diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,128833 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0286513694778567,
+  "eval_steps": 500,
+  "global_step": 18400,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00016591314446887055,
+      "grad_norm": 85.19542448446617,
+      "learning_rate": 3.0000000000000004e-08,
+      "loss": 8.5863,
+      "step": 1
+    },
+    {
+      "epoch": 0.0003318262889377411,
+      "grad_norm": 82.32409405573297,
+      "learning_rate": 6.000000000000001e-08,
+      "loss": 8.473,
+      "step": 2
+    },
+    {
+      "epoch": 0.0004977394334066117,
+      "grad_norm": 87.13591882926195,
+      "learning_rate": 9e-08,
+      "loss": 8.6763,
+      "step": 3
+    },
+    {
+      "epoch": 0.0006636525778754822,
+      "grad_norm": 84.18601565903892,
+      "learning_rate": 1.2000000000000002e-07,
+      "loss": 8.4935,
+      "step": 4
+    },
+    {
+      "epoch": 0.0008295657223443527,
+      "grad_norm": 84.14667729364332,
+      "learning_rate": 1.5e-07,
+      "loss": 8.5695,
+      "step": 5
+    },
+    {
+      "epoch": 0.0009954788668132234,
+      "grad_norm": 84.10731463288944,
+      "learning_rate": 1.8e-07,
+      "loss": 8.5366,
+      "step": 6
+    },
+    {
+      "epoch": 0.0011613920112820938,
+      "grad_norm": 83.28625999969219,
+      "learning_rate": 2.1000000000000003e-07,
+      "loss": 8.5134,
+      "step": 7
+    },
+    {
+      "epoch": 0.0013273051557509644,
+      "grad_norm": 84.44997731146951,
+      "learning_rate": 2.4000000000000003e-07,
+      "loss": 8.5326,
+      "step": 8
+    },
+    {
+      "epoch": 0.0014932183002198348,
+      "grad_norm": 89.14194809970635,
+      "learning_rate": 2.7e-07,
+      "loss": 8.722,
+      "step": 9
+    },
+    {
+      "epoch": 0.0016591314446887055,
+      "grad_norm": 84.83466343564557,
+      "learning_rate": 3e-07,
+      "loss": 8.555,
+      "step": 10
+    },
+    {
+      "epoch": 0.001825044589157576,
+      "grad_norm": 81.56216222466577,
+      "learning_rate": 3.3e-07,
+      "loss": 8.4491,
+      "step": 11
+    },
+    {
+      "epoch": 0.0019909577336264467,
+      "grad_norm": 86.81209990901495,
+      "learning_rate": 3.6e-07,
+      "loss": 8.6371,
+      "step": 12
+    },
+    {
+      "epoch": 0.002156870878095317,
+      "grad_norm": 85.71908305594017,
+      "learning_rate": 3.9e-07,
+      "loss": 8.5535,
+      "step": 13
+    },
+    {
+      "epoch": 0.0023227840225641875,
+      "grad_norm": 86.9623893470603,
+      "learning_rate": 4.2000000000000006e-07,
+      "loss": 8.6119,
+      "step": 14
+    },
+    {
+      "epoch": 0.002488697167033058,
+      "grad_norm": 82.49495302426142,
+      "learning_rate": 4.5000000000000003e-07,
+      "loss": 8.441,
+      "step": 15
+    },
+    {
+      "epoch": 0.002654610311501929,
+      "grad_norm": 81.33695392158589,
+      "learning_rate": 4.800000000000001e-07,
+      "loss": 8.4148,
+      "step": 16
+    },
+    {
+      "epoch": 0.0028205234559707994,
+      "grad_norm": 84.88167396340097,
+      "learning_rate": 5.1e-07,
+      "loss": 8.5835,
+      "step": 17
+    },
+    {
+      "epoch": 0.0029864366004396696,
+      "grad_norm": 79.05682174134483,
+      "learning_rate": 5.4e-07,
+      "loss": 8.3251,
+      "step": 18
+    },
+    {
+      "epoch": 0.0031523497449085403,
+      "grad_norm": 78.38741271121032,
+      "learning_rate": 5.7e-07,
+      "loss": 8.3005,
+      "step": 19
+    },
+    {
+      "epoch": 0.003318262889377411,
+      "grad_norm": 82.4891623112013,
+      "learning_rate": 6e-07,
+      "loss": 8.4621,
+      "step": 20
+    },
+    {
+      "epoch": 0.0034841760338462815,
+      "grad_norm": 79.1659385995681,
+      "learning_rate": 6.300000000000001e-07,
+      "loss": 8.3031,
+      "step": 21
+    },
+    {
+      "epoch": 0.003650089178315152,
+      "grad_norm": 77.06598892419889,
+      "learning_rate": 6.6e-07,
+      "loss": 8.2303,
+      "step": 22
+    },
+    {
+      "epoch": 0.0038160023227840224,
+      "grad_norm": 82.36739641908615,
+      "learning_rate": 6.9e-07,
+      "loss": 8.2115,
+      "step": 23
+    },
+    {
+      "epoch": 0.003981915467252893,
+      "grad_norm": 76.36757247944843,
+      "learning_rate": 7.2e-07,
+      "loss": 8.1456,
+      "step": 24
+    },
+    {
+      "epoch": 0.004147828611721764,
+      "grad_norm": 68.98126779127293,
+      "learning_rate": 7.5e-07,
+      "loss": 7.8079,
+      "step": 25
+    },
+    {
+      "epoch": 0.004313741756190634,
+      "grad_norm": 69.00876196915596,
+      "learning_rate": 7.8e-07,
+      "loss": 7.8411,
+      "step": 26
+    },
+    {
+      "epoch": 0.004479654900659505,
+      "grad_norm": 67.5391863634817,
+      "learning_rate": 8.1e-07,
+      "loss": 7.7531,
+      "step": 27
+    },
+    {
+      "epoch": 0.004645568045128375,
+      "grad_norm": 68.05473702316927,
+      "learning_rate": 8.400000000000001e-07,
+      "loss": 7.7816,
+      "step": 28
+    },
+    {
+      "epoch": 0.004811481189597246,
+      "grad_norm": 67.38920732223586,
+      "learning_rate": 8.7e-07,
+      "loss": 7.7137,
+      "step": 29
+    },
+    {
+      "epoch": 0.004977394334066116,
+      "grad_norm": 66.22135170564259,
+      "learning_rate": 9.000000000000001e-07,
+      "loss": 7.6523,
+      "step": 30
+    },
+    {
+      "epoch": 0.0051433074785349866,
+      "grad_norm": 65.24262292310318,
+      "learning_rate": 9.3e-07,
+      "loss": 7.6215,
+      "step": 31
+    },
+    {
+      "epoch": 0.005309220623003858,
+      "grad_norm": 66.75573857246084,
+      "learning_rate": 9.600000000000001e-07,
+      "loss": 7.6052,
+      "step": 32
+    },
+    {
+      "epoch": 0.005475133767472728,
+      "grad_norm": 66.14125778432644,
+      "learning_rate": 9.9e-07,
+      "loss": 7.4641,
+      "step": 33
+    },
+    {
+      "epoch": 0.005641046911941599,
+      "grad_norm": 62.44214008796182,
+      "learning_rate": 1.02e-06,
+      "loss": 7.0586,
+      "step": 34
+    },
+    {
+      "epoch": 0.005806960056410469,
+      "grad_norm": 59.29700311739906,
+      "learning_rate": 1.0500000000000001e-06,
+      "loss": 6.8323,
+      "step": 35
+    },
+    {
+      "epoch": 0.005972873200879339,
+      "grad_norm": 59.53983946181781,
+      "learning_rate": 1.08e-06,
+      "loss": 6.8458,
+      "step": 36
+    },
+    {
+      "epoch": 0.00613878634534821,
+      "grad_norm": 58.40955787137384,
+      "learning_rate": 1.11e-06,
+      "loss": 6.749,
+      "step": 37
+    },
+    {
+      "epoch": 0.0063046994898170805,
+      "grad_norm": 58.46669985481344,
+      "learning_rate": 1.14e-06,
+      "loss": 6.7341,
+      "step": 38
+    },
+    {
+      "epoch": 0.006470612634285952,
+      "grad_norm": 57.85754721296723,
+      "learning_rate": 1.17e-06,
+      "loss": 6.679,
+      "step": 39
+    },
+    {
+      "epoch": 0.006636525778754822,
+      "grad_norm": 61.03757307715009,
+      "learning_rate": 1.2e-06,
+      "loss": 6.6835,
+      "step": 40
+    },
+    {
+      "epoch": 0.006802438923223692,
+      "grad_norm": 58.679250167783295,
+      "learning_rate": 1.2299999999999999e-06,
+      "loss": 6.6456,
+      "step": 41
+    },
+    {
+      "epoch": 0.006968352067692563,
+      "grad_norm": 58.420472121350315,
+      "learning_rate": 1.2600000000000002e-06,
+      "loss": 6.5997,
+      "step": 42
+    },
+    {
+      "epoch": 0.007134265212161433,
+      "grad_norm": 57.989410831374606,
+      "learning_rate": 1.2900000000000001e-06,
+      "loss": 6.477,
+      "step": 43
+    },
+    {
+      "epoch": 0.007300178356630304,
+      "grad_norm": 57.29419183703238,
+      "learning_rate": 1.32e-06,
+      "loss": 6.3482,
+      "step": 44
+    },
+    {
+      "epoch": 0.0074660915010991745,
+      "grad_norm": 57.64245355217973,
+      "learning_rate": 1.35e-06,
+      "loss": 6.2579,
+      "step": 45
+    },
+    {
+      "epoch": 0.007632004645568045,
+      "grad_norm": 56.91570071531264,
+      "learning_rate": 1.38e-06,
+      "loss": 6.1127,
+      "step": 46
+    },
+    {
+      "epoch": 0.007797917790036916,
+      "grad_norm": 57.511790120005976,
+      "learning_rate": 1.41e-06,
+      "loss": 6.0203,
+      "step": 47
+    },
+    {
+      "epoch": 0.007963830934505787,
+      "grad_norm": 58.13021570410771,
+      "learning_rate": 1.44e-06,
+      "loss": 5.9262,
+      "step": 48
+    },
+    {
+      "epoch": 0.008129744078974656,
+      "grad_norm": 57.938125335380285,
+      "learning_rate": 1.4700000000000001e-06,
+      "loss": 5.8124,
+      "step": 49
+    },
+    {
+      "epoch": 0.008295657223443527,
+      "grad_norm": 57.39264712573751,
+      "learning_rate": 1.5e-06,
+      "loss": 5.6758,
+      "step": 50
+    },
+    {
+      "epoch": 0.008461570367912398,
+      "grad_norm": 57.79052254778045,
+      "learning_rate": 1.5300000000000002e-06,
+      "loss": 5.6168,
+      "step": 51
+    },
+    {
+      "epoch": 0.008627483512381268,
+      "grad_norm": 58.18250706411147,
+      "learning_rate": 1.56e-06,
+      "loss": 5.527,
+      "step": 52
+    },
+    {
+      "epoch": 0.008793396656850139,
+      "grad_norm": 56.91519882334982,
+      "learning_rate": 1.59e-06,
+      "loss": 5.4117,
+      "step": 53
+    },
+    {
+      "epoch": 0.00895930980131901,
+      "grad_norm": 58.80496017496223,
+      "learning_rate": 1.62e-06,
+      "loss": 5.3978,
+      "step": 54
+    },
+    {
+      "epoch": 0.00912522294578788,
+      "grad_norm": 57.514939390872016,
+      "learning_rate": 1.6499999999999999e-06,
+      "loss": 5.2806,
+      "step": 55
+    },
+    {
+      "epoch": 0.00929113609025675,
+      "grad_norm": 58.261230393043675,
+      "learning_rate": 1.6800000000000002e-06,
+      "loss": 5.2305,
+      "step": 56
+    },
+    {
+      "epoch": 0.009457049234725621,
+      "grad_norm": 58.853719732787475,
+      "learning_rate": 1.7100000000000001e-06,
+      "loss": 5.1747,
+      "step": 57
+    },
+    {
+      "epoch": 0.009622962379194492,
+      "grad_norm": 58.44206227327712,
+      "learning_rate": 1.74e-06,
+      "loss": 5.0768,
+      "step": 58
+    },
+    {
+      "epoch": 0.009788875523663362,
+      "grad_norm": 58.39912900993697,
+      "learning_rate": 1.77e-06,
+      "loss": 4.9874,
+      "step": 59
+    },
+    {
+      "epoch": 0.009954788668132233,
+      "grad_norm": 58.19859876404236,
+      "learning_rate": 1.8000000000000001e-06,
+      "loss": 4.8936,
+      "step": 60
+    },
+    {
+      "epoch": 0.010120701812601104,
+      "grad_norm": 58.931458451778994,
+      "learning_rate": 1.83e-06,
+      "loss": 4.8333,
+      "step": 61
+    },
+    {
+      "epoch": 0.010286614957069973,
+      "grad_norm": 58.566598779309345,
+      "learning_rate": 1.86e-06,
+      "loss": 4.7316,
+      "step": 62
+    },
+    {
+      "epoch": 0.010452528101538844,
+      "grad_norm": 59.43204190036222,
+      "learning_rate": 1.89e-06,
+      "loss": 4.6782,
+      "step": 63
+    },
+    {
+      "epoch": 0.010618441246007715,
+      "grad_norm": 59.32237678351381,
+      "learning_rate": 1.9200000000000003e-06,
+      "loss": 4.5875,
+      "step": 64
+    },
+    {
+      "epoch": 0.010784354390476586,
+      "grad_norm": 58.22281010507284,
+      "learning_rate": 1.95e-06,
+      "loss": 4.4805,
+      "step": 65
+    },
+    {
+      "epoch": 0.010950267534945456,
+      "grad_norm": 58.89565065173638,
+      "learning_rate": 1.98e-06,
+      "loss": 4.4417,
+      "step": 66
+    },
+    {
+      "epoch": 0.011116180679414327,
+      "grad_norm": 59.219564213854724,
+      "learning_rate": 2.01e-06,
+      "loss": 4.3646,
+      "step": 67
+    },
+    {
+      "epoch": 0.011282093823883198,
+      "grad_norm": 60.084020905213414,
+      "learning_rate": 2.04e-06,
+      "loss": 4.3387,
+      "step": 68
+    },
+    {
+      "epoch": 0.011448006968352067,
+      "grad_norm": 59.930063371072634,
+      "learning_rate": 2.07e-06,
+      "loss": 4.2256,
+      "step": 69
+    },
+    {
+      "epoch": 0.011613920112820938,
+      "grad_norm": 59.443610975279206,
+      "learning_rate": 2.1000000000000002e-06,
+      "loss": 4.1584,
+      "step": 70
+    },
+    {
+      "epoch": 0.01177983325728981,
+      "grad_norm": 60.46768457622464,
+      "learning_rate": 2.13e-06,
+      "loss": 4.1046,
+      "step": 71
+    },
+    {
+      "epoch": 0.011945746401758679,
+      "grad_norm": 60.923838757674325,
+      "learning_rate": 2.16e-06,
+      "loss": 4.0402,
+      "step": 72
+    },
+    {
+      "epoch": 0.01211165954622755,
+      "grad_norm": 60.56706226907599,
+      "learning_rate": 2.19e-06,
+      "loss": 3.9675,
+      "step": 73
+    },
+    {
+      "epoch": 0.01227757269069642,
+      "grad_norm": 60.52251526885924,
+      "learning_rate": 2.22e-06,
+      "loss": 3.934,
+      "step": 74
+    },
+    {
+      "epoch": 0.012443485835165292,
+      "grad_norm": 60.64687995134201,
+      "learning_rate": 2.25e-06,
+      "loss": 3.8682,
+      "step": 75
+    },
+    {
+      "epoch": 0.012609398979634161,
+      "grad_norm": 59.74345013299483,
+      "learning_rate": 2.28e-06,
+      "loss": 3.7878,
+      "step": 76
+    },
+    {
+      "epoch": 0.012775312124103032,
+      "grad_norm": 62.7371580672841,
+      "learning_rate": 2.31e-06,
+      "loss": 3.7883,
+      "step": 77
+    },
+    {
+      "epoch": 0.012941225268571903,
+      "grad_norm": 61.120581572512556,
+      "learning_rate": 2.34e-06,
+      "loss": 3.6993,
+      "step": 78
+    },
+    {
+      "epoch": 0.013107138413040773,
+      "grad_norm": 61.57800519844952,
+      "learning_rate": 2.3699999999999998e-06,
+      "loss": 3.6369,
+      "step": 79
+    },
+    {
+      "epoch": 0.013273051557509644,
+      "grad_norm": 61.54876895013541,
+      "learning_rate": 2.4e-06,
+      "loss": 3.5785,
+      "step": 80
+    },
+    {
+      "epoch": 0.013438964701978515,
+      "grad_norm": 60.1358841900405,
+      "learning_rate": 2.43e-06,
+      "loss": 3.514,
+      "step": 81
+    },
+    {
+      "epoch": 0.013604877846447384,
+      "grad_norm": 61.297281772956765,
+      "learning_rate": 2.4599999999999997e-06,
+      "loss": 3.4742,
+      "step": 82
+    },
+    {
+      "epoch": 0.013770790990916255,
+      "grad_norm": 61.30923571729013,
+      "learning_rate": 2.4900000000000003e-06,
+      "loss": 3.4433,
+      "step": 83
+    },
+    {
+      "epoch": 0.013936704135385126,
+      "grad_norm": 61.23947604523023,
+      "learning_rate": 2.5200000000000004e-06,
+      "loss": 3.3705,
+      "step": 84
+    },
+    {
+      "epoch": 0.014102617279853997,
+      "grad_norm": 62.39595603092295,
+      "learning_rate": 2.55e-06,
+      "loss": 3.3502,
+      "step": 85
+    },
+    {
+      "epoch": 0.014268530424322867,
+      "grad_norm": 61.53318479023394,
+      "learning_rate": 2.5800000000000003e-06,
+      "loss": 3.2943,
+      "step": 86
+    },
+    {
+      "epoch": 0.014434443568791738,
+      "grad_norm": 61.212619109385486,
+      "learning_rate": 2.61e-06,
+      "loss": 3.2354,
+      "step": 87
+    },
+    {
+      "epoch": 0.014600356713260609,
+      "grad_norm": 60.68820969945481,
+      "learning_rate": 2.64e-06,
+      "loss": 3.1867,
+      "step": 88
+    },
+    {
+      "epoch": 0.014766269857729478,
+      "grad_norm": 61.161255238888366,
+      "learning_rate": 2.6700000000000003e-06,
+      "loss": 3.1363,
+      "step": 89
+    },
+    {
+      "epoch": 0.014932183002198349,
+      "grad_norm": 61.39181889008794,
+      "learning_rate": 2.7e-06,
+      "loss": 3.106,
+      "step": 90
+    },
+    {
+      "epoch": 0.01509809614666722,
+      "grad_norm": 61.510960534577144,
+      "learning_rate": 2.73e-06,
+      "loss": 3.0709,
+      "step": 91
+    },
+    {
+      "epoch": 0.01526400929113609,
+      "grad_norm": 61.403101355512064,
+      "learning_rate": 2.76e-06,
+      "loss": 3.0295,
+      "step": 92
+    },
+    {
+      "epoch": 0.01542992243560496,
+      "grad_norm": 60.92480004727436,
+      "learning_rate": 2.79e-06,
+      "loss": 2.9853,
+      "step": 93
+    },
+    {
+      "epoch": 0.015595835580073832,
+      "grad_norm": 60.38494778358348,
+      "learning_rate": 2.82e-06,
+      "loss": 2.9374,
+      "step": 94
+    },
+    {
+      "epoch": 0.015761748724542703,
+      "grad_norm": 61.24204912886601,
+      "learning_rate": 2.85e-06,
+      "loss": 2.9142,
+      "step": 95
+    },
+    {
+      "epoch": 0.015927661869011574,
+      "grad_norm": 61.79448720849728,
+      "learning_rate": 2.88e-06,
+      "loss": 2.8738,
+      "step": 96
+    },
+    {
+      "epoch": 0.01609357501348044,
+      "grad_norm": 60.44596322113541,
+      "learning_rate": 2.9099999999999997e-06,
+      "loss": 2.8393,
+      "step": 97
+    },
+    {
+      "epoch": 0.016259488157949312,
+      "grad_norm": 61.69496692909939,
+      "learning_rate": 2.9400000000000002e-06,
+      "loss": 2.8091,
+      "step": 98
+    },
+    {
+      "epoch": 0.016425401302418183,
+      "grad_norm": 61.042506818121446,
+      "learning_rate": 2.9700000000000004e-06,
+      "loss": 2.7632,
+      "step": 99
+    },
+    {
+      "epoch": 0.016591314446887055,
+      "grad_norm": 59.83708084484388,
+      "learning_rate": 3e-06,
+      "loss": 2.7211,
+      "step": 100
+    },
+    {
+      "epoch": 0.016757227591355926,
+      "grad_norm": 60.76096104413392,
+      "learning_rate": 3.0300000000000002e-06,
+      "loss": 2.6877,
+      "step": 101
+    },
+    {
+      "epoch": 0.016923140735824797,
+      "grad_norm": 60.503095248644414,
+      "learning_rate": 3.0600000000000003e-06,
+      "loss": 2.6464,
+      "step": 102
+    },
+    {
+      "epoch": 0.017089053880293668,
+      "grad_norm": 60.022580078973974,
+      "learning_rate": 3.09e-06,
+      "loss": 2.6246,
+      "step": 103
+    },
+    {
+      "epoch": 0.017254967024762535,
+      "grad_norm": 60.7786406432538,
+      "learning_rate": 3.12e-06,
+      "loss": 2.5835,
+      "step": 104
+    },
+    {
+      "epoch": 0.017420880169231406,
+      "grad_norm": 60.0134194819714,
+      "learning_rate": 3.15e-06,
+      "loss": 2.548,
+      "step": 105
+    },
+    {
+      "epoch": 0.017586793313700277,
+      "grad_norm": 59.59173672650453,
+      "learning_rate": 3.18e-06,
+      "loss": 2.5352,
+      "step": 106
+    },
+    {
+      "epoch": 0.01775270645816915,
+      "grad_norm": 59.365953677840324,
+      "learning_rate": 3.21e-06,
+      "loss": 2.5026,
+      "step": 107
+    },
+    {
+      "epoch": 0.01791861960263802,
+      "grad_norm": 59.03666897099401,
+      "learning_rate": 3.24e-06,
+      "loss": 2.4604,
+      "step": 108
+    },
+    {
+      "epoch": 0.01808453274710689,
+      "grad_norm": 58.7798482726752,
+      "learning_rate": 3.27e-06,
+      "loss": 2.4486,
+      "step": 109
+    },
+    {
+      "epoch": 0.01825044589157576,
+      "grad_norm": 58.114483374139134,
+      "learning_rate": 3.2999999999999997e-06,
+      "loss": 2.4179,
+      "step": 110
+    },
+    {
+      "epoch": 0.01841635903604463,
+      "grad_norm": 59.80520961291876,
+      "learning_rate": 3.33e-06,
+      "loss": 2.392,
+      "step": 111
+    },
+    {
+      "epoch": 0.0185822721805135,
+      "grad_norm": 58.14478216556001,
+      "learning_rate": 3.3600000000000004e-06,
+      "loss": 2.3406,
+      "step": 112
+    },
+    {
+      "epoch": 0.01874818532498237,
+      "grad_norm": 59.18483586241932,
+      "learning_rate": 3.3899999999999997e-06,
+      "loss": 2.3231,
+      "step": 113
+    },
+    {
+      "epoch": 0.018914098469451242,
+      "grad_norm": 57.28442984470624,
+      "learning_rate": 3.4200000000000003e-06,
+      "loss": 2.2941,
+      "step": 114
+    },
+    {
+      "epoch": 0.019080011613920114,
+      "grad_norm": 57.357039771400295,
+      "learning_rate": 3.4500000000000004e-06,
+      "loss": 2.2816,
+      "step": 115
+    },
+    {
+      "epoch": 0.019245924758388985,
+      "grad_norm": 57.111232218047554,
+      "learning_rate": 3.48e-06,
+      "loss": 2.2342,
+      "step": 116
+    },
+    {
+      "epoch": 0.019411837902857852,
+      "grad_norm": 57.18649743743891,
+      "learning_rate": 3.5100000000000003e-06,
+      "loss": 2.2274,
+      "step": 117
+    },
+    {
+      "epoch": 0.019577751047326723,
+      "grad_norm": 56.32726893984759,
+      "learning_rate": 3.54e-06,
+      "loss": 2.188,
+      "step": 118
+    },
+    {
+      "epoch": 0.019743664191795594,
+      "grad_norm": 56.11126674497503,
+      "learning_rate": 3.57e-06,
+      "loss": 2.1688,
+      "step": 119
+    },
+    {
+      "epoch": 0.019909577336264465,
+      "grad_norm": 56.452361895827636,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 2.1257,
+      "step": 120
+    },
+    {
+      "epoch": 0.020075490480733336,
+      "grad_norm": 54.87188045608243,
+      "learning_rate": 3.63e-06,
+      "loss": 2.1082,
+      "step": 121
+    },
+    {
+      "epoch": 0.020241403625202208,
+      "grad_norm": 55.85330145328899,
+      "learning_rate": 3.66e-06,
+      "loss": 2.1085,
+      "step": 122
+    },
+    {
+      "epoch": 0.02040731676967108,
+      "grad_norm": 54.697357042805265,
+      "learning_rate": 3.69e-06,
+      "loss": 2.0609,
+      "step": 123
+    },
+    {
+      "epoch": 0.020573229914139946,
+      "grad_norm": 54.39711063453628,
+      "learning_rate": 3.72e-06,
+      "loss": 2.0577,
+      "step": 124
+    },
+    {
+      "epoch": 0.020739143058608817,
+      "grad_norm": 54.63336965202357,
+      "learning_rate": 3.75e-06,
+      "loss": 2.0359,
+      "step": 125
+    },
+    {
+      "epoch": 0.02090505620307769,
+      "grad_norm": 54.2613857002437,
+      "learning_rate": 3.78e-06,
+      "loss": 2.0223,
+      "step": 126
+    },
+    {
+      "epoch": 0.02107096934754656,
+      "grad_norm": 54.55239593692058,
+      "learning_rate": 3.81e-06,
+      "loss": 1.9907,
+      "step": 127
+    },
+    {
+      "epoch": 0.02123688249201543,
+      "grad_norm": 54.163024439012574,
+      "learning_rate": 3.8400000000000005e-06,
+      "loss": 1.9572,
+      "step": 128
+    },
+    {
+      "epoch": 0.0214027956364843,
+      "grad_norm": 52.39556608669111,
+      "learning_rate": 3.87e-06,
+      "loss": 1.9547,
+      "step": 129
+    },
+    {
+      "epoch": 0.021568708780953173,
+      "grad_norm": 52.88053635834076,
+      "learning_rate": 3.9e-06,
+      "loss": 1.9174,
+      "step": 130
+    },
+    {
+      "epoch": 0.02173462192542204,
+      "grad_norm": 52.27647350963412,
+      "learning_rate": 3.93e-06,
+      "loss": 1.9085,
+      "step": 131
+    },
+    {
+      "epoch": 0.02190053506989091,
+      "grad_norm": 51.8188123723821,
+      "learning_rate": 3.96e-06,
+      "loss": 1.8921,
+      "step": 132
+    },
+    {
+      "epoch": 0.022066448214359782,
+      "grad_norm": 51.13956466153071,
+      "learning_rate": 3.99e-06,
+      "loss": 1.8695,
+      "step": 133
+    },
+    {
+      "epoch": 0.022232361358828653,
+      "grad_norm": 50.880997478246,
+      "learning_rate": 4.02e-06,
+      "loss": 1.8723,
+      "step": 134
+    },
+    {
+      "epoch": 0.022398274503297524,
+      "grad_norm": 50.17800750552662,
+      "learning_rate": 4.05e-06,
+      "loss": 1.8362,
+      "step": 135
+    },
+    {
+      "epoch": 0.022564187647766396,
+      "grad_norm": 50.25731092529157,
+      "learning_rate": 4.08e-06,
+      "loss": 1.8273,
+      "step": 136
+    },
+    {
+      "epoch": 0.022730100792235263,
+      "grad_norm": 49.557602243937815,
+      "learning_rate": 4.1100000000000005e-06,
+      "loss": 1.7923,
+      "step": 137
+    },
+    {
+      "epoch": 0.022896013936704134,
+      "grad_norm": 49.06910514435632,
+      "learning_rate": 4.14e-06,
+      "loss": 1.7823,
+      "step": 138
+    },
+    {
+      "epoch": 0.023061927081173005,
+      "grad_norm": 48.124932183722606,
+      "learning_rate": 4.17e-06,
+      "loss": 1.7748,
+      "step": 139
+    },
+    {
+      "epoch": 0.023227840225641876,
+      "grad_norm": 48.134667233726226,
+      "learning_rate": 4.2000000000000004e-06,
+      "loss": 1.7639,
+      "step": 140
+    },
+    {
+      "epoch": 0.023393753370110747,
+      "grad_norm": 47.557312841261506,
+      "learning_rate": 4.23e-06,
+      "loss": 1.7324,
+      "step": 141
+    },
+    {
+      "epoch": 0.02355966651457962,
+      "grad_norm": 47.70768250047664,
+      "learning_rate": 4.26e-06,
+      "loss": 1.7218,
+      "step": 142
+    },
+    {
+      "epoch": 0.02372557965904849,
+      "grad_norm": 46.50696652735735,
+      "learning_rate": 4.2900000000000004e-06,
+      "loss": 1.6973,
+      "step": 143
+    },
+    {
+      "epoch": 0.023891492803517357,
+      "grad_norm": 46.08588423958666,
+      "learning_rate": 4.32e-06,
+      "loss": 1.6985,
+      "step": 144
+    },
+    {
+      "epoch": 0.024057405947986228,
+      "grad_norm": 45.56462043474931,
+      "learning_rate": 4.35e-06,
+      "loss": 1.6685,
+      "step": 145
+    },
+    {
+      "epoch": 0.0242233190924551,
+      "grad_norm": 45.97216272149993,
+      "learning_rate": 4.38e-06,
+      "loss": 1.6579,
+      "step": 146
+    },
+    {
+      "epoch": 0.02438923223692397,
+      "grad_norm": 45.46498190474296,
+      "learning_rate": 4.41e-06,
+      "loss": 1.64,
+      "step": 147
+    },
+    {
+      "epoch": 0.02455514538139284,
+      "grad_norm": 44.854534417908845,
+      "learning_rate": 4.44e-06,
+      "loss": 1.624,
+      "step": 148
+    },
+    {
+      "epoch": 0.024721058525861712,
+      "grad_norm": 43.83569267105489,
+      "learning_rate": 4.4699999999999996e-06,
+      "loss": 1.6165,
+      "step": 149
+    },
+    {
+      "epoch": 0.024886971670330583,
+      "grad_norm": 43.48566112246643,
+      "learning_rate": 4.5e-06,
+      "loss": 1.5971,
+      "step": 150
+    },
+    {
+      "epoch": 0.02505288481479945,
+      "grad_norm": 42.91276059636016,
+      "learning_rate": 4.53e-06,
+      "loss": 1.5792,
+      "step": 151
+    },
+    {
+      "epoch": 0.025218797959268322,
+      "grad_norm": 42.606137245138356,
+      "learning_rate": 4.56e-06,
+      "loss": 1.5553,
+      "step": 152
+    },
+    {
+      "epoch": 0.025384711103737193,
+      "grad_norm": 41.4393639959328,
+      "learning_rate": 4.59e-06,
+      "loss": 1.5775,
+      "step": 153
+    },
+    {
+      "epoch": 0.025550624248206064,
+      "grad_norm": 41.54145609070485,
+      "learning_rate": 4.62e-06,
+      "loss": 1.5479,
+      "step": 154
+    },
+    {
+      "epoch": 0.025716537392674935,
+      "grad_norm": 41.887845194447074,
+      "learning_rate": 4.65e-06,
+      "loss": 1.5296,
+      "step": 155
+    },
+    {
+      "epoch": 0.025882450537143806,
+      "grad_norm": 40.933178238844356,
+      "learning_rate": 4.68e-06,
+      "loss": 1.5297,
+      "step": 156
+    },
+    {
+      "epoch": 0.026048363681612677,
+      "grad_norm": 39.94971992763078,
+      "learning_rate": 4.71e-06,
+      "loss": 1.5151,
+      "step": 157
+    },
+    {
+      "epoch": 0.026214276826081545,
+      "grad_norm": 39.35709938410145,
+      "learning_rate": 4.7399999999999995e-06,
+      "loss": 1.5054,
+      "step": 158
+    },
+    {
+      "epoch": 0.026380189970550416,
+      "grad_norm": 38.69670975999757,
+      "learning_rate": 4.77e-06,
+      "loss": 1.4814,
+      "step": 159
+    },
+    {
+      "epoch": 0.026546103115019287,
+      "grad_norm": 38.62419347835826,
+      "learning_rate": 4.8e-06,
+      "loss": 1.4754,
+      "step": 160
+    },
+    {
+      "epoch": 0.02671201625948816,
+      "grad_norm": 37.77998094940736,
+      "learning_rate": 4.8299999999999995e-06,
+      "loss": 1.462,
+      "step": 161
+    },
+    {
+      "epoch": 0.02687792940395703,
+      "grad_norm": 38.187533465091185,
+      "learning_rate": 4.86e-06,
+      "loss": 1.4476,
+      "step": 162
+    },
+    {
+      "epoch": 0.0270438425484259,
+      "grad_norm": 36.94958450784839,
+      "learning_rate": 4.89e-06,
+      "loss": 1.4472,
+      "step": 163
+    },
+    {
+      "epoch": 0.027209755692894768,
+      "grad_norm": 36.82539602249903,
+      "learning_rate": 4.9199999999999995e-06,
+      "loss": 1.4243,
+      "step": 164
+    },
+    {
+      "epoch": 0.02737566883736364,
+      "grad_norm": 35.84655716782342,
+      "learning_rate": 4.950000000000001e-06,
+      "loss": 1.4311,
+      "step": 165
+    },
+    {
+      "epoch": 0.02754158198183251,
+      "grad_norm": 35.976742015954045,
+      "learning_rate": 4.980000000000001e-06,
+      "loss": 1.4,
+      "step": 166
+    },
+    {
+      "epoch": 0.02770749512630138,
+      "grad_norm": 35.117919170195876,
+      "learning_rate": 5.01e-06,
+      "loss": 1.3998,
+      "step": 167
+    },
+    {
+      "epoch": 0.027873408270770252,
+      "grad_norm": 35.170733706617995,
+      "learning_rate": 5.040000000000001e-06,
+      "loss": 1.3726,
+      "step": 168
+    },
+    {
+      "epoch": 0.028039321415239123,
+      "grad_norm": 34.47083347313095,
+      "learning_rate": 5.070000000000001e-06,
+      "loss": 1.381,
+      "step": 169
+    },
+    {
+      "epoch": 0.028205234559707994,
+      "grad_norm": 33.79057750635065,
+      "learning_rate": 5.1e-06,
+      "loss": 1.3674,
+      "step": 170
+    },
+    {
+      "epoch": 0.028371147704176862,
+      "grad_norm": 32.91789621932166,
+      "learning_rate": 5.13e-06,
+      "loss": 1.3759,
+      "step": 171
+    },
+    {
+      "epoch": 0.028537060848645733,
+      "grad_norm": 33.56814465795624,
+      "learning_rate": 5.1600000000000006e-06,
+      "loss": 1.3391,
+      "step": 172
+    },
+    {
+      "epoch": 0.028702973993114604,
+      "grad_norm": 32.6150380329613,
+      "learning_rate": 5.19e-06,
+      "loss": 1.3314,
+      "step": 173
+    },
+    {
+      "epoch": 0.028868887137583475,
+      "grad_norm": 31.955882800699367,
+      "learning_rate": 5.22e-06,
+      "loss": 1.3339,
+      "step": 174
+    },
+    {
+      "epoch": 0.029034800282052346,
+      "grad_norm": 31.40356236669499,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 1.3184,
+      "step": 175
+    },
+    {
+      "epoch": 0.029200713426521217,
+      "grad_norm": 30.681143239334805,
+      "learning_rate": 5.28e-06,
+      "loss": 1.3139,
+      "step": 176
+    },
+    {
+      "epoch": 0.02936662657099009,
+      "grad_norm": 30.21502637398018,
+      "learning_rate": 5.31e-06,
+      "loss": 1.3107,
+      "step": 177
+    },
+    {
+      "epoch": 0.029532539715458956,
+      "grad_norm": 29.92516372692031,
+      "learning_rate": 5.3400000000000005e-06,
+      "loss": 1.3068,
+      "step": 178
+    },
+    {
+      "epoch": 0.029698452859927827,
+      "grad_norm": 30.018033153252357,
+      "learning_rate": 5.37e-06,
+      "loss": 1.2877,
+      "step": 179
+    },
+    {
+      "epoch": 0.029864366004396698,
+      "grad_norm": 29.63626860215389,
+      "learning_rate": 5.4e-06,
+      "loss": 1.2757,
+      "step": 180
+    },
+    {
+      "epoch": 0.03003027914886557,
+      "grad_norm": 29.205952566984713,
+      "learning_rate": 5.4300000000000005e-06,
+      "loss": 1.2624,
+      "step": 181
+    },
+    {
+      "epoch": 0.03019619229333444,
+      "grad_norm": 28.323326976872018,
+      "learning_rate": 5.46e-06,
+      "loss": 1.2636,
+      "step": 182
+    },
+    {
+      "epoch": 0.03036210543780331,
+      "grad_norm": 28.327391694947394,
+      "learning_rate": 5.49e-06,
+      "loss": 1.2445,
+      "step": 183
+    },
+    {
+      "epoch": 0.03052801858227218,
+      "grad_norm": 27.538465732946584,
+      "learning_rate": 5.52e-06,
+      "loss": 1.2553,
+      "step": 184
+    },
+    {
+      "epoch": 0.03069393172674105,
+      "grad_norm": 27.86388136213645,
+      "learning_rate": 5.55e-06,
+      "loss": 1.2187,
+      "step": 185
+    },
+    {
+      "epoch": 0.03085984487120992,
+      "grad_norm": 27.136153982898104,
+      "learning_rate": 5.58e-06,
+      "loss": 1.2264,
+      "step": 186
+    },
+    {
+      "epoch": 0.031025758015678792,
+      "grad_norm": 26.542801786628125,
+      "learning_rate": 5.61e-06,
+      "loss": 1.2271,
+      "step": 187
+    },
+    {
+      "epoch": 0.031191671160147663,
+      "grad_norm": 26.229276061707733,
+      "learning_rate": 5.64e-06,
+      "loss": 1.2213,
+      "step": 188
+    },
+    {
+      "epoch": 0.031357584304616534,
+      "grad_norm": 25.655150823462403,
+      "learning_rate": 5.67e-06,
+      "loss": 1.2131,
+      "step": 189
+    },
+    {
+      "epoch": 0.031523497449085405,
+      "grad_norm": 25.43762363935692,
+      "learning_rate": 5.7e-06,
+      "loss": 1.2009,
+      "step": 190
+    },
+    {
+      "epoch": 0.031689410593554276,
+      "grad_norm": 24.606418616190336,
+      "learning_rate": 5.73e-06,
+      "loss": 1.2105,
+      "step": 191
+    },
+    {
+      "epoch": 0.03185532373802315,
+      "grad_norm": 24.297360245690335,
+      "learning_rate": 5.76e-06,
+      "loss": 1.1902,
+      "step": 192
+    },
+    {
+      "epoch": 0.03202123688249202,
+      "grad_norm": 24.37416138144368,
+      "learning_rate": 5.79e-06,
+      "loss": 1.183,
+      "step": 193
+    },
+    {
+      "epoch": 0.03218715002696088,
+      "grad_norm": 23.58889140972946,
+      "learning_rate": 5.819999999999999e-06,
+      "loss": 1.1951,
+      "step": 194
+    },
+    {
+      "epoch": 0.032353063171429754,
+      "grad_norm": 23.192738038885675,
+      "learning_rate": 5.850000000000001e-06,
+      "loss": 1.1846,
+      "step": 195
+    },
+    {
+      "epoch": 0.032518976315898625,
+      "grad_norm": 23.067808075903493,
+      "learning_rate": 5.8800000000000005e-06,
+      "loss": 1.189,
+      "step": 196
+    },
+    {
+      "epoch": 0.032684889460367496,
+      "grad_norm": 22.129640030521287,
+      "learning_rate": 5.909999999999999e-06,
+      "loss": 1.1902,
+      "step": 197
+    },
+    {
+      "epoch": 0.03285080260483637,
+      "grad_norm": 22.580085955752832,
+      "learning_rate": 5.940000000000001e-06,
+      "loss": 1.142,
+      "step": 198
+    },
+    {
+      "epoch": 0.03301671574930524,
+      "grad_norm": 22.001422436219375,
+      "learning_rate": 5.9700000000000004e-06,
+      "loss": 1.1451,
+      "step": 199
+    },
+    {
+      "epoch": 0.03318262889377411,
+      "grad_norm": 21.511559365669424,
+      "learning_rate": 6e-06,
+      "loss": 1.157,
+      "step": 200
+    },
+    {
+      "epoch": 0.03334854203824298,
+      "grad_norm": 21.350738678737287,
+      "learning_rate": 6.030000000000001e-06,
+      "loss": 1.1548,
+      "step": 201
+    },
+    {
+      "epoch": 0.03351445518271185,
+      "grad_norm": 21.213322714991193,
+      "learning_rate": 6.0600000000000004e-06,
+      "loss": 1.1302,
+      "step": 202
+    },
+    {
+      "epoch": 0.03368036832718072,
+      "grad_norm": 20.96284799616787,
+      "learning_rate": 6.09e-06,
+      "loss": 1.13,
+      "step": 203
+    },
+    {
+      "epoch": 0.03384628147164959,
+      "grad_norm": 20.239673660306355,
+      "learning_rate": 6.120000000000001e-06,
+      "loss": 1.1364,
+      "step": 204
+    },
+    {
+      "epoch": 0.034012194616118464,
+      "grad_norm": 20.140657358498693,
+      "learning_rate": 6.15e-06,
+      "loss": 1.1313,
+      "step": 205
+    },
+    {
+      "epoch": 0.034178107760587335,
+      "grad_norm": 19.752608255172916,
+      "learning_rate": 6.18e-06,
+      "loss": 1.1187,
+      "step": 206
+    },
+    {
+      "epoch": 0.034344020905056206,
+      "grad_norm": 19.351629911713136,
+      "learning_rate": 6.21e-06,
+      "loss": 1.1194,
+      "step": 207
+    },
+    {
+      "epoch": 0.03450993404952507,
+      "grad_norm": 19.1529582255177,
+      "learning_rate": 6.24e-06,
+      "loss": 1.1175,
+      "step": 208
+    },
+    {
+      "epoch": 0.03467584719399394,
+      "grad_norm": 18.921466270786127,
+      "learning_rate": 6.27e-06,
+      "loss": 1.1145,
+      "step": 209
+    },
+    {
+      "epoch": 0.03484176033846281,
+      "grad_norm": 18.439640178944842,
+      "learning_rate": 6.3e-06,
+      "loss": 1.106,
+      "step": 210
+    },
+    {
+      "epoch": 0.035007673482931684,
+      "grad_norm": 18.592213045939108,
+      "learning_rate": 6.33e-06,
+      "loss": 1.0832,
+      "step": 211
+    },
+    {
+      "epoch": 0.035173586627400555,
+      "grad_norm": 18.02782578467052,
+      "learning_rate": 6.36e-06,
+      "loss": 1.1019,
+      "step": 212
+    },
+    {
+      "epoch": 0.035339499771869426,
+      "grad_norm": 18.0166852155525,
+      "learning_rate": 6.39e-06,
+      "loss": 1.0737,
+      "step": 213
+    },
+    {
+      "epoch": 0.0355054129163383,
+      "grad_norm": 17.781239753798864,
+      "learning_rate": 6.42e-06,
+      "loss": 1.0775,
+      "step": 214
+    },
+    {
+      "epoch": 0.03567132606080717,
+      "grad_norm": 17.221262259530782,
+      "learning_rate": 6.45e-06,
+      "loss": 1.1112,
+      "step": 215
+    },
+    {
+      "epoch": 0.03583723920527604,
+      "grad_norm": 16.865632842715378,
+      "learning_rate": 6.48e-06,
+      "loss": 1.0933,
+      "step": 216
+    },
+    {
+      "epoch": 0.03600315234974491,
+      "grad_norm": 16.90995396668004,
+      "learning_rate": 6.51e-06,
+      "loss": 1.0786,
+      "step": 217
+    },
+    {
+      "epoch": 0.03616906549421378,
+      "grad_norm": 16.502587446764466,
+      "learning_rate": 6.54e-06,
+      "loss": 1.0824,
+      "step": 218
+    },
+    {
+      "epoch": 0.03633497863868265,
+      "grad_norm": 16.396720584098915,
+      "learning_rate": 6.57e-06,
+      "loss": 1.0666,
+      "step": 219
+    },
+    {
+      "epoch": 0.03650089178315152,
+      "grad_norm": 16.0958491226786,
+      "learning_rate": 6.5999999999999995e-06,
+      "loss": 1.0587,
+      "step": 220
+    },
+    {
+      "epoch": 0.03666680492762039,
+      "grad_norm": 15.781394542209787,
+      "learning_rate": 6.63e-06,
+      "loss": 1.0913,
+      "step": 221
+    },
+    {
+      "epoch": 0.03683271807208926,
+      "grad_norm": 15.526150669252083,
+      "learning_rate": 6.66e-06,
+      "loss": 1.0685,
+      "step": 222
+    },
+    {
+      "epoch": 0.03699863121655813,
+      "grad_norm": 15.295714129778606,
+      "learning_rate": 6.6899999999999995e-06,
+      "loss": 1.0688,
+      "step": 223
+    },
+    {
+      "epoch": 0.037164544361027,
+      "grad_norm": 14.962595843836096,
+      "learning_rate": 6.720000000000001e-06,
+      "loss": 1.0663,
+      "step": 224
+    },
+    {
+      "epoch": 0.03733045750549587,
+      "grad_norm": 14.845441469454471,
+      "learning_rate": 6.75e-06,
+      "loss": 1.0487,
+      "step": 225
+    },
+    {
+      "epoch": 0.03749637064996474,
+      "grad_norm": 14.669509278037967,
+      "learning_rate": 6.7799999999999995e-06,
+      "loss": 1.058,
+      "step": 226
+    },
+    {
+      "epoch": 0.037662283794433614,
+      "grad_norm": 14.550072253222602,
+      "learning_rate": 6.810000000000001e-06,
+      "loss": 1.0425,
+      "step": 227
+    },
+    {
+      "epoch": 0.037828196938902485,
+      "grad_norm": 14.382438880655664,
+      "learning_rate": 6.840000000000001e-06,
+      "loss": 1.0367,
+      "step": 228
+    },
+    {
+      "epoch": 0.037994110083371356,
+      "grad_norm": 14.274435292405178,
+      "learning_rate": 6.87e-06,
+      "loss": 1.0414,
+      "step": 229
+    },
+    {
+      "epoch": 0.03816002322784023,
+      "grad_norm": 14.07020130770808,
+      "learning_rate": 6.900000000000001e-06,
+      "loss": 1.0264,
+      "step": 230
+    },
+    {
+      "epoch": 0.0383259363723091,
+      "grad_norm": 13.49479267804676,
+      "learning_rate": 6.9300000000000006e-06,
+      "loss": 1.0438,
+      "step": 231
+    },
+    {
+      "epoch": 0.03849184951677797,
+      "grad_norm": 13.554935453999114,
+      "learning_rate": 6.96e-06,
+      "loss": 1.0426,
+      "step": 232
+    },
+    {
+      "epoch": 0.03865776266124684,
+      "grad_norm": 13.55923875969836,
+      "learning_rate": 6.99e-06,
+      "loss": 0.9961,
+      "step": 233
+    },
+    {
+      "epoch": 0.038823675805715704,
+      "grad_norm": 12.9140401944659,
+      "learning_rate": 7.0200000000000006e-06,
+      "loss": 1.0416,
+      "step": 234
+    },
+    {
+      "epoch": 0.038989588950184575,
+      "grad_norm": 13.324650059397225,
+      "learning_rate": 7.05e-06,
+      "loss": 0.9825,
+      "step": 235
+    },
+    {
+      "epoch": 0.03915550209465345,
+      "grad_norm": 12.937116536824265,
+      "learning_rate": 7.08e-06,
+      "loss": 1.0103,
+      "step": 236
+    },
+    {
+      "epoch": 0.03932141523912232,
+      "grad_norm": 12.744639401075105,
+      "learning_rate": 7.1100000000000005e-06,
+      "loss": 1.0198,
+      "step": 237
+    },
+    {
+      "epoch": 0.03948732838359119,
+      "grad_norm": 12.472143264741263,
+      "learning_rate": 7.14e-06,
+      "loss": 1.0117,
+      "step": 238
+    },
+    {
+      "epoch": 0.03965324152806006,
+      "grad_norm": 12.428804263701728,
+      "learning_rate": 7.17e-06,
+      "loss": 0.9982,
+      "step": 239
+    },
+    {
+      "epoch": 0.03981915467252893,
+      "grad_norm": 12.136592772982283,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.0175,
+      "step": 240
+    },
+    {
+      "epoch": 0.0399850678169978,
+      "grad_norm": 12.053549043031309,
+      "learning_rate": 7.23e-06,
+      "loss": 1.0027,
+      "step": 241
+    },
+    {
+      "epoch": 0.04015098096146667,
+      "grad_norm": 11.843388086833002,
+      "learning_rate": 7.26e-06,
+      "loss": 1.0132,
+      "step": 242
+    },
+    {
+      "epoch": 0.040316894105935544,
+      "grad_norm": 11.628322273315547,
+      "learning_rate": 7.2900000000000005e-06,
+      "loss": 1.0017,
+      "step": 243
+    },
+    {
+      "epoch": 0.040482807250404415,
+      "grad_norm": 11.553787753691592,
+      "learning_rate": 7.32e-06,
+      "loss": 0.9988,
+      "step": 244
+    },
+    {
+      "epoch": 0.040648720394873286,
+      "grad_norm": 11.308136922354493,
+      "learning_rate": 7.35e-06,
+      "loss": 1.0267,
+      "step": 245
+    },
+    {
+      "epoch": 0.04081463353934216,
+      "grad_norm": 11.038599061997045,
+      "learning_rate": 7.38e-06,
+      "loss": 1.0145,
+      "step": 246
+    },
+    {
+      "epoch": 0.04098054668381103,
+      "grad_norm": 11.181030653605738,
+      "learning_rate": 7.41e-06,
+      "loss": 0.9889,
+      "step": 247
+    },
+    {
+      "epoch": 0.04114645982827989,
+      "grad_norm": 10.86832281657928,
+      "learning_rate": 7.44e-06,
+      "loss": 0.9852,
+      "step": 248
+    },
+    {
+      "epoch": 0.04131237297274876,
+      "grad_norm": 10.554939136339671,
+      "learning_rate": 7.47e-06,
+      "loss": 0.982,
+      "step": 249
+    },
+    {
+      "epoch": 0.041478286117217635,
+      "grad_norm": 10.568145863035252,
+      "learning_rate": 7.5e-06,
+      "loss": 0.9821,
+      "step": 250
+    },
+    {
+      "epoch": 0.041644199261686506,
+      "grad_norm": 10.308386741012788,
+      "learning_rate": 7.53e-06,
+      "loss": 0.9996,
+      "step": 251
+    },
+    {
+      "epoch": 0.04181011240615538,
+      "grad_norm": 10.372751036409273,
+      "learning_rate": 7.56e-06,
+      "loss": 0.9551,
+      "step": 252
+    },
+    {
+      "epoch": 0.04197602555062425,
+      "grad_norm": 10.072016333214728,
+      "learning_rate": 7.590000000000001e-06,
+      "loss": 0.9793,
+      "step": 253
+    },
+    {
+      "epoch": 0.04214193869509312,
+      "grad_norm": 9.971211572091233,
+      "learning_rate": 7.62e-06,
+      "loss": 0.9783,
+      "step": 254
+    },
+    {
+      "epoch": 0.04230785183956199,
+      "grad_norm": 9.805606829989866,
+      "learning_rate": 7.65e-06,
+      "loss": 0.9777,
+      "step": 255
+    },
+    {
+      "epoch": 0.04247376498403086,
+      "grad_norm": 9.511714368090708,
+      "learning_rate": 7.680000000000001e-06,
+      "loss": 0.9896,
+      "step": 256
+    },
+    {
+      "epoch": 0.04263967812849973,
+      "grad_norm": 9.793257406261617,
+      "learning_rate": 7.71e-06,
+      "loss": 0.9592,
+      "step": 257
+    },
+    {
+      "epoch": 0.0428055912729686,
+      "grad_norm": 9.489765564713549,
+      "learning_rate": 7.74e-06,
+      "loss": 0.9677,
+      "step": 258
+    },
+    {
+      "epoch": 0.042971504417437474,
+      "grad_norm": 9.522682824223127,
+      "learning_rate": 7.77e-06,
+      "loss": 0.9522,
+      "step": 259
+    },
+    {
+      "epoch": 0.043137417561906345,
+      "grad_norm": 9.31316057534312,
+      "learning_rate": 7.8e-06,
+      "loss": 0.9523,
+      "step": 260
+    },
+    {
+      "epoch": 0.04330333070637521,
+      "grad_norm": 9.191006657018935,
+      "learning_rate": 7.83e-06,
+      "loss": 0.9491,
+      "step": 261
+    },
+    {
+      "epoch": 0.04346924385084408,
+      "grad_norm": 9.005774575637536,
+      "learning_rate": 7.86e-06,
+      "loss": 0.9776,
+      "step": 262
+    },
+    {
+      "epoch": 0.04363515699531295,
+      "grad_norm": 8.944376972994409,
+      "learning_rate": 7.89e-06,
+      "loss": 0.9724,
+      "step": 263
+    },
+    {
+      "epoch": 0.04380107013978182,
+      "grad_norm": 8.880616489633164,
+      "learning_rate": 7.92e-06,
+      "loss": 0.9509,
+      "step": 264
+    },
+    {
+      "epoch": 0.043966983284250694,
+      "grad_norm": 10.295937531650397,
+      "learning_rate": 7.95e-06,
+      "loss": 0.9555,
+      "step": 265
+    },
+    {
+      "epoch": 0.044132896428719565,
+      "grad_norm": 8.742470979369012,
+      "learning_rate": 7.98e-06,
+      "loss": 0.9365,
+      "step": 266
+    },
+    {
+      "epoch": 0.044298809573188436,
+      "grad_norm": 8.59286130501365,
+      "learning_rate": 8.01e-06,
+      "loss": 0.9479,
+      "step": 267
+    },
+    {
+      "epoch": 0.04446472271765731,
+      "grad_norm": 8.522351957971027,
+      "learning_rate": 8.04e-06,
+      "loss": 0.9424,
+      "step": 268
+    },
+    {
+      "epoch": 0.04463063586212618,
+      "grad_norm": 8.330738085855412,
+      "learning_rate": 8.069999999999999e-06,
+      "loss": 0.9485,
+      "step": 269
+    },
+    {
+      "epoch": 0.04479654900659505,
+      "grad_norm": 8.18707701163648,
+      "learning_rate": 8.1e-06,
+      "loss": 0.9649,
+      "step": 270
+    },
+    {
+      "epoch": 0.04496246215106392,
+      "grad_norm": 8.222407203836411,
+      "learning_rate": 8.13e-06,
+      "loss": 0.9468,
+      "step": 271
+    },
+    {
+      "epoch": 0.04512837529553279,
+      "grad_norm": 8.094229801443008,
+      "learning_rate": 8.16e-06,
+      "loss": 0.9418,
+      "step": 272
+    },
+    {
+      "epoch": 0.04529428844000166,
+      "grad_norm": 7.940256424590958,
+      "learning_rate": 8.190000000000001e-06,
+      "loss": 0.9553,
+      "step": 273
+    },
+    {
+      "epoch": 0.045460201584470526,
+      "grad_norm": 7.813320194657039,
+      "learning_rate": 8.220000000000001e-06,
+      "loss": 0.9505,
+      "step": 274
+    },
+    {
+      "epoch": 0.0456261147289394,
+      "grad_norm": 7.854878419977774,
+      "learning_rate": 8.25e-06,
+      "loss": 0.9382,
+      "step": 275
+    },
+    {
+      "epoch": 0.04579202787340827,
+      "grad_norm": 7.774670072265875,
+      "learning_rate": 8.28e-06,
+      "loss": 0.9364,
+      "step": 276
+    },
+    {
+      "epoch": 0.04595794101787714,
+      "grad_norm": 7.567068884700512,
+      "learning_rate": 8.31e-06,
+      "loss": 0.9388,
+      "step": 277
+    },
+    {
+      "epoch": 0.04612385416234601,
+      "grad_norm": 7.605183779660865,
+      "learning_rate": 8.34e-06,
+      "loss": 0.9403,
+      "step": 278
+    },
+    {
+      "epoch": 0.04628976730681488,
+      "grad_norm": 7.566454167028441,
+      "learning_rate": 8.370000000000001e-06,
+      "loss": 0.9218,
+      "step": 279
+    },
+    {
+      "epoch": 0.04645568045128375,
+      "grad_norm": 7.3469120826750265,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 0.932,
+      "step": 280
+    },
+    {
+      "epoch": 0.046621593595752624,
+      "grad_norm": 7.26408565369186,
+      "learning_rate": 8.43e-06,
+      "loss": 0.9332,
+      "step": 281
+    },
+    {
+      "epoch": 0.046787506740221495,
+      "grad_norm": 7.126714812653672,
+      "learning_rate": 8.46e-06,
+      "loss": 0.9408,
+      "step": 282
+    },
+    {
+      "epoch": 0.046953419884690366,
+      "grad_norm": 7.160875540112923,
+      "learning_rate": 8.49e-06,
+      "loss": 0.9326,
+      "step": 283
+    },
+    {
+      "epoch": 0.04711933302915924,
+      "grad_norm": 6.989719202314339,
+      "learning_rate": 8.52e-06,
+      "loss": 0.9258,
+      "step": 284
+    },
+    {
+      "epoch": 0.04728524617362811,
+      "grad_norm": 6.897188941855704,
+      "learning_rate": 8.55e-06,
+      "loss": 0.9464,
+      "step": 285
+    },
+    {
+      "epoch": 0.04745115931809698,
+      "grad_norm": 6.841146949131663,
+      "learning_rate": 8.580000000000001e-06,
+      "loss": 0.9452,
+      "step": 286
+    },
+    {
+      "epoch": 0.04761707246256585,
+      "grad_norm": 6.841546310249179,
+      "learning_rate": 8.61e-06,
+      "loss": 0.9178,
+      "step": 287
+    },
+    {
+      "epoch": 0.047782985607034714,
+      "grad_norm": 6.544761686641858,
+      "learning_rate": 8.64e-06,
+      "loss": 0.9603,
+      "step": 288
+    },
+    {
+      "epoch": 0.047948898751503585,
+      "grad_norm": 6.614087506733643,
+      "learning_rate": 8.67e-06,
+      "loss": 0.916,
+      "step": 289
+    },
+    {
+      "epoch": 0.048114811895972456,
+      "grad_norm": 6.6124048640237945,
+      "learning_rate": 8.7e-06,
+      "loss": 0.9129,
+      "step": 290
+    },
+    {
+      "epoch": 0.04828072504044133,
+      "grad_norm": 6.408146101594846,
+      "learning_rate": 8.73e-06,
+      "loss": 0.9225,
+      "step": 291
+    },
+    {
+      "epoch": 0.0484466381849102,
+      "grad_norm": 6.502364355003014,
+      "learning_rate": 8.76e-06,
+      "loss": 0.9031,
+      "step": 292
+    },
+    {
+      "epoch": 0.04861255132937907,
+      "grad_norm": 6.331848917593811,
+      "learning_rate": 8.79e-06,
+      "loss": 0.9251,
+      "step": 293
+    },
+    {
+      "epoch": 0.04877846447384794,
+      "grad_norm": 6.312668466628936,
+      "learning_rate": 8.82e-06,
+      "loss": 0.9046,
+      "step": 294
+    },
+    {
+      "epoch": 0.04894437761831681,
+      "grad_norm": 6.120718833347466,
+      "learning_rate": 8.85e-06,
+      "loss": 0.9387,
+      "step": 295
+    },
+    {
+      "epoch": 0.04911029076278568,
+      "grad_norm": 6.146017531230814,
+      "learning_rate": 8.88e-06,
+      "loss": 0.9065,
+      "step": 296
+    },
+    {
+      "epoch": 0.049276203907254554,
+      "grad_norm": 6.2298783704404475,
+      "learning_rate": 8.91e-06,
+      "loss": 0.8897,
+      "step": 297
+    },
+    {
+      "epoch": 0.049442117051723425,
+      "grad_norm": 6.01881306966524,
+      "learning_rate": 8.939999999999999e-06,
+      "loss": 0.9179,
+      "step": 298
+    },
+    {
+      "epoch": 0.049608030196192296,
+      "grad_norm": 6.018783760260834,
+      "learning_rate": 8.97e-06,
+      "loss": 0.898,
+      "step": 299
+    },
+    {
+      "epoch": 0.04977394334066117,
+      "grad_norm": 5.909905610679987,
+      "learning_rate": 9e-06,
+      "loss": 0.9119,
+      "step": 300
+    },
+    {
+      "epoch": 0.04993985648513003,
+      "grad_norm": 5.734121028769744,
+      "learning_rate": 8.999999975052319e-06,
+      "loss": 0.9342,
+      "step": 301
+    },
+    {
+      "epoch": 0.0501057696295989,
+      "grad_norm": 5.78200438759076,
+      "learning_rate": 8.999999900209277e-06,
+      "loss": 0.9036,
+      "step": 302
+    },
+    {
+      "epoch": 0.05027168277406777,
+      "grad_norm": 5.667277372988322,
+      "learning_rate": 8.999999775470875e-06,
+      "loss": 0.9075,
+      "step": 303
+    },
+    {
+      "epoch": 0.050437595918536644,
+      "grad_norm": 5.686905178208374,
+      "learning_rate": 8.999999600837113e-06,
+      "loss": 0.9196,
+      "step": 304
+    },
+    {
+      "epoch": 0.050603509063005515,
+      "grad_norm": 5.589145090418338,
+      "learning_rate": 8.999999376307994e-06,
+      "loss": 0.9139,
+      "step": 305
+    },
+    {
+      "epoch": 0.050769422207474386,
+      "grad_norm": 5.53466927124937,
+      "learning_rate": 8.99999910188352e-06,
+      "loss": 0.9135,
+      "step": 306
+    },
+    {
+      "epoch": 0.05093533535194326,
+      "grad_norm": 9.215054382271381,
+      "learning_rate": 8.999998777563696e-06,
+      "loss": 0.9075,
+      "step": 307
+    },
+    {
+      "epoch": 0.05110124849641213,
+      "grad_norm": 5.50940374670424,
+      "learning_rate": 8.99999840334852e-06,
+      "loss": 0.906,
+      "step": 308
+    },
+    {
+      "epoch": 0.051267161640881,
+      "grad_norm": 5.315125426110582,
+      "learning_rate": 8.999997979238004e-06,
+      "loss": 0.9362,
+      "step": 309
+    },
+    {
+      "epoch": 0.05143307478534987,
+      "grad_norm": 5.456218191120494,
+      "learning_rate": 8.999997505232148e-06,
+      "loss": 0.8974,
+      "step": 310
+    },
+    {
+      "epoch": 0.05159898792981874,
+      "grad_norm": 5.3458253139310905,
+      "learning_rate": 8.999996981330957e-06,
+      "loss": 0.9052,
+      "step": 311
+    },
+    {
+      "epoch": 0.05176490107428761,
+      "grad_norm": 5.428852344895628,
+      "learning_rate": 8.999996407534438e-06,
+      "loss": 0.8927,
+      "step": 312
+    },
+    {
+      "epoch": 0.051930814218756484,
+      "grad_norm": 5.362891578705491,
+      "learning_rate": 8.999995783842599e-06,
+      "loss": 0.8913,
+      "step": 313
+    },
+    {
+      "epoch": 0.052096727363225355,
+      "grad_norm": 5.2337978422263385,
+      "learning_rate": 8.999995110255444e-06,
+      "loss": 0.9175,
+      "step": 314
+    },
+    {
+      "epoch": 0.05226264050769422,
+      "grad_norm": 5.176887043314268,
+      "learning_rate": 8.999994386772981e-06,
+      "loss": 0.9003,
+      "step": 315
+    },
+    {
+      "epoch": 0.05242855365216309,
+      "grad_norm": 5.184446115994752,
+      "learning_rate": 8.999993613395219e-06,
+      "loss": 0.8943,
+      "step": 316
+    },
+    {
+      "epoch": 0.05259446679663196,
+      "grad_norm": 5.159248082806763,
+      "learning_rate": 8.999992790122167e-06,
+      "loss": 0.8893,
+      "step": 317
+    },
+    {
+      "epoch": 0.05276037994110083,
+      "grad_norm": 5.083647157877446,
+      "learning_rate": 8.999991916953832e-06,
+      "loss": 0.8921,
+      "step": 318
+    },
+    {
+      "epoch": 0.0529262930855697,
+      "grad_norm": 5.0793184864339045,
+      "learning_rate": 8.999990993890227e-06,
+      "loss": 0.8891,
+      "step": 319
+    },
+    {
+      "epoch": 0.053092206230038574,
+      "grad_norm": 4.969318514135246,
+      "learning_rate": 8.99999002093136e-06,
+      "loss": 0.9008,
+      "step": 320
+    },
+    {
+      "epoch": 0.053258119374507445,
+      "grad_norm": 4.866162400058146,
+      "learning_rate": 8.999988998077239e-06,
+      "loss": 0.9074,
+      "step": 321
+    },
+    {
+      "epoch": 0.05342403251897632,
+      "grad_norm": 4.871718580190094,
+      "learning_rate": 8.999987925327882e-06,
+      "loss": 0.9105,
+      "step": 322
+    },
+    {
+      "epoch": 0.05358994566344519,
+      "grad_norm": 4.950770420419117,
+      "learning_rate": 8.999986802683295e-06,
+      "loss": 0.8792,
+      "step": 323
+    },
+    {
+      "epoch": 0.05375585880791406,
+      "grad_norm": 4.8879460632967735,
+      "learning_rate": 8.999985630143494e-06,
+      "loss": 0.8691,
+      "step": 324
+    },
+    {
+      "epoch": 0.05392177195238293,
+      "grad_norm": 4.776820682760442,
+      "learning_rate": 8.999984407708489e-06,
+      "loss": 0.9004,
+      "step": 325
+    },
+    {
+      "epoch": 0.0540876850968518,
+      "grad_norm": 4.734078577215334,
+      "learning_rate": 8.999983135378296e-06,
+      "loss": 0.8811,
+      "step": 326
+    },
+    {
+      "epoch": 0.05425359824132067,
+      "grad_norm": 4.702414692785676,
+      "learning_rate": 8.99998181315293e-06,
+      "loss": 0.9019,
+      "step": 327
+    },
+    {
+      "epoch": 0.054419511385789536,
+      "grad_norm": 4.552715310423402,
+      "learning_rate": 8.999980441032402e-06,
+      "loss": 0.9042,
+      "step": 328
+    },
+    {
+      "epoch": 0.05458542453025841,
+      "grad_norm": 4.507915660100879,
+      "learning_rate": 8.999979019016731e-06,
+      "loss": 0.9119,
+      "step": 329
+    },
+    {
+      "epoch": 0.05475133767472728,
+      "grad_norm": 4.523492938568341,
+      "learning_rate": 8.99997754710593e-06,
+      "loss": 0.8932,
+      "step": 330
+    },
+    {
+      "epoch": 0.05491725081919615,
+      "grad_norm": 4.6297463894007365,
+      "learning_rate": 8.999976025300017e-06,
+      "loss": 0.8477,
+      "step": 331
+    },
+    {
+      "epoch": 0.05508316396366502,
+      "grad_norm": 4.51195869056777,
+      "learning_rate": 8.999974453599007e-06,
+      "loss": 0.8758,
+      "step": 332
+    },
+    {
+      "epoch": 0.05524907710813389,
+      "grad_norm": 4.477645446414166,
+      "learning_rate": 8.99997283200292e-06,
+      "loss": 0.8538,
+      "step": 333
+    },
+    {
+      "epoch": 0.05541499025260276,
+      "grad_norm": 4.393730342387632,
+      "learning_rate": 8.999971160511772e-06,
+      "loss": 0.8964,
+      "step": 334
+    },
+    {
+      "epoch": 0.05558090339707163,
+      "grad_norm": 4.332577691208558,
+      "learning_rate": 8.999969439125582e-06,
+      "loss": 0.9002,
+      "step": 335
+    },
+    {
+      "epoch": 0.055746816541540505,
+      "grad_norm": 4.272904462844758,
+      "learning_rate": 8.999967667844369e-06,
+      "loss": 0.8936,
+      "step": 336
+    },
+    {
+      "epoch": 0.055912729686009376,
+      "grad_norm": 4.284138641618394,
+      "learning_rate": 8.999965846668153e-06,
+      "loss": 0.8806,
+      "step": 337
+    },
+    {
+      "epoch": 0.05607864283047825,
+      "grad_norm": 4.245997951478823,
+      "learning_rate": 8.999963975596955e-06,
+      "loss": 0.8933,
+      "step": 338
+    },
+    {
+      "epoch": 0.05624455597494712,
+      "grad_norm": 4.262350308738906,
+      "learning_rate": 8.999962054630794e-06,
+      "loss": 0.8672,
+      "step": 339
+    },
+    {
+      "epoch": 0.05641046911941599,
+      "grad_norm": 4.218320423707499,
+      "learning_rate": 8.999960083769692e-06,
+      "loss": 0.8697,
+      "step": 340
+    },
+    {
+      "epoch": 0.05657638226388485,
+      "grad_norm": 4.161434572377091,
+      "learning_rate": 8.999958063013673e-06,
+      "loss": 0.8878,
+      "step": 341
+    },
+    {
+      "epoch": 0.056742295408353724,
+      "grad_norm": 4.0848883777831855,
+      "learning_rate": 8.999955992362754e-06,
+      "loss": 0.9028,
+      "step": 342
+    },
+    {
+      "epoch": 0.056908208552822595,
+      "grad_norm": 4.087674741684919,
+      "learning_rate": 8.999953871816964e-06,
+      "loss": 0.884,
+      "step": 343
+    },
+    {
+      "epoch": 0.057074121697291466,
+      "grad_norm": 4.077627298421648,
+      "learning_rate": 8.999951701376323e-06,
+      "loss": 0.8733,
+      "step": 344
+    },
+    {
+      "epoch": 0.05724003484176034,
+      "grad_norm": 4.014185881607656,
+      "learning_rate": 8.999949481040857e-06,
+      "loss": 0.873,
+      "step": 345
+    },
+    {
+      "epoch": 0.05740594798622921,
+      "grad_norm": 4.048242476037428,
+      "learning_rate": 8.999947210810587e-06,
+      "loss": 0.8612,
+      "step": 346
+    },
+    {
+      "epoch": 0.05757186113069808,
+      "grad_norm": 3.9790776791909286,
+      "learning_rate": 8.999944890685543e-06,
+      "loss": 0.8745,
+      "step": 347
+    },
+    {
+      "epoch": 0.05773777427516695,
+      "grad_norm": 3.9412637498937935,
+      "learning_rate": 8.999942520665747e-06,
+      "loss": 0.8751,
+      "step": 348
+    },
+    {
+      "epoch": 0.05790368741963582,
+      "grad_norm": 3.9243342606372216,
+      "learning_rate": 8.999940100751228e-06,
+      "loss": 0.8705,
+      "step": 349
+    },
+    {
+      "epoch": 0.05806960056410469,
+      "grad_norm": 3.8615778390716713,
+      "learning_rate": 8.99993763094201e-06,
+      "loss": 0.8912,
+      "step": 350
+    },
+    {
+      "epoch": 0.058235513708573564,
+      "grad_norm": 3.859140301595901,
+      "learning_rate": 8.999935111238122e-06,
+      "loss": 0.8694,
+      "step": 351
+    },
+    {
+      "epoch": 0.058401426853042435,
+      "grad_norm": 3.822341835829405,
+      "learning_rate": 8.999932541639593e-06,
+      "loss": 0.8742,
+      "step": 352
+    },
+    {
+      "epoch": 0.058567339997511306,
+      "grad_norm": 3.8184180233248943,
+      "learning_rate": 8.99992992214645e-06,
+      "loss": 0.8676,
+      "step": 353
+    },
+    {
+      "epoch": 0.05873325314198018,
+      "grad_norm": 3.801210717065968,
+      "learning_rate": 8.999927252758723e-06,
+      "loss": 0.8662,
+      "step": 354
+    },
+    {
+      "epoch": 0.05889916628644904,
+      "grad_norm": 3.780756273238635,
+      "learning_rate": 8.99992453347644e-06,
+      "loss": 0.8649,
+      "step": 355
+    },
+    {
+      "epoch": 0.05906507943091791,
+      "grad_norm": 3.7474392141127906,
+      "learning_rate": 8.999921764299633e-06,
+      "loss": 0.8583,
+      "step": 356
+    },
+    {
+      "epoch": 0.05923099257538678,
+      "grad_norm": 3.730060223515627,
+      "learning_rate": 8.999918945228331e-06,
+      "loss": 0.8589,
+      "step": 357
+    },
+    {
+      "epoch": 0.059396905719855654,
+      "grad_norm": 3.6811758989495567,
+      "learning_rate": 8.999916076262566e-06,
+      "loss": 0.8568,
+      "step": 358
+    },
+    {
+      "epoch": 0.059562818864324525,
+      "grad_norm": 3.6726514292781505,
+      "learning_rate": 8.999913157402371e-06,
+      "loss": 0.8573,
+      "step": 359
+    },
+    {
+      "epoch": 0.059728732008793396,
+      "grad_norm": 3.595206525156515,
+      "learning_rate": 8.999910188647777e-06,
+      "loss": 0.8742,
+      "step": 360
+    },
+    {
+      "epoch": 0.05989464515326227,
+      "grad_norm": 3.471677475337036,
+      "learning_rate": 8.999907169998819e-06,
+      "loss": 0.8869,
+      "step": 361
+    },
+    {
+      "epoch": 0.06006055829773114,
+      "grad_norm": 3.5659572607511163,
+      "learning_rate": 8.999904101455525e-06,
+      "loss": 0.8576,
+      "step": 362
+    },
+    {
+      "epoch": 0.06022647144220001,
+      "grad_norm": 3.494068972585631,
+      "learning_rate": 8.999900983017934e-06,
+      "loss": 0.8626,
+      "step": 363
+    },
+    {
+      "epoch": 0.06039238458666888,
+      "grad_norm": 3.5366577459151984,
+      "learning_rate": 8.99989781468608e-06,
+      "loss": 0.8641,
+      "step": 364
+    },
+    {
+      "epoch": 0.06055829773113775,
+      "grad_norm": 3.4640552235900026,
+      "learning_rate": 8.999894596459998e-06,
+      "loss": 0.8498,
+      "step": 365
+    },
+    {
+      "epoch": 0.06072421087560662,
+      "grad_norm": 3.4627533254161778,
+      "learning_rate": 8.999891328339722e-06,
+      "loss": 0.8701,
+      "step": 366
+    },
+    {
+      "epoch": 0.060890124020075494,
+      "grad_norm": 3.4543538438907455,
+      "learning_rate": 8.99988801032529e-06,
+      "loss": 0.8525,
+      "step": 367
+    },
+    {
+      "epoch": 0.06105603716454436,
+      "grad_norm": 3.4310968757128895,
+      "learning_rate": 8.999884642416736e-06,
+      "loss": 0.8653,
+      "step": 368
+    },
+    {
+      "epoch": 0.06122195030901323,
+      "grad_norm": 3.3538182281557583,
+      "learning_rate": 8.999881224614101e-06,
+      "loss": 0.8908,
+      "step": 369
+    },
+    {
+      "epoch": 0.0613878634534821,
+      "grad_norm": 3.3905994838285163,
+      "learning_rate": 8.999877756917422e-06,
+      "loss": 0.8521,
+      "step": 370
+    },
+    {
+      "epoch": 0.06155377659795097,
+      "grad_norm": 3.3231583523257604,
+      "learning_rate": 8.999874239326737e-06,
+      "loss": 0.865,
+      "step": 371
+    },
+    {
+      "epoch": 0.06171968974241984,
+      "grad_norm": 3.3486181559656334,
+      "learning_rate": 8.999870671842085e-06,
+      "loss": 0.8565,
+      "step": 372
+    },
+    {
+      "epoch": 0.06188560288688871,
+      "grad_norm": 3.391890189649907,
+      "learning_rate": 8.999867054463503e-06,
+      "loss": 0.8344,
+      "step": 373
+    },
+    {
+      "epoch": 0.062051516031357584,
+      "grad_norm": 3.2932669339414287,
+      "learning_rate": 8.999863387191034e-06,
+      "loss": 0.8668,
+      "step": 374
+    },
+    {
+      "epoch": 0.062217429175826455,
+      "grad_norm": 3.350955000113915,
+      "learning_rate": 8.99985967002472e-06,
+      "loss": 0.8253,
+      "step": 375
+    },
+    {
+      "epoch": 0.062383342320295326,
+      "grad_norm": 3.2330710537829304,
+      "learning_rate": 8.9998559029646e-06,
+      "loss": 0.8684,
+      "step": 376
+    },
+    {
+      "epoch": 0.0625492554647642,
+      "grad_norm": 3.2098638665770243,
+      "learning_rate": 8.999852086010716e-06,
+      "loss": 0.8628,
+      "step": 377
+    },
+    {
+      "epoch": 0.06271516860923307,
+      "grad_norm": 3.1892699537908156,
+      "learning_rate": 8.99984821916311e-06,
+      "loss": 0.8696,
+      "step": 378
+    },
+    {
+      "epoch": 0.06288108175370194,
+      "grad_norm": 3.2334974851490363,
+      "learning_rate": 8.999844302421825e-06,
+      "loss": 0.8387,
+      "step": 379
+    },
+    {
+      "epoch": 0.06304699489817081,
+      "grad_norm": 3.238152271931643,
+      "learning_rate": 8.999840335786906e-06,
+      "loss": 0.8276,
+      "step": 380
+    },
+    {
+      "epoch": 0.06321290804263968,
+      "grad_norm": 3.1331529105356974,
+      "learning_rate": 8.999836319258393e-06,
+      "loss": 0.8512,
+      "step": 381
+    },
+    {
+      "epoch": 0.06337882118710855,
+      "grad_norm": 3.152663308943621,
+      "learning_rate": 8.999832252836337e-06,
+      "loss": 0.8384,
+      "step": 382
+    },
+    {
+      "epoch": 0.06354473433157742,
+      "grad_norm": 3.084432904970283,
+      "learning_rate": 8.999828136520776e-06,
+      "loss": 0.8584,
+      "step": 383
+    },
+    {
+      "epoch": 0.0637106474760463,
+      "grad_norm": 3.1120289343036838,
+      "learning_rate": 8.999823970311762e-06,
+      "loss": 0.8443,
+      "step": 384
+    },
+    {
+      "epoch": 0.06387656062051517,
+      "grad_norm": 3.1157944117751373,
+      "learning_rate": 8.999819754209336e-06,
+      "loss": 0.8355,
+      "step": 385
+    },
+    {
+      "epoch": 0.06404247376498404,
+      "grad_norm": 3.016422516063702,
+      "learning_rate": 8.999815488213547e-06,
+      "loss": 0.8718,
+      "step": 386
+    },
+    {
+      "epoch": 0.06420838690945291,
+      "grad_norm": 3.0280804164647277,
+      "learning_rate": 8.999811172324443e-06,
+      "loss": 0.8525,
+      "step": 387
+    },
+    {
+      "epoch": 0.06437430005392177,
+      "grad_norm": 3.036473880069474,
+      "learning_rate": 8.999806806542072e-06,
+      "loss": 0.8644,
+      "step": 388
+    },
+    {
+      "epoch": 0.06454021319839064,
+      "grad_norm": 2.991716780189957,
+      "learning_rate": 8.99980239086648e-06,
+      "loss": 0.8474,
+      "step": 389
+    },
+    {
+      "epoch": 0.06470612634285951,
+      "grad_norm": 3.0744448937466635,
+      "learning_rate": 8.999797925297717e-06,
+      "loss": 0.8251,
+      "step": 390
+    },
+    {
+      "epoch": 0.06487203948732838,
+      "grad_norm": 2.970959395794443,
+      "learning_rate": 8.999793409835835e-06,
+      "loss": 0.8497,
+      "step": 391
+    },
+    {
+      "epoch": 0.06503795263179725,
+      "grad_norm": 2.942697893932714,
+      "learning_rate": 8.99978884448088e-06,
+      "loss": 0.8497,
+      "step": 392
+    },
+    {
+      "epoch": 0.06520386577626612,
+      "grad_norm": 2.9217535841307374,
+      "learning_rate": 8.999784229232905e-06,
+      "loss": 0.8638,
+      "step": 393
+    },
+    {
+      "epoch": 0.06536977892073499,
+      "grad_norm": 2.9558083426986483,
+      "learning_rate": 8.999779564091961e-06,
+      "loss": 0.8274,
+      "step": 394
+    },
+    {
+      "epoch": 0.06553569206520386,
+      "grad_norm": 2.885361223118148,
+      "learning_rate": 8.999774849058102e-06,
+      "loss": 0.8583,
+      "step": 395
+    },
+    {
+      "epoch": 0.06570160520967273,
+      "grad_norm": 2.8800300061377473,
+      "learning_rate": 8.999770084131374e-06,
+      "loss": 0.8499,
+      "step": 396
+    },
+    {
+      "epoch": 0.0658675183541416,
+      "grad_norm": 2.8665971529078695,
+      "learning_rate": 8.999765269311836e-06,
+      "loss": 0.8489,
+      "step": 397
+    },
+    {
+      "epoch": 0.06603343149861048,
+      "grad_norm": 2.8344547998127103,
+      "learning_rate": 8.999760404599538e-06,
+      "loss": 0.8432,
+      "step": 398
+    },
+    {
+      "epoch": 0.06619934464307935,
+      "grad_norm": 2.839791425822409,
+      "learning_rate": 8.999755489994537e-06,
+      "loss": 0.8407,
+      "step": 399
+    },
+    {
+      "epoch": 0.06636525778754822,
+      "grad_norm": 2.828150592505063,
+      "learning_rate": 8.999750525496884e-06,
+      "loss": 0.8217,
+      "step": 400
+    },
+    {
+      "epoch": 0.06585511054543962,
+      "grad_norm": 2.791449185667969,
+      "learning_rate": 8.99975071828438e-06,
+      "loss": 0.8651,
+      "step": 401
+    },
+    {
+      "epoch": 0.06601933775378235,
+      "grad_norm": 2.799198588104831,
+      "learning_rate": 8.999745757622507e-06,
+      "loss": 0.8615,
+      "step": 402
+    },
+    {
+      "epoch": 0.0661835649621251,
+      "grad_norm": 2.731196238394043,
+      "learning_rate": 8.99974074808897e-06,
+      "loss": 0.87,
+      "step": 403
+    },
+    {
+      "epoch": 0.06634779217046784,
+      "grad_norm": 2.7669579734667407,
+      "learning_rate": 8.999735689683818e-06,
+      "loss": 0.8304,
+      "step": 404
+    },
+    {
+      "epoch": 0.06651201937881059,
+      "grad_norm": 2.6933917829612226,
+      "learning_rate": 8.999730582407112e-06,
+      "loss": 0.8645,
+      "step": 405
+    },
+    {
+      "epoch": 0.06667624658715332,
+      "grad_norm": 2.7037937388444493,
+      "learning_rate": 8.999725426258905e-06,
+      "loss": 0.8522,
+      "step": 406
+    },
+    {
+      "epoch": 0.06684047379549607,
+      "grad_norm": 2.6529365860710947,
+      "learning_rate": 8.999720221239252e-06,
+      "loss": 0.8665,
+      "step": 407
+    },
+    {
+      "epoch": 0.0670047010038388,
+      "grad_norm": 2.6958361564015623,
+      "learning_rate": 8.999714967348209e-06,
+      "loss": 0.8421,
+      "step": 408
+    },
+    {
+      "epoch": 0.06716892821218155,
+      "grad_norm": 2.679623752008986,
+      "learning_rate": 8.999709664585836e-06,
+      "loss": 0.8444,
+      "step": 409
+    },
+    {
+      "epoch": 0.0673331554205243,
+      "grad_norm": 2.6846073911898247,
+      "learning_rate": 8.999704312952188e-06,
+      "loss": 0.8225,
+      "step": 410
+    },
+    {
+      "epoch": 0.06749738262886704,
+      "grad_norm": 2.6357646834424733,
+      "learning_rate": 8.999698912447324e-06,
+      "loss": 0.8428,
+      "step": 411
+    },
+    {
+      "epoch": 0.06766160983720979,
+      "grad_norm": 2.5864163075261515,
+      "learning_rate": 8.999693463071303e-06,
+      "loss": 0.8647,
+      "step": 412
+    },
+    {
+      "epoch": 0.06782583704555252,
+      "grad_norm": 2.649716139853906,
+      "learning_rate": 8.999687964824184e-06,
+      "loss": 0.8336,
+      "step": 413
+    },
+    {
+      "epoch": 0.06799006425389527,
+      "grad_norm": 2.6697160350910982,
+      "learning_rate": 8.999682417706028e-06,
+      "loss": 0.8315,
+      "step": 414
+    },
+    {
+      "epoch": 0.068154291462238,
+      "grad_norm": 2.614544893051611,
+      "learning_rate": 8.999676821716893e-06,
+      "loss": 0.833,
+      "step": 415
+    },
+    {
+      "epoch": 0.06831851867058075,
+      "grad_norm": 2.605311057811457,
+      "learning_rate": 8.99967117685684e-06,
+      "loss": 0.8354,
+      "step": 416
+    },
+    {
+      "epoch": 0.06848274587892349,
+      "grad_norm": 2.512994946017452,
+      "learning_rate": 8.999665483125932e-06,
+      "loss": 0.8527,
+      "step": 417
+    },
+    {
+      "epoch": 0.06864697308726624,
+      "grad_norm": 2.532102472565359,
+      "learning_rate": 8.999659740524227e-06,
+      "loss": 0.8526,
+      "step": 418
+    },
+    {
+      "epoch": 0.06881120029560897,
+      "grad_norm": 2.537552292844787,
+      "learning_rate": 8.999653949051792e-06,
+      "loss": 0.8593,
+      "step": 419
+    },
+    {
+      "epoch": 0.06897542750395172,
+      "grad_norm": 2.520981683215179,
+      "learning_rate": 8.999648108708689e-06,
+      "loss": 0.8409,
+      "step": 420
+    },
+    {
+      "epoch": 0.06913965471229445,
+      "grad_norm": 2.5307279062339947,
+      "learning_rate": 8.999642219494979e-06,
+      "loss": 0.8544,
+      "step": 421
+    },
+    {
+      "epoch": 0.0693038819206372,
+      "grad_norm": 2.513981724612271,
+      "learning_rate": 8.999636281410728e-06,
+      "loss": 0.838,
+      "step": 422
+    },
+    {
+      "epoch": 0.06946810912897994,
+      "grad_norm": 2.5196092111609443,
+      "learning_rate": 8.999630294456001e-06,
+      "loss": 0.822,
+      "step": 423
+    },
+    {
+      "epoch": 0.06963233633732269,
+      "grad_norm": 2.501508735280606,
+      "learning_rate": 8.99962425863086e-06,
+      "loss": 0.814,
+      "step": 424
+    },
+    {
+      "epoch": 0.06979656354566542,
+      "grad_norm": 2.463190118948678,
+      "learning_rate": 8.999618173935375e-06,
+      "loss": 0.821,
+      "step": 425
+    },
+    {
+      "epoch": 0.06996079075400817,
+      "grad_norm": 2.485170388269591,
+      "learning_rate": 8.999612040369608e-06,
+      "loss": 0.8129,
+      "step": 426
+    },
+    {
+      "epoch": 0.07012501796235092,
+      "grad_norm": 2.4535480317174696,
+      "learning_rate": 8.999605857933626e-06,
+      "loss": 0.8204,
+      "step": 427
+    },
+    {
+      "epoch": 0.07028924517069365,
+      "grad_norm": 2.4297331854286934,
+      "learning_rate": 8.999599626627498e-06,
+      "loss": 0.8228,
+      "step": 428
+    },
+    {
+      "epoch": 0.0704534723790364,
+      "grad_norm": 2.412420279962346,
+      "learning_rate": 8.999593346451292e-06,
+      "loss": 0.8317,
+      "step": 429
+    },
+    {
+      "epoch": 0.07061769958737914,
+      "grad_norm": 2.3821068613707364,
+      "learning_rate": 8.999587017405074e-06,
+      "loss": 0.8496,
+      "step": 430
+    },
+    {
+      "epoch": 0.07078192679572189,
+      "grad_norm": 2.4590557883994086,
+      "learning_rate": 8.999580639488916e-06,
+      "loss": 0.8142,
+      "step": 431
+    },
+    {
+      "epoch": 0.07094615400406462,
+      "grad_norm": 2.4087171883288763,
+      "learning_rate": 8.999574212702883e-06,
+      "loss": 0.8401,
+      "step": 432
+    },
+    {
+      "epoch": 0.07111038121240737,
+      "grad_norm": 2.375466734189583,
+      "learning_rate": 8.99956773704705e-06,
+      "loss": 0.8161,
+      "step": 433
+    },
+    {
+      "epoch": 0.0712746084207501,
+      "grad_norm": 2.377908447736706,
+      "learning_rate": 8.999561212521482e-06,
+      "loss": 0.8236,
+      "step": 434
+    },
+    {
+      "epoch": 0.07143883562909285,
+      "grad_norm": 2.3169519503998264,
+      "learning_rate": 8.999554639126252e-06,
+      "loss": 0.8426,
+      "step": 435
+    },
+    {
+      "epoch": 0.07160306283743559,
+      "grad_norm": 2.3362968144154816,
+      "learning_rate": 8.999548016861435e-06,
+      "loss": 0.8359,
+      "step": 436
+    },
+    {
+      "epoch": 0.07176729004577834,
+      "grad_norm": 2.386407267700876,
+      "learning_rate": 8.999541345727095e-06,
+      "loss": 0.8029,
+      "step": 437
+    },
+    {
+      "epoch": 0.07193151725412107,
+      "grad_norm": 2.3200737690092303,
+      "learning_rate": 8.999534625723312e-06,
+      "loss": 0.8193,
+      "step": 438
+    },
+    {
+      "epoch": 0.07209574446246382,
+      "grad_norm": 2.340620712080502,
+      "learning_rate": 8.999527856850155e-06,
+      "loss": 0.8179,
+      "step": 439
+    },
+    {
+      "epoch": 0.07225997167080656,
+      "grad_norm": 2.3005341648037216,
+      "learning_rate": 8.999521039107699e-06,
+      "loss": 0.8352,
+      "step": 440
+    },
+    {
+      "epoch": 0.0724241988791493,
+      "grad_norm": 2.268310807131279,
+      "learning_rate": 8.999514172496018e-06,
+      "loss": 0.8584,
+      "step": 441
+    },
+    {
+      "epoch": 0.07258842608749204,
+      "grad_norm": 2.280231628453319,
+      "learning_rate": 8.999507257015185e-06,
+      "loss": 0.8265,
+      "step": 442
+    },
+    {
+      "epoch": 0.07275265329583479,
+      "grad_norm": 2.2563491434376366,
+      "learning_rate": 8.999500292665276e-06,
+      "loss": 0.8438,
+      "step": 443
+    },
+    {
+      "epoch": 0.07291688050417754,
+      "grad_norm": 2.2687198666635093,
+      "learning_rate": 8.999493279446368e-06,
+      "loss": 0.8342,
+      "step": 444
+    },
+    {
+      "epoch": 0.07308110771252027,
+      "grad_norm": 2.230140831740877,
+      "learning_rate": 8.999486217358537e-06,
+      "loss": 0.8238,
+      "step": 445
+    },
+    {
+      "epoch": 0.07324533492086302,
+      "grad_norm": 2.210105023067696,
+      "learning_rate": 8.999479106401858e-06,
+      "loss": 0.8281,
+      "step": 446
+    },
+    {
+      "epoch": 0.07340956212920575,
+      "grad_norm": 2.22721070539442,
+      "learning_rate": 8.999471946576406e-06,
+      "loss": 0.8255,
+      "step": 447
+    },
+    {
+      "epoch": 0.0735737893375485,
+      "grad_norm": 2.1865500694065423,
+      "learning_rate": 8.999464737882265e-06,
+      "loss": 0.8383,
+      "step": 448
+    },
+    {
+      "epoch": 0.07373801654589124,
+      "grad_norm": 2.2143774580363145,
+      "learning_rate": 8.999457480319511e-06,
+      "loss": 0.8283,
+      "step": 449
+    },
+    {
+      "epoch": 0.07390224375423399,
+      "grad_norm": 2.1718648151396462,
+      "learning_rate": 8.99945017388822e-06,
+      "loss": 0.8467,
+      "step": 450
+    },
+    {
+      "epoch": 0.07406647096257672,
+      "grad_norm": 2.206728470233497,
+      "learning_rate": 8.999442818588473e-06,
+      "loss": 0.8038,
+      "step": 451
+    },
+    {
+      "epoch": 0.07423069817091947,
+      "grad_norm": 2.1780216266278853,
+      "learning_rate": 8.99943541442035e-06,
+      "loss": 0.8327,
+      "step": 452
+    },
+    {
+      "epoch": 0.0743949253792622,
+      "grad_norm": 2.1845803500643184,
+      "learning_rate": 8.999427961383933e-06,
+      "loss": 0.7931,
+      "step": 453
+    },
+    {
+      "epoch": 0.07455915258760495,
+      "grad_norm": 2.13549126985634,
+      "learning_rate": 8.9994204594793e-06,
+      "loss": 0.8301,
+      "step": 454
+    },
+    {
+      "epoch": 0.07472337979594769,
+      "grad_norm": 2.1297114580768404,
+      "learning_rate": 8.999412908706536e-06,
+      "loss": 0.8271,
+      "step": 455
+    },
+    {
+      "epoch": 0.07488760700429044,
+      "grad_norm": 2.1538771659712497,
+      "learning_rate": 8.99940530906572e-06,
+      "loss": 0.8335,
+      "step": 456
+    },
+    {
+      "epoch": 0.07505183421263317,
+      "grad_norm": 2.15071828742556,
+      "learning_rate": 8.999397660556935e-06,
+      "loss": 0.8057,
+      "step": 457
+    },
+    {
+      "epoch": 0.07521606142097592,
+      "grad_norm": 2.093120448722487,
+      "learning_rate": 8.999389963180265e-06,
+      "loss": 0.8367,
+      "step": 458
+    },
+    {
+      "epoch": 0.07538028862931866,
+      "grad_norm": 2.085757684625511,
+      "learning_rate": 8.999382216935793e-06,
