.0786377169147432e-07,
+      "loss": 0.982,
+      "step": 10964
+    },
+    {
+      "epoch": 0.9797395402863717,
+      "grad_norm": 0.41505542397499084,
+      "learning_rate": 1.0691586144199184e-07,
+      "loss": 0.9517,
+      "step": 10965
+    },
+    {
+      "epoch": 0.9798288918176336,
+      "grad_norm": 0.5566489696502686,
+      "learning_rate": 1.059721302701211e-07,
+      "loss": 0.8337,
+      "step": 10966
+    },
+    {
+      "epoch": 0.9799182433488954,
+      "grad_norm": 0.5348109602928162,
+      "learning_rate": 1.0503257825492108e-07,
+      "loss": 1.0199,
+      "step": 10967
+    },
+    {
+      "epoch": 0.9800075948801572,
+      "grad_norm": 0.47354453802108765,
+      "learning_rate": 1.040972054750844e-07,
+      "loss": 0.9287,
+      "step": 10968
+    },
+    {
+      "epoch": 0.9800969464114191,
+      "grad_norm": 0.5585243105888367,
+      "learning_rate": 1.0316601200895948e-07,
+      "loss": 0.9726,
+      "step": 10969
+    },
+    {
+      "epoch": 0.980186297942681,
+      "grad_norm": 0.5595225095748901,
+      "learning_rate": 1.022389979345395e-07,
+      "loss": 0.8816,
+      "step": 10970
+    },
+    {
+      "epoch": 0.9802756494739429,
+      "grad_norm": 0.5292766094207764,
+      "learning_rate": 1.0131616332947346e-07,
+      "loss": 0.9289,
+      "step": 10971
+    },
+    {
+      "epoch": 0.9803650010052047,
+      "grad_norm": 0.42903298139572144,
+      "learning_rate": 1.003975082710662e-07,
+      "loss": 0.9523,
+      "step": 10972
+    },
+    {
+      "epoch": 0.9804543525364666,
+      "grad_norm": 0.43353918194770813,
+      "learning_rate": 9.948303283625615e-08,
+      "loss": 0.9218,
+      "step": 10973
+    },
+    {
+      "epoch": 0.9805437040677285,
+      "grad_norm": 0.6073578596115112,
+      "learning_rate": 9.857273710164871e-08,
+      "loss": 0.9989,
+      "step": 10974
+    },
+    {
+      "epoch": 0.9806330555989903,
+      "grad_norm": 0.4935568869113922,
+      "learning_rate": 9.766662114347736e-08,
+      "loss": 0.9325,
+      "step": 10975
+    },
+    {
+      "epoch": 0.9807224071302522,
+      "grad_norm": 0.6012499332427979,
+      "learning_rate": 9.676468503765357e-08,
+      "loss": 0.9096,
+      "step": 10976
+    },
+    {
+      "epoch": 0.980811758661514,
+      "grad_norm": 0.5922300219535828,
+      "learning_rate": 9.586692885971137e-08,
+      "loss": 0.8704,
+      "step": 10977
+    },
+    {
+      "epoch": 0.9809011101927759,
+      "grad_norm": 0.4871740937232971,
+      "learning_rate": 9.49733526848573e-08,
+      "loss": 0.901,
+      "step": 10978
+    },
+    {
+      "epoch": 0.9809904617240378,
+      "grad_norm": 0.5010733604431152,
+      "learning_rate": 9.4083956587937e-08,
+      "loss": 0.9744,
+      "step": 10979
+    },
+    {
+      "epoch": 0.9810798132552997,
+      "grad_norm": 0.5309439301490784,
+      "learning_rate": 9.319874064344092e-08,
+      "loss": 1.0236,
+      "step": 10980
+    },
+    {
+      "epoch": 0.9811691647865616,
+      "grad_norm": 0.542961597442627,
+      "learning_rate": 9.231770492552084e-08,
+      "loss": 0.9089,
+      "step": 10981
+    },
+    {
+      "epoch": 0.9812585163178233,
+      "grad_norm": 0.5267024636268616,
+      "learning_rate": 9.144084950796772e-08,
+      "loss": 0.9133,
+      "step": 10982
+    },
+    {
+      "epoch": 0.9813478678490852,
+      "grad_norm": 0.48576876521110535,
+      "learning_rate": 9.056817446422839e-08,
+      "loss": 0.9561,
+      "step": 10983
+    },
+    {
+      "epoch": 0.9814372193803471,
+      "grad_norm": 0.4452846944332123,
+      "learning_rate": 8.969967986740546e-08,
+      "loss": 1.0614,
+      "step": 10984
+    },
+    {
+      "epoch": 0.981526570911609,
+      "grad_norm": 0.5701860785484314,
+      "learning_rate": 8.883536579024077e-08,
+      "loss": 0.8198,
+      "step": 10985
+    },
+    {
+      "epoch": 0.9816159224428709,
+      "grad_norm": 0.4147273898124695,
+      "learning_rate": 8.797523230512639e-08,
+      "loss": 0.9686,
+      "step": 10986
+    },
+    {
+      "epoch": 0.9817052739741328,
+      "grad_norm": 0.5088819861412048,
+      "learning_rate": 8.711927948411025e-08,
+      "loss": 0.913,
+      "step": 10987
+    },
+    {
+      "epoch": 0.9817946255053946,
+      "grad_norm": 0.5437602996826172,
+      "learning_rate": 8.626750739888501e-08,
+      "loss": 1.0007,
+      "step": 10988
+    },
+    {
+      "epoch": 0.9818839770366564,
+      "grad_norm": 0.6010596752166748,
+      "learning_rate": 8.541991612080469e-08,
+      "loss": 0.8717,
+      "step": 10989
+    },
+    {
+      "epoch": 0.9819733285679183,
+      "grad_norm": 0.44414857029914856,
+      "learning_rate": 8.457650572085141e-08,
+      "loss": 0.9774,
+      "step": 10990
+    },
+    {
+      "epoch": 0.9820626800991802,
+      "grad_norm": 0.4728522002696991,
+      "learning_rate": 8.373727626967975e-08,
+      "loss": 0.9735,
+      "step": 10991
+    },
+    {
+      "epoch": 0.9821520316304421,
+      "grad_norm": 0.543158233165741,
+      "learning_rate": 8.290222783757795e-08,
+      "loss": 0.8637,
+      "step": 10992
+    },
+    {
+      "epoch": 0.982241383161704,
+      "grad_norm": 0.4106288254261017,
+      "learning_rate": 8.20713604944956e-08,
+      "loss": 0.9912,
+      "step": 10993
+    },
+    {
+      "epoch": 0.9823307346929658,
+      "grad_norm": 0.40964820981025696,
+      "learning_rate": 8.124467431002148e-08,
+      "loss": 0.9706,
+      "step": 10994
+    },
+    {
+      "epoch": 0.9824200862242277,
+      "grad_norm": 0.4523217976093292,
+      "learning_rate": 8.042216935340019e-08,
+      "loss": 0.9801,
+      "step": 10995
+    },
+    {
+      "epoch": 0.9825094377554895,
+      "grad_norm": 0.5060840249061584,
+      "learning_rate": 7.960384569353219e-08,
+      "loss": 0.9665,
+      "step": 10996
+    },
+    {
+      "epoch": 0.9825987892867514,
+      "grad_norm": 0.5814510583877563,
+      "learning_rate": 7.878970339894598e-08,
+      "loss": 0.9288,
+      "step": 10997
+    },
+    {
+      "epoch": 0.9826881408180133,
+      "grad_norm": 0.45717230439186096,
+      "learning_rate": 7.797974253785367e-08,
+      "loss": 0.9257,
+      "step": 10998
+    },
+    {
+      "epoch": 0.9827774923492751,
+      "grad_norm": 0.3922685384750366,
+      "learning_rate": 7.717396317808433e-08,
+      "loss": 0.9438,
+      "step": 10999
+    },
+    {
+      "epoch": 0.982866843880537,
+      "grad_norm": 0.5810117125511169,
+      "learning_rate": 7.637236538713399e-08,
+      "loss": 0.8851,
+      "step": 11000
+    },
+    {
+      "epoch": 0.9829561954117989,
+      "grad_norm": 0.5308172106742859,
+      "learning_rate": 7.557494923214337e-08,
+      "loss": 0.8542,
+      "step": 11001
+    },
+    {
+      "epoch": 0.9830455469430608,
+      "grad_norm": 0.48735228180885315,
+      "learning_rate": 7.478171477990902e-08,
+      "loss": 0.9284,
+      "step": 11002
+    },
+    {
+      "epoch": 0.9831348984743226,
+      "grad_norm": 0.5343984365463257,
+      "learning_rate": 7.399266209687228e-08,
+      "loss": 0.9664,
+      "step": 11003
+    },
+    {
+      "epoch": 0.9832242500055844,
+      "grad_norm": 0.40684744715690613,
+      "learning_rate": 7.32077912491247e-08,
+      "loss": 0.9746,
+      "step": 11004
+    },
+    {
+      "epoch": 0.9833136015368463,
+      "grad_norm": 0.4683247208595276,
+      "learning_rate": 7.242710230240257e-08,
+      "loss": 0.9381,
+      "step": 11005
+    },
+    {
+      "epoch": 0.9834029530681082,
+      "grad_norm": 0.4208630323410034,
+      "learning_rate": 7.165059532210361e-08,
+      "loss": 0.9214,
+      "step": 11006
+    },
+    {
+      "epoch": 0.9834923045993701,
+      "grad_norm": 0.45483431220054626,
+      "learning_rate": 7.087827037325912e-08,
+      "loss": 0.9788,
+      "step": 11007
+    },
+    {
+      "epoch": 0.983581656130632,
+      "grad_norm": 0.6926810145378113,
+      "learning_rate": 7.011012752056733e-08,
+      "loss": 0.826,
+      "step": 11008
+    },
+    {
+      "epoch": 0.9836710076618939,
+      "grad_norm": 0.4510803818702698,
+      "learning_rate": 6.934616682837125e-08,
+      "loss": 0.9017,
+      "step": 11009
+    },
+    {
+      "epoch": 0.9837603591931556,
+      "grad_norm": 0.5115488171577454,
+      "learning_rate": 6.8586388360653e-08,
+      "loss": 0.8919,
+      "step": 11010
+    },
+    {
+      "epoch": 0.9838497107244175,
+      "grad_norm": 0.5512319207191467,
+      "learning_rate": 6.783079218105614e-08,
+      "loss": 0.9227,
+      "step": 11011
+    },
+    {
+      "epoch": 0.9839390622556794,
+      "grad_norm": 0.5396715998649597,
+      "learning_rate": 6.707937835286893e-08,
+      "loss": 0.9214,
+      "step": 11012
+    },
+    {
+      "epoch": 0.9840284137869413,
+      "grad_norm": 0.5042988061904907,
+      "learning_rate": 6.63321469390299e-08,
+      "loss": 1.0184,
+      "step": 11013
+    },
+    {
+      "epoch": 0.9841177653182032,
+      "grad_norm": 0.5353363752365112,
+      "learning_rate": 6.558909800212787e-08,
+      "loss": 0.929,
+      "step": 11014
+    },
+    {
+      "epoch": 0.984207116849465,
+      "grad_norm": 0.5171821117401123,
+      "learning_rate": 6.485023160440195e-08,
+      "loss": 0.8725,
+      "step": 11015
+    },
+    {
+      "epoch": 0.9842964683807269,
+      "grad_norm": 0.44513246417045593,
+      "learning_rate": 6.411554780774154e-08,
+      "loss": 0.9139,
+      "step": 11016
+    },
+    {
+      "epoch": 0.9843858199119887,
+      "grad_norm": 0.4491952061653137,
+      "learning_rate": 6.338504667368072e-08,
+      "loss": 0.9752,
+      "step": 11017
+    },
+    {
+      "epoch": 0.9844751714432506,
+      "grad_norm": 0.47702059149742126,
+      "learning_rate": 6.265872826340946e-08,
+      "loss": 0.9612,
+      "step": 11018
+    },
+    {
+      "epoch": 0.9845645229745125,
+      "grad_norm": 0.6224638223648071,
+      "learning_rate": 6.193659263776242e-08,
+      "loss": 0.9551,
+      "step": 11019
+    },
+    {
+      "epoch": 0.9846538745057744,
+      "grad_norm": 0.521059513092041,
+      "learning_rate": 6.121863985722454e-08,
+      "loss": 0.9127,
+      "step": 11020
+    },
+    {
+      "epoch": 0.9847432260370362,
+      "grad_norm": 0.4040971100330353,
+      "learning_rate": 6.05048699819366e-08,
+      "loss": 0.9459,