+      "loss": 0.8278,
+      "step": 459
+    },
+    {
+      "epoch": 0.0755445158376614,
+      "grad_norm": 2.102360562017555,
+      "learning_rate": 8.999374421823603e-06,
+      "loss": 0.8206,
+      "step": 460
+    },
+    {
+      "epoch": 0.07570874304600414,
+      "grad_norm": 2.061123490022161,
+      "learning_rate": 8.99936657784378e-06,
+      "loss": 0.8178,
+      "step": 461
+    },
+    {
+      "epoch": 0.07587297025434689,
+      "grad_norm": 2.040202576510708,
+      "learning_rate": 8.99935868499641e-06,
+      "loss": 0.8234,
+      "step": 462
+    },
+    {
+      "epoch": 0.07603719746268964,
+      "grad_norm": 2.0925459619084807,
+      "learning_rate": 8.999350743281578e-06,
+      "loss": 0.8011,
+      "step": 463
+    },
+    {
+      "epoch": 0.07620142467103237,
+      "grad_norm": 2.0618958956130506,
+      "learning_rate": 8.999342752699368e-06,
+      "loss": 0.8064,
+      "step": 464
+    },
+    {
+      "epoch": 0.07636565187937512,
+      "grad_norm": 2.0446614226644275,
+      "learning_rate": 8.999334713249872e-06,
+      "loss": 0.8143,
+      "step": 465
+    },
+    {
+      "epoch": 0.07652987908771786,
+      "grad_norm": 2.009335333061511,
+      "learning_rate": 8.999326624933172e-06,
+      "loss": 0.8326,
+      "step": 466
+    },
+    {
+      "epoch": 0.0766941062960606,
+      "grad_norm": 2.0489481976521797,
+      "learning_rate": 8.999318487749358e-06,
+      "loss": 0.7827,
+      "step": 467
+    },
+    {
+      "epoch": 0.07685833350440334,
+      "grad_norm": 2.0221916630599304,
+      "learning_rate": 8.99931030169852e-06,
+      "loss": 0.804,
+      "step": 468
+    },
+    {
+      "epoch": 0.07702256071274609,
+      "grad_norm": 2.029438896224542,
+      "learning_rate": 8.999302066780746e-06,
+      "loss": 0.8087,
+      "step": 469
+    },
+    {
+      "epoch": 0.07718678792108882,
+      "grad_norm": 1.978319845152218,
+      "learning_rate": 8.999293782996124e-06,
+      "loss": 0.821,
+      "step": 470
+    },
+    {
+      "epoch": 0.07735101512943157,
+      "grad_norm": 1.9918589527926724,
+      "learning_rate": 8.999285450344744e-06,
+      "loss": 0.8201,
+      "step": 471
+    },
+    {
+      "epoch": 0.0775152423377743,
+      "grad_norm": 2.0023287957387836,
+      "learning_rate": 8.999277068826698e-06,
+      "loss": 0.8114,
+      "step": 472
+    },
+    {
+      "epoch": 0.07767946954611706,
+      "grad_norm": 1.9735825072492768,
+      "learning_rate": 8.999268638442074e-06,
+      "loss": 0.81,
+      "step": 473
+    },
+    {
+      "epoch": 0.07784369675445979,
+      "grad_norm": 1.9457430616789873,
+      "learning_rate": 8.99926015919097e-06,
+      "loss": 0.8161,
+      "step": 474
+    },
+    {
+      "epoch": 0.07800792396280254,
+      "grad_norm": 1.9501504831433791,
+      "learning_rate": 8.999251631073472e-06,
+      "loss": 0.8233,
+      "step": 475
+    },
+    {
+      "epoch": 0.07817215117114527,
+      "grad_norm": 1.9587112324999745,
+      "learning_rate": 8.999243054089675e-06,
+      "loss": 0.8289,
+      "step": 476
+    },
+    {
+      "epoch": 0.07833637837948802,
+      "grad_norm": 1.9321723467173395,
+      "learning_rate": 8.999234428239671e-06,
+      "loss": 0.81,
+      "step": 477
+    },
+    {
+      "epoch": 0.07850060558783076,
+      "grad_norm": 1.8961717334570505,
+      "learning_rate": 8.999225753523554e-06,
+      "loss": 0.8286,
+      "step": 478
+    },
+    {
+      "epoch": 0.0786648327961735,
+      "grad_norm": 1.8988900894264487,
+      "learning_rate": 8.99921702994142e-06,
+      "loss": 0.8257,
+      "step": 479
+    },
+    {
+      "epoch": 0.07882906000451625,
+      "grad_norm": 1.917437492909705,
+      "learning_rate": 8.999208257493363e-06,
+      "loss": 0.796,
+      "step": 480
+    },
+    {
+      "epoch": 0.07899328721285899,
+      "grad_norm": 1.9162103223989222,
+      "learning_rate": 8.999199436179476e-06,
+      "loss": 0.8117,
+      "step": 481
+    },
+    {
+      "epoch": 0.07915751442120174,
+      "grad_norm": 1.8900698273277583,
+      "learning_rate": 8.999190565999858e-06,
+      "loss": 0.8188,
+      "step": 482
+    },
+    {
+      "epoch": 0.07932174162954447,
+      "grad_norm": 1.8766604711654709,
+      "learning_rate": 8.999181646954602e-06,
+      "loss": 0.8152,
+      "step": 483
+    },
+    {
+      "epoch": 0.07948596883788722,
+      "grad_norm": 1.8868541925904996,
+      "learning_rate": 8.999172679043808e-06,
+      "loss": 0.812,
+      "step": 484
+    },
+    {
+      "epoch": 0.07965019604622996,
+      "grad_norm": 1.8559987483245017,
+      "learning_rate": 8.999163662267572e-06,
+      "loss": 0.8223,
+      "step": 485
+    },
+    {
+      "epoch": 0.0798144232545727,
+      "grad_norm": 1.8423384119854995,
+      "learning_rate": 8.999154596625992e-06,
+      "loss": 0.8195,
+      "step": 486
+    },
+    {
+      "epoch": 0.07997865046291544,
+      "grad_norm": 1.8755155601469158,
+      "learning_rate": 8.999145482119168e-06,
+      "loss": 0.789,
+      "step": 487
+    },
+    {
+      "epoch": 0.08014287767125819,
+      "grad_norm": 1.8578743443690677,
+      "learning_rate": 8.999136318747195e-06,
+      "loss": 0.817,
+      "step": 488
+    },
+    {
+      "epoch": 0.08030710487960092,
+      "grad_norm": 1.8149426139291804,
+      "learning_rate": 8.999127106510175e-06,
+      "loss": 0.8298,
+      "step": 489
+    },
+    {
+      "epoch": 0.08047133208794367,
+      "grad_norm": 1.8071557080745544,
+      "learning_rate": 8.999117845408209e-06,
+      "loss": 0.8156,
+      "step": 490
+    },
+    {
+      "epoch": 0.08063555929628641,
+      "grad_norm": 1.7903215084168274,
+      "learning_rate": 8.999108535441398e-06,
+      "loss": 0.8213,
+      "step": 491
+    },
+    {
+      "epoch": 0.08079978650462916,
+      "grad_norm": 1.8039616892674522,
+      "learning_rate": 8.99909917660984e-06,
+      "loss": 0.8086,
+      "step": 492
+    },
+    {
+      "epoch": 0.08096401371297189,
+      "grad_norm": 1.8228172637206808,
+      "learning_rate": 8.99908976891364e-06,
+      "loss": 0.806,
+      "step": 493
+    },
+    {
+      "epoch": 0.08112824092131464,
+      "grad_norm": 1.808137095151838,
+      "learning_rate": 8.999080312352896e-06,
+      "loss": 0.7975,
+      "step": 494
+    },
+    {
+      "epoch": 0.08129246812965737,
+      "grad_norm": 1.7910413085027803,
+      "learning_rate": 8.999070806927716e-06,
+      "loss": 0.7961,
+      "step": 495
+    },
+    {
+      "epoch": 0.08145669533800012,
+      "grad_norm": 1.8391374609956732,
+      "learning_rate": 8.9990612526382e-06,
+      "loss": 0.7667,
+      "step": 496
+    },
+    {
+      "epoch": 0.08162092254634287,
+      "grad_norm": 1.7825821027741344,
+      "learning_rate": 8.999051649484451e-06,
+      "loss": 0.8082,
+      "step": 497
+    },
+    {
+      "epoch": 0.08178514975468561,
+      "grad_norm": 1.7535944210904948,
+      "learning_rate": 8.999041997466575e-06,
+      "loss": 0.806,
+      "step": 498
+    },
+    {
+      "epoch": 0.08194937696302836,
+      "grad_norm": 1.7940946696819353,
+      "learning_rate": 8.999032296584675e-06,
+      "loss": 0.7794,
+      "step": 499
+    },
+    {
+      "epoch": 0.08211360417137109,
+      "grad_norm": 1.7675061166108443,
+      "learning_rate": 8.999022546838862e-06,
+      "loss": 0.8007,
+      "step": 500
+    },
+    {
+      "epoch": 0.08227783137971384,
+      "grad_norm": 1.7466093753207794,
+      "learning_rate": 8.999012748229235e-06,
+      "loss": 0.8085,
+      "step": 501
+    },
+    {
+      "epoch": 0.08244205858805657,
+      "grad_norm": 1.7384682546038235,
+      "learning_rate": 8.999002900755904e-06,
+      "loss": 0.8029,
+      "step": 502
+    },
+    {
+      "epoch": 0.08260628579639932,
+      "grad_norm": 1.749204038070269,
+      "learning_rate": 8.998993004418973e-06,
+      "loss": 0.8065,
+      "step": 503
+    },
+    {
+      "epoch": 0.08277051300474206,
+      "grad_norm": 1.731827509189124,
+      "learning_rate": 8.998983059218553e-06,
+      "loss": 0.7904,
+      "step": 504
+    },
+    {
+      "epoch": 0.0829347402130848,
+      "grad_norm": 1.713406755440384,
+      "learning_rate": 8.998973065154751e-06,
+      "loss": 0.7929,
+      "step": 505
+    },
+    {
+      "epoch": 0.08309896742142754,
+      "grad_norm": 1.7370353467699013,
+      "learning_rate": 8.998963022227676e-06,
+      "loss": 0.7979,
+      "step": 506
+    },
+    {
+      "epoch": 0.08326319462977029,
+      "grad_norm": 1.722144711965804,
+      "learning_rate": 8.998952930437434e-06,
+      "loss": 0.7864,
+      "step": 507
+    },
+    {
+      "epoch": 0.08342742183811303,
+      "grad_norm": 1.6850288461926786,
+      "learning_rate": 8.998942789784138e-06,
+      "loss": 0.7966,
+      "step": 508
+    },
+    {
+      "epoch": 0.08359164904645577,
+      "grad_norm": 1.7125487489335667,
+      "learning_rate": 8.998932600267896e-06,
+      "loss": 0.789,
+      "step": 509
+    },
+    {
+      "epoch": 0.08375587625479851,
+      "grad_norm": 1.666553952443347,
+      "learning_rate": 8.998922361888821e-06,
+      "loss": 0.8144,
+      "step": 510
+    },
+    {
+      "epoch": 0.08392010346314126,
+      "grad_norm": 1.6766351515133946,
+      "learning_rate": 8.998912074647022e-06,
+      "loss": 0.8144,
+      "step": 511
+    },
+    {
+      "epoch": 0.08408433067148399,
+      "grad_norm": 1.6676136142415157,
+      "learning_rate": 8.998901738542612e-06,
+      "loss": 0.8086,
+      "step": 512
+    },
+    {
+      "epoch": 0.08424855787982674,
+      "grad_norm": 1.6939153245378111,
+      "learning_rate": 8.998891353575703e-06,
+      "loss": 0.7887,
+      "step": 513
+    },
+    {
+      "epoch": 0.08441278508816948,
+      "grad_norm": 1.6367960342037091,
+      "learning_rate": 8.998880919746406e-06,
+      "loss": 0.8217,
+      "step": 514
+    },
+    {
+      "epoch": 0.08457701229651222,
+      "grad_norm": 1.656541292788681,
+      "learning_rate": 8.998870437054837e-06,
+      "loss": 0.7918,
+      "step": 515
+    },
+    {
+      "epoch": 0.08474123950485497,
+      "grad_norm": 1.649968215675082,
+      "learning_rate": 8.99885990550111e-06,
+      "loss": 0.7948,
+      "step": 516
+    },
+    {
+      "epoch": 0.08490546671319771,
+      "grad_norm": 1.637068861746453,
+      "learning_rate": 8.998849325085336e-06,
+      "loss": 0.8022,
+      "step": 517
+    },
+    {
+      "epoch": 0.08506969392154046,
+      "grad_norm": 1.6598278261447834,
+      "learning_rate": 8.998838695807632e-06,
+      "loss": 0.7767,
+      "step": 518
+    },
+    {
+      "epoch": 0.08523392112988319,
+      "grad_norm": 1.6497075059531359,
+      "learning_rate": 8.998828017668115e-06,
+      "loss": 0.7843,
+      "step": 519
+    },
+    {
+      "epoch": 0.08539814833822594,
+      "grad_norm": 1.6048685678000418,
+      "learning_rate": 8.9988172906669e-06,
+      "loss": 0.8041,
+      "step": 520
+    },
+    {
+      "epoch": 0.08556237554656868,
+      "grad_norm": 1.6169953267344268,
+      "learning_rate": 8.998806514804102e-06,
+      "loss": 0.7884,
+      "step": 521
+    },
+    {
+      "epoch": 0.08572660275491142,
+      "grad_norm": 1.6177878656916025,
+      "learning_rate": 8.998795690079838e-06,
+      "loss": 0.7828,
+      "step": 522
+    },
+    {
+      "epoch": 0.08589082996325416,
+      "grad_norm": 1.605566301566824,
+      "learning_rate": 8.998784816494227e-06,
+      "loss": 0.7877,
+      "step": 523
+    },
+    {
+      "epoch": 0.08605505717159691,
+      "grad_norm": 1.5887379483469,
+      "learning_rate": 8.99877389404739e-06,
+      "loss": 0.8035,
+      "step": 524
+    },
+    {
+      "epoch": 0.08621928437993964,
+      "grad_norm": 1.605034282881469,
+      "learning_rate": 8.998762922739438e-06,
+      "loss": 0.793,
+      "step": 525
+    },
+    {
+      "epoch": 0.08638351158828239,
+      "grad_norm": 1.5765227198544196,
+      "learning_rate": 8.998751902570496e-06,
+      "loss": 0.796,
+      "step": 526
+    },
+    {
+      "epoch": 0.08654773879662513,
+      "grad_norm": 1.5863705667950556,
+      "learning_rate": 8.998740833540683e-06,
+      "loss": 0.7796,
+      "step": 527
+    },
+    {
+      "epoch": 0.08671196600496787,
+      "grad_norm": 1.5650967154526163,
+      "learning_rate": 8.998729715650116e-06,
+      "loss": 0.7964,
+      "step": 528
+    },
+    {
+      "epoch": 0.08687619321331061,
+      "grad_norm": 1.5973499387442838,
+      "learning_rate": 8.99871854889892e-06,
+      "loss": 0.7778,
+      "step": 529
+    },
+    {
+      "epoch": 0.08704042042165336,
+      "grad_norm": 1.5399366650437916,
+      "learning_rate": 8.998707333287214e-06,
+      "loss": 0.8062,
+      "step": 530
+    },
+    {
+      "epoch": 0.0872046476299961,
+      "grad_norm": 1.5877605974589395,
+      "learning_rate": 8.99869606881512e-06,
+      "loss": 0.7861,
+      "step": 531
+    },
+    {
+      "epoch": 0.08736887483833884,
+      "grad_norm": 1.5404243830810433,
+      "learning_rate": 8.998684755482762e-06,
+      "loss": 0.7904,
+      "step": 532
+    },
+    {
+      "epoch": 0.08753310204668159,
+      "grad_norm": 1.5293995565013145,
+      "learning_rate": 8.99867339329026e-06,
+      "loss": 0.8052,
+      "step": 533
+    },
+    {
+      "epoch": 0.08769732925502433,
+      "grad_norm": 1.5285645822160316,
+      "learning_rate": 8.99866198223774e-06,
+      "loss": 0.8057,
+      "step": 534
+    },
+    {
+      "epoch": 0.08786155646336707,
+      "grad_norm": 1.5317827095316812,
+      "learning_rate": 8.998650522325322e-06,
+      "loss": 0.7905,
+      "step": 535
+    },
+    {
+      "epoch": 0.08802578367170981,
+      "grad_norm": 1.5487961117886297,
+      "learning_rate": 8.998639013553136e-06,
+      "loss": 0.7814,
+      "step": 536
+    },
+    {
+      "epoch": 0.08819001088005256,
+      "grad_norm": 1.512615812975723,
+      "learning_rate": 8.998627455921304e-06,
+      "loss": 0.794,
+      "step": 537
+    },
+    {
+      "epoch": 0.08835423808839529,
+      "grad_norm": 1.524531472522993,
+      "learning_rate": 8.998615849429952e-06,
+      "loss": 0.7853,
+      "step": 538
+    },
+    {
+      "epoch": 0.08851846529673804,
+      "grad_norm": 1.5113713334392895,
+      "learning_rate": 8.998604194079204e-06,
+      "loss": 0.7985,
+      "step": 539
+    },
+    {
+      "epoch": 0.08868269250508078,
+      "grad_norm": 1.513587345447187,
+      "learning_rate": 8.99859248986919e-06,
+      "loss": 0.7818,
+      "step": 540
+    },
+    {
+      "epoch": 0.08884691971342353,
+      "grad_norm": 1.5025263128659543,
+      "learning_rate": 8.998580736800035e-06,
+      "loss": 0.7828,
+      "step": 541
+    },
+    {
+      "epoch": 0.08901114692176626,
+      "grad_norm": 1.50559719285815,
+      "learning_rate": 8.998568934871868e-06,
+      "loss": 0.7903,
+      "step": 542
+    },
+    {
+      "epoch": 0.08917537413010901,
+      "grad_norm": 1.4894498383121608,
+      "learning_rate": 8.998557084084815e-06,
+      "loss": 0.7731,
+      "step": 543
+    },
+    {
+      "epoch": 0.08933960133845174,
+      "grad_norm": 1.4952299462037122,
+      "learning_rate": 8.998545184439007e-06,
+      "loss": 0.7702,
+      "step": 544
+    },
+    {
+      "epoch": 0.08950382854679449,
+      "grad_norm": 1.4653177315472807,
+      "learning_rate": 8.99853323593457e-06,
+      "loss": 0.7988,
+      "step": 545
+    },
+    {
+      "epoch": 0.08966805575513723,
+      "grad_norm": 1.459127859543925,
+      "learning_rate": 8.99852123857164e-06,
+      "loss": 0.8138,
+      "step": 546
+    },
+    {
+      "epoch": 0.08983228296347998,
+      "grad_norm": 1.4651643994973669,
+      "learning_rate": 8.998509192350342e-06,
+      "loss": 0.7876,
+      "step": 547
+    },
+    {
+      "epoch": 0.08999651017182271,
+      "grad_norm": 1.4676671777660628,
+      "learning_rate": 8.998497097270808e-06,
+      "loss": 0.7933,
+      "step": 548
+    },
+    {
+      "epoch": 0.09016073738016546,
+      "grad_norm": 1.4376367008426452,
+      "learning_rate": 8.998484953333168e-06,
+      "loss": 0.8033,
+      "step": 549
+    },
+    {
+      "epoch": 0.0903249645885082,
+      "grad_norm": 1.4627566923252446,
+      "learning_rate": 8.998472760537557e-06,
+      "loss": 0.7791,
+      "step": 550
+    },
+    {
+      "epoch": 0.09048919179685094,
+      "grad_norm": 1.4303444060969528,
+      "learning_rate": 8.998460518884106e-06,
+      "loss": 0.7941,
+      "step": 551
+    },
+    {
+      "epoch": 0.09065341900519369,
+      "grad_norm": 1.520429285468019,
+      "learning_rate": 8.998448228372947e-06,
+      "loss": 0.7875,
+      "step": 552
+    },
+    {
+      "epoch": 0.09081764621353643,
+      "grad_norm": 1.4467982423594512,
+      "learning_rate": 8.998435889004217e-06,
+      "loss": 0.7599,
+      "step": 553
+    },
+    {
+      "epoch": 0.09098187342187918,
+      "grad_norm": 1.4542459927929825,
+      "learning_rate": 8.998423500778045e-06,
+      "loss": 0.76,
+      "step": 554
+    },
+    {
+      "epoch": 0.09114610063022191,
+      "grad_norm": 1.4334949641597166,
+      "learning_rate": 8.998411063694569e-06,
+      "loss": 0.7816,
+      "step": 555
+    },
+    {
+      "epoch": 0.09131032783856466,
+      "grad_norm": 1.4376273667298758,
+      "learning_rate": 8.99839857775392e-06,
+      "loss": 0.78,
+      "step": 556
+    },
+    {
+      "epoch": 0.0914745550469074,
+      "grad_norm": 1.4244926116912235,
+      "learning_rate": 8.99838604295624e-06,
+      "loss": 0.7887,
+      "step": 557
+    },
+    {
+      "epoch": 0.09163878225525014,
+      "grad_norm": 1.4180874831462564,
+      "learning_rate": 8.99837345930166e-06,
+      "loss": 0.7668,
+      "step": 558
+    },
+    {
+      "epoch": 0.09180300946359288,
+      "grad_norm": 1.402879410935121,
+      "learning_rate": 8.998360826790319e-06,
+      "loss": 0.7842,
+      "step": 559
+    },
+    {
+      "epoch": 0.09196723667193563,
+      "grad_norm": 1.4040836140296928,
+      "learning_rate": 8.998348145422355e-06,
+      "loss": 0.7859,
+      "step": 560
+    },
+    {
+      "epoch": 0.09213146388027836,
+      "grad_norm": 1.4219797527782458,
+      "learning_rate": 8.998335415197903e-06,
+      "loss": 0.7706,
+      "step": 561
+    },
+    {
+      "epoch": 0.09229569108862111,
+      "grad_norm": 1.4503951930362755,
+      "learning_rate": 8.998322636117103e-06,
+      "loss": 0.7518,
+      "step": 562
+    },
+    {
+      "epoch": 0.09245991829696384,
+      "grad_norm": 1.4170028168044975,
+      "learning_rate": 8.998309808180093e-06,
+      "loss": 0.7806,
+      "step": 563
+    },
+    {
+      "epoch": 0.0926241455053066,
+      "grad_norm": 1.3835900975245248,
+      "learning_rate": 8.998296931387013e-06,
+      "loss": 0.7776,
+      "step": 564
+    },
+    {
+      "epoch": 0.09278837271364933,
+      "grad_norm": 1.3880650749493229,
+      "learning_rate": 8.998284005738002e-06,
+      "loss": 0.7716,
+      "step": 565
+    },
+    {
+      "epoch": 0.09295259992199208,
+      "grad_norm": 1.3904515187341628,
+      "learning_rate": 8.998271031233202e-06,
+      "loss": 0.7868,
+      "step": 566
+    },
+    {
+      "epoch": 0.09311682713033481,
+      "grad_norm": 1.3873774301491548,
+      "learning_rate": 8.998258007872753e-06,
+      "loss": 0.7768,
+      "step": 567
+    },
+    {
+      "epoch": 0.09328105433867756,
+      "grad_norm": 1.370753815387451,
+      "learning_rate": 8.998244935656797e-06,
+      "loss": 0.7658,
+      "step": 568
+    },
+    {
+      "epoch": 0.09344528154702031,
+      "grad_norm": 1.4026344155479453,
+      "learning_rate": 8.998231814585473e-06,
+      "loss": 0.7526,
+      "step": 569
+    },
+    {
+      "epoch": 0.09360950875536304,
+      "grad_norm": 1.344243720958496,
+      "learning_rate": 8.998218644658926e-06,
+      "loss": 0.7794,
+      "step": 570
+    },
+    {
+      "epoch": 0.09377373596370579,
+      "grad_norm": 1.3829097281740133,
+      "learning_rate": 8.9982054258773e-06,
+      "loss": 0.7875,
+      "step": 571
+    },
+    {
+      "epoch": 0.09393796317204853,
+      "grad_norm": 1.3604310447903412,
+      "learning_rate": 8.998192158240738e-06,
+      "loss": 0.7869,
+      "step": 572
+    },
+    {
+      "epoch": 0.09410219038039128,
+      "grad_norm": 1.3581920491158632,
+      "learning_rate": 8.998178841749382e-06,
+      "loss": 0.7821,
+      "step": 573
+    },
+    {
+      "epoch": 0.09426641758873401,
+      "grad_norm": 1.3407414437715726,
+      "learning_rate": 8.998165476403378e-06,
+      "loss": 0.78,
+      "step": 574
+    },
+    {
+      "epoch": 0.09443064479707676,
+      "grad_norm": 1.323211309690318,
+      "learning_rate": 8.998152062202874e-06,
+      "loss": 0.7974,
+      "step": 575
+    },
+    {
+      "epoch": 0.0945948720054195,
+      "grad_norm": 1.3516519692522833,
+      "learning_rate": 8.998138599148012e-06,
+      "loss": 0.7649,
+      "step": 576
+    },
+    {
+      "epoch": 0.09475909921376224,
+      "grad_norm": 1.3229131845618651,
+      "learning_rate": 8.998125087238938e-06,
+      "loss": 0.7962,
+      "step": 577
+    },
+    {
+      "epoch": 0.09492332642210498,
+      "grad_norm": 1.307882279240173,
+      "learning_rate": 8.9981115264758e-06,
+      "loss": 0.806,
+      "step": 578
+    },
+    {
+      "epoch": 0.09508755363044773,
+      "grad_norm": 1.3352279874733834,
+      "learning_rate": 8.998097916858747e-06,
+      "loss": 0.774,
+      "step": 579
+    },
+    {
+      "epoch": 0.09525178083879046,
+      "grad_norm": 1.311690507687563,
+      "learning_rate": 8.998084258387923e-06,
+      "loss": 0.7928,
+      "step": 580
+    },
+    {
+      "epoch": 0.09541600804713321,
+      "grad_norm": 1.316169384327891,
+      "learning_rate": 8.99807055106348e-06,
+      "loss": 0.8016,
+      "step": 581
+    },
+    {
+      "epoch": 0.09558023525547595,
+      "grad_norm": 1.3115286012581258,
+      "learning_rate": 8.998056794885563e-06,
+      "loss": 0.7687,
+      "step": 582
+    },
+    {
+      "epoch": 0.0957444624638187,
+      "grad_norm": 1.2973876639193667,
+      "learning_rate": 8.998042989854326e-06,
+      "loss": 0.7957,
+      "step": 583
+    },
+    {
+      "epoch": 0.09590868967216143,
+      "grad_norm": 1.3034852678205227,
+      "learning_rate": 8.998029135969917e-06,
+      "loss": 0.7694,
+      "step": 584
+    },
+    {
+      "epoch": 0.09607291688050418,
+      "grad_norm": 1.2921967384130368,
+      "learning_rate": 8.998015233232484e-06,
+      "loss": 0.7673,
+      "step": 585
+    },
+    {
+      "epoch": 0.09623714408884693,
+      "grad_norm": 1.2906042190034313,
+      "learning_rate": 8.998001281642182e-06,
+      "loss": 0.7768,
+      "step": 586
+    },
+    {
+      "epoch": 0.09640137129718966,
+      "grad_norm": 1.2940931141309069,
+      "learning_rate": 8.99798728119916e-06,
+      "loss": 0.7844,
+      "step": 587
+    },
+    {
+      "epoch": 0.09656559850553241,
+      "grad_norm": 1.284649039797445,
+      "learning_rate": 8.99797323190357e-06,
+      "loss": 0.7789,
+      "step": 588
+    },
+    {
+      "epoch": 0.09672982571387514,
+      "grad_norm": 1.283737139541954,
+      "learning_rate": 8.997959133755566e-06,
+      "loss": 0.7677,
+      "step": 589
+    },
+    {
+      "epoch": 0.0968940529222179,
+      "grad_norm": 1.2736857154419794,
+      "learning_rate": 8.997944986755302e-06,
+      "loss": 0.778,
+      "step": 590
+    },
+    {
+      "epoch": 0.09705828013056063,
+      "grad_norm": 1.2769757024108,
+      "learning_rate": 8.997930790902928e-06,
+      "loss": 0.7819,
+      "step": 591
+    },
+    {
+      "epoch": 0.09722250733890338,
+      "grad_norm": 1.262465276945493,
+      "learning_rate": 8.997916546198599e-06,
+      "loss": 0.7783,
+      "step": 592
+    },
+    {
+      "epoch": 0.09738673454724611,
+      "grad_norm": 1.2663433484817244,
+      "learning_rate": 8.997902252642474e-06,
+      "loss": 0.7661,
+      "step": 593
+    },
+    {
+      "epoch": 0.09755096175558886,
+      "grad_norm": 1.2748735844470442,
+      "learning_rate": 8.997887910234704e-06,
+      "loss": 0.7871,
+      "step": 594
+    },
+    {
+      "epoch": 0.0977151889639316,
+      "grad_norm": 1.2627194359703184,
+      "learning_rate": 8.997873518975445e-06,
+      "loss": 0.7697,
+      "step": 595
+    },
+    {
+      "epoch": 0.09787941617227434,
+      "grad_norm": 1.2458601207727036,
+      "learning_rate": 8.997859078864856e-06,
+      "loss": 0.7761,
+      "step": 596
+    },
+    {
+      "epoch": 0.09804364338061708,
+      "grad_norm": 1.2685292083003645,
+      "learning_rate": 8.99784458990309e-06,
+      "loss": 0.7595,
+      "step": 597
+    },
+    {
+      "epoch": 0.09820787058895983,
+      "grad_norm": 1.2423763258616558,
+      "learning_rate": 8.997830052090308e-06,
+      "loss": 0.7569,
+      "step": 598
+    },
+    {
+      "epoch": 0.09837209779730256,
+      "grad_norm": 1.2284843126416907,
+      "learning_rate": 8.997815465426666e-06,
+      "loss": 0.7691,
+      "step": 599
+    },
+    {
+      "epoch": 0.09853632500564531,
+      "grad_norm": 1.2380214027521046,
+      "learning_rate": 8.997800829912323e-06,
+      "loss": 0.7736,
+      "step": 600
+    },
+    {
+      "epoch": 0.09870055221398805,
+      "grad_norm": 1.2319079389161247,
+      "learning_rate": 8.997786145547437e-06,
+      "loss": 0.7714,
+      "step": 601
+    },
+    {
+      "epoch": 0.0988647794223308,
+      "grad_norm": 1.2327228774043395,
+      "learning_rate": 8.997771412332169e-06,
+      "loss": 0.7796,
+      "step": 602
+    },
+    {
+      "epoch": 0.09902900663067353,
+      "grad_norm": 1.2188386522371906,
+      "learning_rate": 8.997756630266679e-06,
+      "loss": 0.7831,
+      "step": 603
+    },
+    {
+      "epoch": 0.09919323383901628,
+      "grad_norm": 1.198537872248901,
+      "learning_rate": 8.997741799351126e-06,
+      "loss": 0.7919,
+      "step": 604
+    },
+    {
+      "epoch": 0.09935746104735903,
+      "grad_norm": 1.2216480771422678,
+      "learning_rate": 8.997726919585671e-06,
+      "loss": 0.7551,
+      "step": 605
+    },
+    {
+      "epoch": 0.09952168825570176,
+      "grad_norm": 1.225034221391949,
+      "learning_rate": 8.997711990970478e-06,
+      "loss": 0.7663,
+      "step": 606
+    },
+    {
+      "epoch": 0.09968591546404451,
+      "grad_norm": 1.2000663875670825,
+      "learning_rate": 8.997697013505707e-06,
+      "loss": 0.7677,
+      "step": 607
+    },
+    {
+      "epoch": 0.09985014267238725,
+      "grad_norm": 1.2053600381758947,
+      "learning_rate": 8.997681987191521e-06,
+      "loss": 0.7685,
+      "step": 608
+    },
+    {
+      "epoch": 0.10001436988073,
+      "grad_norm": 1.2058066147722575,
+      "learning_rate": 8.997666912028083e-06,
+      "loss": 0.7591,
+      "step": 609
+    },
+    {
+      "epoch": 0.10017859708907273,
+      "grad_norm": 1.201513478456253,
+      "learning_rate": 8.997651788015559e-06,
+      "loss": 0.7716,
+      "step": 610
+    },
+    {
+      "epoch": 0.10034282429741548,
+      "grad_norm": 1.2182501923516547,
+      "learning_rate": 8.997636615154109e-06,
+      "loss": 0.741,
+      "step": 611
+    },
+    {
+      "epoch": 0.10050705150575821,
+      "grad_norm": 1.1980299500822917,
+      "learning_rate": 8.997621393443901e-06,
+      "loss": 0.759,
+      "step": 612
+    },
+    {
+      "epoch": 0.10067127871410096,
+      "grad_norm": 1.2021512642384815,
+      "learning_rate": 8.997606122885102e-06,
+      "loss": 0.7686,
+      "step": 613
+    },
+    {
+      "epoch": 0.1008355059224437,
+      "grad_norm": 1.1891403994893042,
+      "learning_rate": 8.997590803477872e-06,
+      "loss": 0.7735,
+      "step": 614
+    },
+    {
+      "epoch": 0.10099973313078645,
+      "grad_norm": 1.1847699209247626,
+      "learning_rate": 8.997575435222384e-06,
+      "loss": 0.7709,
+      "step": 615
+    },
+    {
+      "epoch": 0.10116396033912918,
+      "grad_norm": 1.1887640824690933,
+      "learning_rate": 8.9975600181188e-06,
+      "loss": 0.7463,
+      "step": 616
+    },
+    {
+      "epoch": 0.10132818754747193,
+      "grad_norm": 1.1738353175527652,
+      "learning_rate": 8.997544552167289e-06,
+      "loss": 0.7775,
+      "step": 617
+    },
+    {
+      "epoch": 0.10149241475581466,
+      "grad_norm": 1.181133780761317,
+      "learning_rate": 8.99752903736802e-06,
+      "loss": 0.7607,
+      "step": 618
+    },
+    {
+      "epoch": 0.10165664196415741,
+      "grad_norm": 1.1674676030003306,
+      "learning_rate": 8.99751347372116e-06,
+      "loss": 0.7616,
+      "step": 619
+    },
+    {
+      "epoch": 0.10182086917250015,
+      "grad_norm": 1.168583211529951,
+      "learning_rate": 8.99749786122688e-06,
+      "loss": 0.7683,
+      "step": 620
+    },
+    {
+      "epoch": 0.1019850963808429,
+      "grad_norm": 1.1622822735095926,
+      "learning_rate": 8.997482199885346e-06,
+      "loss": 0.768,
+      "step": 621
+    },
+    {
+      "epoch": 0.10214932358918564,
+      "grad_norm": 1.1471729076856725,
+      "learning_rate": 8.997466489696732e-06,
+      "loss": 0.7604,
+      "step": 622
+    },
+    {
+      "epoch": 0.10231355079752838,
+      "grad_norm": 1.145057169808049,
+      "learning_rate": 8.997450730661206e-06,
+      "loss": 0.7578,
+      "step": 623
+    },
+    {
+      "epoch": 0.10247777800587113,
+      "grad_norm": 1.1557515410421166,
+      "learning_rate": 8.997434922778942e-06,
+      "loss": 0.7672,
+      "step": 624
+    },
+    {
+      "epoch": 0.10264200521421386,
+      "grad_norm": 1.1355739695048033,
+      "learning_rate": 8.99741906605011e-06,
+      "loss": 0.7653,
+      "step": 625
+    },
+    {
+      "epoch": 0.10280623242255661,
+      "grad_norm": 1.1538565942915397,
+      "learning_rate": 8.99740316047488e-06,
+      "loss": 0.7394,
+      "step": 626
+    },
+    {
+      "epoch": 0.10297045963089935,
+      "grad_norm": 1.1446329691976465,
+      "learning_rate": 8.997387206053427e-06,
+      "loss": 0.7499,
+      "step": 627
+    },
+    {
+      "epoch": 0.1031346868392421,
+      "grad_norm": 1.1537159517502786,
+      "learning_rate": 8.997371202785925e-06,
+      "loss": 0.7422,
+      "step": 628
+    },
+    {
+      "epoch": 0.10329891404758483,
+      "grad_norm": 1.1205865491670322,
+      "learning_rate": 8.997355150672548e-06,
+      "loss": 0.7799,
+      "step": 629
+    },
+    {
+      "epoch": 0.10346314125592758,
+      "grad_norm": 1.138552041683983,
+      "learning_rate": 8.997339049713468e-06,
+      "loss": 0.7533,
+      "step": 630
+    },
+    {
+      "epoch": 0.10362736846427031,
+      "grad_norm": 1.1208591898037377,
+      "learning_rate": 8.99732289990886e-06,
+      "loss": 0.7685,
+      "step": 631
+    },
+    {
+      "epoch": 0.10379159567261306,
+      "grad_norm": 1.122828390596003,
+      "learning_rate": 8.997306701258903e-06,
+      "loss": 0.7533,
+      "step": 632
+    },
+    {
+      "epoch": 0.1039558228809558,
+      "grad_norm": 1.12283473302057,
+      "learning_rate": 8.997290453763768e-06,
+      "loss": 0.748,
+      "step": 633
+    },
+    {
+      "epoch": 0.10412005008929855,
+      "grad_norm": 1.1034866068092488,
+      "learning_rate": 8.997274157423636e-06,
+      "loss": 0.7512,
+      "step": 634
+    },
+    {
+      "epoch": 0.10428427729764128,
+      "grad_norm": 1.1165031140298034,
+      "learning_rate": 8.99725781223868e-06,
+      "loss": 0.7536,
+      "step": 635
+    },
+    {
+      "epoch": 0.10444850450598403,
+      "grad_norm": 1.12908873229894,
+      "learning_rate": 8.99724141820908e-06,
+      "loss": 0.7437,
+      "step": 636
+    },
+    {
+      "epoch": 0.10461273171432676,
+      "grad_norm": 1.122551438784602,
+      "learning_rate": 8.997224975335015e-06,
+      "loss": 0.7525,
+      "step": 637
+    },
+    {
+      "epoch": 0.10477695892266951,
+      "grad_norm": 1.093230994266884,
+      "learning_rate": 8.99720848361666e-06,
+      "loss": 0.7871,
+      "step": 638
+    },
+    {
+      "epoch": 0.10494118613101226,
+      "grad_norm": 1.1022741460479262,
+      "learning_rate": 8.997191943054198e-06,
+      "loss": 0.768,
+      "step": 639
+    },
+    {
+      "epoch": 0.105105413339355,
+      "grad_norm": 1.0910181528443548,
+      "learning_rate": 8.997175353647806e-06,
+      "loss": 0.7549,
+      "step": 640
+    },
+    {
+      "epoch": 0.10526964054769775,
+      "grad_norm": 1.1062114372442224,
+      "learning_rate": 8.997158715397665e-06,
+      "loss": 0.7539,
+      "step": 641
+    },
+    {
+      "epoch": 0.10543386775604048,
+      "grad_norm": 1.103586593071817,
+      "learning_rate": 8.997142028303955e-06,
+      "loss": 0.7544,
+      "step": 642
+    },
+    {
+      "epoch": 0.10559809496438323,
+      "grad_norm": 1.1071233361706303,
+      "learning_rate": 8.99712529236686e-06,
+      "loss": 0.764,
+      "step": 643
+    },
+    {
+      "epoch": 0.10576232217272596,
+      "grad_norm": 1.094482530155215,
+      "learning_rate": 8.99710850758656e-06,
+      "loss": 0.7468,
+      "step": 644
+    },
+    {
+      "epoch": 0.10592654938106871,
+      "grad_norm": 1.078601443848298,
+      "learning_rate": 8.997091673963234e-06,
+      "loss": 0.7576,
+      "step": 645
+    },
+    {
+      "epoch": 0.10609077658941145,
+      "grad_norm": 1.124326466230655,
+      "learning_rate": 8.99707479149707e-06,
+      "loss": 0.7488,
+      "step": 646
+    },
+    {
+      "epoch": 0.1062550037977542,
+      "grad_norm": 1.1065751826100323,
+      "learning_rate": 8.99705786018825e-06,
+      "loss": 0.7552,
+      "step": 647
+    },
+    {
+      "epoch": 0.10641923100609693,
+      "grad_norm": 1.067142366456579,
+      "learning_rate": 8.997040880036956e-06,
+      "loss": 0.771,
+      "step": 648
+    },
+    {
+      "epoch": 0.10658345821443968,
+      "grad_norm": 1.0713690882822862,
+      "learning_rate": 8.997023851043372e-06,
+      "loss": 0.7527,
+      "step": 649
+    },
+    {
+      "epoch": 0.10674768542278242,
+      "grad_norm": 1.0664071164503175,
+      "learning_rate": 8.997006773207687e-06,
+      "loss": 0.7405,
+      "step": 650
+    },
+    {
+      "epoch": 0.10691191263112516,
+      "grad_norm": 1.0590095194012845,
+      "learning_rate": 8.996989646530083e-06,
+      "loss": 0.7711,
+      "step": 651
+    },
+    {
+      "epoch": 0.1070761398394679,
+      "grad_norm": 1.0157260991175592,
+      "learning_rate": 8.996972471010747e-06,
+      "loss": 0.8044,
+      "step": 652
+    },
+    {
+      "epoch": 0.10724036704781065,
+      "grad_norm": 1.0702230692833752,
+      "learning_rate": 8.996955246649866e-06,
+      "loss": 0.7708,
+      "step": 653
+    },
+    {
+      "epoch": 0.10740459425615338,
+      "grad_norm": 1.0415420000470876,
+      "learning_rate": 8.996937973447627e-06,
+      "loss": 0.7648,
+      "step": 654
+    },
+    {
+      "epoch": 0.10756882146449613,
+      "grad_norm": 1.0502831812393232,
+      "learning_rate": 8.996920651404215e-06,
+      "loss": 0.7514,
+      "step": 655
+    },
+    {
+      "epoch": 0.10773304867283887,
+      "grad_norm": 1.0643751230664344,
+      "learning_rate": 8.996903280519821e-06,
+      "loss": 0.7432,
+      "step": 656
+    },
+    {
+      "epoch": 0.10789727588118161,
+      "grad_norm": 1.0611243032791475,
+      "learning_rate": 8.996885860794635e-06,
+      "loss": 0.75,
+      "step": 657
+    },
+    {
+      "epoch": 0.10806150308952436,
+      "grad_norm": 1.0614565864005756,
+      "learning_rate": 8.996868392228843e-06,
+      "loss": 0.7704,
+      "step": 658
+    },
+    {
+      "epoch": 0.1082257302978671,
+      "grad_norm": 1.0522356645726656,
+      "learning_rate": 8.996850874822636e-06,
+      "loss": 0.7514,
+      "step": 659
+    },
+    {
+      "epoch": 0.10838995750620985,
+      "grad_norm": 1.0396641601794014,
+      "learning_rate": 8.996833308576205e-06,
+      "loss": 0.7443,
+      "step": 660
+    },
+    {
+      "epoch": 0.10855418471455258,
+      "grad_norm": 1.0352100207650416,
+      "learning_rate": 8.996815693489738e-06,
+      "loss": 0.7444,
+      "step": 661
+    },
+    {
+      "epoch": 0.10871841192289533,
+      "grad_norm": 1.0440907953492387,
+      "learning_rate": 8.99679802956343e-06,
+      "loss": 0.7588,
+      "step": 662
+    },
+    {
+      "epoch": 0.10888263913123807,
+      "grad_norm": 1.0237415973036992,
+      "learning_rate": 8.99678031679747e-06,
+      "loss": 0.743,
+      "step": 663
+    },
+    {
+      "epoch": 0.10904686633958081,
+      "grad_norm": 1.038219778138841,
+      "learning_rate": 8.996762555192053e-06,
+      "loss": 0.7696,
+      "step": 664
+    },
+    {
+      "epoch": 0.10921109354792355,
+      "grad_norm": 1.0371317655068988,
+      "learning_rate": 8.99674474474737e-06,
+      "loss": 0.741,
+      "step": 665
+    },
+    {
+      "epoch": 0.1093753207562663,
+      "grad_norm": 1.015958023377164,
+      "learning_rate": 8.996726885463612e-06,
+      "loss": 0.7632,
+      "step": 666
+    },
+    {
+      "epoch": 0.10953954796460903,
+      "grad_norm": 1.025225602279085,
+      "learning_rate": 8.996708977340979e-06,
+      "loss": 0.7559,
+      "step": 667
+    },
+    {
+      "epoch": 0.10970377517295178,
+      "grad_norm": 1.0102874868738583,
+      "learning_rate": 8.996691020379662e-06,
+      "loss": 0.7528,
+      "step": 668
+    },
+    {
+      "epoch": 0.10986800238129452,
+      "grad_norm": 1.0171496292001285,
+      "learning_rate": 8.996673014579855e-06,
+      "loss": 0.7476,
+      "step": 669
+    },
+    {
+      "epoch": 0.11003222958963726,
+      "grad_norm": 1.0133586146958062,
+      "learning_rate": 8.996654959941757e-06,
+      "loss": 0.7406,
+      "step": 670
+    },
+    {
+      "epoch": 0.11019645679798,
+      "grad_norm": 1.012184525401541,
+      "learning_rate": 8.99663685646556e-06,
+      "loss": 0.7323,
+      "step": 671
+    },
+    {
+      "epoch": 0.11036068400632275,
+      "grad_norm": 1.0204812192417894,
+      "learning_rate": 8.996618704151464e-06,
+      "loss": 0.7231,
+      "step": 672
+    },
+    {
+      "epoch": 0.11052491121466548,
+      "grad_norm": 1.0394114945935236,
+      "learning_rate": 8.996600502999663e-06,
+      "loss": 0.7318,
+      "step": 673
+    },
+    {
+      "epoch": 0.11068913842300823,
+      "grad_norm": 1.0071324286292476,
+      "learning_rate": 8.99658225301036e-06,
+      "loss": 0.7371,
+      "step": 674
+    },
+    {
+      "epoch": 0.11085336563135098,
+      "grad_norm": 1.0071241514792235,
+      "learning_rate": 8.996563954183747e-06,
+      "loss": 0.7263,
+      "step": 675
+    },
+    {
+      "epoch": 0.11101759283969372,
+      "grad_norm": 0.9849565269875252,
+      "learning_rate": 8.996545606520026e-06,
+      "loss": 0.7488,
+      "step": 676
+    },
+    {
+      "epoch": 0.11118182004803646,
+      "grad_norm": 0.9857903891350793,
+      "learning_rate": 8.996527210019395e-06,
+      "loss": 0.7781,
+      "step": 677
+    },
+    {
+      "epoch": 0.1113460472563792,
+      "grad_norm": 1.0008451419132542,
+      "learning_rate": 8.996508764682056e-06,
+      "loss": 0.7335,
+      "step": 678
+    },
+    {
+      "epoch": 0.11151027446472195,
+      "grad_norm": 0.9904057469198645,
+      "learning_rate": 8.996490270508207e-06,
+      "loss": 0.7509,
+      "step": 679
+    },
+    {
+      "epoch": 0.11167450167306468,
+      "grad_norm": 1.0154428424155522,
+      "learning_rate": 8.99647172749805e-06,
+      "loss": 0.7438,
+      "step": 680
+    },
+    {
+      "epoch": 0.11183872888140743,
+      "grad_norm": 0.9859511271051146,
+      "learning_rate": 8.996453135651785e-06,
+      "loss": 0.7311,
+      "step": 681
+    },
+    {
+      "epoch": 0.11200295608975017,
+      "grad_norm": 1.0172995678678334,
+      "learning_rate": 8.996434494969617e-06,
+      "loss": 0.7522,
+      "step": 682
+    },
+    {
+      "epoch": 0.11216718329809292,
+      "grad_norm": 0.9934227384074349,
+      "learning_rate": 8.996415805451744e-06,
+      "loss": 0.748,
+      "step": 683
+    },
+    {
+      "epoch": 0.11233141050643565,
+      "grad_norm": 0.9809402738647511,
+      "learning_rate": 8.996397067098373e-06,
+      "loss": 0.7384,
+      "step": 684
+    },
+    {
+      "epoch": 0.1124956377147784,
+      "grad_norm": 0.9627480326356829,
+      "learning_rate": 8.996378279909707e-06,
+      "loss": 0.7475,
+      "step": 685
+    },
+    {
+      "epoch": 0.11265986492312113,
+      "grad_norm": 0.9823816568610797,
+      "learning_rate": 8.996359443885948e-06,
+      "loss": 0.7546,
+      "step": 686
+    },
+    {
+      "epoch": 0.11282409213146388,
+      "grad_norm": 0.982802515323171,
+      "learning_rate": 8.9963405590273e-06,
+      "loss": 0.7489,
+      "step": 687
+    },
+    {
+      "epoch": 0.11298831933980662,
+      "grad_norm": 0.9965019206725246,
+      "learning_rate": 8.996321625333972e-06,
+      "loss": 0.7285,
+      "step": 688
+    },
+    {
+      "epoch": 0.11315254654814937,
+      "grad_norm": 1.0211395532769991,
+      "learning_rate": 8.996302642806166e-06,
+      "loss": 0.7529,
+      "step": 689
+    },
+    {
+      "epoch": 0.1133167737564921,
+      "grad_norm": 0.9443212451181419,
+      "learning_rate": 8.996283611444091e-06,
+      "loss": 0.7526,
+      "step": 690
+    },
+    {
+      "epoch": 0.11348100096483485,
+      "grad_norm": 0.9626893407176333,
+      "learning_rate": 8.99626453124795e-06,
+      "loss": 0.7228,
+      "step": 691
+    },
+    {
+      "epoch": 0.11364522817317758,
+      "grad_norm": 0.9730358742605486,
+      "learning_rate": 8.996245402217954e-06,
+      "loss": 0.744,
+      "step": 692
+    },
+    {
+      "epoch": 0.11380945538152033,
+      "grad_norm": 0.9563252686051095,
+      "learning_rate": 8.99622622435431e-06,
+      "loss": 0.7431,
+      "step": 693
+    },
+    {
+      "epoch": 0.11397368258986308,
+      "grad_norm": 0.9554055649452712,
+      "learning_rate": 8.996206997657225e-06,
+      "loss": 0.7432,
+      "step": 694
+    },
+    {
+      "epoch": 0.11413790979820582,
+      "grad_norm": 0.9467956137652699,
+      "learning_rate": 8.996187722126907e-06,
+      "loss": 0.743,
+      "step": 695
+    },
+    {
+      "epoch": 0.11430213700654857,
+      "grad_norm": 0.9489039079960571,
+      "learning_rate": 8.996168397763568e-06,
+      "loss": 0.7494,
+      "step": 696
+    },
+    {
+      "epoch": 0.1144663642148913,
+      "grad_norm": 0.9387244200911926,
+      "learning_rate": 8.996149024567416e-06,
+      "loss": 0.7428,
+      "step": 697
+    },
+    {
+      "epoch": 0.11463059142323405,
+      "grad_norm": 0.9501341353371187,
+      "learning_rate": 8.996129602538664e-06,
+      "loss": 0.7108,
+      "step": 698
+    },
+    {
+      "epoch": 0.11479481863157678,
+      "grad_norm": 0.962206616789228,
+      "learning_rate": 8.996110131677517e-06,
+      "loss": 0.7195,
+      "step": 699
+    },
+    {
+      "epoch": 0.11495904583991953,
+      "grad_norm": 0.9243568911177664,
+      "learning_rate": 8.996090611984192e-06,
+      "loss": 0.7557,
+      "step": 700
+    },
+    {
+      "epoch": 0.11512327304826227,
+      "grad_norm": 0.953170841709303,
+      "learning_rate": 8.996071043458902e-06,
+      "loss": 0.719,
+      "step": 701
+    },
+    {
+      "epoch": 0.11528750025660502,
+      "grad_norm": 0.9386842014299721,
+      "learning_rate": 8.996051426101855e-06,
+      "loss": 0.7436,
+      "step": 702
+    },
+    {
+      "epoch": 0.11545172746494775,
+      "grad_norm": 0.9274541909232491,
+      "learning_rate": 8.996031759913265e-06,
+      "loss": 0.7385,
+      "step": 703
+    },
+    {
+      "epoch": 0.1156159546732905,
+      "grad_norm": 0.9207183074144455,
+      "learning_rate": 8.996012044893347e-06,
+      "loss": 0.7592,
+      "step": 704
+    },
+    {
+      "epoch": 0.11578018188163323,
+      "grad_norm": 0.9062916445772407,
+      "learning_rate": 8.995992281042315e-06,
+      "loss": 0.7615,
+      "step": 705
+    },
+    {
+      "epoch": 0.11594440908997598,
+      "grad_norm": 0.9736278432415078,
+      "learning_rate": 8.995972468360383e-06,
+      "loss": 0.74,
+      "step": 706
+    },
+    {
+      "epoch": 0.11610863629831872,
+      "grad_norm": 0.9065148273263071,
+      "learning_rate": 8.995952606847767e-06,
+      "loss": 0.7568,
+      "step": 707
+    },
+    {
+      "epoch": 0.11627286350666147,
+      "grad_norm": 0.9193357664661844,
+      "learning_rate": 8.995932696504683e-06,
+      "loss": 0.7486,
+      "step": 708
+    },
+    {
+      "epoch": 0.1164370907150042,
+      "grad_norm": 0.9195501174668326,
+      "learning_rate": 8.995912737331345e-06,
+      "loss": 0.7532,
+      "step": 709
+    },
+    {
+      "epoch": 0.11660131792334695,
+      "grad_norm": 0.9507039973612097,
+      "learning_rate": 8.995892729327973e-06,
+      "loss": 0.7613,
+      "step": 710
+    },
+    {
+      "epoch": 0.1167655451316897,
+      "grad_norm": 0.9261363835047515,
+      "learning_rate": 8.995872672494781e-06,
+      "loss": 0.74,
+      "step": 711
+    },
+    {
+      "epoch": 0.11692977234003243,
+      "grad_norm": 0.9060734346127396,
+      "learning_rate": 8.99585256683199e-06,
+      "loss": 0.7368,
+      "step": 712
+    },
+    {
+      "epoch": 0.11709399954837518,
+      "grad_norm": 0.8977667506848039,
+      "learning_rate": 8.995832412339815e-06,
+      "loss": 0.7516,
+      "step": 713
+    },
+    {
+      "epoch": 0.11725822675671792,
+      "grad_norm": 0.8955651223846078,
+      "learning_rate": 8.995812209018479e-06,
+      "loss": 0.7456,
+      "step": 714
+    },
+    {
+      "epoch": 0.11742245396506067,
+      "grad_norm": 0.9030569385935932,
+      "learning_rate": 8.995791956868197e-06,
+      "loss": 0.7176,
+      "step": 715
+    },
+    {
+      "epoch": 0.1175866811734034,
+      "grad_norm": 0.8942305604018785,
+      "learning_rate": 8.995771655889192e-06,
+      "loss": 0.7384,
+      "step": 716
+    },
+    {
+      "epoch": 0.11775090838174615,
+      "grad_norm": 0.9027311362567876,
+      "learning_rate": 8.995751306081683e-06,
+      "loss": 0.7405,
+      "step": 717
+    },
+    {
+      "epoch": 0.11791513559008888,
+      "grad_norm": 0.885311457940408,
+      "learning_rate": 8.995730907445893e-06,
+      "loss": 0.7568,
+      "step": 718
+    },
+    {
+      "epoch": 0.11807936279843163,
+      "grad_norm": 0.9012435000861214,
+      "learning_rate": 8.99571045998204e-06,
+      "loss": 0.7371,
+      "step": 719
+    },
+    {
+      "epoch": 0.11824359000677437,
+      "grad_norm": 0.907146848660689,
+      "learning_rate": 8.99568996369035e-06,
+      "loss": 0.7292,
+      "step": 720
+    },
+    {
+      "epoch": 0.11840781721511712,
+      "grad_norm": 1.0063300221437657,
+      "learning_rate": 8.995669418571042e-06,
+      "loss": 0.7459,
+      "step": 721
+    },
+    {
+      "epoch": 0.11857204442345985,
+      "grad_norm": 0.9072193733429393,
+      "learning_rate": 8.995648824624344e-06,
+      "loss": 0.7195,
+      "step": 722
+    },
+    {
+      "epoch": 0.1187362716318026,
+      "grad_norm": 0.9224186169335864,
+      "learning_rate": 8.995628181850475e-06,
+      "loss": 0.7263,
+      "step": 723
+    },
+    {
+      "epoch": 0.11890049884014534,
+      "grad_norm": 0.8730762718625776,
+      "learning_rate": 8.99560749024966e-06,
+      "loss": 0.7508,
+      "step": 724
+    },
+    {
+      "epoch": 0.11906472604848808,
+      "grad_norm": 0.8982887153940247,
+      "learning_rate": 8.995586749822124e-06,
+      "loss": 0.7269,
+      "step": 725
+    },
+    {
+      "epoch": 0.11922895325683082,
+      "grad_norm": 0.9068289642355488,
+      "learning_rate": 8.995565960568095e-06,
+      "loss": 0.7465,
+      "step": 726
+    },
+    {
+      "epoch": 0.11939318046517357,
+      "grad_norm": 0.8875535394981413,
+      "learning_rate": 8.995545122487797e-06,
+      "loss": 0.72,
+      "step": 727
+    },
+    {
+      "epoch": 0.11955740767351632,
+      "grad_norm": 0.8679336550118347,
+      "learning_rate": 8.995524235581455e-06,
+      "loss": 0.7391,
+      "step": 728
+    },
+    {
+      "epoch": 0.11972163488185905,
+      "grad_norm": 0.8826045646205907,
+      "learning_rate": 8.995503299849294e-06,
+      "loss": 0.7378,
+      "step": 729
+    },
+    {
+      "epoch": 0.1198858620902018,
+      "grad_norm": 0.8701194777603063,
+      "learning_rate": 8.995482315291548e-06,
+      "loss": 0.7364,
+      "step": 730
+    },
+    {
+      "epoch": 0.12005008929854454,
+      "grad_norm": 0.8893542492932401,
+      "learning_rate": 8.99546128190844e-06,
+      "loss": 0.7361,
+      "step": 731
+    },
+    {
+      "epoch": 0.12021431650688728,
+      "grad_norm": 0.8765486658546501,
+      "learning_rate": 8.995440199700199e-06,
+      "loss": 0.7239,
+      "step": 732
+    },
+    {
+      "epoch": 0.12037854371523002,
+      "grad_norm": 0.8761211470503778,
+      "learning_rate": 8.995419068667056e-06,
+      "loss": 0.7118,
+      "step": 733
+    },
+    {
+      "epoch": 0.12054277092357277,
+      "grad_norm": 0.8784561432542338,
+      "learning_rate": 8.995397888809236e-06,
+      "loss": 0.7252,
+      "step": 734
+    },
+    {
+      "epoch": 0.1207069981319155,
+      "grad_norm": 0.885958895044403,
+      "learning_rate": 8.995376660126975e-06,
+      "loss": 0.7289,
+      "step": 735
+    },
+    {
+      "epoch": 0.12087122534025825,
+      "grad_norm": 1.3628303359719938,
+      "learning_rate": 8.995355382620499e-06,
+      "loss": 0.7363,
+      "step": 736
+    },
+    {
+      "epoch": 0.12103545254860099,
+      "grad_norm": 0.8620243858978899,
+      "learning_rate": 8.995334056290042e-06,
+      "loss": 0.7345,
+      "step": 737
+    },
+    {
+      "epoch": 0.12119967975694373,
+      "grad_norm": 0.8616681510393484,
+      "learning_rate": 8.995312681135833e-06,
+      "loss": 0.7369,
+      "step": 738
+    },
+    {
+      "epoch": 0.12136390696528647,
+      "grad_norm": 0.8536335524173874,
+      "learning_rate": 8.995291257158106e-06,
+      "loss": 0.7301,
+      "step": 739
+    },
+    {
+      "epoch": 0.12152813417362922,
+      "grad_norm": 0.8890164010727863,
+      "learning_rate": 8.995269784357094e-06,
+      "loss": 0.7259,
+      "step": 740
+    },
+    {
+      "epoch": 0.12169236138197195,
+      "grad_norm": 0.8584462694033533,
+      "learning_rate": 8.995248262733028e-06,
+      "loss": 0.7211,
+      "step": 741
+    },
+    {
+      "epoch": 0.1218565885903147,
+      "grad_norm": 0.8473041107598641,
+      "learning_rate": 8.995226692286144e-06,
+      "loss": 0.7446,
+      "step": 742
+    },
+    {
+      "epoch": 0.12202081579865744,
+      "grad_norm": 0.857983951564626,
+      "learning_rate": 8.995205073016676e-06,
+      "loss": 0.7486,
+      "step": 743
+    },
+    {
+      "epoch": 0.12218504300700019,
+      "grad_norm": 0.8499021660743089,
+      "learning_rate": 8.995183404924856e-06,
+      "loss": 0.7252,
+      "step": 744
+    },
+    {
+      "epoch": 0.12234927021534292,
+      "grad_norm": 0.8432649022601657,
+      "learning_rate": 8.995161688010924e-06,
+      "loss": 0.7361,
+      "step": 745
+    },
+    {
+      "epoch": 0.12251349742368567,
+      "grad_norm": 0.8349936223743508,
+      "learning_rate": 8.995139922275112e-06,
+      "loss": 0.728,
+      "step": 746
+    },
+    {
+      "epoch": 0.12267772463202842,
+      "grad_norm": 0.8359575337444805,
+      "learning_rate": 8.995118107717658e-06,
+      "loss": 0.7339,
+      "step": 747
+    },
+    {
+      "epoch": 0.12284195184037115,
+      "grad_norm": 0.8491789703211692,
+      "learning_rate": 8.995096244338799e-06,
+      "loss": 0.7273,
+      "step": 748
+    },
+    {
+      "epoch": 0.1230061790487139,
+      "grad_norm": 0.8343756782871875,
+      "learning_rate": 8.995074332138771e-06,
+      "loss": 0.7376,
+      "step": 749
+    },
+    {
+      "epoch": 0.12317040625705664,
+      "grad_norm": 0.8177523292597683,
+      "learning_rate": 8.995052371117816e-06,
+      "loss": 0.7498,
+      "step": 750
+    },
+    {
+      "epoch": 0.12333463346539938,
+      "grad_norm": 0.8275634000310962,
+      "learning_rate": 8.995030361276167e-06,
+      "loss": 0.7181,
+      "step": 751
+    },
+    {
+      "epoch": 0.12349886067374212,
+      "grad_norm": 0.8339533594185612,
+      "learning_rate": 8.995008302614069e-06,
+      "loss": 0.7265,
+      "step": 752
+    },
+    {
+      "epoch": 0.12366308788208487,
+      "grad_norm": 0.8279085022505449,
+      "learning_rate": 8.994986195131755e-06,
+      "loss": 0.7269,
+      "step": 753
+    },
+    {
+      "epoch": 0.1238273150904276,
+      "grad_norm": 0.8346885779308945,
+      "learning_rate": 8.99496403882947e-06,
+      "loss": 0.7239,
+      "step": 754
+    },
+    {
+      "epoch": 0.12399154229877035,
+      "grad_norm": 0.8535778930786089,
+      "learning_rate": 8.994941833707452e-06,
+      "loss": 0.7335,
+      "step": 755
+    },
+    {
+      "epoch": 0.12415576950711309,
+      "grad_norm": 0.8433517095759425,
+      "learning_rate": 8.994919579765944e-06,
+      "loss": 0.7262,
+      "step": 756
+    },
+    {
+      "epoch": 0.12431999671545584,
+      "grad_norm": 0.8115026773168339,
+      "learning_rate": 8.994897277005187e-06,
+      "loss": 0.7543,
+      "step": 757
+    },
+    {
+      "epoch": 0.12448422392379857,
+      "grad_norm": 0.8237859019603188,
+      "learning_rate": 8.994874925425423e-06,
+      "loss": 0.696,
+      "step": 758
+    },
+    {
+      "epoch": 0.12464845113214132,
+      "grad_norm": 0.9508056714088163,
+      "learning_rate": 8.994852525026896e-06,
+      "loss": 0.7313,
+      "step": 759
+    },
+    {
+      "epoch": 0.12481267834048405,
+      "grad_norm": 0.8253401090339175,
+      "learning_rate": 8.994830075809846e-06,
+      "loss": 0.7333,
+      "step": 760
+    },
+    {
+      "epoch": 0.1249769055488268,
+      "grad_norm": 0.8141935870360409,
+      "learning_rate": 8.99480757777452e-06,
+      "loss": 0.7333,
+      "step": 761
+    },
+    {
+      "epoch": 0.12514113275716954,
+      "grad_norm": 0.8136553415480036,
+      "learning_rate": 8.994785030921164e-06,
+      "loss": 0.7293,
+      "step": 762
+    },
+    {
+      "epoch": 0.1253053599655123,
+      "grad_norm": 0.8063988080685636,
+      "learning_rate": 8.994762435250018e-06,
+      "loss": 0.7398,
+      "step": 763
+    },
+    {
+      "epoch": 0.12546958717385504,
+      "grad_norm": 0.8024743885585705,
+      "learning_rate": 8.994739790761329e-06,
+      "loss": 0.7318,
+      "step": 764
+    },
+    {
+      "epoch": 0.12563381438219778,
+      "grad_norm": 0.8210883509073227,
+      "learning_rate": 8.994717097455346e-06,
+      "loss": 0.7128,
+      "step": 765
+    },
+    {
+      "epoch": 0.1257980415905405,
+      "grad_norm": 0.8081208044654137,
+      "learning_rate": 8.994694355332311e-06,
+      "loss": 0.7216,
+      "step": 766
+    },
+    {
+      "epoch": 0.12596226879888325,
+      "grad_norm": 0.8043457567543434,
+      "learning_rate": 8.994671564392472e-06,
+      "loss": 0.729,
+      "step": 767
+    },
+    {
+      "epoch": 0.126126496007226,
+      "grad_norm": 0.8332594138537104,
+      "learning_rate": 8.994648724636081e-06,
+      "loss": 0.7152,
+      "step": 768
+    },
+    {
+      "epoch": 0.12629072321556875,
+      "grad_norm": 0.796400256746438,
+      "learning_rate": 8.994625836063381e-06,
+      "loss": 0.7337,
+      "step": 769
+    },
+    {
+      "epoch": 0.12645495042391147,
+      "grad_norm": 0.7970766048853704,
+      "learning_rate": 8.99460289867462e-06,
+      "loss": 0.7385,
+      "step": 770
+    },
+    {
+      "epoch": 0.12661917763225422,
+      "grad_norm": 0.8168189102976473,
+      "learning_rate": 8.994579912470054e-06,
+      "loss": 0.7207,
+      "step": 771
+    },
+    {
+      "epoch": 0.12678340484059697,
+      "grad_norm": 0.8210337972298519,
+      "learning_rate": 8.994556877449926e-06,
+      "loss": 0.7173,
+      "step": 772
+    },
+    {
+      "epoch": 0.12694763204893972,
+      "grad_norm": 0.7928567401168414,
+      "learning_rate": 8.994533793614486e-06,
+      "loss": 0.7299,
+      "step": 773
+    },
+    {
+      "epoch": 0.12711185925728244,
+      "grad_norm": 0.7867827344107035,
+      "learning_rate": 8.99451066096399e-06,
+      "loss": 0.7327,
+      "step": 774
+    },
+    {
+      "epoch": 0.1272760864656252,
+      "grad_norm": 0.8101031512180691,
+      "learning_rate": 8.994487479498686e-06,
+      "loss": 0.7201,
+      "step": 775
+    },
+    {
+      "epoch": 0.12744031367396794,
+      "grad_norm": 0.8149592408788043,
+      "learning_rate": 8.994464249218824e-06,
+      "loss": 0.7059,
+      "step": 776
+    },
+    {
+      "epoch": 0.12760454088231069,
+      "grad_norm": 0.7976670057578504,
+      "learning_rate": 8.99444097012466e-06,
+      "loss": 0.7234,
+      "step": 777
+    },
+    {
+      "epoch": 0.1277687680906534,
+      "grad_norm": 0.7848442475451943,
+      "learning_rate": 8.994417642216445e-06,
+      "loss": 0.7224,
+      "step": 778
+    },
+    {
+      "epoch": 0.12793299529899615,
+      "grad_norm": 0.7831415532113032,
+      "learning_rate": 8.994394265494434e-06,
+      "loss": 0.7116,
+      "step": 779
+    },
+    {
+      "epoch": 0.1280972225073389,
+      "grad_norm": 0.8042009917600411,
+      "learning_rate": 8.994370839958876e-06,
+      "loss": 0.7249,
+      "step": 780
+    },
+    {
+      "epoch": 0.12826144971568165,
+      "grad_norm": 0.7823737809741357,
+      "learning_rate": 8.994347365610032e-06,
+      "loss": 0.7148,
+      "step": 781
+    },
+    {
+      "epoch": 0.1284256769240244,
+      "grad_norm": 0.7915106748965738,
+      "learning_rate": 8.994323842448152e-06,
+      "loss": 0.7328,
+      "step": 782
+    },
+    {
+      "epoch": 0.12858990413236712,
+      "grad_norm": 0.7721916884200549,
+      "learning_rate": 8.994300270473493e-06,
+      "loss": 0.7112,
+      "step": 783
+    },
+    {
+      "epoch": 0.12875413134070987,
+      "grad_norm": 0.7720807666798172,
+      "learning_rate": 8.994276649686313e-06,
+      "loss": 0.7276,
+      "step": 784
+    },
+    {
+      "epoch": 0.12891835854905262,
+      "grad_norm": 0.7812178056014114,
+      "learning_rate": 8.994252980086865e-06,
+      "loss": 0.7125,
+      "step": 785
+    },
+    {
+      "epoch": 0.12908258575739537,
+      "grad_norm": 0.7552866535606362,
+      "learning_rate": 8.994229261675408e-06,
+      "loss": 0.7384,
+      "step": 786
+    },
+    {
+      "epoch": 0.1292468129657381,
+      "grad_norm": 0.7736822903510917,
+      "learning_rate": 8.9942054944522e-06,
+      "loss": 0.7231,
+      "step": 787
+    },
+    {
+      "epoch": 0.12941104017408084,
+      "grad_norm": 0.7863166996807561,
+      "learning_rate": 8.9941816784175e-06,
+      "loss": 0.7125,
+      "step": 788
+    },
+    {
+      "epoch": 0.1295752673824236,
+      "grad_norm": 0.7584645168372531,
+      "learning_rate": 8.994157813571563e-06,
+      "loss": 0.7218,
+      "step": 789
+    },
+    {
+      "epoch": 0.12973949459076634,
+      "grad_norm": 0.7591392100122629,
+      "learning_rate": 8.99413389991465e-06,
+      "loss": 0.7196,
+      "step": 790
+    },
+    {
+      "epoch": 0.12990372179910906,
+      "grad_norm": 0.7638964528596006,
+      "learning_rate": 8.994109937447023e-06,
+      "loss": 0.742,
+      "step": 791
+    },
+    {
+      "epoch": 0.1300679490074518,
+      "grad_norm": 0.7638417423029302,
+      "learning_rate": 8.994085926168942e-06,
+      "loss": 0.7144,
+      "step": 792
+    },
+    {
+      "epoch": 0.13023217621579455,
+      "grad_norm": 0.7531125912433023,
+      "learning_rate": 8.994061866080662e-06,
+      "loss": 0.7181,
+      "step": 793
+    },
+    {
+      "epoch": 0.1303964034241373,
+      "grad_norm": 0.7638871871583925,
+      "learning_rate": 8.994037757182452e-06,
+      "loss": 0.7263,
+      "step": 794
+    },
+    {
+      "epoch": 0.13056063063248002,
+      "grad_norm": 0.7712535656244461,
+      "learning_rate": 8.99401359947457e-06,
+      "loss": 0.7203,
+      "step": 795
+    },
+    {
+      "epoch": 0.13072485784082277,
+      "grad_norm": 0.7534551679287208,
+      "learning_rate": 8.993989392957277e-06,
+      "loss": 0.7233,
+      "step": 796
+    },
+    {
+      "epoch": 0.13088908504916552,
+      "grad_norm": 0.7541686256824629,
+      "learning_rate": 8.99396513763084e-06,
+      "loss": 0.7191,
+      "step": 797
+    },
+    {
+      "epoch": 0.13105331225750827,
+      "grad_norm": 0.7558776359791881,
+      "learning_rate": 8.993940833495517e-06,
+      "loss": 0.7189,
+      "step": 798
+    },
+    {
+      "epoch": 0.13121753946585102,
+      "grad_norm": 0.7618419103671177,
+      "learning_rate": 8.993916480551577e-06,
+      "loss": 0.7279,
+      "step": 799
+    },
+    {
+      "epoch": 0.13138176667419374,
+      "grad_norm": 0.768280234616204,
+      "learning_rate": 8.993892078799284e-06,
+      "loss": 0.7126,
+      "step": 800
+    },
+    {
+      "epoch": 0.1315459938825365,
+      "grad_norm": 0.7394145525562601,
+      "learning_rate": 8.993867628238901e-06,
+      "loss": 0.7446,
+      "step": 801
+    },
+    {
+      "epoch": 0.13171022109087924,
+      "grad_norm": 0.7615299280436207,
+      "learning_rate": 8.993843128870692e-06,
+      "loss": 0.6987,
+      "step": 802
+    },
+    {
+      "epoch": 0.13187444829922199,
+      "grad_norm": 0.7465231091292661,
+      "learning_rate": 8.993818580694929e-06,
+      "loss": 0.7378,
+      "step": 803
+    },
+    {
+      "epoch": 0.1320386755075647,
+      "grad_norm": 0.7491298101718952,
+      "learning_rate": 8.993793983711872e-06,
+      "loss": 0.6964,
+      "step": 804
+    },
+    {
+      "epoch": 0.13220290271590746,
+      "grad_norm": 0.7326662321957933,
+      "learning_rate": 8.99376933792179e-06,
+      "loss": 0.7346,
+      "step": 805
+    },
+    {
+      "epoch": 0.1323671299242502,
+      "grad_norm": 0.7405957735369632,
+      "learning_rate": 8.993744643324954e-06,
+      "loss": 0.7056,
+      "step": 806
+    },
+    {
+      "epoch": 0.13253135713259295,
+      "grad_norm": 0.7309108095519508,
+      "learning_rate": 8.993719899921629e-06,
+      "loss": 0.7229,
+      "step": 807
+    },
+    {
+      "epoch": 0.13269558434093567,
+      "grad_norm": 0.7965004013613755,
+      "learning_rate": 8.993695107712085e-06,
+      "loss": 0.7272,
+      "step": 808
+    },
+    {
+      "epoch": 0.13285981154927842,
+      "grad_norm": 0.7405587765052916,
+      "learning_rate": 8.993670266696592e-06,
+      "loss": 0.7033,
+      "step": 809
+    },
+    {
+      "epoch": 0.13302403875762117,
+      "grad_norm": 0.738947917878739,
+      "learning_rate": 8.993645376875417e-06,
+      "loss": 0.7037,
+      "step": 810
+    },
+    {
+      "epoch": 0.13318826596596392,
+      "grad_norm": 0.7266519422397545,
+      "learning_rate": 8.993620438248834e-06,
+      "loss": 0.7105,
+      "step": 811
+    },
+    {
+      "epoch": 0.13335249317430664,
+      "grad_norm": 0.7321881237545829,
+      "learning_rate": 8.99359545081711e-06,
+      "loss": 0.719,
+      "step": 812
+    },
+    {
+      "epoch": 0.1335167203826494,
+      "grad_norm": 0.7343760502140673,
+      "learning_rate": 8.99357041458052e-06,
+      "loss": 0.6994,
+      "step": 813
+    },
+    {
+      "epoch": 0.13368094759099214,
+      "grad_norm": 0.7360965253226606,
+      "learning_rate": 8.993545329539331e-06,
+      "loss": 0.7087,
+      "step": 814
+    },
+    {
+      "epoch": 0.1338451747993349,
+      "grad_norm": 0.7312711387682936,
+      "learning_rate": 8.993520195693823e-06,
+      "loss": 0.703,
+      "step": 815
+    },
+    {
+      "epoch": 0.1340094020076776,
+      "grad_norm": 0.7354108968862768,
+      "learning_rate": 8.99349501304426e-06,
+      "loss": 0.7163,
+      "step": 816
+    },
+    {
+      "epoch": 0.13417362921602036,
+      "grad_norm": 0.7263178056025598,
+      "learning_rate": 8.993469781590924e-06,
+      "loss": 0.7257,
+      "step": 817
+    },
+    {
+      "epoch": 0.1343378564243631,
+      "grad_norm": 0.7416571311831478,
+      "learning_rate": 8.993444501334085e-06,
+      "loss": 0.7083,
+      "step": 818
+    },
+    {
+      "epoch": 0.13450208363270585,
+      "grad_norm": 0.7849234501566686,
+      "learning_rate": 8.993419172274016e-06,
+      "loss": 0.7128,
+      "step": 819
+    },
+    {
+      "epoch": 0.1346663108410486,
+      "grad_norm": 0.7178383001551933,
+      "learning_rate": 8.993393794410995e-06,
+      "loss": 0.7131,
+      "step": 820
+    },
+    {
+      "epoch": 0.13483053804939132,
+      "grad_norm": 0.7241642838444384,
+      "learning_rate": 8.993368367745295e-06,
+      "loss": 0.7138,
+      "step": 821
+    },
+    {
+      "epoch": 0.13499476525773407,
+      "grad_norm": 0.7168737582699622,
+      "learning_rate": 8.993342892277194e-06,
+      "loss": 0.7322,
+      "step": 822
+    },
+    {
+      "epoch": 0.13515899246607682,
+      "grad_norm": 0.7120308237386352,
+      "learning_rate": 8.99331736800697e-06,
+      "loss": 0.6974,
+      "step": 823
+    },
+    {
+      "epoch": 0.13532321967441957,
+      "grad_norm": 0.7321566228567261,
+      "learning_rate": 8.993291794934896e-06,
+      "loss": 0.6999,
+      "step": 824
+    },
+    {
+      "epoch": 0.1354874468827623,
+      "grad_norm": 0.7438779269308436,
+      "learning_rate": 8.993266173061255e-06,
+      "loss": 0.6944,
+      "step": 825
+    },
+    {
+      "epoch": 0.13565167409110504,
+      "grad_norm": 0.7118594659041331,
+      "learning_rate": 8.993240502386322e-06,
+      "loss": 0.7189,
+      "step": 826
+    },
+    {
+      "epoch": 0.1358159012994478,
+      "grad_norm": 0.7361564635340935,
+      "learning_rate": 8.993214782910375e-06,
+      "loss": 0.6926,
+      "step": 827
+    },
+    {
+      "epoch": 0.13598012850779054,
+      "grad_norm": 0.711244373700878,
+      "learning_rate": 8.993189014633696e-06,
+      "loss": 0.7185,
+      "step": 828
+    },
+    {
+      "epoch": 0.13614435571613326,
+      "grad_norm": 0.7385931746846988,
+      "learning_rate": 8.993163197556563e-06,
+      "loss": 0.6904,
+      "step": 829
+    },
+    {
+      "epoch": 0.136308582924476,
+      "grad_norm": 0.7135997031691055,
+      "learning_rate": 8.993137331679257e-06,
+      "loss": 0.7108,
+      "step": 830
+    },
+    {
+      "epoch": 0.13647281013281876,
+      "grad_norm": 0.7246704678281216,
+      "learning_rate": 8.993111417002059e-06,
+      "loss": 0.7113,
+      "step": 831
+    },
+    {
+      "epoch": 0.1366370373411615,
+      "grad_norm": 0.7091228636214885,
+      "learning_rate": 8.993085453525251e-06,
+      "loss": 0.7166,
+      "step": 832
+    },
+    {
+      "epoch": 0.13680126454950423,
+      "grad_norm": 0.698307172462183,
+      "learning_rate": 8.993059441249113e-06,
+      "loss": 0.7131,
+      "step": 833
+    },
+    {
+      "epoch": 0.13696549175784697,
+      "grad_norm": 0.7302212267089161,
+      "learning_rate": 8.99303338017393e-06,
+      "loss": 0.7153,
+      "step": 834
+    },
+    {
+      "epoch": 0.13712971896618972,
+      "grad_norm": 0.7133967428699537,
+      "learning_rate": 8.993007270299985e-06,
+      "loss": 0.7199,
+      "step": 835
+    },
+    {
+      "epoch": 0.13729394617453247,
+      "grad_norm": 0.6999377770634104,
+      "learning_rate": 8.992981111627558e-06,
+      "loss": 0.7037,
+      "step": 836
+    },
+    {
+      "epoch": 0.13745817338287522,
+      "grad_norm": 0.7036896301199365,
+      "learning_rate": 8.992954904156937e-06,
+      "loss": 0.702,
+      "step": 837
+    },
+    {
+      "epoch": 0.13762240059121794,
+      "grad_norm": 0.6896374085707714,
+      "learning_rate": 8.992928647888405e-06,
+      "loss": 0.7106,
+      "step": 838
+    },
+    {
+      "epoch": 0.1377866277995607,
+      "grad_norm": 0.7269208950900521,
+      "learning_rate": 8.992902342822247e-06,
+      "loss": 0.7068,
+      "step": 839
+    },
+    {
+      "epoch": 0.13795085500790344,
+      "grad_norm": 0.7046937086091937,
+      "learning_rate": 8.99287598895875e-06,
+      "loss": 0.703,
+      "step": 840
+    },
+    {
+      "epoch": 0.1381150822162462,
+      "grad_norm": 0.7121950225185115,
+      "learning_rate": 8.9928495862982e-06,
+      "loss": 0.7152,
+      "step": 841
+    },
+    {
+      "epoch": 0.1382793094245889,
+      "grad_norm": 0.7026225935912559,
+      "learning_rate": 8.992823134840882e-06,
+      "loss": 0.6939,
+      "step": 842
+    },
+    {
+      "epoch": 0.13844353663293166,
+      "grad_norm": 0.6970719648209621,
+      "learning_rate": 8.992796634587084e-06,
+      "loss": 0.7043,
+      "step": 843
+    },
+    {
+      "epoch": 0.1386077638412744,
+      "grad_norm": 0.6804608164044504,
+      "learning_rate": 8.992770085537095e-06,
+      "loss": 0.7077,
+      "step": 844
+    },
+    {
+      "epoch": 0.13877199104961715,
+      "grad_norm": 0.6956774935332672,
+      "learning_rate": 8.992743487691202e-06,
+      "loss": 0.706,
+      "step": 845
+    },
+    {
+      "epoch": 0.13893621825795988,
+      "grad_norm": 0.6965453226724493,
+      "learning_rate": 8.992716841049694e-06,
+      "loss": 0.6805,
+      "step": 846
+    },
+    {
+      "epoch": 0.13910044546630262,
+      "grad_norm": 0.6834149944374464,
+      "learning_rate": 8.992690145612864e-06,
+      "loss": 0.7184,
+      "step": 847
+    },
+    {
+      "epoch": 0.13926467267464537,
+      "grad_norm": 0.6878178766526616,
+      "learning_rate": 8.992663401380995e-06,
+      "loss": 0.6993,
+      "step": 848
+    },
+    {
+      "epoch": 0.13942889988298812,
+      "grad_norm": 0.6872283526043412,
+      "learning_rate": 8.992636608354383e-06,
+      "loss": 0.6862,
+      "step": 849
+    },
+    {
+      "epoch": 0.13959312709133084,
+      "grad_norm": 0.6963162299008429,
+      "learning_rate": 8.992609766533318e-06,
+      "loss": 0.6932,
+      "step": 850
+    },
+    {
+      "epoch": 0.1397573542996736,
+      "grad_norm": 0.6895674941257026,
+      "learning_rate": 8.992582875918089e-06,
+      "loss": 0.7016,
+      "step": 851
+    },
+    {
+      "epoch": 0.13992158150801634,
+      "grad_norm": 0.6984241181890183,
+      "learning_rate": 8.992555936508991e-06,
+      "loss": 0.7067,
+      "step": 852
+    },
+    {
+      "epoch": 0.1400858087163591,
+      "grad_norm": 0.6905598521116458,
+      "learning_rate": 8.992528948306314e-06,
+      "loss": 0.6937,
+      "step": 853
+    },
+    {
+      "epoch": 0.14025003592470184,
+      "grad_norm": 0.685496447106375,
+      "learning_rate": 8.992501911310354e-06,
+      "loss": 0.6928,
+      "step": 854
+    },
+    {
+      "epoch": 0.14041426313304456,
+      "grad_norm": 0.7186605931410012,
+      "learning_rate": 8.992474825521403e-06,
+      "loss": 0.6888,
+      "step": 855
+    },
+    {
+      "epoch": 0.1405784903413873,
+      "grad_norm": 0.6926681031006332,
+      "learning_rate": 8.992447690939756e-06,
+      "loss": 0.6945,
+      "step": 856
+    },
+    {
+      "epoch": 0.14074271754973006,
+      "grad_norm": 0.6687413046389367,
+      "learning_rate": 8.992420507565707e-06,
+      "loss": 0.6915,
+      "step": 857
+    },
+    {
+      "epoch": 0.1409069447580728,
+      "grad_norm": 0.6750901909722651,
+      "learning_rate": 8.99239327539955e-06,
+      "loss": 0.6972,
+      "step": 858
+    },
+    {
+      "epoch": 0.14107117196641553,
+      "grad_norm": 0.7056162625158083,
+      "learning_rate": 8.992365994441583e-06,
+      "loss": 0.6981,
+      "step": 859
+    },
+    {
+      "epoch": 0.14123539917475827,
+      "grad_norm": 0.6767374335973468,
+      "learning_rate": 8.992338664692101e-06,
+      "loss": 0.704,
+      "step": 860
+    },
+    {
+      "epoch": 0.14139962638310102,
+      "grad_norm": 0.6824925357290131,
+      "learning_rate": 8.992311286151403e-06,
+      "loss": 0.7126,
+      "step": 861
+    },
+    {
+      "epoch": 0.14156385359144377,
+      "grad_norm": 0.6611734657558349,
+      "learning_rate": 8.992283858819784e-06,
+      "loss": 0.7142,
+      "step": 862
+    },
+    {
+      "epoch": 0.1417280807997865,
+      "grad_norm": 0.6632695371380769,
+      "learning_rate": 8.992256382697539e-06,
+      "loss": 0.7232,
+      "step": 863
+    },
+    {
+      "epoch": 0.14189230800812924,
+      "grad_norm": 0.6626421044508288,
+      "learning_rate": 8.992228857784975e-06,
+      "loss": 0.7306,
+      "step": 864
+    },
+    {
+      "epoch": 0.142056535216472,
+      "grad_norm": 0.6765717769175646,
+      "learning_rate": 8.992201284082383e-06,
+      "loss": 0.6982,
+      "step": 865
+    },
+    {
+      "epoch": 0.14222076242481474,
+      "grad_norm": 0.6708039962729273,
+      "learning_rate": 8.992173661590065e-06,
+      "loss": 0.7088,
+      "step": 866
+    },
+    {
+      "epoch": 0.14238498963315746,
+      "grad_norm": 0.6793311754143073,
+      "learning_rate": 8.992145990308323e-06,
+      "loss": 0.7004,
+      "step": 867
+    },
+    {
+      "epoch": 0.1425492168415002,
+      "grad_norm": 0.6657408958013594,
+      "learning_rate": 8.992118270237455e-06,
+      "loss": 0.6799,
+      "step": 868
+    },
+    {
+      "epoch": 0.14271344404984296,
+      "grad_norm": 0.6540758971441215,
+      "learning_rate": 8.992090501377762e-06,
+      "loss": 0.7228,
+      "step": 869
+    },
+    {
+      "epoch": 0.1428776712581857,
+      "grad_norm": 0.6643906221820899,
+      "learning_rate": 8.992062683729548e-06,
+      "loss": 0.7104,
+      "step": 870
+    },
+    {
+      "epoch": 0.14304189846652846,
+      "grad_norm": 0.6989964341191656,
+      "learning_rate": 8.992034817293112e-06,
+      "loss": 0.6858,
+      "step": 871
+    },
+    {
+      "epoch": 0.14320612567487118,
+      "grad_norm": 0.6812360050752753,
+      "learning_rate": 8.99200690206876e-06,
+      "loss": 0.7117,
+      "step": 872
+    },
+    {
+      "epoch": 0.14337035288321393,
+      "grad_norm": 0.6949187113631994,
+      "learning_rate": 8.991978938056793e-06,
+      "loss": 0.7152,
+      "step": 873
+    },
+    {
+      "epoch": 0.14353458009155667,
+      "grad_norm": 0.6577059262682897,
+      "learning_rate": 8.991950925257517e-06,
+      "loss": 0.711,
+      "step": 874
+    },
+    {
+      "epoch": 0.14369880729989942,
+      "grad_norm": 0.662738282110258,
+      "learning_rate": 8.991922863671232e-06,
+      "loss": 0.6881,
+      "step": 875
+    },
+    {
+      "epoch": 0.14386303450824214,
+      "grad_norm": 0.6782128483184631,
+      "learning_rate": 8.991894753298245e-06,
+      "loss": 0.6662,
+      "step": 876
+    },
+    {
+      "epoch": 0.1440272617165849,
+      "grad_norm": 0.6571616469132535,
+      "learning_rate": 8.991866594138862e-06,
+      "loss": 0.7077,
+      "step": 877
+    },
+    {
+      "epoch": 0.14419148892492764,
+      "grad_norm": 0.6516355933069562,
+      "learning_rate": 8.991838386193388e-06,
+      "loss": 0.7009,
+      "step": 878
+    },
+    {
+      "epoch": 0.1443557161332704,
+      "grad_norm": 0.6556308700827564,
+      "learning_rate": 8.991810129462132e-06,
+      "loss": 0.7031,
+      "step": 879
+    },
+    {
+      "epoch": 0.1445199433416131,
+      "grad_norm": 0.6627545749856372,
+      "learning_rate": 8.991781823945398e-06,
+      "loss": 0.6717,
+      "step": 880
+    },
+    {
+      "epoch": 0.14468417054995586,
+      "grad_norm": 0.6482910836902056,
+      "learning_rate": 8.991753469643493e-06,
+      "loss": 0.7217,
+      "step": 881
+    },
+    {
+      "epoch": 0.1448483977582986,
+      "grad_norm": 0.7686915515107932,
+      "learning_rate": 8.991725066556726e-06,
+      "loss": 0.7087,
+      "step": 882
+    },
+    {
+      "epoch": 0.14501262496664136,
+      "grad_norm": 0.633391471242811,
+      "learning_rate": 8.991696614685406e-06,
+      "loss": 0.7121,
+      "step": 883
+    },
+    {
+      "epoch": 0.14517685217498408,
+      "grad_norm": 0.6592509022850245,
+      "learning_rate": 8.99166811402984e-06,
+      "loss": 0.6901,
+      "step": 884
+    },
+    {
+      "epoch": 0.14534107938332683,
+      "grad_norm": 0.6462500462113238,
+      "learning_rate": 8.991639564590342e-06,
+      "loss": 0.6892,
+      "step": 885
+    },
+    {
+      "epoch": 0.14550530659166958,
+      "grad_norm": 0.6449101927159558,
+      "learning_rate": 8.991610966367217e-06,
+      "loss": 0.6986,
+      "step": 886
+    },
+    {
+      "epoch": 0.14566953380001232,
+      "grad_norm": 0.6513519292471238,
+      "learning_rate": 8.991582319360779e-06,
+      "loss": 0.6952,
+      "step": 887
+    },
+    {
+      "epoch": 0.14583376100835507,
+      "grad_norm": 0.6351156522282531,
+      "learning_rate": 8.991553623571336e-06,
+      "loss": 0.6947,
+      "step": 888
+    },
+    {
+      "epoch": 0.1459979882166978,
+      "grad_norm": 0.6463441730942351,
+      "learning_rate": 8.991524878999202e-06,
+      "loss": 0.6855,
+      "step": 889
+    },
+    {
+      "epoch": 0.14616221542504054,
+      "grad_norm": 0.6296524118090603,
+      "learning_rate": 8.99149608564469e-06,
+      "loss": 0.6988,
+      "step": 890
+    },
+    {
+      "epoch": 0.1463264426333833,
+      "grad_norm": 0.6527468073618038,
+      "learning_rate": 8.991467243508111e-06,
+      "loss": 0.6997,
+      "step": 891
+    },
+    {
+      "epoch": 0.14649066984172604,
+      "grad_norm": 0.6337043319103454,
+      "learning_rate": 8.991438352589779e-06,
+      "loss": 0.6963,
+      "step": 892
+    },
+    {
+      "epoch": 0.14665489705006876,
+      "grad_norm": 0.6566657563406406,
+      "learning_rate": 8.991409412890008e-06,
+      "loss": 0.7111,
+      "step": 893
+    },
+    {
+      "epoch": 0.1468191242584115,
+      "grad_norm": 0.6858464246143825,
+      "learning_rate": 8.991380424409111e-06,
+      "loss": 0.7123,
+      "step": 894
+    },
+    {
+      "epoch": 0.14698335146675426,
+      "grad_norm": 0.6529635045203046,
+      "learning_rate": 8.991351387147404e-06,
+      "loss": 0.7247,
+      "step": 895
+    },
+    {
+      "epoch": 0.147147578675097,
+      "grad_norm": 0.6511476919346777,
+      "learning_rate": 8.991322301105202e-06,
+      "loss": 0.7022,
+      "step": 896
+    },
+    {
+      "epoch": 0.14731180588343973,
+      "grad_norm": 0.6335117395711469,
+      "learning_rate": 8.99129316628282e-06,
+      "loss": 0.6852,
+      "step": 897
+    },
+    {
+      "epoch": 0.14747603309178248,
+      "grad_norm": 0.6231917174435052,
+      "learning_rate": 8.991263982680576e-06,
+      "loss": 0.7049,
+      "step": 898
+    },
+    {
+      "epoch": 0.14764026030012523,
+      "grad_norm": 0.6470589103422413,
+      "learning_rate": 8.991234750298787e-06,
+      "loss": 0.7089,
+      "step": 899
+    },
+    {
+      "epoch": 0.14780448750846797,
+      "grad_norm": 0.6358772641933567,
+      "learning_rate": 8.99120546913777e-06,
+      "loss": 0.6738,
+      "step": 900
+    },
+    {
+      "epoch": 0.1479687147168107,
+      "grad_norm": 0.6221324653432067,
+      "learning_rate": 8.991176139197841e-06,
+      "loss": 0.7093,
+      "step": 901
+    },
+    {
+      "epoch": 0.14813294192515344,
+      "grad_norm": 0.6215558642133383,
+      "learning_rate": 8.991146760479323e-06,
+      "loss": 0.6978,
+      "step": 902
+    },
+    {
+      "epoch": 0.1482971691334962,
+      "grad_norm": 0.6263701610045573,
+      "learning_rate": 8.99111733298253e-06,
+      "loss": 0.7108,
+      "step": 903
+    },
+    {