+      "step": 11021
+    },
+    {
+      "epoch": 0.9848325775682981,
+      "grad_norm": 0.6411527395248413,
+      "learning_rate": 5.979528307168414e-08,
+      "loss": 0.9753,
+      "step": 11022
+    },
+    {
+      "epoch": 0.9849219290995599,
+      "grad_norm": 0.469348669052124,
+      "learning_rate": 5.908987918589737e-08,
+      "loss": 0.883,
+      "step": 11023
+    },
+    {
+      "epoch": 0.9850112806308218,
+      "grad_norm": 0.45892199873924255,
+      "learning_rate": 5.838865838366792e-08,
+      "loss": 0.9839,
+      "step": 11024
+    },
+    {
+      "epoch": 0.9851006321620837,
+      "grad_norm": 0.5176712870597839,
+      "learning_rate": 5.769162072373213e-08,
+      "loss": 0.9604,
+      "step": 11025
+    },
+    {
+      "epoch": 0.9851899836933455,
+      "grad_norm": 0.45386213064193726,
+      "learning_rate": 5.699876626446554e-08,
+      "loss": 0.9464,
+      "step": 11026
+    },
+    {
+      "epoch": 0.9852793352246074,
+      "grad_norm": 0.5465309619903564,
+      "learning_rate": 5.6310095063905056e-08,
+      "loss": 0.8722,
+      "step": 11027
+    },
+    {
+      "epoch": 0.9853686867558693,
+      "grad_norm": 0.4664804935455322,
+      "learning_rate": 5.562560717973786e-08,
+      "loss": 0.9386,
+      "step": 11028
+    },
+    {
+      "epoch": 0.9854580382871312,
+      "grad_norm": 0.4405352771282196,
+      "learning_rate": 5.494530266929032e-08,
+      "loss": 0.9666,
+      "step": 11029
+    },
+    {
+      "epoch": 0.985547389818393,
+      "grad_norm": 0.41485795378685,
+      "learning_rate": 5.426918158955574e-08,
+      "loss": 1.0112,
+      "step": 11030
+    },
+    {
+      "epoch": 0.9856367413496548,
+      "grad_norm": 0.45383763313293457,
+      "learning_rate": 5.359724399715549e-08,
+      "loss": 1.0,
+      "step": 11031
+    },
+    {
+      "epoch": 0.9857260928809167,
+      "grad_norm": 0.44838976860046387,
+      "learning_rate": 5.2929489948377876e-08,
+      "loss": 0.926,
+      "step": 11032
+    },
+    {
+      "epoch": 0.9858154444121786,
+      "grad_norm": 0.49654853343963623,
+      "learning_rate": 5.226591949915594e-08,
+      "loss": 0.9731,
+      "step": 11033
+    },
+    {
+      "epoch": 0.9859047959434405,
+      "grad_norm": 0.4356088936328888,
+      "learning_rate": 5.1606532705067436e-08,
+      "loss": 1.0049,
+      "step": 11034
+    },
+    {
+      "epoch": 0.9859941474747024,
+      "grad_norm": 0.5468194484710693,
+      "learning_rate": 5.0951329621340416e-08,
+      "loss": 0.9498,
+      "step": 11035
+    },
+    {
+      "epoch": 0.9860834990059643,
+      "grad_norm": 0.4790467321872711,
+      "learning_rate": 5.0300310302858754e-08,
+      "loss": 0.9816,
+      "step": 11036
+    },
+    {
+      "epoch": 0.986172850537226,
+      "grad_norm": 0.4556531012058258,
+      "learning_rate": 4.965347480415106e-08,
+      "loss": 0.9038,
+      "step": 11037
+    },
+    {
+      "epoch": 0.9862622020684879,
+      "grad_norm": 0.4872002601623535,
+      "learning_rate": 4.901082317940176e-08,
+      "loss": 0.9294,
+      "step": 11038
+    },
+    {
+      "epoch": 0.9863515535997498,
+      "grad_norm": 0.47440412640571594,
+      "learning_rate": 4.837235548242891e-08,
+      "loss": 0.9039,
+      "step": 11039
+    },
+    {
+      "epoch": 0.9864409051310117,
+      "grad_norm": 0.616805374622345,
+      "learning_rate": 4.773807176672307e-08,
+      "loss": 0.9557,
+      "step": 11040
+    },
+    {
+      "epoch": 0.9865302566622736,
+      "grad_norm": 0.6758015751838684,
+      "learning_rate": 4.710797208540285e-08,
+      "loss": 0.8747,
+      "step": 11041
+    },
+    {
+      "epoch": 0.9866196081935354,
+      "grad_norm": 0.5714487433433533,
+      "learning_rate": 4.648205649124826e-08,
+      "loss": 0.9604,
+      "step": 11042
+    },
+    {
+      "epoch": 0.9867089597247973,
+      "grad_norm": 0.5039824843406677,
+      "learning_rate": 4.586032503668958e-08,
+      "loss": 0.949,
+      "step": 11043
+    },
+    {
+      "epoch": 0.9867983112560591,
+      "grad_norm": 0.45489272475242615,
+      "learning_rate": 4.5242777773801816e-08,
+      "loss": 0.9297,
+      "step": 11044
+    },
+    {
+      "epoch": 0.986887662787321,
+      "grad_norm": 0.49669596552848816,
+      "learning_rate": 4.4629414754310264e-08,
+      "loss": 1.0207,
+      "step": 11045
+    },
+    {
+      "epoch": 0.9869770143185829,
+      "grad_norm": 0.4577721059322357,
+      "learning_rate": 4.402023602959604e-08,
+      "loss": 0.9492,
+      "step": 11046
+    },
+    {
+      "epoch": 0.9870663658498448,
+      "grad_norm": 0.4786894619464874,
+      "learning_rate": 4.34152416506739e-08,
+      "loss": 0.9994,
+      "step": 11047
+    },
+    {
+      "epoch": 0.9871557173811066,
+      "grad_norm": 0.49558448791503906,
+      "learning_rate": 4.281443166822552e-08,
+      "loss": 0.928,
+      "step": 11048
+    },
+    {
+      "epoch": 0.9872450689123685,
+      "grad_norm": 0.5031788945198059,
+      "learning_rate": 4.221780613257176e-08,
+      "loss": 0.9041,
+      "step": 11049
+    },
+    {
+      "epoch": 0.9873344204436304,
+      "grad_norm": 0.4300597310066223,
+      "learning_rate": 4.1625365093689306e-08,
+      "loss": 1.0044,
+      "step": 11050
+    },
+    {
+      "epoch": 0.9874237719748922,
+      "grad_norm": 0.5299180746078491,
+      "learning_rate": 4.103710860120513e-08,
+      "loss": 0.9538,
+      "step": 11051
+    },
+    {
+      "epoch": 0.9875131235061541,
+      "grad_norm": 0.4451320469379425,
+      "learning_rate": 4.045303670438538e-08,
+      "loss": 0.9107,
+      "step": 11052
+    },
+    {
+      "epoch": 0.9876024750374159,
+      "grad_norm": 0.4782625138759613,
+      "learning_rate": 3.987314945215204e-08,
+      "loss": 0.962,
+      "step": 11053
+    },
+    {
+      "epoch": 0.9876918265686778,
+      "grad_norm": 0.4866953194141388,
+      "learning_rate": 3.929744689307735e-08,
+      "loss": 0.9365,
+      "step": 11054
+    },
+    {
+      "epoch": 0.9877811780999397,
+      "grad_norm": 0.5117444396018982,
+      "learning_rate": 3.872592907538941e-08,
+      "loss": 0.8583,
+      "step": 11055
+    },
+    {
+      "epoch": 0.9878705296312016,
+      "grad_norm": 0.4728601574897766,
+      "learning_rate": 3.815859604694993e-08,
+      "loss": 0.8894,
+      "step": 11056
+    },
+    {
+      "epoch": 0.9879598811624635,
+      "grad_norm": 0.4523528814315796,
+      "learning_rate": 3.759544785528757e-08,
+      "loss": 1.0043,
+      "step": 11057
+    },
+    {
+      "epoch": 0.9880492326937252,
+      "grad_norm": 0.446228951215744,
+      "learning_rate": 3.7036484547564585e-08,
+      "loss": 0.9598,
+      "step": 11058
+    },
+    {
+      "epoch": 0.9881385842249871,
+      "grad_norm": 0.5366201996803284,
+      "learning_rate": 3.6481706170599094e-08,
+      "loss": 0.8954,
+      "step": 11059
+    },
+    {
+      "epoch": 0.988227935756249,
+      "grad_norm": 0.6366725564002991,
+      "learning_rate": 3.593111277086503e-08,
+      "loss": 0.9405,
+      "step": 11060
+    },
+    {
+      "epoch": 0.9883172872875109,
+      "grad_norm": 0.44401809573173523,
+      "learning_rate": 3.538470439448105e-08,
+      "loss": 0.9049,
+      "step": 11061
+    },
+    {
+      "epoch": 0.9884066388187728,
+      "grad_norm": 0.48204681277275085,
+      "learning_rate": 3.4842481087216104e-08,
+      "loss": 0.8971,
+      "step": 11062
+    },
+    {
+      "epoch": 0.9884959903500347,
+      "grad_norm": 0.4742887616157532,
+      "learning_rate": 3.4304442894478316e-08,
+      "loss": 0.9513,
+      "step": 11063
+    },
+    {
+      "epoch": 0.9885853418812965,
+      "grad_norm": 0.49685031175613403,
+      "learning_rate": 3.377058986134274e-08,
+      "loss": 0.9369,
+      "step": 11064
+    },
+    {
+      "epoch": 0.9886746934125583,
+      "grad_norm": 0.41819649934768677,
+      "learning_rate": 3.324092203251805e-08,
+      "loss": 0.9412,
+      "step": 11065
+    },
+    {
+      "epoch": 0.9887640449438202,
+      "grad_norm": 0.4875706434249878,
+      "learning_rate": 3.271543945237987e-08,
+      "loss": 0.8598,
+      "step": 11066
+    },
+    {
+      "epoch": 0.9888533964750821,
+      "grad_norm": 0.6076412200927734,
+      "learning_rate": 3.219414216493188e-08,
+      "loss": 0.9209,
+      "step": 11067
+    },
+    {
+      "epoch": 0.988942748006344,
+      "grad_norm": 0.5355422496795654,
+      "learning_rate": 3.167703021384471e-08,
+      "loss": 0.9449,
+      "step": 11068
+    },
+    {
+      "epoch": 0.9890320995376058,
+      "grad_norm": 0.48543617129325867,
+      "learning_rate": 3.1164103642428165e-08,
+      "loss": 0.9368,
+      "step": 11069
+    },
+    {
+      "epoch": 0.9891214510688677,
+      "grad_norm": 0.4472845494747162,
+      "learning_rate": 3.0655362493647865e-08,
+      "loss": 0.9731,
+      "step": 11070
+    },
+    {
+      "epoch": 0.9892108026001296,
+      "grad_norm": 0.4854414761066437,
+      "learning_rate": 3.015080681011972e-08,
+      "loss": 0.9416,
+      "step": 11071
+    },
+    {
+      "epoch": 0.9893001541313914,
+      "grad_norm": 0.5297408103942871,
+      "learning_rate": 2.965043663409883e-08,
+      "loss": 0.8901,
+      "step": 11072
+    },
+    {
+      "epoch": 0.9893895056626533,
+      "grad_norm": 0.5325205326080322,
+      "learning_rate": 2.9154252007496108e-08,
+      "loss": 0.8408,
+      "step": 11073
+    },
+    {
+      "epoch": 0.9894788571939152,
+      "grad_norm": 0.42406973242759705,
+      "learning_rate": 2.86622529718783e-08,
+      "loss": 0.9288,
+      "step": 11074
+    },
+    {
+      "epoch": 0.989568208725177,
+      "grad_norm": 0.49581971764564514,
+      "learning_rate": 2.817443956845689e-08,
+      "loss": 0.9434,
+      "step": 11075
+    },
+    {
+      "epoch": 0.9896575602564389,
+      "grad_norm": 0.5132964849472046,
+      "learning_rate": 2.769081183808253e-08,
+      "loss": 0.9745,
+      "step": 11076
+    },
+    {
+      "epoch": 0.9897469117877008,
+      "grad_norm": 0.4875660538673401,
+      "learning_rate": 2.7211369821272816e-08,
+      "loss": 0.9265,
+      "step": 11077
+    },
+    {
+      "epoch": 0.9898362633189627,
+      "grad_norm": 0.46364524960517883,
+      "learning_rate": 2.6736113558178954e-08,
+      "loss": 0.9082,
+      "step": 11078