+      "epoch": 0.14846139634183894,
+      "grad_norm": 0.6296632177864455,
+      "learning_rate": 8.991087856707785e-06,
+      "loss": 0.7151,
+      "step": 904
+    },
+    {
+      "epoch": 0.14862562355018166,
+      "grad_norm": 0.6178781456844099,
+      "learning_rate": 8.991058331655408e-06,
+      "loss": 0.6921,
+      "step": 905
+    },
+    {
+      "epoch": 0.1487898507585244,
+      "grad_norm": 0.6426241622321025,
+      "learning_rate": 8.991028757825718e-06,
+      "loss": 0.6661,
+      "step": 906
+    },
+    {
+      "epoch": 0.14895407796686716,
+      "grad_norm": 0.6315217002269438,
+      "learning_rate": 8.990999135219037e-06,
+      "loss": 0.7029,
+      "step": 907
+    },
+    {
+      "epoch": 0.1491183051752099,
+      "grad_norm": 0.6221289953237676,
+      "learning_rate": 8.990969463835688e-06,
+      "loss": 0.6854,
+      "step": 908
+    },
+    {
+      "epoch": 0.14928253238355266,
+      "grad_norm": 0.6351780842503867,
+      "learning_rate": 8.99093974367599e-06,
+      "loss": 0.6746,
+      "step": 909
+    },
+    {
+      "epoch": 0.14944675959189538,
+      "grad_norm": 0.6162843131557986,
+      "learning_rate": 8.990909974740271e-06,
+      "loss": 0.7063,
+      "step": 910
+    },
+    {
+      "epoch": 0.14961098680023813,
+      "grad_norm": 0.6438716140978041,
+      "learning_rate": 8.990880157028849e-06,
+      "loss": 0.668,
+      "step": 911
+    },
+    {
+      "epoch": 0.14977521400858088,
+      "grad_norm": 0.6738550733623182,
+      "learning_rate": 8.99085029054205e-06,
+      "loss": 0.6723,
+      "step": 912
+    },
+    {
+      "epoch": 0.14993944121692362,
+      "grad_norm": 0.6236064060080636,
+      "learning_rate": 8.990820375280199e-06,
+      "loss": 0.6933,
+      "step": 913
+    },
+    {
+      "epoch": 0.15010366842526635,
+      "grad_norm": 0.8878180053952123,
+      "learning_rate": 8.990790411243622e-06,
+      "loss": 0.7072,
+      "step": 914
+    },
+    {
+      "epoch": 0.1502678956336091,
+      "grad_norm": 0.6040213602808409,
+      "learning_rate": 8.99076039843264e-06,
+      "loss": 0.7265,
+      "step": 915
+    },
+    {
+      "epoch": 0.15043212284195184,
+      "grad_norm": 0.5993272996617645,
+      "learning_rate": 8.990730336847584e-06,
+      "loss": 0.6735,
+      "step": 916
+    },
+    {
+      "epoch": 0.1505963500502946,
+      "grad_norm": 0.6196667818949094,
+      "learning_rate": 8.990700226488775e-06,
+      "loss": 0.683,
+      "step": 917
+    },
+    {
+      "epoch": 0.1507605772586373,
+      "grad_norm": 0.611151140998212,
+      "learning_rate": 8.990670067356546e-06,
+      "loss": 0.6855,
+      "step": 918
+    },
+    {
+      "epoch": 0.15092480446698006,
+      "grad_norm": 0.6124554833223387,
+      "learning_rate": 8.990639859451223e-06,
+      "loss": 0.7316,
+      "step": 919
+    },
+    {
+      "epoch": 0.1510890316753228,
+      "grad_norm": 0.8671036907456195,
+      "learning_rate": 8.99060960277313e-06,
+      "loss": 0.6754,
+      "step": 920
+    },
+    {
+      "epoch": 0.15125325888366556,
+      "grad_norm": 0.604961084181521,
+      "learning_rate": 8.9905792973226e-06,
+      "loss": 0.6957,
+      "step": 921
+    },
+    {
+      "epoch": 0.15141748609200828,
+      "grad_norm": 0.6367685614450236,
+      "learning_rate": 8.99054894309996e-06,
+      "loss": 0.6878,
+      "step": 922
+    },
+    {
+      "epoch": 0.15158171330035103,
+      "grad_norm": 0.6291160157774949,
+      "learning_rate": 8.99051854010554e-06,
+      "loss": 0.6758,
+      "step": 923
+    },
+    {
+      "epoch": 0.15174594050869378,
+      "grad_norm": 0.6186538967618891,
+      "learning_rate": 8.990488088339673e-06,
+      "loss": 0.6851,
+      "step": 924
+    },
+    {
+      "epoch": 0.15191016771703653,
+      "grad_norm": 0.6043423314180405,
+      "learning_rate": 8.990457587802685e-06,
+      "loss": 0.7029,
+      "step": 925
+    },
+    {
+      "epoch": 0.15207439492537927,
+      "grad_norm": 0.6057826193851099,
+      "learning_rate": 8.99042703849491e-06,
+      "loss": 0.6894,
+      "step": 926
+    },
+    {
+      "epoch": 0.152238622133722,
+      "grad_norm": 0.6415804974916999,
+      "learning_rate": 8.990396440416682e-06,
+      "loss": 0.6895,
+      "step": 927
+    },
+    {
+      "epoch": 0.15240284934206474,
+      "grad_norm": 0.590171420782872,
+      "learning_rate": 8.990365793568326e-06,
+      "loss": 0.7064,
+      "step": 928
+    },
+    {
+      "epoch": 0.1525670765504075,
+      "grad_norm": 0.5991023181384277,
+      "learning_rate": 8.990335097950184e-06,
+      "loss": 0.7034,
+      "step": 929
+    },
+    {
+      "epoch": 0.15273130375875024,
+      "grad_norm": 0.5925124030029137,
+      "learning_rate": 8.990304353562582e-06,
+      "loss": 0.6722,
+      "step": 930
+    },
+    {
+      "epoch": 0.15289553096709296,
+      "grad_norm": 0.5879098836185569,
+      "learning_rate": 8.990273560405858e-06,
+      "loss": 0.6778,
+      "step": 931
+    },
+    {
+      "epoch": 0.1530597581754357,
+      "grad_norm": 0.6081890261292409,
+      "learning_rate": 8.990242718480345e-06,
+      "loss": 0.6977,
+      "step": 932
+    },
+    {
+      "epoch": 0.15322398538377846,
+      "grad_norm": 0.5943707439558413,
+      "learning_rate": 8.990211827786377e-06,
+      "loss": 0.6833,
+      "step": 933
+    },
+    {
+      "epoch": 0.1533882125921212,
+      "grad_norm": 0.5957073787044868,
+      "learning_rate": 8.990180888324293e-06,
+      "loss": 0.6769,
+      "step": 934
+    },
+    {
+      "epoch": 0.15355243980046393,
+      "grad_norm": 0.5990803867181086,
+      "learning_rate": 8.990149900094426e-06,
+      "loss": 0.6947,
+      "step": 935
+    },
+    {
+      "epoch": 0.15371666700880668,
+      "grad_norm": 0.5762901620148804,
+      "learning_rate": 8.990118863097113e-06,
+      "loss": 0.6989,
+      "step": 936
+    },
+    {
+      "epoch": 0.15388089421714943,
+      "grad_norm": 0.583179708719448,
+      "learning_rate": 8.990087777332693e-06,
+      "loss": 0.7021,
+      "step": 937
+    },
+    {
+      "epoch": 0.15404512142549218,
+      "grad_norm": 0.5957217968883289,
+      "learning_rate": 8.9900566428015e-06,
+      "loss": 0.6581,
+      "step": 938
+    },
+    {
+      "epoch": 0.1542093486338349,
+      "grad_norm": 0.5768688535477488,
+      "learning_rate": 8.990025459503875e-06,
+      "loss": 0.6909,
+      "step": 939
+    },
+    {
+      "epoch": 0.15437357584217765,
+      "grad_norm": 0.5863263337089208,
+      "learning_rate": 8.989994227440156e-06,
+      "loss": 0.6903,
+      "step": 940
+    },
+    {
+      "epoch": 0.1545378030505204,
+      "grad_norm": 1.449174264009128,
+      "learning_rate": 8.989962946610682e-06,
+      "loss": 0.6777,
+      "step": 941
+    },
+    {
+      "epoch": 0.15470203025886314,
+      "grad_norm": 0.5857095155638259,
+      "learning_rate": 8.989931617015794e-06,
+      "loss": 0.6847,
+      "step": 942
+    },
+    {
+      "epoch": 0.1548662574672059,
+      "grad_norm": 0.6045629484192497,
+      "learning_rate": 8.98990023865583e-06,
+      "loss": 0.6973,
+      "step": 943
+    },
+    {
+      "epoch": 0.1550304846755486,
+      "grad_norm": 0.6271020723122404,
+      "learning_rate": 8.989868811531133e-06,
+      "loss": 0.6822,
+      "step": 944
+    },
+    {
+      "epoch": 0.15519471188389136,
+      "grad_norm": 0.5974097511017354,
+      "learning_rate": 8.989837335642041e-06,
+      "loss": 0.6932,
+      "step": 945
+    },
+    {
+      "epoch": 0.1553589390922341,
+      "grad_norm": 0.6165658939960453,
+      "learning_rate": 8.989805810988899e-06,
+      "loss": 0.7028,
+      "step": 946
+    },
+    {
+      "epoch": 0.15552316630057686,
+      "grad_norm": 0.5750032115880928,
+      "learning_rate": 8.98977423757205e-06,
+      "loss": 0.699,
+      "step": 947
+    },
+    {
+      "epoch": 0.15568739350891958,
+      "grad_norm": 0.5930114131715402,
+      "learning_rate": 8.989742615391835e-06,
+      "loss": 0.6946,
+      "step": 948
+    },
+    {
+      "epoch": 0.15585162071726233,
+      "grad_norm": 0.5821912545770899,
+      "learning_rate": 8.989710944448598e-06,
+      "loss": 0.6944,
+      "step": 949
+    },
+    {
+      "epoch": 0.15601584792560508,
+      "grad_norm": 0.5998707680711169,
+      "learning_rate": 8.989679224742682e-06,
+      "loss": 0.6817,
+      "step": 950
+    },
+    {
+      "epoch": 0.15618007513394783,
+      "grad_norm": 0.784960727758064,
+      "learning_rate": 8.989647456274432e-06,
+      "loss": 0.6667,
+      "step": 951
+    },
+    {
+      "epoch": 0.15634430234229055,
+      "grad_norm": 0.5905550333586939,
+      "learning_rate": 8.989615639044194e-06,
+      "loss": 0.6909,
+      "step": 952
+    },
+    {
+      "epoch": 0.1565085295506333,
+      "grad_norm": 0.5813775463869184,
+      "learning_rate": 8.989583773052312e-06,
+      "loss": 0.6741,
+      "step": 953
+    },
+    {
+      "epoch": 0.15667275675897605,
+      "grad_norm": 0.5859001004681035,
+      "learning_rate": 8.989551858299135e-06,
+      "loss": 0.675,
+      "step": 954
+    },
+    {
+      "epoch": 0.1568369839673188,
+      "grad_norm": 0.5789497437415951,
+      "learning_rate": 8.989519894785007e-06,
+      "loss": 0.6924,
+      "step": 955
+    },
+    {
+      "epoch": 0.15700121117566151,
+      "grad_norm": 0.576924083966074,
+      "learning_rate": 8.989487882510275e-06,
+      "loss": 0.6951,
+      "step": 956
+    },
+    {
+      "epoch": 0.15716543838400426,
+      "grad_norm": 0.5860897339311051,
+      "learning_rate": 8.98945582147529e-06,
+      "loss": 0.6837,
+      "step": 957
+    },
+    {
+      "epoch": 0.157329665592347,
+      "grad_norm": 0.567962554350338,
+      "learning_rate": 8.989423711680394e-06,
+      "loss": 0.6718,
+      "step": 958
+    },
+    {
+      "epoch": 0.15749389280068976,
+      "grad_norm": 0.580251872407634,
+      "learning_rate": 8.989391553125943e-06,
+      "loss": 0.6766,
+      "step": 959
+    },
+    {
+      "epoch": 0.1576581200090325,
+      "grad_norm": 0.5835185656150728,
+      "learning_rate": 8.98935934581228e-06,
+      "loss": 0.6656,
+      "step": 960
+    },
+    {
+      "epoch": 0.15782234721737523,
+      "grad_norm": 0.5680557698537251,
+      "learning_rate": 8.989327089739759e-06,
+      "loss": 0.6886,
+      "step": 961
+    },
+    {
+      "epoch": 0.15798657442571798,
+      "grad_norm": 0.5822249108751449,
+      "learning_rate": 8.98929478490873e-06,
+      "loss": 0.6793,
+      "step": 962
+    },
+    {
+      "epoch": 0.15815080163406073,
+      "grad_norm": 0.6920777043397142,
+      "learning_rate": 8.989262431319541e-06,
+      "loss": 0.68,
+      "step": 963
+    },
+    {
+      "epoch": 0.15831502884240348,
+      "grad_norm": 0.5697926374238244,
+      "learning_rate": 8.989230028972546e-06,
+      "loss": 0.6902,
+      "step": 964
+    },
+    {
+      "epoch": 0.1584792560507462,
+      "grad_norm": 0.5648702262985533,
+      "learning_rate": 8.989197577868095e-06,
+      "loss": 0.6785,
+      "step": 965
+    },
+    {
+      "epoch": 0.15864348325908895,
+      "grad_norm": 0.574611072826276,
+      "learning_rate": 8.989165078006542e-06,
+      "loss": 0.702,
+      "step": 966
+    },
+    {
+      "epoch": 0.1588077104674317,
+      "grad_norm": 0.5653515769606229,
+      "learning_rate": 8.989132529388239e-06,
+      "loss": 0.6805,
+      "step": 967
+    },
+    {
+      "epoch": 0.15897193767577444,
+      "grad_norm": 0.5677479447859182,
+      "learning_rate": 8.989099932013541e-06,
+      "loss": 0.6899,
+      "step": 968
+    },
+    {
+      "epoch": 0.15913616488411716,
+      "grad_norm": 0.5600255179636507,
+      "learning_rate": 8.989067285882801e-06,
+      "loss": 0.6952,
+      "step": 969
+    },
+    {
+      "epoch": 0.1593003920924599,
+      "grad_norm": 0.5645642174349392,
+      "learning_rate": 8.989034590996373e-06,
+      "loss": 0.6797,
+      "step": 970
+    },
+    {
+      "epoch": 0.15946461930080266,
+      "grad_norm": 0.5661098243057704,
+      "learning_rate": 8.989001847354614e-06,
+      "loss": 0.6769,
+      "step": 971
+    },
+    {
+      "epoch": 0.1596288465091454,
+      "grad_norm": 0.5608033372773416,
+      "learning_rate": 8.988969054957878e-06,
+      "loss": 0.6786,
+      "step": 972
+    },
+    {
+      "epoch": 0.15979307371748813,
+      "grad_norm": 0.5780246426412402,
+      "learning_rate": 8.98893621380652e-06,
+      "loss": 0.6693,
+      "step": 973
+    },
+    {
+      "epoch": 0.15995730092583088,
+      "grad_norm": 0.5786252288148208,
+      "learning_rate": 8.9889033239009e-06,
+      "loss": 0.6862,
+      "step": 974
+    },
+    {
+      "epoch": 0.16012152813417363,
+      "grad_norm": 0.7542693728723142,
+      "learning_rate": 8.988870385241371e-06,
+      "loss": 0.6678,
+      "step": 975
+    },
+    {
+      "epoch": 0.16028575534251638,
+      "grad_norm": 0.6306239933124853,
+      "learning_rate": 8.988837397828296e-06,
+      "loss": 0.6757,
+      "step": 976
+    },
+    {
+      "epoch": 0.16044998255085913,
+      "grad_norm": 0.5939312755402465,
+      "learning_rate": 8.988804361662029e-06,
+      "loss": 0.6713,
+      "step": 977
+    },
+    {
+      "epoch": 0.16061420975920185,
+      "grad_norm": 0.5633583256765683,
+      "learning_rate": 8.98877127674293e-06,
+      "loss": 0.6673,
+      "step": 978
+    },
+    {
+      "epoch": 0.1607784369675446,
+      "grad_norm": 0.5800987462350472,
+      "learning_rate": 8.988738143071359e-06,
+      "loss": 0.6605,
+      "step": 979
+    },
+    {
+      "epoch": 0.16094266417588735,
+      "grad_norm": 0.5643582604641704,
+      "learning_rate": 8.988704960647677e-06,
+      "loss": 0.7013,
+      "step": 980
+    },
+    {
+      "epoch": 0.1611068913842301,
+      "grad_norm": 0.5835933494518621,
+      "learning_rate": 8.988671729472241e-06,
+      "loss": 0.6782,
+      "step": 981
+    },
+    {
+      "epoch": 0.16127111859257282,
+      "grad_norm": 0.6788940707481684,
+      "learning_rate": 8.988638449545415e-06,
+      "loss": 0.6844,
+      "step": 982
+    },
+    {
+      "epoch": 0.16143534580091556,
+      "grad_norm": 0.5471636693606157,
+      "learning_rate": 8.988605120867557e-06,
+      "loss": 0.6786,
+      "step": 983
+    },
+    {
+      "epoch": 0.1615995730092583,
+      "grad_norm": 0.5734962283482052,
+      "learning_rate": 8.988571743439032e-06,
+      "loss": 0.6819,
+      "step": 984
+    },
+    {
+      "epoch": 0.16176380021760106,
+      "grad_norm": 0.6161812954255436,
+      "learning_rate": 8.988538317260203e-06,
+      "loss": 0.7011,
+      "step": 985
+    },
+    {
+      "epoch": 0.16192802742594378,
+      "grad_norm": 0.5586729255122475,
+      "learning_rate": 8.988504842331431e-06,
+      "loss": 0.6924,
+      "step": 986
+    },
+    {
+      "epoch": 0.16209225463428653,
+      "grad_norm": 0.5589443674472094,
+      "learning_rate": 8.988471318653081e-06,
+      "loss": 0.6688,
+      "step": 987
+    },
+    {
+      "epoch": 0.16225648184262928,
+      "grad_norm": 0.5725349388200487,
+      "learning_rate": 8.988437746225515e-06,
+      "loss": 0.6859,
+      "step": 988
+    },
+    {
+      "epoch": 0.16242070905097203,
+      "grad_norm": 0.5386632021604499,
+      "learning_rate": 8.9884041250491e-06,
+      "loss": 0.7098,
+      "step": 989
+    },
+    {
+      "epoch": 0.16258493625931475,
+      "grad_norm": 0.5588338765541392,
+      "learning_rate": 8.9883704551242e-06,
+      "loss": 0.6869,
+      "step": 990
+    },
+    {
+      "epoch": 0.1627491634676575,
+      "grad_norm": 0.5620172479210048,
+      "learning_rate": 8.98833673645118e-06,
+      "loss": 0.6801,
+      "step": 991
+    },
+    {
+      "epoch": 0.16291339067600025,
+      "grad_norm": 0.578537129368564,
+      "learning_rate": 8.988302969030409e-06,
+      "loss": 0.6802,
+      "step": 992
+    },
+    {
+      "epoch": 0.163077617884343,
+      "grad_norm": 0.5788833235484687,
+      "learning_rate": 8.98826915286225e-06,
+      "loss": 0.6838,
+      "step": 993
+    },
+    {
+      "epoch": 0.16324184509268574,
+      "grad_norm": 0.5455036420454441,
+      "learning_rate": 8.988235287947074e-06,
+      "loss": 0.684,
+      "step": 994
+    },
+    {
+      "epoch": 0.16340607230102847,
+      "grad_norm": 0.5469684799799748,
+      "learning_rate": 8.988201374285246e-06,
+      "loss": 0.6732,
+      "step": 995
+    },
+    {
+      "epoch": 0.16357029950937121,
+      "grad_norm": 0.5659737004114193,
+      "learning_rate": 8.988167411877134e-06,
+      "loss": 0.6774,
+      "step": 996
+    },
+    {
+      "epoch": 0.16373452671771396,
+      "grad_norm": 0.5472964921114717,
+      "learning_rate": 8.98813340072311e-06,
+      "loss": 0.6561,
+      "step": 997
+    },
+    {
+      "epoch": 0.1638987539260567,
+      "grad_norm": 0.5313159381480278,
+      "learning_rate": 8.98809934082354e-06,
+      "loss": 0.6924,
+      "step": 998
+    },
+    {
+      "epoch": 0.16406298113439943,
+      "grad_norm": 0.5792443966133527,
+      "learning_rate": 8.988065232178799e-06,
+      "loss": 0.6711,
+      "step": 999
+    },
+    {
+      "epoch": 0.16422720834274218,
+      "grad_norm": 0.5545515494905078,
+      "learning_rate": 8.98803107478925e-06,
+      "loss": 0.6831,
+      "step": 1000
+    },
+    {
+      "epoch": 0.16439143555108493,
+      "grad_norm": 0.5573621935545905,
+      "learning_rate": 8.98799686865527e-06,
+      "loss": 0.6991,
+      "step": 1001
+    },
+    {
+      "epoch": 0.16455566275942768,
+      "grad_norm": 0.5491556801369158,
+      "learning_rate": 8.987962613777226e-06,
+      "loss": 0.6638,
+      "step": 1002
+    },
+    {
+      "epoch": 0.1647198899677704,
+      "grad_norm": 0.5392445583149836,
+      "learning_rate": 8.987928310155495e-06,
+      "loss": 0.6652,
+      "step": 1003
+    },
+    {
+      "epoch": 0.16488411717611315,
+      "grad_norm": 0.5348090900057155,
+      "learning_rate": 8.987893957790447e-06,
+      "loss": 0.6828,
+      "step": 1004
+    },
+    {
+      "epoch": 0.1650483443844559,
+      "grad_norm": 0.5577106888054703,
+      "learning_rate": 8.987859556682454e-06,
+      "loss": 0.6719,
+      "step": 1005
+    },
+    {
+      "epoch": 0.16521257159279865,
+      "grad_norm": 0.5363067993234555,
+      "learning_rate": 8.98782510683189e-06,
+      "loss": 0.6803,
+      "step": 1006
+    },
+    {
+      "epoch": 0.16537679880114137,
+      "grad_norm": 0.5603737925264757,
+      "learning_rate": 8.987790608239131e-06,
+      "loss": 0.6825,
+      "step": 1007
+    },
+    {
+      "epoch": 0.16554102600948412,
+      "grad_norm": 0.5411605815062192,
+      "learning_rate": 8.98775606090455e-06,
+      "loss": 0.6625,
+      "step": 1008
+    },
+    {
+      "epoch": 0.16570525321782686,
+      "grad_norm": 0.5489945681101899,
+      "learning_rate": 8.987721464828525e-06,
+      "loss": 0.6695,
+      "step": 1009
+    },
+    {
+      "epoch": 0.1658694804261696,
+      "grad_norm": 0.54397334030031,
+      "learning_rate": 8.987686820011428e-06,
+      "loss": 0.6667,
+      "step": 1010
+    },
+    {
+      "epoch": 0.16603370763451233,
+      "grad_norm": 0.5321484388094383,
+      "learning_rate": 8.987652126453636e-06,
+      "loss": 0.691,
+      "step": 1011
+    },
+    {
+      "epoch": 0.16619793484285508,
+      "grad_norm": 0.5317184209836602,
+      "learning_rate": 8.987617384155527e-06,
+      "loss": 0.6477,
+      "step": 1012
+    },
+    {
+      "epoch": 0.16636216205119783,
+      "grad_norm": 0.5341564431442047,
+      "learning_rate": 8.987582593117478e-06,
+      "loss": 0.6473,
+      "step": 1013
+    },
+    {
+      "epoch": 0.16652638925954058,
+      "grad_norm": 0.5787953947276587,
+      "learning_rate": 8.987547753339868e-06,
+      "loss": 0.6788,
+      "step": 1014
+    },
+    {
+      "epoch": 0.16669061646788333,
+      "grad_norm": 0.5338083228144461,
+      "learning_rate": 8.987512864823073e-06,
+      "loss": 0.6783,
+      "step": 1015
+    },
+    {
+      "epoch": 0.16685484367622605,
+      "grad_norm": 0.5431111870614626,
+      "learning_rate": 8.987477927567475e-06,
+      "loss": 0.6665,
+      "step": 1016
+    },
+    {
+      "epoch": 0.1670190708845688,
+      "grad_norm": 0.5460094365072707,
+      "learning_rate": 8.987442941573448e-06,
+      "loss": 0.665,
+      "step": 1017
+    },
+    {
+      "epoch": 0.16718329809291155,
+      "grad_norm": 0.5427687593109538,
+      "learning_rate": 8.98740790684138e-06,
+      "loss": 0.6562,
+      "step": 1018
+    },
+    {
+      "epoch": 0.1673475253012543,
+      "grad_norm": 0.533396167621092,
+      "learning_rate": 8.987372823371644e-06,
+      "loss": 0.6684,
+      "step": 1019
+    },
+    {
+      "epoch": 0.16751175250959702,
+      "grad_norm": 0.5320714216988802,
+      "learning_rate": 8.987337691164625e-06,
+      "loss": 0.6519,
+      "step": 1020
+    },
+    {
+      "epoch": 0.16767597971793977,
+      "grad_norm": 0.5456380984656337,
+      "learning_rate": 8.987302510220704e-06,
+      "loss": 0.686,
+      "step": 1021
+    },
+    {
+      "epoch": 0.16784020692628251,
+      "grad_norm": 0.5224293963323513,
+      "learning_rate": 8.987267280540264e-06,
+      "loss": 0.6697,
+      "step": 1022
+    },
+    {
+      "epoch": 0.16800443413462526,
+      "grad_norm": 0.5243433616305909,
+      "learning_rate": 8.987232002123685e-06,
+      "loss": 0.6921,
+      "step": 1023
+    },
+    {
+      "epoch": 0.16816866134296798,
+      "grad_norm": 0.5328025582851148,
+      "learning_rate": 8.987196674971352e-06,
+      "loss": 0.6555,
+      "step": 1024
+    },
+    {
+      "epoch": 0.16833288855131073,
+      "grad_norm": 0.5198078640352478,
+      "learning_rate": 8.987161299083647e-06,
+      "loss": 0.6379,
+      "step": 1025
+    },
+    {
+      "epoch": 0.16849711575965348,
+      "grad_norm": 0.5882832847154114,
+      "learning_rate": 8.987125874460957e-06,
+      "loss": 0.6701,
+      "step": 1026
+    },
+    {
+      "epoch": 0.16866134296799623,
+      "grad_norm": 0.5295752447874579,
+      "learning_rate": 8.987090401103665e-06,
+      "loss": 0.6944,
+      "step": 1027
+    },
+    {
+      "epoch": 0.16882557017633895,
+      "grad_norm": 0.5326690138378312,
+      "learning_rate": 8.987054879012156e-06,
+      "loss": 0.6812,
+      "step": 1028
+    },
+    {
+      "epoch": 0.1689897973846817,
+      "grad_norm": 0.5279524083193049,
+      "learning_rate": 8.987019308186818e-06,
+      "loss": 0.664,
+      "step": 1029
+    },
+    {
+      "epoch": 0.16915402459302445,
+      "grad_norm": 0.5696406663324558,
+      "learning_rate": 8.986983688628034e-06,
+      "loss": 0.6615,
+      "step": 1030
+    },
+    {
+      "epoch": 0.1693182518013672,
+      "grad_norm": 0.5237891547003628,
+      "learning_rate": 8.986948020336192e-06,
+      "loss": 0.6643,
+      "step": 1031
+    },
+    {
+      "epoch": 0.16948247900970995,
+      "grad_norm": 0.5398581475126638,
+      "learning_rate": 8.986912303311682e-06,
+      "loss": 0.6935,
+      "step": 1032
+    },
+    {
+      "epoch": 0.16964670621805267,
+      "grad_norm": 0.5257870387792046,
+      "learning_rate": 8.986876537554889e-06,
+      "loss": 0.6688,
+      "step": 1033
+    },
+    {
+      "epoch": 0.16981093342639542,
+      "grad_norm": 0.5358931293606296,
+      "learning_rate": 8.986840723066202e-06,
+      "loss": 0.6756,
+      "step": 1034
+    },
+    {
+      "epoch": 0.16997516063473816,
+      "grad_norm": 0.5429766773598853,
+      "learning_rate": 8.98680485984601e-06,
+      "loss": 0.6706,
+      "step": 1035
+    },
+    {
+      "epoch": 0.1701393878430809,
+      "grad_norm": 0.5115483229740615,
+      "learning_rate": 8.986768947894704e-06,
+      "loss": 0.665,
+      "step": 1036
+    },
+    {
+      "epoch": 0.17030361505142363,
+      "grad_norm": 0.5223361541452932,
+      "learning_rate": 8.98673298721267e-06,
+      "loss": 0.6848,
+      "step": 1037
+    },
+    {
+      "epoch": 0.17046784225976638,
+      "grad_norm": 0.8292942856556398,
+      "learning_rate": 8.986696977800305e-06,
+      "loss": 0.6611,
+      "step": 1038
+    },
+    {
+      "epoch": 0.17063206946810913,
+      "grad_norm": 0.5594058489983992,
+      "learning_rate": 8.986660919657995e-06,
+      "loss": 0.6699,
+      "step": 1039
+    },
+    {
+      "epoch": 0.17079629667645188,
+      "grad_norm": 0.5191782674532504,
+      "learning_rate": 8.986624812786133e-06,
+      "loss": 0.6463,
+      "step": 1040
+    },
+    {
+      "epoch": 0.1709605238847946,
+      "grad_norm": 0.5065778655698908,
+      "learning_rate": 8.986588657185112e-06,
+      "loss": 0.6804,
+      "step": 1041
+    },
+    {
+      "epoch": 0.17112475109313735,
+      "grad_norm": 0.5140861111946393,
+      "learning_rate": 8.986552452855323e-06,
+      "loss": 0.6425,
+      "step": 1042
+    },
+    {
+      "epoch": 0.1712889783014801,
+      "grad_norm": 0.5472505082789985,
+      "learning_rate": 8.98651619979716e-06,
+      "loss": 0.654,
+      "step": 1043
+    },
+    {
+      "epoch": 0.17145320550982285,
+      "grad_norm": 0.5101925390931323,
+      "learning_rate": 8.986479898011019e-06,
+      "loss": 0.685,
+      "step": 1044
+    },
+    {
+      "epoch": 0.17161743271816557,
+      "grad_norm": 0.5206702278219933,
+      "learning_rate": 8.98644354749729e-06,
+      "loss": 0.6693,
+      "step": 1045
+    },
+    {
+      "epoch": 0.17178165992650832,
+      "grad_norm": 0.5126143393833118,
+      "learning_rate": 8.986407148256372e-06,
+      "loss": 0.6799,
+      "step": 1046
+    },
+    {
+      "epoch": 0.17194588713485107,
+      "grad_norm": 0.5331986314895121,
+      "learning_rate": 8.986370700288655e-06,
+      "loss": 0.6638,
+      "step": 1047
+    },
+    {
+      "epoch": 0.17211011434319382,
+      "grad_norm": 0.5073509933218882,
+      "learning_rate": 8.98633420359454e-06,
+      "loss": 0.6802,
+      "step": 1048
+    },
+    {
+      "epoch": 0.17227434155153656,
+      "grad_norm": 0.508473575804819,
+      "learning_rate": 8.986297658174423e-06,
+      "loss": 0.6571,
+      "step": 1049
+    },
+    {
+      "epoch": 0.17243856875987928,
+      "grad_norm": 0.49990662654253826,
+      "learning_rate": 8.986261064028698e-06,
+      "loss": 0.668,
+      "step": 1050
+    },
+    {
+      "epoch": 0.17260279596822203,
+      "grad_norm": 0.5180145085098473,
+      "learning_rate": 8.986224421157764e-06,
+      "loss": 0.6652,
+      "step": 1051
+    },
+    {
+      "epoch": 0.17276702317656478,
+      "grad_norm": 0.5008280735732593,
+      "learning_rate": 8.98618772956202e-06,
+      "loss": 0.6635,
+      "step": 1052
+    },
+    {
+      "epoch": 0.17293125038490753,
+      "grad_norm": 0.525862217902218,
+      "learning_rate": 8.986150989241863e-06,
+      "loss": 0.6604,
+      "step": 1053
+    },
+    {
+      "epoch": 0.17309547759325025,
+      "grad_norm": 0.5041642087839215,
+      "learning_rate": 8.986114200197692e-06,
+      "loss": 0.6637,
+      "step": 1054
+    },
+    {
+      "epoch": 0.173259704801593,
+      "grad_norm": 0.5755431185436396,
+      "learning_rate": 8.986077362429908e-06,
+      "loss": 0.6724,
+      "step": 1055
+    },
+    {
+      "epoch": 0.17342393200993575,
+      "grad_norm": 0.5355683129486976,
+      "learning_rate": 8.986040475938908e-06,
+      "loss": 0.6549,
+      "step": 1056
+    },
+    {
+      "epoch": 0.1735881592182785,
+      "grad_norm": 0.5041556885697243,
+      "learning_rate": 8.986003540725098e-06,
+      "loss": 0.6591,
+      "step": 1057
+    },
+    {
+      "epoch": 0.17375238642662122,
+      "grad_norm": 0.511207002198651,
+      "learning_rate": 8.985966556788873e-06,
+      "loss": 0.6426,
+      "step": 1058
+    },
+    {
+      "epoch": 0.17391661363496397,
+      "grad_norm": 0.49542003624935754,
+      "learning_rate": 8.985929524130638e-06,
+      "loss": 0.6586,
+      "step": 1059
+    },
+    {
+      "epoch": 0.17408084084330672,
+      "grad_norm": 0.5639426152691353,
+      "learning_rate": 8.985892442750796e-06,
+      "loss": 0.6463,
+      "step": 1060
+    },
+    {
+      "epoch": 0.17424506805164947,
+      "grad_norm": 0.5156251092751705,
+      "learning_rate": 8.985855312649749e-06,
+      "loss": 0.701,
+      "step": 1061
+    },
+    {
+      "epoch": 0.1744092952599922,
+      "grad_norm": 0.5004678577815013,
+      "learning_rate": 8.985818133827898e-06,
+      "loss": 0.6876,
+      "step": 1062
+    },
+    {
+      "epoch": 0.17457352246833494,
+      "grad_norm": 0.5368872269813518,
+      "learning_rate": 8.98578090628565e-06,
+      "loss": 0.6573,
+      "step": 1063
+    },
+    {
+      "epoch": 0.17473774967667768,
+      "grad_norm": 0.5012298661884161,
+      "learning_rate": 8.985743630023406e-06,
+      "loss": 0.675,
+      "step": 1064
+    },
+    {
+      "epoch": 0.17490197688502043,
+      "grad_norm": 0.586654914508436,
+      "learning_rate": 8.985706305041575e-06,
+      "loss": 0.6746,
+      "step": 1065
+    },
+    {
+      "epoch": 0.17506620409336318,
+      "grad_norm": 0.5265854279573517,
+      "learning_rate": 8.98566893134056e-06,
+      "loss": 0.6883,
+      "step": 1066
+    },
+    {
+      "epoch": 0.1752304313017059,
+      "grad_norm": 0.49981491800267486,
+      "learning_rate": 8.985631508920767e-06,
+      "loss": 0.6485,
+      "step": 1067
+    },
+    {
+      "epoch": 0.17539465851004865,
+      "grad_norm": 0.4938509609458083,
+      "learning_rate": 8.985594037782602e-06,
+      "loss": 0.6846,
+      "step": 1068
+    },
+    {
+      "epoch": 0.1755588857183914,
+      "grad_norm": 0.5031409162616062,
+      "learning_rate": 8.985556517926472e-06,
+      "loss": 0.6357,
+      "step": 1069
+    },
+    {
+      "epoch": 0.17572311292673415,
+      "grad_norm": 0.5700620669276998,
+      "learning_rate": 8.985518949352786e-06,
+      "loss": 0.6849,
+      "step": 1070
+    },
+    {
+      "epoch": 0.17588734013507687,
+      "grad_norm": 0.5145274475647432,
+      "learning_rate": 8.98548133206195e-06,
+      "loss": 0.6694,
+      "step": 1071
+    },
+    {
+      "epoch": 0.17605156734341962,
+      "grad_norm": 0.5063946464405173,
+      "learning_rate": 8.985443666054375e-06,
+      "loss": 0.678,
+      "step": 1072
+    },
+    {
+      "epoch": 0.17621579455176237,
+      "grad_norm": 0.4916761388400217,
+      "learning_rate": 8.985405951330468e-06,
+      "loss": 0.6829,
+      "step": 1073
+    },
+    {
+      "epoch": 0.17638002176010512,
+      "grad_norm": 0.5289628393025868,
+      "learning_rate": 8.98536818789064e-06,
+      "loss": 0.6685,
+      "step": 1074
+    },
+    {
+      "epoch": 0.17654424896844784,
+      "grad_norm": 0.5643139132504469,
+      "learning_rate": 8.985330375735298e-06,
+      "loss": 0.6627,
+      "step": 1075
+    },
+    {
+      "epoch": 0.17670847617679059,
+      "grad_norm": 0.4814303227261524,
+      "learning_rate": 8.985292514864859e-06,
+      "loss": 0.6559,
+      "step": 1076
+    },
+    {
+      "epoch": 0.17687270338513333,
+      "grad_norm": 0.4974900220408461,
+      "learning_rate": 8.985254605279726e-06,
+      "loss": 0.6716,
+      "step": 1077
+    },
+    {
+      "epoch": 0.17703693059347608,
+      "grad_norm": 0.48748470280103556,
+      "learning_rate": 8.985216646980318e-06,
+      "loss": 0.6554,
+      "step": 1078
+    },
+    {
+      "epoch": 0.1772011578018188,
+      "grad_norm": 0.476700155245185,
+      "learning_rate": 8.985178639967044e-06,
+      "loss": 0.6633,
+      "step": 1079
+    },
+    {
+      "epoch": 0.17736538501016155,
+      "grad_norm": 0.48611980044750863,
+      "learning_rate": 8.985140584240317e-06,
+      "loss": 0.6477,
+      "step": 1080
+    },
+    {
+      "epoch": 0.1775296122185043,
+      "grad_norm": 0.4744923322444135,
+      "learning_rate": 8.985102479800551e-06,
+      "loss": 0.6889,
+      "step": 1081
+    },
+    {
+      "epoch": 0.17769383942684705,
+      "grad_norm": 0.492363316728015,
+      "learning_rate": 8.985064326648157e-06,
+      "loss": 0.6646,
+      "step": 1082
+    },
+    {
+      "epoch": 0.1778580666351898,
+      "grad_norm": 0.5034362431677137,
+      "learning_rate": 8.985026124783554e-06,
+      "loss": 0.6495,
+      "step": 1083
+    },
+    {
+      "epoch": 0.17802229384353252,
+      "grad_norm": 0.5305230978210335,
+      "learning_rate": 8.984987874207156e-06,
+      "loss": 0.6578,
+      "step": 1084
+    },
+    {
+      "epoch": 0.17818652105187527,
+      "grad_norm": 0.49163982241906173,
+      "learning_rate": 8.984949574919374e-06,
+      "loss": 0.67,
+      "step": 1085
+    },
+    {
+      "epoch": 0.17835074826021802,
+      "grad_norm": 0.48484339846155095,
+      "learning_rate": 8.984911226920629e-06,
+      "loss": 0.6795,
+      "step": 1086
+    },
+    {
+      "epoch": 0.17851497546856077,
+      "grad_norm": 0.4746707284972163,
+      "learning_rate": 8.984872830211335e-06,
+      "loss": 0.6693,
+      "step": 1087
+    },
+    {
+      "epoch": 0.1786792026769035,
+      "grad_norm": 0.4815131650072882,
+      "learning_rate": 8.984834384791908e-06,
+      "loss": 0.6627,
+      "step": 1088
+    },
+    {
+      "epoch": 0.17884342988524624,
+      "grad_norm": 0.49934183398618404,
+      "learning_rate": 8.984795890662768e-06,
+      "loss": 0.6705,
+      "step": 1089
+    },
+    {
+      "epoch": 0.17900765709358898,
+      "grad_norm": 0.46901602497900924,
+      "learning_rate": 8.984757347824334e-06,
+      "loss": 0.6822,
+      "step": 1090
+    },
+    {
+      "epoch": 0.17917188430193173,
+      "grad_norm": 0.4841196036749678,
+      "learning_rate": 8.984718756277019e-06,
+      "loss": 0.6612,
+      "step": 1091
+    },
+    {
+      "epoch": 0.17933611151027445,
+      "grad_norm": 0.495840243932423,
+      "learning_rate": 8.984680116021248e-06,
+      "loss": 0.6459,
+      "step": 1092
+    },
+    {
+      "epoch": 0.1795003387186172,
+      "grad_norm": 0.4745846787795327,
+      "learning_rate": 8.98464142705744e-06,
+      "loss": 0.663,
+      "step": 1093
+    },
+    {
+      "epoch": 0.17966456592695995,
+      "grad_norm": 0.4880415845630486,
+      "learning_rate": 8.984602689386013e-06,
+      "loss": 0.6527,
+      "step": 1094
+    },
+    {
+      "epoch": 0.1798287931353027,
+      "grad_norm": 0.48962473221363556,
+      "learning_rate": 8.984563903007389e-06,
+      "loss": 0.6714,
+      "step": 1095
+    },
+    {
+      "epoch": 0.17999302034364542,
+      "grad_norm": 0.483069837873451,
+      "learning_rate": 8.984525067921987e-06,
+      "loss": 0.6776,
+      "step": 1096
+    },
+    {
+      "epoch": 0.18015724755198817,
+      "grad_norm": 0.4767229218630196,
+      "learning_rate": 8.984486184130231e-06,
+      "loss": 0.6615,
+      "step": 1097
+    },
+    {
+      "epoch": 0.18032147476033092,
+      "grad_norm": 0.5064791581734579,
+      "learning_rate": 8.984447251632543e-06,
+      "loss": 0.6765,
+      "step": 1098
+    },
+    {
+      "epoch": 0.18048570196867367,
+      "grad_norm": 0.5345397900086212,
+      "learning_rate": 8.984408270429348e-06,
+      "loss": 0.6513,
+      "step": 1099
+    },
+    {
+      "epoch": 0.1806499291770164,
+      "grad_norm": 0.5296402491323332,
+      "learning_rate": 8.984369240521063e-06,
+      "loss": 0.6468,
+      "step": 1100
+    },
+    {
+      "epoch": 0.18081415638535914,
+      "grad_norm": 0.46787695321689876,
+      "learning_rate": 8.984330161908119e-06,
+      "loss": 0.6772,
+      "step": 1101
+    },
+    {
+      "epoch": 0.18097838359370189,
+      "grad_norm": 0.48925568695329985,
+      "learning_rate": 8.984291034590937e-06,
+      "loss": 0.6857,
+      "step": 1102
+    },
+    {
+      "epoch": 0.18114261080204463,
+      "grad_norm": 0.4790066192561788,
+      "learning_rate": 8.984251858569943e-06,
+      "loss": 0.6709,
+      "step": 1103
+    },
+    {
+      "epoch": 0.18130683801038738,
+      "grad_norm": 0.469440234272998,
+      "learning_rate": 8.98421263384556e-06,
+      "loss": 0.6627,
+      "step": 1104
+    },
+    {
+      "epoch": 0.1814710652187301,
+      "grad_norm": 0.49923220336592083,
+      "learning_rate": 8.984173360418219e-06,
+      "loss": 0.6747,
+      "step": 1105
+    },
+    {
+      "epoch": 0.18163529242707285,
+      "grad_norm": 0.49630959834059857,
+      "learning_rate": 8.98413403828834e-06,
+      "loss": 0.6774,
+      "step": 1106
+    },
+    {
+      "epoch": 0.1817995196354156,
+      "grad_norm": 0.4766594175572563,
+      "learning_rate": 8.984094667456355e-06,
+      "loss": 0.6526,
+      "step": 1107
+    },
+    {
+      "epoch": 0.18196374684375835,
+      "grad_norm": 0.47797237085992367,
+      "learning_rate": 8.98405524792269e-06,
+      "loss": 0.6377,
+      "step": 1108
+    },
+    {
+      "epoch": 0.18212797405210107,
+      "grad_norm": 0.6046432859375188,
+      "learning_rate": 8.984015779687773e-06,
+      "loss": 0.6396,
+      "step": 1109
+    },
+    {
+      "epoch": 0.18229220126044382,
+      "grad_norm": 0.479492420455775,
+      "learning_rate": 8.983976262752034e-06,
+      "loss": 0.6621,
+      "step": 1110
+    },
+    {
+      "epoch": 0.18245642846878657,
+      "grad_norm": 0.4843117988504339,
+      "learning_rate": 8.9839366971159e-06,
+      "loss": 0.6606,
+      "step": 1111
+    },
+    {
+      "epoch": 0.18262065567712932,
+      "grad_norm": 0.487141594800104,
+      "learning_rate": 8.983897082779804e-06,
+      "loss": 0.6644,
+      "step": 1112
+    },
+    {
+      "epoch": 0.18278488288547204,
+      "grad_norm": 0.4785509249745807,
+      "learning_rate": 8.983857419744173e-06,
+      "loss": 0.6592,
+      "step": 1113
+    },
+    {
+      "epoch": 0.1829491100938148,
+      "grad_norm": 0.4779896385199476,
+      "learning_rate": 8.983817708009438e-06,
+      "loss": 0.6618,
+      "step": 1114
+    },
+    {
+      "epoch": 0.18311333730215754,
+      "grad_norm": 0.49850559987657705,
+      "learning_rate": 8.983777947576032e-06,
+      "loss": 0.6579,
+      "step": 1115
+    },
+    {
+      "epoch": 0.18327756451050028,
+      "grad_norm": 0.4635978888264068,
+      "learning_rate": 8.983738138444387e-06,
+      "loss": 0.6794,
+      "step": 1116
+    },
+    {
+      "epoch": 0.183441791718843,
+      "grad_norm": 0.46979893505545445,
+      "learning_rate": 8.98369828061493e-06,
+      "loss": 0.6773,
+      "step": 1117
+    },
+    {
+      "epoch": 0.18360601892718575,
+      "grad_norm": 0.5249699786325431,
+      "learning_rate": 8.983658374088103e-06,
+      "loss": 0.66,
+      "step": 1118
+    },
+    {
+      "epoch": 0.1837702461355285,
+      "grad_norm": 0.6266089286333181,
+      "learning_rate": 8.983618418864334e-06,
+      "loss": 0.6625,
+      "step": 1119
+    },
+    {
+      "epoch": 0.18393447334387125,
+      "grad_norm": 0.47135356277133833,
+      "learning_rate": 8.983578414944056e-06,
+      "loss": 0.6742,
+      "step": 1120
+    },
+    {
+      "epoch": 0.184098700552214,
+      "grad_norm": 0.4939184853353077,
+      "learning_rate": 8.983538362327707e-06,
+      "loss": 0.661,
+      "step": 1121
+    },
+    {
+      "epoch": 0.18426292776055672,
+      "grad_norm": 0.5556973801833033,
+      "learning_rate": 8.98349826101572e-06,
+      "loss": 0.6765,
+      "step": 1122
+    },
+    {
+      "epoch": 0.18442715496889947,
+      "grad_norm": 0.59081209705095,
+      "learning_rate": 8.983458111008528e-06,
+      "loss": 0.6667,
+      "step": 1123
+    },
+    {
+      "epoch": 0.18459138217724222,
+      "grad_norm": 0.6127369186330817,
+      "learning_rate": 8.983417912306573e-06,
+      "loss": 0.6463,
+      "step": 1124
+    },
+    {
+      "epoch": 0.18475560938558497,
+      "grad_norm": 0.4725990782989587,
+      "learning_rate": 8.983377664910287e-06,
+      "loss": 0.6547,
+      "step": 1125
+    },
+    {
+      "epoch": 0.1849198365939277,
+      "grad_norm": 0.459266379649527,
+      "learning_rate": 8.98333736882011e-06,
+      "loss": 0.6441,
+      "step": 1126
+    },
+    {
+      "epoch": 0.18508406380227044,
+      "grad_norm": 0.5021589556181871,
+      "learning_rate": 8.983297024036475e-06,
+      "loss": 0.6618,
+      "step": 1127
+    },
+    {
+      "epoch": 0.1852482910106132,
+      "grad_norm": 0.4731458077349534,
+      "learning_rate": 8.983256630559826e-06,
+      "loss": 0.6343,
+      "step": 1128
+    },
+    {
+      "epoch": 0.18541251821895594,
+      "grad_norm": 0.46076652722107214,
+      "learning_rate": 8.983216188390598e-06,
+      "loss": 0.6571,
+      "step": 1129
+    },
+    {
+      "epoch": 0.18557674542729866,
+      "grad_norm": 0.4729510638590709,
+      "learning_rate": 8.98317569752923e-06,
+      "loss": 0.6551,
+      "step": 1130
+    },
+    {
+      "epoch": 0.1857409726356414,
+      "grad_norm": 0.4672436396771921,
+      "learning_rate": 8.983135157976166e-06,
+      "loss": 0.6526,
+      "step": 1131
+    },
+    {
+      "epoch": 0.18590519984398415,
+      "grad_norm": 0.5362518055595471,
+      "learning_rate": 8.983094569731842e-06,
+      "loss": 0.6495,
+      "step": 1132
+    },
+    {
+      "epoch": 0.1860694270523269,
+      "grad_norm": 0.4974174360111951,
+      "learning_rate": 8.9830539327967e-06,
+      "loss": 0.6796,
+      "step": 1133
+    },
+    {
+      "epoch": 0.18623365426066962,
+      "grad_norm": 0.5038650088289387,
+      "learning_rate": 8.983013247171182e-06,
+      "loss": 0.6395,
+      "step": 1134
+    },
+    {
+      "epoch": 0.18639788146901237,
+      "grad_norm": 0.46062899373893684,
+      "learning_rate": 8.98297251285573e-06,
+      "loss": 0.692,
+      "step": 1135
+    },
+    {
+      "epoch": 0.18656210867735512,
+      "grad_norm": 0.5124254676856769,
+      "learning_rate": 8.982931729850786e-06,
+      "loss": 0.6724,
+      "step": 1136
+    },
+    {
+      "epoch": 0.18672633588569787,
+      "grad_norm": 0.4507316596529217,
+      "learning_rate": 8.98289089815679e-06,
+      "loss": 0.6534,
+      "step": 1137
+    },
+    {
+      "epoch": 0.18689056309404062,
+      "grad_norm": 0.5255609623079383,
+      "learning_rate": 8.98285001777419e-06,
+      "loss": 0.6297,
+      "step": 1138
+    },
+    {
+      "epoch": 0.18705479030238334,
+      "grad_norm": 0.4753157263680939,
+      "learning_rate": 8.98280908870343e-06,
+      "loss": 0.6602,
+      "step": 1139
+    },
+    {
+      "epoch": 0.1872190175107261,
+      "grad_norm": 0.4699605583978582,
+      "learning_rate": 8.98276811094495e-06,
+      "loss": 0.6736,
+      "step": 1140
+    },
+    {
+      "epoch": 0.18738324471906884,
+      "grad_norm": 0.46174823365599904,
+      "learning_rate": 8.9827270844992e-06,
+      "loss": 0.6678,
+      "step": 1141
+    },
+    {
+      "epoch": 0.18754747192741159,
+      "grad_norm": 0.46428061282059035,
+      "learning_rate": 8.982686009366622e-06,
+      "loss": 0.6789,
+      "step": 1142
+    },
+    {
+      "epoch": 0.1877116991357543,
+      "grad_norm": 0.4699014887684446,
+      "learning_rate": 8.982644885547666e-06,
+      "loss": 0.6732,
+      "step": 1143
+    },
+    {
+      "epoch": 0.18787592634409706,
+      "grad_norm": 0.4811773724914709,
+      "learning_rate": 8.982603713042773e-06,
+      "loss": 0.66,
+      "step": 1144
+    },
+    {
+      "epoch": 0.1880401535524398,
+      "grad_norm": 0.45321971267257205,
+      "learning_rate": 8.982562491852394e-06,
+      "loss": 0.6625,
+      "step": 1145
+    },
+    {
+      "epoch": 0.18820438076078255,
+      "grad_norm": 0.6330573920154635,
+      "learning_rate": 8.982521221976978e-06,
+      "loss": 0.6692,
+      "step": 1146
+    },
+    {
+      "epoch": 0.18836860796912527,
+      "grad_norm": 0.45518725569387647,
+      "learning_rate": 8.98247990341697e-06,
+      "loss": 0.6435,
+      "step": 1147
+    },
+    {
+      "epoch": 0.18853283517746802,
+      "grad_norm": 0.47309733075163013,
+      "learning_rate": 8.982438536172819e-06,
+      "loss": 0.6296,
+      "step": 1148
+    },
+    {
+      "epoch": 0.18869706238581077,
+      "grad_norm": 0.475995730507989,
+      "learning_rate": 8.982397120244977e-06,
+      "loss": 0.6272,
+      "step": 1149
+    },
+    {
+      "epoch": 0.18886128959415352,
+      "grad_norm": 0.4547023805476552,
+      "learning_rate": 8.982355655633892e-06,
+      "loss": 0.6477,
+      "step": 1150
+    },
+    {
+      "epoch": 0.18902551680249624,
+      "grad_norm": 0.4725682510670995,
+      "learning_rate": 8.982314142340014e-06,
+      "loss": 0.6653,
+      "step": 1151
+    },
+    {
+      "epoch": 0.189189744010839,
+      "grad_norm": 0.5627057072719744,
+      "learning_rate": 8.982272580363796e-06,
+      "loss": 0.6545,
+      "step": 1152
+    },
+    {
+      "epoch": 0.18935397121918174,
+      "grad_norm": 0.5006530922999425,
+      "learning_rate": 8.982230969705685e-06,
+      "loss": 0.6545,
+      "step": 1153
+    },
+    {
+      "epoch": 0.1895181984275245,
+      "grad_norm": 0.4547374521034954,
+      "learning_rate": 8.982189310366138e-06,
+      "loss": 0.6717,
+      "step": 1154
+    },
+    {
+      "epoch": 0.18968242563586724,
+      "grad_norm": 0.445107036433399,
+      "learning_rate": 8.982147602345605e-06,
+      "loss": 0.661,
+      "step": 1155
+    },
+    {
+      "epoch": 0.18984665284420996,
+      "grad_norm": 0.4561286392537167,
+      "learning_rate": 8.982105845644539e-06,
+      "loss": 0.6447,
+      "step": 1156
+    },
+    {
+      "epoch": 0.1900108800525527,
+      "grad_norm": 0.5216727923195004,
+      "learning_rate": 8.982064040263394e-06,
+      "loss": 0.6586,
+      "step": 1157
+    },
+    {
+      "epoch": 0.19017510726089545,
+      "grad_norm": 0.45996861833590036,
+      "learning_rate": 8.982022186202623e-06,
+      "loss": 0.6618,
+      "step": 1158
+    },
+    {
+      "epoch": 0.1903393344692382,
+      "grad_norm": 0.44939371658844723,
+      "learning_rate": 8.981980283462681e-06,
+      "loss": 0.6705,
+      "step": 1159
+    },
+    {
+      "epoch": 0.19050356167758092,
+      "grad_norm": 0.48425712732679077,
+      "learning_rate": 8.981938332044024e-06,
+      "loss": 0.6684,
+      "step": 1160
+    },
+    {
+      "epoch": 0.19066778888592367,
+      "grad_norm": 0.4645137102326744,
+      "learning_rate": 8.981896331947108e-06,
+      "loss": 0.6696,
+      "step": 1161
+    },
+    {
+      "epoch": 0.19083201609426642,
+      "grad_norm": 0.4695085651143102,
+      "learning_rate": 8.981854283172386e-06,
+      "loss": 0.6632,
+      "step": 1162
+    },
+    {
+      "epoch": 0.19099624330260917,
+      "grad_norm": 0.45627726789169143,
+      "learning_rate": 8.981812185720319e-06,
+      "loss": 0.6774,
+      "step": 1163
+    },
+    {
+      "epoch": 0.1911604705109519,
+      "grad_norm": 0.4508546900179419,
+      "learning_rate": 8.98177003959136e-06,
+      "loss": 0.6495,
+      "step": 1164
+    },
+    {
+      "epoch": 0.19132469771929464,
+      "grad_norm": 0.4676794625199336,
+      "learning_rate": 8.981727844785972e-06,
+      "loss": 0.6505,
+      "step": 1165
+    },
+    {
+      "epoch": 0.1914889249276374,
+      "grad_norm": 0.46561161225429254,
+      "learning_rate": 8.981685601304608e-06,
+      "loss": 0.639,
+      "step": 1166
+    },
+    {
+      "epoch": 0.19165315213598014,
+      "grad_norm": 0.46003161609163923,
+      "learning_rate": 8.98164330914773e-06,
+      "loss": 0.659,
+      "step": 1167
+    },
+    {
+      "epoch": 0.19181737934432286,
+      "grad_norm": 0.4750628926671919,
+      "learning_rate": 8.981600968315796e-06,
+      "loss": 0.6494,
+      "step": 1168
+    },
+    {
+      "epoch": 0.1919816065526656,
+      "grad_norm": 0.4766948623132289,
+      "learning_rate": 8.981558578809265e-06,
+      "loss": 0.6408,
+      "step": 1169
+    },
+    {
+      "epoch": 0.19214583376100836,
+      "grad_norm": 0.5949849501521395,
+      "learning_rate": 8.9815161406286e-06,
+      "loss": 0.6377,
+      "step": 1170
+    },
+    {
+      "epoch": 0.1923100609693511,
+      "grad_norm": 0.4315864051018545,
+      "learning_rate": 8.98147365377426e-06,
+      "loss": 0.657,
+      "step": 1171
+    },
+    {
+      "epoch": 0.19247428817769385,
+      "grad_norm": 0.48620239608886057,
+      "learning_rate": 8.981431118246707e-06,
+      "loss": 0.6484,
+      "step": 1172
+    },
+    {
+      "epoch": 0.19263851538603657,
+      "grad_norm": 0.48009521513705844,
+      "learning_rate": 8.981388534046403e-06,
+      "loss": 0.661,
+      "step": 1173
+    },
+    {
+      "epoch": 0.19280274259437932,
+      "grad_norm": 0.46661786912772535,
+      "learning_rate": 8.981345901173812e-06,
+      "loss": 0.6553,
+      "step": 1174
+    },
+    {
+      "epoch": 0.19296696980272207,
+      "grad_norm": 0.4670500707072574,
+      "learning_rate": 8.981303219629392e-06,
+      "loss": 0.6443,
+      "step": 1175
+    },
+    {
+      "epoch": 0.19313119701106482,
+      "grad_norm": 0.4409917364955242,
+      "learning_rate": 8.981260489413613e-06,
+      "loss": 0.6619,
+      "step": 1176
+    },
+    {
+      "epoch": 0.19329542421940754,
+      "grad_norm": 0.4711657126739569,
+      "learning_rate": 8.981217710526935e-06,
+      "loss": 0.635,
+      "step": 1177
+    },
+    {
+      "epoch": 0.1934596514277503,
+      "grad_norm": 0.5696321006657865,
+      "learning_rate": 8.981174882969823e-06,
+      "loss": 0.6625,
+      "step": 1178
+    },
+    {
+      "epoch": 0.19362387863609304,
+      "grad_norm": 0.46026048794687807,
+      "learning_rate": 8.981132006742745e-06,
+      "loss": 0.6392,
+      "step": 1179
+    },
+    {
+      "epoch": 0.1937881058444358,
+      "grad_norm": 0.44618034160175857,
+      "learning_rate": 8.981089081846164e-06,
+      "loss": 0.6793,
+      "step": 1180
+    },
+    {
+      "epoch": 0.1939523330527785,
+      "grad_norm": 0.466485326789076,
+      "learning_rate": 8.981046108280545e-06,
+      "loss": 0.6483,
+      "step": 1181
+    },
+    {
+      "epoch": 0.19411656026112126,
+      "grad_norm": 0.48215706406724307,
+      "learning_rate": 8.981003086046358e-06,
+      "loss": 0.6493,
+      "step": 1182
+    },
+    {
+      "epoch": 0.194280787469464,
+      "grad_norm": 0.44683277849330777,
+      "learning_rate": 8.980960015144068e-06,
+      "loss": 0.6471,
+      "step": 1183
+    },
+    {
+      "epoch": 0.19444501467780675,
+      "grad_norm": 0.4492007111991035,
+      "learning_rate": 8.980916895574143e-06,
+      "loss": 0.6608,
+      "step": 1184
+    },
+    {
+      "epoch": 0.19460924188614948,
+      "grad_norm": 0.43166191142444144,
+      "learning_rate": 8.980873727337053e-06,
+      "loss": 0.6674,
+      "step": 1185
+    },
+    {
+      "epoch": 0.19477346909449222,
+      "grad_norm": 0.43998146007104594,
+      "learning_rate": 8.980830510433266e-06,
+      "loss": 0.6579,
+      "step": 1186
+    },
+    {
+      "epoch": 0.19493769630283497,
+      "grad_norm": 0.48354840890327583,
+      "learning_rate": 8.98078724486325e-06,
+      "loss": 0.6475,
+      "step": 1187
+    },
+    {
+      "epoch": 0.19510192351117772,
+      "grad_norm": 0.48479443389772425,
+      "learning_rate": 8.980743930627477e-06,
+      "loss": 0.6576,
+      "step": 1188
+    },
+    {
+      "epoch": 0.19526615071952047,
+      "grad_norm": 0.46889288135317325,
+      "learning_rate": 8.980700567726415e-06,
+      "loss": 0.6674,
+      "step": 1189
+    },
+    {
+      "epoch": 0.1954303779278632,
+      "grad_norm": 0.4363016948383347,
+      "learning_rate": 8.980657156160538e-06,
+      "loss": 0.6426,
+      "step": 1190
+    },
+    {
+      "epoch": 0.19559460513620594,
+      "grad_norm": 0.4856980442738892,
+      "learning_rate": 8.980613695930315e-06,
+      "loss": 0.6763,
+      "step": 1191
+    },
+    {
+      "epoch": 0.1957588323445487,
+      "grad_norm": 0.43424616977986746,
+      "learning_rate": 8.98057018703622e-06,
+      "loss": 0.6549,
+      "step": 1192
+    },
+    {
+      "epoch": 0.19592305955289144,
+      "grad_norm": 0.4315225641477105,
+      "learning_rate": 8.980526629478724e-06,
+      "loss": 0.633,
+      "step": 1193
+    },
+    {
+      "epoch": 0.19608728676123416,
+      "grad_norm": 0.44353238665548256,
+      "learning_rate": 8.9804830232583e-06,
+      "loss": 0.6314,
+      "step": 1194
+    },
+    {
+      "epoch": 0.1962515139695769,
+      "grad_norm": 0.42772453875967764,
+      "learning_rate": 8.980439368375423e-06,
+      "loss": 0.6608,
+      "step": 1195
+    },
+    {
+      "epoch": 0.19641574117791966,
+      "grad_norm": 0.47642585813722443,
+      "learning_rate": 8.980395664830566e-06,
+      "loss": 0.6469,
+      "step": 1196
+    },
+    {
+      "epoch": 0.1965799683862624,
+      "grad_norm": 0.4585709399028893,
+      "learning_rate": 8.980351912624204e-06,
+      "loss": 0.6409,
+      "step": 1197
+    },
+    {
+      "epoch": 0.19674419559460513,
+      "grad_norm": 0.4683407067017671,
+      "learning_rate": 8.980308111756812e-06,
+      "loss": 0.6343,
+      "step": 1198
+    },
+    {
+      "epoch": 0.19690842280294787,
+      "grad_norm": 0.45598990783869575,
+      "learning_rate": 8.980264262228865e-06,
+      "loss": 0.6723,
+      "step": 1199
+    },
+    {
+      "epoch": 0.19707265001129062,
+      "grad_norm": 0.4441138325488558,
+      "learning_rate": 8.980220364040843e-06,
+      "loss": 0.642,
+      "step": 1200
+    },
+    {
+      "epoch": 0.19723687721963337,
+      "grad_norm": 0.4372372641941166,
+      "learning_rate": 8.980176417193217e-06,
+      "loss": 0.6613,
+      "step": 1201
+    },
+    {
+      "epoch": 0.1974011044279761,
+      "grad_norm": 0.43527778983518434,
+      "learning_rate": 8.980132421686467e-06,
+      "loss": 0.6413,
+      "step": 1202
+    },
+    {
+      "epoch": 0.19756533163631884,
+      "grad_norm": 0.4318187764560961,
+      "learning_rate": 8.980088377521073e-06,
+      "loss": 0.6488,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1977295588446616,
+      "grad_norm": 0.45481144116691763,
+      "learning_rate": 8.98004428469751e-06,
+      "loss": 0.6527,
+      "step": 1204
+    },
+    {
+      "epoch": 0.19789378605300434,
+      "grad_norm": 0.5522025822262004,
+      "learning_rate": 8.98000014321626e-06,
+      "loss": 0.6489,
+      "step": 1205
+    },
+    {
+      "epoch": 0.19805801326134706,
+      "grad_norm": 0.41859957738652587,
+      "learning_rate": 8.9799559530778e-06,
+      "loss": 0.6368,
+      "step": 1206
+    },
+    {
+      "epoch": 0.1982222404696898,
+      "grad_norm": 0.45326825381870833,
+      "learning_rate": 8.97991171428261e-06,
+      "loss": 0.6596,
+      "step": 1207
+    },
+    {
+      "epoch": 0.19838646767803256,
+      "grad_norm": 0.44391403918832606,
+      "learning_rate": 8.979867426831171e-06,
+      "loss": 0.6476,
+      "step": 1208
+    },
+    {
+      "epoch": 0.1985506948863753,
+      "grad_norm": 0.43255445491586364,
+      "learning_rate": 8.979823090723966e-06,
+      "loss": 0.6301,
+      "step": 1209
+    },
+    {
+      "epoch": 0.19871492209471806,
+      "grad_norm": 0.428228400413717,
+      "learning_rate": 8.979778705961471e-06,
+      "loss": 0.6779,
+      "step": 1210
+    },
+    {
+      "epoch": 0.19887914930306078,
+      "grad_norm": 0.461878401668105,
+      "learning_rate": 8.979734272544175e-06,
+      "loss": 0.648,
+      "step": 1211
+    },
+    {
+      "epoch": 0.19904337651140352,
+      "grad_norm": 0.4311866229558285,
+      "learning_rate": 8.979689790472556e-06,
+      "loss": 0.6456,
+      "step": 1212
+    },
+    {
+      "epoch": 0.19920760371974627,
+      "grad_norm": 0.41531054144954355,
+      "learning_rate": 8.979645259747098e-06,
+      "loss": 0.6407,
+      "step": 1213
+    },
+    {
+      "epoch": 0.19937183092808902,
+      "grad_norm": 0.4262317469216339,
+      "learning_rate": 8.979600680368286e-06,
+      "loss": 0.6243,
+      "step": 1214
+    },
+    {
+      "epoch": 0.19953605813643174,
+      "grad_norm": 0.4314545771479194,
+      "learning_rate": 8.979556052336605e-06,
+      "loss": 0.6493,
+      "step": 1215
+    },
+    {
+      "epoch": 0.1997002853447745,
+      "grad_norm": 0.44815070308467875,
+      "learning_rate": 8.979511375652535e-06,
+      "loss": 0.6438,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19986451255311724,
+      "grad_norm": 0.4422403594628592,
+      "learning_rate": 8.979466650316565e-06,
+      "loss": 0.669,
+      "step": 1217
+    },
+    {
+      "epoch": 0.20002873976146,
+      "grad_norm": 0.42824265290355684,
+      "learning_rate": 8.97942187632918e-06,
+      "loss": 0.6579,
+      "step": 1218
+    },
+    {
+      "epoch": 0.2001929669698027,
+      "grad_norm": 0.543273528497163,
+      "learning_rate": 8.979377053690867e-06,
+      "loss": 0.6373,
+      "step": 1219
+    },
+    {
+      "epoch": 0.20035719417814546,
+      "grad_norm": 0.4230607481140885,
+      "learning_rate": 8.979332182402111e-06,
+      "loss": 0.6431,
+      "step": 1220
+    },
+    {
+      "epoch": 0.2005214213864882,
+      "grad_norm": 0.4362433490235076,
+      "learning_rate": 8.979287262463403e-06,
+      "loss": 0.6501,
+      "step": 1221
+    },
+    {
+      "epoch": 0.20068564859483096,
+      "grad_norm": 0.4451428401814509,
+      "learning_rate": 8.979242293875225e-06,
+      "loss": 0.6557,
+      "step": 1222
+    },
+    {
+      "epoch": 0.20084987580317368,
+      "grad_norm": 0.4352292485824094,
+      "learning_rate": 8.979197276638071e-06,
+      "loss": 0.6598,
+      "step": 1223
+    },
+    {
+      "epoch": 0.20101410301151643,
+      "grad_norm": 0.4261401748689992,
+      "learning_rate": 8.979152210752427e-06,
+      "loss": 0.6318,
+      "step": 1224
+    },
+    {
+      "epoch": 0.20117833021985917,
+      "grad_norm": 0.41345651752110774,
+      "learning_rate": 8.979107096218781e-06,
+      "loss": 0.6701,
+      "step": 1225
+    },
+    {
+      "epoch": 0.20134255742820192,
+      "grad_norm": 0.4257173704920539,
+      "learning_rate": 8.979061933037629e-06,
+      "loss": 0.6462,
+      "step": 1226
+    },
+    {
+      "epoch": 0.20150678463654467,
+      "grad_norm": 0.4402291803372949,
+      "learning_rate": 8.979016721209456e-06,
+      "loss": 0.6405,
+      "step": 1227
+    },
+    {
+      "epoch": 0.2016710118448874,
+      "grad_norm": 0.4149097427684321,
+      "learning_rate": 8.978971460734753e-06,
+      "loss": 0.6203,
+      "step": 1228
+    },
+    {
+      "epoch": 0.20183523905323014,
+      "grad_norm": 0.48308972389337906,
+      "learning_rate": 8.978926151614014e-06,
+      "loss": 0.6642,
+      "step": 1229
+    },
+    {
+      "epoch": 0.2019994662615729,
+      "grad_norm": 0.4331467017777629,
+      "learning_rate": 8.978880793847732e-06,
+      "loss": 0.6395,
+      "step": 1230
+    },
+    {
+      "epoch": 0.20216369346991564,
+      "grad_norm": 0.43266826431991046,
+      "learning_rate": 8.978835387436396e-06,
+      "loss": 0.6434,
+      "step": 1231
+    },
+    {
+      "epoch": 0.20232792067825836,
+      "grad_norm": 0.4460728101045431,
+      "learning_rate": 8.978789932380501e-06,
+      "loss": 0.6286,
+      "step": 1232
+    },
+    {
+      "epoch": 0.2024921478866011,
+      "grad_norm": 0.43293087715759737,
+      "learning_rate": 8.978744428680543e-06,
+      "loss": 0.6546,
+      "step": 1233
+    },
+    {
+      "epoch": 0.20265637509494386,
+      "grad_norm": 0.42924541630533913,
+      "learning_rate": 8.978698876337011e-06,
+      "loss": 0.6446,
+      "step": 1234
+    },
+    {
+      "epoch": 0.2028206023032866,
+      "grad_norm": 0.4151254920163802,
+      "learning_rate": 8.978653275350405e-06,
+      "loss": 0.6631,
+      "step": 1235
+    },
+    {
+      "epoch": 0.20298482951162933,
+      "grad_norm": 0.44360392632935175,
+      "learning_rate": 8.978607625721219e-06,
+      "loss": 0.6303,
+      "step": 1236
+    },
+    {
+      "epoch": 0.20314905671997208,
+      "grad_norm": 0.4277824314710478,
+      "learning_rate": 8.978561927449946e-06,
+      "loss": 0.6467,
+      "step": 1237
+    },
+    {
+      "epoch": 0.20331328392831483,
+      "grad_norm": 0.4366290288251897,
+      "learning_rate": 8.978516180537083e-06,
+      "loss": 0.6476,
+      "step": 1238
+    },
+    {
+      "epoch": 0.20347751113665757,
+      "grad_norm": 0.42651600201106177,
+      "learning_rate": 8.97847038498313e-06,
+      "loss": 0.6514,
+      "step": 1239
+    },
+    {
+      "epoch": 0.2036417383450003,
+      "grad_norm": 0.4251031339705655,
+      "learning_rate": 8.978424540788583e-06,
+      "loss": 0.6734,
+      "step": 1240
+    },
+    {
+      "epoch": 0.20380596555334304,
+      "grad_norm": 0.549156680303186,
+      "learning_rate": 8.978378647953937e-06,
+      "loss": 0.6291,
+      "step": 1241
+    },
+    {
+      "epoch": 0.2039701927616858,
+      "grad_norm": 0.4224197425754509,
+      "learning_rate": 8.978332706479694e-06,
+      "loss": 0.6516,
+      "step": 1242
+    },
+    {
+      "epoch": 0.20413441997002854,
+      "grad_norm": 0.42435648487983524,
+      "learning_rate": 8.978286716366352e-06,
+      "loss": 0.6401,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2042986471783713,
+      "grad_norm": 0.4218890161029301,
+      "learning_rate": 8.97824067761441e-06,
+      "loss": 0.6332,
+      "step": 1244
+    },
+    {
+      "epoch": 0.204462874386714,
+      "grad_norm": 0.4598458015905371,
+      "learning_rate": 8.978194590224367e-06,
+      "loss": 0.636,
+      "step": 1245
+    },
+    {
+      "epoch": 0.20462710159505676,
+      "grad_norm": 0.4494394320075245,
+      "learning_rate": 8.978148454196728e-06,
+      "loss": 0.6628,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2047913288033995,
+      "grad_norm": 0.45629173674361884,
+      "learning_rate": 8.978102269531988e-06,
+      "loss": 0.6434,
+      "step": 1247
+    },
+    {
+      "epoch": 0.20495555601174226,
+      "grad_norm": 0.42740668248808533,
+      "learning_rate": 8.978056036230651e-06,
+      "loss": 0.6408,
+      "step": 1248
+    },
+    {
+      "epoch": 0.20511978322008498,
+      "grad_norm": 0.45719887031605844,
+      "learning_rate": 8.978009754293221e-06,
+      "loss": 0.6407,
+      "step": 1249
+    },
+    {
+      "epoch": 0.20528401042842773,
+      "grad_norm": 0.4352265634714779,
+      "learning_rate": 8.9779634237202e-06,
+      "loss": 0.6371,
+      "step": 1250
+    },
+    {
+      "epoch": 0.20544823763677048,
+      "grad_norm": 0.42210549828374605,
+      "learning_rate": 8.97791704451209e-06,
+      "loss": 0.6408,
+      "step": 1251
+    },
+    {
+      "epoch": 0.20561246484511322,
+      "grad_norm": 0.40181174335862413,
+      "learning_rate": 8.977870616669395e-06,
+      "loss": 0.6494,
+      "step": 1252
+    },
+    {
+      "epoch": 0.20577669205345595,
+      "grad_norm": 0.40188184471409283,
+      "learning_rate": 8.97782414019262e-06,
+      "loss": 0.6593,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2059409192617987,
+      "grad_norm": 0.410659768636049,
+      "learning_rate": 8.977777615082268e-06,
+      "loss": 0.6452,
+      "step": 1254
+    },
+    {
+      "epoch": 0.20610514647014144,
+      "grad_norm": 0.4235122270844502,
+      "learning_rate": 8.977731041338847e-06,
+      "loss": 0.6438,
+      "step": 1255
+    },
+    {
+      "epoch": 0.2062693736784842,
+      "grad_norm": 0.44846490809871864,
+      "learning_rate": 8.97768441896286e-06,
+      "loss": 0.6454,
+      "step": 1256
+    },
+    {
+      "epoch": 0.2064336008868269,
+      "grad_norm": 0.42416716591523856,
+      "learning_rate": 8.977637747954815e-06,
+      "loss": 0.6342,
+      "step": 1257
+    },
+    {
+      "epoch": 0.20659782809516966,
+      "grad_norm": 0.4198893083929536,
+      "learning_rate": 8.977591028315221e-06,
+      "loss": 0.6533,
+      "step": 1258
+    },
+    {
+      "epoch": 0.2067620553035124,
+      "grad_norm": 0.4455138571329966,
+      "learning_rate": 8.97754426004458e-06,
+      "loss": 0.6586,
+      "step": 1259
+    },
+    {
+      "epoch": 0.20692628251185516,
+      "grad_norm": 0.4142993504597112,
+      "learning_rate": 8.977497443143405e-06,
+      "loss": 0.6525,
+      "step": 1260
+    },
+    {
+      "epoch": 0.2070905097201979,
+      "grad_norm": 0.42824370833772935,
+      "learning_rate": 8.9774505776122e-06,
+      "loss": 0.6228,
+      "step": 1261
+    },
+    {
+      "epoch": 0.20725473692854063,
+      "grad_norm": 0.424782198551827,
+      "learning_rate": 8.977403663451478e-06,
+      "loss": 0.6432,
+      "step": 1262
+    },
+    {
+      "epoch": 0.20741896413688338,
+      "grad_norm": 0.41832663825833505,
+      "learning_rate": 8.977356700661749e-06,
+      "loss": 0.6567,
+      "step": 1263
+    },
+    {
+      "epoch": 0.20758319134522613,
+      "grad_norm": 0.45774058372764936,
+      "learning_rate": 8.977309689243519e-06,
+      "loss": 0.6594,
+      "step": 1264
+    },
+    {
+      "epoch": 0.20774741855356887,
+      "grad_norm": 0.4315936364515443,
+      "learning_rate": 8.9772626291973e-06,
+      "loss": 0.6348,
+      "step": 1265
+    },
+    {
+      "epoch": 0.2079116457619116,
+      "grad_norm": 0.4070764707847252,
+      "learning_rate": 8.977215520523605e-06,
+      "loss": 0.6427,
+      "step": 1266
+    },
+    {
+      "epoch": 0.20807587297025434,
+      "grad_norm": 0.4162334688331903,
+      "learning_rate": 8.977168363222944e-06,
+      "loss": 0.6605,
+      "step": 1267
+    },
+    {
+      "epoch": 0.2082401001785971,
+      "grad_norm": 0.4053266968546352,
+      "learning_rate": 8.977121157295831e-06,
+      "loss": 0.6579,
+      "step": 1268
+    },
+    {
+      "epoch": 0.20840432738693984,
+      "grad_norm": 0.43978146816965014,
+      "learning_rate": 8.977073902742775e-06,
+      "loss": 0.6334,
+      "step": 1269
+    },
+    {
+      "epoch": 0.20856855459528256,
+      "grad_norm": 0.40390746582339854,
+      "learning_rate": 8.977026599564294e-06,
+      "loss": 0.6526,
+      "step": 1270
+    },
+    {
+      "epoch": 0.2087327818036253,
+      "grad_norm": 0.3978119140838034,
+      "learning_rate": 8.976979247760898e-06,
+      "loss": 0.6379,
+      "step": 1271
+    },
+    {
+      "epoch": 0.20889700901196806,
+      "grad_norm": 0.400050972879593,
+      "learning_rate": 8.976931847333104e-06,
+      "loss": 0.634,
+      "step": 1272
+    },
+    {
+      "epoch": 0.2090612362203108,
+      "grad_norm": 0.43290124573046884,
+      "learning_rate": 8.976884398281424e-06,
+      "loss": 0.6635,
+      "step": 1273
+    },
+    {
+      "epoch": 0.20922546342865353,
+      "grad_norm": 0.4087546614663093,
+      "learning_rate": 8.976836900606375e-06,
+      "loss": 0.6506,
+      "step": 1274
+    },
+    {
+      "epoch": 0.20938969063699628,
+      "grad_norm": 0.3974013674929746,
+      "learning_rate": 8.976789354308471e-06,
+      "loss": 0.6425,
+      "step": 1275
+    },
+    {
+      "epoch": 0.20955391784533903,
+      "grad_norm": 0.519176836639908,
+      "learning_rate": 8.976741759388233e-06,
+      "loss": 0.6533,
+      "step": 1276
+    },
+    {
+      "epoch": 0.20971814505368178,
+      "grad_norm": 0.4212920607770516,
+      "learning_rate": 8.976694115846174e-06,
+      "loss": 0.6414,
+      "step": 1277
+    },
+    {
+      "epoch": 0.20988237226202452,
+      "grad_norm": 0.42102429097091515,
+      "learning_rate": 8.976646423682811e-06,
+      "loss": 0.6277,
+      "step": 1278
+    },
+    {
+      "epoch": 0.21004659947036725,
+      "grad_norm": 0.4265483499282997,
+      "learning_rate": 8.976598682898665e-06,
+      "loss": 0.6527,
+      "step": 1279
+    },
+    {
+      "epoch": 0.21021082667871,
+      "grad_norm": 0.44527848666988024,
+      "learning_rate": 8.976550893494252e-06,
+      "loss": 0.6275,
+      "step": 1280
+    },
+    {
+      "epoch": 0.21037505388705274,
+      "grad_norm": 0.4074471380026174,
+      "learning_rate": 8.976503055470093e-06,
+      "loss": 0.6325,
+      "step": 1281
+    },
+    {
+      "epoch": 0.2105392810953955,
+      "grad_norm": 0.42635054736052974,
+      "learning_rate": 8.976455168826705e-06,
+      "loss": 0.6404,
+      "step": 1282
+    },
+    {
+      "epoch": 0.2107035083037382,
+      "grad_norm": 0.5478927850514103,
+      "learning_rate": 8.97640723356461e-06,
+      "loss": 0.6493,
+      "step": 1283
+    },
+    {
+      "epoch": 0.21086773551208096,
+      "grad_norm": 0.4516746938788672,
+      "learning_rate": 8.976359249684329e-06,
+      "loss": 0.6431,
+      "step": 1284
+    },
+    {
+      "epoch": 0.2110319627204237,
+      "grad_norm": 0.4591186560356215,
+      "learning_rate": 8.976311217186384e-06,
+      "loss": 0.6636,
+      "step": 1285
+    },
+    {
+      "epoch": 0.21119618992876646,
+      "grad_norm": 0.42161671968749,
+      "learning_rate": 8.976263136071294e-06,
+      "loss": 0.6388,
+      "step": 1286
+    },
+    {
+      "epoch": 0.21136041713710918,
+      "grad_norm": 0.4294902286359102,
+      "learning_rate": 8.97621500633958e-06,
+      "loss": 0.6307,
+      "step": 1287
+    },
+    {
+      "epoch": 0.21152464434545193,
+      "grad_norm": 0.47729600920882603,
+      "learning_rate": 8.97616682799177e-06,
+      "loss": 0.664,
+      "step": 1288
+    },
+    {
+      "epoch": 0.21168887155379468,
+      "grad_norm": 0.40274636860562796,
+      "learning_rate": 8.976118601028382e-06,
+      "loss": 0.6676,
+      "step": 1289
+    },
+    {
+      "epoch": 0.21185309876213743,
+      "grad_norm": 0.41214219381854394,
+      "learning_rate": 8.976070325449942e-06,
+      "loss": 0.6405,
+      "step": 1290
+    },
+    {
+      "epoch": 0.21201732597048015,
+      "grad_norm": 0.41848751377730636,
+      "learning_rate": 8.976022001256977e-06,
+      "loss": 0.6483,
+      "step": 1291
+    },
+    {
+      "epoch": 0.2121815531788229,
+      "grad_norm": 0.42500778911619475,
+      "learning_rate": 8.975973628450006e-06,
+      "loss": 0.6406,
+      "step": 1292
+    },
+    {
+      "epoch": 0.21234578038716564,
+      "grad_norm": 0.4119091349723487,
+      "learning_rate": 8.97592520702956e-06,
+      "loss": 0.6458,
+      "step": 1293
+    },
+    {
+      "epoch": 0.2125100075955084,
+      "grad_norm": 0.44255627064099584,
+      "learning_rate": 8.975876736996163e-06,
+      "loss": 0.64,
+      "step": 1294
+    },
+    {
+      "epoch": 0.21267423480385111,
+      "grad_norm": 0.43848087034148364,
+      "learning_rate": 8.97582821835034e-06,
+      "loss": 0.6463,
+      "step": 1295
+    },
+    {
+      "epoch": 0.21283846201219386,
+      "grad_norm": 0.42192516605366814,
+      "learning_rate": 8.975779651092618e-06,
+      "loss": 0.6413,
+      "step": 1296
+    },
+    {
+      "epoch": 0.2130026892205366,
+      "grad_norm": 0.4323607525943481,
+      "learning_rate": 8.975731035223526e-06,
+      "loss": 0.6407,
+      "step": 1297
+    },
+    {
+      "epoch": 0.21316691642887936,
+      "grad_norm": 0.3968768309143591,
+      "learning_rate": 8.975682370743592e-06,
+      "loss": 0.626,
+      "step": 1298
+    },
+    {
+      "epoch": 0.2133311436372221,
+      "grad_norm": 0.40895837748550445,
+      "learning_rate": 8.975633657653344e-06,
+      "loss": 0.6436,
+      "step": 1299
+    },
+    {
+      "epoch": 0.21349537084556483,
+      "grad_norm": 0.4110881048866647,
+      "learning_rate": 8.97558489595331e-06,
+      "loss": 0.6084,
+      "step": 1300
+    },
+    {
+      "epoch": 0.21365959805390758,
+      "grad_norm": 0.7065067394013022,
+      "learning_rate": 8.975536085644022e-06,
+      "loss": 0.6485,
+      "step": 1301
+    },
+    {
+      "epoch": 0.21382382526225033,
+      "grad_norm": 0.5630818303342476,
+      "learning_rate": 8.975487226726007e-06,
+      "loss": 0.6415,
+      "step": 1302
+    },
+    {
+      "epoch": 0.21398805247059308,
+      "grad_norm": 0.42991690147388806,
+      "learning_rate": 8.975438319199798e-06,
+      "loss": 0.6244,
+      "step": 1303
+    },
+    {
+      "epoch": 0.2141522796789358,
+      "grad_norm": 0.3916322356920595,
+      "learning_rate": 8.975389363065928e-06,
+      "loss": 0.6474,
+      "step": 1304
+    },
+    {
+      "epoch": 0.21431650688727855,
+      "grad_norm": 0.4040633626617589,
+      "learning_rate": 8.975340358324925e-06,
+      "loss": 0.6482,
+      "step": 1305
+    },
+    {
+      "epoch": 0.2144807340956213,
+      "grad_norm": 0.39442006030905546,
+      "learning_rate": 8.97529130497732e-06,
+      "loss": 0.6329,
+      "step": 1306
+    },
+    {
+      "epoch": 0.21464496130396404,
+      "grad_norm": 0.628544271221126,
+      "learning_rate": 8.975242203023652e-06,
+      "loss": 0.6529,
+      "step": 1307
+    },
+    {
+      "epoch": 0.21480918851230676,
+      "grad_norm": 0.38401898520257327,
+      "learning_rate": 8.97519305246445e-06,
+      "loss": 0.6534,
+      "step": 1308
+    },
+    {
+      "epoch": 0.2149734157206495,
+      "grad_norm": 0.39648552275477433,
+      "learning_rate": 8.975143853300246e-06,
+      "loss": 0.6424,
+      "step": 1309
+    },
+    {
+      "epoch": 0.21513764292899226,
+      "grad_norm": 0.3998211133640743,
+      "learning_rate": 8.975094605531577e-06,
+      "loss": 0.6135,
+      "step": 1310
+    },
+    {
+      "epoch": 0.215301870137335,
+      "grad_norm": 0.41497948541509916,
+      "learning_rate": 8.975045309158978e-06,
+      "loss": 0.6564,
+      "step": 1311
+    },
+    {
+      "epoch": 0.21546609734567773,
+      "grad_norm": 0.4072266315384676,
+      "learning_rate": 8.974995964182987e-06,
+      "loss": 0.6239,
+      "step": 1312
+    },
+    {
+      "epoch": 0.21563032455402048,
+      "grad_norm": 0.4158843054553293,
+      "learning_rate": 8.974946570604135e-06,
+      "loss": 0.6197,
+      "step": 1313
+    },
+    {
+      "epoch": 0.21579455176236323,
+      "grad_norm": 0.42555213454131813,
+      "learning_rate": 8.97489712842296e-06,
+      "loss": 0.6357,
+      "step": 1314
+    },
+    {
+      "epoch": 0.21595877897070598,
+      "grad_norm": 0.4344847335024281,
+      "learning_rate": 8.97484763764e-06,
+      "loss": 0.6538,
+      "step": 1315
+    },
+    {
+      "epoch": 0.21612300617904873,
+      "grad_norm": 0.40231338795547006,
+      "learning_rate": 8.974798098255793e-06,
+      "loss": 0.6438,
+      "step": 1316
+    },
+    {
+      "epoch": 0.21628723338739145,
+      "grad_norm": 0.401018844856692,
+      "learning_rate": 8.974748510270874e-06,
+      "loss": 0.6219,
+      "step": 1317
+    },
+    {
+      "epoch": 0.2164514605957342,
+      "grad_norm": 0.3879562862518085,
+      "learning_rate": 8.974698873685786e-06,
+      "loss": 0.6432,
+      "step": 1318
+    },
+    {
+      "epoch": 0.21661568780407695,
+      "grad_norm": 0.3916001285671798,
+      "learning_rate": 8.974649188501065e-06,
+      "loss": 0.6266,
+      "step": 1319
+    },
+    {
+      "epoch": 0.2167799150124197,
+      "grad_norm": 0.43067968795936457,
+      "learning_rate": 8.974599454717248e-06,
+      "loss": 0.6507,
+      "step": 1320
+    },
+    {
+      "epoch": 0.21694414222076241,
+      "grad_norm": 0.3984580712524322,
+      "learning_rate": 8.974549672334883e-06,
+      "loss": 0.6419,
+      "step": 1321
+    },
+    {
+      "epoch": 0.21710836942910516,
+      "grad_norm": 0.46140477849198946,
+      "learning_rate": 8.974499841354504e-06,
+      "loss": 0.6416,
+      "step": 1322
+    },
+    {
+      "epoch": 0.2172725966374479,
+      "grad_norm": 0.459126770981474,
+      "learning_rate": 8.974449961776656e-06,
+      "loss": 0.6478,
+      "step": 1323
+    },
+    {
+      "epoch": 0.21743682384579066,
+      "grad_norm": 0.46092127260237753,
+      "learning_rate": 8.974400033601878e-06,
+      "loss": 0.6424,
+      "step": 1324
+    },
+    {
+      "epoch": 0.21760105105413338,
+      "grad_norm": 0.39361748300881266,
+      "learning_rate": 8.974350056830712e-06,
+      "loss": 0.6444,
+      "step": 1325
+    },
+    {
+      "epoch": 0.21776527826247613,
+      "grad_norm": 0.4046619648292167,
+      "learning_rate": 8.974300031463704e-06,
+      "loss": 0.6576,
+      "step": 1326
+    },
+    {
+      "epoch": 0.21792950547081888,
+      "grad_norm": 0.37804314723939025,
+      "learning_rate": 8.974249957501395e-06,
+      "loss": 0.6309,
+      "step": 1327
+    },
+    {
+      "epoch": 0.21809373267916163,
+      "grad_norm": 0.5167184663185815,
+      "learning_rate": 8.97419983494433e-06,
+      "loss": 0.643,
+      "step": 1328
+    },
+    {
+      "epoch": 0.21825795988750435,
+      "grad_norm": 0.37870371535857766,
+      "learning_rate": 8.974149663793053e-06,
+      "loss": 0.6322,
+      "step": 1329
+    },
+    {
+      "epoch": 0.2184221870958471,
+      "grad_norm": 0.39915021911898146,
+      "learning_rate": 8.974099444048108e-06,
+      "loss": 0.6457,
+      "step": 1330
+    },
+    {
+      "epoch": 0.21858641430418985,
+      "grad_norm": 0.4006828964084616,
+      "learning_rate": 8.97404917571004e-06,
+      "loss": 0.6247,
+      "step": 1331
+    },
+    {
+      "epoch": 0.2187506415125326,
+      "grad_norm": 0.40900509712567223,
+      "learning_rate": 8.973998858779397e-06,
+      "loss": 0.6388,
+      "step": 1332
+    },
+    {
+      "epoch": 0.21891486872087534,
+      "grad_norm": 0.454939516696798,
+      "learning_rate": 8.973948493256727e-06,
+      "loss": 0.629,
+      "step": 1333
+    },
+    {
+      "epoch": 0.21907909592921807,
+      "grad_norm": 0.37761379656488037,
+      "learning_rate": 8.973898079142573e-06,
+      "loss": 0.6359,
+      "step": 1334
+    },
+    {
+      "epoch": 0.2192433231375608,
+      "grad_norm": 0.4672563224075992,
+      "learning_rate": 8.973847616437483e-06,
+      "loss": 0.6387,
+      "step": 1335
+    },
+    {
+      "epoch": 0.21940755034590356,
+      "grad_norm": 0.3981227122118326,
+      "learning_rate": 8.973797105142005e-06,
+      "loss": 0.6317,
+      "step": 1336
+    },
+    {
+      "epoch": 0.2195717775542463,
+      "grad_norm": 0.45785924142754797,
+      "learning_rate": 8.973746545256692e-06,
+      "loss": 0.6275,
+      "step": 1337
+    },
+    {
+      "epoch": 0.21973600476258903,
+      "grad_norm": 0.40208418946268193,
+      "learning_rate": 8.973695936782088e-06,
+      "loss": 0.6649,
+      "step": 1338
+    },
+    {
+      "epoch": 0.21990023197093178,
+      "grad_norm": 0.4196465854058847,
+      "learning_rate": 8.973645279718746e-06,
+      "loss": 0.65,
+      "step": 1339
+    },
+    {
+      "epoch": 0.22006445917927453,
+      "grad_norm": 0.39709401064649014,
+      "learning_rate": 8.973594574067214e-06,
+      "loss": 0.6219,
+      "step": 1340
+    },
+    {
+      "epoch": 0.22022868638761728,
+      "grad_norm": 0.4317979361442743,
+      "learning_rate": 8.973543819828042e-06,
+      "loss": 0.6352,
+      "step": 1341
+    },
+    {
+      "epoch": 0.22039291359596,
+      "grad_norm": 0.3996012626147995,
+      "learning_rate": 8.973493017001785e-06,
+      "loss": 0.6504,
+      "step": 1342
+    },
+    {
+      "epoch": 0.22055714080430275,
+      "grad_norm": 0.401605897362349,
+      "learning_rate": 8.973442165588993e-06,
+      "loss": 0.6564,
+      "step": 1343
+    },
+    {
+      "epoch": 0.2207213680126455,
+      "grad_norm": 0.4211815055879324,
+      "learning_rate": 8.973391265590215e-06,
+      "loss": 0.6377,
+      "step": 1344
+    },
+    {
+      "epoch": 0.22088559522098825,
+      "grad_norm": 0.4147590208542743,
+      "learning_rate": 8.97334031700601e-06,
+      "loss": 0.6358,
+      "step": 1345
+    },
+    {
+      "epoch": 0.22104982242933097,
+      "grad_norm": 0.4041410755841317,
+      "learning_rate": 8.973289319836924e-06,
+      "loss": 0.6345,
+      "step": 1346
+    },
+    {
+      "epoch": 0.22121404963767372,
+      "grad_norm": 0.4419733735874853,
+      "learning_rate": 8.973238274083517e-06,
+      "loss": 0.6309,
+      "step": 1347
+    },
+    {
+      "epoch": 0.22137827684601646,
+      "grad_norm": 0.4438396420085765,
+      "learning_rate": 8.973187179746341e-06,
+      "loss": 0.6456,
+      "step": 1348
+    },
+    {
+      "epoch": 0.2215425040543592,
+      "grad_norm": 0.40548180198977596,
+      "learning_rate": 8.973136036825952e-06,
+      "loss": 0.6448,
+      "step": 1349
+    },
+    {
+      "epoch": 0.22170673126270196,
+      "grad_norm": 0.3776693195760448,
+      "learning_rate": 8.973084845322905e-06,
+      "loss": 0.6267,
+      "step": 1350
+    },
+    {
+      "epoch": 0.22187095847104468,
+      "grad_norm": 0.3914134341854077,
+      "learning_rate": 8.973033605237754e-06,
+      "loss": 0.6064,
+      "step": 1351
+    },
+    {
+      "epoch": 0.22203518567938743,
+      "grad_norm": 0.4008878109833888,
+      "learning_rate": 8.972982316571059e-06,
+      "loss": 0.6412,
+      "step": 1352
+    },
+    {
+      "epoch": 0.22219941288773018,
+      "grad_norm": 0.3947443304069219,
+      "learning_rate": 8.972930979323373e-06,
+      "loss": 0.6296,
+      "step": 1353
+    },
+    {
+      "epoch": 0.22236364009607293,
+      "grad_norm": 0.40428731395639134,
+      "learning_rate": 8.972879593495257e-06,
+      "loss": 0.6468,
+      "step": 1354
+    },
+    {
+      "epoch": 0.22252786730441565,
+      "grad_norm": 0.42559973840758325,
+      "learning_rate": 8.972828159087268e-06,
+      "loss": 0.636,
+      "step": 1355
+    },
+    {
+      "epoch": 0.2226920945127584,
+      "grad_norm": 0.38558193013481934,
+      "learning_rate": 8.972776676099965e-06,
+      "loss": 0.6267,
+      "step": 1356
+    },
+    {
+      "epoch": 0.22285632172110115,
+      "grad_norm": 0.3942742258512122,
+      "learning_rate": 8.972725144533905e-06,
+      "loss": 0.6095,
+      "step": 1357
+    },
+    {
+      "epoch": 0.2230205489294439,
+      "grad_norm": 0.39813829783312527,
+      "learning_rate": 8.972673564389651e-06,
+      "loss": 0.6206,
+      "step": 1358
+    },
+    {
+      "epoch": 0.22318477613778662,
+      "grad_norm": 0.424281039810817,
+      "learning_rate": 8.972621935667763e-06,
+      "loss": 0.6455,
+      "step": 1359
+    },
+    {
+      "epoch": 0.22334900334612937,
+      "grad_norm": 0.3908176395636074,
+      "learning_rate": 8.972570258368797e-06,
+      "loss": 0.6242,
+      "step": 1360
+    },
+    {
+      "epoch": 0.22351323055447211,
+      "grad_norm": 0.4052328537533125,
+      "learning_rate": 8.972518532493319e-06,
+      "loss": 0.6209,
+      "step": 1361
+    },
+    {
+      "epoch": 0.22367745776281486,
+      "grad_norm": 0.39202858379796474,
+      "learning_rate": 8.97246675804189e-06,
+      "loss": 0.6256,
+      "step": 1362
+    },
+    {
+      "epoch": 0.22384168497115758,
+      "grad_norm": 0.3800992157787325,
+      "learning_rate": 8.97241493501507e-06,
+      "loss": 0.6232,
+      "step": 1363
+    },
+    {
+      "epoch": 0.22400591217950033,
+      "grad_norm": 0.3889099579798512,
+      "learning_rate": 8.972363063413424e-06,
+      "loss": 0.642,
+      "step": 1364
+    },
+    {
+      "epoch": 0.22417013938784308,
+      "grad_norm": 0.43843337554692713,
+      "learning_rate": 8.972311143237516e-06,
+      "loss": 0.6309,
+      "step": 1365
+    },
+    {
+      "epoch": 0.22433436659618583,
+      "grad_norm": 0.38555050369799626,
+      "learning_rate": 8.972259174487908e-06,
+      "loss": 0.6175,
+      "step": 1366
+    },
+    {
+      "epoch": 0.22449859380452858,
+      "grad_norm": 0.3954594559012571,
+      "learning_rate": 8.972207157165167e-06,
+      "loss": 0.6332,
+      "step": 1367
+    },
+    {
+      "epoch": 0.2246628210128713,
+      "grad_norm": 0.397471975775458,
+      "learning_rate": 8.972155091269854e-06,
+      "loss": 0.6291,
+      "step": 1368
+    },
+    {
+      "epoch": 0.22482704822121405,
+      "grad_norm": 0.3954234981146683,
+      "learning_rate": 8.972102976802537e-06,
+      "loss": 0.6497,
+      "step": 1369
+    },
+    {
+      "epoch": 0.2249912754295568,
+      "grad_norm": 0.4398435558984372,
+      "learning_rate": 8.972050813763783e-06,
+      "loss": 0.6282,
+      "step": 1370
+    },
+    {
+      "epoch": 0.22515550263789955,
+      "grad_norm": 1.1371437397465023,
+      "learning_rate": 8.971998602154156e-06,
+      "loss": 0.6283,
+      "step": 1371
+    },
+    {
+      "epoch": 0.22531972984624227,
+      "grad_norm": 0.42734572284380834,
+      "learning_rate": 8.971946341974225e-06,
+      "loss": 0.6498,
+      "step": 1372
+    },
+    {
+      "epoch": 0.22548395705458502,
+      "grad_norm": 0.372006683716092,
+      "learning_rate": 8.971894033224556e-06,
+      "loss": 0.6317,
+      "step": 1373
+    },
+    {
+      "epoch": 0.22564818426292776,
+      "grad_norm": 0.5365999450195464,
+      "learning_rate": 8.97184167590572e-06,
+      "loss": 0.6398,
+      "step": 1374
+    },
+    {
+      "epoch": 0.2258124114712705,
+      "grad_norm": 0.4513109754424501,
+      "learning_rate": 8.971789270018282e-06,
+      "loss": 0.6516,
+      "step": 1375
+    },
+    {
+      "epoch": 0.22597663867961323,
+      "grad_norm": 0.3876777628138284,
+      "learning_rate": 8.971736815562813e-06,
+      "loss": 0.6239,
+      "step": 1376
+    },
+    {
+      "epoch": 0.22614086588795598,
+      "grad_norm": 0.39030172173320044,
+      "learning_rate": 8.971684312539884e-06,
+      "loss": 0.6362,
+      "step": 1377
+    },
+    {
+      "epoch": 0.22630509309629873,
+      "grad_norm": 0.5022741874530742,
+      "learning_rate": 8.971631760950062e-06,
+      "loss": 0.6391,
+      "step": 1378
+    },
+    {
+      "epoch": 0.22646932030464148,
+      "grad_norm": 0.3966665973762089,
+      "learning_rate": 8.971579160793921e-06,
+      "loss": 0.6279,
+      "step": 1379
+    },
+    {
+      "epoch": 0.2266335475129842,
+      "grad_norm": 0.46552566540421336,
+      "learning_rate": 8.971526512072028e-06,
+      "loss": 0.633,
+      "step": 1380
+    },
+    {
+      "epoch": 0.22679777472132695,
+      "grad_norm": 0.3887481021652706,
+      "learning_rate": 8.971473814784961e-06,
+      "loss": 0.6316,
+      "step": 1381
+    },
+    {
+      "epoch": 0.2269620019296697,
+      "grad_norm": 0.3981403897977426,
+      "learning_rate": 8.971421068933289e-06,
+      "loss": 0.6297,
+      "step": 1382
+    },
+    {
+      "epoch": 0.22712622913801245,
+      "grad_norm": 0.3760925511936663,
+      "learning_rate": 8.971368274517584e-06,
+      "loss": 0.6518,
+      "step": 1383
+    },
+    {
+      "epoch": 0.22729045634635517,
+      "grad_norm": 0.41634568057723575,
+      "learning_rate": 8.971315431538419e-06,
+      "loss": 0.6507,
+      "step": 1384
+    },
+    {
+      "epoch": 0.22745468355469792,
+      "grad_norm": 0.3767648106691863,
+      "learning_rate": 8.971262539996371e-06,
+      "loss": 0.6219,
+      "step": 1385
+    },
+    {
+      "epoch": 0.22761891076304067,
+      "grad_norm": 0.37280796908592456,
+      "learning_rate": 8.971209599892012e-06,
+      "loss": 0.63,
+      "step": 1386
+    },
+    {
+      "epoch": 0.22778313797138341,
+      "grad_norm": 0.39063293180757414,
+      "learning_rate": 8.971156611225918e-06,
+      "loss": 0.6354,
+      "step": 1387
+    },
+    {
+      "epoch": 0.22794736517972616,
+      "grad_norm": 0.4130871584775029,
+      "learning_rate": 8.971103573998664e-06,
+      "loss": 0.6215,
+      "step": 1388
+    },
+    {
+      "epoch": 0.22811159238806888,
+      "grad_norm": 0.3697250690243639,
+      "learning_rate": 8.971050488210827e-06,
+      "loss": 0.6509,
+      "step": 1389
+    },
+    {
+      "epoch": 0.22827581959641163,
+      "grad_norm": 0.40985315624852925,
+      "learning_rate": 8.97099735386298e-06,
+      "loss": 0.6223,
+      "step": 1390
+    },
+    {
+      "epoch": 0.22844004680475438,
+      "grad_norm": 0.41753388559467336,
+      "learning_rate": 8.970944170955705e-06,
+      "loss": 0.6308,
+      "step": 1391
+    },
+    {
+      "epoch": 0.22860427401309713,
+      "grad_norm": 0.37172418439885546,
+      "learning_rate": 8.970890939489577e-06,
+      "loss": 0.6378,
+      "step": 1392
+    },
+    {
+      "epoch": 0.22876850122143985,
+      "grad_norm": 0.3799761140290008,
+      "learning_rate": 8.970837659465175e-06,
+      "loss": 0.62,
+      "step": 1393
+    },
+    {
+      "epoch": 0.2289327284297826,
+      "grad_norm": 0.3637974123542842,
+      "learning_rate": 8.970784330883077e-06,
+      "loss": 0.6414,
+      "step": 1394
+    },
+    {
+      "epoch": 0.22909695563812535,
+      "grad_norm": 0.39145712168539487,
+      "learning_rate": 8.970730953743865e-06,
+      "loss": 0.6259,
+      "step": 1395
+    },
+    {
+      "epoch": 0.2292611828464681,
+      "grad_norm": 0.3747433458407329,
+      "learning_rate": 8.970677528048112e-06,
+      "loss": 0.629,
+      "step": 1396
+    },
+    {
+      "epoch": 0.22942541005481082,
+      "grad_norm": 0.40593368367010124,
+      "learning_rate": 8.970624053796405e-06,
+      "loss": 0.6545,
+      "step": 1397
+    },
+    {
+      "epoch": 0.22958963726315357,
+      "grad_norm": 0.3814870268904145,
+      "learning_rate": 8.970570530989322e-06,
+      "loss": 0.6356,
+      "step": 1398
+    },
+    {
+      "epoch": 0.22975386447149632,
+      "grad_norm": 0.37393431918398545,
+      "learning_rate": 8.970516959627445e-06,
+      "loss": 0.6149,
+      "step": 1399
+    },
+    {
+      "epoch": 0.22991809167983907,
+      "grad_norm": 0.4234768187758134,
+      "learning_rate": 8.970463339711354e-06,
+      "loss": 0.6213,
+      "step": 1400
+    },
+    {
+      "epoch": 0.23008231888818179,
+      "grad_norm": 0.39010381977649883,
+      "learning_rate": 8.970409671241635e-06,
+      "loss": 0.6239,
+      "step": 1401
+    },
+    {
+      "epoch": 0.23024654609652453,
+      "grad_norm": 0.4059667118236035,
+      "learning_rate": 8.970355954218866e-06,
+      "loss": 0.6457,
+      "step": 1402
+    },
+    {
+      "epoch": 0.23041077330486728,
+      "grad_norm": 0.36347685311176176,
+      "learning_rate": 8.970302188643634e-06,
+      "loss": 0.6366,
+      "step": 1403
+    },
+    {
+      "epoch": 0.23057500051321003,
+      "grad_norm": 0.4016516068154081,
+      "learning_rate": 8.970248374516523e-06,
+      "loss": 0.6155,
+      "step": 1404
+    },
+    {
+      "epoch": 0.23073922772155278,
+      "grad_norm": 0.3738165759888122,
+      "learning_rate": 8.970194511838116e-06,
+      "loss": 0.6297,
+      "step": 1405
+    },
+    {
+      "epoch": 0.2309034549298955,
+      "grad_norm": 0.38775548720178016,
+      "learning_rate": 8.970140600608998e-06,
+      "loss": 0.631,
+      "step": 1406
+    },
+    {
+      "epoch": 0.23106768213823825,
+      "grad_norm": 0.36651163646368323,
+      "learning_rate": 8.970086640829755e-06,
+      "loss": 0.6235,
+      "step": 1407
+    },
+    {
+      "epoch": 0.231231909346581,
+      "grad_norm": 0.388177161793234,
+      "learning_rate": 8.970032632500974e-06,
+      "loss": 0.6447,
+      "step": 1408
+    },
+    {
+      "epoch": 0.23139613655492375,
+      "grad_norm": 0.38342963166847494,
+      "learning_rate": 8.96997857562324e-06,
+      "loss": 0.6446,
+      "step": 1409
+    },
+    {
+      "epoch": 0.23156036376326647,
+      "grad_norm": 0.35965242348385945,
+      "learning_rate": 8.969924470197141e-06,
+      "loss": 0.6211,
+      "step": 1410
+    },
+    {
+      "epoch": 0.23172459097160922,
+      "grad_norm": 0.40581904246250583,
+      "learning_rate": 8.969870316223264e-06,
+      "loss": 0.6089,
+      "step": 1411
+    },
+    {
+      "epoch": 0.23188881817995197,
+      "grad_norm": 0.36406413748044436,
+      "learning_rate": 8.969816113702198e-06,
+      "loss": 0.6302,
+      "step": 1412
+    },
+    {
+      "epoch": 0.23205304538829472,
+      "grad_norm": 0.36016891310093596,
+      "learning_rate": 8.969761862634532e-06,
+      "loss": 0.6301,
+      "step": 1413
+    },
+    {
+      "epoch": 0.23221727259663744,
+      "grad_norm": 0.38497253061331427,
+      "learning_rate": 8.969707563020854e-06,
+      "loss": 0.6381,
+      "step": 1414
+    },
+    {
+      "epoch": 0.23238149980498018,
+      "grad_norm": 0.3728601291907777,
+      "learning_rate": 8.969653214861753e-06,
+      "loss": 0.6181,
+      "step": 1415
+    },
+    {
+      "epoch": 0.23254572701332293,
+      "grad_norm": 0.3667182768200828,
+      "learning_rate": 8.969598818157824e-06,
+      "loss": 0.6175,
+      "step": 1416
+    },
+    {
+      "epoch": 0.23270995422166568,
+      "grad_norm": 0.40991591837455477,
+      "learning_rate": 8.969544372909651e-06,
+      "loss": 0.6357,
+      "step": 1417
+    },
+    {
+      "epoch": 0.2328741814300084,
+      "grad_norm": 0.3830261589112887,
+      "learning_rate": 8.96948987911783e-06,
+      "loss": 0.6263,
+      "step": 1418
+    },
+    {
+      "epoch": 0.23303840863835115,
+      "grad_norm": 0.37342955461632715,
+      "learning_rate": 8.969435336782951e-06,
+      "loss": 0.651,
+      "step": 1419
+    },
+    {
+      "epoch": 0.2332026358466939,
+      "grad_norm": 0.38302602639530275,
+      "learning_rate": 8.969380745905607e-06,
+      "loss": 0.6147,
+      "step": 1420
+    },
+    {
+      "epoch": 0.23336686305503665,
+      "grad_norm": 0.3801283833275303,
+      "learning_rate": 8.969326106486392e-06,
+      "loss": 0.5939,
+      "step": 1421
+    },
+    {
+      "epoch": 0.2335310902633794,
+      "grad_norm": 0.3829133639241409,
+      "learning_rate": 8.969271418525897e-06,
+      "loss": 0.6029,
+      "step": 1422
+    },
+    {
+      "epoch": 0.23369531747172212,
+      "grad_norm": 0.37812529413156654,
+      "learning_rate": 8.969216682024718e-06,
+      "loss": 0.6403,
+      "step": 1423
+    },
+    {
+      "epoch": 0.23385954468006487,
+      "grad_norm": 0.37094348293191975,
+      "learning_rate": 8.969161896983448e-06,
+      "loss": 0.6171,
+      "step": 1424
+    },
+    {
+      "epoch": 0.23402377188840762,
+      "grad_norm": 0.35854811021498156,
+      "learning_rate": 8.969107063402682e-06,
+      "loss": 0.6213,
+      "step": 1425
+    },
+    {
+      "epoch": 0.23418799909675037,
+      "grad_norm": 0.3600943551683208,
+      "learning_rate": 8.969052181283017e-06,
+      "loss": 0.6146,
+      "step": 1426
+    },
+    {
+      "epoch": 0.2343522263050931,
+      "grad_norm": 0.37828891860958147,
+      "learning_rate": 8.968997250625048e-06,
+      "loss": 0.6205,
+      "step": 1427
+    },
+    {
+      "epoch": 0.23451645351343584,
+      "grad_norm": 0.3753629445499872,
+      "learning_rate": 8.968942271429375e-06,
+      "loss": 0.6267,
+      "step": 1428
+    },
+    {
+      "epoch": 0.23468068072177858,
+      "grad_norm": 0.35650819622201213,
+      "learning_rate": 8.968887243696589e-06,
+      "loss": 0.6282,
+      "step": 1429
+    },
+    {
+      "epoch": 0.23484490793012133,
+      "grad_norm": 0.38875352987788536,
+      "learning_rate": 8.96883216742729e-06,
+      "loss": 0.6156,
+      "step": 1430
+    },
+    {
+      "epoch": 0.23500913513846405,
+      "grad_norm": 0.3817434000607439,
+      "learning_rate": 8.96877704262208e-06,
+      "loss": 0.6388,
+      "step": 1431
+    },
+    {
+      "epoch": 0.2351733623468068,
+      "grad_norm": 0.3873016030270038,
+      "learning_rate": 8.968721869281552e-06,
+      "loss": 0.6163,
+      "step": 1432
+    },
+    {
+      "epoch": 0.23533758955514955,
+      "grad_norm": 0.4079351706434482,
+      "learning_rate": 8.96866664740631e-06,
+      "loss": 0.6485,
+      "step": 1433
+    },
+    {
+      "epoch": 0.2355018167634923,
+      "grad_norm": 0.38317014691327517,
+      "learning_rate": 8.968611376996949e-06,
+      "loss": 0.5968,
+      "step": 1434
+    },
+    {
+      "epoch": 0.23566604397183502,
+      "grad_norm": 0.3862892534452918,
+      "learning_rate": 8.968556058054075e-06,
+      "loss": 0.6228,
+      "step": 1435
+    },
+    {
+      "epoch": 0.23583027118017777,
+      "grad_norm": 0.3513593263777787,
+      "learning_rate": 8.968500690578285e-06,
+      "loss": 0.6188,
+      "step": 1436
+    },
+    {
+      "epoch": 0.23599449838852052,
+      "grad_norm": 0.36867544120784373,
+      "learning_rate": 8.968445274570179e-06,
+      "loss": 0.6263,
+      "step": 1437
+    },
+    {
+      "epoch": 0.23615872559686327,
+      "grad_norm": 0.36938216239379995,
+      "learning_rate": 8.968389810030362e-06,
+      "loss": 0.6164,
+      "step": 1438
+    },
+    {
+      "epoch": 0.23632295280520602,
+      "grad_norm": 0.3727449831108178,
+      "learning_rate": 8.968334296959436e-06,
+      "loss": 0.6085,
+      "step": 1439
+    },
+    {
+      "epoch": 0.23648718001354874,
+      "grad_norm": 0.3662499381596673,
+      "learning_rate": 8.968278735358003e-06,
+      "loss": 0.6023,
+      "step": 1440
+    },
+    {
+      "epoch": 0.23665140722189149,
+      "grad_norm": 0.40628718930364516,
+      "learning_rate": 8.968223125226667e-06,
+      "loss": 0.6171,
+      "step": 1441
+    },
+    {
+      "epoch": 0.23681563443023423,
+      "grad_norm": 0.37037082262637283,
+      "learning_rate": 8.96816746656603e-06,
+      "loss": 0.6368,
+      "step": 1442
+    },
+    {
+      "epoch": 0.23697986163857698,
+      "grad_norm": 0.3549645143515296,
+      "learning_rate": 8.968111759376699e-06,
+      "loss": 0.6334,
+      "step": 1443
+    },
+    {
+      "epoch": 0.2371440888469197,
+      "grad_norm": 0.3798618217383637,
+      "learning_rate": 8.96805600365928e-06,
+      "loss": 0.6191,
+      "step": 1444
+    },
+    {
+      "epoch": 0.23730831605526245,
+      "grad_norm": 0.36686801549322573,
+      "learning_rate": 8.968000199414376e-06,
+      "loss": 0.6356,
+      "step": 1445
+    },
+    {
+      "epoch": 0.2374725432636052,
+      "grad_norm": 0.35942437209207495,
+      "learning_rate": 8.967944346642592e-06,
+      "loss": 0.6144,
+      "step": 1446
+    },
+    {
+      "epoch": 0.23763677047194795,
+      "grad_norm": 0.36600431669189964,
+      "learning_rate": 8.96788844534454e-06,
+      "loss": 0.6198,
+      "step": 1447
+    },
+    {
+      "epoch": 0.23780099768029067,
+      "grad_norm": 0.367127385380099,
+      "learning_rate": 8.967832495520822e-06,
+      "loss": 0.6204,
+      "step": 1448
+    },
+    {
+      "epoch": 0.23796522488863342,
+      "grad_norm": 0.37739288988592845,
+      "learning_rate": 8.967776497172046e-06,
+      "loss": 0.6286,
+      "step": 1449
+    },
+    {
+      "epoch": 0.23812945209697617,
+      "grad_norm": 0.3694809371169283,
+      "learning_rate": 8.967720450298822e-06,
+      "loss": 0.6329,
+      "step": 1450
+    },
+    {
+      "epoch": 0.23829367930531892,
+      "grad_norm": 0.39873053139399506,
+      "learning_rate": 8.967664354901759e-06,
+      "loss": 0.607,
+      "step": 1451
+    },
+    {
+      "epoch": 0.23845790651366164,
+      "grad_norm": 0.37174093067137187,
+      "learning_rate": 8.967608210981466e-06,
+      "loss": 0.6297,
+      "step": 1452
+    },
+    {
+      "epoch": 0.2386221337220044,
+      "grad_norm": 0.3655257783754846,
+      "learning_rate": 8.967552018538552e-06,
+      "loss": 0.614,
+      "step": 1453
+    },
+    {
+      "epoch": 0.23878636093034714,
+      "grad_norm": 0.3486334444422297,
+      "learning_rate": 8.967495777573626e-06,
+      "loss": 0.612,
+      "step": 1454
+    },
+    {
+      "epoch": 0.23895058813868988,
+      "grad_norm": 0.38653414135810354,
+      "learning_rate": 8.967439488087303e-06,
+      "loss": 0.6373,
+      "step": 1455
+    },
+    {
+      "epoch": 0.23911481534703263,
+      "grad_norm": 0.4166657864417276,
+      "learning_rate": 8.967383150080191e-06,
+      "loss": 0.6172,
+      "step": 1456
+    },
+    {
+      "epoch": 0.23927904255537535,
+      "grad_norm": 0.36625495836800076,
+      "learning_rate": 8.967326763552901e-06,
+      "loss": 0.6105,
+      "step": 1457
+    },
+    {
+      "epoch": 0.2394432697637181,
+      "grad_norm": 0.35693092319100034,
+      "learning_rate": 8.96727032850605e-06,
+      "loss": 0.6361,
+      "step": 1458
+    },
+    {
+      "epoch": 0.23960749697206085,
+      "grad_norm": 0.4668329077344547,
+      "learning_rate": 8.967213844940246e-06,
+      "loss": 0.6507,
+      "step": 1459
+    },
+    {
+      "epoch": 0.2397717241804036,
+      "grad_norm": 0.36293669801834133,
+      "learning_rate": 8.967157312856105e-06,
+      "loss": 0.629,
+      "step": 1460
+    },
+    {
+      "epoch": 0.23993595138874632,
+      "grad_norm": 0.39226552096477374,
+      "learning_rate": 8.96710073225424e-06,
+      "loss": 0.6246,
+      "step": 1461
+    },
+    {
+      "epoch": 0.24010017859708907,
+      "grad_norm": 0.3767099803854531,
+      "learning_rate": 8.967044103135266e-06,
+      "loss": 0.6157,
+      "step": 1462
+    },
+    {
+      "epoch": 0.24026440580543182,
+      "grad_norm": 0.3738464526239834,
+      "learning_rate": 8.966987425499798e-06,
+      "loss": 0.6338,
+      "step": 1463
+    },
+    {
+      "epoch": 0.24042863301377457,
+      "grad_norm": 0.3924322055790276,
+      "learning_rate": 8.966930699348453e-06,
+      "loss": 0.6097,
+      "step": 1464
+    },
+    {
+      "epoch": 0.2405928602221173,
+      "grad_norm": 0.40128496672548014,
+      "learning_rate": 8.966873924681845e-06,
+      "loss": 0.6409,
+      "step": 1465
+    },
+    {
+      "epoch": 0.24075708743046004,
+      "grad_norm": 0.35113853348291707,
+      "learning_rate": 8.96681710150059e-06,
+      "loss": 0.6347,
+      "step": 1466
+    },
+    {
+      "epoch": 0.24092131463880279,
+      "grad_norm": 0.3820894843457969,
+      "learning_rate": 8.96676022980531e-06,
+      "loss": 0.6058,
+      "step": 1467
+    },
+    {
+      "epoch": 0.24108554184714553,
+      "grad_norm": 0.36199118035878114,
+      "learning_rate": 8.966703309596615e-06,
+      "loss": 0.6356,
+      "step": 1468
+    },
+    {
+      "epoch": 0.24124976905548826,
+      "grad_norm": 0.3540001836120606,
+      "learning_rate": 8.966646340875129e-06,
+      "loss": 0.619,
+      "step": 1469
+    },
+    {
+      "epoch": 0.241413996263831,
+      "grad_norm": 0.38920950337400373,
+      "learning_rate": 8.96658932364147e-06,
+      "loss": 0.6122,
+      "step": 1470
+    },
+    {
+      "epoch": 0.24157822347217375,
+      "grad_norm": 0.35862870218740506,
+      "learning_rate": 8.966532257896256e-06,
+      "loss": 0.5999,
+      "step": 1471
+    },
+    {
+      "epoch": 0.2417424506805165,
+      "grad_norm": 0.40346544693249303,
+      "learning_rate": 8.966475143640108e-06,
+      "loss": 0.6389,
+      "step": 1472
+    },
+    {
+      "epoch": 0.24190667788885925,
+      "grad_norm": 0.35802924018053983,
+      "learning_rate": 8.966417980873644e-06,
+      "loss": 0.656,
+      "step": 1473
+    },
+    {
+      "epoch": 0.24207090509720197,
+      "grad_norm": 0.35416797749163403,
+      "learning_rate": 8.966360769597487e-06,
+      "loss": 0.5871,
+      "step": 1474
+    },
+    {
+      "epoch": 0.24223513230554472,
+      "grad_norm": 0.38394310761980127,
+      "learning_rate": 8.966303509812259e-06,
+      "loss": 0.6278,
+      "step": 1475
+    },
+    {
+      "epoch": 0.24239935951388747,
+      "grad_norm": 0.37542608512409903,
+      "learning_rate": 8.966246201518577e-06,
+      "loss": 0.6451,
+      "step": 1476
+    },
+    {
+      "epoch": 0.24256358672223022,
+      "grad_norm": 0.50798389472191,
+      "learning_rate": 8.96618884471707e-06,
+      "loss": 0.6169,
+      "step": 1477
+    },
+    {
+      "epoch": 0.24272781393057294,
+      "grad_norm": 0.37174062893917387,
+      "learning_rate": 8.966131439408357e-06,
+      "loss": 0.6373,
+      "step": 1478
+    },
+    {
+      "epoch": 0.2428920411389157,
+      "grad_norm": 0.4209079011176253,
+      "learning_rate": 8.966073985593063e-06,
+      "loss": 0.6235,
+      "step": 1479
+    },
+    {
+      "epoch": 0.24305626834725844,
+      "grad_norm": 0.41783270170305736,
+      "learning_rate": 8.966016483271813e-06,
+      "loss": 0.6291,
+      "step": 1480
+    },
+    {
+      "epoch": 0.24322049555560118,
+      "grad_norm": 0.3578979030166597,
+      "learning_rate": 8.965958932445228e-06,
+      "loss": 0.6075,
+      "step": 1481
+    },
+    {
+      "epoch": 0.2433847227639439,
+      "grad_norm": 0.37588742363710526,
+      "learning_rate": 8.965901333113936e-06,
+      "loss": 0.6172,
+      "step": 1482
+    },
+    {
+      "epoch": 0.24354894997228665,
+      "grad_norm": 0.43019135611502246,
+      "learning_rate": 8.965843685278561e-06,
+      "loss": 0.6152,
+      "step": 1483
+    },
+    {
+      "epoch": 0.2437131771806294,
+      "grad_norm": 0.3713307328258515,
+      "learning_rate": 8.965785988939728e-06,
+      "loss": 0.633,
+      "step": 1484
+    },
+    {
+      "epoch": 0.24387740438897215,
+      "grad_norm": 0.38538363095385886,
+      "learning_rate": 8.96572824409807e-06,
+      "loss": 0.6103,
+      "step": 1485
+    },
+    {
+      "epoch": 0.24404163159731487,
+      "grad_norm": 0.42959307234492405,
+      "learning_rate": 8.965670450754205e-06,
+      "loss": 0.6159,
+      "step": 1486
+    },
+    {
+      "epoch": 0.24420585880565762,
+      "grad_norm": 0.3677298238888804,
+      "learning_rate": 8.965612608908767e-06,
+      "loss": 0.6242,
+      "step": 1487
+    },
+    {
+      "epoch": 0.24437008601400037,
+      "grad_norm": 0.3700124245084724,
+      "learning_rate": 8.965554718562383e-06,
+      "loss": 0.632,
+      "step": 1488
+    },
+    {
+      "epoch": 0.24453431322234312,
+      "grad_norm": 0.37162697278412393,
+      "learning_rate": 8.965496779715681e-06,
+      "loss": 0.6291,
+      "step": 1489
+    },
+    {
+      "epoch": 0.24469854043068584,
+      "grad_norm": 0.5932246129868997,
+      "learning_rate": 8.965438792369291e-06,
+      "loss": 0.6218,
+      "step": 1490
+    },
+    {
+      "epoch": 0.2448627676390286,
+      "grad_norm": 0.35796971024046653,
+      "learning_rate": 8.965380756523842e-06,
+      "loss": 0.6301,
+      "step": 1491
+    },
+    {
+      "epoch": 0.24502699484737134,
+      "grad_norm": 0.3888508026013573,
+      "learning_rate": 8.965322672179964e-06,
+      "loss": 0.607,
+      "step": 1492
+    },
+    {
+      "epoch": 0.2451912220557141,
+      "grad_norm": 0.3782937892080753,
+      "learning_rate": 8.96526453933829e-06,
+      "loss": 0.6456,
+      "step": 1493
+    },
+    {
+      "epoch": 0.24535544926405684,
+      "grad_norm": 0.3799106905153972,
+      "learning_rate": 8.965206357999449e-06,
+      "loss": 0.6294,
+      "step": 1494
+    },
+    {
+      "epoch": 0.24551967647239956,
+      "grad_norm": 0.39762851482762696,
+      "learning_rate": 8.965148128164074e-06,
+      "loss": 0.6399,
+      "step": 1495
+    },
+    {
+      "epoch": 0.2456839036807423,
+      "grad_norm": 0.4416226219612057,
+      "learning_rate": 8.965089849832796e-06,
+      "loss": 0.6251,
+      "step": 1496
+    },
+    {
+      "epoch": 0.24584813088908505,
+      "grad_norm": 0.3889843392209306,
+      "learning_rate": 8.96503152300625e-06,
+      "loss": 0.6497,
+      "step": 1497
+    },
+    {
+      "epoch": 0.2460123580974278,
+      "grad_norm": 0.35215446037387954,
+      "learning_rate": 8.964973147685069e-06,
+      "loss": 0.633,
+      "step": 1498
+    },
+    {
+      "epoch": 0.24617658530577052,
+      "grad_norm": 0.45650337654446665,
+      "learning_rate": 8.964914723869886e-06,
+      "loss": 0.6042,
+      "step": 1499
+    },
+    {
+      "epoch": 0.24634081251411327,
+      "grad_norm": 0.36286389615728926,
+      "learning_rate": 8.964856251561336e-06,
+      "loss": 0.6341,
+      "step": 1500
+    },
+    {
+      "epoch": 0.24650503972245602,
+      "grad_norm": 0.3593756064119613,
+      "learning_rate": 8.964797730760055e-06,
+      "loss": 0.6197,
+      "step": 1501
+    },
+    {
+      "epoch": 0.24666926693079877,
+      "grad_norm": 0.3771052035230812,
+      "learning_rate": 8.964739161466678e-06,
+      "loss": 0.6242,
+      "step": 1502
+    },
+    {
+      "epoch": 0.2468334941391415,
+      "grad_norm": 0.3605637264989002,
+      "learning_rate": 8.96468054368184e-06,
+      "loss": 0.6094,
+      "step": 1503
+    },
+    {
+      "epoch": 0.24699772134748424,
+      "grad_norm": 0.3626333911931281,
+      "learning_rate": 8.964621877406181e-06,
+      "loss": 0.6348,
+      "step": 1504
+    },
+    {
+      "epoch": 0.247161948555827,
+      "grad_norm": 0.38088499252044117,
+      "learning_rate": 8.964563162640334e-06,
+      "loss": 0.6268,
+      "step": 1505
+    },
+    {
+      "epoch": 0.24732617576416974,
+      "grad_norm": 0.36603008668404546,
+      "learning_rate": 8.964504399384938e-06,
+      "loss": 0.6265,
+      "step": 1506
+    },
+    {
+      "epoch": 0.24749040297251246,
+      "grad_norm": 0.3627778415298624,
+      "learning_rate": 8.964445587640633e-06,
+      "loss": 0.6557,
+      "step": 1507
+    },
+    {
+      "epoch": 0.2476546301808552,
+      "grad_norm": 0.3976038856166393,
+      "learning_rate": 8.964386727408055e-06,
+      "loss": 0.6295,
+      "step": 1508
+    },
+    {
+      "epoch": 0.24781885738919796,
+      "grad_norm": 0.35128896141586324,
+      "learning_rate": 8.964327818687847e-06,
+      "loss": 0.629,
+      "step": 1509
+    },
+    {
+      "epoch": 0.2479830845975407,
+      "grad_norm": 0.3408244422741637,
+      "learning_rate": 8.964268861480645e-06,
+      "loss": 0.614,
+      "step": 1510
+    },
+    {
+      "epoch": 0.24814731180588345,
+      "grad_norm": 0.35929482815511576,
+      "learning_rate": 8.964209855787091e-06,
+      "loss": 0.5934,
+      "step": 1511
+    },
+    {
+      "epoch": 0.24831153901422617,
+      "grad_norm": 0.46183852703188355,
+      "learning_rate": 8.964150801607825e-06,
+      "loss": 0.6054,
+      "step": 1512
+    },
+    {
+      "epoch": 0.24847576622256892,
+      "grad_norm": 0.37059564147077584,
+      "learning_rate": 8.96409169894349e-06,
+      "loss": 0.6184,
+      "step": 1513
+    },
+    {
+      "epoch": 0.24863999343091167,
+      "grad_norm": 0.3629530929389863,
+      "learning_rate": 8.964032547794728e-06,
+      "loss": 0.6242,
+      "step": 1514
+    },
+    {
+      "epoch": 0.24880422063925442,
+      "grad_norm": 0.36798673186434855,
+      "learning_rate": 8.96397334816218e-06,
+      "loss": 0.6034,
+      "step": 1515
+    },
+    {
+      "epoch": 0.24896844784759714,
+      "grad_norm": 0.3943059155021281,
+      "learning_rate": 8.963914100046489e-06,
+      "loss": 0.6471,
+      "step": 1516
+    },
+    {
+      "epoch": 0.2491326750559399,
+      "grad_norm": 0.4348402454752103,
+      "learning_rate": 8.963854803448301e-06,
+      "loss": 0.5869,
+      "step": 1517
+    },
+    {
+      "epoch": 0.24929690226428264,
+      "grad_norm": 0.400012642981447,
+      "learning_rate": 8.963795458368254e-06,
+      "loss": 0.6282,
+      "step": 1518
+    },
+    {
+      "epoch": 0.2494611294726254,
+      "grad_norm": 0.39046182025528064,
+      "learning_rate": 8.963736064807e-06,
+      "loss": 0.6466,
+      "step": 1519
+    },
+    {
+      "epoch": 0.2496253566809681,
+      "grad_norm": 0.35844392595007163,
+      "learning_rate": 8.963676622765179e-06,
+      "loss": 0.6256,
+      "step": 1520
+    },
+    {
+      "epoch": 0.24978958388931086,
+      "grad_norm": 0.35242439852874236,
+      "learning_rate": 8.963617132243439e-06,
+      "loss": 0.5967,
+      "step": 1521
+    },
+    {
+      "epoch": 0.2499538110976536,
+      "grad_norm": 0.35656035849610795,
+      "learning_rate": 8.963557593242424e-06,
+      "loss": 0.6198,
+      "step": 1522
+    },
+    {
+      "epoch": 0.2501180383059963,
+      "grad_norm": 0.38132355048447236,
+      "learning_rate": 8.963498005762783e-06,
+      "loss": 0.6431,
+      "step": 1523
+    },
+    {
+      "epoch": 0.2502822655143391,
+      "grad_norm": 0.3537836522540743,
+      "learning_rate": 8.963438369805163e-06,
+      "loss": 0.6209,
+      "step": 1524
+    },
+    {
+      "epoch": 0.2504464927226818,
+      "grad_norm": 0.3807854751881578,
+      "learning_rate": 8.963378685370209e-06,
+      "loss": 0.614,
+      "step": 1525
+    },
+    {
+      "epoch": 0.2506107199310246,
+      "grad_norm": 0.3506963314976677,
+      "learning_rate": 8.963318952458571e-06,
+      "loss": 0.607,
+      "step": 1526
+    },
+    {
+      "epoch": 0.2507749471393673,
+      "grad_norm": 0.3478461020759691,
+      "learning_rate": 8.9632591710709e-06,
+      "loss": 0.641,
+      "step": 1527
+    },
+    {
+      "epoch": 0.25093917434771007,
+      "grad_norm": 0.3530580403738584,
+      "learning_rate": 8.963199341207842e-06,
+      "loss": 0.6202,
+      "step": 1528
+    },
+    {
+      "epoch": 0.2511034015560528,
+      "grad_norm": 0.39033985727780796,
+      "learning_rate": 8.963139462870049e-06,
+      "loss": 0.6342,
+      "step": 1529
+    },
+    {
+      "epoch": 0.25126762876439557,
+      "grad_norm": 0.34915595980919556,
+      "learning_rate": 8.963079536058168e-06,
+      "loss": 0.6036,
+      "step": 1530
+    },
+    {
+      "epoch": 0.25143185597273826,
+      "grad_norm": 0.41289312093272673,
+      "learning_rate": 8.963019560772856e-06,
+      "loss": 0.6,
+      "step": 1531
+    },
+    {
+      "epoch": 0.251596083181081,
+      "grad_norm": 0.3579650538537075,
+      "learning_rate": 8.962959537014757e-06,
+      "loss": 0.6271,
+      "step": 1532
+    },
+    {
+      "epoch": 0.25176031038942376,
+      "grad_norm": 0.358747837059559,
+      "learning_rate": 8.962899464784528e-06,
+      "loss": 0.6079,
+      "step": 1533
+    },
+    {
+      "epoch": 0.2519245375977665,
+      "grad_norm": 0.3630698663213305,
+      "learning_rate": 8.962839344082818e-06,
+      "loss": 0.6221,
+      "step": 1534
+    },
+    {
+      "epoch": 0.25208876480610926,
+      "grad_norm": 0.3587173079907215,
+      "learning_rate": 8.962779174910283e-06,
+      "loss": 0.6049,
+      "step": 1535
+    },
+    {
+      "epoch": 0.252252992014452,
+      "grad_norm": 0.3627910673327825,
+      "learning_rate": 8.962718957267576e-06,
+      "loss": 0.627,
+      "step": 1536
+    },
+    {
+      "epoch": 0.25241721922279475,
+      "grad_norm": 0.3543705245438461,
+      "learning_rate": 8.962658691155351e-06,
+      "loss": 0.6004,
+      "step": 1537
+    },
+    {
+      "epoch": 0.2525814464311375,
+      "grad_norm": 0.36309273648911605,
+      "learning_rate": 8.96259837657426e-06,
+      "loss": 0.6173,
+      "step": 1538
+    },
+    {
+      "epoch": 0.2527456736394802,
+      "grad_norm": 0.3777501728160246,
+      "learning_rate": 8.962538013524963e-06,
+      "loss": 0.5753,
+      "step": 1539
+    },
+    {
+      "epoch": 0.25290990084782294,
+      "grad_norm": 0.3592769791035358,
+      "learning_rate": 8.96247760200811e-06,
+      "loss": 0.6148,
+      "step": 1540
+    },
+    {
+      "epoch": 0.2530741280561657,
+      "grad_norm": 0.36539569711437503,
+      "learning_rate": 8.96241714202436e-06,
+      "loss": 0.636,
+      "step": 1541
+    },
+    {
+      "epoch": 0.25323835526450844,
+      "grad_norm": 0.3581273145789663,
+      "learning_rate": 8.962356633574368e-06,
+      "loss": 0.5825,
+      "step": 1542
+    },
+    {
+      "epoch": 0.2534025824728512,
+      "grad_norm": 0.4135080963610115,
+      "learning_rate": 8.962296076658795e-06,
+      "loss": 0.6252,
+      "step": 1543
+    },
+    {
+      "epoch": 0.25356680968119394,
+      "grad_norm": 0.33935214150807924,
+      "learning_rate": 8.962235471278298e-06,
+      "loss": 0.6465,
+      "step": 1544
+    },
+    {
+      "epoch": 0.2537310368895367,
+      "grad_norm": 0.4749931745427001,
+      "learning_rate": 8.962174817433531e-06,
+      "loss": 0.6142,
+      "step": 1545
+    },
+    {
+      "epoch": 0.25389526409787944,
+      "grad_norm": 0.42031769394767504,
+      "learning_rate": 8.962114115125154e-06,
+      "loss": 0.63,
+      "step": 1546
+    },
+    {
+      "epoch": 0.2540594913062222,
+      "grad_norm": 0.3804360525221743,
+      "learning_rate": 8.962053364353831e-06,
+      "loss": 0.6312,
+      "step": 1547
+    },
+    {
+      "epoch": 0.2542237185145649,
+      "grad_norm": 0.3524668319377424,
+      "learning_rate": 8.961992565120216e-06,
+      "loss": 0.6088,
+      "step": 1548
+    },
+    {
+      "epoch": 0.2543879457229076,
+      "grad_norm": 0.3458219025452087,
+      "learning_rate": 8.961931717424973e-06,
+      "loss": 0.6192,
+      "step": 1549
+    },
+    {
+      "epoch": 0.2545521729312504,
+      "grad_norm": 0.35274115975192566,
+      "learning_rate": 8.96187082126876e-06,
+      "loss": 0.6129,
+      "step": 1550
+    },
+    {
+      "epoch": 0.2547164001395931,
+      "grad_norm": 0.3626793254051729,
+      "learning_rate": 8.96180987665224e-06,
+      "loss": 0.6305,
+      "step": 1551
+    },
+    {
+      "epoch": 0.2548806273479359,
+      "grad_norm": 0.3848176233620632,
+      "learning_rate": 8.961748883576077e-06,
+      "loss": 0.589,
+      "step": 1552
+    },
+    {
+      "epoch": 0.2550448545562786,
+      "grad_norm": 0.366125994685682,
+      "learning_rate": 8.96168784204093e-06,
+      "loss": 0.6528,
+      "step": 1553
+    },
+    {
+      "epoch": 0.25520908176462137,
+      "grad_norm": 0.3322249818553472,
+      "learning_rate": 8.961626752047464e-06,
+      "loss": 0.641,
+      "step": 1554
+    },
+    {
+      "epoch": 0.2553733089729641,
+      "grad_norm": 0.39701211769550127,
+      "learning_rate": 8.96156561359634e-06,
+      "loss": 0.6347,
+      "step": 1555
+    },
+    {
+      "epoch": 0.2555375361813068,
+      "grad_norm": 0.331778398473829,
+      "learning_rate": 8.961504426688226e-06,
+      "loss": 0.5913,
+      "step": 1556
+    },
+    {
+      "epoch": 0.25570176338964956,
+      "grad_norm": 0.35888908022149046,
+      "learning_rate": 8.961443191323783e-06,
+      "loss": 0.6081,
+      "step": 1557
+    },
+    {
+      "epoch": 0.2558659905979923,
+      "grad_norm": 0.3535893722927766,
+      "learning_rate": 8.961381907503678e-06,
+      "loss": 0.6385,
+      "step": 1558
+    },
+    {
+      "epoch": 0.25603021780633506,
+      "grad_norm": 0.33617576269857896,
+      "learning_rate": 8.961320575228577e-06,
+      "loss": 0.6137,
+      "step": 1559
+    },
+    {
+      "epoch": 0.2561944450146778,
+      "grad_norm": 0.4050222822206187,
+      "learning_rate": 8.961259194499144e-06,
+      "loss": 0.6156,
+      "step": 1560
+    },
+    {
+      "epoch": 0.25635867222302056,
+      "grad_norm": 0.35052205808117587,
+      "learning_rate": 8.961197765316048e-06,
+      "loss": 0.599,
+      "step": 1561
+    },
+    {
+      "epoch": 0.2565228994313633,
+      "grad_norm": 0.39858862462836475,
+      "learning_rate": 8.961136287679955e-06,
+      "loss": 0.611,
+      "step": 1562
+    },
+    {
+      "epoch": 0.25668712663970605,
+      "grad_norm": 0.3749625887686892,
+      "learning_rate": 8.96107476159153e-06,
+      "loss": 0.6183,
+      "step": 1563
+    },
+    {
+      "epoch": 0.2568513538480488,
+      "grad_norm": 0.36936476849359495,
+      "learning_rate": 8.961013187051448e-06,
+      "loss": 0.6082,
+      "step": 1564
+    },
+    {
+      "epoch": 0.2570155810563915,
+      "grad_norm": 0.34295965567698833,
+      "learning_rate": 8.96095156406037e-06,
+      "loss": 0.6076,
+      "step": 1565
+    },
+    {
+      "epoch": 0.25717980826473424,
+      "grad_norm": 0.35256410051497583,
+      "learning_rate": 8.960889892618972e-06,
+      "loss": 0.6208,
+      "step": 1566
+    },
+    {
+      "epoch": 0.257344035473077,
+      "grad_norm": 0.5784333725990156,
+      "learning_rate": 8.960828172727918e-06,
+      "loss": 0.6236,
+      "step": 1567
+    },
+    {
+      "epoch": 0.25750826268141974,
+      "grad_norm": 0.3489605554526544,
+      "learning_rate": 8.960766404387882e-06,
+      "loss": 0.64,
+      "step": 1568
+    },
+    {
+      "epoch": 0.2576724898897625,
+      "grad_norm": 0.3415995007524165,
+      "learning_rate": 8.960704587599537e-06,
+      "loss": 0.638,
+      "step": 1569
+    },
+    {
+      "epoch": 0.25783671709810524,
+      "grad_norm": 0.3820929372774659,
+      "learning_rate": 8.960642722363548e-06,
+      "loss": 0.6488,
+      "step": 1570
+    },
+    {
+      "epoch": 0.258000944306448,
+      "grad_norm": 0.38911326483657277,
+      "learning_rate": 8.960580808680592e-06,
+      "loss": 0.5972,
+      "step": 1571
+    },
+    {
+      "epoch": 0.25816517151479074,
+      "grad_norm": 0.47819526662868234,
+      "learning_rate": 8.96051884655134e-06,
+      "loss": 0.6315,
+      "step": 1572
+    },
+    {
+      "epoch": 0.25832939872313343,
+      "grad_norm": 0.38675435762044147,
+      "learning_rate": 8.960456835976463e-06,
+      "loss": 0.6121,
+      "step": 1573
+    },
+    {
+      "epoch": 0.2584936259314762,
+      "grad_norm": 0.36691627542081623,
+      "learning_rate": 8.96039477695664e-06,
+      "loss": 0.6224,
+      "step": 1574
+    },
+    {
+      "epoch": 0.2586578531398189,
+      "grad_norm": 0.3273576946150552,
+      "learning_rate": 8.960332669492536e-06,
+      "loss": 0.6082,
+      "step": 1575
+    },
+    {
+      "epoch": 0.2588220803481617,
+      "grad_norm": 0.3771741764322628,
+      "learning_rate": 8.960270513584835e-06,
+      "loss": 0.6154,
+      "step": 1576
+    },
+    {
+      "epoch": 0.2589863075565044,
+      "grad_norm": 0.3504930691326926,
+      "learning_rate": 8.960208309234205e-06,
+      "loss": 0.6115,
+      "step": 1577
+    },
+    {
+      "epoch": 0.2591505347648472,
+      "grad_norm": 0.3221176517577065,
+      "learning_rate": 8.960146056441327e-06,
+      "loss": 0.6233,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2593147619731899,
+      "grad_norm": 0.34553595989460767,
+      "learning_rate": 8.960083755206874e-06,
+      "loss": 0.6314,
+      "step": 1579
+    },
+    {
+      "epoch": 0.25947898918153267,
+      "grad_norm": 0.35748357758110466,
+      "learning_rate": 8.960021405531523e-06,
+      "loss": 0.6069,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2596432163898754,
+      "grad_norm": 0.4473671425734154,
+      "learning_rate": 8.959959007415951e-06,
+      "loss": 0.6417,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2598074435982181,
+      "grad_norm": 0.3510174264321415,
+      "learning_rate": 8.959896560860838e-06,
+      "loss": 0.629,
+      "step": 1582
+    },
+    {
+      "epoch": 0.25997167080656086,
+      "grad_norm": 0.36410783759248944,
+      "learning_rate": 8.959834065866857e-06,
+      "loss": 0.5946,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2601358980149036,
+      "grad_norm": 0.3512116883938849,
+      "learning_rate": 8.959771522434693e-06,
+      "loss": 0.609,
+      "step": 1584
+    },
+    {
+      "epoch": 0.26030012522324636,
+      "grad_norm": 0.36875418431071966,
+      "learning_rate": 8.959708930565021e-06,
+      "loss": 0.6243,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2604643524315891,
+      "grad_norm": 0.40873019656468806,
+      "learning_rate": 8.959646290258523e-06,
+      "loss": 0.6009,
+      "step": 1586
+    },
+    {
+      "epoch": 0.26062857963993186,
+      "grad_norm": 0.33934926723741654,
+      "learning_rate": 8.959583601515878e-06,
+      "loss": 0.6143,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2607928068482746,
+      "grad_norm": 0.36594161286713034,
+      "learning_rate": 8.959520864337769e-06,
+      "loss": 0.5963,
+      "step": 1588
+    },
+    {
+      "epoch": 0.26095703405661735,
+      "grad_norm": 0.35446975771035094,
+      "learning_rate": 8.959458078724875e-06,
+      "loss": 0.6133,
+      "step": 1589
+    },
+    {
+      "epoch": 0.26112126126496005,
+      "grad_norm": 0.33119122466208173,
+      "learning_rate": 8.959395244677878e-06,
+      "loss": 0.6029,
+      "step": 1590
+    },
+    {
+      "epoch": 0.2612854884733028,
+      "grad_norm": 0.3558849986573231,
+      "learning_rate": 8.959332362197461e-06,
+      "loss": 0.5961,
+      "step": 1591
+    },
+    {
+      "epoch": 0.26144971568164554,
+      "grad_norm": 0.4300437747977983,
+      "learning_rate": 8.959269431284309e-06,
+      "loss": 0.6265,
+      "step": 1592
+    },
+    {
+      "epoch": 0.2616139428899883,
+      "grad_norm": 0.33313374885784347,
+      "learning_rate": 8.959206451939102e-06,
+      "loss": 0.6335,
+      "step": 1593
+    },
+    {
+      "epoch": 0.26177817009833104,
+      "grad_norm": 0.36400831389369626,
+      "learning_rate": 8.959143424162526e-06,
+      "loss": 0.6284,
+      "step": 1594
+    },
+    {
+      "epoch": 0.2619423973066738,
+      "grad_norm": 0.3464235716370396,
+      "learning_rate": 8.959080347955264e-06,
+      "loss": 0.6023,
+      "step": 1595
+    },
+    {
+      "epoch": 0.26210662451501654,
+      "grad_norm": 0.39376900324065406,
+      "learning_rate": 8.959017223318005e-06,
+      "loss": 0.6145,
+      "step": 1596
+    },
+    {
+      "epoch": 0.2622708517233593,
+      "grad_norm": 0.34821762378354826,
+      "learning_rate": 8.95895405025143e-06,
+      "loss": 0.6114,
+      "step": 1597
+    },
+    {
+      "epoch": 0.26243507893170204,
+      "grad_norm": 0.3463710949430958,
+      "learning_rate": 8.958890828756229e-06,
+      "loss": 0.6386,
+      "step": 1598
+    },
+    {
+      "epoch": 0.26259930614004473,
+      "grad_norm": 0.3247606140038734,
+      "learning_rate": 8.958827558833084e-06,
+      "loss": 0.6303,
+      "step": 1599
+    },
+    {
+      "epoch": 0.2627635333483875,
+      "grad_norm": 0.3406923782871825,
+      "learning_rate": 8.958764240482686e-06,
+      "loss": 0.5948,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26292776055673023,
+      "grad_norm": 0.3627730588083312,
+      "learning_rate": 8.958700873705721e-06,
+      "loss": 0.6232,
+      "step": 1601
+    },
+    {
+      "epoch": 0.263091987765073,
+      "grad_norm": 0.35207234573189816,
+      "learning_rate": 8.958637458502879e-06,
+      "loss": 0.6102,
+      "step": 1602
+    },
+    {
+      "epoch": 0.2632562149734157,
+      "grad_norm": 0.3569459596649669,
+      "learning_rate": 8.958573994874846e-06,
+      "loss": 0.6313,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2634204421817585,
+      "grad_norm": 0.4495054405643325,
+      "learning_rate": 8.958510482822314e-06,
+      "loss": 0.6,
+      "step": 1604
+    },
+    {
+      "epoch": 0.2635846693901012,
+      "grad_norm": 0.38731324914519233,
+      "learning_rate": 8.95844692234597e-06,
+      "loss": 0.6405,
+      "step": 1605
+    },
+    {
+      "epoch": 0.26374889659844397,
+      "grad_norm": 0.3338592407113862,
+      "learning_rate": 8.958383313446508e-06,
+      "loss": 0.6234,
+      "step": 1606
+    },
+    {
+      "epoch": 0.26391312380678666,
+      "grad_norm": 0.33659403725843934,
+      "learning_rate": 8.958319656124615e-06,
+      "loss": 0.6163,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2640773510151294,
+      "grad_norm": 0.34801294714515657,
+      "learning_rate": 8.958255950380986e-06,
+      "loss": 0.6201,
+      "step": 1608
+    },
+    {
+      "epoch": 0.26424157822347216,
+      "grad_norm": 0.33754657169064073,
+      "learning_rate": 8.958192196216309e-06,
+      "loss": 0.6342,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2644058054318149,
+      "grad_norm": 0.3224965637155876,
+      "learning_rate": 8.958128393631279e-06,
+      "loss": 0.6122,
+      "step": 1610
+    },
+    {
+      "epoch": 0.26457003264015766,
+      "grad_norm": 0.33135459212029433,
+      "learning_rate": 8.958064542626589e-06,
+      "loss": 0.5957,
+      "step": 1611
+    },
+    {
+      "epoch": 0.2647342598485004,
+      "grad_norm": 0.34651185946628477,
+      "learning_rate": 8.958000643202932e-06,
+      "loss": 0.5911,
+      "step": 1612
+    },
+    {
+      "epoch": 0.26489848705684316,
+      "grad_norm": 0.3637450672410526,
+      "learning_rate": 8.957936695361001e-06,
+      "loss": 0.6292,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2650627142651859,
+      "grad_norm": 0.34550569205067416,
+      "learning_rate": 8.957872699101492e-06,
+      "loss": 0.6216,
+      "step": 1614
+    },
+    {
+      "epoch": 0.26522694147352865,
+      "grad_norm": 0.3360151957193412,
+      "learning_rate": 8.9578086544251e-06,
+      "loss": 0.6141,
+      "step": 1615
+    },
+    {
+      "epoch": 0.26539116868187135,
+      "grad_norm": 0.3322104258097254,
+      "learning_rate": 8.957744561332521e-06,
+      "loss": 0.6043,
+      "step": 1616
+    },
+    {
+      "epoch": 0.2655553958902141,
+      "grad_norm": 0.3507509351350694,
+      "learning_rate": 8.957680419824448e-06,
+      "loss": 0.6475,
+      "step": 1617
+    },
+    {
+      "epoch": 0.26571962309855685,
+      "grad_norm": 0.3477772878053762,
+      "learning_rate": 8.95761622990158e-06,
+      "loss": 0.601,
+      "step": 1618
+    },
+    {
+      "epoch": 0.2658838503068996,
+      "grad_norm": 0.3976960245557338,
+      "learning_rate": 8.957551991564617e-06,
+      "loss": 0.6343,
+      "step": 1619
+    },
+    {
+      "epoch": 0.26604807751524234,
+      "grad_norm": 0.34982742819917306,
+      "learning_rate": 8.957487704814252e-06,
+      "loss": 0.6665,
+      "step": 1620
+    },
+    {
+      "epoch": 0.2662123047235851,
+      "grad_norm": 0.3406041837951626,
+      "learning_rate": 8.957423369651183e-06,
+      "loss": 0.6283,
+      "step": 1621
+    },
+    {
+      "epoch": 0.26637653193192784,
+      "grad_norm": 0.3279996924316886,
+      "learning_rate": 8.957358986076113e-06,
+      "loss": 0.6057,
+      "step": 1622
+    },
+    {
+      "epoch": 0.2665407591402706,
+      "grad_norm": 0.3672845856678108,
+      "learning_rate": 8.957294554089738e-06,
+      "loss": 0.6201,
+      "step": 1623
+    },
+    {
+      "epoch": 0.2667049863486133,
+      "grad_norm": 0.39245027772202284,
+      "learning_rate": 8.957230073692759e-06,
+      "loss": 0.6082,
+      "step": 1624
+    },
+    {
+      "epoch": 0.26686921355695603,
+      "grad_norm": 0.40103050047766803,
+      "learning_rate": 8.957165544885875e-06,
+      "loss": 0.5781,
+      "step": 1625
+    },
+    {
+      "epoch": 0.2670334407652988,
+      "grad_norm": 0.33885876819639205,
+      "learning_rate": 8.957100967669791e-06,
+      "loss": 0.6334,
+      "step": 1626
+    },
+    {
+      "epoch": 0.26719766797364153,
+      "grad_norm": 0.3583568872434457,
+      "learning_rate": 8.957036342045203e-06,
+      "loss": 0.6053,
+      "step": 1627
+    },
+    {
+      "epoch": 0.2673618951819843,
+      "grad_norm": 0.3518879810969012,
+      "learning_rate": 8.956971668012817e-06,
+      "loss": 0.6216,
+      "step": 1628
+    },
+    {
+      "epoch": 0.267526122390327,
+      "grad_norm": 0.3381543177932053,
+      "learning_rate": 8.95690694557333e-06,
+      "loss": 0.6232,
+      "step": 1629
+    },
+    {
+      "epoch": 0.2676903495986698,
+      "grad_norm": 0.441450413777482,
+      "learning_rate": 8.95684217472745e-06,
+      "loss": 0.6225,
+      "step": 1630
+    },
+    {
+      "epoch": 0.2678545768070125,
+      "grad_norm": 0.35541320952449645,
+      "learning_rate": 8.956777355475881e-06,
+      "loss": 0.6389,
+      "step": 1631
+    },
+    {
+      "epoch": 0.2680188040153552,
+      "grad_norm": 0.4207779627283632,
+      "learning_rate": 8.956712487819323e-06,
+      "loss": 0.632,
+      "step": 1632
+    },
+    {
+      "epoch": 0.26818303122369797,
+      "grad_norm": 0.3320596414586848,
+      "learning_rate": 8.956647571758485e-06,
+      "loss": 0.6146,
+      "step": 1633
+    },
+    {
+      "epoch": 0.2683472584320407,
+      "grad_norm": 0.3565184586735187,
+      "learning_rate": 8.956582607294067e-06,
+      "loss": 0.6196,
+      "step": 1634
+    },
+    {
+      "epoch": 0.26851148564038346,
+      "grad_norm": 0.32988657768521945,
+      "learning_rate": 8.956517594426778e-06,
+      "loss": 0.6184,
+      "step": 1635
+    },
+    {
+      "epoch": 0.2686757128487262,
+      "grad_norm": 0.33061850241523516,
+      "learning_rate": 8.956452533157325e-06,
+      "loss": 0.5958,
+      "step": 1636
+    },
+    {
+      "epoch": 0.26883994005706896,
+      "grad_norm": 0.3852551983617042,
+      "learning_rate": 8.95638742348641e-06,
+      "loss": 0.6365,
+      "step": 1637
+    },
+    {
+      "epoch": 0.2690041672654117,
+      "grad_norm": 0.35206933111559585,
+      "learning_rate": 8.956322265414746e-06,
+      "loss": 0.6004,
+      "step": 1638
+    },
+    {
+      "epoch": 0.26916839447375446,
+      "grad_norm": 0.3427069503065173,
+      "learning_rate": 8.956257058943036e-06,
+      "loss": 0.6185,
+      "step": 1639
+    },
+    {
+      "epoch": 0.2693326216820972,
+      "grad_norm": 0.3406012616435544,
+      "learning_rate": 8.95619180407199e-06,
+      "loss": 0.606,
+      "step": 1640
+    },
+    {
+      "epoch": 0.2694968488904399,
+      "grad_norm": 0.3597673860314228,
+      "learning_rate": 8.956126500802318e-06,
+      "loss": 0.6304,
+      "step": 1641
+    },
+    {
+      "epoch": 0.26966107609878265,
+      "grad_norm": 0.3146073190261586,
+      "learning_rate": 8.956061149134725e-06,
+      "loss": 0.6177,
+      "step": 1642
+    },
+    {
+      "epoch": 0.2698253033071254,
+      "grad_norm": 0.3424703951143541,
+      "learning_rate": 8.955995749069926e-06,
+      "loss": 0.6211,
+      "step": 1643
+    },
+    {
+      "epoch": 0.26998953051546815,
+      "grad_norm": 0.35081835808818557,
+      "learning_rate": 8.955930300608629e-06,
+      "loss": 0.598,
+      "step": 1644
+    },
+    {
+      "epoch": 0.2701537577238109,
+      "grad_norm": 0.4085725713510005,
+      "learning_rate": 8.955864803751546e-06,
+      "loss": 0.6,
+      "step": 1645
+    },
+    {
+      "epoch": 0.27031798493215364,
+      "grad_norm": 0.3245057016062214,
+      "learning_rate": 8.955799258499384e-06,
+      "loss": 0.6016,
+      "step": 1646
+    },
+    {
+      "epoch": 0.2704822121404964,
+      "grad_norm": 0.3429782149104829,
+      "learning_rate": 8.95573366485286e-06,
+      "loss": 0.6029,
+      "step": 1647
+    },
+    {
+      "epoch": 0.27064643934883914,
+      "grad_norm": 0.3637562624540342,
+      "learning_rate": 8.955668022812687e-06,
+      "loss": 0.6142,
+      "step": 1648
+    },
+    {
+      "epoch": 0.27081066655718183,
+      "grad_norm": 0.4237638762114716,
+      "learning_rate": 8.955602332379572e-06,
+      "loss": 0.5921,
+      "step": 1649
+    },
+    {
+      "epoch": 0.2709748937655246,
+      "grad_norm": 0.39003068117908357,
+      "learning_rate": 8.955536593554232e-06,
+      "loss": 0.5899,
+      "step": 1650
+    },
+    {
+      "epoch": 0.27113912097386733,
+      "grad_norm": 0.37048027690848173,
+      "learning_rate": 8.955470806337382e-06,
+      "loss": 0.6276,
+      "step": 1651
+    },
+    {
+      "epoch": 0.2713033481822101,
+      "grad_norm": 0.3597344209193968,
+      "learning_rate": 8.955404970729736e-06,
+      "loss": 0.6175,
+      "step": 1652
+    },
+    {
+      "epoch": 0.27146757539055283,
+      "grad_norm": 0.3367753913122933,
+      "learning_rate": 8.955339086732009e-06,
+      "loss": 0.6233,
+      "step": 1653
+    },
+    {
+      "epoch": 0.2716318025988956,
+      "grad_norm": 0.3767695965340454,
+      "learning_rate": 8.955273154344914e-06,
+      "loss": 0.6275,
+      "step": 1654
+    },
+    {
+      "epoch": 0.2717960298072383,
+      "grad_norm": 0.35115066088714475,
+      "learning_rate": 8.95520717356917e-06,
+      "loss": 0.6262,
+      "step": 1655
+    },
+    {
+      "epoch": 0.2719602570155811,
+      "grad_norm": 0.35017560014066323,
+      "learning_rate": 8.955141144405493e-06,
+      "loss": 0.5956,
+      "step": 1656
+    },
+    {
+      "epoch": 0.2721244842239238,
+      "grad_norm": 0.33377410466349333,
+      "learning_rate": 8.9550750668546e-06,
+      "loss": 0.59,
+      "step": 1657
+    },
+    {
+      "epoch": 0.2722887114322665,
+      "grad_norm": 0.3590803976413437,
+      "learning_rate": 8.955008940917208e-06,
+      "loss": 0.6097,
+      "step": 1658
+    },
+    {
+      "epoch": 0.27245293864060927,
+      "grad_norm": 0.36554099156071396,
+      "learning_rate": 8.954942766594036e-06,
+      "loss": 0.6201,
+      "step": 1659
+    },
+    {
+      "epoch": 0.272617165848952,
+      "grad_norm": 0.3378974550030977,
+      "learning_rate": 8.954876543885802e-06,
+      "loss": 0.5875,
+      "step": 1660
+    },
+    {
+      "epoch": 0.27278139305729476,
+      "grad_norm": 0.32781104553889123,
+      "learning_rate": 8.954810272793227e-06,
+      "loss": 0.6002,
+      "step": 1661
+    },
+    {
+      "epoch": 0.2729456202656375,
+      "grad_norm": 0.398398331687394,
+      "learning_rate": 8.954743953317029e-06,
+      "loss": 0.62,
+      "step": 1662
+    },
+    {
+      "epoch": 0.27310984747398026,
+      "grad_norm": 0.5063519634001101,
+      "learning_rate": 8.95467758545793e-06,
+      "loss": 0.604,
+      "step": 1663
+    },
+    {
+      "epoch": 0.273274074682323,
+      "grad_norm": 0.3998293658682676,
+      "learning_rate": 8.954611169216646e-06,
+      "loss": 0.6273,
+      "step": 1664
+    },
+    {
+      "epoch": 0.27343830189066576,
+      "grad_norm": 0.6269815667257993,
+      "learning_rate": 8.954544704593904e-06,
+      "loss": 0.6311,
+      "step": 1665
+    },
+    {
+      "epoch": 0.27360252909900845,
+      "grad_norm": 0.32155955625909083,
+      "learning_rate": 8.954478191590425e-06,
+      "loss": 0.6042,
+      "step": 1666
+    },
+    {
+      "epoch": 0.2737667563073512,
+      "grad_norm": 0.35671629044279696,
+      "learning_rate": 8.95441163020693e-06,
+      "loss": 0.6325,
+      "step": 1667
+    },
+    {
+      "epoch": 0.27393098351569395,
+      "grad_norm": 0.3193642963859553,
+      "learning_rate": 8.954345020444141e-06,
+      "loss": 0.6272,
+      "step": 1668
+    },
+    {
+      "epoch": 0.2740952107240367,
+      "grad_norm": 0.3555838716484768,
+      "learning_rate": 8.954278362302783e-06,
+      "loss": 0.5956,
+      "step": 1669
+    },
+    {
+      "epoch": 0.27425943793237945,
+      "grad_norm": 0.49064088111583565,
+      "learning_rate": 8.954211655783579e-06,
+      "loss": 0.6205,
+      "step": 1670
+    },
+    {
+      "epoch": 0.2744236651407222,
+      "grad_norm": 0.3711693322859996,
+      "learning_rate": 8.954144900887255e-06,
+      "loss": 0.6149,
+      "step": 1671
+    },
+    {
+      "epoch": 0.27458789234906494,
+      "grad_norm": 0.33828977223496476,
+      "learning_rate": 8.954078097614534e-06,
+      "loss": 0.595,
+      "step": 1672
+    },
+    {
+      "epoch": 0.2747521195574077,
+      "grad_norm": 0.3603675345755279,
+      "learning_rate": 8.954011245966145e-06,
+      "loss": 0.6123,
+      "step": 1673
+    },
+    {
+      "epoch": 0.27491634676575044,
+      "grad_norm": 0.3342987882526401,
+      "learning_rate": 8.953944345942809e-06,
+      "loss": 0.6064,
+      "step": 1674
+    },
+    {
+      "epoch": 0.27508057397409313,
+      "grad_norm": 0.33840786908343246,
+      "learning_rate": 8.953877397545255e-06,
+      "loss": 0.6147,
+      "step": 1675
+    },
+    {
+      "epoch": 0.2752448011824359,
+      "grad_norm": 0.3361362202891087,
+      "learning_rate": 8.953810400774213e-06,
+      "loss": 0.612,
+      "step": 1676
+    },
+    {
+      "epoch": 0.27540902839077863,
+      "grad_norm": 0.3615623201654782,
+      "learning_rate": 8.953743355630406e-06,
+      "loss": 0.6197,
+      "step": 1677
+    },
+    {
+      "epoch": 0.2755732555991214,
+      "grad_norm": 0.34209041694780395,
+      "learning_rate": 8.953676262114565e-06,
+      "loss": 0.6493,
+      "step": 1678
+    },
+    {
+      "epoch": 0.27573748280746413,
+      "grad_norm": 0.325938108900595,
+      "learning_rate": 8.95360912022742e-06,
+      "loss": 0.6057,
+      "step": 1679
+    },
+    {
+      "epoch": 0.2759017100158069,
+      "grad_norm": 0.38745143087531064,
+      "learning_rate": 8.953541929969696e-06,
+      "loss": 0.6295,
+      "step": 1680
+    },
+    {
+      "epoch": 0.2760659372241496,
+      "grad_norm": 0.3810022825723158,
+      "learning_rate": 8.953474691342126e-06,
+      "loss": 0.5934,
+      "step": 1681
+    },
+    {
+      "epoch": 0.2762301644324924,
+      "grad_norm": 0.39295072393329594,
+      "learning_rate": 8.953407404345437e-06,
+      "loss": 0.6321,
+      "step": 1682
+    },
+    {
+      "epoch": 0.27639439164083507,
+      "grad_norm": 0.35390570166315877,
+      "learning_rate": 8.953340068980363e-06,
+      "loss": 0.5874,
+      "step": 1683
+    },
+    {
+      "epoch": 0.2765586188491778,
+      "grad_norm": 0.34976927367936533,
+      "learning_rate": 8.953272685247636e-06,
+      "loss": 0.6143,
+      "step": 1684
+    },
+    {
+      "epoch": 0.27672284605752057,
+      "grad_norm": 0.3327787798882458,
+      "learning_rate": 8.953205253147985e-06,
+      "loss": 0.6057,
+      "step": 1685
+    },
+    {
+      "epoch": 0.2768870732658633,
+      "grad_norm": 0.3415951518427282,
+      "learning_rate": 8.953137772682144e-06,
+      "loss": 0.6334,
+      "step": 1686
+    },
+    {
+      "epoch": 0.27705130047420606,
+      "grad_norm": 0.3384035832213004,
+      "learning_rate": 8.953070243850843e-06,
+      "loss": 0.6103,
+      "step": 1687
+    },
+    {
+      "epoch": 0.2772155276825488,
+      "grad_norm": 0.3495169277940791,
+      "learning_rate": 8.953002666654822e-06,
+      "loss": 0.6097,
+      "step": 1688
+    },
+    {
+      "epoch": 0.27737975489089156,
+      "grad_norm": 0.3416538910156951,
+      "learning_rate": 8.952935041094809e-06,
+      "loss": 0.5985,
+      "step": 1689
+    },
+    {
+      "epoch": 0.2775439820992343,
+      "grad_norm": 0.3451977176545778,
+      "learning_rate": 8.95286736717154e-06,
+      "loss": 0.5766,
+      "step": 1690
+    },
+    {
+      "epoch": 0.27770820930757706,
+      "grad_norm": 0.32525419888437135,
+      "learning_rate": 8.95279964488575e-06,
+      "loss": 0.6011,
+      "step": 1691
+    },
+    {
+      "epoch": 0.27787243651591975,
+      "grad_norm": 0.39000414414804097,
+      "learning_rate": 8.952731874238176e-06,
+      "loss": 0.6029,
+      "step": 1692
+    },
+    {
+      "epoch": 0.2780366637242625,
+      "grad_norm": 0.3393591338584082,
+      "learning_rate": 8.952664055229553e-06,
+      "loss": 0.5948,
+      "step": 1693
+    },
+    {
+      "epoch": 0.27820089093260525,
+      "grad_norm": 0.3986866139500602,
+      "learning_rate": 8.952596187860617e-06,
+      "loss": 0.6006,
+      "step": 1694
+    },
+    {
+      "epoch": 0.278365118140948,
+      "grad_norm": 0.3098858325727656,
+      "learning_rate": 8.952528272132107e-06,
+      "loss": 0.6008,
+      "step": 1695
+    },
+    {
+      "epoch": 0.27852934534929075,
+      "grad_norm": 0.3572611724356876,
+      "learning_rate": 8.952460308044756e-06,
+      "loss": 0.5998,
+      "step": 1696
+    },
+    {
+      "epoch": 0.2786935725576335,
+      "grad_norm": 0.4291924390919209,
+      "learning_rate": 8.952392295599309e-06,
+      "loss": 0.6133,
+      "step": 1697
+    },
+    {
+      "epoch": 0.27885779976597624,
+      "grad_norm": 0.34816840940610366,
+      "learning_rate": 8.9523242347965e-06,
+      "loss": 0.5985,
+      "step": 1698
+    },
+    {
+      "epoch": 0.279022026974319,
+      "grad_norm": 0.34022234855467215,
+      "learning_rate": 8.952256125637069e-06,
+      "loss": 0.6103,
+      "step": 1699
+    },
+    {
+      "epoch": 0.2791862541826617,
+      "grad_norm": 0.3639374654161306,
+      "learning_rate": 8.952187968121755e-06,
+      "loss": 0.6076,
+      "step": 1700
+    },
+    {
+      "epoch": 0.27935048139100443,
+      "grad_norm": 0.3338247300715134,
+      "learning_rate": 8.952119762251299e-06,
+      "loss": 0.6114,
+      "step": 1701
+    },
+    {
+      "epoch": 0.2795147085993472,
+      "grad_norm": 0.31947936911146907,
+      "learning_rate": 8.952051508026443e-06,
+      "loss": 0.616,
+      "step": 1702
+    },
+    {
+      "epoch": 0.27967893580768993,
+      "grad_norm": 0.34279451006593414,
+      "learning_rate": 8.951983205447928e-06,
+      "loss": 0.6179,
+      "step": 1703
+    },
+    {
+      "epoch": 0.2798431630160327,
+      "grad_norm": 0.45582975634168327,
+      "learning_rate": 8.951914854516495e-06,
+      "loss": 0.6173,
+      "step": 1704
+    },
+    {
+      "epoch": 0.28000739022437543,
+      "grad_norm": 0.3419181751540804,
+      "learning_rate": 8.951846455232888e-06,
+      "loss": 0.6024,
+      "step": 1705
+    },
+    {
+      "epoch": 0.2801716174327182,
+      "grad_norm": 0.3593734235508695,
+      "learning_rate": 8.951778007597848e-06,
+      "loss": 0.5989,
+      "step": 1706
+    },
+    {
+      "epoch": 0.2803358446410609,
+      "grad_norm": 0.3511409384421343,
+      "learning_rate": 8.951709511612116e-06,
+      "loss": 0.592,
+      "step": 1707
+    },
+    {
+      "epoch": 0.2805000718494037,
+      "grad_norm": 0.3501672881668432,
+      "learning_rate": 8.95164096727644e-06,
+      "loss": 0.5928,
+      "step": 1708
+    },
+    {
+      "epoch": 0.28066429905774637,
+      "grad_norm": 0.3727919134182136,
+      "learning_rate": 8.951572374591564e-06,
+      "loss": 0.5893,
+      "step": 1709
+    },
+    {
+      "epoch": 0.2808285262660891,
+      "grad_norm": 0.3280328400507155,
+      "learning_rate": 8.951503733558232e-06,
+      "loss": 0.6162,
+      "step": 1710
+    },
+    {
+      "epoch": 0.28099275347443187,
+      "grad_norm": 0.3775465194021973,
+      "learning_rate": 8.951435044177191e-06,
+      "loss": 0.6215,
+      "step": 1711
+    },
+    {
+      "epoch": 0.2811569806827746,
+      "grad_norm": 0.33441012730815717,
+      "learning_rate": 8.951366306449184e-06,
+      "loss": 0.5886,
+      "step": 1712
+    },
+    {
+      "epoch": 0.28132120789111736,
+      "grad_norm": 0.34134012357709204,
+      "learning_rate": 8.95129752037496e-06,
+      "loss": 0.5977,
+      "step": 1713
+    },
+    {
+      "epoch": 0.2814854350994601,
+      "grad_norm": 0.33093093198335205,
+      "learning_rate": 8.951228685955265e-06,
+      "loss": 0.6345,
+      "step": 1714
+    },
+    {
+      "epoch": 0.28164966230780286,
+      "grad_norm": 0.40993061330630076,
+      "learning_rate": 8.951159803190848e-06,
+      "loss": 0.6178,
+      "step": 1715
+    },
+    {
+      "epoch": 0.2818138895161456,
+      "grad_norm": 0.3963971197606208,
+      "learning_rate": 8.951090872082457e-06,
+      "loss": 0.6059,
+      "step": 1716
+    },
+    {
+      "epoch": 0.2819781167244883,
+      "grad_norm": 0.34012540670581837,
+      "learning_rate": 8.951021892630839e-06,
+      "loss": 0.6247,
+      "step": 1717
+    },
+    {
+      "epoch": 0.28214234393283105,
+      "grad_norm": 0.32786048540396723,
+      "learning_rate": 8.950952864836743e-06,
+      "loss": 0.6105,
+      "step": 1718
+    },
+    {
+      "epoch": 0.2823065711411738,
+      "grad_norm": 0.5082922315473971,
+      "learning_rate": 8.95088378870092e-06,
+      "loss": 0.6049,
+      "step": 1719
+    },
+    {
+      "epoch": 0.28247079834951655,
+      "grad_norm": 0.3094071280048266,
+      "learning_rate": 8.95081466422412e-06,
+      "loss": 0.5871,
+      "step": 1720
+    },
+    {
+      "epoch": 0.2826350255578593,
+      "grad_norm": 0.32426351466574627,
+      "learning_rate": 8.950745491407095e-06,
+      "loss": 0.5966,
+      "step": 1721
+    },
+    {
+      "epoch": 0.28279925276620205,
+      "grad_norm": 0.31977137099864866,
+      "learning_rate": 8.950676270250593e-06,
+      "loss": 0.6026,
+      "step": 1722
+    },
+    {
+      "epoch": 0.2829634799745448,
+      "grad_norm": 0.341677966380917,
+      "learning_rate": 8.95060700075537e-06,
+      "loss": 0.6146,
+      "step": 1723
+    },
+    {
+      "epoch": 0.28312770718288754,
+      "grad_norm": 0.34463367244022197,
+      "learning_rate": 8.950537682922175e-06,
+      "loss": 0.6037,
+      "step": 1724
+    },
+    {
+      "epoch": 0.2832919343912303,
+      "grad_norm": 0.3238216529925343,
+      "learning_rate": 8.950468316751763e-06,
+      "loss": 0.605,
+      "step": 1725
+    },
+    {
+      "epoch": 0.283456161599573,
+      "grad_norm": 0.3491065555826677,
+      "learning_rate": 8.950398902244884e-06,
+      "loss": 0.6054,
+      "step": 1726
+    },
+    {
+      "epoch": 0.28362038880791574,
+      "grad_norm": 0.3248877879087777,
+      "learning_rate": 8.950329439402296e-06,
+      "loss": 0.5972,
+      "step": 1727
+    },
+    {
+      "epoch": 0.2837846160162585,
+      "grad_norm": 0.3444958318564493,
+      "learning_rate": 8.950259928224753e-06,
+      "loss": 0.5953,
+      "step": 1728
+    },
+    {
+      "epoch": 0.28394884322460123,
+      "grad_norm": 0.32530493542474337,
+      "learning_rate": 8.950190368713007e-06,
+      "loss": 0.6184,
+      "step": 1729
+    },
+    {
+      "epoch": 0.284113070432944,
+      "grad_norm": 0.3321681672983501,
+      "learning_rate": 8.950120760867817e-06,
+      "loss": 0.598,
+      "step": 1730
+    },
+    {
+      "epoch": 0.28427729764128673,
+      "grad_norm": 0.33431613907622554,
+      "learning_rate": 8.950051104689933e-06,
+      "loss": 0.6233,
+      "step": 1731
+    },
+    {
+      "epoch": 0.2844415248496295,
+      "grad_norm": 0.33533763383067433,
+      "learning_rate": 8.94998140018012e-06,
+      "loss": 0.6041,
+      "step": 1732
+    },
+    {
+      "epoch": 0.2846057520579722,
+      "grad_norm": 0.3504748214822058,
+      "learning_rate": 8.949911647339128e-06,
+      "loss": 0.6167,
+      "step": 1733
+    },
+    {
+      "epoch": 0.2847699792663149,
+      "grad_norm": 0.4093185842862643,
+      "learning_rate": 8.94984184616772e-06,
+      "loss": 0.5901,
+      "step": 1734
+    },
+    {
+      "epoch": 0.28493420647465767,
+      "grad_norm": 0.3544777274138171,
+      "learning_rate": 8.94977199666665e-06,
+      "loss": 0.6031,
+      "step": 1735
+    },
+    {
+      "epoch": 0.2850984336830004,
+      "grad_norm": 0.3149036493882386,
+      "learning_rate": 8.949702098836679e-06,
+      "loss": 0.5978,
+      "step": 1736
+    },
+    {
+      "epoch": 0.28526266089134317,
+      "grad_norm": 0.32806889942804984,
+      "learning_rate": 8.949632152678564e-06,
+      "loss": 0.6175,
+      "step": 1737
+    },
+    {
+      "epoch": 0.2854268880996859,
+      "grad_norm": 0.33769469010145153,
+      "learning_rate": 8.949562158193067e-06,
+      "loss": 0.6051,
+      "step": 1738
+    },
+    {
+      "epoch": 0.28559111530802866,
+      "grad_norm": 0.321457597624339,
+      "learning_rate": 8.949492115380947e-06,
+      "loss": 0.6147,
+      "step": 1739
+    },
+    {
+      "epoch": 0.2857553425163714,
+      "grad_norm": 0.35122328128449154,
+      "learning_rate": 8.949422024242963e-06,
+      "loss": 0.5977,
+      "step": 1740
+    },
+    {
+      "epoch": 0.28591956972471416,
+      "grad_norm": 0.3853149359679044,
+      "learning_rate": 8.949351884779882e-06,
+      "loss": 0.6193,
+      "step": 1741
+    },
+    {
+      "epoch": 0.2860837969330569,
+      "grad_norm": 0.4647759735760696,
+      "learning_rate": 8.949281696992459e-06,
+      "loss": 0.6024,
+      "step": 1742
+    },
+    {
+      "epoch": 0.2862480241413996,
+      "grad_norm": 0.31943303091838704,
+      "learning_rate": 8.94921146088146e-06,
+      "loss": 0.5835,
+      "step": 1743
+    },
+    {
+      "epoch": 0.28641225134974235,
+      "grad_norm": 0.3386314610382678,
+      "learning_rate": 8.949141176447648e-06,
+      "loss": 0.6309,
+      "step": 1744
+    },
+    {
+      "epoch": 0.2865764785580851,
+      "grad_norm": 0.3223619040890537,
+      "learning_rate": 8.949070843691785e-06,
+      "loss": 0.6206,
+      "step": 1745
+    },
+    {
+      "epoch": 0.28674070576642785,
+      "grad_norm": 0.3460822464622039,
+      "learning_rate": 8.949000462614634e-06,
+      "loss": 0.5956,
+      "step": 1746
+    },
+    {
+      "epoch": 0.2869049329747706,
+      "grad_norm": 0.3360222306588388,
+      "learning_rate": 8.948930033216963e-06,
+      "loss": 0.5946,
+      "step": 1747
+    },
+    {
+      "epoch": 0.28706916018311335,
+      "grad_norm": 0.35259283103540384,
+      "learning_rate": 8.948859555499533e-06,
+      "loss": 0.596,
+      "step": 1748
+    },
+    {
+      "epoch": 0.2872333873914561,
+      "grad_norm": 0.34166466778665,
+      "learning_rate": 8.948789029463112e-06,
+      "loss": 0.5831,
+      "step": 1749
+    },
+    {
+      "epoch": 0.28739761459979885,
+      "grad_norm": 0.31738289979786777,
+      "learning_rate": 8.948718455108464e-06,
+      "loss": 0.6179,
+      "step": 1750
+    },
+    {
+      "epoch": 0.28756184180814154,
+      "grad_norm": 0.32852270318439086,
+      "learning_rate": 8.948647832436357e-06,
+      "loss": 0.6015,
+      "step": 1751
+    },
+    {
+      "epoch": 0.2877260690164843,
+      "grad_norm": 0.38106270430454836,
+      "learning_rate": 8.948577161447558e-06,
+      "loss": 0.6165,
+      "step": 1752
+    },
+    {
+      "epoch": 0.28789029622482704,
+      "grad_norm": 0.34254010965813875,
+      "learning_rate": 8.948506442142834e-06,
+      "loss": 0.6043,
+      "step": 1753
+    },
+    {
+      "epoch": 0.2880545234331698,
+      "grad_norm": 0.3124040810333103,
+      "learning_rate": 8.948435674522954e-06,
+      "loss": 0.5979,
+      "step": 1754
+    },
+    {
+      "epoch": 0.28821875064151253,
+      "grad_norm": 0.3204292251504406,
+      "learning_rate": 8.948364858588684e-06,
+      "loss": 0.5847,
+      "step": 1755
+    },
+    {
+      "epoch": 0.2883829778498553,
+      "grad_norm": 0.32046035935302963,
+      "learning_rate": 8.948293994340797e-06,
+      "loss": 0.6032,
+      "step": 1756
+    },
+    {
+      "epoch": 0.28854720505819803,
+      "grad_norm": 0.3408352541517514,
+      "learning_rate": 8.948223081780062e-06,
+      "loss": 0.6159,
+      "step": 1757
+    },
+    {
+      "epoch": 0.2887114322665408,
+      "grad_norm": 0.5290664262676616,
+      "learning_rate": 8.948152120907245e-06,
+      "loss": 0.6169,
+      "step": 1758
+    },
+    {
+      "epoch": 0.28887565947488353,
+      "grad_norm": 0.3400090749950359,
+      "learning_rate": 8.948081111723122e-06,
+      "loss": 0.6176,
+      "step": 1759
+    },
+    {
+      "epoch": 0.2890398866832262,
+      "grad_norm": 0.48512254987990944,
+      "learning_rate": 8.94801005422846e-06,
+      "loss": 0.6045,
+      "step": 1760
+    },
+    {
+      "epoch": 0.28920411389156897,
+      "grad_norm": 0.32696720216752834,
+      "learning_rate": 8.947938948424033e-06,
+      "loss": 0.611,
+      "step": 1761
+    },
+    {
+      "epoch": 0.2893683410999117,
+      "grad_norm": 0.3348572131814443,
+      "learning_rate": 8.947867794310612e-06,
+      "loss": 0.5976,
+      "step": 1762
+    },
+    {
+      "epoch": 0.28953256830825447,
+      "grad_norm": 0.3463666631002641,
+      "learning_rate": 8.947796591888971e-06,
+      "loss": 0.6093,
+      "step": 1763
+    },
+    {
+      "epoch": 0.2896967955165972,
+      "grad_norm": 0.37184836370878455,
+      "learning_rate": 8.947725341159884e-06,
+      "loss": 0.577,
+      "step": 1764
+    },
+    {
+      "epoch": 0.28986102272493997,
+      "grad_norm": 0.3554557818232329,
+      "learning_rate": 8.947654042124124e-06,
+      "loss": 0.628,
+      "step": 1765
+    },
+    {
+      "epoch": 0.2900252499332827,
+      "grad_norm": 0.34625018438648913,
+      "learning_rate": 8.947582694782464e-06,
+      "loss": 0.5874,
+      "step": 1766
+    },
+    {
+      "epoch": 0.29018947714162546,
+      "grad_norm": 0.3405480851791619,
+      "learning_rate": 8.947511299135681e-06,
+      "loss": 0.6213,
+      "step": 1767
+    },
+    {
+      "epoch": 0.29035370434996816,
+      "grad_norm": 0.33999734574851553,
+      "learning_rate": 8.947439855184548e-06,
+      "loss": 0.5855,
+      "step": 1768
+    },
+    {
+      "epoch": 0.2905179315583109,
+      "grad_norm": 0.33180325232357516,
+      "learning_rate": 8.947368362929844e-06,
+      "loss": 0.6249,
+      "step": 1769
+    },
+    {
+      "epoch": 0.29068215876665365,
+      "grad_norm": 0.33592557868625156,
+      "learning_rate": 8.947296822372344e-06,
+      "loss": 0.6044,
+      "step": 1770
+    },
+    {
+      "epoch": 0.2908463859749964,
+      "grad_norm": 0.3836129375299099,
+      "learning_rate": 8.947225233512824e-06,
+      "loss": 0.6128,
+      "step": 1771
+    },
+    {
+      "epoch": 0.29101061318333915,
+      "grad_norm": 0.35129704586375776,
+      "learning_rate": 8.947153596352064e-06,
+      "loss": 0.6037,
+      "step": 1772
+    },
+    {
+      "epoch": 0.2911748403916819,
+      "grad_norm": 0.3359681523693811,
+      "learning_rate": 8.947081910890838e-06,
+      "loss": 0.5972,
+      "step": 1773
+    },
+    {
+      "epoch": 0.29133906760002465,
+      "grad_norm": 0.4025206425833003,
+      "learning_rate": 8.947010177129929e-06,
+      "loss": 0.6035,
+      "step": 1774
+    },
+    {
+      "epoch": 0.2915032948083674,
+      "grad_norm": 0.3219472625707683,
+      "learning_rate": 8.946938395070115e-06,
+      "loss": 0.6012,
+      "step": 1775
+    },
+    {
+      "epoch": 0.29166752201671015,
+      "grad_norm": 0.34762377131835476,
+      "learning_rate": 8.946866564712174e-06,
+      "loss": 0.5764,
+      "step": 1776
+    },
+    {
+      "epoch": 0.29183174922505284,
+      "grad_norm": 0.33217069383444225,
+      "learning_rate": 8.946794686056886e-06,
+      "loss": 0.6019,
+      "step": 1777
+    },
+    {
+      "epoch": 0.2919959764333956,
+      "grad_norm": 0.31309510597308177,
+      "learning_rate": 8.946722759105034e-06,
+      "loss": 0.5969,
+      "step": 1778
+    },
+    {
+      "epoch": 0.29216020364173834,
+      "grad_norm": 0.3426072418753475,
+      "learning_rate": 8.946650783857395e-06,
+      "loss": 0.5952,
+      "step": 1779
+    },
+    {
+      "epoch": 0.2923244308500811,
+      "grad_norm": 0.3422905765735522,
+      "learning_rate": 8.946578760314758e-06,
+      "loss": 0.5999,
+      "step": 1780
+    },
+    {
+      "epoch": 0.29248865805842383,
+      "grad_norm": 0.32223052931910556,
+      "learning_rate": 8.946506688477896e-06,
+      "loss": 0.5891,
+      "step": 1781
+    },
+    {
+      "epoch": 0.2926528852667666,
+      "grad_norm": 0.34014514460328144,
+      "learning_rate": 8.9464345683476e-06,
+      "loss": 0.6022,
+      "step": 1782
+    },
+    {
+      "epoch": 0.29281711247510933,
+      "grad_norm": 0.4222426807348207,
+      "learning_rate": 8.94636239992465e-06,
+      "loss": 0.5949,
+      "step": 1783
+    },
+    {
+      "epoch": 0.2929813396834521,
+      "grad_norm": 1.4806624577700644,
+      "learning_rate": 8.946290183209829e-06,
+      "loss": 0.6048,
+      "step": 1784
+    },
+    {
+      "epoch": 0.2931455668917948,
+      "grad_norm": 0.43374222283990266,
+      "learning_rate": 8.946217918203922e-06,
+      "loss": 0.6213,
+      "step": 1785
+    },
+    {
+      "epoch": 0.2933097941001375,
+      "grad_norm": 0.33943359501404785,
+      "learning_rate": 8.946145604907712e-06,
+      "loss": 0.5849,
+      "step": 1786
+    },
+    {
+      "epoch": 0.29347402130848027,
+      "grad_norm": 0.34687657406918043,