+    },
+    {
+      "epoch": 0.9899256148502245,
+      "grad_norm": 0.4320114254951477,
+      "learning_rate": 2.626504308861355e-08,
+      "loss": 0.8796,
+      "step": 11079
+    },
+    {
+      "epoch": 0.9900149663814863,
+      "grad_norm": 0.471098393201828,
+      "learning_rate": 2.5798158452033927e-08,
+      "loss": 0.8859,
+      "step": 11080
+    },
+    {
+      "epoch": 0.9901043179127482,
+      "grad_norm": 0.5627569556236267,
+      "learning_rate": 2.533545968754214e-08,
+      "loss": 0.8719,
+      "step": 11081
+    },
+    {
+      "epoch": 0.9901936694440101,
+      "grad_norm": 0.4802547097206116,
+      "learning_rate": 2.4876946833901628e-08,
+      "loss": 0.949,
+      "step": 11082
+    },
+    {
+      "epoch": 0.990283020975272,
+      "grad_norm": 0.4991259276866913,
+      "learning_rate": 2.442261992950945e-08,
+      "loss": 0.8312,
+      "step": 11083
+    },
+    {
+      "epoch": 0.9903723725065339,
+      "grad_norm": 0.4693763256072998,
+      "learning_rate": 2.3972479012429605e-08,
+      "loss": 0.9392,
+      "step": 11084
+    },
+    {
+      "epoch": 0.9904617240377956,
+      "grad_norm": 0.6424983143806458,
+      "learning_rate": 2.3526524120354164e-08,
+      "loss": 0.8377,
+      "step": 11085
+    },
+    {
+      "epoch": 0.9905510755690575,
+      "grad_norm": 0.601347029209137,
+      "learning_rate": 2.3084755290647685e-08,
+      "loss": 0.8502,
+      "step": 11086
+    },
+    {
+      "epoch": 0.9906404271003194,
+      "grad_norm": 0.43913698196411133,
+      "learning_rate": 2.264717256030835e-08,
+      "loss": 0.9496,
+      "step": 11087
+    },
+    {
+      "epoch": 0.9907297786315813,
+      "grad_norm": 0.6155788898468018,
+      "learning_rate": 2.2213775965984617e-08,
+      "loss": 0.9325,
+      "step": 11088
+    },
+    {
+      "epoch": 0.9908191301628432,
+      "grad_norm": 0.48678359389305115,
+      "learning_rate": 2.1784565543986337e-08,
+      "loss": 0.9062,
+      "step": 11089
+    },
+    {
+      "epoch": 0.9909084816941051,
+      "grad_norm": 0.5303758978843689,
+      "learning_rate": 2.135954133025697e-08,
+      "loss": 0.8908,
+      "step": 11090
+    },
+    {
+      "epoch": 0.9909978332253669,
+      "grad_norm": 0.4228883385658264,
+      "learning_rate": 2.093870336040138e-08,
+      "loss": 0.9731,
+      "step": 11091
+    },
+    {
+      "epoch": 0.9910871847566287,
+      "grad_norm": 0.5669239163398743,
+      "learning_rate": 2.052205166966914e-08,
+      "loss": 0.9193,
+      "step": 11092
+    },
+    {
+      "epoch": 0.9911765362878906,
+      "grad_norm": 0.5962837338447571,
+      "learning_rate": 2.0109586292960116e-08,
+      "loss": 0.8922,
+      "step": 11093
+    },
+    {
+      "epoch": 0.9912658878191525,
+      "grad_norm": 0.47855356335639954,
+      "learning_rate": 1.9701307264818893e-08,
+      "loss": 0.8838,
+      "step": 11094
+    },
+    {
+      "epoch": 0.9913552393504144,
+      "grad_norm": 0.5598346590995789,
+      "learning_rate": 1.929721461944034e-08,
+      "loss": 0.8942,
+      "step": 11095
+    },
+    {
+      "epoch": 0.9914445908816762,
+      "grad_norm": 0.45872122049331665,
+      "learning_rate": 1.889730839068071e-08,
+      "loss": 1.0056,
+      "step": 11096
+    },
+    {
+      "epoch": 0.9915339424129381,
+      "grad_norm": 0.5087199807167053,
+      "learning_rate": 1.8501588612029887e-08,
+      "loss": 0.8756,
+      "step": 11097
+    },
+    {
+      "epoch": 0.9916232939442,
+      "grad_norm": 0.4826347529888153,
+      "learning_rate": 1.8110055316633567e-08,
+      "loss": 0.8969,
+      "step": 11098
+    },
+    {
+      "epoch": 0.9917126454754618,
+      "grad_norm": 0.4179272949695587,
+      "learning_rate": 1.7722708537293296e-08,
+      "loss": 0.9523,
+      "step": 11099
+    },
+    {
+      "epoch": 0.9918019970067237,
+      "grad_norm": 0.44711118936538696,
+      "learning_rate": 1.7339548306449794e-08,
+      "loss": 0.9084,
+      "step": 11100
+    },
+    {
+      "epoch": 0.9918913485379856,
+      "grad_norm": 0.4555767774581909,
+      "learning_rate": 1.69605746561885e-08,
+      "loss": 0.9573,
+      "step": 11101
+    },
+    {
+      "epoch": 0.9919807000692474,
+      "grad_norm": 0.5947062373161316,
+      "learning_rate": 1.6585787618267346e-08,
+      "loss": 0.8424,
+      "step": 11102
+    },
+    {
+      "epoch": 0.9920700516005093,
+      "grad_norm": 0.4358064532279968,
+      "learning_rate": 1.6215187224066787e-08,
+      "loss": 0.911,
+      "step": 11103
+    },
+    {
+      "epoch": 0.9921594031317712,
+      "grad_norm": 0.5996667742729187,
+      "learning_rate": 1.5848773504634207e-08,
+      "loss": 0.8718,
+      "step": 11104
+    },
+    {
+      "epoch": 0.9922487546630331,
+      "grad_norm": 0.4151008129119873,
+      "learning_rate": 1.5486546490661724e-08,
+      "loss": 0.9816,
+      "step": 11105
+    },
+    {
+      "epoch": 0.9923381061942949,
+      "grad_norm": 0.5637631416320801,
+      "learning_rate": 1.5128506212486183e-08,
+      "loss": 0.9081,
+      "step": 11106
+    },
+    {
+      "epoch": 0.9924274577255567,
+      "grad_norm": 0.46075546741485596,
+      "learning_rate": 1.4774652700100256e-08,
+      "loss": 0.8814,
+      "step": 11107
+    },
+    {
+      "epoch": 0.9925168092568186,
+      "grad_norm": 0.46446409821510315,
+      "learning_rate": 1.4424985983141348e-08,
+      "loss": 0.9606,
+      "step": 11108
+    },
+    {
+      "epoch": 0.9926061607880805,
+      "grad_norm": 0.4711821377277374,
+      "learning_rate": 1.4079506090891592e-08,
+      "loss": 0.932,
+      "step": 11109
+    },
+    {
+      "epoch": 0.9926955123193424,
+      "grad_norm": 0.5269238352775574,
+      "learning_rate": 1.3738213052300053e-08,
+      "loss": 0.8762,
+      "step": 11110
+    },
+    {
+      "epoch": 0.9927848638506043,
+      "grad_norm": 0.4487016499042511,
+      "learning_rate": 1.3401106895949422e-08,
+      "loss": 0.9347,
+      "step": 11111
+    },
+    {
+      "epoch": 0.9928742153818662,
+      "grad_norm": 0.6086486577987671,
+      "learning_rate": 1.3068187650072672e-08,
+      "loss": 0.8741,
+      "step": 11112
+    },
+    {
+      "epoch": 0.9929635669131279,
+      "grad_norm": 0.540021538734436,
+      "learning_rate": 1.2739455342558603e-08,
+      "loss": 0.8542,
+      "step": 11113
+    },
+    {
+      "epoch": 0.9930529184443898,
+      "grad_norm": 0.43035051226615906,
+      "learning_rate": 1.241491000094075e-08,
+      "loss": 0.8728,
+      "step": 11114
+    },
+    {
+      "epoch": 0.9931422699756517,
+      "grad_norm": 0.4739319980144501,
+      "learning_rate": 1.2094551652402919e-08,
+      "loss": 0.9627,
+      "step": 11115
+    },
+    {
+      "epoch": 0.9932316215069136,
+      "grad_norm": 0.42306095361709595,
+      "learning_rate": 1.1778380323779203e-08,
+      "loss": 0.9101,
+      "step": 11116
+    },
+    {
+      "epoch": 0.9933209730381755,
+      "grad_norm": 0.46345555782318115,
+      "learning_rate": 1.1466396041553973e-08,
+      "loss": 0.8924,
+      "step": 11117
+    },
+    {
+      "epoch": 0.9934103245694373,
+      "grad_norm": 0.4990481734275818,
+      "learning_rate": 1.1158598831856326e-08,
+      "loss": 0.9019,
+      "step": 11118
+    },
+    {
+      "epoch": 0.9934996761006992,
+      "grad_norm": 0.4848730266094208,
+      "learning_rate": 1.0854988720471193e-08,
+      "loss": 0.9512,
+      "step": 11119
+    },
+    {
+      "epoch": 0.993589027631961,
+      "grad_norm": 0.446789413690567,
+      "learning_rate": 1.0555565732822681e-08,
+      "loss": 0.951,
+      "step": 11120
+    },
+    {
+      "epoch": 0.9936783791632229,
+      "grad_norm": 0.5458630323410034,
+      "learning_rate": 1.0260329893996278e-08,
+      "loss": 0.9909,
+      "step": 11121
+    },
+    {
+      "epoch": 0.9937677306944848,
+      "grad_norm": 0.4132474660873413,
+      "learning_rate": 9.969281228722204e-09,
+      "loss": 0.9564,
+      "step": 11122
+    },
+    {
+      "epoch": 0.9938570822257466,
+      "grad_norm": 0.47372639179229736,
+      "learning_rate": 9.682419761369855e-09,
+      "loss": 0.9558,
+      "step": 11123
+    },
+    {
+      "epoch": 0.9939464337570085,
+      "grad_norm": 0.48265764117240906,
+      "learning_rate": 9.399745515981106e-09,
+      "loss": 0.94,
+      "step": 11124
+    },
+    {
+      "epoch": 0.9940357852882704,
+      "grad_norm": 0.4469437003135681,
+      "learning_rate": 9.121258516220366e-09,
+      "loss": 0.9356,
+      "step": 11125
+    },
+    {
+      "epoch": 0.9941251368195323,
+      "grad_norm": 0.553846538066864,
+      "learning_rate": 8.846958785418968e-09,
+      "loss": 0.941,
+      "step": 11126
+    },
+    {
+      "epoch": 0.9942144883507941,
+      "grad_norm": 0.46764692664146423,
+      "learning_rate": 8.57684634655298e-09,
+      "loss": 0.981,
+      "step": 11127
+    },
+    {
+      "epoch": 0.994303839882056,
+      "grad_norm": 0.5349944233894348,
+      "learning_rate": 8.310921222243195e-09,
+      "loss": 0.9157,
+      "step": 11128
+    },
+    {
+      "epoch": 0.9943931914133178,
+      "grad_norm": 0.47920531034469604,
+      "learning_rate": 8.04918343477179e-09,
+      "loss": 0.9246,
+      "step": 11129
+    },
+    {
+      "epoch": 0.9944825429445797,
+      "grad_norm": 0.4930896759033203,
+      "learning_rate": 7.791633006054567e-09,
+      "loss": 0.8867,
+      "step": 11130
+    },
+    {
+      "epoch": 0.9945718944758416,
+      "grad_norm": 0.5266461372375488,
+      "learning_rate": 7.538269957668709e-09,
+      "loss": 0.8601,
+      "step": 11131
+    },
+    {
+      "epoch": 0.9946612460071035,
+      "grad_norm": 0.5032413601875305,
+      "learning_rate": 7.2890943108305796e-09,
+      "loss": 0.9025,
+      "step": 11132
+    },
+    {
+      "epoch": 0.9947505975383654,
+      "grad_norm": 0.49191898107528687,
+      "learning_rate": 7.0441060864179235e-09,
+      "loss": 0.8893,
+      "step": 11133
+    },
+    {
+      "epoch": 0.9948399490696271,
+      "grad_norm": 0.46208131313323975,
+      "learning_rate": 6.80330530494766e-09,
+      "loss": 0.876,
+      "step": 11134
+    },
+    {
+      "epoch": 0.994929300600889,
+      "grad_norm": 0.5641262531280518,
+      "learning_rate": 6.566691986592543e-09,
+      "loss": 0.889,
+      "step": 11135