+      "learning_rate": 8.94607324332199e-06,
+      "loss": 0.5989,
+      "step": 1787
+    },
+    {
+      "epoch": 0.293638248516823,
+      "grad_norm": 0.35147792904951136,
+      "learning_rate": 8.946000833447535e-06,
+      "loss": 0.6301,
+      "step": 1788
+    },
+    {
+      "epoch": 0.29380247572516577,
+      "grad_norm": 0.3120533026712586,
+      "learning_rate": 8.945928375285139e-06,
+      "loss": 0.6007,
+      "step": 1789
+    },
+    {
+      "epoch": 0.2939667029335085,
+      "grad_norm": 0.32762054284041,
+      "learning_rate": 8.945855868835584e-06,
+      "loss": 0.6263,
+      "step": 1790
+    },
+    {
+      "epoch": 0.29413093014185127,
+      "grad_norm": 0.319177490918817,
+      "learning_rate": 8.945783314099663e-06,
+      "loss": 0.5937,
+      "step": 1791
+    },
+    {
+      "epoch": 0.294295157350194,
+      "grad_norm": 0.3426538853948598,
+      "learning_rate": 8.94571071107816e-06,
+      "loss": 0.5915,
+      "step": 1792
+    },
+    {
+      "epoch": 0.29445938455853676,
+      "grad_norm": 0.3534302275740654,
+      "learning_rate": 8.945638059771864e-06,
+      "loss": 0.6086,
+      "step": 1793
+    },
+    {
+      "epoch": 0.29462361176687946,
+      "grad_norm": 0.4119987675520305,
+      "learning_rate": 8.945565360181566e-06,
+      "loss": 0.5817,
+      "step": 1794
+    },
+    {
+      "epoch": 0.2947878389752222,
+      "grad_norm": 0.33739850349835476,
+      "learning_rate": 8.945492612308053e-06,
+      "loss": 0.6046,
+      "step": 1795
+    },
+    {
+      "epoch": 0.29495206618356495,
+      "grad_norm": 0.31484597834060024,
+      "learning_rate": 8.945419816152118e-06,
+      "loss": 0.6109,
+      "step": 1796
+    },
+    {
+      "epoch": 0.2951162933919077,
+      "grad_norm": 0.3359396360573391,
+      "learning_rate": 8.945346971714548e-06,
+      "loss": 0.6,
+      "step": 1797
+    },
+    {
+      "epoch": 0.29528052060025045,
+      "grad_norm": 0.3273419658246956,
+      "learning_rate": 8.945274078996139e-06,
+      "loss": 0.5885,
+      "step": 1798
+    },
+    {
+      "epoch": 0.2954447478085932,
+      "grad_norm": 0.361277697562596,
+      "learning_rate": 8.945201137997677e-06,
+      "loss": 0.5726,
+      "step": 1799
+    },
+    {
+      "epoch": 0.29560897501693595,
+      "grad_norm": 0.3692868752877909,
+      "learning_rate": 8.94512814871996e-06,
+      "loss": 0.593,
+      "step": 1800
+    },
+    {
+      "epoch": 0.2957732022252787,
+      "grad_norm": 0.4303230153835513,
+      "learning_rate": 8.945055111163776e-06,
+      "loss": 0.5879,
+      "step": 1801
+    },
+    {
+      "epoch": 0.2959374294336214,
+      "grad_norm": 0.36640985236501644,
+      "learning_rate": 8.94498202532992e-06,
+      "loss": 0.6146,
+      "step": 1802
+    },
+    {
+      "epoch": 0.29610165664196414,
+      "grad_norm": 0.38608955906589143,
+      "learning_rate": 8.944908891219187e-06,
+      "loss": 0.5973,
+      "step": 1803
+    },
+    {
+      "epoch": 0.2962658838503069,
+      "grad_norm": 0.345799158653238,
+      "learning_rate": 8.94483570883237e-06,
+      "loss": 0.5899,
+      "step": 1804
+    },
+    {
+      "epoch": 0.29643011105864964,
+      "grad_norm": 0.38703871745196605,
+      "learning_rate": 8.944762478170264e-06,
+      "loss": 0.6125,
+      "step": 1805
+    },
+    {
+      "epoch": 0.2965943382669924,
+      "grad_norm": 0.4328536946290052,
+      "learning_rate": 8.944689199233665e-06,
+      "loss": 0.6108,
+      "step": 1806
+    },
+    {
+      "epoch": 0.29675856547533513,
+      "grad_norm": 0.41970722512068465,
+      "learning_rate": 8.944615872023367e-06,
+      "loss": 0.616,
+      "step": 1807
+    },
+    {
+      "epoch": 0.2969227926836779,
+      "grad_norm": 0.36376878975146204,
+      "learning_rate": 8.944542496540167e-06,
+      "loss": 0.5927,
+      "step": 1808
+    },
+    {
+      "epoch": 0.29708701989202063,
+      "grad_norm": 0.3833035639036213,
+      "learning_rate": 8.944469072784864e-06,
+      "loss": 0.5912,
+      "step": 1809
+    },
+    {
+      "epoch": 0.2972512471003633,
+      "grad_norm": 0.36238546383545633,
+      "learning_rate": 8.944395600758255e-06,
+      "loss": 0.6036,
+      "step": 1810
+    },
+    {
+      "epoch": 0.2974154743087061,
+      "grad_norm": 0.3480258093184134,
+      "learning_rate": 8.944322080461137e-06,
+      "loss": 0.616,
+      "step": 1811
+    },
+    {
+      "epoch": 0.2975797015170488,
+      "grad_norm": 0.3480417118481595,
+      "learning_rate": 8.944248511894307e-06,
+      "loss": 0.6033,
+      "step": 1812
+    },
+    {
+      "epoch": 0.29774392872539157,
+      "grad_norm": 0.33712664821835,
+      "learning_rate": 8.944174895058567e-06,
+      "loss": 0.5945,
+      "step": 1813
+    },
+    {
+      "epoch": 0.2979081559337343,
+      "grad_norm": 0.313027470384797,
+      "learning_rate": 8.944101229954714e-06,
+      "loss": 0.5916,
+      "step": 1814
+    },
+    {
+      "epoch": 0.29807238314207707,
+      "grad_norm": 0.3439276047463889,
+      "learning_rate": 8.94402751658355e-06,
+      "loss": 0.6182,
+      "step": 1815
+    },
+    {
+      "epoch": 0.2982366103504198,
+      "grad_norm": 0.36083957952146445,
+      "learning_rate": 8.943953754945874e-06,
+      "loss": 0.6012,
+      "step": 1816
+    },
+    {
+      "epoch": 0.29840083755876257,
+      "grad_norm": 0.3693169795087396,
+      "learning_rate": 8.943879945042488e-06,
+      "loss": 0.6028,
+      "step": 1817
+    },
+    {
+      "epoch": 0.2985650647671053,
+      "grad_norm": 0.35025691555948874,
+      "learning_rate": 8.943806086874195e-06,
+      "loss": 0.6178,
+      "step": 1818
+    },
+    {
+      "epoch": 0.298729291975448,
+      "grad_norm": 0.30773657041387825,
+      "learning_rate": 8.943732180441794e-06,
+      "loss": 0.6097,
+      "step": 1819
+    },
+    {
+      "epoch": 0.29889351918379076,
+      "grad_norm": 0.3707759522224546,
+      "learning_rate": 8.94365822574609e-06,
+      "loss": 0.5885,
+      "step": 1820
+    },
+    {
+      "epoch": 0.2990577463921335,
+      "grad_norm": 0.35508095599295475,
+      "learning_rate": 8.943584222787888e-06,
+      "loss": 0.6205,
+      "step": 1821
+    },
+    {
+      "epoch": 0.29922197360047625,
+      "grad_norm": 0.33789169443655837,
+      "learning_rate": 8.943510171567986e-06,
+      "loss": 0.6159,
+      "step": 1822
+    },
+    {
+      "epoch": 0.299386200808819,
+      "grad_norm": 0.37893216238689686,
+      "learning_rate": 8.943436072087195e-06,
+      "loss": 0.6163,
+      "step": 1823
+    },
+    {
+      "epoch": 0.29955042801716175,
+      "grad_norm": 0.3149645776504594,
+      "learning_rate": 8.943361924346313e-06,
+      "loss": 0.5986,
+      "step": 1824
+    },
+    {
+      "epoch": 0.2997146552255045,
+      "grad_norm": 0.2981620792000357,
+      "learning_rate": 8.943287728346151e-06,
+      "loss": 0.5902,
+      "step": 1825
+    },
+    {
+      "epoch": 0.29987888243384725,
+      "grad_norm": 0.3527275097586558,
+      "learning_rate": 8.943213484087512e-06,
+      "loss": 0.6147,
+      "step": 1826
+    },
+    {
+      "epoch": 0.30004310964218994,
+      "grad_norm": 0.3273132809209595,
+      "learning_rate": 8.943139191571203e-06,
+      "loss": 0.6007,
+      "step": 1827
+    },
+    {
+      "epoch": 0.3002073368505327,
+      "grad_norm": 0.37250810652301003,
+      "learning_rate": 8.943064850798031e-06,
+      "loss": 0.5959,
+      "step": 1828
+    },
+    {
+      "epoch": 0.30037156405887544,
+      "grad_norm": 0.33906589441259377,
+      "learning_rate": 8.942990461768805e-06,
+      "loss": 0.5965,
+      "step": 1829
+    },
+    {
+      "epoch": 0.3005357912672182,
+      "grad_norm": 0.313977528622849,
+      "learning_rate": 8.94291602448433e-06,
+      "loss": 0.5909,
+      "step": 1830
+    },
+    {
+      "epoch": 0.30070001847556094,
+      "grad_norm": 0.35447541373215136,
+      "learning_rate": 8.942841538945415e-06,
+      "loss": 0.5942,
+      "step": 1831
+    },
+    {
+      "epoch": 0.3008642456839037,
+      "grad_norm": 0.3184211563502309,
+      "learning_rate": 8.94276700515287e-06,
+      "loss": 0.6229,
+      "step": 1832
+    },
+    {
+      "epoch": 0.30102847289224643,
+      "grad_norm": 0.3279512969542947,
+      "learning_rate": 8.942692423107506e-06,
+      "loss": 0.5918,
+      "step": 1833
+    },
+    {
+      "epoch": 0.3011927001005892,
+      "grad_norm": 0.3862938159831846,
+      "learning_rate": 8.94261779281013e-06,
+      "loss": 0.631,
+      "step": 1834
+    },
+    {
+      "epoch": 0.30135692730893193,
+      "grad_norm": 0.9431248577218457,
+      "learning_rate": 8.942543114261552e-06,
+      "loss": 0.5946,
+      "step": 1835
+    },
+    {
+      "epoch": 0.3015211545172746,
+      "grad_norm": 0.37485001496427883,
+      "learning_rate": 8.942468387462588e-06,
+      "loss": 0.6025,
+      "step": 1836
+    },
+    {
+      "epoch": 0.3016853817256174,
+      "grad_norm": 0.32292549370149765,
+      "learning_rate": 8.942393612414045e-06,
+      "loss": 0.6107,
+      "step": 1837
+    },
+    {
+      "epoch": 0.3018496089339601,
+      "grad_norm": 0.3677915274414714,
+      "learning_rate": 8.942318789116736e-06,
+      "loss": 0.5722,
+      "step": 1838
+    },
+    {
+      "epoch": 0.30201383614230287,
+      "grad_norm": 0.31430531461192124,
+      "learning_rate": 8.942243917571474e-06,
+      "loss": 0.5984,
+      "step": 1839
+    },
+    {
+      "epoch": 0.3021780633506456,
+      "grad_norm": 0.320090326229002,
+      "learning_rate": 8.942168997779075e-06,
+      "loss": 0.5669,
+      "step": 1840
+    },
+    {
+      "epoch": 0.30234229055898837,
+      "grad_norm": 0.4661611655990814,
+      "learning_rate": 8.942094029740347e-06,
+      "loss": 0.5949,
+      "step": 1841
+    },
+    {
+      "epoch": 0.3025065177673311,
+      "grad_norm": 0.35595856183511837,
+      "learning_rate": 8.94201901345611e-06,
+      "loss": 0.6392,
+      "step": 1842
+    },
+    {
+      "epoch": 0.30267074497567387,
+      "grad_norm": 0.34468323175899396,
+      "learning_rate": 8.941943948927175e-06,
+      "loss": 0.6319,
+      "step": 1843
+    },
+    {
+      "epoch": 0.30283497218401656,
+      "grad_norm": 0.3283822883044266,
+      "learning_rate": 8.94186883615436e-06,
+      "loss": 0.633,
+      "step": 1844
+    },
+    {
+      "epoch": 0.3029991993923593,
+      "grad_norm": 0.3211390353728499,
+      "learning_rate": 8.941793675138477e-06,
+      "loss": 0.5856,
+      "step": 1845
+    },
+    {
+      "epoch": 0.30316342660070206,
+      "grad_norm": 0.38373627857810344,
+      "learning_rate": 8.941718465880344e-06,
+      "loss": 0.6072,
+      "step": 1846
+    },
+    {
+      "epoch": 0.3033276538090448,
+      "grad_norm": 0.4349614186299253,
+      "learning_rate": 8.941643208380781e-06,
+      "loss": 0.5782,
+      "step": 1847
+    },
+    {
+      "epoch": 0.30349188101738755,
+      "grad_norm": 0.33020652969405023,
+      "learning_rate": 8.941567902640602e-06,
+      "loss": 0.5849,
+      "step": 1848
+    },
+    {
+      "epoch": 0.3036561082257303,
+      "grad_norm": 0.3768588711697932,
+      "learning_rate": 8.941492548660625e-06,
+      "loss": 0.5899,
+      "step": 1849
+    },
+    {
+      "epoch": 0.30382033543407305,
+      "grad_norm": 0.3160245478293392,
+      "learning_rate": 8.94141714644167e-06,
+      "loss": 0.5968,
+      "step": 1850
+    },
+    {
+      "epoch": 0.3039845626424158,
+      "grad_norm": 0.34318955817414776,
+      "learning_rate": 8.941341695984554e-06,
+      "loss": 0.6093,
+      "step": 1851
+    },
+    {
+      "epoch": 0.30414878985075855,
+      "grad_norm": 0.44061105391992383,
+      "learning_rate": 8.941266197290098e-06,
+      "loss": 0.5914,
+      "step": 1852
+    },
+    {
+      "epoch": 0.30431301705910124,
+      "grad_norm": 0.32870210797503463,
+      "learning_rate": 8.941190650359121e-06,
+      "loss": 0.624,
+      "step": 1853
+    },
+    {
+      "epoch": 0.304477244267444,
+      "grad_norm": 0.30637297274135705,
+      "learning_rate": 8.941115055192444e-06,
+      "loss": 0.6058,
+      "step": 1854
+    },
+    {
+      "epoch": 0.30464147147578674,
+      "grad_norm": 0.32762818726888737,
+      "learning_rate": 8.941039411790888e-06,
+      "loss": 0.6015,
+      "step": 1855
+    },
+    {
+      "epoch": 0.3048056986841295,
+      "grad_norm": 0.3731775719566123,
+      "learning_rate": 8.940963720155276e-06,
+      "loss": 0.603,
+      "step": 1856
+    },
+    {
+      "epoch": 0.30496992589247224,
+      "grad_norm": 0.346190637525773,
+      "learning_rate": 8.940887980286428e-06,
+      "loss": 0.5789,
+      "step": 1857
+    },
+    {
+      "epoch": 0.305134153100815,
+      "grad_norm": 0.36066973926340856,
+      "learning_rate": 8.940812192185166e-06,
+      "loss": 0.6123,
+      "step": 1858
+    },
+    {
+      "epoch": 0.30529838030915774,
+      "grad_norm": 0.34442243351119006,
+      "learning_rate": 8.940736355852316e-06,
+      "loss": 0.5858,
+      "step": 1859
+    },
+    {
+      "epoch": 0.3054626075175005,
+      "grad_norm": 0.30796542041483227,
+      "learning_rate": 8.9406604712887e-06,
+      "loss": 0.6035,
+      "step": 1860
+    },
+    {
+      "epoch": 0.3056268347258432,
+      "grad_norm": 0.3052034273976577,
+      "learning_rate": 8.94058453849514e-06,
+      "loss": 0.5995,
+      "step": 1861
+    },
+    {
+      "epoch": 0.3057910619341859,
+      "grad_norm": 0.3136793732786086,
+      "learning_rate": 8.940508557472466e-06,
+      "loss": 0.6061,
+      "step": 1862
+    },
+    {
+      "epoch": 0.3059552891425287,
+      "grad_norm": 0.3088594677892597,
+      "learning_rate": 8.940432528221499e-06,
+      "loss": 0.5707,
+      "step": 1863
+    },
+    {
+      "epoch": 0.3061195163508714,
+      "grad_norm": 0.33833389134184366,
+      "learning_rate": 8.940356450743065e-06,
+      "loss": 0.5839,
+      "step": 1864
+    },
+    {
+      "epoch": 0.30628374355921417,
+      "grad_norm": 0.3889948728445,
+      "learning_rate": 8.940280325037992e-06,
+      "loss": 0.606,
+      "step": 1865
+    },
+    {
+      "epoch": 0.3064479707675569,
+      "grad_norm": 0.30634134506963684,
+      "learning_rate": 8.940204151107106e-06,
+      "loss": 0.5845,
+      "step": 1866
+    },
+    {
+      "epoch": 0.30661219797589967,
+      "grad_norm": 0.33450251258079644,
+      "learning_rate": 8.940127928951235e-06,
+      "loss": 0.615,
+      "step": 1867
+    },
+    {
+      "epoch": 0.3067764251842424,
+      "grad_norm": 0.33318859011481017,
+      "learning_rate": 8.940051658571205e-06,
+      "loss": 0.5912,
+      "step": 1868
+    },
+    {
+      "epoch": 0.30694065239258517,
+      "grad_norm": 0.33002590261509496,
+      "learning_rate": 8.939975339967846e-06,
+      "loss": 0.6006,
+      "step": 1869
+    },
+    {
+      "epoch": 0.30710487960092786,
+      "grad_norm": 0.381727301651218,
+      "learning_rate": 8.939898973141987e-06,
+      "loss": 0.5911,
+      "step": 1870
+    },
+    {
+      "epoch": 0.3072691068092706,
+      "grad_norm": 0.3523516962800485,
+      "learning_rate": 8.939822558094456e-06,
+      "loss": 0.6086,
+      "step": 1871
+    },
+    {
+      "epoch": 0.30743333401761336,
+      "grad_norm": 0.3417375521609287,
+      "learning_rate": 8.939746094826085e-06,
+      "loss": 0.5858,
+      "step": 1872
+    },
+    {
+      "epoch": 0.3075975612259561,
+      "grad_norm": 0.3338860766579545,
+      "learning_rate": 8.939669583337703e-06,
+      "loss": 0.61,
+      "step": 1873
+    },
+    {
+      "epoch": 0.30776178843429886,
+      "grad_norm": 0.32800363065689425,
+      "learning_rate": 8.93959302363014e-06,
+      "loss": 0.6232,
+      "step": 1874
+    },
+    {
+      "epoch": 0.3079260156426416,
+      "grad_norm": 0.3791074578049409,
+      "learning_rate": 8.939516415704231e-06,
+      "loss": 0.6074,
+      "step": 1875
+    },
+    {
+      "epoch": 0.30809024285098435,
+      "grad_norm": 0.32136934807801004,
+      "learning_rate": 8.939439759560805e-06,
+      "loss": 0.6395,
+      "step": 1876
+    },
+    {
+      "epoch": 0.3082544700593271,
+      "grad_norm": 0.3308724841591282,
+      "learning_rate": 8.939363055200693e-06,
+      "loss": 0.5923,
+      "step": 1877
+    },
+    {
+      "epoch": 0.3084186972676698,
+      "grad_norm": 0.3039170961510034,
+      "learning_rate": 8.939286302624734e-06,
+      "loss": 0.5927,
+      "step": 1878
+    },
+    {
+      "epoch": 0.30858292447601254,
+      "grad_norm": 0.3139854071425278,
+      "learning_rate": 8.939209501833755e-06,
+      "loss": 0.6029,
+      "step": 1879
+    },
+    {
+      "epoch": 0.3087471516843553,
+      "grad_norm": 0.32362328635050663,
+      "learning_rate": 8.939132652828596e-06,
+      "loss": 0.5967,
+      "step": 1880
+    },
+    {
+      "epoch": 0.30891137889269804,
+      "grad_norm": 0.30376582477672504,
+      "learning_rate": 8.939055755610087e-06,
+      "loss": 0.6067,
+      "step": 1881
+    },
+    {
+      "epoch": 0.3090756061010408,
+      "grad_norm": 0.3088804302395554,
+      "learning_rate": 8.938978810179066e-06,
+      "loss": 0.5968,
+      "step": 1882
+    },
+    {
+      "epoch": 0.30923983330938354,
+      "grad_norm": 0.3182885237046062,
+      "learning_rate": 8.938901816536367e-06,
+      "loss": 0.6052,
+      "step": 1883
+    },
+    {
+      "epoch": 0.3094040605177263,
+      "grad_norm": 0.334401122678781,
+      "learning_rate": 8.938824774682829e-06,
+      "loss": 0.5915,
+      "step": 1884
+    },
+    {
+      "epoch": 0.30956828772606904,
+      "grad_norm": 0.33028895160076666,
+      "learning_rate": 8.938747684619284e-06,
+      "loss": 0.609,
+      "step": 1885
+    },
+    {
+      "epoch": 0.3097325149344118,
+      "grad_norm": 0.3116708839296472,
+      "learning_rate": 8.938670546346574e-06,
+      "loss": 0.5787,
+      "step": 1886
+    },
+    {
+      "epoch": 0.3098967421427545,
+      "grad_norm": 0.29246076671142657,
+      "learning_rate": 8.938593359865533e-06,
+      "loss": 0.5822,
+      "step": 1887
+    },
+    {
+      "epoch": 0.3100609693510972,
+      "grad_norm": 0.3222058679707449,
+      "learning_rate": 8.938516125177002e-06,
+      "loss": 0.6057,
+      "step": 1888
+    },
+    {
+      "epoch": 0.31022519655944,
+      "grad_norm": 0.3983985286357778,
+      "learning_rate": 8.938438842281819e-06,
+      "loss": 0.61,
+      "step": 1889
+    },
+    {
+      "epoch": 0.3103894237677827,
+      "grad_norm": 0.360508351280119,
+      "learning_rate": 8.938361511180823e-06,
+      "loss": 0.5955,
+      "step": 1890
+    },
+    {
+      "epoch": 0.3105536509761255,
+      "grad_norm": 0.333327409913908,
+      "learning_rate": 8.938284131874856e-06,
+      "loss": 0.5741,
+      "step": 1891
+    },
+    {
+      "epoch": 0.3107178781844682,
+      "grad_norm": 0.32873570321422285,
+      "learning_rate": 8.938206704364754e-06,
+      "loss": 0.5719,
+      "step": 1892
+    },
+    {
+      "epoch": 0.31088210539281097,
+      "grad_norm": 0.33211514766374295,
+      "learning_rate": 8.938129228651361e-06,
+      "loss": 0.5712,
+      "step": 1893
+    },
+    {
+      "epoch": 0.3110463326011537,
+      "grad_norm": 0.37887895546631883,
+      "learning_rate": 8.938051704735521e-06,
+      "loss": 0.5768,
+      "step": 1894
+    },
+    {
+      "epoch": 0.3112105598094964,
+      "grad_norm": 0.3034162951190907,
+      "learning_rate": 8.93797413261807e-06,
+      "loss": 0.6036,
+      "step": 1895
+    },
+    {
+      "epoch": 0.31137478701783916,
+      "grad_norm": 0.2969354168845284,
+      "learning_rate": 8.937896512299854e-06,
+      "loss": 0.5692,
+      "step": 1896
+    },
+    {
+      "epoch": 0.3115390142261819,
+      "grad_norm": 0.3212729936642514,
+      "learning_rate": 8.937818843781717e-06,
+      "loss": 0.5753,
+      "step": 1897
+    },
+    {
+      "epoch": 0.31170324143452466,
+      "grad_norm": 0.31974316763883187,
+      "learning_rate": 8.9377411270645e-06,
+      "loss": 0.593,
+      "step": 1898
+    },
+    {
+      "epoch": 0.3118674686428674,
+      "grad_norm": 0.427602217890422,
+      "learning_rate": 8.937663362149048e-06,
+      "loss": 0.6011,
+      "step": 1899
+    },
+    {
+      "epoch": 0.31203169585121016,
+      "grad_norm": 0.32321832135134826,
+      "learning_rate": 8.937585549036207e-06,
+      "loss": 0.5819,
+      "step": 1900
+    },
+    {
+      "epoch": 0.3121959230595529,
+      "grad_norm": 0.3977562088177384,
+      "learning_rate": 8.93750768772682e-06,
+      "loss": 0.5982,
+      "step": 1901
+    },
+    {
+      "epoch": 0.31236015026789565,
+      "grad_norm": 0.34126358407384527,
+      "learning_rate": 8.937429778221733e-06,
+      "loss": 0.6179,
+      "step": 1902
+    },
+    {
+      "epoch": 0.3125243774762384,
+      "grad_norm": 0.34038697596903955,
+      "learning_rate": 8.937351820521793e-06,
+      "loss": 0.5826,
+      "step": 1903
+    },
+    {
+      "epoch": 0.3126886046845811,
+      "grad_norm": 0.382868319329483,
+      "learning_rate": 8.937273814627848e-06,
+      "loss": 0.5834,
+      "step": 1904
+    },
+    {
+      "epoch": 0.31285283189292384,
+      "grad_norm": 0.32282662203740864,
+      "learning_rate": 8.937195760540742e-06,
+      "loss": 0.5991,
+      "step": 1905
+    },
+    {
+      "epoch": 0.3130170591012666,
+      "grad_norm": 0.319661355229016,
+      "learning_rate": 8.937117658261326e-06,
+      "loss": 0.5726,
+      "step": 1906
+    },
+    {
+      "epoch": 0.31318128630960934,
+      "grad_norm": 0.34430211300551833,
+      "learning_rate": 8.937039507790446e-06,
+      "loss": 0.5983,
+      "step": 1907
+    },
+    {
+      "epoch": 0.3133455135179521,
+      "grad_norm": 0.38993287400775706,
+      "learning_rate": 8.936961309128951e-06,
+      "loss": 0.5928,
+      "step": 1908
+    },
+    {
+      "epoch": 0.31350974072629484,
+      "grad_norm": 0.4499590084430486,
+      "learning_rate": 8.93688306227769e-06,
+      "loss": 0.6112,
+      "step": 1909
+    },
+    {
+      "epoch": 0.3136739679346376,
+      "grad_norm": 0.35068628407856295,
+      "learning_rate": 8.936804767237515e-06,
+      "loss": 0.582,
+      "step": 1910
+    },
+    {
+      "epoch": 0.31383819514298034,
+      "grad_norm": 0.32678889633656605,
+      "learning_rate": 8.936726424009275e-06,
+      "loss": 0.5945,
+      "step": 1911
+    },
+    {
+      "epoch": 0.31400242235132303,
+      "grad_norm": 0.3427217699640625,
+      "learning_rate": 8.93664803259382e-06,
+      "loss": 0.6034,
+      "step": 1912
+    },
+    {
+      "epoch": 0.3141666495596658,
+      "grad_norm": 0.3750821944895772,
+      "learning_rate": 8.936569592992003e-06,
+      "loss": 0.5906,
+      "step": 1913
+    },
+    {
+      "epoch": 0.3143308767680085,
+      "grad_norm": 0.3733784173782264,
+      "learning_rate": 8.936491105204675e-06,
+      "loss": 0.6176,
+      "step": 1914
+    },
+    {
+      "epoch": 0.3144951039763513,
+      "grad_norm": 0.3463030952400915,
+      "learning_rate": 8.936412569232689e-06,
+      "loss": 0.5843,
+      "step": 1915
+    },
+    {
+      "epoch": 0.314659331184694,
+      "grad_norm": 0.3750041213901857,
+      "learning_rate": 8.936333985076898e-06,
+      "loss": 0.6075,
+      "step": 1916
+    },
+    {
+      "epoch": 0.3148235583930368,
+      "grad_norm": 0.3245755348831032,
+      "learning_rate": 8.936255352738155e-06,
+      "loss": 0.5803,
+      "step": 1917
+    },
+    {
+      "epoch": 0.3149877856013795,
+      "grad_norm": 0.3287569137043314,
+      "learning_rate": 8.936176672217314e-06,
+      "loss": 0.5633,
+      "step": 1918
+    },
+    {
+      "epoch": 0.31515201280972227,
+      "grad_norm": 0.34395092289289697,
+      "learning_rate": 8.936097943515229e-06,
+      "loss": 0.6011,
+      "step": 1919
+    },
+    {
+      "epoch": 0.315316240018065,
+      "grad_norm": 0.339337024599154,
+      "learning_rate": 8.936019166632757e-06,
+      "loss": 0.5962,
+      "step": 1920
+    },
+    {
+      "epoch": 0.3154804672264077,
+      "grad_norm": 0.38368448835939173,
+      "learning_rate": 8.935940341570752e-06,
+      "loss": 0.5859,
+      "step": 1921
+    },
+    {
+      "epoch": 0.31564469443475046,
+      "grad_norm": 0.36731857860519107,
+      "learning_rate": 8.93586146833007e-06,
+      "loss": 0.5962,
+      "step": 1922
+    },
+    {
+      "epoch": 0.3158089216430932,
+      "grad_norm": 0.32553923326385664,
+      "learning_rate": 8.935782546911568e-06,
+      "loss": 0.5815,
+      "step": 1923
+    },
+    {
+      "epoch": 0.31597314885143596,
+      "grad_norm": 0.32593913183393547,
+      "learning_rate": 8.935703577316105e-06,
+      "loss": 0.5998,
+      "step": 1924
+    },
+    {
+      "epoch": 0.3161373760597787,
+      "grad_norm": 0.3369591114430366,
+      "learning_rate": 8.935624559544534e-06,
+      "loss": 0.5987,
+      "step": 1925
+    },
+    {
+      "epoch": 0.31630160326812146,
+      "grad_norm": 0.34102164790399514,
+      "learning_rate": 8.935545493597719e-06,
+      "loss": 0.5898,
+      "step": 1926
+    },
+    {
+      "epoch": 0.3164658304764642,
+      "grad_norm": 0.44631235074440223,
+      "learning_rate": 8.935466379476515e-06,
+      "loss": 0.6025,
+      "step": 1927
+    },
+    {
+      "epoch": 0.31663005768480695,
+      "grad_norm": 0.3522044856839867,
+      "learning_rate": 8.93538721718178e-06,
+      "loss": 0.6083,
+      "step": 1928
+    },
+    {
+      "epoch": 0.31679428489314965,
+      "grad_norm": 0.3472773544596263,
+      "learning_rate": 8.93530800671438e-06,
+      "loss": 0.5754,
+      "step": 1929
+    },
+    {
+      "epoch": 0.3169585121014924,
+      "grad_norm": 0.4215209510654946,
+      "learning_rate": 8.93522874807517e-06,
+      "loss": 0.6199,
+      "step": 1930
+    },
+    {
+      "epoch": 0.31712273930983514,
+      "grad_norm": 0.33209722753968446,
+      "learning_rate": 8.935149441265012e-06,
+      "loss": 0.6037,
+      "step": 1931
+    },
+    {
+      "epoch": 0.3172869665181779,
+      "grad_norm": 0.5349339542080866,
+      "learning_rate": 8.935070086284766e-06,
+      "loss": 0.587,
+      "step": 1932
+    },
+    {
+      "epoch": 0.31745119372652064,
+      "grad_norm": 0.3410062528635434,
+      "learning_rate": 8.934990683135297e-06,
+      "loss": 0.5932,
+      "step": 1933
+    },
+    {
+      "epoch": 0.3176154209348634,
+      "grad_norm": 0.3408786107621934,
+      "learning_rate": 8.934911231817464e-06,
+      "loss": 0.609,
+      "step": 1934
+    },
+    {
+      "epoch": 0.31777964814320614,
+      "grad_norm": 0.3219443504322176,
+      "learning_rate": 8.934831732332133e-06,
+      "loss": 0.5974,
+      "step": 1935
+    },
+    {
+      "epoch": 0.3179438753515489,
+      "grad_norm": 0.40622542122388705,
+      "learning_rate": 8.934752184680166e-06,
+      "loss": 0.5862,
+      "step": 1936
+    },
+    {
+      "epoch": 0.31810810255989164,
+      "grad_norm": 0.32480511259589073,
+      "learning_rate": 8.934672588862426e-06,
+      "loss": 0.564,
+      "step": 1937
+    },
+    {
+      "epoch": 0.31827232976823433,
+      "grad_norm": 0.3076637039890817,
+      "learning_rate": 8.93459294487978e-06,
+      "loss": 0.5945,
+      "step": 1938
+    },
+    {
+      "epoch": 0.3184365569765771,
+      "grad_norm": 0.33152318749385223,
+      "learning_rate": 8.934513252733091e-06,
+      "loss": 0.5637,
+      "step": 1939
+    },
+    {
+      "epoch": 0.3186007841849198,
+      "grad_norm": 0.3318923807049237,
+      "learning_rate": 8.934433512423224e-06,
+      "loss": 0.6095,
+      "step": 1940
+    },
+    {
+      "epoch": 0.3187650113932626,
+      "grad_norm": 0.37782491111307404,
+      "learning_rate": 8.934353723951049e-06,
+      "loss": 0.6057,
+      "step": 1941
+    },
+    {
+      "epoch": 0.3189292386016053,
+      "grad_norm": 0.36785165188784136,
+      "learning_rate": 8.934273887317427e-06,
+      "loss": 0.6034,
+      "step": 1942
+    },
+    {
+      "epoch": 0.3190934658099481,
+      "grad_norm": 0.3129470298856957,
+      "learning_rate": 8.93419400252323e-06,
+      "loss": 0.6043,
+      "step": 1943
+    },
+    {
+      "epoch": 0.3192576930182908,
+      "grad_norm": 0.36195379839111386,
+      "learning_rate": 8.934114069569321e-06,
+      "loss": 0.5935,
+      "step": 1944
+    },
+    {
+      "epoch": 0.31942192022663357,
+      "grad_norm": 0.34969714763824405,
+      "learning_rate": 8.934034088456573e-06,
+      "loss": 0.582,
+      "step": 1945
+    },
+    {
+      "epoch": 0.31958614743497626,
+      "grad_norm": 0.3143958189825889,
+      "learning_rate": 8.93395405918585e-06,
+      "loss": 0.5904,
+      "step": 1946
+    },
+    {
+      "epoch": 0.319750374643319,
+      "grad_norm": 0.3676003778377262,
+      "learning_rate": 8.933873981758026e-06,
+      "loss": 0.5981,
+      "step": 1947
+    },
+    {
+      "epoch": 0.31991460185166176,
+      "grad_norm": 0.29483469959922937,
+      "learning_rate": 8.933793856173966e-06,
+      "loss": 0.6051,
+      "step": 1948
+    },
+    {
+      "epoch": 0.3200788290600045,
+      "grad_norm": 0.32753721487164505,
+      "learning_rate": 8.933713682434545e-06,
+      "loss": 0.5924,
+      "step": 1949
+    },
+    {
+      "epoch": 0.32024305626834726,
+      "grad_norm": 0.33750303418011973,
+      "learning_rate": 8.93363346054063e-06,
+      "loss": 0.5933,
+      "step": 1950
+    },
+    {
+      "epoch": 0.32040728347669,
+      "grad_norm": 0.31414417859725996,
+      "learning_rate": 8.933553190493092e-06,
+      "loss": 0.5742,
+      "step": 1951
+    },
+    {
+      "epoch": 0.32057151068503276,
+      "grad_norm": 0.3123828711273527,
+      "learning_rate": 8.933472872292805e-06,
+      "loss": 0.5843,
+      "step": 1952
+    },
+    {
+      "epoch": 0.3207357378933755,
+      "grad_norm": 0.4409976229317403,
+      "learning_rate": 8.933392505940643e-06,
+      "loss": 0.5793,
+      "step": 1953
+    },
+    {
+      "epoch": 0.32089996510171825,
+      "grad_norm": 0.30482982633884265,
+      "learning_rate": 8.933312091437473e-06,
+      "loss": 0.5888,
+      "step": 1954
+    },
+    {
+      "epoch": 0.32106419231006095,
+      "grad_norm": 0.4277019006668714,
+      "learning_rate": 8.933231628784174e-06,
+      "loss": 0.5929,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3212284195184037,
+      "grad_norm": 0.35975629174938006,
+      "learning_rate": 8.933151117981617e-06,
+      "loss": 0.5996,
+      "step": 1956
+    },
+    {
+      "epoch": 0.32139264672674644,
+      "grad_norm": 0.33774220319360165,
+      "learning_rate": 8.933070559030678e-06,
+      "loss": 0.6049,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3215568739350892,
+      "grad_norm": 0.34493943664636884,
+      "learning_rate": 8.93298995193223e-06,
+      "loss": 0.5892,
+      "step": 1958
+    },
+    {
+      "epoch": 0.32172110114343194,
+      "grad_norm": 0.3969135510220126,
+      "learning_rate": 8.93290929668715e-06,
+      "loss": 0.598,
+      "step": 1959
+    },
+    {
+      "epoch": 0.3218853283517747,
+      "grad_norm": 0.3308507690751659,
+      "learning_rate": 8.932828593296315e-06,
+      "loss": 0.5872,
+      "step": 1960
+    },
+    {
+      "epoch": 0.32204955556011744,
+      "grad_norm": 0.33602909569345507,
+      "learning_rate": 8.9327478417606e-06,
+      "loss": 0.5769,
+      "step": 1961
+    },
+    {
+      "epoch": 0.3222137827684602,
+      "grad_norm": 0.3928903161584532,
+      "learning_rate": 8.932667042080881e-06,
+      "loss": 0.5826,
+      "step": 1962
+    },
+    {
+      "epoch": 0.3223780099768029,
+      "grad_norm": 0.3503559259230569,
+      "learning_rate": 8.932586194258038e-06,
+      "loss": 0.5986,
+      "step": 1963
+    },
+    {
+      "epoch": 0.32254223718514563,
+      "grad_norm": 0.3413850568357385,
+      "learning_rate": 8.932505298292945e-06,
+      "loss": 0.572,
+      "step": 1964
+    },
+    {
+      "epoch": 0.3227064643934884,
+      "grad_norm": 0.33703606476691533,
+      "learning_rate": 8.932424354186486e-06,
+      "loss": 0.5834,
+      "step": 1965
+    },
+    {
+      "epoch": 0.32287069160183113,
+      "grad_norm": 0.3213463785765971,
+      "learning_rate": 8.932343361939538e-06,
+      "loss": 0.601,
+      "step": 1966
+    },
+    {
+      "epoch": 0.3230349188101739,
+      "grad_norm": 0.30437587809999017,
+      "learning_rate": 8.93226232155298e-06,
+      "loss": 0.591,
+      "step": 1967
+    },
+    {
+      "epoch": 0.3231991460185166,
+      "grad_norm": 0.44379718365544524,
+      "learning_rate": 8.932181233027692e-06,
+      "loss": 0.573,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3233633732268594,
+      "grad_norm": 0.44330142263154393,
+      "learning_rate": 8.932100096364554e-06,
+      "loss": 0.6035,
+      "step": 1969
+    },
+    {
+      "epoch": 0.3235276004352021,
+      "grad_norm": 0.3358028448333551,
+      "learning_rate": 8.93201891156445e-06,
+      "loss": 0.5983,
+      "step": 1970
+    },
+    {
+      "epoch": 0.32369182764354487,
+      "grad_norm": 0.366694044935094,
+      "learning_rate": 8.931937678628258e-06,
+      "loss": 0.6014,
+      "step": 1971
+    },
+    {
+      "epoch": 0.32385605485188756,
+      "grad_norm": 0.31338557422441116,
+      "learning_rate": 8.931856397556865e-06,
+      "loss": 0.5766,
+      "step": 1972
+    },
+    {
+      "epoch": 0.3240202820602303,
+      "grad_norm": 0.2903187762457536,
+      "learning_rate": 8.931775068351149e-06,
+      "loss": 0.5789,
+      "step": 1973
+    },
+    {
+      "epoch": 0.32418450926857306,
+      "grad_norm": 0.3254260429214987,
+      "learning_rate": 8.931693691011998e-06,
+      "loss": 0.5892,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3243487364769158,
+      "grad_norm": 0.3428275678284653,
+      "learning_rate": 8.931612265540291e-06,
+      "loss": 0.5984,
+      "step": 1975
+    },
+    {
+      "epoch": 0.32451296368525856,
+      "grad_norm": 0.30397612315020456,
+      "learning_rate": 8.931530791936915e-06,
+      "loss": 0.5831,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3246771908936013,
+      "grad_norm": 0.4572089137256494,
+      "learning_rate": 8.931449270202756e-06,
+      "loss": 0.5819,
+      "step": 1977
+    },
+    {
+      "epoch": 0.32484141810194406,
+      "grad_norm": 0.36966275950806693,
+      "learning_rate": 8.931367700338696e-06,
+      "loss": 0.5889,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3250056453102868,
+      "grad_norm": 0.3415927987872072,
+      "learning_rate": 8.931286082345625e-06,
+      "loss": 0.5923,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3251698725186295,
+      "grad_norm": 0.3627702364467653,
+      "learning_rate": 8.931204416224426e-06,
+      "loss": 0.6089,
+      "step": 1980
+    },
+    {
+      "epoch": 0.32533409972697225,
+      "grad_norm": 0.340328174457014,
+      "learning_rate": 8.931122701975987e-06,
+      "loss": 0.5923,
+      "step": 1981
+    },
+    {
+      "epoch": 0.325498326935315,
+      "grad_norm": 0.32978423355398195,
+      "learning_rate": 8.931040939601196e-06,
+      "loss": 0.5914,
+      "step": 1982
+    },
+    {
+      "epoch": 0.32566255414365775,
+      "grad_norm": 0.3543467620738116,
+      "learning_rate": 8.930959129100941e-06,
+      "loss": 0.5821,
+      "step": 1983
+    },
+    {
+      "epoch": 0.3258267813520005,
+      "grad_norm": 0.5390093529285701,
+      "learning_rate": 8.930877270476112e-06,
+      "loss": 0.5872,
+      "step": 1984
+    },
+    {
+      "epoch": 0.32599100856034324,
+      "grad_norm": 0.3290275508548655,
+      "learning_rate": 8.930795363727595e-06,
+      "loss": 0.588,
+      "step": 1985
+    },
+    {
+      "epoch": 0.326155235768686,
+      "grad_norm": 0.31186173084525426,
+      "learning_rate": 8.930713408856281e-06,
+      "loss": 0.6012,
+      "step": 1986
+    },
+    {
+      "epoch": 0.32631946297702874,
+      "grad_norm": 0.32941053882022925,
+      "learning_rate": 8.930631405863059e-06,
+      "loss": 0.5816,
+      "step": 1987
+    },
+    {
+      "epoch": 0.3264836901853715,
+      "grad_norm": 0.3222654111456884,
+      "learning_rate": 8.930549354748822e-06,
+      "loss": 0.5898,
+      "step": 1988
+    },
+    {
+      "epoch": 0.3266479173937142,
+      "grad_norm": 0.3292664467169063,
+      "learning_rate": 8.930467255514461e-06,
+      "loss": 0.5889,
+      "step": 1989
+    },
+    {
+      "epoch": 0.32681214460205693,
+      "grad_norm": 0.3743999278918246,
+      "learning_rate": 8.930385108160867e-06,
+      "loss": 0.6068,
+      "step": 1990
+    },
+    {
+      "epoch": 0.3269763718103997,
+      "grad_norm": 0.31297732477605195,
+      "learning_rate": 8.93030291268893e-06,
+      "loss": 0.603,
+      "step": 1991
+    },
+    {
+      "epoch": 0.32714059901874243,
+      "grad_norm": 0.33774822533754145,
+      "learning_rate": 8.930220669099544e-06,
+      "loss": 0.5755,
+      "step": 1992
+    },
+    {
+      "epoch": 0.3273048262270852,
+      "grad_norm": 0.3403085046460094,
+      "learning_rate": 8.930138377393604e-06,
+      "loss": 0.5925,
+      "step": 1993
+    },
+    {
+      "epoch": 0.3274690534354279,
+      "grad_norm": 0.48517931840712203,
+      "learning_rate": 8.930056037572002e-06,
+      "loss": 0.5873,
+      "step": 1994
+    },
+    {
+      "epoch": 0.3276332806437707,
+      "grad_norm": 0.30851335436470095,
+      "learning_rate": 8.929973649635633e-06,
+      "loss": 0.5701,
+      "step": 1995
+    },
+    {
+      "epoch": 0.3277975078521134,
+      "grad_norm": 0.3451118082788753,
+      "learning_rate": 8.929891213585391e-06,
+      "loss": 0.5815,
+      "step": 1996
+    },
+    {
+      "epoch": 0.3279617350604561,
+      "grad_norm": 0.3343156174433525,
+      "learning_rate": 8.929808729422172e-06,
+      "loss": 0.5835,
+      "step": 1997
+    },
+    {
+      "epoch": 0.32812596226879887,
+      "grad_norm": 0.31129871158266786,
+      "learning_rate": 8.929726197146872e-06,
+      "loss": 0.5938,
+      "step": 1998
+    },
+    {
+      "epoch": 0.3282901894771416,
+      "grad_norm": 0.30126632572026746,
+      "learning_rate": 8.92964361676039e-06,
+      "loss": 0.5936,
+      "step": 1999
+    },
+    {
+      "epoch": 0.32845441668548436,
+      "grad_norm": 0.5638600322215224,
+      "learning_rate": 8.929560988263617e-06,
+      "loss": 0.5845,
+      "step": 2000
+    },
+    {
+      "epoch": 0.3286186438938271,
+      "grad_norm": 0.3554185177476191,
+      "learning_rate": 8.929478311657455e-06,
+      "loss": 0.5588,
+      "step": 2001
+    },
+    {
+      "epoch": 0.32878287110216986,
+      "grad_norm": 0.33610749379303184,
+      "learning_rate": 8.9293955869428e-06,
+      "loss": 0.5824,
+      "step": 2002
+    },
+    {
+      "epoch": 0.3289470983105126,
+      "grad_norm": 0.8752215708398972,
+      "learning_rate": 8.929312814120551e-06,
+      "loss": 0.6125,
+      "step": 2003
+    },
+    {
+      "epoch": 0.32911132551885536,
+      "grad_norm": 0.3846513114226988,
+      "learning_rate": 8.929229993191608e-06,
+      "loss": 0.5778,
+      "step": 2004
+    },
+    {
+      "epoch": 0.32927555272719805,
+      "grad_norm": 0.3111081213681579,
+      "learning_rate": 8.929147124156869e-06,
+      "loss": 0.5772,
+      "step": 2005
+    },
+    {
+      "epoch": 0.3294397799355408,
+      "grad_norm": 0.3280468513105951,
+      "learning_rate": 8.929064207017233e-06,
+      "loss": 0.5855,
+      "step": 2006
+    },
+    {
+      "epoch": 0.32960400714388355,
+      "grad_norm": 0.32684896198790714,
+      "learning_rate": 8.928981241773603e-06,
+      "loss": 0.5722,
+      "step": 2007
+    },
+    {
+      "epoch": 0.3297682343522263,
+      "grad_norm": 0.32752500433087867,
+      "learning_rate": 8.92889822842688e-06,
+      "loss": 0.5977,
+      "step": 2008
+    },
+    {
+      "epoch": 0.32993246156056905,
+      "grad_norm": 0.3142469749217832,
+      "learning_rate": 8.928815166977964e-06,
+      "loss": 0.5972,
+      "step": 2009
+    },
+    {
+      "epoch": 0.3300966887689118,
+      "grad_norm": 0.34216065189885336,
+      "learning_rate": 8.928732057427757e-06,
+      "loss": 0.6017,
+      "step": 2010
+    },
+    {
+      "epoch": 0.33026091597725454,
+      "grad_norm": 0.33522533076340355,
+      "learning_rate": 8.928648899777165e-06,
+      "loss": 0.5883,
+      "step": 2011
+    },
+    {
+      "epoch": 0.3304251431855973,
+      "grad_norm": 0.4104881167330931,
+      "learning_rate": 8.928565694027086e-06,
+      "loss": 0.5868,
+      "step": 2012
+    },
+    {
+      "epoch": 0.33058937039394004,
+      "grad_norm": 0.323178569938277,
+      "learning_rate": 8.928482440178428e-06,
+      "loss": 0.6021,
+      "step": 2013
+    },
+    {
+      "epoch": 0.33075359760228273,
+      "grad_norm": 0.3731243392644556,
+      "learning_rate": 8.928399138232094e-06,
+      "loss": 0.577,
+      "step": 2014
+    },
+    {
+      "epoch": 0.3309178248106255,
+      "grad_norm": 0.30986603573501476,
+      "learning_rate": 8.928315788188989e-06,
+      "loss": 0.6136,
+      "step": 2015
+    },
+    {
+      "epoch": 0.33108205201896823,
+      "grad_norm": 0.33958219563081765,
+      "learning_rate": 8.928232390050015e-06,
+      "loss": 0.5737,
+      "step": 2016
+    },
+    {
+      "epoch": 0.331246279227311,
+      "grad_norm": 0.328627168041723,
+      "learning_rate": 8.928148943816084e-06,
+      "loss": 0.5868,
+      "step": 2017
+    },
+    {
+      "epoch": 0.33141050643565373,
+      "grad_norm": 0.3176779588770702,
+      "learning_rate": 8.928065449488096e-06,
+      "loss": 0.5769,
+      "step": 2018
+    },
+    {
+      "epoch": 0.3315747336439965,
+      "grad_norm": 0.41809285685467873,
+      "learning_rate": 8.927981907066961e-06,
+      "loss": 0.5899,
+      "step": 2019
+    },
+    {
+      "epoch": 0.3317389608523392,
+      "grad_norm": 0.36709251595651227,
+      "learning_rate": 8.927898316553586e-06,
+      "loss": 0.604,
+      "step": 2020
+    },
+    {
+      "epoch": 0.331903188060682,
+      "grad_norm": 0.3274955175203132,
+      "learning_rate": 8.927814677948879e-06,
+      "loss": 0.6012,
+      "step": 2021
+    },
+    {
+      "epoch": 0.33206741526902467,
+      "grad_norm": 0.2832864673661248,
+      "learning_rate": 8.92773099125375e-06,
+      "loss": 0.5942,
+      "step": 2022
+    },
+    {
+      "epoch": 0.3322316424773674,
+      "grad_norm": 0.30603924362675944,
+      "learning_rate": 8.927647256469104e-06,
+      "loss": 0.5925,
+      "step": 2023
+    },
+    {
+      "epoch": 0.33239586968571017,
+      "grad_norm": 0.33180800499144236,
+      "learning_rate": 8.927563473595853e-06,
+      "loss": 0.5823,
+      "step": 2024
+    },
+    {
+      "epoch": 0.3325600968940529,
+      "grad_norm": 0.31278550416452017,
+      "learning_rate": 8.927479642634906e-06,
+      "loss": 0.5746,
+      "step": 2025
+    },
+    {
+      "epoch": 0.33272432410239566,
+      "grad_norm": 0.6370773833935756,
+      "learning_rate": 8.927395763587175e-06,
+      "loss": 0.5944,
+      "step": 2026
+    },
+    {
+      "epoch": 0.3328885513107384,
+      "grad_norm": 0.3793789597731592,
+      "learning_rate": 8.927311836453569e-06,
+      "loss": 0.6057,
+      "step": 2027
+    },
+    {
+      "epoch": 0.33305277851908116,
+      "grad_norm": 0.32422037386343944,
+      "learning_rate": 8.927227861235002e-06,
+      "loss": 0.5886,
+      "step": 2028
+    },
+    {
+      "epoch": 0.3332170057274239,
+      "grad_norm": 0.41822762137690095,
+      "learning_rate": 8.927143837932384e-06,
+      "loss": 0.5775,
+      "step": 2029
+    },
+    {
+      "epoch": 0.33338123293576666,
+      "grad_norm": 0.3339592815475642,
+      "learning_rate": 8.927059766546627e-06,
+      "loss": 0.5943,
+      "step": 2030
+    },
+    {
+      "epoch": 0.33354546014410935,
+      "grad_norm": 0.4645135259033762,
+      "learning_rate": 8.926975647078648e-06,
+      "loss": 0.6036,
+      "step": 2031
+    },
+    {
+      "epoch": 0.3337096873524521,
+      "grad_norm": 0.30158131377366604,
+      "learning_rate": 8.926891479529356e-06,
+      "loss": 0.5819,
+      "step": 2032
+    },
+    {
+      "epoch": 0.33387391456079485,
+      "grad_norm": 0.46237162147584415,
+      "learning_rate": 8.926807263899665e-06,
+      "loss": 0.5877,
+      "step": 2033
+    },
+    {
+      "epoch": 0.3340381417691376,
+      "grad_norm": 0.3169681472015116,
+      "learning_rate": 8.926723000190496e-06,
+      "loss": 0.59,
+      "step": 2034
+    },
+    {
+      "epoch": 0.33420236897748035,
+      "grad_norm": 0.3515972769278677,
+      "learning_rate": 8.926638688402759e-06,
+      "loss": 0.5732,
+      "step": 2035
+    },
+    {
+      "epoch": 0.3343665961858231,
+      "grad_norm": 0.32805529347869533,
+      "learning_rate": 8.926554328537368e-06,
+      "loss": 0.5825,
+      "step": 2036
+    },
+    {
+      "epoch": 0.33453082339416584,
+      "grad_norm": 0.32908781830163236,
+      "learning_rate": 8.926469920595243e-06,
+      "loss": 0.5955,
+      "step": 2037
+    },
+    {
+      "epoch": 0.3346950506025086,
+      "grad_norm": 0.8047963527089921,
+      "learning_rate": 8.9263854645773e-06,
+      "loss": 0.5999,
+      "step": 2038
+    },
+    {
+      "epoch": 0.3348592778108513,
+      "grad_norm": 0.2904460170060317,
+      "learning_rate": 8.926300960484457e-06,
+      "loss": 0.585,
+      "step": 2039
+    },
+    {
+      "epoch": 0.33502350501919403,
+      "grad_norm": 0.32512733927564513,
+      "learning_rate": 8.92621640831763e-06,
+      "loss": 0.6103,
+      "step": 2040
+    },
+    {
+      "epoch": 0.3351877322275368,
+      "grad_norm": 0.620086893708396,
+      "learning_rate": 8.926131808077737e-06,
+      "loss": 0.613,
+      "step": 2041
+    },
+    {
+      "epoch": 0.33535195943587953,
+      "grad_norm": 0.3439753681608022,
+      "learning_rate": 8.926047159765699e-06,
+      "loss": 0.5954,
+      "step": 2042
+    },
+    {
+      "epoch": 0.3355161866442223,
+      "grad_norm": 0.30302759955620057,
+      "learning_rate": 8.925962463382433e-06,
+      "loss": 0.588,
+      "step": 2043
+    },
+    {
+      "epoch": 0.33568041385256503,
+      "grad_norm": 0.34048898804704547,
+      "learning_rate": 8.92587771892886e-06,
+      "loss": 0.6006,
+      "step": 2044
+    },
+    {
+      "epoch": 0.3358446410609078,
+      "grad_norm": 0.35733144461693095,
+      "learning_rate": 8.925792926405903e-06,
+      "loss": 0.5708,
+      "step": 2045
+    },
+    {
+      "epoch": 0.3360088682692505,
+      "grad_norm": 0.3100918297713711,
+      "learning_rate": 8.925708085814478e-06,
+      "loss": 0.579,
+      "step": 2046
+    },
+    {
+      "epoch": 0.3361730954775933,
+      "grad_norm": 0.31010557216776774,
+      "learning_rate": 8.925623197155508e-06,
+      "loss": 0.6043,
+      "step": 2047
+    },
+    {
+      "epoch": 0.33633732268593597,
+      "grad_norm": 0.3207717425849682,
+      "learning_rate": 8.925538260429919e-06,
+      "loss": 0.5791,
+      "step": 2048
+    },
+    {
+      "epoch": 0.3365015498942787,
+      "grad_norm": 0.46303927724673305,
+      "learning_rate": 8.925453275638628e-06,
+      "loss": 0.5981,
+      "step": 2049
+    },
+    {
+      "epoch": 0.33666577710262147,
+      "grad_norm": 0.32519157260027265,
+      "learning_rate": 8.925368242782562e-06,
+      "loss": 0.5773,
+      "step": 2050
+    },
+    {
+      "epoch": 0.3368300043109642,
+      "grad_norm": 0.30398409620669065,
+      "learning_rate": 8.925283161862642e-06,
+      "loss": 0.5894,
+      "step": 2051
+    },
+    {
+      "epoch": 0.33699423151930696,
+      "grad_norm": 0.2974031839313766,
+      "learning_rate": 8.925198032879793e-06,
+      "loss": 0.5957,
+      "step": 2052
+    },
+    {
+      "epoch": 0.3371584587276497,
+      "grad_norm": 0.3404861792754053,
+      "learning_rate": 8.925112855834939e-06,
+      "loss": 0.5979,
+      "step": 2053
+    },
+    {
+      "epoch": 0.33732268593599246,
+      "grad_norm": 0.3542777008958305,
+      "learning_rate": 8.925027630729007e-06,
+      "loss": 0.5731,
+      "step": 2054
+    },
+    {
+      "epoch": 0.3374869131443352,
+      "grad_norm": 0.29838686215896404,
+      "learning_rate": 8.92494235756292e-06,
+      "loss": 0.5914,
+      "step": 2055
+    },
+    {
+      "epoch": 0.3376511403526779,
+      "grad_norm": 0.30489826458776687,
+      "learning_rate": 8.924857036337606e-06,
+      "loss": 0.5885,
+      "step": 2056
+    },
+    {
+      "epoch": 0.33781536756102065,
+      "grad_norm": 0.34373941171524325,
+      "learning_rate": 8.924771667053991e-06,
+      "loss": 0.5691,
+      "step": 2057
+    },
+    {
+      "epoch": 0.3379795947693634,
+      "grad_norm": 0.3235745424873555,
+      "learning_rate": 8.924686249713002e-06,
+      "loss": 0.5791,
+      "step": 2058
+    },
+    {
+      "epoch": 0.33814382197770615,
+      "grad_norm": 0.45911687221923925,
+      "learning_rate": 8.924600784315568e-06,
+      "loss": 0.6007,
+      "step": 2059
+    },
+    {
+      "epoch": 0.3383080491860489,
+      "grad_norm": 0.3303513848889647,
+      "learning_rate": 8.924515270862615e-06,
+      "loss": 0.5877,
+      "step": 2060
+    },
+    {
+      "epoch": 0.33847227639439165,
+      "grad_norm": 0.3146634573861978,
+      "learning_rate": 8.924429709355075e-06,
+      "loss": 0.5985,
+      "step": 2061
+    },
+    {
+      "epoch": 0.3386365036027344,
+      "grad_norm": 0.35605464455064945,
+      "learning_rate": 8.924344099793873e-06,
+      "loss": 0.5658,
+      "step": 2062
+    },
+    {
+      "epoch": 0.33880073081107714,
+      "grad_norm": 0.3348179585377794,
+      "learning_rate": 8.924258442179942e-06,
+      "loss": 0.6065,
+      "step": 2063
+    },
+    {
+      "epoch": 0.3389649580194199,
+      "grad_norm": 0.3194465381879545,
+      "learning_rate": 8.924172736514213e-06,
+      "loss": 0.5856,
+      "step": 2064
+    },
+    {
+      "epoch": 0.3391291852277626,
+      "grad_norm": 0.36449000613032345,
+      "learning_rate": 8.924086982797612e-06,
+      "loss": 0.5774,
+      "step": 2065
+    },
+    {
+      "epoch": 0.33929341243610533,
+      "grad_norm": 0.3880607217082634,
+      "learning_rate": 8.924001181031077e-06,
+      "loss": 0.566,
+      "step": 2066
+    },
+    {
+      "epoch": 0.3394576396444481,
+      "grad_norm": 0.287273007995822,
+      "learning_rate": 8.923915331215534e-06,
+      "loss": 0.5615,
+      "step": 2067
+    },
+    {
+      "epoch": 0.33962186685279083,
+      "grad_norm": 0.3312610820471845,
+      "learning_rate": 8.923829433351919e-06,
+      "loss": 0.5675,
+      "step": 2068
+    },
+    {
+      "epoch": 0.3397860940611336,
+      "grad_norm": 0.35310045879408936,
+      "learning_rate": 8.923743487441164e-06,
+      "loss": 0.5948,
+      "step": 2069
+    },
+    {
+      "epoch": 0.33995032126947633,
+      "grad_norm": 0.33574837233858923,
+      "learning_rate": 8.923657493484203e-06,
+      "loss": 0.5896,
+      "step": 2070
+    },
+    {
+      "epoch": 0.3401145484778191,
+      "grad_norm": 0.29104720814747725,
+      "learning_rate": 8.923571451481967e-06,
+      "loss": 0.5728,
+      "step": 2071
+    },
+    {
+      "epoch": 0.3402787756861618,
+      "grad_norm": 0.3733366968858962,
+      "learning_rate": 8.923485361435397e-06,
+      "loss": 0.5897,
+      "step": 2072
+    },
+    {
+      "epoch": 0.3404430028945045,
+      "grad_norm": 0.31009524646895975,
+      "learning_rate": 8.92339922334542e-06,
+      "loss": 0.5529,
+      "step": 2073
+    },
+    {
+      "epoch": 0.34060723010284727,
+      "grad_norm": 0.2956517683672337,
+      "learning_rate": 8.923313037212977e-06,
+      "loss": 0.6152,
+      "step": 2074
+    },
+    {
+      "epoch": 0.34077145731119,
+      "grad_norm": 0.3332125958408152,
+      "learning_rate": 8.923226803039e-06,
+      "loss": 0.6156,
+      "step": 2075
+    },
+    {
+      "epoch": 0.34093568451953277,
+      "grad_norm": 0.3180632021721819,
+      "learning_rate": 8.923140520824432e-06,
+      "loss": 0.583,
+      "step": 2076
+    },
+    {
+      "epoch": 0.3410999117278755,
+      "grad_norm": 0.35586058964375455,
+      "learning_rate": 8.923054190570204e-06,
+      "loss": 0.5819,
+      "step": 2077
+    },
+    {
+      "epoch": 0.34126413893621826,
+      "grad_norm": 0.3998068688551633,
+      "learning_rate": 8.922967812277256e-06,
+      "loss": 0.5787,
+      "step": 2078
+    },
+    {
+      "epoch": 0.341428366144561,
+      "grad_norm": 0.30357534998154256,
+      "learning_rate": 8.922881385946526e-06,
+      "loss": 0.5593,
+      "step": 2079
+    },
+    {
+      "epoch": 0.34159259335290376,
+      "grad_norm": 0.5561696146995321,
+      "learning_rate": 8.922794911578954e-06,
+      "loss": 0.5973,
+      "step": 2080
+    },
+    {
+      "epoch": 0.3417568205612465,
+      "grad_norm": 0.40185317262477926,
+      "learning_rate": 8.922708389175476e-06,
+      "loss": 0.5947,
+      "step": 2081
+    },
+    {
+      "epoch": 0.3419210477695892,
+      "grad_norm": 0.30390626963582257,
+      "learning_rate": 8.922621818737033e-06,
+      "loss": 0.5678,
+      "step": 2082
+    },
+    {
+      "epoch": 0.34208527497793195,
+      "grad_norm": 0.3413233665139855,
+      "learning_rate": 8.922535200264568e-06,
+      "loss": 0.5982,
+      "step": 2083
+    },
+    {
+      "epoch": 0.3422495021862747,
+      "grad_norm": 0.28579448814830777,
+      "learning_rate": 8.922448533759017e-06,
+      "loss": 0.5848,
+      "step": 2084
+    },
+    {
+      "epoch": 0.34241372939461745,
+      "grad_norm": 0.3204345352536137,
+      "learning_rate": 8.922361819221326e-06,
+      "loss": 0.5757,
+      "step": 2085
+    },
+    {
+      "epoch": 0.3425779566029602,
+      "grad_norm": 0.3142059988489119,
+      "learning_rate": 8.922275056652434e-06,
+      "loss": 0.5934,
+      "step": 2086
+    },
+    {
+      "epoch": 0.34274218381130295,
+      "grad_norm": 0.3307567711086455,
+      "learning_rate": 8.922188246053284e-06,
+      "loss": 0.5934,
+      "step": 2087
+    },
+    {
+      "epoch": 0.3429064110196457,
+      "grad_norm": 0.31046467756377955,
+      "learning_rate": 8.922101387424818e-06,
+      "loss": 0.5882,
+      "step": 2088
+    },
+    {
+      "epoch": 0.34307063822798844,
+      "grad_norm": 0.3820278829809434,
+      "learning_rate": 8.922014480767981e-06,
+      "loss": 0.5954,
+      "step": 2089
+    },
+    {
+      "epoch": 0.34323486543633114,
+      "grad_norm": 0.335264516354752,
+      "learning_rate": 8.921927526083716e-06,
+      "loss": 0.5987,
+      "step": 2090
+    },
+    {
+      "epoch": 0.3433990926446739,
+      "grad_norm": 0.35109540453615623,
+      "learning_rate": 8.921840523372967e-06,
+      "loss": 0.5886,
+      "step": 2091
+    },
+    {
+      "epoch": 0.34356331985301664,
+      "grad_norm": 0.3404371948804148,
+      "learning_rate": 8.92175347263668e-06,
+      "loss": 0.612,
+      "step": 2092
+    },
+    {
+      "epoch": 0.3437275470613594,
+      "grad_norm": 0.3226414596776337,
+      "learning_rate": 8.9216663738758e-06,
+      "loss": 0.5992,
+      "step": 2093
+    },
+    {
+      "epoch": 0.34389177426970213,
+      "grad_norm": 0.3638859412166717,
+      "learning_rate": 8.921579227091272e-06,
+      "loss": 0.5853,
+      "step": 2094
+    },
+    {
+      "epoch": 0.3440560014780449,
+      "grad_norm": 0.29899585682019175,
+      "learning_rate": 8.921492032284043e-06,
+      "loss": 0.5919,
+      "step": 2095
+    },
+    {
+      "epoch": 0.34422022868638763,
+      "grad_norm": 0.3039192178553265,
+      "learning_rate": 8.921404789455061e-06,
+      "loss": 0.5763,
+      "step": 2096
+    },
+    {
+      "epoch": 0.3443844558947304,
+      "grad_norm": 0.3465925369057025,
+      "learning_rate": 8.921317498605274e-06,
+      "loss": 0.593,
+      "step": 2097
+    },
+    {
+      "epoch": 0.34454868310307313,
+      "grad_norm": 0.3332403460427041,
+      "learning_rate": 8.921230159735627e-06,
+      "loss": 0.6263,
+      "step": 2098
+    },
+    {
+      "epoch": 0.3447129103114158,
+      "grad_norm": 0.3097195369061601,
+      "learning_rate": 8.921142772847073e-06,
+      "loss": 0.5772,
+      "step": 2099
+    },
+    {
+      "epoch": 0.34487713751975857,
+      "grad_norm": 0.2961247616477229,
+      "learning_rate": 8.921055337940556e-06,
+      "loss": 0.5517,
+      "step": 2100
+    },
+    {
+      "epoch": 0.3450413647281013,
+      "grad_norm": 0.3645824498707989,
+      "learning_rate": 8.92096785501703e-06,
+      "loss": 0.568,
+      "step": 2101
+    },
+    {
+      "epoch": 0.34520559193644407,
+      "grad_norm": 0.3490365788059674,
+      "learning_rate": 8.920880324077443e-06,
+      "loss": 0.5953,
+      "step": 2102
+    },
+    {
+      "epoch": 0.3453698191447868,
+      "grad_norm": 0.3786183967790828,
+      "learning_rate": 8.920792745122747e-06,
+      "loss": 0.6042,
+      "step": 2103
+    },
+    {
+      "epoch": 0.34553404635312956,
+      "grad_norm": 0.3290400716459685,
+      "learning_rate": 8.92070511815389e-06,
+      "loss": 0.5824,
+      "step": 2104
+    },
+    {
+      "epoch": 0.3456982735614723,
+      "grad_norm": 0.34834661911635995,
+      "learning_rate": 8.920617443171828e-06,
+      "loss": 0.5908,
+      "step": 2105
+    },
+    {
+      "epoch": 0.34586250076981506,
+      "grad_norm": 0.37268445468777833,
+      "learning_rate": 8.920529720177512e-06,
+      "loss": 0.5783,
+      "step": 2106
+    },
+    {
+      "epoch": 0.34602672797815776,
+      "grad_norm": 0.31664534703004893,
+      "learning_rate": 8.92044194917189e-06,
+      "loss": 0.5798,
+      "step": 2107
+    },
+    {
+      "epoch": 0.3461909551865005,
+      "grad_norm": 0.3582045293008781,
+      "learning_rate": 8.920354130155924e-06,
+      "loss": 0.5913,
+      "step": 2108
+    },
+    {
+      "epoch": 0.34635518239484325,
+      "grad_norm": 0.5543690484261287,
+      "learning_rate": 8.92026626313056e-06,
+      "loss": 0.5881,
+      "step": 2109
+    },
+    {
+      "epoch": 0.346519409603186,
+      "grad_norm": 0.3208990175255881,
+      "learning_rate": 8.920178348096756e-06,
+      "loss": 0.5996,
+      "step": 2110
+    },
+    {
+      "epoch": 0.34668363681152875,
+      "grad_norm": 0.3016388152725111,
+      "learning_rate": 8.920090385055468e-06,
+      "loss": 0.6002,
+      "step": 2111
+    },
+    {
+      "epoch": 0.3468478640198715,
+      "grad_norm": 0.3602146787451213,
+      "learning_rate": 8.920002374007648e-06,
+      "loss": 0.5964,
+      "step": 2112
+    },
+    {
+      "epoch": 0.34701209122821425,
+      "grad_norm": 0.35331652465559116,
+      "learning_rate": 8.919914314954255e-06,
+      "loss": 0.5855,
+      "step": 2113
+    },
+    {
+      "epoch": 0.347176318436557,
+      "grad_norm": 0.30569388964646355,
+      "learning_rate": 8.919826207896243e-06,
+      "loss": 0.6103,
+      "step": 2114
+    },
+    {
+      "epoch": 0.34734054564489975,
+      "grad_norm": 0.3339711679529885,
+      "learning_rate": 8.919738052834569e-06,
+      "loss": 0.5792,
+      "step": 2115
+    },
+    {
+      "epoch": 0.34750477285324244,
+      "grad_norm": 0.31919676404834385,
+      "learning_rate": 8.919649849770193e-06,
+      "loss": 0.5673,
+      "step": 2116
+    },
+    {
+      "epoch": 0.3476690000615852,
+      "grad_norm": 0.3503739553239841,
+      "learning_rate": 8.91956159870407e-06,
+      "loss": 0.586,
+      "step": 2117
+    },
+    {
+      "epoch": 0.34783322726992794,
+      "grad_norm": 0.3263957410349147,
+      "learning_rate": 8.919473299637159e-06,
+      "loss": 0.5933,
+      "step": 2118
+    },
+    {
+      "epoch": 0.3479974544782707,
+      "grad_norm": 0.3206121735433827,
+      "learning_rate": 8.919384952570423e-06,
+      "loss": 0.5619,
+      "step": 2119
+    },
+    {
+      "epoch": 0.34816168168661343,
+      "grad_norm": 0.3672111356514065,
+      "learning_rate": 8.919296557504816e-06,
+      "loss": 0.5986,
+      "step": 2120
+    },
+    {
+      "epoch": 0.3483259088949562,
+      "grad_norm": 0.31992702542573537,
+      "learning_rate": 8.9192081144413e-06,
+      "loss": 0.5877,
+      "step": 2121
+    },
+    {
+      "epoch": 0.34849013610329893,
+      "grad_norm": 0.5187842764216102,
+      "learning_rate": 8.919119623380837e-06,
+      "loss": 0.5838,
+      "step": 2122
+    },
+    {
+      "epoch": 0.3486543633116417,
+      "grad_norm": 0.28609000136348256,
+      "learning_rate": 8.919031084324387e-06,
+      "loss": 0.5767,
+      "step": 2123
+    },
+    {
+      "epoch": 0.3488185905199844,
+      "grad_norm": 0.3176274746766672,
+      "learning_rate": 8.918942497272911e-06,
+      "loss": 0.5679,
+      "step": 2124
+    },
+    {
+      "epoch": 0.3489828177283271,
+      "grad_norm": 0.4791596535673938,
+      "learning_rate": 8.918853862227372e-06,
+      "loss": 0.5807,
+      "step": 2125
+    },
+    {
+      "epoch": 0.34914704493666987,
+      "grad_norm": 0.3121921995990465,
+      "learning_rate": 8.918765179188733e-06,
+      "loss": 0.5492,
+      "step": 2126
+    },
+    {
+      "epoch": 0.3493112721450126,
+      "grad_norm": 0.34081370981264214,
+      "learning_rate": 8.918676448157957e-06,
+      "loss": 0.5876,
+      "step": 2127
+    },
+    {
+      "epoch": 0.34947549935335537,
+      "grad_norm": 0.36611718567205803,
+      "learning_rate": 8.918587669136007e-06,
+      "loss": 0.5729,
+      "step": 2128
+    },
+    {
+      "epoch": 0.3496397265616981,
+      "grad_norm": 0.3122800636082036,
+      "learning_rate": 8.918498842123846e-06,
+      "loss": 0.568,
+      "step": 2129
+    },
+    {
+      "epoch": 0.34980395377004087,
+      "grad_norm": 0.33777800341019615,
+      "learning_rate": 8.918409967122443e-06,
+      "loss": 0.5903,
+      "step": 2130
+    },
+    {
+      "epoch": 0.3499681809783836,
+      "grad_norm": 0.3359295272643457,
+      "learning_rate": 8.91832104413276e-06,
+      "loss": 0.5659,
+      "step": 2131
+    },
+    {
+      "epoch": 0.35013240818672636,
+      "grad_norm": 0.3429887840973727,
+      "learning_rate": 8.918232073155762e-06,
+      "loss": 0.581,
+      "step": 2132
+    },
+    {
+      "epoch": 0.35029663539506906,
+      "grad_norm": 0.33042086580264834,
+      "learning_rate": 8.918143054192417e-06,
+      "loss": 0.592,
+      "step": 2133
+    },
+    {
+      "epoch": 0.3504608626034118,
+      "grad_norm": 0.3318582273308789,
+      "learning_rate": 8.918053987243692e-06,
+      "loss": 0.5953,
+      "step": 2134
+    },
+    {
+      "epoch": 0.35062508981175455,
+      "grad_norm": 0.35486222754880553,
+      "learning_rate": 8.917964872310555e-06,
+      "loss": 0.5634,
+      "step": 2135
+    },
+    {
+      "epoch": 0.3507893170200973,
+      "grad_norm": 0.2850475030435862,
+      "learning_rate": 8.91787570939397e-06,
+      "loss": 0.5726,
+      "step": 2136
+    },
+    {
+      "epoch": 0.35095354422844005,
+      "grad_norm": 0.47896412101886765,
+      "learning_rate": 8.917786498494912e-06,
+      "loss": 0.5956,
+      "step": 2137
+    },
+    {
+      "epoch": 0.3511177714367828,
+      "grad_norm": 0.3116157303094224,
+      "learning_rate": 8.917697239614343e-06,
+      "loss": 0.5734,
+      "step": 2138
+    },
+    {
+      "epoch": 0.35128199864512555,
+      "grad_norm": 0.30719351202702805,
+      "learning_rate": 8.917607932753237e-06,
+      "loss": 0.5558,
+      "step": 2139
+    },
+    {
+      "epoch": 0.3514462258534683,
+      "grad_norm": 0.3747383902637673,
+      "learning_rate": 8.917518577912562e-06,
+      "loss": 0.5724,
+      "step": 2140
+    },
+    {
+      "epoch": 0.351610453061811,
+      "grad_norm": 0.33882358226497106,
+      "learning_rate": 8.91742917509329e-06,
+      "loss": 0.5962,
+      "step": 2141
+    },
+    {
+      "epoch": 0.35177468027015374,
+      "grad_norm": 0.36082268139826695,
+      "learning_rate": 8.917339724296391e-06,
+      "loss": 0.5844,
+      "step": 2142
+    },
+    {
+      "epoch": 0.3519389074784965,
+      "grad_norm": 0.30132427096962017,
+      "learning_rate": 8.917250225522834e-06,
+      "loss": 0.5746,
+      "step": 2143
+    },
+    {
+      "epoch": 0.35210313468683924,
+      "grad_norm": 0.3129525632905135,
+      "learning_rate": 8.917160678773597e-06,
+      "loss": 0.5752,
+      "step": 2144
+    },
+    {
+      "epoch": 0.352267361895182,
+      "grad_norm": 0.3219843814868087,
+      "learning_rate": 8.917071084049647e-06,
+      "loss": 0.5968,
+      "step": 2145
+    },
+    {
+      "epoch": 0.35243158910352473,
+      "grad_norm": 0.3258109865865092,
+      "learning_rate": 8.91698144135196e-06,
+      "loss": 0.6133,
+      "step": 2146
+    },
+    {
+      "epoch": 0.3525958163118675,
+      "grad_norm": 0.3090823359585868,
+      "learning_rate": 8.916891750681508e-06,
+      "loss": 0.5866,
+      "step": 2147
+    },
+    {
+      "epoch": 0.35276004352021023,
+      "grad_norm": 0.32487573794202323,
+      "learning_rate": 8.916802012039267e-06,
+      "loss": 0.5595,
+      "step": 2148
+    },
+    {
+      "epoch": 0.352924270728553,
+      "grad_norm": 0.31033278772936684,
+      "learning_rate": 8.916712225426208e-06,
+      "loss": 0.5872,
+      "step": 2149
+    },
+    {
+      "epoch": 0.3530884979368957,
+      "grad_norm": 0.2833943755067742,
+      "learning_rate": 8.916622390843312e-06,
+      "loss": 0.5687,
+      "step": 2150
+    },
+    {
+      "epoch": 0.3532527251452384,
+      "grad_norm": 0.33627350008595963,
+      "learning_rate": 8.916532508291549e-06,
+      "loss": 0.5878,
+      "step": 2151
+    },
+    {
+      "epoch": 0.35341695235358117,
+      "grad_norm": 0.3469802742069985,
+      "learning_rate": 8.9164425777719e-06,
+      "loss": 0.6043,
+      "step": 2152
+    },
+    {
+      "epoch": 0.3535811795619239,
+      "grad_norm": 0.31575761311555894,
+      "learning_rate": 8.916352599285338e-06,
+      "loss": 0.5815,
+      "step": 2153
+    },
+    {
+      "epoch": 0.35374540677026667,
+      "grad_norm": 0.3045497966721819,
+      "learning_rate": 8.916262572832842e-06,
+      "loss": 0.5663,
+      "step": 2154
+    },
+    {
+      "epoch": 0.3539096339786094,
+      "grad_norm": 0.31121959786989556,
+      "learning_rate": 8.916172498415389e-06,
+      "loss": 0.5729,
+      "step": 2155
+    },
+    {
+      "epoch": 0.35407386118695217,
+      "grad_norm": 0.3627412802197949,
+      "learning_rate": 8.916082376033958e-06,
+      "loss": 0.5644,
+      "step": 2156
+    },
+    {
+      "epoch": 0.3542380883952949,
+      "grad_norm": 0.31693514999285666,
+      "learning_rate": 8.915992205689529e-06,
+      "loss": 0.5936,
+      "step": 2157
+    },
+    {
+      "epoch": 0.3544023156036376,
+      "grad_norm": 0.3855064420166456,
+      "learning_rate": 8.915901987383078e-06,
+      "loss": 0.5771,
+      "step": 2158
+    },
+    {
+      "epoch": 0.35456654281198036,
+      "grad_norm": 0.32831398862385386,
+      "learning_rate": 8.915811721115588e-06,
+      "loss": 0.6016,
+      "step": 2159
+    },
+    {
+      "epoch": 0.3547307700203231,
+      "grad_norm": 0.32057668168581144,
+      "learning_rate": 8.915721406888037e-06,
+      "loss": 0.582,
+      "step": 2160
+    },
+    {
+      "epoch": 0.35489499722866585,
+      "grad_norm": 0.3029317139589128,
+      "learning_rate": 8.915631044701408e-06,
+      "loss": 0.5853,
+      "step": 2161
+    },
+    {
+      "epoch": 0.3550592244370086,
+      "grad_norm": 0.3683147589988735,
+      "learning_rate": 8.915540634556681e-06,
+      "loss": 0.5618,
+      "step": 2162
+    },
+    {
+      "epoch": 0.35522345164535135,
+      "grad_norm": 0.32250564191378006,
+      "learning_rate": 8.915450176454838e-06,
+      "loss": 0.5949,
+      "step": 2163
+    },
+    {
+      "epoch": 0.3553876788536941,
+      "grad_norm": 0.34769282222308573,
+      "learning_rate": 8.915359670396863e-06,
+      "loss": 0.6056,
+      "step": 2164
+    },
+    {
+      "epoch": 0.35555190606203685,
+      "grad_norm": 0.32794507485016755,
+      "learning_rate": 8.915269116383736e-06,
+      "loss": 0.5667,
+      "step": 2165
+    },
+    {
+      "epoch": 0.3557161332703796,
+      "grad_norm": 0.3592878992842856,
+      "learning_rate": 8.915178514416443e-06,
+      "loss": 0.5734,
+      "step": 2166
+    },
+    {
+      "epoch": 0.3558803604787223,
+      "grad_norm": 0.31418949050424083,
+      "learning_rate": 8.91508786449597e-06,
+      "loss": 0.5647,
+      "step": 2167
+    },
+    {
+      "epoch": 0.35604458768706504,
+      "grad_norm": 0.41684345813587714,
+      "learning_rate": 8.914997166623295e-06,
+      "loss": 0.5661,
+      "step": 2168
+    },
+    {
+      "epoch": 0.3562088148954078,
+      "grad_norm": 0.33388499752938794,
+      "learning_rate": 8.91490642079941e-06,
+      "loss": 0.5724,
+      "step": 2169
+    },
+    {
+      "epoch": 0.35637304210375054,
+      "grad_norm": 0.30404975902522224,
+      "learning_rate": 8.914815627025295e-06,
+      "loss": 0.5754,
+      "step": 2170
+    },
+    {
+      "epoch": 0.3565372693120933,
+      "grad_norm": 0.3150577755264073,
+      "learning_rate": 8.914724785301942e-06,
+      "loss": 0.5769,
+      "step": 2171
+    },
+    {
+      "epoch": 0.35670149652043603,
+      "grad_norm": 0.3201787714157396,
+      "learning_rate": 8.91463389563033e-06,
+      "loss": 0.6045,
+      "step": 2172
+    },
+    {
+      "epoch": 0.3568657237287788,
+      "grad_norm": 0.32275307016076826,
+      "learning_rate": 8.914542958011454e-06,
+      "loss": 0.5873,
+      "step": 2173
+    },
+    {
+      "epoch": 0.35702995093712153,
+      "grad_norm": 0.29979459104094425,
+      "learning_rate": 8.914451972446297e-06,
+      "loss": 0.5786,
+      "step": 2174
+    },
+    {
+      "epoch": 0.3571941781454642,
+      "grad_norm": 0.625963371230108,
+      "learning_rate": 8.914360938935849e-06,
+      "loss": 0.5633,
+      "step": 2175
+    },
+    {
+      "epoch": 0.357358405353807,
+      "grad_norm": 0.3087502313684936,
+      "learning_rate": 8.914269857481098e-06,
+      "loss": 0.5823,
+      "step": 2176
+    },
+    {
+      "epoch": 0.3575226325621497,
+      "grad_norm": 0.32683326534937834,
+      "learning_rate": 8.914178728083031e-06,
+      "loss": 0.5778,
+      "step": 2177
+    },
+    {
+      "epoch": 0.35768685977049247,
+      "grad_norm": 0.7880735018196897,
+      "learning_rate": 8.914087550742643e-06,
+      "loss": 0.5851,
+      "step": 2178
+    },
+    {
+      "epoch": 0.3578510869788352,
+      "grad_norm": 0.3278097326230743,
+      "learning_rate": 8.91399632546092e-06,
+      "loss": 0.5914,
+      "step": 2179
+    },
+    {
+      "epoch": 0.35801531418717797,
+      "grad_norm": 0.35430290321922725,
+      "learning_rate": 8.913905052238854e-06,
+      "loss": 0.6058,
+      "step": 2180
+    },
+    {
+      "epoch": 0.3581795413955207,
+      "grad_norm": 1.0396310534287811,
+      "learning_rate": 8.913813731077437e-06,
+      "loss": 0.583,
+      "step": 2181
+    },
+    {
+      "epoch": 0.35834376860386347,
+      "grad_norm": 0.3060473871714537,
+      "learning_rate": 8.91372236197766e-06,
+      "loss": 0.5871,
+      "step": 2182
+    },
+    {
+      "epoch": 0.3585079958122062,
+      "grad_norm": 0.33881404244223196,
+      "learning_rate": 8.913630944940516e-06,
+      "loss": 0.5872,
+      "step": 2183
+    },
+    {
+      "epoch": 0.3586722230205489,
+      "grad_norm": 0.2973344128005135,
+      "learning_rate": 8.913539479966997e-06,
+      "loss": 0.5696,
+      "step": 2184
+    },
+    {
+      "epoch": 0.35883645022889166,
+      "grad_norm": 0.35184436201426394,
+      "learning_rate": 8.913447967058097e-06,
+      "loss": 0.5965,
+      "step": 2185
+    },
+    {
+      "epoch": 0.3590006774372344,
+      "grad_norm": 0.3020841641575663,
+      "learning_rate": 8.913356406214809e-06,
+      "loss": 0.562,
+      "step": 2186
+    },
+    {
+      "epoch": 0.35916490464557715,
+      "grad_norm": 0.4027338067146939,
+      "learning_rate": 8.91326479743813e-06,
+      "loss": 0.5637,
+      "step": 2187
+    },
+    {
+      "epoch": 0.3593291318539199,
+      "grad_norm": 0.297962563186807,
+      "learning_rate": 8.913173140729051e-06,
+      "loss": 0.584,
+      "step": 2188
+    },
+    {
+      "epoch": 0.35949335906226265,
+      "grad_norm": 0.49567057600017395,
+      "learning_rate": 8.913081436088573e-06,
+      "loss": 0.5698,
+      "step": 2189
+    },
+    {
+      "epoch": 0.3596575862706054,
+      "grad_norm": 0.3139466119859544,
+      "learning_rate": 8.912989683517686e-06,
+      "loss": 0.581,
+      "step": 2190
+    },
+    {
+      "epoch": 0.35982181347894815,
+      "grad_norm": 0.3222467409727742,
+      "learning_rate": 8.91289788301739e-06,
+      "loss": 0.5692,
+      "step": 2191
+    },
+    {
+      "epoch": 0.35998604068729084,
+      "grad_norm": 0.3132899288065772,
+      "learning_rate": 8.912806034588682e-06,
+      "loss": 0.5716,
+      "step": 2192
+    },
+    {
+      "epoch": 0.3601502678956336,
+      "grad_norm": 0.311666768458647,
+      "learning_rate": 8.912714138232558e-06,
+      "loss": 0.5707,
+      "step": 2193
+    },
+    {
+      "epoch": 0.36031449510397634,
+      "grad_norm": 0.318265739153915,
+      "learning_rate": 8.912622193950016e-06,
+      "loss": 0.6014,
+      "step": 2194
+    },
+    {
+      "epoch": 0.3604787223123191,
+      "grad_norm": 0.33405275857775724,
+      "learning_rate": 8.912530201742057e-06,
+      "loss": 0.5796,
+      "step": 2195
+    },
+    {
+      "epoch": 0.36064294952066184,
+      "grad_norm": 0.32054434013299965,
+      "learning_rate": 8.912438161609678e-06,
+      "loss": 0.5737,
+      "step": 2196
+    },
+    {
+      "epoch": 0.3608071767290046,
+      "grad_norm": 0.4906329903361965,
+      "learning_rate": 8.912346073553882e-06,
+      "loss": 0.5959,
+      "step": 2197
+    },
+    {
+      "epoch": 0.36097140393734733,
+      "grad_norm": 0.6300646881591123,
+      "learning_rate": 8.912253937575663e-06,
+      "loss": 0.5996,
+      "step": 2198
+    },
+    {
+      "epoch": 0.3611356311456901,
+      "grad_norm": 0.32082175087580206,
+      "learning_rate": 8.912161753676026e-06,
+      "loss": 0.5648,
+      "step": 2199
+    },
+    {
+      "epoch": 0.3612998583540328,
+      "grad_norm": 0.3143112057835378,
+      "learning_rate": 8.912069521855971e-06,
+      "loss": 0.5699,
+      "step": 2200
+    },
+    {
+      "epoch": 0.3614640855623755,
+      "grad_norm": 0.3295077023756852,
+      "learning_rate": 8.911977242116502e-06,
+      "loss": 0.5756,
+      "step": 2201
+    },
+    {
+      "epoch": 0.3616283127707183,
+      "grad_norm": 0.33517369326913704,
+      "learning_rate": 8.911884914458618e-06,
+      "loss": 0.5851,
+      "step": 2202
+    },
+    {
+      "epoch": 0.361792539979061,
+      "grad_norm": 0.33051215135799833,
+      "learning_rate": 8.911792538883323e-06,
+      "loss": 0.5722,
+      "step": 2203
+    },
+    {
+      "epoch": 0.36195676718740377,
+      "grad_norm": 0.43832424126361785,
+      "learning_rate": 8.91170011539162e-06,
+      "loss": 0.5514,
+      "step": 2204
+    },
+    {
+      "epoch": 0.3621209943957465,
+      "grad_norm": 0.37071372638553013,
+      "learning_rate": 8.911607643984513e-06,
+      "loss": 0.5964,
+      "step": 2205
+    },
+    {
+      "epoch": 0.36228522160408927,
+      "grad_norm": 0.4086726046467989,
+      "learning_rate": 8.91151512466301e-06,
+      "loss": 0.5983,
+      "step": 2206
+    },
+    {
+      "epoch": 0.362449448812432,
+      "grad_norm": 0.3148352832312435,
+      "learning_rate": 8.91142255742811e-06,
+      "loss": 0.5952,
+      "step": 2207
+    },
+    {
+      "epoch": 0.36261367602077477,
+      "grad_norm": 0.30697433895023185,
+      "learning_rate": 8.91132994228082e-06,
+      "loss": 0.5793,
+      "step": 2208
+    },
+    {
+      "epoch": 0.36277790322911746,
+      "grad_norm": 0.3825985510743253,
+      "learning_rate": 8.911237279222148e-06,
+      "loss": 0.5774,
+      "step": 2209
+    },
+    {
+      "epoch": 0.3629421304374602,
+      "grad_norm": 0.34276141854757924,
+      "learning_rate": 8.911144568253097e-06,
+      "loss": 0.5758,
+      "step": 2210
+    },
+    {
+      "epoch": 0.36310635764580296,
+      "grad_norm": 0.3142215471342378,
+      "learning_rate": 8.911051809374677e-06,
+      "loss": 0.5792,
+      "step": 2211
+    },
+    {
+      "epoch": 0.3632705848541457,
+      "grad_norm": 0.30903065304902716,
+      "learning_rate": 8.910959002587895e-06,
+      "loss": 0.5866,
+      "step": 2212
+    },
+    {
+      "epoch": 0.36343481206248845,
+      "grad_norm": 0.31135335714393814,
+      "learning_rate": 8.910866147893758e-06,
+      "loss": 0.5868,
+      "step": 2213
+    },
+    {
+      "epoch": 0.3635990392708312,
+      "grad_norm": 0.3764236674444635,
+      "learning_rate": 8.910773245293275e-06,
+      "loss": 0.5889,
+      "step": 2214
+    },
+    {
+      "epoch": 0.36376326647917395,
+      "grad_norm": 0.40469774842863443,
+      "learning_rate": 8.910680294787455e-06,
+      "loss": 0.566,
+      "step": 2215
+    },
+    {
+      "epoch": 0.3639274936875167,
+      "grad_norm": 0.3145755387386865,
+      "learning_rate": 8.910587296377308e-06,
+      "loss": 0.5788,
+      "step": 2216
+    },
+    {
+      "epoch": 0.3640917208958594,
+      "grad_norm": 0.35885193000909843,
+      "learning_rate": 8.910494250063844e-06,
+      "loss": 0.5677,
+      "step": 2217
+    },
+    {
+      "epoch": 0.36425594810420214,
+      "grad_norm": 0.3651850799070969,
+      "learning_rate": 8.910401155848072e-06,
+      "loss": 0.5735,
+      "step": 2218
+    },
+    {
+      "epoch": 0.3644201753125449,
+      "grad_norm": 0.29711733435878707,
+      "learning_rate": 8.910308013731004e-06,
+      "loss": 0.5525,
+      "step": 2219
+    },
+    {
+      "epoch": 0.36458440252088764,
+      "grad_norm": 0.3236062586888815,
+      "learning_rate": 8.910214823713652e-06,
+      "loss": 0.592,
+      "step": 2220
+    },
+    {
+      "epoch": 0.3647486297292304,
+      "grad_norm": 0.33402951253417174,
+      "learning_rate": 8.910121585797028e-06,
+      "loss": 0.5915,
+      "step": 2221
+    },
+    {
+      "epoch": 0.36491285693757314,
+      "grad_norm": 0.29300268294595966,
+      "learning_rate": 8.910028299982145e-06,
+      "loss": 0.5813,
+      "step": 2222
+    },
+    {
+      "epoch": 0.3650770841459159,
+      "grad_norm": 0.3091653000574013,
+      "learning_rate": 8.909934966270016e-06,
+      "loss": 0.5804,
+      "step": 2223
+    },
+    {
+      "epoch": 0.36524131135425864,
+      "grad_norm": 0.3079107397662537,
+      "learning_rate": 8.909841584661654e-06,
+      "loss": 0.5748,
+      "step": 2224
+    },
+    {
+      "epoch": 0.3654055385626014,
+      "grad_norm": 0.34259914622682275,
+      "learning_rate": 8.909748155158074e-06,
+      "loss": 0.5796,
+      "step": 2225
+    },
+    {
+      "epoch": 0.3655697657709441,
+      "grad_norm": 0.4153545557740463,
+      "learning_rate": 8.90965467776029e-06,
+      "loss": 0.5734,
+      "step": 2226
+    },
+    {
+      "epoch": 0.3657339929792868,
+      "grad_norm": 0.31071013419276294,
+      "learning_rate": 8.909561152469317e-06,
+      "loss": 0.5905,
+      "step": 2227
+    },
+    {
+      "epoch": 0.3658982201876296,
+      "grad_norm": 0.3471544762094001,
+      "learning_rate": 8.909467579286173e-06,
+      "loss": 0.5755,
+      "step": 2228
+    },
+    {
+      "epoch": 0.3660624473959723,
+      "grad_norm": 0.3216553400318287,
+      "learning_rate": 8.909373958211872e-06,
+      "loss": 0.5566,
+      "step": 2229
+    },
+    {
+      "epoch": 0.3662266746043151,
+      "grad_norm": 0.3292521251192245,
+      "learning_rate": 8.909280289247431e-06,
+      "loss": 0.5803,
+      "step": 2230
+    },
+    {
+      "epoch": 0.3663909018126578,
+      "grad_norm": 0.3344410080798811,
+      "learning_rate": 8.90918657239387e-06,
+      "loss": 0.5777,
+      "step": 2231
+    },
+    {
+      "epoch": 0.36655512902100057,
+      "grad_norm": 0.2941625541245194,
+      "learning_rate": 8.909092807652202e-06,
+      "loss": 0.6021,
+      "step": 2232
+    },
+    {
+      "epoch": 0.3667193562293433,
+      "grad_norm": 0.35239833707095725,
+      "learning_rate": 8.908998995023449e-06,
+      "loss": 0.5848,
+      "step": 2233
+    },
+    {
+      "epoch": 0.366883583437686,
+      "grad_norm": 0.34132808655561586,
+      "learning_rate": 8.908905134508631e-06,
+      "loss": 0.5662,
+      "step": 2234
+    },
+    {
+      "epoch": 0.36704781064602876,
+      "grad_norm": 0.5185030895359983,
+      "learning_rate": 8.908811226108764e-06,
+      "loss": 0.5913,
+      "step": 2235
+    },
+    {
+      "epoch": 0.3672120378543715,
+      "grad_norm": 0.30832542073870384,
+      "learning_rate": 8.90871726982487e-06,
+      "loss": 0.6005,
+      "step": 2236
+    },
+    {
+      "epoch": 0.36737626506271426,
+      "grad_norm": 0.32855997531791475,
+      "learning_rate": 8.908623265657966e-06,
+      "loss": 0.5756,
+      "step": 2237
+    },
+    {
+      "epoch": 0.367540492271057,
+      "grad_norm": 0.30127305901284296,
+      "learning_rate": 8.90852921360908e-06,
+      "loss": 0.5881,
+      "step": 2238
+    },
+    {
+      "epoch": 0.36770471947939976,
+      "grad_norm": 0.2993924136214141,
+      "learning_rate": 8.908435113679226e-06,
+      "loss": 0.5682,
+      "step": 2239
+    },
+    {
+      "epoch": 0.3678689466877425,
+      "grad_norm": 0.3406570591462043,
+      "learning_rate": 8.908340965869432e-06,
+      "loss": 0.5739,
+      "step": 2240
+    },
+    {
+      "epoch": 0.36803317389608525,
+      "grad_norm": 0.33425544277809804,
+      "learning_rate": 8.908246770180716e-06,
+      "loss": 0.5857,
+      "step": 2241
+    },
+    {
+      "epoch": 0.368197401104428,
+      "grad_norm": 0.3471600230497452,
+      "learning_rate": 8.908152526614104e-06,
+      "loss": 0.5689,
+      "step": 2242
+    },
+    {
+      "epoch": 0.3683616283127707,
+      "grad_norm": 0.31874708068845486,
+      "learning_rate": 8.90805823517062e-06,
+      "loss": 0.5846,
+      "step": 2243
+    },
+    {
+      "epoch": 0.36852585552111344,
+      "grad_norm": 0.3397332860106541,
+      "learning_rate": 8.907963895851282e-06,
+      "loss": 0.5983,
+      "step": 2244
+    },
+    {
+      "epoch": 0.3686900827294562,
+      "grad_norm": 0.3454421184183436,
+      "learning_rate": 8.907869508657122e-06,
+      "loss": 0.5703,
+      "step": 2245
+    },
+    {
+      "epoch": 0.36885430993779894,
+      "grad_norm": 0.31162260048303864,
+      "learning_rate": 8.907775073589163e-06,
+      "loss": 0.5857,
+      "step": 2246
+    },
+    {
+      "epoch": 0.3690185371461417,
+      "grad_norm": 0.326196691003036,
+      "learning_rate": 8.907680590648429e-06,
+      "loss": 0.5868,
+      "step": 2247
+    },
+    {
+      "epoch": 0.36918276435448444,
+      "grad_norm": 0.3186926099650805,
+      "learning_rate": 8.907586059835948e-06,
+      "loss": 0.6031,
+      "step": 2248
+    },
+    {
+      "epoch": 0.3693469915628272,
+      "grad_norm": 0.36454682926659043,
+      "learning_rate": 8.907491481152747e-06,
+      "loss": 0.601,
+      "step": 2249
+    },
+    {
+      "epoch": 0.36951121877116994,
+      "grad_norm": 0.3407203715992103,
+      "learning_rate": 8.90739685459985e-06,
+      "loss": 0.5824,
+      "step": 2250
+    },
+    {
+      "epoch": 0.36967544597951263,
+      "grad_norm": 0.27972540167378696,
+      "learning_rate": 8.907302180178286e-06,
+      "loss": 0.5686,
+      "step": 2251
+    },
+    {
+      "epoch": 0.3698396731878554,
+      "grad_norm": 0.3208480967722726,
+      "learning_rate": 8.907207457889087e-06,
+      "loss": 0.5886,
+      "step": 2252
+    },
+    {
+      "epoch": 0.3700039003961981,
+      "grad_norm": 0.3041652148393244,
+      "learning_rate": 8.907112687733278e-06,
+      "loss": 0.582,
+      "step": 2253
+    },
+    {
+      "epoch": 0.3701681276045409,
+      "grad_norm": 0.3174281025630536,
+      "learning_rate": 8.907017869711888e-06,
+      "loss": 0.5669,
+      "step": 2254
+    },
+    {
+      "epoch": 0.3703323548128836,
+      "grad_norm": 0.2875806634656067,
+      "learning_rate": 8.906923003825949e-06,
+      "loss": 0.5741,
+      "step": 2255
+    },
+    {
+      "epoch": 0.3704965820212264,
+      "grad_norm": 0.3209729400410392,
+      "learning_rate": 8.90682809007649e-06,
+      "loss": 0.5761,
+      "step": 2256
+    },
+    {
+      "epoch": 0.3706608092295691,
+      "grad_norm": 0.3300546032520837,
+      "learning_rate": 8.906733128464541e-06,
+      "loss": 0.5587,
+      "step": 2257
+    },
+    {
+      "epoch": 0.37082503643791187,
+      "grad_norm": 0.291617247856832,
+      "learning_rate": 8.906638118991137e-06,
+      "loss": 0.5801,
+      "step": 2258
+    },
+    {
+      "epoch": 0.3709892636462546,
+      "grad_norm": 0.35873974625749094,
+      "learning_rate": 8.906543061657306e-06,
+      "loss": 0.5652,
+      "step": 2259
+    },
+    {
+      "epoch": 0.3711534908545973,
+      "grad_norm": 0.2995194684429335,
+      "learning_rate": 8.906447956464082e-06,
+      "loss": 0.5818,
+      "step": 2260
+    },
+    {
+      "epoch": 0.37131771806294006,
+      "grad_norm": 0.28613397814477437,
+      "learning_rate": 8.906352803412499e-06,
+      "loss": 0.5497,
+      "step": 2261
+    },
+    {
+      "epoch": 0.3714819452712828,
+      "grad_norm": 0.35714670764867573,
+      "learning_rate": 8.906257602503589e-06,
+      "loss": 0.5854,
+      "step": 2262
+    },
+    {
+      "epoch": 0.37164617247962556,
+      "grad_norm": 0.314102556822799,
+      "learning_rate": 8.906162353738385e-06,
+      "loss": 0.5927,
+      "step": 2263
+    },
+    {
+      "epoch": 0.3718103996879683,
+      "grad_norm": 0.31461753863692804,
+      "learning_rate": 8.906067057117924e-06,
+      "loss": 0.559,
+      "step": 2264
+    },
+    {
+      "epoch": 0.37197462689631106,
+      "grad_norm": 0.3202526289895275,
+      "learning_rate": 8.905971712643238e-06,
+      "loss": 0.5908,
+      "step": 2265
+    },
+    {
+      "epoch": 0.3721388541046538,
+      "grad_norm": 0.29543179263334973,
+      "learning_rate": 8.905876320315367e-06,
+      "loss": 0.5434,
+      "step": 2266
+    },
+    {
+      "epoch": 0.37230308131299655,
+      "grad_norm": 0.3062964276019493,
+      "learning_rate": 8.905780880135343e-06,
+      "loss": 0.5745,
+      "step": 2267
+    },
+    {
+      "epoch": 0.37246730852133925,
+      "grad_norm": 0.28951540649656626,
+      "learning_rate": 8.905685392104203e-06,
+      "loss": 0.587,
+      "step": 2268
+    },
+    {
+      "epoch": 0.372631535729682,
+      "grad_norm": 0.3192492106256812,
+      "learning_rate": 8.905589856222985e-06,
+      "loss": 0.5776,
+      "step": 2269
+    },
+    {
+      "epoch": 0.37279576293802474,
+      "grad_norm": 0.33482973901264274,
+      "learning_rate": 8.905494272492728e-06,
+      "loss": 0.5691,
+      "step": 2270
+    },
+    {
+      "epoch": 0.3729599901463675,
+      "grad_norm": 0.2875150584599915,
+      "learning_rate": 8.905398640914468e-06,
+      "loss": 0.5927,
+      "step": 2271
+    },
+    {
+      "epoch": 0.37312421735471024,
+      "grad_norm": 0.3076711414341829,
+      "learning_rate": 8.905302961489245e-06,
+      "loss": 0.5755,
+      "step": 2272
+    },
+    {
+      "epoch": 0.373288444563053,
+      "grad_norm": 0.29479734170962885,
+      "learning_rate": 8.905207234218098e-06,
+      "loss": 0.5966,
+      "step": 2273
+    },
+    {
+      "epoch": 0.37345267177139574,
+      "grad_norm": 0.28667253454249403,
+      "learning_rate": 8.905111459102065e-06,
+      "loss": 0.5847,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3736168989797385,
+      "grad_norm": 0.37883062566583625,
+      "learning_rate": 8.905015636142189e-06,
+      "loss": 0.5881,
+      "step": 2275
+    },
+    {
+      "epoch": 0.37378112618808124,
+      "grad_norm": 0.29767160803116754,
+      "learning_rate": 8.904919765339508e-06,
+      "loss": 0.5627,
+      "step": 2276
+    },
+    {
+      "epoch": 0.37394535339642393,
+      "grad_norm": 0.3155297745770699,
+      "learning_rate": 8.904823846695065e-06,
+      "loss": 0.5693,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3741095806047667,
+      "grad_norm": 0.30626645197578384,
+      "learning_rate": 8.904727880209902e-06,
+      "loss": 0.5739,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3742738078131094,
+      "grad_norm": 0.2870421361283258,
+      "learning_rate": 8.904631865885059e-06,
+      "loss": 0.564,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3744380350214522,
+      "grad_norm": 0.3041713714865128,
+      "learning_rate": 8.904535803721581e-06,
+      "loss": 0.581,
+      "step": 2280
+    },
+    {
+      "epoch": 0.3746022622297949,
+      "grad_norm": 0.3342741300499172,
+      "learning_rate": 8.904439693720511e-06,
+      "loss": 0.5672,
+      "step": 2281
+    },
+    {
+      "epoch": 0.3747664894381377,
+      "grad_norm": 0.29876866247680856,
+      "learning_rate": 8.904343535882892e-06,
+      "loss": 0.5563,
+      "step": 2282
+    },
+    {
+      "epoch": 0.3749307166464804,
+      "grad_norm": 0.3272388991501618,
+      "learning_rate": 8.90424733020977e-06,
+      "loss": 0.5733,
+      "step": 2283
+    },
+    {
+      "epoch": 0.37509494385482317,
+      "grad_norm": 0.29504478377135096,
+      "learning_rate": 8.904151076702187e-06,
+      "loss": 0.5582,
+      "step": 2284
+    },
+    {
+      "epoch": 0.37525917106316586,
+      "grad_norm": 0.32151438651497766,
+      "learning_rate": 8.904054775361191e-06,
+      "loss": 0.5808,
+      "step": 2285
+    },
+    {
+      "epoch": 0.3754233982715086,
+      "grad_norm": 0.29408465592424793,
+      "learning_rate": 8.903958426187827e-06,
+      "loss": 0.6031,
+      "step": 2286
+    },
+    {
+      "epoch": 0.37558762547985136,
+      "grad_norm": 0.39603717369803254,
+      "learning_rate": 8.903862029183144e-06,
+      "loss": 0.5646,
+      "step": 2287
+    },
+    {
+      "epoch": 0.3757518526881941,
+      "grad_norm": 0.27728565310231973,
+      "learning_rate": 8.903765584348183e-06,
+      "loss": 0.5764,
+      "step": 2288
+    },
+    {
+      "epoch": 0.37591607989653686,
+      "grad_norm": 0.3228980017218485,
+      "learning_rate": 8.903669091683996e-06,
+      "loss": 0.5758,
+      "step": 2289
+    },
+    {
+      "epoch": 0.3760803071048796,
+      "grad_norm": 0.27396890144473407,
+      "learning_rate": 8.903572551191633e-06,
+      "loss": 0.5904,
+      "step": 2290
+    },
+    {
+      "epoch": 0.37624453431322236,
+      "grad_norm": 0.33219915905771785,
+      "learning_rate": 8.903475962872135e-06,
+      "loss": 0.588,
+      "step": 2291
+    },
+    {
+      "epoch": 0.3764087615215651,
+      "grad_norm": 0.29442976439850005,
+      "learning_rate": 8.903379326726559e-06,
+      "loss": 0.5726,
+      "step": 2292
+    },
+    {
+      "epoch": 0.37657298872990785,
+      "grad_norm": 0.27379664077338145,
+      "learning_rate": 8.903282642755948e-06,
+      "loss": 0.5569,
+      "step": 2293
+    },
+    {
+      "epoch": 0.37673721593825055,
+      "grad_norm": 0.309011494993213,
+      "learning_rate": 8.903185910961358e-06,
+      "loss": 0.5866,
+      "step": 2294
+    },
+    {
+      "epoch": 0.3769014431465933,
+      "grad_norm": 0.34027174363674284,
+      "learning_rate": 8.903089131343835e-06,
+      "loss": 0.5794,
+      "step": 2295
+    },
+    {
+      "epoch": 0.37706567035493604,
+      "grad_norm": 0.2904502113115749,
+      "learning_rate": 8.902992303904435e-06,
+      "loss": 0.6026,
+      "step": 2296
+    },
+    {
+      "epoch": 0.3772298975632788,
+      "grad_norm": 0.38170797705184983,
+      "learning_rate": 8.902895428644203e-06,
+      "loss": 0.5842,
+      "step": 2297
+    },
+    {
+      "epoch": 0.37739412477162154,
+      "grad_norm": 0.31501550061600586,
+      "learning_rate": 8.902798505564198e-06,
+      "loss": 0.5805,
+      "step": 2298
+    },
+    {
+      "epoch": 0.3775583519799643,
+      "grad_norm": 0.33472044182482413,
+      "learning_rate": 8.902701534665467e-06,
+      "loss": 0.596,
+      "step": 2299
+    },
+    {
+      "epoch": 0.37772257918830704,
+      "grad_norm": 0.29453287123397076,
+      "learning_rate": 8.902604515949067e-06,
+      "loss": 0.5808,
+      "step": 2300
+    },
+    {
+      "epoch": 0.3778868063966498,
+      "grad_norm": 0.3003032769783216,
+      "learning_rate": 8.90250744941605e-06,
+      "loss": 0.595,
+      "step": 2301
+    },
+    {
+      "epoch": 0.3780510336049925,
+      "grad_norm": 0.33000419779183837,
+      "learning_rate": 8.90241033506747e-06,
+      "loss": 0.5461,
+      "step": 2302
+    },
+    {
+      "epoch": 0.37821526081333523,
+      "grad_norm": 0.44537302571064064,
+      "learning_rate": 8.902313172904383e-06,
+      "loss": 0.5843,
+      "step": 2303
+    },
+    {
+      "epoch": 0.378379488021678,
+      "grad_norm": 0.27528935880347627,
+      "learning_rate": 8.902215962927844e-06,
+      "loss": 0.5868,
+      "step": 2304
+    },
+    {
+      "epoch": 0.3785437152300207,
+      "grad_norm": 0.30776693033578106,
+      "learning_rate": 8.902118705138908e-06,
+      "loss": 0.5892,
+      "step": 2305
+    },
+    {
+      "epoch": 0.3787079424383635,
+      "grad_norm": 0.4128017693757589,
+      "learning_rate": 8.90202139953863e-06,
+      "loss": 0.5512,
+      "step": 2306
+    },
+    {
+      "epoch": 0.3788721696467062,
+      "grad_norm": 0.29928028112716526,
+      "learning_rate": 8.901924046128072e-06,
+      "loss": 0.5752,
+      "step": 2307
+    },
+    {
+      "epoch": 0.379036396855049,
+      "grad_norm": 0.2854681450585411,
+      "learning_rate": 8.901826644908287e-06,
+      "loss": 0.5643,
+      "step": 2308
+    },
+    {
+      "epoch": 0.3792006240633917,
+      "grad_norm": 0.33566978961389754,
+      "learning_rate": 8.901729195880332e-06,
+      "loss": 0.5894,
+      "step": 2309
+    },
+    {
+      "epoch": 0.37936485127173447,
+      "grad_norm": 0.3031308669451596,
+      "learning_rate": 8.90163169904527e-06,
+      "loss": 0.5769,
+      "step": 2310
+    },
+    {
+      "epoch": 0.37952907848007716,
+      "grad_norm": 0.3232660855580926,
+      "learning_rate": 8.901534154404154e-06,
+      "loss": 0.5928,
+      "step": 2311
+    },
+    {
+      "epoch": 0.3796933056884199,
+      "grad_norm": 0.3708012017505623,
+      "learning_rate": 8.901436561958048e-06,
+      "loss": 0.5831,
+      "step": 2312
+    },
+    {
+      "epoch": 0.37985753289676266,
+      "grad_norm": 0.31052685319187673,
+      "learning_rate": 8.90133892170801e-06,
+      "loss": 0.5587,
+      "step": 2313
+    },
+    {
+      "epoch": 0.3800217601051054,
+      "grad_norm": 0.32187830125576383,
+      "learning_rate": 8.901241233655103e-06,
+      "loss": 0.5717,
+      "step": 2314
+    },
+    {
+      "epoch": 0.38018598731344816,
+      "grad_norm": 0.43759240032581914,
+      "learning_rate": 8.901143497800383e-06,
+      "loss": 0.5554,
+      "step": 2315
+    },
+    {
+      "epoch": 0.3803502145217909,
+      "grad_norm": 0.44601495555487336,
+      "learning_rate": 8.901045714144916e-06,
+      "loss": 0.5956,
+      "step": 2316
+    },
+    {
+      "epoch": 0.38051444173013366,
+      "grad_norm": 0.33538126500137966,
+      "learning_rate": 8.900947882689763e-06,
+      "loss": 0.567,
+      "step": 2317
+    },
+    {
+      "epoch": 0.3806786689384764,
+      "grad_norm": 0.31306894560268406,
+      "learning_rate": 8.900850003435985e-06,
+      "loss": 0.5624,
+      "step": 2318
+    },
+    {
+      "epoch": 0.3808428961468191,
+      "grad_norm": 0.3111850571560804,
+      "learning_rate": 8.900752076384648e-06,
+      "loss": 0.5789,
+      "step": 2319
+    },
+    {
+      "epoch": 0.38100712335516185,
+      "grad_norm": 0.3030241102758526,
+      "learning_rate": 8.900654101536811e-06,
+      "loss": 0.5693,
+      "step": 2320
+    },
+    {
+      "epoch": 0.3811713505635046,
+      "grad_norm": 0.33699935182531765,
+      "learning_rate": 8.900556078893542e-06,
+      "loss": 0.5959,
+      "step": 2321
+    },
+    {
+      "epoch": 0.38133557777184734,
+      "grad_norm": 0.3544044560119914,
+      "learning_rate": 8.900458008455905e-06,
+      "loss": 0.5713,
+      "step": 2322
+    },
+    {
+      "epoch": 0.3814998049801901,
+      "grad_norm": 0.3177507153530851,
+      "learning_rate": 8.900359890224965e-06,
+      "loss": 0.5984,
+      "step": 2323
+    },
+    {
+      "epoch": 0.38166403218853284,
+      "grad_norm": 0.3175167714364075,
+      "learning_rate": 8.900261724201786e-06,
+      "loss": 0.5915,
+      "step": 2324
+    },
+    {
+      "epoch": 0.3818282593968756,
+      "grad_norm": 0.4009151151987348,
+      "learning_rate": 8.900163510387436e-06,
+      "loss": 0.5518,
+      "step": 2325
+    },
+    {
+      "epoch": 0.38199248660521834,
+      "grad_norm": 0.3754271825257931,
+      "learning_rate": 8.900065248782981e-06,
+      "loss": 0.6039,
+      "step": 2326
+    },
+    {
+      "epoch": 0.3821567138135611,
+      "grad_norm": 0.32691754192502154,
+      "learning_rate": 8.899966939389488e-06,
+      "loss": 0.5507,
+      "step": 2327
+    },
+    {
+      "epoch": 0.3823209410219038,
+      "grad_norm": 0.28998835197412,
+      "learning_rate": 8.899868582208024e-06,
+      "loss": 0.5652,
+      "step": 2328
+    },
+    {
+      "epoch": 0.38248516823024653,
+      "grad_norm": 0.3133390970464335,
+      "learning_rate": 8.89977017723966e-06,
+      "loss": 0.5767,
+      "step": 2329
+    },
+    {
+      "epoch": 0.3826493954385893,
+      "grad_norm": 0.27878076670015955,
+      "learning_rate": 8.899671724485463e-06,
+      "loss": 0.5876,
+      "step": 2330
+    },
+    {
+      "epoch": 0.38281362264693203,
+      "grad_norm": 0.3268019065018182,
+      "learning_rate": 8.899573223946502e-06,
+      "loss": 0.5673,
+      "step": 2331
+    },
+    {
+      "epoch": 0.3829778498552748,
+      "grad_norm": 0.373397306799112,
+      "learning_rate": 8.899474675623847e-06,
+      "loss": 0.5799,
+      "step": 2332
+    },
+    {
+      "epoch": 0.3831420770636175,
+      "grad_norm": 0.2975325766072855,
+      "learning_rate": 8.89937607951857e-06,
+      "loss": 0.5762,
+      "step": 2333
+    },
+    {
+      "epoch": 0.3833063042719603,
+      "grad_norm": 0.28508597358780635,
+      "learning_rate": 8.899277435631738e-06,
+      "loss": 0.5858,
+      "step": 2334
+    },
+    {
+      "epoch": 0.383470531480303,
+      "grad_norm": 0.4014745772227359,
+      "learning_rate": 8.899178743964426e-06,
+      "loss": 0.5544,
+      "step": 2335
+    },
+    {
+      "epoch": 0.3836347586886457,
+      "grad_norm": 0.32360223924724896,
+      "learning_rate": 8.899080004517704e-06,
+      "loss": 0.5718,
+      "step": 2336
+    },
+    {
+      "epoch": 0.38379898589698846,
+      "grad_norm": 0.33739313880409866,
+      "learning_rate": 8.898981217292645e-06,
+      "loss": 0.5623,
+      "step": 2337
+    },
+    {
+      "epoch": 0.3839632131053312,
+      "grad_norm": 0.3172906694171004,
+      "learning_rate": 8.898882382290323e-06,
+      "loss": 0.5796,
+      "step": 2338
+    },
+    {
+      "epoch": 0.38412744031367396,
+      "grad_norm": 0.35624301408414016,
+      "learning_rate": 8.89878349951181e-06,
+      "loss": 0.5828,
+      "step": 2339
+    },
+    {
+      "epoch": 0.3842916675220167,
+      "grad_norm": 0.3434168572205274,
+      "learning_rate": 8.89868456895818e-06,
+      "loss": 0.5946,
+      "step": 2340
+    },
+    {
+      "epoch": 0.38445589473035946,
+      "grad_norm": 0.3153538586937991,
+      "learning_rate": 8.898585590630508e-06,
+      "loss": 0.593,
+      "step": 2341
+    },
+    {
+      "epoch": 0.3846201219387022,
+      "grad_norm": 0.296511438451358,
+      "learning_rate": 8.89848656452987e-06,
+      "loss": 0.5789,
+      "step": 2342
+    },
+    {
+      "epoch": 0.38478434914704496,
+      "grad_norm": 0.42450348545740313,
+      "learning_rate": 8.898387490657339e-06,
+      "loss": 0.5797,
+      "step": 2343
+    },
+    {
+      "epoch": 0.3849485763553877,
+      "grad_norm": 0.33579453521532565,
+      "learning_rate": 8.898288369013993e-06,
+      "loss": 0.5962,
+      "step": 2344
+    },
+    {
+      "epoch": 0.3851128035637304,
+      "grad_norm": 0.30616008807308986,
+      "learning_rate": 8.898189199600907e-06,
+      "loss": 0.5813,
+      "step": 2345
+    },
+    {
+      "epoch": 0.38527703077207315,
+      "grad_norm": 0.36008697096086706,
+      "learning_rate": 8.89808998241916e-06,
+      "loss": 0.5833,
+      "step": 2346
+    },
+    {
+      "epoch": 0.3854412579804159,
+      "grad_norm": 0.30608114375905615,
+      "learning_rate": 8.897990717469828e-06,
+      "loss": 0.5581,
+      "step": 2347
+    },
+    {
+      "epoch": 0.38560548518875865,
+      "grad_norm": 0.29685505491002356,
+      "learning_rate": 8.89789140475399e-06,
+      "loss": 0.577,
+      "step": 2348
+    },
+    {
+      "epoch": 0.3857697123971014,
+      "grad_norm": 0.3025244608028047,
+      "learning_rate": 8.897792044272724e-06,
+      "loss": 0.5567,
+      "step": 2349
+    },
+    {
+      "epoch": 0.38593393960544414,
+      "grad_norm": 0.2933770229759006,
+      "learning_rate": 8.897692636027112e-06,
+      "loss": 0.5592,
+      "step": 2350
+    },
+    {
+      "epoch": 0.3860981668137869,
+      "grad_norm": 0.3232920174640066,
+      "learning_rate": 8.89759318001823e-06,
+      "loss": 0.5668,
+      "step": 2351
+    },
+    {
+      "epoch": 0.38626239402212964,
+      "grad_norm": 0.3288617268332227,
+      "learning_rate": 8.897493676247158e-06,
+      "loss": 0.5708,
+      "step": 2352
+    },
+    {
+      "epoch": 0.38642662123047233,
+      "grad_norm": 0.36540336033645326,
+      "learning_rate": 8.897394124714979e-06,
+      "loss": 0.5605,
+      "step": 2353
+    },
+    {
+      "epoch": 0.3865908484388151,
+      "grad_norm": 0.38478628819551197,
+      "learning_rate": 8.897294525422773e-06,
+      "loss": 0.5824,
+      "step": 2354
+    },
+    {
+      "epoch": 0.38675507564715783,
+      "grad_norm": 0.3239640598548664,
+      "learning_rate": 8.897194878371623e-06,
+      "loss": 0.5794,
+      "step": 2355
+    },
+    {
+      "epoch": 0.3869193028555006,
+      "grad_norm": 0.31349793588556835,
+      "learning_rate": 8.897095183562609e-06,
+      "loss": 0.5636,
+      "step": 2356
+    },
+    {
+      "epoch": 0.38708353006384333,
+      "grad_norm": 0.3038538132161622,
+      "learning_rate": 8.896995440996816e-06,
+      "loss": 0.5791,
+      "step": 2357
+    },
+    {
+      "epoch": 0.3872477572721861,
+      "grad_norm": 0.32051112577468094,
+      "learning_rate": 8.896895650675327e-06,
+      "loss": 0.5791,
+      "step": 2358
+    },
+    {
+      "epoch": 0.3874119844805288,
+      "grad_norm": 0.44838209737281975,
+      "learning_rate": 8.896795812599224e-06,
+      "loss": 0.5915,
+      "step": 2359
+    },
+    {
+      "epoch": 0.3875762116888716,
+      "grad_norm": 0.34659678851515363,
+      "learning_rate": 8.896695926769594e-06,
+      "loss": 0.5726,
+      "step": 2360
+    },
+    {
+      "epoch": 0.3877404388972143,
+      "grad_norm": 0.2887359708809979,
+      "learning_rate": 8.89659599318752e-06,
+      "loss": 0.5854,
+      "step": 2361
+    },
+    {
+      "epoch": 0.387904666105557,
+      "grad_norm": 0.3166935243544558,
+      "learning_rate": 8.896496011854087e-06,
+      "loss": 0.582,
+      "step": 2362
+    },
+    {
+      "epoch": 0.38806889331389977,
+      "grad_norm": 0.2931378659260637,
+      "learning_rate": 8.896395982770382e-06,
+      "loss": 0.5595,
+      "step": 2363
+    },
+    {
+      "epoch": 0.3882331205222425,
+      "grad_norm": 0.35930294339419266,
+      "learning_rate": 8.896295905937492e-06,
+      "loss": 0.5803,
+      "step": 2364
+    },
+    {
+      "epoch": 0.38839734773058526,
+      "grad_norm": 0.3112712071807427,
+      "learning_rate": 8.896195781356502e-06,
+      "loss": 0.5941,
+      "step": 2365
+    },
+    {
+      "epoch": 0.388561574938928,
+      "grad_norm": 0.2942703416825157,
+      "learning_rate": 8.896095609028501e-06,
+      "loss": 0.5783,
+      "step": 2366
+    },
+    {
+      "epoch": 0.38872580214727076,
+      "grad_norm": 0.3352361065283778,
+      "learning_rate": 8.895995388954577e-06,
+      "loss": 0.559,
+      "step": 2367
+    },
+    {
+      "epoch": 0.3888900293556135,
+      "grad_norm": 0.2917202118240349,
+      "learning_rate": 8.895895121135819e-06,
+      "loss": 0.5731,
+      "step": 2368
+    },
+    {
+      "epoch": 0.38905425656395626,
+      "grad_norm": 0.30769072743532866,
+      "learning_rate": 8.895794805573313e-06,
+      "loss": 0.5623,
+      "step": 2369
+    },
+    {
+      "epoch": 0.38921848377229895,
+      "grad_norm": 0.2893103915184653,
+      "learning_rate": 8.89569444226815e-06,
+      "loss": 0.5903,
+      "step": 2370
+    },
+    {
+      "epoch": 0.3893827109806417,
+      "grad_norm": 0.29857635371328933,
+      "learning_rate": 8.895594031221423e-06,
+      "loss": 0.5766,
+      "step": 2371
+    },
+    {
+      "epoch": 0.38954693818898445,
+      "grad_norm": 0.5167094917444112,
+      "learning_rate": 8.895493572434218e-06,
+      "loss": 0.5776,
+      "step": 2372
+    },
+    {
+      "epoch": 0.3897111653973272,
+      "grad_norm": 0.358184481637941,
+      "learning_rate": 8.89539306590763e-06,
+      "loss": 0.595,
+      "step": 2373
+    },
+    {
+      "epoch": 0.38987539260566995,
+      "grad_norm": 1.3947990141046052,
+      "learning_rate": 8.895292511642748e-06,
+      "loss": 0.5791,
+      "step": 2374
+    },
+    {
+      "epoch": 0.3900396198140127,
+      "grad_norm": 0.33588084377988386,
+      "learning_rate": 8.895191909640665e-06,
+      "loss": 0.5948,
+      "step": 2375
+    },
+    {
+      "epoch": 0.39020384702235544,
+      "grad_norm": 0.42192446454016386,
+      "learning_rate": 8.895091259902472e-06,
+      "loss": 0.5804,
+      "step": 2376
+    },
+    {
+      "epoch": 0.3903680742306982,
+      "grad_norm": 0.3417045707043655,
+      "learning_rate": 8.894990562429265e-06,
+      "loss": 0.5712,
+      "step": 2377
+    },
+    {
+      "epoch": 0.39053230143904094,
+      "grad_norm": 0.32637787783707956,
+      "learning_rate": 8.894889817222138e-06,
+      "loss": 0.59,
+      "step": 2378
+    },
+    {
+      "epoch": 0.39069652864738363,
+      "grad_norm": 0.3589500141121288,
+      "learning_rate": 8.894789024282181e-06,
+      "loss": 0.578,
+      "step": 2379
+    },
+    {
+      "epoch": 0.3908607558557264,
+      "grad_norm": 0.3401379685320087,
+      "learning_rate": 8.894688183610494e-06,
+      "loss": 0.5698,
+      "step": 2380
+    },
+    {
+      "epoch": 0.39102498306406913,
+      "grad_norm": 0.31926923165948984,
+      "learning_rate": 8.894587295208167e-06,
+      "loss": 0.5883,
+      "step": 2381
+    },
+    {
+      "epoch": 0.3911892102724119,
+      "grad_norm": 0.3215207946196231,
+      "learning_rate": 8.8944863590763e-06,
+      "loss": 0.5943,
+      "step": 2382
+    },
+    {
+      "epoch": 0.39135343748075463,
+      "grad_norm": 0.3453997776935591,
+      "learning_rate": 8.894385375215987e-06,
+      "loss": 0.5744,
+      "step": 2383
+    },
+    {
+      "epoch": 0.3915176646890974,
+      "grad_norm": 0.3461759429970046,
+      "learning_rate": 8.894284343628326e-06,
+      "loss": 0.5682,
+      "step": 2384
+    },
+    {
+      "epoch": 0.3916818918974401,
+      "grad_norm": 0.33305618278830995,
+      "learning_rate": 8.89418326431441e-06,
+      "loss": 0.6009,
+      "step": 2385
+    },
+    {
+      "epoch": 0.3918461191057829,
+      "grad_norm": 0.34941470397732544,
+      "learning_rate": 8.894082137275344e-06,
+      "loss": 0.5922,
+      "step": 2386
+    },
+    {
+      "epoch": 0.39201034631412557,
+      "grad_norm": 0.33931275242971576,
+      "learning_rate": 8.893980962512224e-06,
+      "loss": 0.5559,
+      "step": 2387
+    },
+    {
+      "epoch": 0.3921745735224683,
+      "grad_norm": 0.43197181831368625,
+      "learning_rate": 8.893879740026146e-06,
+      "loss": 0.5432,
+      "step": 2388
+    },
+    {
+      "epoch": 0.39233880073081107,
+      "grad_norm": 0.4445672617305598,
+      "learning_rate": 8.893778469818211e-06,
+      "loss": 0.5849,
+      "step": 2389
+    },
+    {
+      "epoch": 0.3925030279391538,
+      "grad_norm": 0.30662557274934144,
+      "learning_rate": 8.893677151889517e-06,
+      "loss": 0.5479,
+      "step": 2390
+    },
+    {
+      "epoch": 0.39266725514749656,
+      "grad_norm": 0.35586055709357794,
+      "learning_rate": 8.893575786241168e-06,
+      "loss": 0.5713,
+      "step": 2391
+    },
+    {
+      "epoch": 0.3928314823558393,
+      "grad_norm": 0.3416283784802478,
+      "learning_rate": 8.893474372874264e-06,
+      "loss": 0.5692,
+      "step": 2392
+    },
+    {
+      "epoch": 0.39299570956418206,
+      "grad_norm": 0.37964865251282476,
+      "learning_rate": 8.893372911789904e-06,
+      "loss": 0.5615,
+      "step": 2393
+    },
+    {
+      "epoch": 0.3931599367725248,
+      "grad_norm": 0.30654274690443206,
+      "learning_rate": 8.893271402989193e-06,
+      "loss": 0.5878,
+      "step": 2394
+    },
+    {
+      "epoch": 0.3933241639808675,
+      "grad_norm": 0.3098855741105631,
+      "learning_rate": 8.893169846473233e-06,
+      "loss": 0.586,
+      "step": 2395
+    },
+    {
+      "epoch": 0.39348839118921025,
+      "grad_norm": 0.3284469381158421,
+      "learning_rate": 8.893068242243126e-06,
+      "loss": 0.5656,
+      "step": 2396
+    },
+    {
+      "epoch": 0.393652618397553,
+      "grad_norm": 0.3457270508908126,
+      "learning_rate": 8.892966590299975e-06,
+      "loss": 0.5804,
+      "step": 2397
+    },
+    {
+      "epoch": 0.39381684560589575,
+      "grad_norm": 0.3824770008300639,
+      "learning_rate": 8.892864890644882e-06,
+      "loss": 0.5445,
+      "step": 2398
+    },
+    {
+      "epoch": 0.3939810728142385,
+      "grad_norm": 0.2890234530985488,
+      "learning_rate": 8.892763143278958e-06,
+      "loss": 0.5591,
+      "step": 2399
+    },
+    {
+      "epoch": 0.39414530002258125,
+      "grad_norm": 0.2999344078752705,
+      "learning_rate": 8.892661348203304e-06,
+      "loss": 0.5752,
+      "step": 2400
+    },
+    {
+      "epoch": 0.394309527230924,
+      "grad_norm": 0.2966419849742581,
+      "learning_rate": 8.892559505419023e-06,
+      "loss": 0.5673,
+      "step": 2401
+    },
+    {
+      "epoch": 0.39447375443926674,
+      "grad_norm": 0.3193420409937522,
+      "learning_rate": 8.892457614927228e-06,
+      "loss": 0.5612,
+      "step": 2402
+    },
+    {
+      "epoch": 0.3946379816476095,
+      "grad_norm": 0.3291083690324307,
+      "learning_rate": 8.89235567672902e-06,
+      "loss": 0.5676,
+      "step": 2403
+    },
+    {
+      "epoch": 0.3948022088559522,
+      "grad_norm": 0.3398061099802675,
+      "learning_rate": 8.892253690825507e-06,
+      "loss": 0.5447,
+      "step": 2404
+    },
+    {
+      "epoch": 0.39496643606429493,
+      "grad_norm": 0.436882175126899,
+      "learning_rate": 8.892151657217799e-06,
+      "loss": 0.5661,
+      "step": 2405
+    },
+    {
+      "epoch": 0.3951306632726377,
+      "grad_norm": 0.9383550668483664,
+      "learning_rate": 8.892049575907003e-06,
+      "loss": 0.5763,
+      "step": 2406
+    },
+    {
+      "epoch": 0.39529489048098043,
+      "grad_norm": 0.3557356759695701,
+      "learning_rate": 8.891947446894224e-06,
+      "loss": 0.5721,
+      "step": 2407
+    },
+    {
+      "epoch": 0.3954591176893232,
+      "grad_norm": 0.40604062140883007,
+      "learning_rate": 8.891845270180578e-06,
+      "loss": 0.577,
+      "step": 2408
+    },
+    {
+      "epoch": 0.39562334489766593,
+      "grad_norm": 0.3646391321735781,
+      "learning_rate": 8.89174304576717e-06,
+      "loss": 0.5777,
+      "step": 2409
+    },
+    {
+      "epoch": 0.3957875721060087,
+      "grad_norm": 0.3310293733714123,
+      "learning_rate": 8.891640773655112e-06,
+      "loss": 0.5613,
+      "step": 2410
+    },
+    {
+      "epoch": 0.3959517993143514,
+      "grad_norm": 0.4039936546858141,
+      "learning_rate": 8.891538453845515e-06,
+      "loss": 0.5776,
+      "step": 2411
+    },
+    {
+      "epoch": 0.3961160265226941,
+      "grad_norm": 0.39043339053206516,
+      "learning_rate": 8.891436086339489e-06,
+      "loss": 0.5647,
+      "step": 2412
+    },
+    {
+      "epoch": 0.39628025373103687,
+      "grad_norm": 0.3213534464575592,
+      "learning_rate": 8.891333671138146e-06,
+      "loss": 0.5711,
+      "step": 2413
+    },
+    {
+      "epoch": 0.3964444809393796,
+      "grad_norm": 0.31102891837891017,
+      "learning_rate": 8.8912312082426e-06,
+      "loss": 0.5837,
+      "step": 2414
+    },
+    {
+      "epoch": 0.39660870814772237,
+      "grad_norm": 0.3046339082276663,
+      "learning_rate": 8.891128697653962e-06,
+      "loss": 0.5554,
+      "step": 2415
+    },
+    {
+      "epoch": 0.3967729353560651,
+      "grad_norm": 0.31346347835751853,
+      "learning_rate": 8.891026139373347e-06,
+      "loss": 0.5792,
+      "step": 2416
+    },
+    {
+      "epoch": 0.39693716256440786,
+      "grad_norm": 0.3290758388183897,
+      "learning_rate": 8.890923533401866e-06,
+      "loss": 0.5893,
+      "step": 2417
+    },
+    {
+      "epoch": 0.3971013897727506,
+      "grad_norm": 0.32363239445426245,
+      "learning_rate": 8.890820879740636e-06,
+      "loss": 0.5926,
+      "step": 2418
+    },
+    {
+      "epoch": 0.39726561698109336,
+      "grad_norm": 0.3397402640059377,
+      "learning_rate": 8.890718178390772e-06,
+      "loss": 0.5573,
+      "step": 2419
+    },
+    {
+      "epoch": 0.3974298441894361,
+      "grad_norm": 0.33337105589432475,
+      "learning_rate": 8.89061542935339e-06,
+      "loss": 0.5788,
+      "step": 2420
+    },
+    {
+      "epoch": 0.3975940713977788,
+      "grad_norm": 0.3040636722728571,
+      "learning_rate": 8.890512632629603e-06,
+      "loss": 0.5735,
+      "step": 2421
+    },
+    {
+      "epoch": 0.39775829860612155,
+      "grad_norm": 0.3486589140163701,
+      "learning_rate": 8.89040978822053e-06,
+      "loss": 0.5698,
+      "step": 2422
+    },
+    {
+      "epoch": 0.3979225258144643,
+      "grad_norm": 0.33530465006472016,
+      "learning_rate": 8.890306896127285e-06,
+      "loss": 0.5824,
+      "step": 2423
+    },
+    {
+      "epoch": 0.39808675302280705,
+      "grad_norm": 0.416663784834214,
+      "learning_rate": 8.890203956350989e-06,
+      "loss": 0.5673,
+      "step": 2424
+    },
+    {
+      "epoch": 0.3982509802311498,
+      "grad_norm": 0.31480558408985104,
+      "learning_rate": 8.89010096889276e-06,
+      "loss": 0.5696,
+      "step": 2425
+    },
+    {
+      "epoch": 0.39841520743949255,
+      "grad_norm": 0.3087118076140605,
+      "learning_rate": 8.889997933753713e-06,
+      "loss": 0.5764,
+      "step": 2426
+    },
+    {
+      "epoch": 0.3985794346478353,
+      "grad_norm": 0.33372436198265837,
+      "learning_rate": 8.88989485093497e-06,
+      "loss": 0.5898,
+      "step": 2427
+    },
+    {
+      "epoch": 0.39874366185617804,
+      "grad_norm": 0.35881328155788444,
+      "learning_rate": 8.88979172043765e-06,
+      "loss": 0.5731,
+      "step": 2428
+    },
+    {
+      "epoch": 0.39890788906452074,
+      "grad_norm": 0.30641486846057925,
+      "learning_rate": 8.889688542262872e-06,
+      "loss": 0.5621,
+      "step": 2429
+    },
+    {
+      "epoch": 0.3990721162728635,
+      "grad_norm": 0.3188642436177513,
+      "learning_rate": 8.889585316411759e-06,
+      "loss": 0.5958,
+      "step": 2430
+    },
+    {
+      "epoch": 0.39923634348120624,
+      "grad_norm": 0.3586864652327105,
+      "learning_rate": 8.88948204288543e-06,
+      "loss": 0.5696,
+      "step": 2431
+    },
+    {
+      "epoch": 0.399400570689549,
+      "grad_norm": 0.32446607347051026,
+      "learning_rate": 8.889378721685008e-06,
+      "loss": 0.5801,
+      "step": 2432
+    },
+    {
+      "epoch": 0.39956479789789173,
+      "grad_norm": 0.3091050200854687,
+      "learning_rate": 8.889275352811614e-06,
+      "loss": 0.5882,
+      "step": 2433
+    },
+    {
+      "epoch": 0.3997290251062345,
+      "grad_norm": 0.4026782667136338,
+      "learning_rate": 8.889171936266373e-06,
+      "loss": 0.5563,
+      "step": 2434
+    },
+    {
+      "epoch": 0.39989325231457723,
+      "grad_norm": 0.3830299878740365,
+      "learning_rate": 8.889068472050405e-06,
+      "loss": 0.5828,
+      "step": 2435
+    },
+    {
+      "epoch": 0.40005747952292,
+      "grad_norm": 0.35592783342450557,
+      "learning_rate": 8.888964960164833e-06,
+      "loss": 0.5734,
+      "step": 2436
+    },
+    {
+      "epoch": 0.4002217067312627,
+      "grad_norm": 0.31061412781069053,
+      "learning_rate": 8.888861400610786e-06,
+      "loss": 0.5967,
+      "step": 2437
+    },
+    {
+      "epoch": 0.4003859339396054,
+      "grad_norm": 0.35688423964889243,
+      "learning_rate": 8.888757793389384e-06,
+      "loss": 0.5738,
+      "step": 2438
+    },
+    {
+      "epoch": 0.40055016114794817,
+      "grad_norm": 0.30702503029090233,
+      "learning_rate": 8.888654138501756e-06,
+      "loss": 0.5848,
+      "step": 2439
+    },
+    {
+      "epoch": 0.4007143883562909,
+      "grad_norm": 0.4147640287492882,
+      "learning_rate": 8.888550435949027e-06,
+      "loss": 0.5759,
+      "step": 2440
+    },
+    {
+      "epoch": 0.40087861556463367,
+      "grad_norm": 0.3049971946792247,
+      "learning_rate": 8.888446685732321e-06,
+      "loss": 0.5723,
+      "step": 2441
+    },
+    {
+      "epoch": 0.4010428427729764,
+      "grad_norm": 0.2979275672901426,
+      "learning_rate": 8.888342887852767e-06,
+      "loss": 0.5836,
+      "step": 2442
+    },
+    {
+      "epoch": 0.40120706998131916,
+      "grad_norm": 0.31789230808393515,
+      "learning_rate": 8.88823904231149e-06,
+      "loss": 0.5758,
+      "step": 2443
+    },
+    {
+      "epoch": 0.4013712971896619,
+      "grad_norm": 0.2969921000323856,
+      "learning_rate": 8.888135149109623e-06,
+      "loss": 0.5829,
+      "step": 2444
+    },
+    {
+      "epoch": 0.40153552439800466,
+      "grad_norm": 0.32598594902859673,
+      "learning_rate": 8.888031208248288e-06,
+      "loss": 0.5533,
+      "step": 2445
+    },
+    {
+      "epoch": 0.40169975160634736,
+      "grad_norm": 0.3055143376194248,
+      "learning_rate": 8.887927219728618e-06,
+      "loss": 0.5745,
+      "step": 2446
+    },
+    {
+      "epoch": 0.4018639788146901,
+      "grad_norm": 0.3290788171475274,
+      "learning_rate": 8.887823183551741e-06,
+      "loss": 0.5758,
+      "step": 2447
+    },
+    {
+      "epoch": 0.40202820602303285,
+      "grad_norm": 0.3092126257798601,
+      "learning_rate": 8.887719099718788e-06,
+      "loss": 0.5568,
+      "step": 2448
+    },
+    {
+      "epoch": 0.4021924332313756,
+      "grad_norm": 0.31991904996752984,
+      "learning_rate": 8.887614968230888e-06,
+      "loss": 0.5766,
+      "step": 2449
+    },
+    {
+      "epoch": 0.40235666043971835,
+      "grad_norm": 0.2857049219979474,
+      "learning_rate": 8.887510789089173e-06,
+      "loss": 0.5905,
+      "step": 2450
+    },
+    {
+      "epoch": 0.4025208876480611,
+      "grad_norm": 0.3171922371979184,
+      "learning_rate": 8.887406562294774e-06,
+      "loss": 0.5537,
+      "step": 2451
+    },
+    {
+      "epoch": 0.40268511485640385,
+      "grad_norm": 0.3223877063261562,
+      "learning_rate": 8.887302287848822e-06,
+      "loss": 0.578,
+      "step": 2452
+    },
+    {
+      "epoch": 0.4028493420647466,
+      "grad_norm": 0.2973553918001752,
+      "learning_rate": 8.887197965752452e-06,
+      "loss": 0.5664,
+      "step": 2453
+    },
+    {
+      "epoch": 0.40301356927308934,
+      "grad_norm": 0.2895585733719346,
+      "learning_rate": 8.887093596006794e-06,
+      "loss": 0.5725,
+      "step": 2454
+    },
+    {
+      "epoch": 0.40317779648143204,
+      "grad_norm": 0.33434995224312225,
+      "learning_rate": 8.886989178612985e-06,
+      "loss": 0.5889,
+      "step": 2455
+    },
+    {
+      "epoch": 0.4033420236897748,
+      "grad_norm": 0.31888170721341874,
+      "learning_rate": 8.886884713572157e-06,
+      "loss": 0.5661,
+      "step": 2456
+    },
+    {
+      "epoch": 0.40350625089811754,
+      "grad_norm": 0.2937884224933003,
+      "learning_rate": 8.886780200885444e-06,
+      "loss": 0.5522,
+      "step": 2457
+    },
+    {
+      "epoch": 0.4036704781064603,
+      "grad_norm": 0.3394152732680013,
+      "learning_rate": 8.886675640553981e-06,
+      "loss": 0.5622,
+      "step": 2458
+    },
+    {
+      "epoch": 0.40383470531480303,
+      "grad_norm": 0.32314358205589594,
+      "learning_rate": 8.886571032578906e-06,
+      "loss": 0.5775,
+      "step": 2459
+    },
+    {
+      "epoch": 0.4039989325231458,
+      "grad_norm": 0.40212521213858143,
+      "learning_rate": 8.886466376961355e-06,
+      "loss": 0.5769,
+      "step": 2460
+    },
+    {
+      "epoch": 0.40416315973148853,
+      "grad_norm": 0.3058600193689216,
+      "learning_rate": 8.886361673702463e-06,
+      "loss": 0.582,
+      "step": 2461
+    },
+    {
+      "epoch": 0.4043273869398313,
+      "grad_norm": 0.27062081866814136,
+      "learning_rate": 8.886256922803368e-06,
+      "loss": 0.5452,
+      "step": 2462
+    },
+    {
+      "epoch": 0.404491614148174,
+      "grad_norm": 0.3095302292598992,
+      "learning_rate": 8.886152124265205e-06,
+      "loss": 0.5701,
+      "step": 2463
+    },
+    {
+      "epoch": 0.4046558413565167,
+      "grad_norm": 0.3196683063197642,
+      "learning_rate": 8.886047278089117e-06,
+      "loss": 0.5688,
+      "step": 2464
+    },
+    {
+      "epoch": 0.40482006856485947,
+      "grad_norm": 0.3747769365378931,
+      "learning_rate": 8.885942384276238e-06,
+      "loss": 0.5602,
+      "step": 2465
+    },
+    {
+      "epoch": 0.4049842957732022,
+      "grad_norm": 0.3314657584893606,
+      "learning_rate": 8.88583744282771e-06,
+      "loss": 0.568,
+      "step": 2466
+    },
+    {
+      "epoch": 0.40514852298154497,
+      "grad_norm": 0.2921241951052027,
+      "learning_rate": 8.885732453744673e-06,
+      "loss": 0.5694,
+      "step": 2467
+    },
+    {
+      "epoch": 0.4053127501898877,
+      "grad_norm": 0.31684539227316827,
+      "learning_rate": 8.885627417028266e-06,
+      "loss": 0.5706,
+      "step": 2468
+    },
+    {
+      "epoch": 0.40547697739823046,
+      "grad_norm": 0.29919740238742754,
+      "learning_rate": 8.885522332679632e-06,
+      "loss": 0.5611,
+      "step": 2469
+    },
+    {
+      "epoch": 0.4056412046065732,
+      "grad_norm": 0.2927552985408856,
+      "learning_rate": 8.88541720069991e-06,
+      "loss": 0.5614,
+      "step": 2470
+    },
+    {
+      "epoch": 0.40580543181491596,
+      "grad_norm": 0.33353867719732694,
+      "learning_rate": 8.885312021090242e-06,
+      "loss": 0.5638,
+      "step": 2471
+    },
+    {
+      "epoch": 0.40596965902325866,
+      "grad_norm": 0.2925862099678924,
+      "learning_rate": 8.885206793851771e-06,
+      "loss": 0.5577,
+      "step": 2472
+    },
+    {
+      "epoch": 0.4061338862316014,
+      "grad_norm": 0.31469465854680573,
+      "learning_rate": 8.88510151898564e-06,
+      "loss": 0.5658,
+      "step": 2473
+    },
+    {
+      "epoch": 0.40629811343994415,
+      "grad_norm": 0.3254583303924738,
+      "learning_rate": 8.884996196492992e-06,
+      "loss": 0.5728,
+      "step": 2474
+    },
+    {
+      "epoch": 0.4064623406482869,
+      "grad_norm": 0.3475809377362272,
+      "learning_rate": 8.88489082637497e-06,
+      "loss": 0.5621,
+      "step": 2475
+    },
+    {
+      "epoch": 0.40662656785662965,
+      "grad_norm": 0.29271987278075334,
+      "learning_rate": 8.88478540863272e-06,
+      "loss": 0.5867,
+      "step": 2476
+    },
+    {
+      "epoch": 0.4067907950649724,
+      "grad_norm": 0.29606614439458234,
+      "learning_rate": 8.884679943267387e-06,
+      "loss": 0.5621,
+      "step": 2477
+    },
+    {
+      "epoch": 0.40695502227331515,
+      "grad_norm": 0.2990525558360778,
+      "learning_rate": 8.884574430280117e-06,
+      "loss": 0.5771,
+      "step": 2478
+    },
+    {
+      "epoch": 0.4071192494816579,
+      "grad_norm": 0.34870302304095335,
+      "learning_rate": 8.884468869672053e-06,
+      "loss": 0.5587,
+      "step": 2479
+    },
+    {
+      "epoch": 0.4072834766900006,
+      "grad_norm": 0.3581627456016973,
+      "learning_rate": 8.884363261444344e-06,
+      "loss": 0.566,
+      "step": 2480
+    },
+    {
+      "epoch": 0.40744770389834334,
+      "grad_norm": 0.35359658435859415,
+      "learning_rate": 8.884257605598137e-06,
+      "loss": 0.5745,
+      "step": 2481
+    },
+    {
+      "epoch": 0.4076119311066861,
+      "grad_norm": 0.31570900922366557,
+      "learning_rate": 8.884151902134578e-06,
+      "loss": 0.5746,
+      "step": 2482
+    },
+    {
+      "epoch": 0.40777615831502884,
+      "grad_norm": 0.35211170928524965,
+      "learning_rate": 8.884046151054815e-06,
+      "loss": 0.5819,
+      "step": 2483
+    },
+    {
+      "epoch": 0.4079403855233716,
+      "grad_norm": 0.3246151812151191,
+      "learning_rate": 8.883940352359998e-06,
+      "loss": 0.5808,
+      "step": 2484
+    },
+    {
+      "epoch": 0.40810461273171433,
+      "grad_norm": 0.29286391071193524,
+      "learning_rate": 8.883834506051277e-06,
+      "loss": 0.5618,
+      "step": 2485
+    },
+    {
+      "epoch": 0.4082688399400571,
+      "grad_norm": 0.38869425964352805,
+      "learning_rate": 8.883728612129799e-06,
+      "loss": 0.5944,
+      "step": 2486
+    },
+    {
+      "epoch": 0.40843306714839983,
+      "grad_norm": 0.3357922498523062,
+      "learning_rate": 8.883622670596715e-06,
+      "loss": 0.5485,
+      "step": 2487
+    },
+    {
+      "epoch": 0.4085972943567426,
+      "grad_norm": 0.3037126347902224,
+      "learning_rate": 8.883516681453177e-06,
+      "loss": 0.5801,
+      "step": 2488
+    },
+    {
+      "epoch": 0.4087615215650853,
+      "grad_norm": 0.32469023718462664,
+      "learning_rate": 8.883410644700335e-06,
+      "loss": 0.5674,
+      "step": 2489
+    },
+    {
+      "epoch": 0.408925748773428,
+      "grad_norm": 0.31429830880564624,
+      "learning_rate": 8.88330456033934e-06,
+      "loss": 0.5506,
+      "step": 2490
+    },
+    {
+      "epoch": 0.40908997598177077,
+      "grad_norm": 0.28926492543203663,
+      "learning_rate": 8.883198428371346e-06,
+      "loss": 0.584,
+      "step": 2491
+    },
+    {
+      "epoch": 0.4092542031901135,
+      "grad_norm": 0.33964002500396295,
+      "learning_rate": 8.883092248797503e-06,
+      "loss": 0.5619,
+      "step": 2492
+    },
+    {
+      "epoch": 0.40941843039845627,
+      "grad_norm": 0.30098148533924857,
+      "learning_rate": 8.882986021618967e-06,
+      "loss": 0.5504,
+      "step": 2493
+    },
+    {
+      "epoch": 0.409582657606799,
+      "grad_norm": 0.3575580059528442,
+      "learning_rate": 8.88287974683689e-06,
+      "loss": 0.5601,
+      "step": 2494
+    },
+    {
+      "epoch": 0.40974688481514177,
+      "grad_norm": 0.32069671607176764,
+      "learning_rate": 8.882773424452427e-06,
+      "loss": 0.5295,
+      "step": 2495
+    },
+    {
+      "epoch": 0.4099111120234845,
+      "grad_norm": 0.30232051278302446,
+      "learning_rate": 8.882667054466731e-06,
+      "loss": 0.5796,
+      "step": 2496
+    },
+    {
+      "epoch": 0.4100753392318272,
+      "grad_norm": 0.3708374336609579,
+      "learning_rate": 8.88256063688096e-06,
+      "loss": 0.5743,
+      "step": 2497
+    },
+    {
+      "epoch": 0.41023956644016996,
+      "grad_norm": 0.3032029622742079,
+      "learning_rate": 8.88245417169627e-06,
+      "loss": 0.5754,
+      "step": 2498
+    },
+    {
+      "epoch": 0.4104037936485127,
+      "grad_norm": 0.2867613437909591,
+      "learning_rate": 8.882347658913814e-06,
+      "loss": 0.5713,
+      "step": 2499
+    },
+    {
+      "epoch": 0.41056802085685545,
+      "grad_norm": 0.32926370873977717,
+      "learning_rate": 8.882241098534751e-06,
+      "loss": 0.5838,
+      "step": 2500
+    },
+    {
+      "epoch": 0.4107322480651982,
+      "grad_norm": 0.3428683835495346,
+      "learning_rate": 8.88213449056024e-06,
+      "loss": 0.5651,
+      "step": 2501
+    },
+    {
+      "epoch": 0.41089647527354095,
+      "grad_norm": 0.28973504149209034,
+      "learning_rate": 8.882027834991435e-06,
+      "loss": 0.5739,
+      "step": 2502
+    },
+    {
+      "epoch": 0.4110607024818837,
+      "grad_norm": 0.2797815273126128,
+      "learning_rate": 8.881921131829497e-06,
+      "loss": 0.5723,
+      "step": 2503
+    },
+    {
+      "epoch": 0.41122492969022645,
+      "grad_norm": 0.4145755184754829,
+      "learning_rate": 8.881814381075583e-06,
+      "loss": 0.5595,
+      "step": 2504
+    },
+    {
+      "epoch": 0.4113891568985692,
+      "grad_norm": 0.3331349868005999,
+      "learning_rate": 8.881707582730855e-06,
+      "loss": 0.5694,
+      "step": 2505
+    },
+    {
+      "epoch": 0.4115533841069119,
+      "grad_norm": 0.30051933126756675,
+      "learning_rate": 8.881600736796473e-06,
+      "loss": 0.5705,
+      "step": 2506
+    },
+    {
+      "epoch": 0.41171761131525464,
+      "grad_norm": 0.33450947684165244,
+      "learning_rate": 8.881493843273595e-06,
+      "loss": 0.5733,
+      "step": 2507
+    },
+    {
+      "epoch": 0.4118818385235974,
+      "grad_norm": 0.32963924885381335,
+      "learning_rate": 8.881386902163382e-06,
+      "loss": 0.5658,
+      "step": 2508
+    },
+    {
+      "epoch": 0.41204606573194014,
+      "grad_norm": 0.35930224711053116,
+      "learning_rate": 8.881279913466997e-06,
+      "loss": 0.5658,
+      "step": 2509
+    },
+    {
+      "epoch": 0.4122102929402829,
+      "grad_norm": 0.3704447240612607,
+      "learning_rate": 8.881172877185601e-06,
+      "loss": 0.5525,
+      "step": 2510
+    },
+    {
+      "epoch": 0.41237452014862563,
+      "grad_norm": 0.30949854959121004,
+      "learning_rate": 8.881065793320358e-06,
+      "loss": 0.5693,
+      "step": 2511
+    },
+    {
+      "epoch": 0.4125387473569684,
+      "grad_norm": 0.32021726342404155,
+      "learning_rate": 8.880958661872431e-06,
+      "loss": 0.5729,
+      "step": 2512
+    },
+    {
+      "epoch": 0.41270297456531113,
+      "grad_norm": 0.298710533844321,
+      "learning_rate": 8.88085148284298e-06,
+      "loss": 0.5395,
+      "step": 2513
+    },
+    {
+      "epoch": 0.4128672017736538,
+      "grad_norm": 0.3740569254074185,
+      "learning_rate": 8.880744256233175e-06,
+      "loss": 0.5472,
+      "step": 2514
+    },
+    {
+      "epoch": 0.4130314289819966,
+      "grad_norm": 0.4018513304426621,
+      "learning_rate": 8.880636982044176e-06,
+      "loss": 0.5624,
+      "step": 2515
+    },
+    {
+      "epoch": 0.4131956561903393,
+      "grad_norm": 0.40970239846974127,
+      "learning_rate": 8.88052966027715e-06,
+      "loss": 0.5564,
+      "step": 2516
+    },
+    {
+      "epoch": 0.41335988339868207,
+      "grad_norm": 0.3117304225668632,
+      "learning_rate": 8.88042229093326e-06,
+      "loss": 0.54,
+      "step": 2517
+    },
+    {
+      "epoch": 0.4135241106070248,
+      "grad_norm": 0.28437155849049806,
+      "learning_rate": 8.880314874013674e-06,
+      "loss": 0.5711,
+      "step": 2518
+    },
+    {
+      "epoch": 0.41368833781536757,
+      "grad_norm": 0.3332562664647129,
+      "learning_rate": 8.88020740951956e-06,
+      "loss": 0.5559,
+      "step": 2519
+    },
+    {
+      "epoch": 0.4138525650237103,
+      "grad_norm": 0.5483444520242049,
+      "learning_rate": 8.880099897452086e-06,
+      "loss": 0.5722,
+      "step": 2520
+    },
+    {
+      "epoch": 0.41401679223205307,
+      "grad_norm": 0.3011635778727716,
+      "learning_rate": 8.879992337812416e-06,
+      "loss": 0.5684,
+      "step": 2521
+    },
+    {
+      "epoch": 0.4141810194403958,
+      "grad_norm": 0.31450829980721595,
+      "learning_rate": 8.879884730601718e-06,
+      "loss": 0.5698,
+      "step": 2522
+    },
+    {
+      "epoch": 0.4143452466487385,
+      "grad_norm": 0.31001996453014075,
+      "learning_rate": 8.879777075821165e-06,
+      "loss": 0.5342,
+      "step": 2523
+    },
+    {
+      "epoch": 0.41450947385708126,
+      "grad_norm": 0.3139993732413725,
+      "learning_rate": 8.879669373471923e-06,
+      "loss": 0.5567,
+      "step": 2524
+    },
+    {
+      "epoch": 0.414673701065424,
+      "grad_norm": 0.3203420354656011,
+      "learning_rate": 8.879561623555163e-06,
+      "loss": 0.5657,
+      "step": 2525
+    },
+    {
+      "epoch": 0.41483792827376675,
+      "grad_norm": 0.3108136022021269,
+      "learning_rate": 8.879453826072055e-06,
+      "loss": 0.5815,
+      "step": 2526
+    },
+    {
+      "epoch": 0.4150021554821095,
+      "grad_norm": 0.30562850254753904,
+      "learning_rate": 8.879345981023769e-06,
+      "loss": 0.5765,
+      "step": 2527
+    },
+    {
+      "epoch": 0.41516638269045225,
+      "grad_norm": 0.3901832252063882,
+      "learning_rate": 8.879238088411476e-06,
+      "loss": 0.5606,
+      "step": 2528
+    },
+    {
+      "epoch": 0.415330609898795,
+      "grad_norm": 0.3017152351908569,
+      "learning_rate": 8.879130148236351e-06,
+      "loss": 0.6051,
+      "step": 2529
+    },
+    {
+      "epoch": 0.41549483710713775,
+      "grad_norm": 0.37648168466183224,
+      "learning_rate": 8.879022160499563e-06,
+      "loss": 0.557,
+      "step": 2530
+    },
+    {
+      "epoch": 0.41565906431548044,
+      "grad_norm": 0.31899551450864727,
+      "learning_rate": 8.878914125202287e-06,
+      "loss": 0.5608,
+      "step": 2531
+    },
+    {
+      "epoch": 0.4158232915238232,
+      "grad_norm": 0.2937937843776762,
+      "learning_rate": 8.878806042345693e-06,
+      "loss": 0.5608,
+      "step": 2532
+    },
+    {
+      "epoch": 0.41598751873216594,
+      "grad_norm": 0.3343905076548578,
+      "learning_rate": 8.878697911930959e-06,
+      "loss": 0.5666,
+      "step": 2533
+    },
+    {
+      "epoch": 0.4161517459405087,
+      "grad_norm": 0.37676843413898936,
+      "learning_rate": 8.878589733959256e-06,
+      "loss": 0.554,
+      "step": 2534
+    },
+    {
+      "epoch": 0.41631597314885144,
+      "grad_norm": 0.3158506997271894,
+      "learning_rate": 8.878481508431762e-06,
+      "loss": 0.56,
+      "step": 2535
+    },
+    {
+      "epoch": 0.4164802003571942,
+      "grad_norm": 0.31815058752542835,
+      "learning_rate": 8.87837323534965e-06,
+      "loss": 0.5791,
+      "step": 2536
+    },
+    {
+      "epoch": 0.41664442756553693,
+      "grad_norm": 0.4371060424111749,
+      "learning_rate": 8.878264914714098e-06,
+      "loss": 0.5815,
+      "step": 2537
+    },
+    {
+      "epoch": 0.4168086547738797,
+      "grad_norm": 0.2983880649857614,
+      "learning_rate": 8.878156546526282e-06,
+      "loss": 0.5604,
+      "step": 2538
+    },
+    {
+      "epoch": 0.41697288198222243,
+      "grad_norm": 0.34073001572633294,
+      "learning_rate": 8.878048130787376e-06,
+      "loss": 0.5813,
+      "step": 2539
+    },
+    {
+      "epoch": 0.4171371091905651,
+      "grad_norm": 0.3624505324000262,
+      "learning_rate": 8.877939667498561e-06,
+      "loss": 0.551,
+      "step": 2540
+    },
+    {
+      "epoch": 0.4173013363989079,
+      "grad_norm": 0.3038346714696093,
+      "learning_rate": 8.877831156661015e-06,
+      "loss": 0.5392,
+      "step": 2541
+    },
+    {
+      "epoch": 0.4174655636072506,
+      "grad_norm": 0.3447650018068754,
+      "learning_rate": 8.877722598275915e-06,
+      "loss": 0.5765,
+      "step": 2542
+    },
+    {
+      "epoch": 0.41762979081559337,
+      "grad_norm": 0.30392547582003404,
+      "learning_rate": 8.87761399234444e-06,
+      "loss": 0.5616,
+      "step": 2543
+    },
+    {
+      "epoch": 0.4177940180239361,
+      "grad_norm": 0.3887953041088847,
+      "learning_rate": 8.87750533886777e-06,
+      "loss": 0.5483,
+      "step": 2544
+    },
+    {
+      "epoch": 0.41795824523227887,
+      "grad_norm": 0.3289178164106548,
+      "learning_rate": 8.877396637847085e-06,
+      "loss": 0.5337,
+      "step": 2545
+    },
+    {
+      "epoch": 0.4181224724406216,
+      "grad_norm": 0.332748619678136,
+      "learning_rate": 8.877287889283566e-06,
+      "loss": 0.5682,
+      "step": 2546
+    },
+    {
+      "epoch": 0.41828669964896437,
+      "grad_norm": 0.3044178943644956,
+      "learning_rate": 8.877179093178394e-06,
+      "loss": 0.5424,
+      "step": 2547
+    },
+    {
+      "epoch": 0.41845092685730706,
+      "grad_norm": 0.3173245691997307,
+      "learning_rate": 8.87707024953275e-06,
+      "loss": 0.5936,
+      "step": 2548
+    },
+    {
+      "epoch": 0.4186151540656498,
+      "grad_norm": 0.265613088884184,
+      "learning_rate": 8.876961358347819e-06,
+      "loss": 0.561,
+      "step": 2549
+    },
+    {
+      "epoch": 0.41877938127399256,
+      "grad_norm": 0.3146875818714378,
+      "learning_rate": 8.876852419624777e-06,
+      "loss": 0.5787,
+      "step": 2550
+    },
+    {
+      "epoch": 0.4189436084823353,
+      "grad_norm": 0.37599575888562486,
+      "learning_rate": 8.876743433364814e-06,
+      "loss": 0.5568,
+      "step": 2551
+    },
+    {
+      "epoch": 0.41910783569067805,
+      "grad_norm": 0.3598518156246321,
+      "learning_rate": 8.876634399569111e-06,
+      "loss": 0.5856,
+      "step": 2552
+    },
+    {
+      "epoch": 0.4192720628990208,
+      "grad_norm": 0.30311498096956535,
+      "learning_rate": 8.876525318238852e-06,
+      "loss": 0.5876,
+      "step": 2553
+    },
+    {
+      "epoch": 0.41943629010736355,
+      "grad_norm": 0.3171555277490769,
+      "learning_rate": 8.876416189375222e-06,
+      "loss": 0.5866,
+      "step": 2554
+    },
+    {
+      "epoch": 0.4196005173157063,
+      "grad_norm": 0.3690366608054597,
+      "learning_rate": 8.876307012979409e-06,
+      "loss": 0.5691,
+      "step": 2555
+    },
+    {
+      "epoch": 0.41976474452404905,
+      "grad_norm": 0.2956348303710773,
+      "learning_rate": 8.876197789052593e-06,
+      "loss": 0.5774,
+      "step": 2556
+    },
+    {
+      "epoch": 0.41992897173239174,
+      "grad_norm": 0.30313111772520873,
+      "learning_rate": 8.876088517595964e-06,
+      "loss": 0.5626,
+      "step": 2557
+    },
+    {
+      "epoch": 0.4200931989407345,
+      "grad_norm": 0.29748543393772187,
+      "learning_rate": 8.875979198610709e-06,
+      "loss": 0.585,
+      "step": 2558
+    },
+    {
+      "epoch": 0.42025742614907724,
+      "grad_norm": 0.35983239485183327,
+      "learning_rate": 8.875869832098014e-06,
+      "loss": 0.5698,
+      "step": 2559
+    },
+    {
+      "epoch": 0.42042165335742,
+      "grad_norm": 0.2952903149801761,
+      "learning_rate": 8.875760418059067e-06,
+      "loss": 0.5602,
+      "step": 2560
+    },
+    {
+      "epoch": 0.42058588056576274,
+      "grad_norm": 0.2748700755222291,
+      "learning_rate": 8.875650956495058e-06,
+      "loss": 0.5571,
+      "step": 2561
+    },
+    {
+      "epoch": 0.4207501077741055,
+      "grad_norm": 0.30142460059542353,
+      "learning_rate": 8.875541447407174e-06,
+      "loss": 0.5815,
+      "step": 2562
+    },
+    {
+      "epoch": 0.42091433498244823,
+      "grad_norm": 0.30208405047227427,
+      "learning_rate": 8.875431890796603e-06,
+      "loss": 0.5731,
+      "step": 2563
+    },
+    {
+      "epoch": 0.421078562190791,
+      "grad_norm": 0.3118637667702491,
+      "learning_rate": 8.875322286664538e-06,
+      "loss": 0.5474,
+      "step": 2564
+    },
+    {
+      "epoch": 0.4212427893991337,
+      "grad_norm": 0.317291042191685,
+      "learning_rate": 8.87521263501217e-06,
+      "loss": 0.5446,
+      "step": 2565
+    },
+    {
+      "epoch": 0.4214070166074764,
+      "grad_norm": 0.35441933886963456,
+      "learning_rate": 8.875102935840687e-06,
+      "loss": 0.5516,
+      "step": 2566
+    },
+    {
+      "epoch": 0.4215712438158192,
+      "grad_norm": 0.3204356969177786,
+      "learning_rate": 8.874993189151281e-06,
+      "loss": 0.5661,
+      "step": 2567
+    },
+    {
+      "epoch": 0.4217354710241619,
+      "grad_norm": 0.30987334724965443,
+      "learning_rate": 8.874883394945145e-06,
+      "loss": 0.5629,
+      "step": 2568
+    },
+    {
+      "epoch": 0.42189969823250467,
+      "grad_norm": 0.32477041715743676,
+      "learning_rate": 8.87477355322347e-06,
+      "loss": 0.5627,
+      "step": 2569
+    },
+    {
+      "epoch": 0.4220639254408474,
+      "grad_norm": 0.2954548449250752,
+      "learning_rate": 8.874663663987452e-06,
+      "loss": 0.5503,
+      "step": 2570
+    },
+    {
+      "epoch": 0.42222815264919017,
+      "grad_norm": 0.3245662283540012,
+      "learning_rate": 8.874553727238281e-06,
+      "loss": 0.5747,
+      "step": 2571
+    },
+    {
+      "epoch": 0.4223923798575329,
+      "grad_norm": 0.3987350454378436,
+      "learning_rate": 8.874443742977154e-06,
+      "loss": 0.5561,
+      "step": 2572
+    },
+    {
+      "epoch": 0.4225566070658756,
+      "grad_norm": 0.3345985015745256,
+      "learning_rate": 8.874333711205264e-06,
+      "loss": 0.5555,
+      "step": 2573
+    },
+    {
+      "epoch": 0.42272083427421836,
+      "grad_norm": 0.28819563482433996,
+      "learning_rate": 8.874223631923804e-06,
+      "loss": 0.5392,
+      "step": 2574
+    },
+    {
+      "epoch": 0.4228850614825611,
+      "grad_norm": 0.3202474861777573,
+      "learning_rate": 8.874113505133974e-06,
+      "loss": 0.565,
+      "step": 2575
+    },
+    {
+      "epoch": 0.42304928869090386,
+      "grad_norm": 0.2790382154947942,
+      "learning_rate": 8.874003330836966e-06,
+      "loss": 0.5766,
+      "step": 2576
+    },
+    {
+      "epoch": 0.4232135158992466,
+      "grad_norm": 0.44535350715320376,
+      "learning_rate": 8.87389310903398e-06,
+      "loss": 0.5744,
+      "step": 2577
+    },
+    {
+      "epoch": 0.42337774310758935,
+      "grad_norm": 0.3184028856113947,
+      "learning_rate": 8.87378283972621e-06,
+      "loss": 0.5816,
+      "step": 2578
+    },
+    {
+      "epoch": 0.4235419703159321,
+      "grad_norm": 0.2922792339457821,
+      "learning_rate": 8.873672522914856e-06,
+      "loss": 0.5509,
+      "step": 2579
+    },
+    {
+      "epoch": 0.42370619752427485,
+      "grad_norm": 0.454554563345644,
+      "learning_rate": 8.873562158601116e-06,
+      "loss": 0.5594,
+      "step": 2580
+    },
+    {
+      "epoch": 0.4238704247326176,
+      "grad_norm": 0.34907028522108435,
+      "learning_rate": 8.873451746786186e-06,
+      "loss": 0.5655,
+      "step": 2581
+    },
+    {
+      "epoch": 0.4240346519409603,
+      "grad_norm": 0.3298719872719342,
+      "learning_rate": 8.873341287471269e-06,
+      "loss": 0.5629,
+      "step": 2582
+    },
+    {
+      "epoch": 0.42419887914930304,
+      "grad_norm": 0.3389124908725632,
+      "learning_rate": 8.873230780657562e-06,
+      "loss": 0.5878,
+      "step": 2583
+    },
+    {
+      "epoch": 0.4243631063576458,
+      "grad_norm": 0.37188737857676285,
+      "learning_rate": 8.873120226346266e-06,
+      "loss": 0.5831,
+      "step": 2584
+    },
+    {
+      "epoch": 0.42452733356598854,
+      "grad_norm": 0.34890335602420697,
+      "learning_rate": 8.873009624538582e-06,
+      "loss": 0.5722,
+      "step": 2585
+    },
+    {
+      "epoch": 0.4246915607743313,
+      "grad_norm": 0.328354898484246,
+      "learning_rate": 8.872898975235711e-06,
+      "loss": 0.5569,
+      "step": 2586
+    },
+    {
+      "epoch": 0.42485578798267404,
+      "grad_norm": 0.312327677555695,
+      "learning_rate": 8.872788278438854e-06,
+      "loss": 0.5674,
+      "step": 2587
+    },
+    {
+      "epoch": 0.4250200151910168,
+      "grad_norm": 0.32296344711606056,
+      "learning_rate": 8.872677534149215e-06,
+      "loss": 0.559,
+      "step": 2588
+    },
+    {
+      "epoch": 0.42518424239935954,
+      "grad_norm": 0.36315755635387575,
+      "learning_rate": 8.872566742367995e-06,
+      "loss": 0.5895,
+      "step": 2589
+    },
+    {
+      "epoch": 0.42534846960770223,
+      "grad_norm": 0.33097917263767107,
+      "learning_rate": 8.8724559030964e-06,
+      "loss": 0.578,
+      "step": 2590
+    },
+    {
+      "epoch": 0.425512696816045,
+      "grad_norm": 0.315094597471828,
+      "learning_rate": 8.87234501633563e-06,
+      "loss": 0.5504,
+      "step": 2591
+    },
+    {
+      "epoch": 0.4256769240243877,
+      "grad_norm": 0.45635596872249246,
+      "learning_rate": 8.87223408208689e-06,
+      "loss": 0.5482,
+      "step": 2592
+    },
+    {
+      "epoch": 0.4258411512327305,
+      "grad_norm": 0.4193262239236399,
+      "learning_rate": 8.872123100351389e-06,
+      "loss": 0.5688,
+      "step": 2593
+    },
+    {
+      "epoch": 0.4260053784410732,
+      "grad_norm": 0.4273899125136232,
+      "learning_rate": 8.87201207113033e-06,
+      "loss": 0.5796,
+      "step": 2594
+    },
+    {
+      "epoch": 0.426169605649416,
+      "grad_norm": 0.3917537813124731,
+      "learning_rate": 8.871900994424917e-06,
+      "loss": 0.5666,
+      "step": 2595
+    },
+    {
+      "epoch": 0.4263338328577587,
+      "grad_norm": 0.3335513432900561,
+      "learning_rate": 8.871789870236358e-06,
+      "loss": 0.5536,
+      "step": 2596
+    },
+    {
+      "epoch": 0.42649806006610147,
+      "grad_norm": 0.3463487402113218,
+      "learning_rate": 8.87167869856586e-06,
+      "loss": 0.5663,
+      "step": 2597
+    },
+    {
+      "epoch": 0.4266622872744442,
+      "grad_norm": 0.3872490703111275,
+      "learning_rate": 8.87156747941463e-06,
+      "loss": 0.5594,
+      "step": 2598
+    },
+    {
+      "epoch": 0.4268265144827869,
+      "grad_norm": 0.31074493917689083,
+      "learning_rate": 8.871456212783874e-06,
+      "loss": 0.5922,
+      "step": 2599
+    },
+    {
+      "epoch": 0.42699074169112966,
+      "grad_norm": 0.31115722805904655,
+      "learning_rate": 8.871344898674806e-06,
+      "loss": 0.5681,
+      "step": 2600
+    },
+    {
+      "epoch": 0.4271549688994724,
+      "grad_norm": 0.30377485723430814,
+      "learning_rate": 8.87123353708863e-06,
+      "loss": 0.5751,
+      "step": 2601
+    },
+    {
+      "epoch": 0.42731919610781516,
+      "grad_norm": 0.3443330184625631,
+      "learning_rate": 8.871122128026559e-06,
+      "loss": 0.5477,
+      "step": 2602
+    },
+    {
+      "epoch": 0.4274834233161579,
+      "grad_norm": 0.3160677382382284,
+      "learning_rate": 8.871010671489798e-06,
+      "loss": 0.551,
+      "step": 2603
+    },
+    {
+      "epoch": 0.42764765052450066,
+      "grad_norm": 0.3089935106670285,
+      "learning_rate": 8.870899167479561e-06,
+      "loss": 0.5573,
+      "step": 2604
+    },
+    {
+      "epoch": 0.4278118777328434,
+      "grad_norm": 0.29792130232279923,
+      "learning_rate": 8.87078761599706e-06,
+      "loss": 0.5449,
+      "step": 2605
+    },
+    {
+      "epoch": 0.42797610494118615,
+      "grad_norm": 0.3416750043491599,
+      "learning_rate": 8.870676017043506e-06,
+      "loss": 0.5558,
+      "step": 2606
+    },
+    {
+      "epoch": 0.42814033214952885,
+      "grad_norm": 0.3137377292747842,
+      "learning_rate": 8.870564370620109e-06,
+      "loss": 0.5785,
+      "step": 2607
+    },
+    {
+      "epoch": 0.4283045593578716,
+      "grad_norm": 0.337439787190629,
+      "learning_rate": 8.870452676728082e-06,
+      "loss": 0.5636,
+      "step": 2608
+    },
+    {
+      "epoch": 0.42846878656621434,
+      "grad_norm": 0.3449765466085809,
+      "learning_rate": 8.870340935368641e-06,
+      "loss": 0.5613,
+      "step": 2609
+    },
+    {
+      "epoch": 0.4286330137745571,
+      "grad_norm": 0.3239998293778881,
+      "learning_rate": 8.870229146542996e-06,
+      "loss": 0.5427,
+      "step": 2610
+    },
+    {
+      "epoch": 0.42879724098289984,
+      "grad_norm": 0.3629879882971418,
+      "learning_rate": 8.870117310252364e-06,
+      "loss": 0.5485,
+      "step": 2611
+    },
+    {
+      "epoch": 0.4289614681912426,
+      "grad_norm": 0.322609355069675,
+      "learning_rate": 8.870005426497957e-06,
+      "loss": 0.5946,
+      "step": 2612
+    },
+    {
+      "epoch": 0.42912569539958534,
+      "grad_norm": 0.2760721542583468,
+      "learning_rate": 8.869893495280993e-06,
+      "loss": 0.545,
+      "step": 2613
+    },
+    {
+      "epoch": 0.4292899226079281,
+      "grad_norm": 0.377724350549196,
+      "learning_rate": 8.869781516602686e-06,
+      "loss": 0.5727,
+      "step": 2614
+    },
+    {
+      "epoch": 0.42945414981627084,
+      "grad_norm": 0.3233053561286611,
+      "learning_rate": 8.869669490464253e-06,
+      "loss": 0.5629,
+      "step": 2615
+    },
+    {
+      "epoch": 0.42961837702461353,
+      "grad_norm": 0.35088644192144613,
+      "learning_rate": 8.869557416866907e-06,
+      "loss": 0.561,
+      "step": 2616
+    },
+    {
+      "epoch": 0.4297826042329563,
+      "grad_norm": 0.3447050844031401,
+      "learning_rate": 8.86944529581187e-06,
+      "loss": 0.572,
+      "step": 2617
+    },
+    {
+      "epoch": 0.429946831441299,
+      "grad_norm": 0.34469094004572337,
+      "learning_rate": 8.86933312730036e-06,
+      "loss": 0.5506,
+      "step": 2618
+    },
+    {
+      "epoch": 0.4301110586496418,
+      "grad_norm": 0.327264522482107,
+      "learning_rate": 8.869220911333591e-06,
+      "loss": 0.5374,
+      "step": 2619
+    },
+    {
+      "epoch": 0.4302752858579845,
+      "grad_norm": 0.32108065910291667,
+      "learning_rate": 8.869108647912786e-06,
+      "loss": 0.5562,
+      "step": 2620
+    },
+    {
+      "epoch": 0.4304395130663273,
+      "grad_norm": 0.3075435330753905,
+      "learning_rate": 8.868996337039163e-06,
+      "loss": 0.5655,
+      "step": 2621
+    },
+    {
+      "epoch": 0.43060374027467,
+      "grad_norm": 0.34795979712051084,
+      "learning_rate": 8.868883978713939e-06,
+      "loss": 0.5661,
+      "step": 2622
+    },
+    {
+      "epoch": 0.43076796748301277,
+      "grad_norm": 0.36050340443618495,
+      "learning_rate": 8.868771572938337e-06,
+      "loss": 0.5624,
+      "step": 2623
+    },
+    {
+      "epoch": 0.43093219469135546,
+      "grad_norm": 0.3918416774697456,
+      "learning_rate": 8.868659119713579e-06,
+      "loss": 0.5749,
+      "step": 2624
+    },
+    {
+      "epoch": 0.4310964218996982,
+      "grad_norm": 0.3457358807465997,
+      "learning_rate": 8.868546619040884e-06,
+      "loss": 0.5643,
+      "step": 2625
+    },
+    {
+      "epoch": 0.43126064910804096,
+      "grad_norm": 0.39287093615546537,
+      "learning_rate": 8.868434070921473e-06,
+      "loss": 0.5588,
+      "step": 2626
+    },
+    {
+      "epoch": 0.4314248763163837,
+      "grad_norm": 0.33227836409954387,
+      "learning_rate": 8.868321475356572e-06,
+      "loss": 0.531,
+      "step": 2627
+    },
+    {
+      "epoch": 0.43158910352472646,
+      "grad_norm": 0.30996877615907675,
+      "learning_rate": 8.868208832347401e-06,
+      "loss": 0.54,
+      "step": 2628
+    },
+    {
+      "epoch": 0.4317533307330692,
+      "grad_norm": 0.29705507963891703,
+      "learning_rate": 8.868096141895186e-06,
+      "loss": 0.5647,
+      "step": 2629
+    },
+    {
+      "epoch": 0.43191755794141196,
+      "grad_norm": 0.3145590224011055,
+      "learning_rate": 8.867983404001147e-06,
+      "loss": 0.575,
+      "step": 2630
+    },
+    {
+      "epoch": 0.4320817851497547,
+      "grad_norm": 0.2984999228772706,
+      "learning_rate": 8.867870618666512e-06,
+      "loss": 0.564,
+      "step": 2631
+    },
+    {
+      "epoch": 0.43224601235809745,
+      "grad_norm": 0.3948551709555983,
+      "learning_rate": 8.867757785892506e-06,
+      "loss": 0.56,
+      "step": 2632
+    },
+    {
+      "epoch": 0.43241023956644015,
+      "grad_norm": 0.313192693394057,
+      "learning_rate": 8.86764490568035e-06,
+      "loss": 0.5537,
+      "step": 2633
+    },
+    {
+      "epoch": 0.4325744667747829,
+      "grad_norm": 0.4653518718560589,
+      "learning_rate": 8.867531978031276e-06,
+      "loss": 0.5714,
+      "step": 2634
+    },
+    {
+      "epoch": 0.43273869398312564,
+      "grad_norm": 0.3742502734817775,
+      "learning_rate": 8.867419002946505e-06,
+      "loss": 0.5631,
+      "step": 2635
+    },
+    {
+      "epoch": 0.4329029211914684,
+      "grad_norm": 0.3232307295194687,
+      "learning_rate": 8.867305980427268e-06,
+      "loss": 0.5529,
+      "step": 2636
+    },
+    {
+      "epoch": 0.43306714839981114,
+      "grad_norm": 0.2807991745055016,
+      "learning_rate": 8.867192910474792e-06,
+      "loss": 0.5406,
+      "step": 2637
+    },
+    {
+      "epoch": 0.4332313756081539,
+      "grad_norm": 0.297296622670886,
+      "learning_rate": 8.867079793090304e-06,
+      "loss": 0.5506,
+      "step": 2638
+    },
+    {
+      "epoch": 0.43339560281649664,
+      "grad_norm": 0.3542755078188722,
+      "learning_rate": 8.866966628275032e-06,
+      "loss": 0.5684,
+      "step": 2639
+    },
+    {
+      "epoch": 0.4335598300248394,
+      "grad_norm": 0.2744566073730852,
+      "learning_rate": 8.866853416030206e-06,
+      "loss": 0.5396,
+      "step": 2640
+    },
+    {
+      "epoch": 0.4337240572331821,
+      "grad_norm": 0.2959244532919851,
+      "learning_rate": 8.866740156357056e-06,
+      "loss": 0.5702,
+      "step": 2641
+    },
+    {
+      "epoch": 0.43388828444152483,
+      "grad_norm": 0.3268923575661857,
+      "learning_rate": 8.866626849256812e-06,
+      "loss": 0.5568,
+      "step": 2642
+    },
+    {
+      "epoch": 0.4340525116498676,
+      "grad_norm": 0.30432906161931816,
+      "learning_rate": 8.866513494730702e-06,
+      "loss": 0.54,
+      "step": 2643
+    },
+    {
+      "epoch": 0.4342167388582103,
+      "grad_norm": 0.33743946703434036,
+      "learning_rate": 8.866400092779963e-06,
+      "loss": 0.5553,
+      "step": 2644
+    },
+    {
+      "epoch": 0.4343809660665531,
+      "grad_norm": 0.2873723233387204,
+      "learning_rate": 8.866286643405819e-06,
+      "loss": 0.5603,
+      "step": 2645
+    },
+    {
+      "epoch": 0.4345451932748958,
+      "grad_norm": 0.4160177869919669,
+      "learning_rate": 8.866173146609509e-06,
+      "loss": 0.5581,
+      "step": 2646
+    },
+    {
+      "epoch": 0.4347094204832386,
+      "grad_norm": 0.3620215325482711,
+      "learning_rate": 8.866059602392262e-06,
+      "loss": 0.5538,
+      "step": 2647
+    },
+    {
+      "epoch": 0.4348736476915813,
+      "grad_norm": 0.34398345578712186,
+      "learning_rate": 8.865946010755313e-06,
+      "loss": 0.5617,
+      "step": 2648
+    },
+    {
+      "epoch": 0.43503787489992407,
+      "grad_norm": 0.4050944944383113,
+      "learning_rate": 8.865832371699894e-06,
+      "loss": 0.5628,
+      "step": 2649
+    },
+    {
+      "epoch": 0.43520210210826676,
+      "grad_norm": 0.34906966357449254,
+      "learning_rate": 8.86571868522724e-06,
+      "loss": 0.5485,
+      "step": 2650
+    },
+    {
+      "epoch": 0.4353663293166095,
+      "grad_norm": 0.2968598963085859,
+      "learning_rate": 8.865604951338585e-06,
+      "loss": 0.5592,
+      "step": 2651
+    },
+    {
+      "epoch": 0.43553055652495226,
+      "grad_norm": 0.2987516697302997,
+      "learning_rate": 8.865491170035166e-06,
+      "loss": 0.5615,
+      "step": 2652
+    },
+    {
+      "epoch": 0.435694783733295,
+      "grad_norm": 0.3500301965790752,
+      "learning_rate": 8.865377341318218e-06,
+      "loss": 0.5892,
+      "step": 2653
+    },
+    {
+      "epoch": 0.43585901094163776,
+      "grad_norm": 0.3929052238635463,
+      "learning_rate": 8.865263465188977e-06,
+      "loss": 0.5495,
+      "step": 2654
+    },
+    {
+      "epoch": 0.4360232381499805,
+      "grad_norm": 0.3113448223634506,
+      "learning_rate": 8.865149541648679e-06,
+      "loss": 0.5663,
+      "step": 2655
+    },
+    {
+      "epoch": 0.43618746535832326,
+      "grad_norm": 0.31364147135723414,
+      "learning_rate": 8.865035570698563e-06,
+      "loss": 0.5871,
+      "step": 2656
+    },
+    {
+      "epoch": 0.436351692566666,
+      "grad_norm": 0.31789727183776045,
+      "learning_rate": 8.864921552339866e-06,
+      "loss": 0.5898,
+      "step": 2657
+    },
+    {
+      "epoch": 0.4365159197750087,
+      "grad_norm": 0.2991212298735117,
+      "learning_rate": 8.864807486573827e-06,
+      "loss": 0.5535,
+      "step": 2658
+    },
+    {
+      "epoch": 0.43668014698335145,
+      "grad_norm": 0.31462832495510007,
+      "learning_rate": 8.864693373401684e-06,
+      "loss": 0.5674,
+      "step": 2659
+    },
+    {
+      "epoch": 0.4368443741916942,
+      "grad_norm": 0.3690385907683935,
+      "learning_rate": 8.864579212824676e-06,
+      "loss": 0.5641,
+      "step": 2660
+    },
+    {
+      "epoch": 0.43700860140003694,
+      "grad_norm": 0.30516447961861315,
+      "learning_rate": 8.864465004844045e-06,
+      "loss": 0.5619,
+      "step": 2661
+    },
+    {
+      "epoch": 0.4371728286083797,
+      "grad_norm": 0.32974499017869374,
+      "learning_rate": 8.864350749461027e-06,
+      "loss": 0.5534,
+      "step": 2662
+    },
+    {
+      "epoch": 0.43733705581672244,
+      "grad_norm": 0.5630108302746223,
+      "learning_rate": 8.864236446676871e-06,
+      "loss": 0.5881,
+      "step": 2663
+    },
+    {
+      "epoch": 0.4375012830250652,
+      "grad_norm": 0.33965305823725117,
+      "learning_rate": 8.864122096492808e-06,
+      "loss": 0.5615,
+      "step": 2664
+    },
+    {
+      "epoch": 0.43766551023340794,
+      "grad_norm": 0.2980421255968811,
+      "learning_rate": 8.86400769891009e-06,
+      "loss": 0.5415,
+      "step": 2665
+    },
+    {
+      "epoch": 0.4378297374417507,
+      "grad_norm": 0.33674308958283006,
+      "learning_rate": 8.863893253929951e-06,
+      "loss": 0.5721,
+      "step": 2666
+    },
+    {
+      "epoch": 0.4379939646500934,
+      "grad_norm": 0.37123811762089753,
+      "learning_rate": 8.86377876155364e-06,
+      "loss": 0.5859,
+      "step": 2667
+    },
+    {
+      "epoch": 0.43815819185843613,
+      "grad_norm": 0.3360935049247868,
+      "learning_rate": 8.863664221782397e-06,
+      "loss": 0.5466,
+      "step": 2668
+    },
+    {
+      "epoch": 0.4383224190667789,
+      "grad_norm": 0.3194927447043426,
+      "learning_rate": 8.863549634617467e-06,
+      "loss": 0.5725,
+      "step": 2669
+    },
+    {
+      "epoch": 0.4384866462751216,
+      "grad_norm": 0.31690358845503236,
+      "learning_rate": 8.863435000060097e-06,