+    },
+    {
+      "epoch": 0.9950186521321509,
+      "grad_norm": 0.4556792080402374,
+      "learning_rate": 6.334266151164503e-09,
+      "loss": 0.933,
+      "step": 11136
+    },
+    {
+      "epoch": 0.9951080036634128,
+      "grad_norm": 0.439849317073822,
+      "learning_rate": 6.106027818136851e-09,
+      "loss": 0.9767,
+      "step": 11137
+    },
+    {
+      "epoch": 0.9951973551946747,
+      "grad_norm": 0.5620716214179993,
+      "learning_rate": 5.881977006622075e-09,
+      "loss": 0.8875,
+      "step": 11138
+    },
+    {
+      "epoch": 0.9952867067259366,
+      "grad_norm": 0.4618760943412781,
+      "learning_rate": 5.662113735394048e-09,
+      "loss": 0.9241,
+      "step": 11139
+    },
+    {
+      "epoch": 0.9953760582571984,
+      "grad_norm": 0.4364456236362457,
+      "learning_rate": 5.4464380228658184e-09,
+      "loss": 0.8809,
+      "step": 11140
+    },
+    {
+      "epoch": 0.9954654097884602,
+      "grad_norm": 0.5377599000930786,
+      "learning_rate": 5.234949887106266e-09,
+      "loss": 0.9302,
+      "step": 11141
+    },
+    {
+      "epoch": 0.9955547613197221,
+      "grad_norm": 0.4686380624771118,
+      "learning_rate": 5.0276493458178976e-09,
+      "loss": 0.9593,
+      "step": 11142
+    },
+    {
+      "epoch": 0.995644112850984,
+      "grad_norm": 0.4569147527217865,
+      "learning_rate": 4.8245364163757025e-09,
+      "loss": 0.8769,
+      "step": 11143
+    },
+    {
+      "epoch": 0.9957334643822459,
+      "grad_norm": 0.6316559910774231,
+      "learning_rate": 4.6256111157882976e-09,
+      "loss": 0.9268,
+      "step": 11144
+    },
+    {
+      "epoch": 0.9958228159135077,
+      "grad_norm": 0.6083618402481079,
+      "learning_rate": 4.430873460720131e-09,
+      "loss": 0.8524,
+      "step": 11145
+    },
+    {
+      "epoch": 0.9959121674447696,
+      "grad_norm": 0.42265886068344116,
+      "learning_rate": 4.2403234674803785e-09,
+      "loss": 0.9046,
+      "step": 11146
+    },
+    {
+      "epoch": 0.9960015189760314,
+      "grad_norm": 0.4713272750377655,
+      "learning_rate": 4.053961152028496e-09,
+      "loss": 0.9119,
+      "step": 11147
+    },
+    {
+      "epoch": 0.9960908705072933,
+      "grad_norm": 0.5426056385040283,
+      "learning_rate": 3.871786529974219e-09,
+      "loss": 0.901,
+      "step": 11148
+    },
+    {
+      "epoch": 0.9961802220385552,
+      "grad_norm": 0.46449586749076843,
+      "learning_rate": 3.6937996165831156e-09,
+      "loss": 0.9605,
+      "step": 11149
+    },
+    {
+      "epoch": 0.996269573569817,
+      "grad_norm": 0.4397094249725342,
+      "learning_rate": 3.5200004267543773e-09,
+      "loss": 0.9556,
+      "step": 11150
+    },
+    {
+      "epoch": 0.9963589251010789,
+      "grad_norm": 0.47000452876091003,
+      "learning_rate": 3.3503889750485796e-09,
+      "loss": 0.9636,
+      "step": 11151
+    },
+    {
+      "epoch": 0.9964482766323408,
+      "grad_norm": 0.48919039964675903,
+      "learning_rate": 3.184965275676577e-09,
+      "loss": 0.8975,
+      "step": 11152
+    },
+    {
+      "epoch": 0.9965376281636027,
+      "grad_norm": 0.46157464385032654,
+      "learning_rate": 3.0237293424939527e-09,
+      "loss": 0.8855,
+      "step": 11153
+    },
+    {
+      "epoch": 0.9966269796948645,
+      "grad_norm": 0.4459165036678314,
+      "learning_rate": 2.8666811890010193e-09,
+      "loss": 0.9519,
+      "step": 11154
+    },
+    {
+      "epoch": 0.9967163312261264,
+      "grad_norm": 0.5475977063179016,
+      "learning_rate": 2.713820828353919e-09,
+      "loss": 0.8692,
+      "step": 11155
+    },
+    {
+      "epoch": 0.9968056827573882,
+      "grad_norm": 0.40559154748916626,
+      "learning_rate": 2.5651482733535237e-09,
+      "loss": 0.9232,
+      "step": 11156
+    },
+    {
+      "epoch": 0.9968950342886501,
+      "grad_norm": 0.4425158202648163,
+      "learning_rate": 2.4206635364620867e-09,
+      "loss": 0.916,
+      "step": 11157
+    },
+    {
+      "epoch": 0.996984385819912,
+      "grad_norm": 0.4523138105869293,
+      "learning_rate": 2.2803666297754876e-09,
+      "loss": 0.8618,
+      "step": 11158
+    },
+    {
+      "epoch": 0.9970737373511739,
+      "grad_norm": 0.5861170291900635,
+      "learning_rate": 2.144257565045438e-09,
+      "loss": 0.8668,
+      "step": 11159
+    },
+    {
+      "epoch": 0.9971630888824358,
+      "grad_norm": 0.5498065948486328,
+      "learning_rate": 2.012336353668376e-09,
+      "loss": 0.8823,
+      "step": 11160
+    },
+    {
+      "epoch": 0.9972524404136975,
+      "grad_norm": 0.5954715013504028,
+      "learning_rate": 1.884603006702124e-09,
+      "loss": 0.8704,
+      "step": 11161
+    },
+    {
+      "epoch": 0.9973417919449594,
+      "grad_norm": 0.5423356294631958,
+      "learning_rate": 1.7610575348436798e-09,
+      "loss": 1.0066,
+      "step": 11162
+    },
+    {
+      "epoch": 0.9974311434762213,
+      "grad_norm": 0.51182621717453,
+      "learning_rate": 1.6416999484347716e-09,
+      "loss": 0.9438,
+      "step": 11163
+    },
+    {
+      "epoch": 0.9975204950074832,
+      "grad_norm": 0.5101816654205322,