+      "loss": 0.5396,
+      "step": 2670
+    },
+    {
+      "epoch": 0.4386508734834644,
+      "grad_norm": 0.3191798894096308,
+      "learning_rate": 8.863320318111528e-06,
+      "loss": 0.5577,
+      "step": 2671
+    },
+    {
+      "epoch": 0.4388151006918071,
+      "grad_norm": 0.4654210766977454,
+      "learning_rate": 8.863205588773007e-06,
+      "loss": 0.5724,
+      "step": 2672
+    },
+    {
+      "epoch": 0.4389793279001499,
+      "grad_norm": 0.3422961348923824,
+      "learning_rate": 8.863090812045783e-06,
+      "loss": 0.5847,
+      "step": 2673
+    },
+    {
+      "epoch": 0.4391435551084926,
+      "grad_norm": 0.3121440943186785,
+      "learning_rate": 8.862975987931097e-06,
+      "loss": 0.5774,
+      "step": 2674
+    },
+    {
+      "epoch": 0.4393077823168353,
+      "grad_norm": 0.3614058733108181,
+      "learning_rate": 8.8628611164302e-06,
+      "loss": 0.5538,
+      "step": 2675
+    },
+    {
+      "epoch": 0.43947200952517806,
+      "grad_norm": 0.3305418031431199,
+      "learning_rate": 8.862746197544341e-06,
+      "loss": 0.5522,
+      "step": 2676
+    },
+    {
+      "epoch": 0.4396362367335208,
+      "grad_norm": 0.31728827829254425,
+      "learning_rate": 8.862631231274764e-06,
+      "loss": 0.5637,
+      "step": 2677
+    },
+    {
+      "epoch": 0.43980046394186356,
+      "grad_norm": 0.2984532306242673,
+      "learning_rate": 8.862516217622721e-06,
+      "loss": 0.5741,
+      "step": 2678
+    },
+    {
+      "epoch": 0.4399646911502063,
+      "grad_norm": 0.30044448216886954,
+      "learning_rate": 8.862401156589457e-06,
+      "loss": 0.5649,
+      "step": 2679
+    },
+    {
+      "epoch": 0.44012891835854906,
+      "grad_norm": 0.33355290409354743,
+      "learning_rate": 8.862286048176227e-06,
+      "loss": 0.5435,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4402931455668918,
+      "grad_norm": 0.305282881254181,
+      "learning_rate": 8.862170892384278e-06,
+      "loss": 0.5473,
+      "step": 2681
+    },
+    {
+      "epoch": 0.44045737277523456,
+      "grad_norm": 0.31605141038354306,
+      "learning_rate": 8.86205568921486e-06,
+      "loss": 0.5179,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4406215999835773,
+      "grad_norm": 0.36881933440627185,
+      "learning_rate": 8.861940438669227e-06,
+      "loss": 0.591,
+      "step": 2683
+    },
+    {
+      "epoch": 0.44078582719192,
+      "grad_norm": 0.3375529244437988,
+      "learning_rate": 8.86182514074863e-06,
+      "loss": 0.5791,
+      "step": 2684
+    },
+    {
+      "epoch": 0.44095005440026275,
+      "grad_norm": 0.30314726667864034,
+      "learning_rate": 8.861709795454319e-06,
+      "loss": 0.5557,
+      "step": 2685
+    },
+    {
+      "epoch": 0.4411142816086055,
+      "grad_norm": 0.3364679818060467,
+      "learning_rate": 8.86159440278755e-06,
+      "loss": 0.5733,
+      "step": 2686
+    },
+    {
+      "epoch": 0.44127850881694825,
+      "grad_norm": 0.39220116051738163,
+      "learning_rate": 8.861478962749572e-06,
+      "loss": 0.5579,
+      "step": 2687
+    },
+    {
+      "epoch": 0.441442736025291,
+      "grad_norm": 0.34252772123554887,
+      "learning_rate": 8.861363475341642e-06,
+      "loss": 0.5544,
+      "step": 2688
+    },
+    {
+      "epoch": 0.44160696323363374,
+      "grad_norm": 0.33108150991015783,
+      "learning_rate": 8.861247940565015e-06,
+      "loss": 0.5761,
+      "step": 2689
+    },
+    {
+      "epoch": 0.4417711904419765,
+      "grad_norm": 0.32834599672177206,
+      "learning_rate": 8.861132358420943e-06,
+      "loss": 0.5662,
+      "step": 2690
+    },
+    {
+      "epoch": 0.44193541765031924,
+      "grad_norm": 0.32244811454221667,
+      "learning_rate": 8.861016728910683e-06,
+      "loss": 0.5671,
+      "step": 2691
+    },
+    {
+      "epoch": 0.44209964485866193,
+      "grad_norm": 0.3597249482708189,
+      "learning_rate": 8.860901052035492e-06,
+      "loss": 0.5613,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4422638720670047,
+      "grad_norm": 0.37744052906600656,
+      "learning_rate": 8.860785327796625e-06,
+      "loss": 0.553,
+      "step": 2693
+    },
+    {
+      "epoch": 0.44242809927534743,
+      "grad_norm": 0.3052348493632045,
+      "learning_rate": 8.860669556195338e-06,
+      "loss": 0.5457,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4425923264836902,
+      "grad_norm": 0.30573461825823206,
+      "learning_rate": 8.860553737232889e-06,
+      "loss": 0.5613,
+      "step": 2695
+    },
+    {
+      "epoch": 0.44275655369203293,
+      "grad_norm": 0.3007325457721176,
+      "learning_rate": 8.860437870910537e-06,
+      "loss": 0.5534,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4429207809003757,
+      "grad_norm": 0.2869011044903269,
+      "learning_rate": 8.86032195722954e-06,
+      "loss": 0.5542,
+      "step": 2697
+    },
+    {
+      "epoch": 0.4430850081087184,
+      "grad_norm": 0.5107075695902422,
+      "learning_rate": 8.860205996191155e-06,
+      "loss": 0.5081,
+      "step": 2698
+    },
+    {
+      "epoch": 0.4432492353170612,
+      "grad_norm": 0.33253967123335504,
+      "learning_rate": 8.860089987796643e-06,
+      "loss": 0.5554,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4434134625254039,
+      "grad_norm": 0.42343864420968474,
+      "learning_rate": 8.859973932047267e-06,
+      "loss": 0.5545,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4435776897337466,
+      "grad_norm": 0.30685102611881854,
+      "learning_rate": 8.85985782894428e-06,
+      "loss": 0.5412,
+      "step": 2701
+    },
+    {
+      "epoch": 0.44374191694208936,
+      "grad_norm": 0.3169560527008105,
+      "learning_rate": 8.85974167848895e-06,
+      "loss": 0.5784,
+      "step": 2702
+    },
+    {
+      "epoch": 0.4439061441504321,
+      "grad_norm": 0.3084651380614442,
+      "learning_rate": 8.859625480682535e-06,
+      "loss": 0.5317,
+      "step": 2703
+    },
+    {
+      "epoch": 0.44407037135877486,
+      "grad_norm": 0.441293933174961,
+      "learning_rate": 8.8595092355263e-06,
+      "loss": 0.543,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4442345985671176,
+      "grad_norm": 0.46942179938637835,
+      "learning_rate": 8.859392943021504e-06,
+      "loss": 0.5627,
+      "step": 2705
+    },
+    {
+      "epoch": 0.44439882577546036,
+      "grad_norm": 0.293267545918469,
+      "learning_rate": 8.859276603169412e-06,
+      "loss": 0.5479,
+      "step": 2706
+    },
+    {
+      "epoch": 0.4445630529838031,
+      "grad_norm": 0.45510996453392444,
+      "learning_rate": 8.859160215971286e-06,
+      "loss": 0.5778,
+      "step": 2707
+    },
+    {
+      "epoch": 0.44472728019214586,
+      "grad_norm": 0.32926696538846945,
+      "learning_rate": 8.859043781428393e-06,
+      "loss": 0.5417,
+      "step": 2708
+    },
+    {
+      "epoch": 0.44489150740048855,
+      "grad_norm": 0.33787931129775456,
+      "learning_rate": 8.858927299541995e-06,
+      "loss": 0.572,
+      "step": 2709
+    },
+    {
+      "epoch": 0.4450557346088313,
+      "grad_norm": 0.40369557531155686,
+      "learning_rate": 8.858810770313358e-06,
+      "loss": 0.5943,
+      "step": 2710
+    },
+    {
+      "epoch": 0.44521996181717405,
+      "grad_norm": 0.33877275852459676,
+      "learning_rate": 8.858694193743747e-06,
+      "loss": 0.5494,
+      "step": 2711
+    },
+    {
+      "epoch": 0.4453841890255168,
+      "grad_norm": 0.3597032289417445,
+      "learning_rate": 8.85857756983443e-06,
+      "loss": 0.5811,
+      "step": 2712
+    },
+    {
+      "epoch": 0.44554841623385955,
+      "grad_norm": 0.34694433353018983,
+      "learning_rate": 8.858460898586671e-06,
+      "loss": 0.5472,
+      "step": 2713
+    },
+    {
+      "epoch": 0.4457126434422023,
+      "grad_norm": 0.3595141962113896,
+      "learning_rate": 8.858344180001738e-06,
+      "loss": 0.5738,
+      "step": 2714
+    },
+    {
+      "epoch": 0.44587687065054504,
+      "grad_norm": 0.31463064675320646,
+      "learning_rate": 8.8582274140809e-06,
+      "loss": 0.5648,
+      "step": 2715
+    },
+    {
+      "epoch": 0.4460410978588878,
+      "grad_norm": 0.34440317093930706,
+      "learning_rate": 8.858110600825425e-06,
+      "loss": 0.5654,
+      "step": 2716
+    },
+    {
+      "epoch": 0.44620532506723054,
+      "grad_norm": 0.31860356038557297,
+      "learning_rate": 8.857993740236582e-06,
+      "loss": 0.5553,
+      "step": 2717
+    },
+    {
+      "epoch": 0.44636955227557323,
+      "grad_norm": 0.3105546506846929,
+      "learning_rate": 8.857876832315636e-06,
+      "loss": 0.5647,
+      "step": 2718
+    },
+    {
+      "epoch": 0.446533779483916,
+      "grad_norm": 0.33831142671969106,
+      "learning_rate": 8.857759877063863e-06,
+      "loss": 0.5479,
+      "step": 2719
+    },
+    {
+      "epoch": 0.44669800669225873,
+      "grad_norm": 0.3358614278897194,
+      "learning_rate": 8.857642874482528e-06,
+      "loss": 0.5792,
+      "step": 2720
+    },
+    {
+      "epoch": 0.4468622339006015,
+      "grad_norm": 1.239182763799466,
+      "learning_rate": 8.857525824572906e-06,
+      "loss": 0.5551,
+      "step": 2721
+    },
+    {
+      "epoch": 0.44702646110894423,
+      "grad_norm": 0.41414189926956013,
+      "learning_rate": 8.857408727336265e-06,
+      "loss": 0.5596,
+      "step": 2722
+    },
+    {
+      "epoch": 0.447190688317287,
+      "grad_norm": 0.3364069633037803,
+      "learning_rate": 8.857291582773878e-06,
+      "loss": 0.5507,
+      "step": 2723
+    },
+    {
+      "epoch": 0.4473549155256297,
+      "grad_norm": 0.32342124716147996,
+      "learning_rate": 8.857174390887019e-06,
+      "loss": 0.5608,
+      "step": 2724
+    },
+    {
+      "epoch": 0.4475191427339725,
+      "grad_norm": 0.34676827689048356,
+      "learning_rate": 8.85705715167696e-06,
+      "loss": 0.5582,
+      "step": 2725
+    },
+    {
+      "epoch": 0.44768336994231517,
+      "grad_norm": 0.35153982996698413,
+      "learning_rate": 8.856939865144971e-06,
+      "loss": 0.5515,
+      "step": 2726
+    },
+    {
+      "epoch": 0.4478475971506579,
+      "grad_norm": 0.2938845084834155,
+      "learning_rate": 8.856822531292329e-06,
+      "loss": 0.5429,
+      "step": 2727
+    },
+    {
+      "epoch": 0.44801182435900067,
+      "grad_norm": 0.2910178771667892,
+      "learning_rate": 8.856705150120308e-06,
+      "loss": 0.5645,
+      "step": 2728
+    },
+    {
+      "epoch": 0.4481760515673434,
+      "grad_norm": 0.2918827196955868,
+      "learning_rate": 8.856587721630182e-06,
+      "loss": 0.5451,
+      "step": 2729
+    },
+    {
+      "epoch": 0.44834027877568616,
+      "grad_norm": 0.3394189888352664,
+      "learning_rate": 8.856470245823227e-06,
+      "loss": 0.5605,
+      "step": 2730
+    },
+    {
+      "epoch": 0.4485045059840289,
+      "grad_norm": 0.33940159120607644,
+      "learning_rate": 8.85635272270072e-06,
+      "loss": 0.5491,
+      "step": 2731
+    },
+    {
+      "epoch": 0.44866873319237166,
+      "grad_norm": 0.36680822587198164,
+      "learning_rate": 8.856235152263938e-06,
+      "loss": 0.5681,
+      "step": 2732
+    },
+    {
+      "epoch": 0.4488329604007144,
+      "grad_norm": 0.3334810084593633,
+      "learning_rate": 8.856117534514154e-06,
+      "loss": 0.5606,
+      "step": 2733
+    },
+    {
+      "epoch": 0.44899718760905716,
+      "grad_norm": 0.29740348475239803,
+      "learning_rate": 8.855999869452647e-06,
+      "loss": 0.5647,
+      "step": 2734
+    },
+    {
+      "epoch": 0.44916141481739985,
+      "grad_norm": 0.32512754587660186,
+      "learning_rate": 8.855882157080697e-06,
+      "loss": 0.5652,
+      "step": 2735
+    },
+    {
+      "epoch": 0.4493256420257426,
+      "grad_norm": 0.3561551765070881,
+      "learning_rate": 8.85576439739958e-06,
+      "loss": 0.5773,
+      "step": 2736
+    },
+    {
+      "epoch": 0.44948986923408535,
+      "grad_norm": 0.3304245407015048,
+      "learning_rate": 8.855646590410578e-06,
+      "loss": 0.5666,
+      "step": 2737
+    },
+    {
+      "epoch": 0.4496540964424281,
+      "grad_norm": 0.3332153957003678,
+      "learning_rate": 8.855528736114969e-06,
+      "loss": 0.5686,
+      "step": 2738
+    },
+    {
+      "epoch": 0.44981832365077085,
+      "grad_norm": 0.3014454733597397,
+      "learning_rate": 8.85541083451403e-06,
+      "loss": 0.5547,
+      "step": 2739
+    },
+    {
+      "epoch": 0.4499825508591136,
+      "grad_norm": 0.3068041232696349,
+      "learning_rate": 8.855292885609045e-06,
+      "loss": 0.5625,
+      "step": 2740
+    },
+    {
+      "epoch": 0.45014677806745634,
+      "grad_norm": 0.3670493909898495,
+      "learning_rate": 8.855174889401295e-06,
+      "loss": 0.5724,
+      "step": 2741
+    },
+    {
+      "epoch": 0.4503110052757991,
+      "grad_norm": 0.3600172039749998,
+      "learning_rate": 8.85505684589206e-06,
+      "loss": 0.552,
+      "step": 2742
+    },
+    {
+      "epoch": 0.4504752324841418,
+      "grad_norm": 0.34510224190707856,
+      "learning_rate": 8.854938755082624e-06,
+      "loss": 0.5542,
+      "step": 2743
+    },
+    {
+      "epoch": 0.45063945969248453,
+      "grad_norm": 0.3354210595356994,
+      "learning_rate": 8.854820616974267e-06,
+      "loss": 0.5706,
+      "step": 2744
+    },
+    {
+      "epoch": 0.4508036869008273,
+      "grad_norm": 0.27484837623978225,
+      "learning_rate": 8.854702431568276e-06,
+      "loss": 0.5536,
+      "step": 2745
+    },
+    {
+      "epoch": 0.45096791410917003,
+      "grad_norm": 0.2772875379629504,
+      "learning_rate": 8.85458419886593e-06,
+      "loss": 0.5519,
+      "step": 2746
+    },
+    {
+      "epoch": 0.4511321413175128,
+      "grad_norm": 0.36745553176616597,
+      "learning_rate": 8.854465918868516e-06,
+      "loss": 0.5691,
+      "step": 2747
+    },
+    {
+      "epoch": 0.45129636852585553,
+      "grad_norm": 0.9592406829660379,
+      "learning_rate": 8.854347591577319e-06,
+      "loss": 0.5453,
+      "step": 2748
+    },
+    {
+      "epoch": 0.4514605957341983,
+      "grad_norm": 0.37135239233609174,
+      "learning_rate": 8.85422921699362e-06,
+      "loss": 0.5541,
+      "step": 2749
+    },
+    {
+      "epoch": 0.451624822942541,
+      "grad_norm": 0.3242742529723335,
+      "learning_rate": 8.85411079511871e-06,
+      "loss": 0.5644,
+      "step": 2750
+    },
+    {
+      "epoch": 0.4517890501508838,
+      "grad_norm": 0.323580754011984,
+      "learning_rate": 8.853992325953872e-06,
+      "loss": 0.5593,
+      "step": 2751
+    },
+    {
+      "epoch": 0.45195327735922647,
+      "grad_norm": 0.3877844317699021,
+      "learning_rate": 8.853873809500395e-06,
+      "loss": 0.5353,
+      "step": 2752
+    },
+    {
+      "epoch": 0.4521175045675692,
+      "grad_norm": 0.32987127792086013,
+      "learning_rate": 8.853755245759564e-06,
+      "loss": 0.547,
+      "step": 2753
+    },
+    {
+      "epoch": 0.45228173177591197,
+      "grad_norm": 0.3243082577170755,
+      "learning_rate": 8.853636634732668e-06,
+      "loss": 0.5534,
+      "step": 2754
+    },
+    {
+      "epoch": 0.4524459589842547,
+      "grad_norm": 0.31835431916955587,
+      "learning_rate": 8.853517976420993e-06,
+      "loss": 0.5437,
+      "step": 2755
+    },
+    {
+      "epoch": 0.45261018619259746,
+      "grad_norm": 0.34976967541051596,
+      "learning_rate": 8.85339927082583e-06,
+      "loss": 0.5577,
+      "step": 2756
+    },
+    {
+      "epoch": 0.4527744134009402,
+      "grad_norm": 0.2923139255600356,
+      "learning_rate": 8.853280517948468e-06,
+      "loss": 0.5569,
+      "step": 2757
+    },
+    {
+      "epoch": 0.45293864060928296,
+      "grad_norm": 0.30552005384252057,
+      "learning_rate": 8.853161717790197e-06,
+      "loss": 0.5416,
+      "step": 2758
+    },
+    {
+      "epoch": 0.4531028678176257,
+      "grad_norm": 0.3990802742968821,
+      "learning_rate": 8.853042870352308e-06,
+      "loss": 0.5465,
+      "step": 2759
+    },
+    {
+      "epoch": 0.4532670950259684,
+      "grad_norm": 0.3461464555447581,
+      "learning_rate": 8.852923975636089e-06,
+      "loss": 0.5565,
+      "step": 2760
+    },
+    {
+      "epoch": 0.45343132223431115,
+      "grad_norm": 0.4068062311659384,
+      "learning_rate": 8.852805033642834e-06,
+      "loss": 0.5293,
+      "step": 2761
+    },
+    {
+      "epoch": 0.4535955494426539,
+      "grad_norm": 0.4050717095926476,
+      "learning_rate": 8.852686044373831e-06,
+      "loss": 0.5512,
+      "step": 2762
+    },
+    {
+      "epoch": 0.45375977665099665,
+      "grad_norm": 0.3106886645730793,
+      "learning_rate": 8.852567007830378e-06,
+      "loss": 0.566,
+      "step": 2763
+    },
+    {
+      "epoch": 0.4539240038593394,
+      "grad_norm": 0.3465009114201763,
+      "learning_rate": 8.852447924013763e-06,
+      "loss": 0.5429,
+      "step": 2764
+    },
+    {
+      "epoch": 0.45408823106768215,
+      "grad_norm": 0.34100205206268885,
+      "learning_rate": 8.852328792925284e-06,
+      "loss": 0.5562,
+      "step": 2765
+    },
+    {
+      "epoch": 0.4542524582760249,
+      "grad_norm": 0.3373659034184963,
+      "learning_rate": 8.85220961456623e-06,
+      "loss": 0.5508,
+      "step": 2766
+    },
+    {
+      "epoch": 0.45441668548436764,
+      "grad_norm": 0.32662364480543243,
+      "learning_rate": 8.852090388937899e-06,
+      "loss": 0.5574,
+      "step": 2767
+    },
+    {
+      "epoch": 0.45458091269271034,
+      "grad_norm": 0.29597270835282974,
+      "learning_rate": 8.851971116041582e-06,
+      "loss": 0.5619,
+      "step": 2768
+    },
+    {
+      "epoch": 0.4547451399010531,
+      "grad_norm": 0.3269485670857866,
+      "learning_rate": 8.85185179587858e-06,
+      "loss": 0.5336,
+      "step": 2769
+    },
+    {
+      "epoch": 0.45490936710939583,
+      "grad_norm": 0.3079376099727638,
+      "learning_rate": 8.851732428450183e-06,
+      "loss": 0.549,
+      "step": 2770
+    },
+    {
+      "epoch": 0.4550735943177386,
+      "grad_norm": 0.3205362156914591,
+      "learning_rate": 8.851613013757693e-06,
+      "loss": 0.5441,
+      "step": 2771
+    },
+    {
+      "epoch": 0.45523782152608133,
+      "grad_norm": 0.32782562576015545,
+      "learning_rate": 8.851493551802403e-06,
+      "loss": 0.5498,
+      "step": 2772
+    },
+    {
+      "epoch": 0.4554020487344241,
+      "grad_norm": 0.36441285832394543,
+      "learning_rate": 8.851374042585612e-06,
+      "loss": 0.5473,
+      "step": 2773
+    },
+    {
+      "epoch": 0.45556627594276683,
+      "grad_norm": 0.3225326006599982,
+      "learning_rate": 8.851254486108616e-06,
+      "loss": 0.5283,
+      "step": 2774
+    },
+    {
+      "epoch": 0.4557305031511096,
+      "grad_norm": 0.29349747631366513,
+      "learning_rate": 8.851134882372716e-06,
+      "loss": 0.5582,
+      "step": 2775
+    },
+    {
+      "epoch": 0.4558947303594523,
+      "grad_norm": 0.3107685467178685,
+      "learning_rate": 8.851015231379211e-06,
+      "loss": 0.5504,
+      "step": 2776
+    },
+    {
+      "epoch": 0.456058957567795,
+      "grad_norm": 0.31413826669046374,
+      "learning_rate": 8.8508955331294e-06,
+      "loss": 0.5159,
+      "step": 2777
+    },
+    {
+      "epoch": 0.45622318477613777,
+      "grad_norm": 0.2936011119768247,
+      "learning_rate": 8.850775787624584e-06,
+      "loss": 0.539,
+      "step": 2778
+    },
+    {
+      "epoch": 0.4563874119844805,
+      "grad_norm": 0.3475480393438803,
+      "learning_rate": 8.85065599486606e-06,
+      "loss": 0.5642,
+      "step": 2779
+    },
+    {
+      "epoch": 0.45655163919282327,
+      "grad_norm": 0.3098780378515171,
+      "learning_rate": 8.850536154855132e-06,
+      "loss": 0.5429,
+      "step": 2780
+    },
+    {
+      "epoch": 0.456715866401166,
+      "grad_norm": 0.9929195452826118,
+      "learning_rate": 8.850416267593102e-06,
+      "loss": 0.5698,
+      "step": 2781
+    },
+    {
+      "epoch": 0.45688009360950876,
+      "grad_norm": 0.3244883281188489,
+      "learning_rate": 8.85029633308127e-06,
+      "loss": 0.5742,
+      "step": 2782
+    },
+    {
+      "epoch": 0.4570443208178515,
+      "grad_norm": 0.31113246989096616,
+      "learning_rate": 8.85017635132094e-06,
+      "loss": 0.5397,
+      "step": 2783
+    },
+    {
+      "epoch": 0.45720854802619426,
+      "grad_norm": 0.31519864743587955,
+      "learning_rate": 8.850056322313414e-06,
+      "loss": 0.5792,
+      "step": 2784
+    },
+    {
+      "epoch": 0.45737277523453695,
+      "grad_norm": 0.3145460315844349,
+      "learning_rate": 8.849936246059998e-06,
+      "loss": 0.5623,
+      "step": 2785
+    },
+    {
+      "epoch": 0.4575370024428797,
+      "grad_norm": 0.38866288927128867,
+      "learning_rate": 8.849816122561993e-06,
+      "loss": 0.5444,
+      "step": 2786
+    },
+    {
+      "epoch": 0.45770122965122245,
+      "grad_norm": 0.32096724495160234,
+      "learning_rate": 8.849695951820707e-06,
+      "loss": 0.5496,
+      "step": 2787
+    },
+    {
+      "epoch": 0.4578654568595652,
+      "grad_norm": 0.37573771878896706,
+      "learning_rate": 8.849575733837444e-06,
+      "loss": 0.5418,
+      "step": 2788
+    },
+    {
+      "epoch": 0.45802968406790795,
+      "grad_norm": 0.27925464921147647,
+      "learning_rate": 8.849455468613506e-06,
+      "loss": 0.5458,
+      "step": 2789
+    },
+    {
+      "epoch": 0.4581939112762507,
+      "grad_norm": 0.36654175597987393,
+      "learning_rate": 8.849335156150205e-06,
+      "loss": 0.5602,
+      "step": 2790
+    },
+    {
+      "epoch": 0.45835813848459345,
+      "grad_norm": 0.3155475116274339,
+      "learning_rate": 8.849214796448844e-06,
+      "loss": 0.5824,
+      "step": 2791
+    },
+    {
+      "epoch": 0.4585223656929362,
+      "grad_norm": 0.3428136888193705,
+      "learning_rate": 8.84909438951073e-06,
+      "loss": 0.569,
+      "step": 2792
+    },
+    {
+      "epoch": 0.45868659290127894,
+      "grad_norm": 0.30583378806886213,
+      "learning_rate": 8.848973935337174e-06,
+      "loss": 0.5545,
+      "step": 2793
+    },
+    {
+      "epoch": 0.45885082010962164,
+      "grad_norm": 0.33735566982609017,
+      "learning_rate": 8.848853433929482e-06,
+      "loss": 0.5572,
+      "step": 2794
+    },
+    {
+      "epoch": 0.4590150473179644,
+      "grad_norm": 0.31361528788938636,
+      "learning_rate": 8.848732885288963e-06,
+      "loss": 0.5697,
+      "step": 2795
+    },
+    {
+      "epoch": 0.45917927452630714,
+      "grad_norm": 0.34545031796156106,
+      "learning_rate": 8.848612289416926e-06,
+      "loss": 0.5642,
+      "step": 2796
+    },
+    {
+      "epoch": 0.4593435017346499,
+      "grad_norm": 0.2874345855470007,
+      "learning_rate": 8.84849164631468e-06,
+      "loss": 0.5849,
+      "step": 2797
+    },
+    {
+      "epoch": 0.45950772894299263,
+      "grad_norm": 0.28314439970383,
+      "learning_rate": 8.848370955983539e-06,
+      "loss": 0.5566,
+      "step": 2798
+    },
+    {
+      "epoch": 0.4596719561513354,
+      "grad_norm": 0.29925600402316954,
+      "learning_rate": 8.848250218424809e-06,
+      "loss": 0.5598,
+      "step": 2799
+    },
+    {
+      "epoch": 0.45983618335967813,
+      "grad_norm": 0.5002875320880814,
+      "learning_rate": 8.848129433639803e-06,
+      "loss": 0.5434,
+      "step": 2800
+    },
+    {
+      "epoch": 0.4600004105680209,
+      "grad_norm": 0.33733296405596075,
+      "learning_rate": 8.848008601629834e-06,
+      "loss": 0.5406,
+      "step": 2801
+    },
+    {
+      "epoch": 0.46016463777636357,
+      "grad_norm": 0.3162926750411412,
+      "learning_rate": 8.847887722396215e-06,
+      "loss": 0.5539,
+      "step": 2802
+    },
+    {
+      "epoch": 0.4603288649847063,
+      "grad_norm": 0.3281447732451233,
+      "learning_rate": 8.847766795940256e-06,
+      "loss": 0.5747,
+      "step": 2803
+    },
+    {
+      "epoch": 0.46049309219304907,
+      "grad_norm": 0.2805684954794172,
+      "learning_rate": 8.847645822263274e-06,
+      "loss": 0.5665,
+      "step": 2804
+    },
+    {
+      "epoch": 0.4606573194013918,
+      "grad_norm": 0.39677613691474567,
+      "learning_rate": 8.847524801366579e-06,
+      "loss": 0.5454,
+      "step": 2805
+    },
+    {
+      "epoch": 0.46082154660973457,
+      "grad_norm": 0.29768665118106963,
+      "learning_rate": 8.847403733251488e-06,
+      "loss": 0.5619,
+      "step": 2806
+    },
+    {
+      "epoch": 0.4609857738180773,
+      "grad_norm": 0.2719942035991359,
+      "learning_rate": 8.847282617919317e-06,
+      "loss": 0.5527,
+      "step": 2807
+    },
+    {
+      "epoch": 0.46115000102642006,
+      "grad_norm": 0.30476150813606573,
+      "learning_rate": 8.847161455371376e-06,
+      "loss": 0.5661,
+      "step": 2808
+    },
+    {
+      "epoch": 0.4613142282347628,
+      "grad_norm": 0.35212052120108744,
+      "learning_rate": 8.847040245608987e-06,
+      "loss": 0.54,
+      "step": 2809
+    },
+    {
+      "epoch": 0.46147845544310556,
+      "grad_norm": 0.3856979779368862,
+      "learning_rate": 8.846918988633464e-06,
+      "loss": 0.5946,
+      "step": 2810
+    },
+    {
+      "epoch": 0.46164268265144826,
+      "grad_norm": 0.28956054679877297,
+      "learning_rate": 8.846797684446123e-06,
+      "loss": 0.5461,
+      "step": 2811
+    },
+    {
+      "epoch": 0.461806909859791,
+      "grad_norm": 0.3947231541849336,
+      "learning_rate": 8.846676333048283e-06,
+      "loss": 0.5781,
+      "step": 2812
+    },
+    {
+      "epoch": 0.46197113706813375,
+      "grad_norm": 0.3223342533441565,
+      "learning_rate": 8.84655493444126e-06,
+      "loss": 0.5637,
+      "step": 2813
+    },
+    {
+      "epoch": 0.4621353642764765,
+      "grad_norm": 0.31670445757816623,
+      "learning_rate": 8.846433488626376e-06,
+      "loss": 0.5548,
+      "step": 2814
+    },
+    {
+      "epoch": 0.46229959148481925,
+      "grad_norm": 0.3614008756161484,
+      "learning_rate": 8.846311995604947e-06,
+      "loss": 0.5659,
+      "step": 2815
+    },
+    {
+      "epoch": 0.462463818693162,
+      "grad_norm": 0.3072418723805468,
+      "learning_rate": 8.846190455378293e-06,
+      "loss": 0.5395,
+      "step": 2816
+    },
+    {
+      "epoch": 0.46262804590150475,
+      "grad_norm": 0.3025970724441588,
+      "learning_rate": 8.846068867947736e-06,
+      "loss": 0.545,
+      "step": 2817
+    },
+    {
+      "epoch": 0.4627922731098475,
+      "grad_norm": 0.4316806453711984,
+      "learning_rate": 8.845947233314595e-06,
+      "loss": 0.5384,
+      "step": 2818
+    },
+    {
+      "epoch": 0.4629565003181902,
+      "grad_norm": 0.3162804536736171,
+      "learning_rate": 8.845825551480192e-06,
+      "loss": 0.5714,
+      "step": 2819
+    },
+    {
+      "epoch": 0.46312072752653294,
+      "grad_norm": 0.3311951157979156,
+      "learning_rate": 8.845703822445845e-06,
+      "loss": 0.5648,
+      "step": 2820
+    },
+    {
+      "epoch": 0.4632849547348757,
+      "grad_norm": 0.3077557161039132,
+      "learning_rate": 8.84558204621288e-06,
+      "loss": 0.5737,
+      "step": 2821
+    },
+    {
+      "epoch": 0.46344918194321844,
+      "grad_norm": 0.4303477814296107,
+      "learning_rate": 8.845460222782619e-06,
+      "loss": 0.5653,
+      "step": 2822
+    },
+    {
+      "epoch": 0.4636134091515612,
+      "grad_norm": 0.29648758662065006,
+      "learning_rate": 8.845338352156384e-06,
+      "loss": 0.561,
+      "step": 2823
+    },
+    {
+      "epoch": 0.46377763635990393,
+      "grad_norm": 0.3183803997527791,
+      "learning_rate": 8.8452164343355e-06,
+      "loss": 0.5502,
+      "step": 2824
+    },
+    {
+      "epoch": 0.4639418635682467,
+      "grad_norm": 0.3179317015924949,
+      "learning_rate": 8.845094469321291e-06,
+      "loss": 0.5111,
+      "step": 2825
+    },
+    {
+      "epoch": 0.46410609077658943,
+      "grad_norm": 0.33356525833540174,
+      "learning_rate": 8.84497245711508e-06,
+      "loss": 0.5532,
+      "step": 2826
+    },
+    {
+      "epoch": 0.4642703179849322,
+      "grad_norm": 0.35637651622204897,
+      "learning_rate": 8.844850397718193e-06,
+      "loss": 0.5457,
+      "step": 2827
+    },
+    {
+      "epoch": 0.4644345451932749,
+      "grad_norm": 0.3510786554960796,
+      "learning_rate": 8.844728291131956e-06,
+      "loss": 0.542,
+      "step": 2828
+    },
+    {
+      "epoch": 0.4645987724016176,
+      "grad_norm": 0.36430876563423503,
+      "learning_rate": 8.844606137357697e-06,
+      "loss": 0.5329,
+      "step": 2829
+    },
+    {
+      "epoch": 0.46476299960996037,
+      "grad_norm": 0.28596151687840704,
+      "learning_rate": 8.84448393639674e-06,
+      "loss": 0.5466,
+      "step": 2830
+    },
+    {
+      "epoch": 0.4649272268183031,
+      "grad_norm": 0.3332646439130557,
+      "learning_rate": 8.844361688250412e-06,
+      "loss": 0.5234,
+      "step": 2831
+    },
+    {
+      "epoch": 0.46509145402664587,
+      "grad_norm": 0.29528217400431656,
+      "learning_rate": 8.844239392920044e-06,
+      "loss": 0.5533,
+      "step": 2832
+    },
+    {
+      "epoch": 0.4652556812349886,
+      "grad_norm": 0.3045475455711421,
+      "learning_rate": 8.844117050406958e-06,
+      "loss": 0.5516,
+      "step": 2833
+    },
+    {
+      "epoch": 0.46541990844333136,
+      "grad_norm": 0.3673341114723952,
+      "learning_rate": 8.84399466071249e-06,
+      "loss": 0.5628,
+      "step": 2834
+    },
+    {
+      "epoch": 0.4655841356516741,
+      "grad_norm": 0.36352857411555345,
+      "learning_rate": 8.843872223837964e-06,
+      "loss": 0.5527,
+      "step": 2835
+    },
+    {
+      "epoch": 0.4657483628600168,
+      "grad_norm": 0.37350387535762136,
+      "learning_rate": 8.843749739784714e-06,
+      "loss": 0.563,
+      "step": 2836
+    },
+    {
+      "epoch": 0.46591259006835956,
+      "grad_norm": 0.31369805940776757,
+      "learning_rate": 8.843627208554067e-06,
+      "loss": 0.5599,
+      "step": 2837
+    },
+    {
+      "epoch": 0.4660768172767023,
+      "grad_norm": 0.30836598525731435,
+      "learning_rate": 8.843504630147356e-06,
+      "loss": 0.5575,
+      "step": 2838
+    },
+    {
+      "epoch": 0.46624104448504505,
+      "grad_norm": 0.3666029081716213,
+      "learning_rate": 8.843382004565909e-06,
+      "loss": 0.544,
+      "step": 2839
+    },
+    {
+      "epoch": 0.4664052716933878,
+      "grad_norm": 0.32683584986669917,
+      "learning_rate": 8.843259331811062e-06,
+      "loss": 0.5926,
+      "step": 2840
+    },
+    {
+      "epoch": 0.46656949890173055,
+      "grad_norm": 0.3935470331834398,
+      "learning_rate": 8.843136611884145e-06,
+      "loss": 0.5502,
+      "step": 2841
+    },
+    {
+      "epoch": 0.4667337261100733,
+      "grad_norm": 0.32710167529269035,
+      "learning_rate": 8.843013844786491e-06,
+      "loss": 0.5599,
+      "step": 2842
+    },
+    {
+      "epoch": 0.46689795331841605,
+      "grad_norm": 0.3596882455911912,
+      "learning_rate": 8.842891030519434e-06,
+      "loss": 0.551,
+      "step": 2843
+    },
+    {
+      "epoch": 0.4670621805267588,
+      "grad_norm": 0.29837305973530415,
+      "learning_rate": 8.842768169084309e-06,
+      "loss": 0.5616,
+      "step": 2844
+    },
+    {
+      "epoch": 0.4672264077351015,
+      "grad_norm": 0.36261886857746567,
+      "learning_rate": 8.842645260482446e-06,
+      "loss": 0.546,
+      "step": 2845
+    },
+    {
+      "epoch": 0.46739063494344424,
+      "grad_norm": 0.3239781103541487,
+      "learning_rate": 8.842522304715184e-06,
+      "loss": 0.5493,
+      "step": 2846
+    },
+    {
+      "epoch": 0.467554862151787,
+      "grad_norm": 0.3022618320478039,
+      "learning_rate": 8.842399301783859e-06,
+      "loss": 0.5563,
+      "step": 2847
+    },
+    {
+      "epoch": 0.46771908936012974,
+      "grad_norm": 0.3324394490245069,
+      "learning_rate": 8.842276251689804e-06,
+      "loss": 0.5467,
+      "step": 2848
+    },
+    {
+      "epoch": 0.4678833165684725,
+      "grad_norm": 0.34005289281067436,
+      "learning_rate": 8.842153154434357e-06,
+      "loss": 0.5507,
+      "step": 2849
+    },
+    {
+      "epoch": 0.46804754377681523,
+      "grad_norm": 0.30758197825127814,
+      "learning_rate": 8.842030010018855e-06,
+      "loss": 0.5617,
+      "step": 2850
+    },
+    {
+      "epoch": 0.468211770985158,
+      "grad_norm": 0.3485397690878994,
+      "learning_rate": 8.841906818444634e-06,
+      "loss": 0.5414,
+      "step": 2851
+    },
+    {
+      "epoch": 0.46837599819350073,
+      "grad_norm": 0.28465131392759013,
+      "learning_rate": 8.841783579713033e-06,
+      "loss": 0.5549,
+      "step": 2852
+    },
+    {
+      "epoch": 0.4685402254018434,
+      "grad_norm": 0.340827843779612,
+      "learning_rate": 8.841660293825392e-06,
+      "loss": 0.5542,
+      "step": 2853
+    },
+    {
+      "epoch": 0.4687044526101862,
+      "grad_norm": 0.3642631778373032,
+      "learning_rate": 8.841536960783047e-06,
+      "loss": 0.5422,
+      "step": 2854
+    },
+    {
+      "epoch": 0.4688686798185289,
+      "grad_norm": 0.3350444847989983,
+      "learning_rate": 8.84141358058734e-06,
+      "loss": 0.5582,
+      "step": 2855
+    },
+    {
+      "epoch": 0.46903290702687167,
+      "grad_norm": 0.32037747228626573,
+      "learning_rate": 8.84129015323961e-06,
+      "loss": 0.5511,
+      "step": 2856
+    },
+    {
+      "epoch": 0.4691971342352144,
+      "grad_norm": 0.32535957757017653,
+      "learning_rate": 8.841166678741197e-06,
+      "loss": 0.5542,
+      "step": 2857
+    },
+    {
+      "epoch": 0.46936136144355717,
+      "grad_norm": 0.3612115613975925,
+      "learning_rate": 8.841043157093444e-06,
+      "loss": 0.5645,
+      "step": 2858
+    },
+    {
+      "epoch": 0.4695255886518999,
+      "grad_norm": 0.3032820965069931,
+      "learning_rate": 8.840919588297691e-06,
+      "loss": 0.5561,
+      "step": 2859
+    },
+    {
+      "epoch": 0.46968981586024267,
+      "grad_norm": 0.34091583425847477,
+      "learning_rate": 8.84079597235528e-06,
+      "loss": 0.5539,
+      "step": 2860
+    },
+    {
+      "epoch": 0.4698540430685854,
+      "grad_norm": 0.3366327725164345,
+      "learning_rate": 8.840672309267553e-06,
+      "loss": 0.5512,
+      "step": 2861
+    },
+    {
+      "epoch": 0.4700182702769281,
+      "grad_norm": 0.33509050371301097,
+      "learning_rate": 8.840548599035857e-06,
+      "loss": 0.5806,
+      "step": 2862
+    },
+    {
+      "epoch": 0.47018249748527086,
+      "grad_norm": 0.3043078378641758,
+      "learning_rate": 8.84042484166153e-06,
+      "loss": 0.5549,
+      "step": 2863
+    },
+    {
+      "epoch": 0.4703467246936136,
+      "grad_norm": 0.2911581650021685,
+      "learning_rate": 8.840301037145919e-06,
+      "loss": 0.5385,
+      "step": 2864
+    },
+    {
+      "epoch": 0.47051095190195635,
+      "grad_norm": 0.33595882669664007,
+      "learning_rate": 8.840177185490369e-06,
+      "loss": 0.5284,
+      "step": 2865
+    },
+    {
+      "epoch": 0.4706751791102991,
+      "grad_norm": 0.3617199149406533,
+      "learning_rate": 8.840053286696224e-06,
+      "loss": 0.5446,
+      "step": 2866
+    },
+    {
+      "epoch": 0.47083940631864185,
+      "grad_norm": 0.30372519781604457,
+      "learning_rate": 8.839929340764832e-06,
+      "loss": 0.5728,
+      "step": 2867
+    },
+    {
+      "epoch": 0.4710036335269846,
+      "grad_norm": 0.31933667718002556,
+      "learning_rate": 8.839805347697536e-06,
+      "loss": 0.5518,
+      "step": 2868
+    },
+    {
+      "epoch": 0.47116786073532735,
+      "grad_norm": 0.3966424493600261,
+      "learning_rate": 8.839681307495685e-06,
+      "loss": 0.5743,
+      "step": 2869
+    },
+    {
+      "epoch": 0.47133208794367004,
+      "grad_norm": 0.3348971739237751,
+      "learning_rate": 8.839557220160626e-06,
+      "loss": 0.5524,
+      "step": 2870
+    },
+    {
+      "epoch": 0.4714963151520128,
+      "grad_norm": 0.3213892824965565,
+      "learning_rate": 8.839433085693704e-06,
+      "loss": 0.5628,
+      "step": 2871
+    },
+    {
+      "epoch": 0.47166054236035554,
+      "grad_norm": 0.4390491443368386,
+      "learning_rate": 8.839308904096272e-06,
+      "loss": 0.5376,
+      "step": 2872
+    },
+    {
+      "epoch": 0.4718247695686983,
+      "grad_norm": 0.32059743684089986,
+      "learning_rate": 8.839184675369672e-06,
+      "loss": 0.5614,
+      "step": 2873
+    },
+    {
+      "epoch": 0.47198899677704104,
+      "grad_norm": 0.3058018710793769,
+      "learning_rate": 8.83906039951526e-06,
+      "loss": 0.5359,
+      "step": 2874
+    },
+    {
+      "epoch": 0.4721532239853838,
+      "grad_norm": 0.2804240335544641,
+      "learning_rate": 8.838936076534381e-06,
+      "loss": 0.5442,
+      "step": 2875
+    },
+    {
+      "epoch": 0.47231745119372653,
+      "grad_norm": 0.3746199995295414,
+      "learning_rate": 8.83881170642839e-06,
+      "loss": 0.5763,
+      "step": 2876
+    },
+    {
+      "epoch": 0.4724816784020693,
+      "grad_norm": 0.43755114764082026,
+      "learning_rate": 8.838687289198634e-06,
+      "loss": 0.5677,
+      "step": 2877
+    },
+    {
+      "epoch": 0.47264590561041203,
+      "grad_norm": 0.30021384425688985,
+      "learning_rate": 8.838562824846464e-06,
+      "loss": 0.5354,
+      "step": 2878
+    },
+    {
+      "epoch": 0.4728101328187547,
+      "grad_norm": 0.2982429736947675,
+      "learning_rate": 8.838438313373234e-06,
+      "loss": 0.5523,
+      "step": 2879
+    },
+    {
+      "epoch": 0.4729743600270975,
+      "grad_norm": 0.3139337209816409,
+      "learning_rate": 8.838313754780297e-06,
+      "loss": 0.5442,
+      "step": 2880
+    },
+    {
+      "epoch": 0.4731385872354402,
+      "grad_norm": 0.38187862880767065,
+      "learning_rate": 8.838189149069004e-06,
+      "loss": 0.5596,
+      "step": 2881
+    },
+    {
+      "epoch": 0.47330281444378297,
+      "grad_norm": 0.35478374350718284,
+      "learning_rate": 8.838064496240706e-06,
+      "loss": 0.5492,
+      "step": 2882
+    },
+    {
+      "epoch": 0.4734670416521257,
+      "grad_norm": 0.3149790656321742,
+      "learning_rate": 8.837939796296762e-06,
+      "loss": 0.5453,
+      "step": 2883
+    },
+    {
+      "epoch": 0.47363126886046847,
+      "grad_norm": 0.349800653583477,
+      "learning_rate": 8.837815049238523e-06,
+      "loss": 0.5303,
+      "step": 2884
+    },
+    {
+      "epoch": 0.4737954960688112,
+      "grad_norm": 0.3199958228053064,
+      "learning_rate": 8.837690255067346e-06,
+      "loss": 0.5516,
+      "step": 2885
+    },
+    {
+      "epoch": 0.47395972327715397,
+      "grad_norm": 0.3614633246024179,
+      "learning_rate": 8.837565413784583e-06,
+      "loss": 0.5503,
+      "step": 2886
+    },
+    {
+      "epoch": 0.47412395048549666,
+      "grad_norm": 0.41022846573910837,
+      "learning_rate": 8.837440525391593e-06,
+      "loss": 0.5704,
+      "step": 2887
+    },
+    {
+      "epoch": 0.4742881776938394,
+      "grad_norm": 0.32146513999845583,
+      "learning_rate": 8.83731558988973e-06,
+      "loss": 0.5593,
+      "step": 2888
+    },
+    {
+      "epoch": 0.47445240490218216,
+      "grad_norm": 0.3522243215874715,
+      "learning_rate": 8.837190607280355e-06,
+      "loss": 0.5535,
+      "step": 2889
+    },
+    {
+      "epoch": 0.4746166321105249,
+      "grad_norm": 0.34027532771479646,
+      "learning_rate": 8.83706557756482e-06,
+      "loss": 0.5422,
+      "step": 2890
+    },
+    {
+      "epoch": 0.47478085931886765,
+      "grad_norm": 0.3531730299805744,
+      "learning_rate": 8.836940500744489e-06,
+      "loss": 0.5617,
+      "step": 2891
+    },
+    {
+      "epoch": 0.4749450865272104,
+      "grad_norm": 0.3674374072994261,
+      "learning_rate": 8.836815376820715e-06,
+      "loss": 0.5592,
+      "step": 2892
+    },
+    {
+      "epoch": 0.47510931373555315,
+      "grad_norm": 0.32970365360706744,
+      "learning_rate": 8.836690205794858e-06,
+      "loss": 0.5676,
+      "step": 2893
+    },
+    {
+      "epoch": 0.4752735409438959,
+      "grad_norm": 0.39642473826995,
+      "learning_rate": 8.836564987668281e-06,
+      "loss": 0.5787,
+      "step": 2894
+    },
+    {
+      "epoch": 0.47543776815223865,
+      "grad_norm": 0.2868884450572186,
+      "learning_rate": 8.836439722442341e-06,
+      "loss": 0.5578,
+      "step": 2895
+    },
+    {
+      "epoch": 0.47560199536058134,
+      "grad_norm": 0.3071135606555718,
+      "learning_rate": 8.8363144101184e-06,
+      "loss": 0.5446,
+      "step": 2896
+    },
+    {
+      "epoch": 0.4757662225689241,
+      "grad_norm": 0.4085356980863807,
+      "learning_rate": 8.836189050697817e-06,
+      "loss": 0.5494,
+      "step": 2897
+    },
+    {
+      "epoch": 0.47593044977726684,
+      "grad_norm": 0.29666396729431577,
+      "learning_rate": 8.836063644181954e-06,
+      "loss": 0.544,
+      "step": 2898
+    },
+    {
+      "epoch": 0.4760946769856096,
+      "grad_norm": 0.30255551130327596,
+      "learning_rate": 8.835938190572174e-06,
+      "loss": 0.5499,
+      "step": 2899
+    },
+    {
+      "epoch": 0.47625890419395234,
+      "grad_norm": 0.2806984477691293,
+      "learning_rate": 8.83581268986984e-06,
+      "loss": 0.5648,
+      "step": 2900
+    },
+    {
+      "epoch": 0.4764231314022951,
+      "grad_norm": 0.3263273596646186,
+      "learning_rate": 8.835687142076314e-06,
+      "loss": 0.5561,
+      "step": 2901
+    },
+    {
+      "epoch": 0.47658735861063783,
+      "grad_norm": 0.3790842531002415,
+      "learning_rate": 8.83556154719296e-06,
+      "loss": 0.5744,
+      "step": 2902
+    },
+    {
+      "epoch": 0.4767515858189806,
+      "grad_norm": 0.3417983950853606,
+      "learning_rate": 8.835435905221142e-06,
+      "loss": 0.5652,
+      "step": 2903
+    },
+    {
+      "epoch": 0.4769158130273233,
+      "grad_norm": 0.29218721475019915,
+      "learning_rate": 8.835310216162224e-06,
+      "loss": 0.5554,
+      "step": 2904
+    },
+    {
+      "epoch": 0.477080040235666,
+      "grad_norm": 0.31113339800938206,
+      "learning_rate": 8.835184480017572e-06,
+      "loss": 0.5521,
+      "step": 2905
+    },
+    {
+      "epoch": 0.4772442674440088,
+      "grad_norm": 0.2985823425378832,
+      "learning_rate": 8.835058696788552e-06,
+      "loss": 0.5459,
+      "step": 2906
+    },
+    {
+      "epoch": 0.4774084946523515,
+      "grad_norm": 0.35116196870091576,
+      "learning_rate": 8.834932866476531e-06,
+      "loss": 0.5569,
+      "step": 2907
+    },
+    {
+      "epoch": 0.47757272186069427,
+      "grad_norm": 0.285228916766138,
+      "learning_rate": 8.83480698908287e-06,
+      "loss": 0.5476,
+      "step": 2908
+    },
+    {
+      "epoch": 0.477736949069037,
+      "grad_norm": 0.32355675164680775,
+      "learning_rate": 8.834681064608944e-06,
+      "loss": 0.5616,
+      "step": 2909
+    },
+    {
+      "epoch": 0.47790117627737977,
+      "grad_norm": 0.29250415203008606,
+      "learning_rate": 8.834555093056114e-06,
+      "loss": 0.5452,
+      "step": 2910
+    },
+    {
+      "epoch": 0.4780654034857225,
+      "grad_norm": 0.4086216732635031,
+      "learning_rate": 8.834429074425752e-06,
+      "loss": 0.5887,
+      "step": 2911
+    },
+    {
+      "epoch": 0.47822963069406527,
+      "grad_norm": 0.30486039535028525,
+      "learning_rate": 8.834303008719226e-06,
+      "loss": 0.5586,
+      "step": 2912
+    },
+    {
+      "epoch": 0.47839385790240796,
+      "grad_norm": 0.392026031748003,
+      "learning_rate": 8.834176895937906e-06,
+      "loss": 0.5454,
+      "step": 2913
+    },
+    {
+      "epoch": 0.4785580851107507,
+      "grad_norm": 0.32606961737462525,
+      "learning_rate": 8.834050736083158e-06,
+      "loss": 0.5565,
+      "step": 2914
+    },
+    {
+      "epoch": 0.47872231231909346,
+      "grad_norm": 0.3567403007077362,
+      "learning_rate": 8.833924529156357e-06,
+      "loss": 0.5579,
+      "step": 2915
+    },
+    {
+      "epoch": 0.4788865395274362,
+      "grad_norm": 0.2962722445309099,
+      "learning_rate": 8.833798275158871e-06,
+      "loss": 0.5517,
+      "step": 2916
+    },
+    {
+      "epoch": 0.47905076673577895,
+      "grad_norm": 0.31457620394082064,
+      "learning_rate": 8.83367197409207e-06,
+      "loss": 0.5654,
+      "step": 2917
+    },
+    {
+      "epoch": 0.4792149939441217,
+      "grad_norm": 0.36662242627563446,
+      "learning_rate": 8.833545625957332e-06,
+      "loss": 0.552,
+      "step": 2918
+    },
+    {
+      "epoch": 0.47937922115246445,
+      "grad_norm": 0.28945439867043166,
+      "learning_rate": 8.833419230756021e-06,
+      "loss": 0.5611,
+      "step": 2919
+    },
+    {
+      "epoch": 0.4795434483608072,
+      "grad_norm": 0.3090300015262434,
+      "learning_rate": 8.833292788489517e-06,
+      "loss": 0.57,
+      "step": 2920
+    },
+    {
+      "epoch": 0.4797076755691499,
+      "grad_norm": 0.3367798755701906,
+      "learning_rate": 8.833166299159187e-06,
+      "loss": 0.5635,
+      "step": 2921
+    },
+    {
+      "epoch": 0.47987190277749264,
+      "grad_norm": 0.3211307845952408,
+      "learning_rate": 8.833039762766408e-06,
+      "loss": 0.539,
+      "step": 2922
+    },
+    {
+      "epoch": 0.4800361299858354,
+      "grad_norm": 0.3269164013348773,
+      "learning_rate": 8.832913179312555e-06,
+      "loss": 0.5481,
+      "step": 2923
+    },
+    {
+      "epoch": 0.48020035719417814,
+      "grad_norm": 0.3193091915753343,
+      "learning_rate": 8.832786548799002e-06,
+      "loss": 0.5685,
+      "step": 2924
+    },
+    {
+      "epoch": 0.4803645844025209,
+      "grad_norm": 0.29715881654797105,
+      "learning_rate": 8.832659871227124e-06,
+      "loss": 0.547,
+      "step": 2925
+    },
+    {
+      "epoch": 0.48052881161086364,
+      "grad_norm": 0.3092509709334806,
+      "learning_rate": 8.832533146598297e-06,
+      "loss": 0.5407,
+      "step": 2926
+    },
+    {
+      "epoch": 0.4806930388192064,
+      "grad_norm": 0.2997878748298698,
+      "learning_rate": 8.832406374913896e-06,
+      "loss": 0.5437,
+      "step": 2927
+    },
+    {
+      "epoch": 0.48085726602754914,
+      "grad_norm": 0.36833955082132547,
+      "learning_rate": 8.832279556175302e-06,
+      "loss": 0.5591,
+      "step": 2928
+    },
+    {
+      "epoch": 0.4810214932358919,
+      "grad_norm": 0.31188670229780074,
+      "learning_rate": 8.832152690383887e-06,
+      "loss": 0.5523,
+      "step": 2929
+    },
+    {
+      "epoch": 0.4811857204442346,
+      "grad_norm": 0.31498192500648686,
+      "learning_rate": 8.832025777541032e-06,
+      "loss": 0.5367,
+      "step": 2930
+    },
+    {
+      "epoch": 0.4813499476525773,
+      "grad_norm": 0.32468516735093983,
+      "learning_rate": 8.831898817648116e-06,
+      "loss": 0.5561,
+      "step": 2931
+    },
+    {
+      "epoch": 0.4815141748609201,
+      "grad_norm": 0.33711950838323035,
+      "learning_rate": 8.831771810706518e-06,
+      "loss": 0.5711,
+      "step": 2932
+    },
+    {
+      "epoch": 0.4816784020692628,
+      "grad_norm": 0.3467531673707492,
+      "learning_rate": 8.831644756717614e-06,
+      "loss": 0.576,
+      "step": 2933
+    },
+    {
+      "epoch": 0.48184262927760557,
+      "grad_norm": 0.3587735987166933,
+      "learning_rate": 8.831517655682787e-06,
+      "loss": 0.5603,
+      "step": 2934
+    },
+    {
+      "epoch": 0.4820068564859483,
+      "grad_norm": 0.3479101996126524,
+      "learning_rate": 8.831390507603416e-06,
+      "loss": 0.5607,
+      "step": 2935
+    },
+    {
+      "epoch": 0.48217108369429107,
+      "grad_norm": 0.35567025483515724,
+      "learning_rate": 8.831263312480883e-06,
+      "loss": 0.5428,
+      "step": 2936
+    },
+    {
+      "epoch": 0.4823353109026338,
+      "grad_norm": 0.34050479161407216,
+      "learning_rate": 8.831136070316568e-06,
+      "loss": 0.5335,
+      "step": 2937
+    },
+    {
+      "epoch": 0.4824995381109765,
+      "grad_norm": 0.34844624545525676,
+      "learning_rate": 8.831008781111855e-06,
+      "loss": 0.5609,
+      "step": 2938
+    },
+    {
+      "epoch": 0.48266376531931926,
+      "grad_norm": 0.3534685183853488,
+      "learning_rate": 8.830881444868126e-06,
+      "loss": 0.5601,
+      "step": 2939
+    },
+    {
+      "epoch": 0.482827992527662,
+      "grad_norm": 0.3576134623563205,
+      "learning_rate": 8.830754061586764e-06,
+      "loss": 0.5622,
+      "step": 2940
+    },
+    {
+      "epoch": 0.48299221973600476,
+      "grad_norm": 0.37002257796135896,
+      "learning_rate": 8.83062663126915e-06,
+      "loss": 0.556,
+      "step": 2941
+    },
+    {
+      "epoch": 0.4831564469443475,
+      "grad_norm": 0.37832760131778265,
+      "learning_rate": 8.830499153916671e-06,
+      "loss": 0.5463,
+      "step": 2942
+    },
+    {
+      "epoch": 0.48332067415269026,
+      "grad_norm": 0.3257646983345067,
+      "learning_rate": 8.83037162953071e-06,
+      "loss": 0.5626,
+      "step": 2943
+    },
+    {
+      "epoch": 0.483484901361033,
+      "grad_norm": 0.34700830090226,
+      "learning_rate": 8.830244058112655e-06,
+      "loss": 0.5396,
+      "step": 2944
+    },
+    {
+      "epoch": 0.48364912856937575,
+      "grad_norm": 0.31036138655536344,
+      "learning_rate": 8.830116439663887e-06,
+      "loss": 0.5197,
+      "step": 2945
+    },
+    {
+      "epoch": 0.4838133557777185,
+      "grad_norm": 0.3565916198778085,
+      "learning_rate": 8.829988774185794e-06,
+      "loss": 0.5376,
+      "step": 2946
+    },
+    {
+      "epoch": 0.4839775829860612,
+      "grad_norm": 0.3681633251091734,
+      "learning_rate": 8.829861061679763e-06,
+      "loss": 0.5404,
+      "step": 2947
+    },
+    {
+      "epoch": 0.48414181019440394,
+      "grad_norm": 0.4760323163689969,
+      "learning_rate": 8.829733302147182e-06,
+      "loss": 0.5564,
+      "step": 2948
+    },
+    {
+      "epoch": 0.4843060374027467,
+      "grad_norm": 0.3142530374093069,
+      "learning_rate": 8.829605495589436e-06,
+      "loss": 0.5424,
+      "step": 2949
+    },
+    {
+      "epoch": 0.48447026461108944,
+      "grad_norm": 0.395721637651938,
+      "learning_rate": 8.829477642007915e-06,
+      "loss": 0.5428,
+      "step": 2950
+    },
+    {
+      "epoch": 0.4846344918194322,
+      "grad_norm": 0.3317668352816898,
+      "learning_rate": 8.82934974140401e-06,
+      "loss": 0.5413,
+      "step": 2951
+    },
+    {
+      "epoch": 0.48479871902777494,
+      "grad_norm": 0.4043660326447801,
+      "learning_rate": 8.829221793779102e-06,
+      "loss": 0.5645,
+      "step": 2952
+    },
+    {
+      "epoch": 0.4849629462361177,
+      "grad_norm": 0.31928485059252737,
+      "learning_rate": 8.82909379913459e-06,
+      "loss": 0.5396,
+      "step": 2953
+    },
+    {
+      "epoch": 0.48512717344446044,
+      "grad_norm": 0.30272961197459414,
+      "learning_rate": 8.828965757471858e-06,
+      "loss": 0.5611,
+      "step": 2954
+    },
+    {
+      "epoch": 0.48529140065280313,
+      "grad_norm": 0.3219097780968554,
+      "learning_rate": 8.8288376687923e-06,
+      "loss": 0.5495,
+      "step": 2955
+    },
+    {
+      "epoch": 0.4854556278611459,
+      "grad_norm": 0.3068594785573637,
+      "learning_rate": 8.828709533097304e-06,
+      "loss": 0.527,
+      "step": 2956
+    },
+    {
+      "epoch": 0.4856198550694886,
+      "grad_norm": 0.3286151563554068,
+      "learning_rate": 8.828581350388267e-06,
+      "loss": 0.5453,
+      "step": 2957
+    },
+    {
+      "epoch": 0.4857840822778314,
+      "grad_norm": 0.32326057654047574,
+      "learning_rate": 8.828453120666574e-06,
+      "loss": 0.5433,
+      "step": 2958
+    },
+    {
+      "epoch": 0.4859483094861741,
+      "grad_norm": 0.341029084787711,
+      "learning_rate": 8.828324843933625e-06,
+      "loss": 0.5644,
+      "step": 2959
+    },
+    {
+      "epoch": 0.4861125366945169,
+      "grad_norm": 0.34016806957422147,
+      "learning_rate": 8.828196520190807e-06,
+      "loss": 0.5556,
+      "step": 2960
+    },
+    {
+      "epoch": 0.4862767639028596,
+      "grad_norm": 0.4028159993767315,
+      "learning_rate": 8.828068149439518e-06,
+      "loss": 0.5325,
+      "step": 2961
+    },
+    {
+      "epoch": 0.48644099111120237,
+      "grad_norm": 0.2984402049931922,
+      "learning_rate": 8.82793973168115e-06,
+      "loss": 0.5751,
+      "step": 2962
+    },
+    {
+      "epoch": 0.48660521831954506,
+      "grad_norm": 0.33616040754201904,
+      "learning_rate": 8.827811266917099e-06,
+      "loss": 0.5424,
+      "step": 2963
+    },
+    {
+      "epoch": 0.4867694455278878,
+      "grad_norm": 0.333707218309075,
+      "learning_rate": 8.827682755148757e-06,
+      "loss": 0.5424,
+      "step": 2964
+    },
+    {
+      "epoch": 0.48693367273623056,
+      "grad_norm": 0.3336325173722268,
+      "learning_rate": 8.827554196377525e-06,
+      "loss": 0.5617,
+      "step": 2965
+    },
+    {
+      "epoch": 0.4870978999445733,
+      "grad_norm": 0.3299202105690531,
+      "learning_rate": 8.827425590604796e-06,
+      "loss": 0.5422,
+      "step": 2966
+    },
+    {
+      "epoch": 0.48726212715291606,
+      "grad_norm": 0.3506142749597406,
+      "learning_rate": 8.827296937831969e-06,
+      "loss": 0.5609,
+      "step": 2967
+    },
+    {
+      "epoch": 0.4874263543612588,
+      "grad_norm": 0.44404974906674444,
+      "learning_rate": 8.82716823806044e-06,
+      "loss": 0.5782,
+      "step": 2968
+    },
+    {
+      "epoch": 0.48759058156960156,
+      "grad_norm": 0.30580865087991876,
+      "learning_rate": 8.827039491291604e-06,
+      "loss": 0.5581,
+      "step": 2969
+    },
+    {
+      "epoch": 0.4877548087779443,
+      "grad_norm": 0.3048111574103392,
+      "learning_rate": 8.826910697526862e-06,
+      "loss": 0.5343,
+      "step": 2970
+    },
+    {
+      "epoch": 0.48791903598628705,
+      "grad_norm": 0.3281501305555785,
+      "learning_rate": 8.826781856767614e-06,
+      "loss": 0.5627,
+      "step": 2971
+    },
+    {
+      "epoch": 0.48808326319462975,
+      "grad_norm": 0.3026997019538971,
+      "learning_rate": 8.826652969015258e-06,
+      "loss": 0.579,
+      "step": 2972
+    },
+    {
+      "epoch": 0.4882474904029725,
+      "grad_norm": 0.33950303423306283,
+      "learning_rate": 8.826524034271194e-06,
+      "loss": 0.5491,
+      "step": 2973
+    },
+    {
+      "epoch": 0.48841171761131524,
+      "grad_norm": 0.3662426620093517,
+      "learning_rate": 8.82639505253682e-06,
+      "loss": 0.5609,
+      "step": 2974
+    },
+    {
+      "epoch": 0.488575944819658,
+      "grad_norm": 0.3504254683156696,
+      "learning_rate": 8.826266023813543e-06,
+      "loss": 0.5358,
+      "step": 2975
+    },
+    {
+      "epoch": 0.48874017202800074,
+      "grad_norm": 0.362624051096286,
+      "learning_rate": 8.826136948102757e-06,
+      "loss": 0.5515,
+      "step": 2976
+    },
+    {
+      "epoch": 0.4889043992363435,
+      "grad_norm": 0.3303718575483711,
+      "learning_rate": 8.82600782540587e-06,
+      "loss": 0.5562,
+      "step": 2977
+    },
+    {
+      "epoch": 0.48906862644468624,
+      "grad_norm": 0.35238845644920747,
+      "learning_rate": 8.825878655724279e-06,
+      "loss": 0.5385,
+      "step": 2978
+    },
+    {
+      "epoch": 0.489232853653029,
+      "grad_norm": 0.31177874457750715,
+      "learning_rate": 8.825749439059393e-06,
+      "loss": 0.5396,
+      "step": 2979
+    },
+    {
+      "epoch": 0.4893970808613717,
+      "grad_norm": 0.34109682668123126,
+      "learning_rate": 8.825620175412609e-06,
+      "loss": 0.5505,
+      "step": 2980
+    },
+    {
+      "epoch": 0.48956130806971443,
+      "grad_norm": 0.3620326005023896,
+      "learning_rate": 8.825490864785336e-06,
+      "loss": 0.5778,
+      "step": 2981
+    },
+    {
+      "epoch": 0.4897255352780572,
+      "grad_norm": 0.5979291508659067,
+      "learning_rate": 8.825361507178977e-06,
+      "loss": 0.5518,
+      "step": 2982
+    },
+    {
+      "epoch": 0.4898897624863999,
+      "grad_norm": 0.29689712347868064,
+      "learning_rate": 8.825232102594935e-06,
+      "loss": 0.5817,
+      "step": 2983
+    },
+    {
+      "epoch": 0.4900539896947427,
+      "grad_norm": 0.34548738955931996,
+      "learning_rate": 8.825102651034617e-06,
+      "loss": 0.5585,
+      "step": 2984
+    },
+    {
+      "epoch": 0.4902182169030854,
+      "grad_norm": 0.3554958070227503,
+      "learning_rate": 8.82497315249943e-06,
+      "loss": 0.5488,
+      "step": 2985
+    },
+    {
+      "epoch": 0.4903824441114282,
+      "grad_norm": 0.2904946631395032,
+      "learning_rate": 8.82484360699078e-06,
+      "loss": 0.5386,
+      "step": 2986
+    },
+    {
+      "epoch": 0.4905466713197709,
+      "grad_norm": 0.3229198644119859,
+      "learning_rate": 8.824714014510071e-06,
+      "loss": 0.5558,
+      "step": 2987
+    },
+    {
+      "epoch": 0.49071089852811367,
+      "grad_norm": 0.36384501410130143,
+      "learning_rate": 8.824584375058713e-06,
+      "loss": 0.5466,
+      "step": 2988
+    },
+    {
+      "epoch": 0.49087512573645636,
+      "grad_norm": 0.3753164950066053,
+      "learning_rate": 8.824454688638116e-06,
+      "loss": 0.5401,
+      "step": 2989
+    },
+    {
+      "epoch": 0.4910393529447991,
+      "grad_norm": 0.36709523119655635,
+      "learning_rate": 8.824324955249685e-06,
+      "loss": 0.5366,
+      "step": 2990
+    },
+    {
+      "epoch": 0.49120358015314186,
+      "grad_norm": 0.36824250049715346,
+      "learning_rate": 8.82419517489483e-06,
+      "loss": 0.5468,
+      "step": 2991
+    },
+    {
+      "epoch": 0.4913678073614846,
+      "grad_norm": 0.41026285040761973,
+      "learning_rate": 8.824065347574962e-06,
+      "loss": 0.5796,
+      "step": 2992
+    },
+    {
+      "epoch": 0.49153203456982736,
+      "grad_norm": 0.36743089732858497,
+      "learning_rate": 8.82393547329149e-06,
+      "loss": 0.5691,
+      "step": 2993
+    },
+    {
+      "epoch": 0.4916962617781701,
+      "grad_norm": 0.3573127649092164,
+      "learning_rate": 8.823805552045824e-06,
+      "loss": 0.5782,
+      "step": 2994
+    },
+    {
+      "epoch": 0.49186048898651286,
+      "grad_norm": 0.4711373569935757,
+      "learning_rate": 8.823675583839375e-06,
+      "loss": 0.5448,
+      "step": 2995
+    },
+    {
+      "epoch": 0.4920247161948556,
+      "grad_norm": 0.3467146120478822,
+      "learning_rate": 8.823545568673556e-06,
+      "loss": 0.5785,
+      "step": 2996
+    },
+    {
+      "epoch": 0.4921889434031983,
+      "grad_norm": 0.35151193970461303,
+      "learning_rate": 8.823415506549779e-06,
+      "loss": 0.5703,
+      "step": 2997
+    },
+    {
+      "epoch": 0.49235317061154105,
+      "grad_norm": 0.3620038956032762,
+      "learning_rate": 8.823285397469455e-06,
+      "loss": 0.5713,
+      "step": 2998
+    },
+    {
+      "epoch": 0.4925173978198838,
+      "grad_norm": 0.3607731749175902,
+      "learning_rate": 8.823155241434e-06,
+      "loss": 0.5556,
+      "step": 2999
+    },
+    {
+      "epoch": 0.49268162502822654,
+      "grad_norm": 0.32757022768540983,
+      "learning_rate": 8.823025038444823e-06,
+      "loss": 0.5531,
+      "step": 3000
+    },
+    {
+      "epoch": 0.4928458522365693,
+      "grad_norm": 0.3781611992714109,
+      "learning_rate": 8.822894788503342e-06,
+      "loss": 0.5505,
+      "step": 3001
+    },
+    {
+      "epoch": 0.49301007944491204,
+      "grad_norm": 0.3676888297654473,
+      "learning_rate": 8.82276449161097e-06,
+      "loss": 0.5646,
+      "step": 3002
+    },
+    {
+      "epoch": 0.4931743066532548,
+      "grad_norm": 0.3467636715211463,
+      "learning_rate": 8.822634147769123e-06,
+      "loss": 0.538,
+      "step": 3003
+    },
+    {
+      "epoch": 0.49333853386159754,
+      "grad_norm": 0.29828305218985435,
+      "learning_rate": 8.822503756979217e-06,
+      "loss": 0.5407,
+      "step": 3004
+    },
+    {
+      "epoch": 0.4935027610699403,
+      "grad_norm": 0.3722421038012637,
+      "learning_rate": 8.822373319242666e-06,
+      "loss": 0.553,
+      "step": 3005
+    },
+    {
+      "epoch": 0.493666988278283,
+      "grad_norm": 0.3964608753926012,
+      "learning_rate": 8.822242834560888e-06,
+      "loss": 0.5506,
+      "step": 3006
+    },
+    {
+      "epoch": 0.49383121548662573,
+      "grad_norm": 0.3038259436819525,
+      "learning_rate": 8.822112302935302e-06,
+      "loss": 0.5528,
+      "step": 3007
+    },
+    {
+      "epoch": 0.4939954426949685,
+      "grad_norm": 0.31767582517383786,
+      "learning_rate": 8.821981724367322e-06,
+      "loss": 0.5335,
+      "step": 3008
+    },
+    {
+      "epoch": 0.4941596699033112,
+      "grad_norm": 0.2613418127902927,
+      "learning_rate": 8.82185109885837e-06,
+      "loss": 0.5279,
+      "step": 3009
+    },
+    {
+      "epoch": 0.494323897111654,
+      "grad_norm": 0.3858067232180448,
+      "learning_rate": 8.821720426409862e-06,
+      "loss": 0.5577,
+      "step": 3010
+    },
+    {
+      "epoch": 0.4944881243199967,
+      "grad_norm": 0.34249742513275216,
+      "learning_rate": 8.821589707023218e-06,
+      "loss": 0.5404,
+      "step": 3011
+    },
+    {
+      "epoch": 0.4946523515283395,
+      "grad_norm": 0.3078017073081959,
+      "learning_rate": 8.821458940699858e-06,
+      "loss": 0.5441,
+      "step": 3012
+    },
+    {
+      "epoch": 0.4948165787366822,
+      "grad_norm": 0.4729428717813342,
+      "learning_rate": 8.821328127441202e-06,
+      "loss": 0.5871,
+      "step": 3013
+    },
+    {
+      "epoch": 0.4949808059450249,
+      "grad_norm": 0.344819112206208,
+      "learning_rate": 8.821197267248673e-06,
+      "loss": 0.5371,
+      "step": 3014
+    },
+    {
+      "epoch": 0.49514503315336766,
+      "grad_norm": 0.5182067984676952,
+      "learning_rate": 8.821066360123687e-06,
+      "loss": 0.554,
+      "step": 3015
+    },
+    {
+      "epoch": 0.4953092603617104,
+      "grad_norm": 0.5446474279763543,
+      "learning_rate": 8.820935406067672e-06,
+      "loss": 0.5833,
+      "step": 3016
+    },
+    {
+      "epoch": 0.49547348757005316,
+      "grad_norm": 0.2829439671030965,
+      "learning_rate": 8.820804405082045e-06,
+      "loss": 0.5415,
+      "step": 3017
+    },
+    {
+      "epoch": 0.4956377147783959,
+      "grad_norm": 0.3880941527869299,
+      "learning_rate": 8.820673357168232e-06,
+      "loss": 0.5579,
+      "step": 3018
+    },
+    {
+      "epoch": 0.49580194198673866,
+      "grad_norm": 0.29511019081794293,
+      "learning_rate": 8.820542262327655e-06,
+      "loss": 0.54,
+      "step": 3019
+    },
+    {
+      "epoch": 0.4959661691950814,
+      "grad_norm": 0.337603608717014,
+      "learning_rate": 8.820411120561738e-06,
+      "loss": 0.5414,
+      "step": 3020
+    },
+    {
+      "epoch": 0.49613039640342416,
+      "grad_norm": 0.36921638040684235,
+      "learning_rate": 8.820279931871906e-06,
+      "loss": 0.5572,
+      "step": 3021
+    },
+    {
+      "epoch": 0.4962946236117669,
+      "grad_norm": 0.34506745881486306,
+      "learning_rate": 8.820148696259584e-06,
+      "loss": 0.5594,
+      "step": 3022
+    },
+    {
+      "epoch": 0.4964588508201096,
+      "grad_norm": 0.35208611833465026,
+      "learning_rate": 8.820017413726196e-06,
+      "loss": 0.5487,
+      "step": 3023
+    },
+    {
+      "epoch": 0.49662307802845235,
+      "grad_norm": 0.3274229204168676,
+      "learning_rate": 8.819886084273168e-06,
+      "loss": 0.5649,
+      "step": 3024
+    },
+    {
+      "epoch": 0.4967873052367951,
+      "grad_norm": 0.3208518695214698,
+      "learning_rate": 8.819754707901928e-06,
+      "loss": 0.5461,
+      "step": 3025
+    },
+    {
+      "epoch": 0.49695153244513784,
+      "grad_norm": 0.3198716314130458,
+      "learning_rate": 8.819623284613901e-06,
+      "loss": 0.5608,
+      "step": 3026
+    },
+    {
+      "epoch": 0.4971157596534806,
+      "grad_norm": 0.4017507102525848,
+      "learning_rate": 8.819491814410516e-06,
+      "loss": 0.5156,
+      "step": 3027
+    },
+    {
+      "epoch": 0.49727998686182334,
+      "grad_norm": 0.3032437885755787,
+      "learning_rate": 8.819360297293199e-06,
+      "loss": 0.5362,
+      "step": 3028
+    },
+    {
+      "epoch": 0.4974442140701661,
+      "grad_norm": 0.3414840714435411,
+      "learning_rate": 8.81922873326338e-06,
+      "loss": 0.5451,
+      "step": 3029
+    },
+    {
+      "epoch": 0.49760844127850884,
+      "grad_norm": 0.35225110432503826,
+      "learning_rate": 8.819097122322488e-06,
+      "loss": 0.5209,
+      "step": 3030
+    },
+    {
+      "epoch": 0.49777266848685153,
+      "grad_norm": 0.3239762292518163,
+      "learning_rate": 8.81896546447195e-06,
+      "loss": 0.572,
+      "step": 3031
+    },
+    {
+      "epoch": 0.4979368956951943,
+      "grad_norm": 0.4545041404536738,
+      "learning_rate": 8.8188337597132e-06,
+      "loss": 0.5595,
+      "step": 3032
+    },
+    {
+      "epoch": 0.49810112290353703,
+      "grad_norm": 0.35087259800795967,
+      "learning_rate": 8.818702008047666e-06,
+      "loss": 0.5503,
+      "step": 3033
+    },
+    {
+      "epoch": 0.4982653501118798,
+      "grad_norm": 0.3518407108267486,
+      "learning_rate": 8.818570209476777e-06,
+      "loss": 0.5565,
+      "step": 3034
+    },
+    {
+      "epoch": 0.4984295773202225,
+      "grad_norm": 0.377200938642081,
+      "learning_rate": 8.81843836400197e-06,
+      "loss": 0.5663,
+      "step": 3035
+    },
+    {
+      "epoch": 0.4985938045285653,
+      "grad_norm": 0.33495384047527427,
+      "learning_rate": 8.818306471624672e-06,
+      "loss": 0.558,
+      "step": 3036
+    },
+    {
+      "epoch": 0.498758031736908,
+      "grad_norm": 0.3370327966629263,
+      "learning_rate": 8.818174532346315e-06,
+      "loss": 0.5431,
+      "step": 3037
+    },
+    {
+      "epoch": 0.4989222589452508,
+      "grad_norm": 0.31485961232651577,
+      "learning_rate": 8.818042546168336e-06,
+      "loss": 0.5452,
+      "step": 3038
+    },
+    {
+      "epoch": 0.4990864861535935,
+      "grad_norm": 1.0283110791160432,
+      "learning_rate": 8.817910513092168e-06,
+      "loss": 0.555,
+      "step": 3039
+    },
+    {
+      "epoch": 0.4992507133619362,
+      "grad_norm": 0.3370249745346664,
+      "learning_rate": 8.81777843311924e-06,
+      "loss": 0.5565,
+      "step": 3040
+    },
+    {
+      "epoch": 0.49941494057027896,
+      "grad_norm": 0.3697422926403468,
+      "learning_rate": 8.817646306250992e-06,
+      "loss": 0.5376,
+      "step": 3041
+    },
+    {
+      "epoch": 0.4995791677786217,
+      "grad_norm": 0.33821750551255797,
+      "learning_rate": 8.817514132488858e-06,
+      "loss": 0.5741,
+      "step": 3042
+    },
+    {
+      "epoch": 0.49974339498696446,
+      "grad_norm": 0.333623396144671,
+      "learning_rate": 8.817381911834272e-06,
+      "loss": 0.537,
+      "step": 3043
+    },
+    {
+      "epoch": 0.4999076221953072,
+      "grad_norm": 0.5552665670609216,
+      "learning_rate": 8.817249644288669e-06,
+      "loss": 0.5366,
+      "step": 3044
+    },
+    {
+      "epoch": 0.5000718494036499,
+      "grad_norm": 0.3361543209358446,
+      "learning_rate": 8.817117329853489e-06,
+      "loss": 0.5571,
+      "step": 3045
+    },
+    {
+      "epoch": 0.5002360766119927,
+      "grad_norm": 0.3305629109517608,
+      "learning_rate": 8.816984968530167e-06,
+      "loss": 0.5147,
+      "step": 3046
+    },
+    {
+      "epoch": 0.5004003038203354,
+      "grad_norm": 0.3732536451895665,
+      "learning_rate": 8.816852560320142e-06,
+      "loss": 0.5593,
+      "step": 3047
+    },
+    {
+      "epoch": 0.5005645310286782,
+      "grad_norm": 0.3253389659113194,
+      "learning_rate": 8.816720105224851e-06,
+      "loss": 0.5649,
+      "step": 3048
+    },
+    {
+      "epoch": 0.5007287582370209,
+      "grad_norm": 0.5642123979492453,
+      "learning_rate": 8.81658760324573e-06,
+      "loss": 0.5608,
+      "step": 3049
+    },
+    {
+      "epoch": 0.5008929854453636,
+      "grad_norm": 0.34823482199959965,
+      "learning_rate": 8.816455054384224e-06,
+      "loss": 0.5371,
+      "step": 3050
+    },
+    {
+      "epoch": 0.5010572126537064,
+      "grad_norm": 0.2841807878282881,
+      "learning_rate": 8.816322458641767e-06,
+      "loss": 0.577,
+      "step": 3051
+    },
+    {
+      "epoch": 0.5012214398620491,
+      "grad_norm": 0.42949070216811547,
+      "learning_rate": 8.816189816019802e-06,
+      "loss": 0.5596,
+      "step": 3052
+    },
+    {
+      "epoch": 0.5013856670703919,
+      "grad_norm": 0.37820954212665636,
+      "learning_rate": 8.816057126519769e-06,
+      "loss": 0.5555,
+      "step": 3053
+    },
+    {
+      "epoch": 0.5015498942787346,
+      "grad_norm": 0.33884019153259126,
+      "learning_rate": 8.815924390143108e-06,
+      "loss": 0.5125,
+      "step": 3054
+    },
+    {
+      "epoch": 0.5017141214870774,
+      "grad_norm": 0.3024625122784817,
+      "learning_rate": 8.815791606891265e-06,
+      "loss": 0.5431,
+      "step": 3055
+    },
+    {
+      "epoch": 0.5018783486954201,
+      "grad_norm": 0.31699884443526544,
+      "learning_rate": 8.815658776765675e-06,
+      "loss": 0.5392,
+      "step": 3056
+    },
+    {
+      "epoch": 0.5020425759037629,
+      "grad_norm": 0.3264450614548681,
+      "learning_rate": 8.815525899767788e-06,
+      "loss": 0.5455,
+      "step": 3057
+    },
+    {
+      "epoch": 0.5022068031121056,
+      "grad_norm": 0.3043521417126159,
+      "learning_rate": 8.815392975899042e-06,
+      "loss": 0.546,
+      "step": 3058
+    },
+    {
+      "epoch": 0.5023710303204484,
+      "grad_norm": 0.2830419102096985,
+      "learning_rate": 8.815260005160884e-06,
+      "loss": 0.5405,
+      "step": 3059
+    },
+    {
+      "epoch": 0.5025352575287911,
+      "grad_norm": 0.3197247856576631,
+      "learning_rate": 8.815126987554755e-06,
+      "loss": 0.5653,
+      "step": 3060
+    },
+    {
+      "epoch": 0.5026994847371338,
+      "grad_norm": 0.3388722907872103,
+      "learning_rate": 8.814993923082102e-06,
+      "loss": 0.5546,
+      "step": 3061
+    },
+    {
+      "epoch": 0.5028637119454765,
+      "grad_norm": 0.29425536996130774,
+      "learning_rate": 8.81486081174437e-06,
+      "loss": 0.5499,
+      "step": 3062
+    },
+    {
+      "epoch": 0.5030279391538193,
+      "grad_norm": 0.34880282593568984,
+      "learning_rate": 8.814727653543005e-06,
+      "loss": 0.5848,
+      "step": 3063
+    },
+    {
+      "epoch": 0.503192166362162,
+      "grad_norm": 0.3310747337461077,
+      "learning_rate": 8.814594448479452e-06,
+      "loss": 0.551,
+      "step": 3064
+    },
+    {
+      "epoch": 0.5033563935705048,
+      "grad_norm": 0.3895534185008042,
+      "learning_rate": 8.814461196555156e-06,
+      "loss": 0.5314,
+      "step": 3065
+    },
+    {
+      "epoch": 0.5035206207788475,
+      "grad_norm": 0.3101567026796809,
+      "learning_rate": 8.81432789777157e-06,
+      "loss": 0.5445,
+      "step": 3066
+    },
+    {
+      "epoch": 0.5036848479871903,
+      "grad_norm": 0.28256966020552643,
+      "learning_rate": 8.814194552130136e-06,
+      "loss": 0.5586,
+      "step": 3067
+    },
+    {
+      "epoch": 0.503849075195533,
+      "grad_norm": 0.29522749873378323,
+      "learning_rate": 8.814061159632306e-06,
+      "loss": 0.5617,
+      "step": 3068
+    },
+    {
+      "epoch": 0.5040133024038758,
+      "grad_norm": 0.35241166914489735,
+      "learning_rate": 8.813927720279526e-06,
+      "loss": 0.5617,
+      "step": 3069
+    },
+    {
+      "epoch": 0.5041775296122185,
+      "grad_norm": 0.3227727829669039,
+      "learning_rate": 8.813794234073247e-06,
+      "loss": 0.5513,
+      "step": 3070
+    },
+    {
+      "epoch": 0.5043417568205613,
+      "grad_norm": 0.35211764287996394,
+      "learning_rate": 8.813660701014918e-06,
+      "loss": 0.553,
+      "step": 3071
+    },
+    {
+      "epoch": 0.504505984028904,
+      "grad_norm": 0.2820677190638171,
+      "learning_rate": 8.813527121105991e-06,
+      "loss": 0.5366,
+      "step": 3072
+    },
+    {
+      "epoch": 0.5046702112372468,
+      "grad_norm": 0.3628674686715253,
+      "learning_rate": 8.813393494347915e-06,
+      "loss": 0.5504,
+      "step": 3073
+    },
+    {
+      "epoch": 0.5048344384455895,
+      "grad_norm": 0.3344707001677626,
+      "learning_rate": 8.813259820742143e-06,
+      "loss": 0.5411,
+      "step": 3074
+    },
+    {
+      "epoch": 0.5049986656539323,
+      "grad_norm": 0.33865193193614024,
+      "learning_rate": 8.813126100290124e-06,
+      "loss": 0.543,
+      "step": 3075
+    },
+    {
+      "epoch": 0.505162892862275,
+      "grad_norm": 0.34283850872753413,
+      "learning_rate": 8.812992332993312e-06,
+      "loss": 0.5304,
+      "step": 3076
+    },
+    {
+      "epoch": 0.5053271200706178,
+      "grad_norm": 0.3125953167598692,
+      "learning_rate": 8.81285851885316e-06,
+      "loss": 0.5779,
+      "step": 3077
+    },
+    {
+      "epoch": 0.5054913472789604,
+      "grad_norm": 0.3253589557042883,
+      "learning_rate": 8.812724657871124e-06,
+      "loss": 0.544,
+      "step": 3078
+    },
+    {
+      "epoch": 0.5056555744873031,
+      "grad_norm": 0.32809582790623654,
+      "learning_rate": 8.812590750048651e-06,
+      "loss": 0.5484,
+      "step": 3079
+    },
+    {
+      "epoch": 0.5058198016956459,
+      "grad_norm": 0.3000554576014139,
+      "learning_rate": 8.8124567953872e-06,
+      "loss": 0.5455,
+      "step": 3080
+    },
+    {
+      "epoch": 0.5059840289039886,
+      "grad_norm": 0.4201987809012978,
+      "learning_rate": 8.812322793888229e-06,
+      "loss": 0.542,
+      "step": 3081
+    },
+    {
+      "epoch": 0.5061482561123314,
+      "grad_norm": 0.3110305012514798,
+      "learning_rate": 8.812188745553186e-06,
+      "loss": 0.5443,
+      "step": 3082
+    },
+    {
+      "epoch": 0.5063124833206741,
+      "grad_norm": 0.31745465529222217,
+      "learning_rate": 8.812054650383533e-06,
+      "loss": 0.5374,
+      "step": 3083
+    },
+    {
+      "epoch": 0.5064767105290169,
+      "grad_norm": 0.28706529458952756,
+      "learning_rate": 8.811920508380722e-06,
+      "loss": 0.5796,
+      "step": 3084
+    },
+    {
+      "epoch": 0.5066409377373596,
+      "grad_norm": 0.31694659994111607,
+      "learning_rate": 8.811786319546213e-06,
+      "loss": 0.568,
+      "step": 3085
+    },
+    {
+      "epoch": 0.5068051649457024,
+      "grad_norm": 0.3311771175932345,
+      "learning_rate": 8.81165208388146e-06,
+      "loss": 0.541,
+      "step": 3086
+    },
+    {
+      "epoch": 0.5069693921540451,
+      "grad_norm": 0.2894350217712809,
+      "learning_rate": 8.811517801387926e-06,
+      "loss": 0.5511,
+      "step": 3087
+    },
+    {
+      "epoch": 0.5071336193623879,
+      "grad_norm": 0.3198518620985994,
+      "learning_rate": 8.811383472067066e-06,
+      "loss": 0.5487,
+      "step": 3088
+    },
+    {
+      "epoch": 0.5072978465707306,
+      "grad_norm": 0.2919190229017156,
+      "learning_rate": 8.811249095920339e-06,
+      "loss": 0.5515,
+      "step": 3089
+    },
+    {
+      "epoch": 0.5074620737790734,
+      "grad_norm": 0.3323556812127161,
+      "learning_rate": 8.811114672949207e-06,
+      "loss": 0.5363,
+      "step": 3090
+    },
+    {
+      "epoch": 0.5076263009874161,
+      "grad_norm": 0.28621504004096965,
+      "learning_rate": 8.810980203155126e-06,
+      "loss": 0.5152,
+      "step": 3091
+    },
+    {
+      "epoch": 0.5077905281957589,
+      "grad_norm": 0.30244379133479443,
+      "learning_rate": 8.81084568653956e-06,
+      "loss": 0.5592,
+      "step": 3092
+    },
+    {
+      "epoch": 0.5079547554041016,
+      "grad_norm": 0.32967015947935796,
+      "learning_rate": 8.810711123103967e-06,
+      "loss": 0.5378,
+      "step": 3093
+    },
+    {
+      "epoch": 0.5081189826124444,
+      "grad_norm": 0.32233218351029747,
+      "learning_rate": 8.810576512849812e-06,
+      "loss": 0.5701,
+      "step": 3094
+    },
+    {
+      "epoch": 0.508283209820787,
+      "grad_norm": 0.3038805312926043,
+      "learning_rate": 8.810441855778554e-06,
+      "loss": 0.5369,
+      "step": 3095
+    },
+    {
+      "epoch": 0.5084474370291298,
+      "grad_norm": 0.3675474725686056,
+      "learning_rate": 8.810307151891658e-06,
+      "loss": 0.5435,
+      "step": 3096
+    },
+    {
+      "epoch": 0.5086116642374725,
+      "grad_norm": 0.3031891362387634,
+      "learning_rate": 8.810172401190583e-06,
+      "loss": 0.553,
+      "step": 3097
+    },
+    {
+      "epoch": 0.5087758914458153,
+      "grad_norm": 0.3125909011850417,
+      "learning_rate": 8.810037603676797e-06,
+      "loss": 0.5349,
+      "step": 3098
+    },
+    {
+      "epoch": 0.508940118654158,
+      "grad_norm": 0.5022551060954056,
+      "learning_rate": 8.809902759351761e-06,
+      "loss": 0.5645,
+      "step": 3099
+    },
+    {
+      "epoch": 0.5091043458625008,
+      "grad_norm": 0.3055638918113333,
+      "learning_rate": 8.809767868216941e-06,
+      "loss": 0.548,
+      "step": 3100
+    },
+    {
+      "epoch": 0.5092685730708435,
+      "grad_norm": 0.2874713939919956,
+      "learning_rate": 8.809632930273801e-06,
+      "loss": 0.561,
+      "step": 3101
+    },
+    {
+      "epoch": 0.5094328002791862,
+      "grad_norm": 0.4250919250625114,
+      "learning_rate": 8.809497945523808e-06,
+      "loss": 0.5645,
+      "step": 3102
+    },
+    {
+      "epoch": 0.509597027487529,
+      "grad_norm": 0.35550987220631075,
+      "learning_rate": 8.809362913968428e-06,
+      "loss": 0.5403,
+      "step": 3103
+    },
+    {
+      "epoch": 0.5097612546958717,
+      "grad_norm": 0.3443953804934167,
+      "learning_rate": 8.809227835609127e-06,
+      "loss": 0.5413,
+      "step": 3104
+    },
+    {
+      "epoch": 0.5099254819042145,
+      "grad_norm": 0.2839313401806877,
+      "learning_rate": 8.80909271044737e-06,
+      "loss": 0.5491,
+      "step": 3105
+    },
+    {
+      "epoch": 0.5100897091125572,
+      "grad_norm": 0.2967153264099807,
+      "learning_rate": 8.808957538484629e-06,
+      "loss": 0.5488,
+      "step": 3106
+    },
+    {
+      "epoch": 0.5102539363209,
+      "grad_norm": 0.28848535562090943,
+      "learning_rate": 8.808822319722367e-06,
+      "loss": 0.5588,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5104181635292427,
+      "grad_norm": 0.2830657128227046,
+      "learning_rate": 8.808687054162057e-06,
+      "loss": 0.557,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5105823907375855,
+      "grad_norm": 0.39693335600192026,
+      "learning_rate": 8.808551741805167e-06,
+      "loss": 0.5537,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5107466179459282,
+      "grad_norm": 0.3161563688662857,
+      "learning_rate": 8.808416382653165e-06,
+      "loss": 0.5418,
+      "step": 3110
+    },
+    {
+      "epoch": 0.510910845154271,
+      "grad_norm": 0.299459534024022,
+      "learning_rate": 8.808280976707522e-06,
+      "loss": 0.5751,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5110750723626136,
+      "grad_norm": 0.3302741452682689,
+      "learning_rate": 8.80814552396971e-06,