+      "learning_rate": 1.5265302574785089e-09,
+      "loss": 1.0102,
+      "step": 11164
+    },
+    {
+      "epoch": 0.9976098465387451,
+      "grad_norm": 0.44059595465660095,
+      "learning_rate": 1.4155484716227296e-09,
+      "loss": 0.9013,
+      "step": 11165
+    },
+    {
+      "epoch": 0.997699198070007,
+      "grad_norm": 0.4549104869365692,
+      "learning_rate": 1.3087546001600004e-09,
+      "loss": 0.9969,
+      "step": 11166
+    },
+    {
+      "epoch": 0.9977885496012688,
+      "grad_norm": 0.6448225378990173,
+      "learning_rate": 1.2061486520387188e-09,
+      "loss": 0.8289,
+      "step": 11167
+    },
+    {
+      "epoch": 0.9978779011325306,
+      "grad_norm": 0.5743524432182312,
+      "learning_rate": 1.1077306358520112e-09,
+      "loss": 0.8468,
+      "step": 11168
+    },
+    {
+      "epoch": 0.9979672526637925,
+      "grad_norm": 0.5093139410018921,
+      "learning_rate": 1.0135005598432835e-09,
+      "loss": 0.9001,
+      "step": 11169
+    },
+    {
+      "epoch": 0.9980566041950544,
+      "grad_norm": 0.5424085855484009,
+      "learning_rate": 9.2345843190067e-10,
+      "loss": 0.8673,
+      "step": 11170
+    },
+    {
+      "epoch": 0.9981459557263163,
+      "grad_norm": 0.4926496744155884,
+      "learning_rate": 8.376042595736877e-10,
+      "loss": 0.9348,
+      "step": 11171
+    },
+    {
+      "epoch": 0.9982353072575781,
+      "grad_norm": 0.5678501129150391,
+      "learning_rate": 7.559380500454794e-10,
+      "loss": 0.9215,
+      "step": 11172
+    },
+    {
+      "epoch": 0.99832465878884,
+      "grad_norm": 0.5548151135444641,
+      "learning_rate": 6.784598101661211e-10,
+      "loss": 0.9721,
+      "step": 11173
+    },
+    {
+      "epoch": 0.9984140103201019,
+      "grad_norm": 0.5497487783432007,
+      "learning_rate": 6.051695464193152e-10,
+      "loss": 0.8651,
+      "step": 11174
+    },
+    {
+      "epoch": 0.9985033618513637,
+      "grad_norm": 0.6130139827728271,
+      "learning_rate": 5.360672649445952e-10,
+      "loss": 0.8772,
+      "step": 11175
+    },
+    {
+      "epoch": 0.9985927133826256,
+      "grad_norm": 0.6136727333068848,
+      "learning_rate": 4.711529715262231e-10,
+      "loss": 0.7813,
+      "step": 11176
+    },
+    {
+      "epoch": 0.9986820649138874,
+      "grad_norm": 0.5682715773582458,
+      "learning_rate": 4.104266716098426e-10,
+      "loss": 0.9652,
+      "step": 11177
+    },
+    {
+      "epoch": 0.9987714164451493,
+      "grad_norm": 0.4911440908908844,
+      "learning_rate": 3.538883702747242e-10,
+      "loss": 0.9506,
+      "step": 11178
+    },
+    {
+      "epoch": 0.9988607679764112,
+      "grad_norm": 0.48630785942077637,
+      "learning_rate": 3.015380722615202e-10,
+      "loss": 1.0029,
+      "step": 11179
+    },
+    {
+      "epoch": 0.9989501195076731,
+      "grad_norm": 0.4619069993495941,
+      "learning_rate": 2.533757819556115e-10,
+      "loss": 0.9795,
+      "step": 11180
+    },
+    {
+      "epoch": 0.999039471038935,
+      "grad_norm": 0.47902753949165344,
+      "learning_rate": 2.0940150338155662e-10,
+      "loss": 0.9322,
+      "step": 11181
+    },
+    {
+      "epoch": 0.9991288225701968,
+      "grad_norm": 0.4930683970451355,
+      "learning_rate": 1.6961524023639819e-10,
+      "loss": 0.9295,
+      "step": 11182
+    },
+    {
+      "epoch": 0.9992181741014586,
+      "grad_norm": 0.6377449035644531,
+      "learning_rate": 1.3401699583970306e-10,
+      "loss": 0.8337,
+      "step": 11183
+    },
+    {
+      "epoch": 0.9993075256327205,
+      "grad_norm": 0.47255298495292664,
+      "learning_rate": 1.026067731835223e-10,
+      "loss": 0.9339,
+      "step": 11184
+    },
+    {
+      "epoch": 0.9993968771639824,
+      "grad_norm": 0.5022866129875183,
+      "learning_rate": 7.538457489353334e-11,
+      "loss": 0.892,
+      "step": 11185
+    },
+    {
+      "epoch": 0.9994862286952443,
+      "grad_norm": 0.47073841094970703,
+      "learning_rate": 5.2350403251244516e-11,
+      "loss": 0.9153,
+      "step": 11186
+    },
+    {
+      "epoch": 0.9995755802265062,
+      "grad_norm": 0.5592496395111084,
+      "learning_rate": 3.3504260182892765e-11,
+      "loss": 0.9546,
+      "step": 11187
+    },
+    {
+      "epoch": 0.999664931757768,
+      "grad_norm": 0.41335591673851013,
+      "learning_rate": 1.8846147276097014e-11,
+      "loss": 0.9724,
+      "step": 11188
+    },
+    {
+      "epoch": 0.9997542832890298,
+      "grad_norm": 0.6354795694351196,
+      "learning_rate": 8.376065746551476e-12,
+      "loss": 0.8885,
+      "step": 11189
+    },
+    {
+      "epoch": 0.9998436348202917,
+      "grad_norm": 0.4729311466217041,
+      "learning_rate": 2.094016482434569e-12,
+      "loss": 0.897,
+      "step": 11190
+    },
+    {
+      "epoch": 0.9999329863515536,
+      "grad_norm": 0.5076917409896851,
+      "learning_rate": 0.0,
+      "loss": 0.9054,
+      "step": 11191
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 11191,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.597312254069965e+18,
+  "train_batch_size": 16,
+  "trial_name": null,
+  "trial_params": null
+}