+      "loss": 0.5722,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5112392995709564,
+      "grad_norm": 0.4430503083050463,
+      "learning_rate": 8.808010024441198e-06,
+      "loss": 0.5413,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5114035267792991,
+      "grad_norm": 0.32648783038577595,
+      "learning_rate": 8.80787447812346e-06,
+      "loss": 0.5405,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5115677539876419,
+      "grad_norm": 0.300303509573739,
+      "learning_rate": 8.807738885017965e-06,
+      "loss": 0.5455,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5117319811959846,
+      "grad_norm": 0.3694340171302308,
+      "learning_rate": 8.807603245126187e-06,
+      "loss": 0.545,
+      "step": 3116
+    },
+    {
+      "epoch": 0.5118962084043274,
+      "grad_norm": 0.3788879023858456,
+      "learning_rate": 8.807467558449603e-06,
+      "loss": 0.5359,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5120604356126701,
+      "grad_norm": 0.353354588272609,
+      "learning_rate": 8.80733182498968e-06,
+      "loss": 0.5488,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5122246628210129,
+      "grad_norm": 0.3408319876373525,
+      "learning_rate": 8.807196044747897e-06,
+      "loss": 0.5485,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5123888900293556,
+      "grad_norm": 0.39035712683504437,
+      "learning_rate": 8.807060217725726e-06,
+      "loss": 0.5464,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5125531172376984,
+      "grad_norm": 0.3533902186234627,
+      "learning_rate": 8.806924343924644e-06,
+      "loss": 0.5289,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5127173444460411,
+      "grad_norm": 0.5125086187864211,
+      "learning_rate": 8.806788423346127e-06,
+      "loss": 0.5434,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5128815716543839,
+      "grad_norm": 0.7203539322855015,
+      "learning_rate": 8.806652455991651e-06,
+      "loss": 0.5588,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5130457988627266,
+      "grad_norm": 0.36552585640486723,
+      "learning_rate": 8.80651644186269e-06,
+      "loss": 0.5523,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5132100260710694,
+      "grad_norm": 0.31504873528001304,
+      "learning_rate": 8.806380380960725e-06,
+      "loss": 0.5414,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5133742532794121,
+      "grad_norm": 0.3246787661246523,
+      "learning_rate": 8.806244273287233e-06,
+      "loss": 0.5418,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5135384804877549,
+      "grad_norm": 0.3715899215366805,
+      "learning_rate": 8.806108118843688e-06,
+      "loss": 0.5657,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5137027076960976,
+      "grad_norm": 0.4586182447833959,
+      "learning_rate": 8.805971917631575e-06,
+      "loss": 0.5338,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5138669349044402,
+      "grad_norm": 0.2998730656450566,
+      "learning_rate": 8.80583566965237e-06,
+      "loss": 0.567,
+      "step": 3129
+    },
+    {
+      "epoch": 0.514031162112783,
+      "grad_norm": 0.3358753869110285,
+      "learning_rate": 8.805699374907553e-06,
+      "loss": 0.5329,
+      "step": 3130
+    },
+    {
+      "epoch": 0.5141953893211257,
+      "grad_norm": 0.3623082881810261,
+      "learning_rate": 8.805563033398604e-06,
+      "loss": 0.5481,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5143596165294685,
+      "grad_norm": 0.41586552825851203,
+      "learning_rate": 8.805426645127005e-06,
+      "loss": 0.5347,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5145238437378112,
+      "grad_norm": 0.3389036505262228,
+      "learning_rate": 8.805290210094238e-06,
+      "loss": 0.5432,
+      "step": 3133
+    },
+    {
+      "epoch": 0.514688070946154,
+      "grad_norm": 0.33220166906284515,
+      "learning_rate": 8.80515372830178e-06,
+      "loss": 0.5737,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5148522981544967,
+      "grad_norm": 1.4240426803228405,
+      "learning_rate": 8.80501719975112e-06,
+      "loss": 0.563,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5150165253628395,
+      "grad_norm": 0.34283390659735813,
+      "learning_rate": 8.804880624443737e-06,
+      "loss": 0.5487,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5151807525711822,
+      "grad_norm": 0.3625409825436214,
+      "learning_rate": 8.804744002381114e-06,
+      "loss": 0.558,
+      "step": 3137
+    },
+    {
+      "epoch": 0.515344979779525,
+      "grad_norm": 0.39081632339485356,
+      "learning_rate": 8.804607333564737e-06,
+      "loss": 0.5447,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5155092069878677,
+      "grad_norm": 0.3236707823666088,
+      "learning_rate": 8.804470617996088e-06,
+      "loss": 0.5647,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5156734341962105,
+      "grad_norm": 0.42790371855431897,
+      "learning_rate": 8.804333855676653e-06,
+      "loss": 0.5436,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5158376614045532,
+      "grad_norm": 0.3510945198863832,
+      "learning_rate": 8.804197046607918e-06,
+      "loss": 0.595,
+      "step": 3141
+    },
+    {
+      "epoch": 0.516001888612896,
+      "grad_norm": 0.34592814315247516,
+      "learning_rate": 8.80406019079137e-06,
+      "loss": 0.5549,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5161661158212387,
+      "grad_norm": 0.3300648787481515,
+      "learning_rate": 8.803923288228492e-06,
+      "loss": 0.5512,
+      "step": 3143
+    },
+    {
+      "epoch": 0.5163303430295815,
+      "grad_norm": 0.3367035380808404,
+      "learning_rate": 8.803786338920773e-06,
+      "loss": 0.5349,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5164945702379242,
+      "grad_norm": 0.40066400824553233,
+      "learning_rate": 8.803649342869698e-06,
+      "loss": 0.5556,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5166587974462669,
+      "grad_norm": 0.311816919024996,
+      "learning_rate": 8.803512300076759e-06,
+      "loss": 0.5589,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5168230246546096,
+      "grad_norm": 0.31924308074256,
+      "learning_rate": 8.803375210543442e-06,
+      "loss": 0.5502,
+      "step": 3147
+    },
+    {
+      "epoch": 0.5169872518629524,
+      "grad_norm": 0.3624853958484826,
+      "learning_rate": 8.803238074271237e-06,
+      "loss": 0.5562,
+      "step": 3148
+    },
+    {
+      "epoch": 0.5171514790712951,
+      "grad_norm": 0.3150090211544387,
+      "learning_rate": 8.803100891261632e-06,
+      "loss": 0.534,
+      "step": 3149
+    },
+    {
+      "epoch": 0.5173157062796379,
+      "grad_norm": 0.30714145840950313,
+      "learning_rate": 8.802963661516117e-06,
+      "loss": 0.5609,
+      "step": 3150
+    },
+    {
+      "epoch": 0.5174799334879806,
+      "grad_norm": 0.41808965641054635,
+      "learning_rate": 8.802826385036183e-06,
+      "loss": 0.5277,
+      "step": 3151
+    },
+    {
+      "epoch": 0.5176441606963234,
+      "grad_norm": 0.33805758218512644,
+      "learning_rate": 8.802689061823322e-06,
+      "loss": 0.5744,
+      "step": 3152
+    },
+    {
+      "epoch": 0.5178083879046661,
+      "grad_norm": 0.32757272518208763,
+      "learning_rate": 8.802551691879024e-06,
+      "loss": 0.5434,
+      "step": 3153
+    },
+    {
+      "epoch": 0.5179726151130088,
+      "grad_norm": 0.30549027755496716,
+      "learning_rate": 8.802414275204783e-06,
+      "loss": 0.5388,
+      "step": 3154
+    },
+    {
+      "epoch": 0.5181368423213516,
+      "grad_norm": 0.5121294354664127,
+      "learning_rate": 8.802276811802089e-06,
+      "loss": 0.5559,
+      "step": 3155
+    },
+    {
+      "epoch": 0.5183010695296943,
+      "grad_norm": 0.4360850175903607,
+      "learning_rate": 8.802139301672434e-06,
+      "loss": 0.5508,
+      "step": 3156
+    },
+    {
+      "epoch": 0.5184652967380371,
+      "grad_norm": 0.3109791520360334,
+      "learning_rate": 8.802001744817315e-06,
+      "loss": 0.5582,
+      "step": 3157
+    },
+    {
+      "epoch": 0.5186295239463798,
+      "grad_norm": 0.34541983264627407,
+      "learning_rate": 8.801864141238225e-06,
+      "loss": 0.5425,
+      "step": 3158
+    },
+    {
+      "epoch": 0.5187937511547226,
+      "grad_norm": 0.44077911149552174,
+      "learning_rate": 8.801726490936658e-06,
+      "loss": 0.5333,
+      "step": 3159
+    },
+    {
+      "epoch": 0.5189579783630653,
+      "grad_norm": 0.3175152055000214,
+      "learning_rate": 8.801588793914108e-06,
+      "loss": 0.5347,
+      "step": 3160
+    },
+    {
+      "epoch": 0.5191222055714081,
+      "grad_norm": 0.33110520939463356,
+      "learning_rate": 8.801451050172072e-06,
+      "loss": 0.5367,
+      "step": 3161
+    },
+    {
+      "epoch": 0.5192864327797508,
+      "grad_norm": 0.6786090232297229,
+      "learning_rate": 8.801313259712045e-06,
+      "loss": 0.5379,
+      "step": 3162
+    },
+    {
+      "epoch": 0.5194506599880935,
+      "grad_norm": 0.32345532898552576,
+      "learning_rate": 8.801175422535524e-06,
+      "loss": 0.5689,
+      "step": 3163
+    },
+    {
+      "epoch": 0.5196148871964362,
+      "grad_norm": 0.29288324836950214,
+      "learning_rate": 8.801037538644008e-06,
+      "loss": 0.552,
+      "step": 3164
+    },
+    {
+      "epoch": 0.519779114404779,
+      "grad_norm": 0.3060045843978916,
+      "learning_rate": 8.80089960803899e-06,
+      "loss": 0.549,
+      "step": 3165
+    },
+    {
+      "epoch": 0.5199433416131217,
+      "grad_norm": 0.3997067399866202,
+      "learning_rate": 8.800761630721973e-06,
+      "loss": 0.5322,
+      "step": 3166
+    },
+    {
+      "epoch": 0.5201075688214645,
+      "grad_norm": 0.2952695042762905,
+      "learning_rate": 8.800623606694453e-06,
+      "loss": 0.5631,
+      "step": 3167
+    },
+    {
+      "epoch": 0.5202717960298072,
+      "grad_norm": 0.33202602894268374,
+      "learning_rate": 8.800485535957928e-06,
+      "loss": 0.5355,
+      "step": 3168
+    },
+    {
+      "epoch": 0.52043602323815,
+      "grad_norm": 0.3053471478926584,
+      "learning_rate": 8.8003474185139e-06,
+      "loss": 0.5239,
+      "step": 3169
+    },
+    {
+      "epoch": 0.5206002504464927,
+      "grad_norm": 0.44546827302209424,
+      "learning_rate": 8.80020925436387e-06,
+      "loss": 0.5446,
+      "step": 3170
+    },
+    {
+      "epoch": 0.5207644776548355,
+      "grad_norm": 0.3232464754044946,
+      "learning_rate": 8.800071043509333e-06,
+      "loss": 0.5625,
+      "step": 3171
+    },
+    {
+      "epoch": 0.5209287048631782,
+      "grad_norm": 0.31924534819311456,
+      "learning_rate": 8.799932785951797e-06,
+      "loss": 0.5548,
+      "step": 3172
+    },
+    {
+      "epoch": 0.521092932071521,
+      "grad_norm": 0.8093812270735015,
+      "learning_rate": 8.799794481692757e-06,
+      "loss": 0.5512,
+      "step": 3173
+    },
+    {
+      "epoch": 0.5212571592798637,
+      "grad_norm": 0.353912459104798,
+      "learning_rate": 8.79965613073372e-06,
+      "loss": 0.5355,
+      "step": 3174
+    },
+    {
+      "epoch": 0.5214213864882065,
+      "grad_norm": 17.314905034002273,
+      "learning_rate": 8.799517733076186e-06,
+      "loss": 0.5549,
+      "step": 3175
+    },
+    {
+      "epoch": 0.5215856136965492,
+      "grad_norm": 0.4253045024687482,
+      "learning_rate": 8.799379288721663e-06,
+      "loss": 0.5465,
+      "step": 3176
+    },
+    {
+      "epoch": 0.521749840904892,
+      "grad_norm": 0.5363599916050024,
+      "learning_rate": 8.799240797671648e-06,
+      "loss": 0.5398,
+      "step": 3177
+    },
+    {
+      "epoch": 0.5219140681132347,
+      "grad_norm": 0.5429239713433921,
+      "learning_rate": 8.799102259927648e-06,
+      "loss": 0.5547,
+      "step": 3178
+    },
+    {
+      "epoch": 0.5220782953215775,
+      "grad_norm": 0.888409605205552,
+      "learning_rate": 8.798963675491168e-06,
+      "loss": 0.569,
+      "step": 3179
+    },
+    {
+      "epoch": 0.5222425225299201,
+      "grad_norm": 0.8292470249496997,
+      "learning_rate": 8.798825044363714e-06,
+      "loss": 0.5491,
+      "step": 3180
+    },
+    {
+      "epoch": 0.5224067497382628,
+      "grad_norm": 0.8524348696719072,
+      "learning_rate": 8.79868636654679e-06,
+      "loss": 0.5565,
+      "step": 3181
+    },
+    {
+      "epoch": 0.5225709769466056,
+      "grad_norm": 0.7638273678405479,
+      "learning_rate": 8.798547642041903e-06,
+      "loss": 0.5655,
+      "step": 3182
+    },
+    {
+      "epoch": 0.5227352041549483,
+      "grad_norm": 0.5794743123076325,
+      "learning_rate": 8.798408870850557e-06,
+      "loss": 0.5791,
+      "step": 3183
+    },
+    {
+      "epoch": 0.5228994313632911,
+      "grad_norm": 0.6652418445715474,
+      "learning_rate": 8.798270052974265e-06,
+      "loss": 0.5817,
+      "step": 3184
+    },
+    {
+      "epoch": 0.5230636585716338,
+      "grad_norm": 0.6033023612512277,
+      "learning_rate": 8.79813118841453e-06,
+      "loss": 0.5602,
+      "step": 3185
+    },
+    {
+      "epoch": 0.5232278857799766,
+      "grad_norm": 0.7713085937031505,
+      "learning_rate": 8.79799227717286e-06,
+      "loss": 0.5467,
+      "step": 3186
+    },
+    {
+      "epoch": 0.5233921129883193,
+      "grad_norm": 0.5954572400751407,
+      "learning_rate": 8.797853319250767e-06,
+      "loss": 0.5632,
+      "step": 3187
+    },
+    {
+      "epoch": 0.5235563401966621,
+      "grad_norm": 0.47584740248527363,
+      "learning_rate": 8.797714314649757e-06,
+      "loss": 0.5594,
+      "step": 3188
+    },
+    {
+      "epoch": 0.5237205674050048,
+      "grad_norm": 0.5639314104879384,
+      "learning_rate": 8.797575263371343e-06,
+      "loss": 0.5555,
+      "step": 3189
+    },
+    {
+      "epoch": 0.5238847946133476,
+      "grad_norm": 0.5285776579354926,
+      "learning_rate": 8.797436165417032e-06,
+      "loss": 0.5671,
+      "step": 3190
+    },
+    {
+      "epoch": 0.5240490218216903,
+      "grad_norm": 0.43872512757835813,
+      "learning_rate": 8.797297020788336e-06,
+      "loss": 0.5555,
+      "step": 3191
+    },
+    {
+      "epoch": 0.5242132490300331,
+      "grad_norm": 0.5753312286290805,
+      "learning_rate": 8.797157829486767e-06,
+      "loss": 0.5585,
+      "step": 3192
+    },
+    {
+      "epoch": 0.5243774762383758,
+      "grad_norm": 0.38476151861139246,
+      "learning_rate": 8.797018591513837e-06,
+      "loss": 0.5576,
+      "step": 3193
+    },
+    {
+      "epoch": 0.5245417034467186,
+      "grad_norm": 0.43061245503098006,
+      "learning_rate": 8.796879306871056e-06,
+      "loss": 0.58,
+      "step": 3194
+    },
+    {
+      "epoch": 0.5247059306550613,
+      "grad_norm": 0.3378612386322587,
+      "learning_rate": 8.79673997555994e-06,
+      "loss": 0.552,
+      "step": 3195
+    },
+    {
+      "epoch": 0.5248701578634041,
+      "grad_norm": 0.3335234350164391,
+      "learning_rate": 8.796600597581998e-06,
+      "loss": 0.5599,
+      "step": 3196
+    },
+    {
+      "epoch": 0.5250343850717467,
+      "grad_norm": 0.5837684658025116,
+      "learning_rate": 8.796461172938749e-06,
+      "loss": 0.564,
+      "step": 3197
+    },
+    {
+      "epoch": 0.5251986122800895,
+      "grad_norm": 0.34310523436981255,
+      "learning_rate": 8.796321701631702e-06,
+      "loss": 0.541,
+      "step": 3198
+    },
+    {
+      "epoch": 0.5253628394884322,
+      "grad_norm": 0.44627037677002795,
+      "learning_rate": 8.796182183662376e-06,
+      "loss": 0.5639,
+      "step": 3199
+    },
+    {
+      "epoch": 0.525527066696775,
+      "grad_norm": 0.3762255974042171,
+      "learning_rate": 8.796042619032283e-06,
+      "loss": 0.5529,
+      "step": 3200
+    },
+    {
+      "epoch": 0.5256912939051177,
+      "grad_norm": 0.3437555157560286,
+      "learning_rate": 8.795903007742941e-06,
+      "loss": 0.5632,
+      "step": 3201
+    },
+    {
+      "epoch": 0.5258555211134605,
+      "grad_norm": 0.3317101998553545,
+      "learning_rate": 8.795763349795866e-06,
+      "loss": 0.5524,
+      "step": 3202
+    },
+    {
+      "epoch": 0.5260197483218032,
+      "grad_norm": 0.3979361730906255,
+      "learning_rate": 8.795623645192574e-06,
+      "loss": 0.5464,
+      "step": 3203
+    },
+    {
+      "epoch": 0.526183975530146,
+      "grad_norm": 0.3619951010888624,
+      "learning_rate": 8.795483893934584e-06,
+      "loss": 0.5404,
+      "step": 3204
+    },
+    {
+      "epoch": 0.5263482027384887,
+      "grad_norm": 0.3117407096018107,
+      "learning_rate": 8.795344096023411e-06,
+      "loss": 0.5376,
+      "step": 3205
+    },
+    {
+      "epoch": 0.5265124299468315,
+      "grad_norm": 0.3293924524709304,
+      "learning_rate": 8.795204251460576e-06,
+      "loss": 0.5306,
+      "step": 3206
+    },
+    {
+      "epoch": 0.5266766571551742,
+      "grad_norm": 0.41805058863513644,
+      "learning_rate": 8.795064360247598e-06,
+      "loss": 0.5646,
+      "step": 3207
+    },
+    {
+      "epoch": 0.526840884363517,
+      "grad_norm": 0.3826750759835131,
+      "learning_rate": 8.794924422385995e-06,
+      "loss": 0.5386,
+      "step": 3208
+    },
+    {
+      "epoch": 0.5270051115718597,
+      "grad_norm": 0.3430815145020039,
+      "learning_rate": 8.794784437877286e-06,
+      "loss": 0.5521,
+      "step": 3209
+    },
+    {
+      "epoch": 0.5271693387802024,
+      "grad_norm": 0.5506840435783226,
+      "learning_rate": 8.794644406722993e-06,
+      "loss": 0.5466,
+      "step": 3210
+    },
+    {
+      "epoch": 0.5273335659885452,
+      "grad_norm": 0.2992384367198502,
+      "learning_rate": 8.794504328924636e-06,
+      "loss": 0.5506,
+      "step": 3211
+    },
+    {
+      "epoch": 0.5274977931968879,
+      "grad_norm": 0.3029752344219687,
+      "learning_rate": 8.794364204483736e-06,
+      "loss": 0.5484,
+      "step": 3212
+    },
+    {
+      "epoch": 0.5276620204052307,
+      "grad_norm": 0.2948964578526369,
+      "learning_rate": 8.794224033401818e-06,
+      "loss": 0.5462,
+      "step": 3213
+    },
+    {
+      "epoch": 0.5278262476135733,
+      "grad_norm": 0.366601371536449,
+      "learning_rate": 8.794083815680402e-06,
+      "loss": 0.5599,
+      "step": 3214
+    },
+    {
+      "epoch": 0.5279904748219161,
+      "grad_norm": 0.3498632389083059,
+      "learning_rate": 8.79394355132101e-06,
+      "loss": 0.5383,
+      "step": 3215
+    },
+    {
+      "epoch": 0.5281547020302588,
+      "grad_norm": 0.31505091855348366,
+      "learning_rate": 8.793803240325165e-06,
+      "loss": 0.5631,
+      "step": 3216
+    },
+    {
+      "epoch": 0.5283189292386016,
+      "grad_norm": 0.42965003549284014,
+      "learning_rate": 8.793662882694394e-06,
+      "loss": 0.5419,
+      "step": 3217
+    },
+    {
+      "epoch": 0.5284831564469443,
+      "grad_norm": 0.29650964340269054,
+      "learning_rate": 8.79352247843022e-06,
+      "loss": 0.5551,
+      "step": 3218
+    },
+    {
+      "epoch": 0.5286473836552871,
+      "grad_norm": 0.29571046480753704,
+      "learning_rate": 8.793382027534167e-06,
+      "loss": 0.5402,
+      "step": 3219
+    },
+    {
+      "epoch": 0.5288116108636298,
+      "grad_norm": 0.3534211821619501,
+      "learning_rate": 8.79324153000776e-06,
+      "loss": 0.56,
+      "step": 3220
+    },
+    {
+      "epoch": 0.5289758380719726,
+      "grad_norm": 0.3105684543180674,
+      "learning_rate": 8.793100985852527e-06,
+      "loss": 0.5344,
+      "step": 3221
+    },
+    {
+      "epoch": 0.5291400652803153,
+      "grad_norm": 0.32396286394780704,
+      "learning_rate": 8.792960395069993e-06,
+      "loss": 0.5344,
+      "step": 3222
+    },
+    {
+      "epoch": 0.5293042924886581,
+      "grad_norm": 0.40299790637446253,
+      "learning_rate": 8.792819757661686e-06,
+      "loss": 0.5337,
+      "step": 3223
+    },
+    {
+      "epoch": 0.5294685196970008,
+      "grad_norm": 0.2948307048801837,
+      "learning_rate": 8.792679073629132e-06,
+      "loss": 0.5372,
+      "step": 3224
+    },
+    {
+      "epoch": 0.5296327469053436,
+      "grad_norm": 0.33790169922370855,
+      "learning_rate": 8.792538342973862e-06,
+      "loss": 0.5311,
+      "step": 3225
+    },
+    {
+      "epoch": 0.5297969741136863,
+      "grad_norm": 0.30645198851218913,
+      "learning_rate": 8.792397565697399e-06,
+      "loss": 0.543,
+      "step": 3226
+    },
+    {
+      "epoch": 0.5299612013220291,
+      "grad_norm": 0.3628053695130395,
+      "learning_rate": 8.792256741801277e-06,
+      "loss": 0.5559,
+      "step": 3227
+    },
+    {
+      "epoch": 0.5301254285303718,
+      "grad_norm": 0.27857795489535925,
+      "learning_rate": 8.792115871287025e-06,
+      "loss": 0.53,
+      "step": 3228
+    },
+    {
+      "epoch": 0.5302896557387146,
+      "grad_norm": 0.3071228087006437,
+      "learning_rate": 8.79197495415617e-06,
+      "loss": 0.5541,
+      "step": 3229
+    },
+    {
+      "epoch": 0.5304538829470573,
+      "grad_norm": 0.3325204943756406,
+      "learning_rate": 8.791833990410246e-06,
+      "loss": 0.5422,
+      "step": 3230
+    },
+    {
+      "epoch": 0.5306181101554,
+      "grad_norm": 0.3669488938803181,
+      "learning_rate": 8.79169298005078e-06,
+      "loss": 0.5363,
+      "step": 3231
+    },
+    {
+      "epoch": 0.5307823373637427,
+      "grad_norm": 0.27794447125357075,
+      "learning_rate": 8.791551923079308e-06,
+      "loss": 0.5678,
+      "step": 3232
+    },
+    {
+      "epoch": 0.5309465645720854,
+      "grad_norm": 0.3167738475291457,
+      "learning_rate": 8.791410819497359e-06,
+      "loss": 0.54,
+      "step": 3233
+    },
+    {
+      "epoch": 0.5311107917804282,
+      "grad_norm": 0.39694196446332014,
+      "learning_rate": 8.791269669306465e-06,
+      "loss": 0.5468,
+      "step": 3234
+    },
+    {
+      "epoch": 0.5312750189887709,
+      "grad_norm": 0.32582945645680483,
+      "learning_rate": 8.791128472508163e-06,
+      "loss": 0.5498,
+      "step": 3235
+    },
+    {
+      "epoch": 0.5314392461971137,
+      "grad_norm": 0.30304838745817103,
+      "learning_rate": 8.790987229103981e-06,
+      "loss": 0.5399,
+      "step": 3236
+    },
+    {
+      "epoch": 0.5316034734054564,
+      "grad_norm": 0.3316043075793242,
+      "learning_rate": 8.790845939095456e-06,
+      "loss": 0.5433,
+      "step": 3237
+    },
+    {
+      "epoch": 0.5317677006137992,
+      "grad_norm": 0.3544344684280817,
+      "learning_rate": 8.790704602484125e-06,
+      "loss": 0.5629,
+      "step": 3238
+    },
+    {
+      "epoch": 0.5319319278221419,
+      "grad_norm": 0.2916093235123261,
+      "learning_rate": 8.790563219271518e-06,
+      "loss": 0.5545,
+      "step": 3239
+    },
+    {
+      "epoch": 0.5320961550304847,
+      "grad_norm": 0.45549622656485855,
+      "learning_rate": 8.790421789459175e-06,
+      "loss": 0.5399,
+      "step": 3240
+    },
+    {
+      "epoch": 0.5322603822388274,
+      "grad_norm": 0.33583247440668695,
+      "learning_rate": 8.79028031304863e-06,
+      "loss": 0.5452,
+      "step": 3241
+    },
+    {
+      "epoch": 0.5324246094471702,
+      "grad_norm": 0.30414857412185614,
+      "learning_rate": 8.79013879004142e-06,
+      "loss": 0.5533,
+      "step": 3242
+    },
+    {
+      "epoch": 0.5325888366555129,
+      "grad_norm": 0.3661648266690582,
+      "learning_rate": 8.78999722043908e-06,
+      "loss": 0.5316,
+      "step": 3243
+    },
+    {
+      "epoch": 0.5327530638638557,
+      "grad_norm": 0.2943779954560458,
+      "learning_rate": 8.78985560424315e-06,
+      "loss": 0.5439,
+      "step": 3244
+    },
+    {
+      "epoch": 0.5329172910721984,
+      "grad_norm": 0.6613520472875823,
+      "learning_rate": 8.789713941455168e-06,
+      "loss": 0.5534,
+      "step": 3245
+    },
+    {
+      "epoch": 0.5330815182805412,
+      "grad_norm": 0.37859318983449436,
+      "learning_rate": 8.789572232076671e-06,
+      "loss": 0.5552,
+      "step": 3246
+    },
+    {
+      "epoch": 0.5332457454888838,
+      "grad_norm": 0.32920543399704716,
+      "learning_rate": 8.789430476109201e-06,
+      "loss": 0.5392,
+      "step": 3247
+    },
+    {
+      "epoch": 0.5334099726972266,
+      "grad_norm": 0.36207175639388633,
+      "learning_rate": 8.789288673554296e-06,
+      "loss": 0.5584,
+      "step": 3248
+    },
+    {
+      "epoch": 0.5335741999055693,
+      "grad_norm": 0.3265540316957601,
+      "learning_rate": 8.789146824413494e-06,
+      "loss": 0.5497,
+      "step": 3249
+    },
+    {
+      "epoch": 0.5337384271139121,
+      "grad_norm": 0.7945446107991943,
+      "learning_rate": 8.789004928688339e-06,
+      "loss": 0.5217,
+      "step": 3250
+    },
+    {
+      "epoch": 0.5339026543222548,
+      "grad_norm": 0.3216038639309901,
+      "learning_rate": 8.78886298638037e-06,
+      "loss": 0.5409,
+      "step": 3251
+    },
+    {
+      "epoch": 0.5340668815305976,
+      "grad_norm": 0.29287927055837376,
+      "learning_rate": 8.78872099749113e-06,
+      "loss": 0.5481,
+      "step": 3252
+    },
+    {
+      "epoch": 0.5342311087389403,
+      "grad_norm": 0.29317563619901427,
+      "learning_rate": 8.78857896202216e-06,
+      "loss": 0.548,
+      "step": 3253
+    },
+    {
+      "epoch": 0.5343953359472831,
+      "grad_norm": 0.739804214926253,
+      "learning_rate": 8.788436879975003e-06,
+      "loss": 0.5477,
+      "step": 3254
+    },
+    {
+      "epoch": 0.5345595631556258,
+      "grad_norm": 0.3462570710581254,
+      "learning_rate": 8.788294751351201e-06,
+      "loss": 0.5446,
+      "step": 3255
+    },
+    {
+      "epoch": 0.5347237903639686,
+      "grad_norm": 0.35379963473107207,
+      "learning_rate": 8.7881525761523e-06,
+      "loss": 0.5492,
+      "step": 3256
+    },
+    {
+      "epoch": 0.5348880175723113,
+      "grad_norm": 0.6842229018746369,
+      "learning_rate": 8.788010354379844e-06,
+      "loss": 0.5425,
+      "step": 3257
+    },
+    {
+      "epoch": 0.535052244780654,
+      "grad_norm": 0.3078870199163517,
+      "learning_rate": 8.787868086035374e-06,
+      "loss": 0.5589,
+      "step": 3258
+    },
+    {
+      "epoch": 0.5352164719889968,
+      "grad_norm": 0.3013589290107262,
+      "learning_rate": 8.78772577112044e-06,
+      "loss": 0.5458,
+      "step": 3259
+    },
+    {
+      "epoch": 0.5353806991973395,
+      "grad_norm": 0.29817945458234185,
+      "learning_rate": 8.787583409636587e-06,
+      "loss": 0.5515,
+      "step": 3260
+    },
+    {
+      "epoch": 0.5355449264056823,
+      "grad_norm": 0.5011852342901882,
+      "learning_rate": 8.787441001585356e-06,
+      "loss": 0.5421,
+      "step": 3261
+    },
+    {
+      "epoch": 0.535709153614025,
+      "grad_norm": 0.3648740653731427,
+      "learning_rate": 8.787298546968301e-06,
+      "loss": 0.5653,
+      "step": 3262
+    },
+    {
+      "epoch": 0.5358733808223678,
+      "grad_norm": 0.29821976509943804,
+      "learning_rate": 8.787156045786963e-06,
+      "loss": 0.5378,
+      "step": 3263
+    },
+    {
+      "epoch": 0.5360376080307104,
+      "grad_norm": 0.3126612686731939,
+      "learning_rate": 8.787013498042896e-06,
+      "loss": 0.5413,
+      "step": 3264
+    },
+    {
+      "epoch": 0.5362018352390532,
+      "grad_norm": 0.4119496533178666,
+      "learning_rate": 8.786870903737641e-06,
+      "loss": 0.5681,
+      "step": 3265
+    },
+    {
+      "epoch": 0.5363660624473959,
+      "grad_norm": 0.3168742881680851,
+      "learning_rate": 8.78672826287275e-06,
+      "loss": 0.5539,
+      "step": 3266
+    },
+    {
+      "epoch": 0.5365302896557387,
+      "grad_norm": 0.31014627546022744,
+      "learning_rate": 8.786585575449775e-06,
+      "loss": 0.5537,
+      "step": 3267
+    },
+    {
+      "epoch": 0.5366945168640814,
+      "grad_norm": 0.4724267454615436,
+      "learning_rate": 8.786442841470261e-06,
+      "loss": 0.55,
+      "step": 3268
+    },
+    {
+      "epoch": 0.5368587440724242,
+      "grad_norm": 0.33371309506311503,
+      "learning_rate": 8.786300060935761e-06,
+      "loss": 0.5482,
+      "step": 3269
+    },
+    {
+      "epoch": 0.5370229712807669,
+      "grad_norm": 0.3795166104581432,
+      "learning_rate": 8.786157233847827e-06,
+      "loss": 0.5255,
+      "step": 3270
+    },
+    {
+      "epoch": 0.5371871984891097,
+      "grad_norm": 0.4632986698184538,
+      "learning_rate": 8.786014360208008e-06,
+      "loss": 0.5316,
+      "step": 3271
+    },
+    {
+      "epoch": 0.5373514256974524,
+      "grad_norm": 0.3120794912916995,
+      "learning_rate": 8.785871440017854e-06,
+      "loss": 0.5272,
+      "step": 3272
+    },
+    {
+      "epoch": 0.5375156529057952,
+      "grad_norm": 0.3266812219370469,
+      "learning_rate": 8.785728473278922e-06,
+      "loss": 0.5395,
+      "step": 3273
+    },
+    {
+      "epoch": 0.5376798801141379,
+      "grad_norm": 0.3667428118373131,
+      "learning_rate": 8.78558545999276e-06,
+      "loss": 0.5372,
+      "step": 3274
+    },
+    {
+      "epoch": 0.5378441073224807,
+      "grad_norm": 0.3770025621124436,
+      "learning_rate": 8.785442400160925e-06,
+      "loss": 0.5321,
+      "step": 3275
+    },
+    {
+      "epoch": 0.5380083345308234,
+      "grad_norm": 0.3108514386539361,
+      "learning_rate": 8.785299293784968e-06,
+      "loss": 0.581,
+      "step": 3276
+    },
+    {
+      "epoch": 0.5381725617391662,
+      "grad_norm": 0.32926200102815384,
+      "learning_rate": 8.785156140866444e-06,
+      "loss": 0.5551,
+      "step": 3277
+    },
+    {
+      "epoch": 0.5383367889475089,
+      "grad_norm": 0.29382084808673803,
+      "learning_rate": 8.785012941406911e-06,
+      "loss": 0.5505,
+      "step": 3278
+    },
+    {
+      "epoch": 0.5385010161558517,
+      "grad_norm": 0.296915713846914,
+      "learning_rate": 8.78486969540792e-06,
+      "loss": 0.5403,
+      "step": 3279
+    },
+    {
+      "epoch": 0.5386652433641944,
+      "grad_norm": 0.3364779005663321,
+      "learning_rate": 8.784726402871028e-06,
+      "loss": 0.5507,
+      "step": 3280
+    },
+    {
+      "epoch": 0.538829470572537,
+      "grad_norm": 0.31712966506772305,
+      "learning_rate": 8.78458306379779e-06,
+      "loss": 0.5481,
+      "step": 3281
+    },
+    {
+      "epoch": 0.5389936977808798,
+      "grad_norm": 0.4194875222865034,
+      "learning_rate": 8.784439678189769e-06,
+      "loss": 0.5419,
+      "step": 3282
+    },
+    {
+      "epoch": 0.5391579249892225,
+      "grad_norm": 0.36369565463198017,
+      "learning_rate": 8.784296246048515e-06,
+      "loss": 0.538,
+      "step": 3283
+    },
+    {
+      "epoch": 0.5393221521975653,
+      "grad_norm": 0.33327635860855287,
+      "learning_rate": 8.784152767375589e-06,
+      "loss": 0.5786,
+      "step": 3284
+    },
+    {
+      "epoch": 0.539486379405908,
+      "grad_norm": 0.31341028441266205,
+      "learning_rate": 8.784009242172548e-06,
+      "loss": 0.5383,
+      "step": 3285
+    },
+    {
+      "epoch": 0.5396506066142508,
+      "grad_norm": 0.363543182076021,
+      "learning_rate": 8.783865670440954e-06,
+      "loss": 0.524,
+      "step": 3286
+    },
+    {
+      "epoch": 0.5398148338225935,
+      "grad_norm": 0.3023715111443661,
+      "learning_rate": 8.783722052182361e-06,
+      "loss": 0.54,
+      "step": 3287
+    },
+    {
+      "epoch": 0.5399790610309363,
+      "grad_norm": 0.3172844189567488,
+      "learning_rate": 8.783578387398333e-06,
+      "loss": 0.5319,
+      "step": 3288
+    },
+    {
+      "epoch": 0.540143288239279,
+      "grad_norm": 0.315659171219246,
+      "learning_rate": 8.78343467609043e-06,
+      "loss": 0.5714,
+      "step": 3289
+    },
+    {
+      "epoch": 0.5403075154476218,
+      "grad_norm": 0.7465434186042995,
+      "learning_rate": 8.783290918260212e-06,
+      "loss": 0.5532,
+      "step": 3290
+    },
+    {
+      "epoch": 0.5404717426559645,
+      "grad_norm": 0.3801930075543929,
+      "learning_rate": 8.78314711390924e-06,
+      "loss": 0.5595,
+      "step": 3291
+    },
+    {
+      "epoch": 0.5406359698643073,
+      "grad_norm": 0.35849452095029644,
+      "learning_rate": 8.783003263039077e-06,
+      "loss": 0.558,
+      "step": 3292
+    },
+    {
+      "epoch": 0.54080019707265,
+      "grad_norm": 0.3286836188938242,
+      "learning_rate": 8.782859365651284e-06,
+      "loss": 0.549,
+      "step": 3293
+    },
+    {
+      "epoch": 0.5409644242809928,
+      "grad_norm": 0.3544088100080713,
+      "learning_rate": 8.782715421747424e-06,
+      "loss": 0.5263,
+      "step": 3294
+    },
+    {
+      "epoch": 0.5411286514893355,
+      "grad_norm": 0.3012499735294153,
+      "learning_rate": 8.782571431329062e-06,
+      "loss": 0.5493,
+      "step": 3295
+    },
+    {
+      "epoch": 0.5412928786976783,
+      "grad_norm": 0.2862519026155833,
+      "learning_rate": 8.78242739439776e-06,
+      "loss": 0.5529,
+      "step": 3296
+    },
+    {
+      "epoch": 0.541457105906021,
+      "grad_norm": 0.34212920789775564,
+      "learning_rate": 8.782283310955084e-06,
+      "loss": 0.5156,
+      "step": 3297
+    },
+    {
+      "epoch": 0.5416213331143637,
+      "grad_norm": 0.2874868127529915,
+      "learning_rate": 8.782139181002598e-06,
+      "loss": 0.5347,
+      "step": 3298
+    },
+    {
+      "epoch": 0.5417855603227064,
+      "grad_norm": 0.30620036288558916,
+      "learning_rate": 8.781995004541866e-06,
+      "loss": 0.5562,
+      "step": 3299
+    },
+    {
+      "epoch": 0.5419497875310492,
+      "grad_norm": 0.3328961656461178,
+      "learning_rate": 8.781850781574458e-06,
+      "loss": 0.5402,
+      "step": 3300
+    },
+    {
+      "epoch": 0.5421140147393919,
+      "grad_norm": 0.4321499881343302,
+      "learning_rate": 8.781706512101936e-06,
+      "loss": 0.5424,
+      "step": 3301
+    },
+    {
+      "epoch": 0.5422782419477347,
+      "grad_norm": 0.2959437492978648,
+      "learning_rate": 8.781562196125868e-06,
+      "loss": 0.5616,
+      "step": 3302
+    },
+    {
+      "epoch": 0.5424424691560774,
+      "grad_norm": 0.34091956215548647,
+      "learning_rate": 8.781417833647823e-06,
+      "loss": 0.533,
+      "step": 3303
+    },
+    {
+      "epoch": 0.5426066963644202,
+      "grad_norm": 0.2969866679071415,
+      "learning_rate": 8.781273424669368e-06,
+      "loss": 0.5295,
+      "step": 3304
+    },
+    {
+      "epoch": 0.5427709235727629,
+      "grad_norm": 0.33197289823642345,
+      "learning_rate": 8.78112896919207e-06,
+      "loss": 0.5319,
+      "step": 3305
+    },
+    {
+      "epoch": 0.5429351507811057,
+      "grad_norm": 0.30111409556440877,
+      "learning_rate": 8.780984467217503e-06,
+      "loss": 0.5497,
+      "step": 3306
+    },
+    {
+      "epoch": 0.5430993779894484,
+      "grad_norm": 0.3166579827962453,
+      "learning_rate": 8.78083991874723e-06,
+      "loss": 0.5573,
+      "step": 3307
+    },
+    {
+      "epoch": 0.5432636051977912,
+      "grad_norm": 0.39090836439548915,
+      "learning_rate": 8.780695323782823e-06,
+      "loss": 0.5588,
+      "step": 3308
+    },
+    {
+      "epoch": 0.5434278324061339,
+      "grad_norm": 0.41882139487374753,
+      "learning_rate": 8.780550682325853e-06,
+      "loss": 0.5407,
+      "step": 3309
+    },
+    {
+      "epoch": 0.5435920596144767,
+      "grad_norm": 0.351388516687854,
+      "learning_rate": 8.780405994377893e-06,
+      "loss": 0.561,
+      "step": 3310
+    },
+    {
+      "epoch": 0.5437562868228194,
+      "grad_norm": 0.5146337082689596,
+      "learning_rate": 8.780261259940511e-06,
+      "loss": 0.5512,
+      "step": 3311
+    },
+    {
+      "epoch": 0.5439205140311622,
+      "grad_norm": 0.3632997394671109,
+      "learning_rate": 8.780116479015283e-06,
+      "loss": 0.5423,
+      "step": 3312
+    },
+    {
+      "epoch": 0.5440847412395049,
+      "grad_norm": 0.31828930879474143,
+      "learning_rate": 8.779971651603776e-06,
+      "loss": 0.5395,
+      "step": 3313
+    },
+    {
+      "epoch": 0.5442489684478476,
+      "grad_norm": 0.3052865887699414,
+      "learning_rate": 8.779826777707568e-06,
+      "loss": 0.5373,
+      "step": 3314
+    },
+    {
+      "epoch": 0.5444131956561903,
+      "grad_norm": 0.28649306849827694,
+      "learning_rate": 8.77968185732823e-06,
+      "loss": 0.5406,
+      "step": 3315
+    },
+    {
+      "epoch": 0.544577422864533,
+      "grad_norm": 0.3262919874735506,
+      "learning_rate": 8.779536890467336e-06,
+      "loss": 0.5552,
+      "step": 3316
+    },
+    {
+      "epoch": 0.5447416500728758,
+      "grad_norm": 0.28064816936119535,
+      "learning_rate": 8.77939187712646e-06,
+      "loss": 0.5448,
+      "step": 3317
+    },
+    {
+      "epoch": 0.5449058772812185,
+      "grad_norm": 0.2905610645984247,
+      "learning_rate": 8.77924681730718e-06,
+      "loss": 0.5514,
+      "step": 3318
+    },
+    {
+      "epoch": 0.5450701044895613,
+      "grad_norm": 0.3582455870421614,
+      "learning_rate": 8.779101711011067e-06,
+      "loss": 0.5543,
+      "step": 3319
+    },
+    {
+      "epoch": 0.545234331697904,
+      "grad_norm": 0.2956549075881862,
+      "learning_rate": 8.7789565582397e-06,
+      "loss": 0.528,
+      "step": 3320
+    },
+    {
+      "epoch": 0.5453985589062468,
+      "grad_norm": 0.32122379476319896,
+      "learning_rate": 8.778811358994655e-06,
+      "loss": 0.5133,
+      "step": 3321
+    },
+    {
+      "epoch": 0.5455627861145895,
+      "grad_norm": 0.2971563406062923,
+      "learning_rate": 8.77866611327751e-06,
+      "loss": 0.5245,
+      "step": 3322
+    },
+    {
+      "epoch": 0.5457270133229323,
+      "grad_norm": 0.3102675621709384,
+      "learning_rate": 8.77852082108984e-06,
+      "loss": 0.538,
+      "step": 3323
+    },
+    {
+      "epoch": 0.545891240531275,
+      "grad_norm": 0.2870978359981991,
+      "learning_rate": 8.778375482433226e-06,
+      "loss": 0.5384,
+      "step": 3324
+    },
+    {
+      "epoch": 0.5460554677396178,
+      "grad_norm": 0.2900289483988109,
+      "learning_rate": 8.778230097309243e-06,
+      "loss": 0.542,
+      "step": 3325
+    },
+    {
+      "epoch": 0.5462196949479605,
+      "grad_norm": 0.32092240664718624,
+      "learning_rate": 8.778084665719473e-06,
+      "loss": 0.5591,
+      "step": 3326
+    },
+    {
+      "epoch": 0.5463839221563033,
+      "grad_norm": 0.3126772582717247,
+      "learning_rate": 8.777939187665495e-06,
+      "loss": 0.5469,
+      "step": 3327
+    },
+    {
+      "epoch": 0.546548149364646,
+      "grad_norm": 0.3857668440217911,
+      "learning_rate": 8.777793663148888e-06,
+      "loss": 0.5633,
+      "step": 3328
+    },
+    {
+      "epoch": 0.5467123765729888,
+      "grad_norm": 0.28195626253107636,
+      "learning_rate": 8.777648092171232e-06,
+      "loss": 0.5439,
+      "step": 3329
+    },
+    {
+      "epoch": 0.5468766037813315,
+      "grad_norm": 0.36064556386630914,
+      "learning_rate": 8.777502474734109e-06,
+      "loss": 0.5489,
+      "step": 3330
+    },
+    {
+      "epoch": 0.5470408309896743,
+      "grad_norm": 0.33236569642633484,
+      "learning_rate": 8.777356810839102e-06,
+      "loss": 0.5251,
+      "step": 3331
+    },
+    {
+      "epoch": 0.5472050581980169,
+      "grad_norm": 0.5230949434342357,
+      "learning_rate": 8.77721110048779e-06,
+      "loss": 0.5652,
+      "step": 3332
+    },
+    {
+      "epoch": 0.5473692854063597,
+      "grad_norm": 0.2952566421849118,
+      "learning_rate": 8.77706534368176e-06,
+      "loss": 0.5377,
+      "step": 3333
+    },
+    {
+      "epoch": 0.5475335126147024,
+      "grad_norm": 0.35344945699330593,
+      "learning_rate": 8.77691954042259e-06,
+      "loss": 0.5225,
+      "step": 3334
+    },
+    {
+      "epoch": 0.5476977398230451,
+      "grad_norm": 0.32436302971883846,
+      "learning_rate": 8.776773690711866e-06,
+      "loss": 0.5521,
+      "step": 3335
+    },
+    {
+      "epoch": 0.5478619670313879,
+      "grad_norm": 0.30968709714872467,
+      "learning_rate": 8.776627794551174e-06,
+      "loss": 0.5473,
+      "step": 3336
+    },
+    {
+      "epoch": 0.5480261942397306,
+      "grad_norm": 0.292355032883547,
+      "learning_rate": 8.776481851942094e-06,
+      "loss": 0.5296,
+      "step": 3337
+    },
+    {
+      "epoch": 0.5481904214480734,
+      "grad_norm": 0.30650548477899137,
+      "learning_rate": 8.776335862886216e-06,
+      "loss": 0.5486,
+      "step": 3338
+    },
+    {
+      "epoch": 0.5483546486564161,
+      "grad_norm": 0.30828921012684446,
+      "learning_rate": 8.776189827385121e-06,
+      "loss": 0.5424,
+      "step": 3339
+    },
+    {
+      "epoch": 0.5485188758647589,
+      "grad_norm": 0.30623149799210697,
+      "learning_rate": 8.776043745440398e-06,
+      "loss": 0.5563,
+      "step": 3340
+    },
+    {
+      "epoch": 0.5486831030731016,
+      "grad_norm": 0.320878900454465,
+      "learning_rate": 8.775897617053633e-06,
+      "loss": 0.5434,
+      "step": 3341
+    },
+    {
+      "epoch": 0.5488473302814444,
+      "grad_norm": 0.3160907340766199,
+      "learning_rate": 8.775751442226412e-06,
+      "loss": 0.5266,
+      "step": 3342
+    },
+    {
+      "epoch": 0.5490115574897871,
+      "grad_norm": 0.2940009688601457,
+      "learning_rate": 8.775605220960325e-06,
+      "loss": 0.5591,
+      "step": 3343
+    },
+    {
+      "epoch": 0.5491757846981299,
+      "grad_norm": 0.2756229954361331,
+      "learning_rate": 8.775458953256958e-06,
+      "loss": 0.5422,
+      "step": 3344
+    },
+    {
+      "epoch": 0.5493400119064726,
+      "grad_norm": 0.4017540228900945,
+      "learning_rate": 8.7753126391179e-06,
+      "loss": 0.5378,
+      "step": 3345
+    },
+    {
+      "epoch": 0.5495042391148154,
+      "grad_norm": 0.30553436345975726,
+      "learning_rate": 8.775166278544742e-06,
+      "loss": 0.5207,
+      "step": 3346
+    },
+    {
+      "epoch": 0.5496684663231581,
+      "grad_norm": 0.3855818478613099,
+      "learning_rate": 8.775019871539071e-06,
+      "loss": 0.5377,
+      "step": 3347
+    },
+    {
+      "epoch": 0.5498326935315009,
+      "grad_norm": 0.26781930874015525,
+      "learning_rate": 8.774873418102477e-06,
+      "loss": 0.5154,
+      "step": 3348
+    },
+    {
+      "epoch": 0.5499969207398435,
+      "grad_norm": 0.5037312157429292,
+      "learning_rate": 8.774726918236553e-06,
+      "loss": 0.5387,
+      "step": 3349
+    },
+    {
+      "epoch": 0.5501611479481863,
+      "grad_norm": 0.37168947952148945,
+      "learning_rate": 8.774580371942888e-06,
+      "loss": 0.5503,
+      "step": 3350
+    },
+    {
+      "epoch": 0.550325375156529,
+      "grad_norm": 0.33315012217900974,
+      "learning_rate": 8.774433779223076e-06,
+      "loss": 0.5407,
+      "step": 3351
+    },
+    {
+      "epoch": 0.5504896023648718,
+      "grad_norm": 0.3247936341233711,
+      "learning_rate": 8.774287140078708e-06,
+      "loss": 0.5522,
+      "step": 3352
+    },
+    {
+      "epoch": 0.5506538295732145,
+      "grad_norm": 0.4965710840359514,
+      "learning_rate": 8.774140454511375e-06,
+      "loss": 0.5416,
+      "step": 3353
+    },
+    {
+      "epoch": 0.5508180567815573,
+      "grad_norm": 0.33188955600441644,
+      "learning_rate": 8.773993722522672e-06,
+      "loss": 0.5555,
+      "step": 3354
+    },
+    {
+      "epoch": 0.5509822839899,
+      "grad_norm": 0.418317451310977,
+      "learning_rate": 8.773846944114192e-06,
+      "loss": 0.5425,
+      "step": 3355
+    },
+    {
+      "epoch": 0.5511465111982428,
+      "grad_norm": 0.2825039647828183,
+      "learning_rate": 8.773700119287528e-06,
+      "loss": 0.5707,
+      "step": 3356
+    },
+    {
+      "epoch": 0.5513107384065855,
+      "grad_norm": 0.3538687145304638,
+      "learning_rate": 8.773553248044278e-06,
+      "loss": 0.5685,
+      "step": 3357
+    },
+    {
+      "epoch": 0.5514749656149283,
+      "grad_norm": 0.42920655053335216,
+      "learning_rate": 8.773406330386034e-06,
+      "loss": 0.5295,
+      "step": 3358
+    },
+    {
+      "epoch": 0.551639192823271,
+      "grad_norm": 0.32412456816316443,
+      "learning_rate": 8.773259366314393e-06,
+      "loss": 0.5574,
+      "step": 3359
+    },
+    {
+      "epoch": 0.5518034200316138,
+      "grad_norm": 0.2797144018953076,
+      "learning_rate": 8.77311235583095e-06,
+      "loss": 0.5669,
+      "step": 3360
+    },
+    {
+      "epoch": 0.5519676472399565,
+      "grad_norm": 0.37403853631233064,
+      "learning_rate": 8.772965298937305e-06,
+      "loss": 0.5285,
+      "step": 3361
+    },
+    {
+      "epoch": 0.5521318744482993,
+      "grad_norm": 0.28490995583623363,
+      "learning_rate": 8.772818195635052e-06,
+      "loss": 0.5413,
+      "step": 3362
+    },
+    {
+      "epoch": 0.552296101656642,
+      "grad_norm": 0.32396743410008255,
+      "learning_rate": 8.772671045925788e-06,
+      "loss": 0.5585,
+      "step": 3363
+    },
+    {
+      "epoch": 0.5524603288649848,
+      "grad_norm": 0.3075785289042439,
+      "learning_rate": 8.772523849811114e-06,
+      "loss": 0.5535,
+      "step": 3364
+    },
+    {
+      "epoch": 0.5526245560733275,
+      "grad_norm": 0.4027775951653619,
+      "learning_rate": 8.772376607292627e-06,
+      "loss": 0.55,
+      "step": 3365
+    },
+    {
+      "epoch": 0.5527887832816701,
+      "grad_norm": 0.385492345624032,
+      "learning_rate": 8.772229318371927e-06,
+      "loss": 0.5506,
+      "step": 3366
+    },
+    {
+      "epoch": 0.5529530104900129,
+      "grad_norm": 0.33761364893706824,
+      "learning_rate": 8.772081983050611e-06,
+      "loss": 0.5321,
+      "step": 3367
+    },
+    {
+      "epoch": 0.5531172376983556,
+      "grad_norm": 0.3403673283677389,
+      "learning_rate": 8.771934601330285e-06,
+      "loss": 0.5451,
+      "step": 3368
+    },
+    {
+      "epoch": 0.5532814649066984,
+      "grad_norm": 0.35126406168503904,
+      "learning_rate": 8.771787173212545e-06,
+      "loss": 0.5395,
+      "step": 3369
+    },
+    {
+      "epoch": 0.5534456921150411,
+      "grad_norm": 0.36250462085917223,
+      "learning_rate": 8.771639698698993e-06,
+      "loss": 0.5538,
+      "step": 3370
+    },
+    {
+      "epoch": 0.5536099193233839,
+      "grad_norm": 0.28299198501272577,
+      "learning_rate": 8.77149217779123e-06,
+      "loss": 0.5542,
+      "step": 3371
+    },
+    {
+      "epoch": 0.5537741465317266,
+      "grad_norm": 0.3021177801777963,
+      "learning_rate": 8.77134461049086e-06,
+      "loss": 0.5455,
+      "step": 3372
+    },
+    {
+      "epoch": 0.5539383737400694,
+      "grad_norm": 0.33274453416677413,
+      "learning_rate": 8.771196996799486e-06,
+      "loss": 0.5501,
+      "step": 3373
+    },
+    {
+      "epoch": 0.5541026009484121,
+      "grad_norm": 0.31709314075267925,
+      "learning_rate": 8.77104933671871e-06,
+      "loss": 0.5319,
+      "step": 3374
+    },
+    {
+      "epoch": 0.5542668281567549,
+      "grad_norm": 0.38946982014827064,
+      "learning_rate": 8.770901630250137e-06,
+      "loss": 0.5438,
+      "step": 3375
+    },
+    {
+      "epoch": 0.5544310553650976,
+      "grad_norm": 0.3106343141375796,
+      "learning_rate": 8.77075387739537e-06,
+      "loss": 0.537,
+      "step": 3376
+    },
+    {
+      "epoch": 0.5545952825734404,
+      "grad_norm": 0.40028112123615217,
+      "learning_rate": 8.770606078156013e-06,
+      "loss": 0.5553,
+      "step": 3377
+    },
+    {
+      "epoch": 0.5547595097817831,
+      "grad_norm": 0.3187911975290353,
+      "learning_rate": 8.770458232533672e-06,
+      "loss": 0.5554,
+      "step": 3378
+    },
+    {
+      "epoch": 0.5549237369901259,
+      "grad_norm": 0.3378804598434121,
+      "learning_rate": 8.770310340529954e-06,
+      "loss": 0.5533,
+      "step": 3379
+    },
+    {
+      "epoch": 0.5550879641984686,
+      "grad_norm": 0.3401695168568818,
+      "learning_rate": 8.770162402146465e-06,
+      "loss": 0.5408,
+      "step": 3380
+    },
+    {
+      "epoch": 0.5552521914068114,
+      "grad_norm": 0.29459484154398957,
+      "learning_rate": 8.77001441738481e-06,
+      "loss": 0.5198,
+      "step": 3381
+    },
+    {
+      "epoch": 0.5554164186151541,
+      "grad_norm": 0.3459913273509143,
+      "learning_rate": 8.769866386246596e-06,
+      "loss": 0.5393,
+      "step": 3382
+    },
+    {
+      "epoch": 0.5555806458234968,
+      "grad_norm": 0.3211088301865988,
+      "learning_rate": 8.769718308733434e-06,
+      "loss": 0.5494,
+      "step": 3383
+    },
+    {
+      "epoch": 0.5557448730318395,
+      "grad_norm": 0.3254018054944106,
+      "learning_rate": 8.769570184846929e-06,
+      "loss": 0.5429,
+      "step": 3384
+    },
+    {
+      "epoch": 0.5559091002401823,
+      "grad_norm": 0.3114123751834169,
+      "learning_rate": 8.769422014588692e-06,
+      "loss": 0.5479,
+      "step": 3385
+    },
+    {
+      "epoch": 0.556073327448525,
+      "grad_norm": 0.3236795283927829,
+      "learning_rate": 8.769273797960331e-06,
+      "loss": 0.5375,
+      "step": 3386
+    },
+    {
+      "epoch": 0.5562375546568677,
+      "grad_norm": 0.4880689140211738,
+      "learning_rate": 8.769125534963457e-06,
+      "loss": 0.5387,
+      "step": 3387
+    },
+    {
+      "epoch": 0.5564017818652105,
+      "grad_norm": 0.31117332262719577,
+      "learning_rate": 8.768977225599679e-06,
+      "loss": 0.5455,
+      "step": 3388
+    },
+    {
+      "epoch": 0.5565660090735532,
+      "grad_norm": 0.2786661939353369,
+      "learning_rate": 8.768828869870609e-06,
+      "loss": 0.5511,
+      "step": 3389
+    },
+    {
+      "epoch": 0.556730236281896,
+      "grad_norm": 0.31635464932598195,
+      "learning_rate": 8.768680467777857e-06,
+      "loss": 0.5336,
+      "step": 3390
+    },
+    {
+      "epoch": 0.5568944634902387,
+      "grad_norm": 0.29765570525059454,
+      "learning_rate": 8.768532019323034e-06,
+      "loss": 0.5561,
+      "step": 3391
+    },
+    {
+      "epoch": 0.5570586906985815,
+      "grad_norm": 0.3243154243746134,
+      "learning_rate": 8.768383524507754e-06,
+      "loss": 0.5381,
+      "step": 3392
+    },
+    {
+      "epoch": 0.5572229179069242,
+      "grad_norm": 0.2893446275664096,
+      "learning_rate": 8.76823498333363e-06,
+      "loss": 0.5556,
+      "step": 3393
+    },
+    {
+      "epoch": 0.557387145115267,
+      "grad_norm": 0.3273044449065161,
+      "learning_rate": 8.768086395802274e-06,
+      "loss": 0.5476,
+      "step": 3394
+    },
+    {
+      "epoch": 0.5575513723236097,
+      "grad_norm": 0.3267822607622496,
+      "learning_rate": 8.767937761915302e-06,
+      "loss": 0.5467,
+      "step": 3395
+    },
+    {
+      "epoch": 0.5577155995319525,
+      "grad_norm": 0.3321830374515757,
+      "learning_rate": 8.767789081674324e-06,
+      "loss": 0.5442,
+      "step": 3396
+    },
+    {
+      "epoch": 0.5578798267402952,
+      "grad_norm": 0.361249384322939,
+      "learning_rate": 8.767640355080962e-06,
+      "loss": 0.5374,
+      "step": 3397
+    },
+    {
+      "epoch": 0.558044053948638,
+      "grad_norm": 0.3154351904554402,
+      "learning_rate": 8.767491582136823e-06,
+      "loss": 0.536,
+      "step": 3398
+    },
+    {
+      "epoch": 0.5582082811569807,
+      "grad_norm": 0.2980191259728029,
+      "learning_rate": 8.767342762843529e-06,
+      "loss": 0.5449,
+      "step": 3399
+    },
+    {
+      "epoch": 0.5583725083653234,
+      "grad_norm": 0.39564078503411465,
+      "learning_rate": 8.767193897202692e-06,
+      "loss": 0.5377,
+      "step": 3400
+    },
+    {
+      "epoch": 0.5585367355736661,
+      "grad_norm": 0.3517990839488776,
+      "learning_rate": 8.767044985215933e-06,
+      "loss": 0.5404,
+      "step": 3401
+    },
+    {
+      "epoch": 0.5587009627820089,
+      "grad_norm": 0.4155353751005951,
+      "learning_rate": 8.766896026884868e-06,
+      "loss": 0.5333,
+      "step": 3402
+    },
+    {
+      "epoch": 0.5588651899903516,
+      "grad_norm": 0.3017636592925083,
+      "learning_rate": 8.766747022211112e-06,
+      "loss": 0.5676,
+      "step": 3403
+    },
+    {
+      "epoch": 0.5590294171986944,
+      "grad_norm": 0.33995408666281135,
+      "learning_rate": 8.766597971196285e-06,
+      "loss": 0.5374,
+      "step": 3404
+    },
+    {
+      "epoch": 0.5591936444070371,
+      "grad_norm": 0.2795836141353966,
+      "learning_rate": 8.766448873842009e-06,
+      "loss": 0.5438,
+      "step": 3405
+    },
+    {
+      "epoch": 0.5593578716153799,
+      "grad_norm": 0.35753173927761156,
+      "learning_rate": 8.766299730149898e-06,
+      "loss": 0.5552,
+      "step": 3406
+    },
+    {
+      "epoch": 0.5595220988237226,
+      "grad_norm": 0.3248618335926127,
+      "learning_rate": 8.766150540121576e-06,
+      "loss": 0.532,
+      "step": 3407
+    },
+    {
+      "epoch": 0.5596863260320654,
+      "grad_norm": 0.36621704636882757,
+      "learning_rate": 8.766001303758661e-06,
+      "loss": 0.5654,
+      "step": 3408
+    },
+    {
+      "epoch": 0.5598505532404081,
+      "grad_norm": 0.3491811217207003,
+      "learning_rate": 8.765852021062774e-06,
+      "loss": 0.5283,
+      "step": 3409
+    },
+    {
+      "epoch": 0.5600147804487509,
+      "grad_norm": 0.3545037992141468,
+      "learning_rate": 8.765702692035539e-06,
+      "loss": 0.548,
+      "step": 3410
+    },
+    {
+      "epoch": 0.5601790076570936,
+      "grad_norm": 0.2763757716992341,
+      "learning_rate": 8.765553316678574e-06,
+      "loss": 0.5585,
+      "step": 3411
+    },
+    {
+      "epoch": 0.5603432348654364,
+      "grad_norm": 0.30894970057288423,
+      "learning_rate": 8.765403894993503e-06,
+      "loss": 0.5697,
+      "step": 3412
+    },
+    {
+      "epoch": 0.5605074620737791,
+      "grad_norm": 0.30395462426760694,
+      "learning_rate": 8.765254426981951e-06,
+      "loss": 0.5392,
+      "step": 3413
+    },
+    {
+      "epoch": 0.5606716892821219,
+      "grad_norm": 0.2938324142755099,
+      "learning_rate": 8.765104912645538e-06,
+      "loss": 0.5445,
+      "step": 3414
+    },
+    {
+      "epoch": 0.5608359164904646,
+      "grad_norm": 0.33410474841600196,
+      "learning_rate": 8.76495535198589e-06,
+      "loss": 0.533,
+      "step": 3415
+    },
+    {
+      "epoch": 0.5610001436988074,
+      "grad_norm": 0.3183614582303699,
+      "learning_rate": 8.764805745004632e-06,
+      "loss": 0.5414,
+      "step": 3416
+    },
+    {
+      "epoch": 0.56116437090715,
+      "grad_norm": 0.3107793870235241,
+      "learning_rate": 8.764656091703386e-06,
+      "loss": 0.5563,
+      "step": 3417
+    },
+    {
+      "epoch": 0.5613285981154927,
+      "grad_norm": 0.28256412488593424,
+      "learning_rate": 8.76450639208378e-06,
+      "loss": 0.5213,
+      "step": 3418
+    },
+    {
+      "epoch": 0.5614928253238355,
+      "grad_norm": 0.3291646446986307,
+      "learning_rate": 8.764356646147437e-06,
+      "loss": 0.5298,
+      "step": 3419
+    },
+    {
+      "epoch": 0.5616570525321782,
+      "grad_norm": 0.29015251711779344,
+      "learning_rate": 8.764206853895987e-06,
+      "loss": 0.5451,
+      "step": 3420
+    },
+    {
+      "epoch": 0.561821279740521,
+      "grad_norm": 0.3669840500783633,
+      "learning_rate": 8.764057015331054e-06,
+      "loss": 0.5599,
+      "step": 3421
+    },
+    {
+      "epoch": 0.5619855069488637,
+      "grad_norm": 0.39324670708032294,
+      "learning_rate": 8.763907130454267e-06,
+      "loss": 0.5375,
+      "step": 3422
+    },
+    {
+      "epoch": 0.5621497341572065,
+      "grad_norm": 0.31065568764508283,
+      "learning_rate": 8.763757199267253e-06,
+      "loss": 0.5408,
+      "step": 3423
+    },
+    {
+      "epoch": 0.5623139613655492,
+      "grad_norm": 0.3830704592320809,
+      "learning_rate": 8.763607221771643e-06,
+      "loss": 0.5215,
+      "step": 3424
+    },
+    {
+      "epoch": 0.562478188573892,
+      "grad_norm": 0.38348921195787256,
+      "learning_rate": 8.763457197969061e-06,
+      "loss": 0.5264,
+      "step": 3425
+    },
+    {
+      "epoch": 0.5626424157822347,
+      "grad_norm": 0.37148155697446655,
+      "learning_rate": 8.76330712786114e-06,
+      "loss": 0.5582,
+      "step": 3426
+    },
+    {
+      "epoch": 0.5628066429905775,
+      "grad_norm": 0.4273592131591784,
+      "learning_rate": 8.76315701144951e-06,
+      "loss": 0.5652,
+      "step": 3427
+    },
+    {
+      "epoch": 0.5629708701989202,
+      "grad_norm": 0.36466575092615616,
+      "learning_rate": 8.7630068487358e-06,
+      "loss": 0.5272,
+      "step": 3428
+    },
+    {
+      "epoch": 0.563135097407263,
+      "grad_norm": 0.3627460014212349,
+      "learning_rate": 8.762856639721642e-06,
+      "loss": 0.5487,
+      "step": 3429
+    },
+    {
+      "epoch": 0.5632993246156057,
+      "grad_norm": 0.8673504397637322,
+      "learning_rate": 8.762706384408666e-06,
+      "loss": 0.5456,
+      "step": 3430
+    },
+    {
+      "epoch": 0.5634635518239485,
+      "grad_norm": 0.32618874303521883,
+      "learning_rate": 8.762556082798505e-06,
+      "loss": 0.5374,
+      "step": 3431
+    },
+    {
+      "epoch": 0.5636277790322912,
+      "grad_norm": 0.3913592007145045,
+      "learning_rate": 8.76240573489279e-06,
+      "loss": 0.5377,
+      "step": 3432
+    },
+    {
+      "epoch": 0.563792006240634,
+      "grad_norm": 0.34399391895825815,
+      "learning_rate": 8.762255340693156e-06,
+      "loss": 0.5405,
+      "step": 3433
+    },
+    {
+      "epoch": 0.5639562334489766,
+      "grad_norm": 0.307384899591098,
+      "learning_rate": 8.762104900201235e-06,
+      "loss": 0.5346,
+      "step": 3434
+    },
+    {
+      "epoch": 0.5641204606573194,
+      "grad_norm": 0.2968959663698474,
+      "learning_rate": 8.761954413418663e-06,
+      "loss": 0.5568,
+      "step": 3435
+    },
+    {
+      "epoch": 0.5642846878656621,
+      "grad_norm": 0.3949150892211441,
+      "learning_rate": 8.761803880347073e-06,
+      "loss": 0.5486,
+      "step": 3436
+    },
+    {
+      "epoch": 0.5644489150740049,
+      "grad_norm": 0.33926890627687606,
+      "learning_rate": 8.761653300988097e-06,
+      "loss": 0.5197,
+      "step": 3437
+    },
+    {
+      "epoch": 0.5646131422823476,
+      "grad_norm": 0.3172017138331228,
+      "learning_rate": 8.761502675343375e-06,
+      "loss": 0.5471,
+      "step": 3438
+    },
+    {
+      "epoch": 0.5647773694906904,
+      "grad_norm": 0.418356801651984,
+      "learning_rate": 8.761352003414541e-06,
+      "loss": 0.5321,
+      "step": 3439
+    },
+    {
+      "epoch": 0.5649415966990331,
+      "grad_norm": 0.3079378893504073,
+      "learning_rate": 8.761201285203232e-06,
+      "loss": 0.5321,
+      "step": 3440
+    },
+    {
+      "epoch": 0.5651058239073758,
+      "grad_norm": 0.37210351753249876,
+      "learning_rate": 8.761050520711083e-06,
+      "loss": 0.5329,
+      "step": 3441
+    },
+    {
+      "epoch": 0.5652700511157186,
+      "grad_norm": 0.2805648721097993,
+      "learning_rate": 8.760899709939735e-06,
+      "loss": 0.5186,
+      "step": 3442
+    },
+    {
+      "epoch": 0.5654342783240613,
+      "grad_norm": 0.47212938527808185,
+      "learning_rate": 8.760748852890824e-06,
+      "loss": 0.5474,
+      "step": 3443
+    },
+    {
+      "epoch": 0.5655985055324041,
+      "grad_norm": 0.342366746899832,
+      "learning_rate": 8.760597949565988e-06,
+      "loss": 0.5356,
+      "step": 3444
+    },
+    {
+      "epoch": 0.5657627327407468,
+      "grad_norm": 0.30910313660004984,
+      "learning_rate": 8.760446999966866e-06,
+      "loss": 0.5389,
+      "step": 3445
+    },
+    {
+      "epoch": 0.5659269599490896,
+      "grad_norm": 0.3183907922282123,
+      "learning_rate": 8.760296004095098e-06,
+      "loss": 0.526,
+      "step": 3446
+    },
+    {
+      "epoch": 0.5660911871574323,
+      "grad_norm": 0.3900742021203354,
+      "learning_rate": 8.760144961952324e-06,
+      "loss": 0.538,
+      "step": 3447
+    },
+    {
+      "epoch": 0.5662554143657751,
+      "grad_norm": 0.4494570545768727,
+      "learning_rate": 8.759993873540184e-06,
+      "loss": 0.5409,
+      "step": 3448
+    },
+    {
+      "epoch": 0.5664196415741178,
+      "grad_norm": 0.3344372117969043,
+      "learning_rate": 8.75984273886032e-06,
+      "loss": 0.5416,
+      "step": 3449
+    },
+    {
+      "epoch": 0.5665838687824606,
+      "grad_norm": 0.28815799555494404,
+      "learning_rate": 8.759691557914372e-06,
+      "loss": 0.5321,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5667480959908032,
+      "grad_norm": 0.3427716429030344,
+      "learning_rate": 8.759540330703983e-06,
+      "loss": 0.5218,
+      "step": 3451
+    },
+    {
+      "epoch": 0.566912323199146,
+      "grad_norm": 0.3535468139773774,
+      "learning_rate": 8.759389057230795e-06,
+      "loss": 0.5311,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5670765504074887,
+      "grad_norm": 0.37238689749646064,
+      "learning_rate": 8.759237737496451e-06,
+      "loss": 0.5626,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5672407776158315,
+      "grad_norm": 0.4913391693392319,
+      "learning_rate": 8.759086371502595e-06,
+      "loss": 0.5505,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5674050048241742,
+      "grad_norm": 0.37181041963439404,
+      "learning_rate": 8.75893495925087e-06,
+      "loss": 0.5414,
+      "step": 3455
+    },
+    {
+      "epoch": 0.567569232032517,
+      "grad_norm": 0.31509365349628793,
+      "learning_rate": 8.758783500742922e-06,
+      "loss": 0.55,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5677334592408597,
+      "grad_norm": 0.3368764395024524,
+      "learning_rate": 8.758631995980395e-06,
+      "loss": 0.5384,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5678976864492025,
+      "grad_norm": 0.42083886653211133,
+      "learning_rate": 8.758480444964933e-06,
+      "loss": 0.55,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5680619136575452,
+      "grad_norm": 0.30635925898925725,
+      "learning_rate": 8.758328847698185e-06,
+      "loss": 0.5211,
+      "step": 3459
+    },
+    {
+      "epoch": 0.568226140865888,
+      "grad_norm": 0.31708343966830854,
+      "learning_rate": 8.758177204181797e-06,
+      "loss": 0.5579,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5683903680742307,
+      "grad_norm": 0.36409414468119855,
+      "learning_rate": 8.758025514417415e-06,
+      "loss": 0.5375,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5685545952825735,
+      "grad_norm": 0.3112449064143783,
+      "learning_rate": 8.757873778406686e-06,
+      "loss": 0.5374,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5687188224909162,
+      "grad_norm": 0.3608556254888521,
+      "learning_rate": 8.757721996151258e-06,
+      "loss": 0.5528,
+      "step": 3463
+    },
+    {
+      "epoch": 0.568883049699259,
+      "grad_norm": 0.3114835902741878,
+      "learning_rate": 8.757570167652781e-06,
+      "loss": 0.5478,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5690472769076017,
+      "grad_norm": 0.4473767595398761,
+      "learning_rate": 8.757418292912902e-06,
+      "loss": 0.5407,
+      "step": 3465
+    },
+    {
+      "epoch": 0.5692115041159445,
+      "grad_norm": 0.30253085478675174,
+      "learning_rate": 8.757266371933272e-06,
+      "loss": 0.5603,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5693757313242872,
+      "grad_norm": 0.3115092905114874,
+      "learning_rate": 8.75711440471554e-06,
+      "loss": 0.5284,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5695399585326298,
+      "grad_norm": 0.30839473910587395,
+      "learning_rate": 8.756962391261358e-06,
+      "loss": 0.5443,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5697041857409726,
+      "grad_norm": 0.28102093112639204,
+      "learning_rate": 8.756810331572375e-06,
+      "loss": 0.5476,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5698684129493153,
+      "grad_norm": 0.3302176543184028,
+      "learning_rate": 8.756658225650245e-06,
+      "loss": 0.5431,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5700326401576581,
+      "grad_norm": 0.3365665104771197,
+      "learning_rate": 8.75650607349662e-06,
+      "loss": 0.5262,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5701968673660008,
+      "grad_norm": 0.3067227588817041,
+      "learning_rate": 8.75635387511315e-06,
+      "loss": 0.5384,
+      "step": 3472
+    },
+    {
+      "epoch": 0.5703610945743436,
+      "grad_norm": 0.33218329012978515,
+      "learning_rate": 8.756201630501487e-06,
+      "loss": 0.5332,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5705253217826863,
+      "grad_norm": 0.3713379909457561,
+      "learning_rate": 8.756049339663288e-06,
+      "loss": 0.5466,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5706895489910291,
+      "grad_norm": 0.34441687097450313,
+      "learning_rate": 8.755897002600207e-06,
+      "loss": 0.5541,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5708537761993718,
+      "grad_norm": 0.7228301451209881,
+      "learning_rate": 8.755744619313896e-06,
+      "loss": 0.5376,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5710180034077146,
+      "grad_norm": 0.2999297342940657,
+      "learning_rate": 8.755592189806012e-06,
+      "loss": 0.5443,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5711822306160573,
+      "grad_norm": 0.39362953316784804,
+      "learning_rate": 8.75543971407821e-06,
+      "loss": 0.535,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5713464578244001,
+      "grad_norm": 0.4111607829748149,
+      "learning_rate": 8.755287192132145e-06,
+      "loss": 0.5611,
+      "step": 3479
+    },
+    {
+      "epoch": 0.5715106850327428,
+      "grad_norm": 0.29266563450602223,
+      "learning_rate": 8.755134623969477e-06,
+      "loss": 0.5634,
+      "step": 3480
+    },
+    {
+      "epoch": 0.5716749122410856,
+      "grad_norm": 0.32732512271978664,
+      "learning_rate": 8.75498200959186e-06,
+      "loss": 0.543,
+      "step": 3481
+    },
+    {
+      "epoch": 0.5718391394494283,
+      "grad_norm": 0.46875550296652735,
+      "learning_rate": 8.754829349000948e-06,
+      "loss": 0.5457,
+      "step": 3482
+    },
+    {
+      "epoch": 0.5720033666577711,
+      "grad_norm": 0.33783116396267615,
+      "learning_rate": 8.754676642198407e-06,
+      "loss": 0.526,
+      "step": 3483
+    },
+    {
+      "epoch": 0.5721675938661138,
+      "grad_norm": 0.5346132082744581,
+      "learning_rate": 8.75452388918589e-06,
+      "loss": 0.5329,
+      "step": 3484
+    },
+    {
+      "epoch": 0.5723318210744565,
+      "grad_norm": 0.2764576251732501,
+      "learning_rate": 8.754371089965058e-06,
+      "loss": 0.5519,
+      "step": 3485
+    },
+    {
+      "epoch": 0.5724960482827992,
+      "grad_norm": 0.2688202687009742,
+      "learning_rate": 8.75421824453757e-06,
+      "loss": 0.5399,
+      "step": 3486
+    },
+    {
+      "epoch": 0.572660275491142,
+      "grad_norm": 0.3093609334958058,
+      "learning_rate": 8.754065352905087e-06,
+      "loss": 0.5277,
+      "step": 3487
+    },
+    {
+      "epoch": 0.5728245026994847,
+      "grad_norm": 0.39119778064404526,
+      "learning_rate": 8.753912415069269e-06,
+      "loss": 0.5498,
+      "step": 3488
+    },
+    {
+      "epoch": 0.5729887299078275,
+      "grad_norm": 0.2955026812716169,
+      "learning_rate": 8.753759431031775e-06,
+      "loss": 0.5499,
+      "step": 3489
+    },
+    {
+      "epoch": 0.5731529571161702,
+      "grad_norm": 0.3113087610249926,
+      "learning_rate": 8.753606400794271e-06,
+      "loss": 0.5452,
+      "step": 3490
+    },
+    {
+      "epoch": 0.573317184324513,
+      "grad_norm": 0.3611344907617437,
+      "learning_rate": 8.753453324358416e-06,
+      "loss": 0.5348,
+      "step": 3491
+    },
+    {
+      "epoch": 0.5734814115328557,
+      "grad_norm": 0.3344023185058532,
+      "learning_rate": 8.753300201725872e-06,
+      "loss": 0.5429,
+      "step": 3492
+    },
+    {
+      "epoch": 0.5736456387411984,
+      "grad_norm": 0.304956411481048,
+      "learning_rate": 8.753147032898303e-06,
+      "loss": 0.5482,
+      "step": 3493
+    },
+    {
+      "epoch": 0.5738098659495412,
+      "grad_norm": 0.30165741270268126,
+      "learning_rate": 8.752993817877373e-06,
+      "loss": 0.5328,
+      "step": 3494
+    },
+    {
+      "epoch": 0.573974093157884,
+      "grad_norm": 0.33741706948502814,
+      "learning_rate": 8.752840556664747e-06,
+      "loss": 0.5189,
+      "step": 3495
+    },
+    {
+      "epoch": 0.5741383203662267,
+      "grad_norm": 0.3701100738767769,
+      "learning_rate": 8.752687249262087e-06,
+      "loss": 0.5623,
+      "step": 3496
+    },
+    {
+      "epoch": 0.5743025475745694,
+      "grad_norm": 0.3169045858154691,
+      "learning_rate": 8.75253389567106e-06,
+      "loss": 0.5277,
+      "step": 3497
+    },
+    {
+      "epoch": 0.5744667747829122,
+      "grad_norm": 0.3854837052550634,
+      "learning_rate": 8.75238049589333e-06,
+      "loss": 0.5461,
+      "step": 3498
+    },
+    {
+      "epoch": 0.5746310019912549,
+      "grad_norm": 0.35234654736307097,
+      "learning_rate": 8.752227049930566e-06,
+      "loss": 0.5478,
+      "step": 3499
+    },
+    {
+      "epoch": 0.5747952291995977,
+      "grad_norm": 0.43861512127403823,
+      "learning_rate": 8.75207355778443e-06,
+      "loss": 0.5158,
+      "step": 3500
+    },
+    {
+      "epoch": 0.5749594564079404,
+      "grad_norm": 0.5687892409034156,
+      "learning_rate": 8.751920019456594e-06,
+      "loss": 0.5426,
+      "step": 3501
+    },
+    {
+      "epoch": 0.5751236836162831,
+      "grad_norm": 0.321497927550389,
+      "learning_rate": 8.75176643494872e-06,
+      "loss": 0.5448,
+      "step": 3502
+    },
+    {
+      "epoch": 0.5752879108246258,
+      "grad_norm": 0.3619816521146462,
+      "learning_rate": 8.751612804262483e-06,
+      "loss": 0.539,
+      "step": 3503
+    },
+    {
+      "epoch": 0.5754521380329686,
+      "grad_norm": 0.32956518716772837,
+      "learning_rate": 8.751459127399548e-06,
+      "loss": 0.5492,
+      "step": 3504
+    },
+    {
+      "epoch": 0.5756163652413113,
+      "grad_norm": 0.3437166953073345,
+      "learning_rate": 8.751305404361582e-06,
+      "loss": 0.5374,
+      "step": 3505
+    },
+    {
+      "epoch": 0.5757805924496541,
+      "grad_norm": 0.32115500488296933,
+      "learning_rate": 8.751151635150255e-06,
+      "loss": 0.5278,
+      "step": 3506
+    },
+    {
+      "epoch": 0.5759448196579968,
+      "grad_norm": 0.347119267249423,
+      "learning_rate": 8.750997819767241e-06,
+      "loss": 0.5529,
+      "step": 3507
+    },
+    {
+      "epoch": 0.5761090468663396,
+      "grad_norm": 0.28964439064190384,
+      "learning_rate": 8.75084395821421e-06,
+      "loss": 0.5135,
+      "step": 3508
+    },
+    {
+      "epoch": 0.5762732740746823,
+      "grad_norm": 0.3267292561169078,
+      "learning_rate": 8.750690050492828e-06,
+      "loss": 0.5232,
+      "step": 3509
+    },
+    {
+      "epoch": 0.5764375012830251,
+      "grad_norm": 0.31301380438742404,
+      "learning_rate": 8.750536096604772e-06,
+      "loss": 0.5444,
+      "step": 3510
+    },
+    {
+      "epoch": 0.5766017284913678,
+      "grad_norm": 0.3211593964134133,
+      "learning_rate": 8.750382096551711e-06,
+      "loss": 0.5541,
+      "step": 3511
+    },
+    {
+      "epoch": 0.5767659556997106,
+      "grad_norm": 0.2826221728840829,
+      "learning_rate": 8.750228050335319e-06,
+      "loss": 0.5424,
+      "step": 3512
+    },
+    {
+      "epoch": 0.5769301829080533,
+      "grad_norm": 0.436341101746728,
+      "learning_rate": 8.750073957957269e-06,
+      "loss": 0.5366,
+      "step": 3513
+    },
+    {
+      "epoch": 0.5770944101163961,
+      "grad_norm": 0.38049927921661986,
+      "learning_rate": 8.749919819419234e-06,
+      "loss": 0.5241,
+      "step": 3514
+    },
+    {
+      "epoch": 0.5772586373247388,
+      "grad_norm": 0.32937459946163317,
+      "learning_rate": 8.749765634722889e-06,
+      "loss": 0.5361,
+      "step": 3515
+    },
+    {
+      "epoch": 0.5774228645330816,
+      "grad_norm": 0.3247301226127869,
+      "learning_rate": 8.749611403869907e-06,
+      "loss": 0.5296,
+      "step": 3516
+    },
+    {
+      "epoch": 0.5775870917414243,
+      "grad_norm": 0.3040428174782979,
+      "learning_rate": 8.749457126861965e-06,
+      "loss": 0.5287,
+      "step": 3517
+    },
+    {
+      "epoch": 0.5777513189497671,
+      "grad_norm": 0.33533069093525125,
+      "learning_rate": 8.749302803700735e-06,
+      "loss": 0.5494,
+      "step": 3518
+    },
+    {
+      "epoch": 0.5779155461581097,
+      "grad_norm": 0.3993043789739712,
+      "learning_rate": 8.7491484343879e-06,
+      "loss": 0.5533,
+      "step": 3519
+    },
+    {
+      "epoch": 0.5780797733664524,
+      "grad_norm": 0.37167067761157563,
+      "learning_rate": 8.748994018925129e-06,
+      "loss": 0.5457,
+      "step": 3520
+    },
+    {
+      "epoch": 0.5782440005747952,
+      "grad_norm": 0.29226128233424004,
+      "learning_rate": 8.748839557314105e-06,
+      "loss": 0.5426,
+      "step": 3521
+    },
+    {
+      "epoch": 0.5784082277831379,
+      "grad_norm": 0.3363682700498467,
+      "learning_rate": 8.748685049556502e-06,
+      "loss": 0.5322,
+      "step": 3522
+    },
+    {
+      "epoch": 0.5785724549914807,
+      "grad_norm": 0.6507199888257422,
+      "learning_rate": 8.748530495653999e-06,
+      "loss": 0.553,
+      "step": 3523
+    },
+    {
+      "epoch": 0.5787366821998234,
+      "grad_norm": 0.33296678939243896,
+      "learning_rate": 8.748375895608275e-06,
+      "loss": 0.5268,
+      "step": 3524
+    },
+    {
+      "epoch": 0.5789009094081662,
+      "grad_norm": 0.29283720540204305,
+      "learning_rate": 8.748221249421009e-06,
+      "loss": 0.5463,
+      "step": 3525
+    },
+    {
+      "epoch": 0.5790651366165089,
+      "grad_norm": 0.3050167584945107,
+      "learning_rate": 8.74806655709388e-06,
+      "loss": 0.549,
+      "step": 3526
+    },
+    {
+      "epoch": 0.5792293638248517,
+      "grad_norm": 0.34419849382255013,
+      "learning_rate": 8.74791181862857e-06,
+      "loss": 0.5378,
+      "step": 3527
+    },
+    {
+      "epoch": 0.5793935910331944,
+      "grad_norm": 0.3661112951777841,
+      "learning_rate": 8.747757034026757e-06,
+      "loss": 0.5397,
+      "step": 3528
+    },
+    {
+      "epoch": 0.5795578182415372,
+      "grad_norm": 0.29887150740963486,
+      "learning_rate": 8.747602203290124e-06,
+      "loss": 0.5328,
+      "step": 3529
+    },
+    {
+      "epoch": 0.5797220454498799,
+      "grad_norm": 0.46824862966054376,
+      "learning_rate": 8.747447326420352e-06,
+      "loss": 0.5405,
+      "step": 3530
+    },
+    {
+      "epoch": 0.5798862726582227,
+      "grad_norm": 0.29399612928502583,
+      "learning_rate": 8.747292403419123e-06,
+      "loss": 0.5435,
+      "step": 3531
+    },
+    {
+      "epoch": 0.5800504998665654,
+      "grad_norm": 0.310842296672758,
+      "learning_rate": 8.74713743428812e-06,
+      "loss": 0.5184,
+      "step": 3532
+    },
+    {
+      "epoch": 0.5802147270749082,
+      "grad_norm": 0.4061885724828876,
+      "learning_rate": 8.746982419029025e-06,
+      "loss": 0.5321,
+      "step": 3533
+    },
+    {
+      "epoch": 0.5803789542832509,
+      "grad_norm": 0.4215721203763714,
+      "learning_rate": 8.746827357643524e-06,
+      "loss": 0.5578,
+      "step": 3534
+    },
+    {
+      "epoch": 0.5805431814915937,
+      "grad_norm": 0.320833189661272,
+      "learning_rate": 8.746672250133299e-06,
+      "loss": 0.5444,
+      "step": 3535
+    },
+    {
+      "epoch": 0.5807074086999363,
+      "grad_norm": 0.30776465075514364,
+      "learning_rate": 8.746517096500034e-06,
+      "loss": 0.554,
+      "step": 3536
+    },
+    {
+      "epoch": 0.5808716359082791,
+      "grad_norm": 0.3897926875015399,
+      "learning_rate": 8.746361896745416e-06,
+      "loss": 0.5526,
+      "step": 3537
+    },
+    {
+      "epoch": 0.5810358631166218,
+      "grad_norm": 0.31542044643027906,
+      "learning_rate": 8.74620665087113e-06,
+      "loss": 0.5212,
+      "step": 3538
+    },
+    {
+      "epoch": 0.5812000903249646,
+      "grad_norm": 0.33704459885406324,
+      "learning_rate": 8.746051358878863e-06,
+      "loss": 0.5216,
+      "step": 3539
+    },
+    {
+      "epoch": 0.5813643175333073,
+      "grad_norm": 0.3362121370815365,
+      "learning_rate": 8.745896020770298e-06,
+      "loss": 0.5601,
+      "step": 3540
+    },
+    {
+      "epoch": 0.58152854474165,
+      "grad_norm": 0.39125110963458015,
+      "learning_rate": 8.745740636547128e-06,
+      "loss": 0.5317,
+      "step": 3541
+    },
+    {
+      "epoch": 0.5816927719499928,
+      "grad_norm": 0.30491752320771665,
+      "learning_rate": 8.745585206211037e-06,
+      "loss": 0.5494,
+      "step": 3542
+    },
+    {
+      "epoch": 0.5818569991583356,
+      "grad_norm": 0.4110951329926245,
+      "learning_rate": 8.745429729763711e-06,
+      "loss": 0.5618,
+      "step": 3543
+    },
+    {
+      "epoch": 0.5820212263666783,
+      "grad_norm": 0.343279992216792,
+      "learning_rate": 8.745274207206844e-06,
+      "loss": 0.5209,
+      "step": 3544
+    },
+    {
+      "epoch": 0.582185453575021,
+      "grad_norm": 0.32229992763426724,
+      "learning_rate": 8.745118638542121e-06,
+      "loss": 0.5143,
+      "step": 3545
+    },
+    {
+      "epoch": 0.5823496807833638,
+      "grad_norm": 0.41394397944438854,
+      "learning_rate": 8.744963023771233e-06,
+      "loss": 0.5621,
+      "step": 3546
+    },
+    {
+      "epoch": 0.5825139079917065,
+      "grad_norm": 0.4164012584557086,
+      "learning_rate": 8.74480736289587e-06,
+      "loss": 0.554,
+      "step": 3547
+    },
+    {
+      "epoch": 0.5826781352000493,
+      "grad_norm": 0.3226663377624633,
+      "learning_rate": 8.744651655917724e-06,
+      "loss": 0.549,
+      "step": 3548
+    },
+    {
+      "epoch": 0.582842362408392,
+      "grad_norm": 0.37795580482924596,
+      "learning_rate": 8.744495902838483e-06,
+      "loss": 0.5176,
+      "step": 3549
+    },
+    {
+      "epoch": 0.5830065896167348,
+      "grad_norm": 0.3317217876574988,
+      "learning_rate": 8.744340103659841e-06,
+      "loss": 0.5235,
+      "step": 3550
+    },
+    {
+      "epoch": 0.5831708168250775,
+      "grad_norm": 0.32555740709280667,
+      "learning_rate": 8.74418425838349e-06,
+      "loss": 0.5547,
+      "step": 3551
+    },
+    {
+      "epoch": 0.5833350440334203,
+      "grad_norm": 0.4069952879795932,
+      "learning_rate": 8.744028367011122e-06,
+      "loss": 0.5065,
+      "step": 3552
+    },
+    {
+      "epoch": 0.5834992712417629,
+      "grad_norm": 0.32360525266398305,
+      "learning_rate": 8.74387242954443e-06,
+      "loss": 0.5353,
+      "step": 3553
+    },
+    {
+      "epoch": 0.5836634984501057,
+      "grad_norm": 0.3948501021982254,
+      "learning_rate": 8.74371644598511e-06,
+      "loss": 0.5251,
+      "step": 3554
+    },
+    {
+      "epoch": 0.5838277256584484,
+      "grad_norm": 0.3584701735245281,
+      "learning_rate": 8.743560416334852e-06,
+      "loss": 0.5521,
+      "step": 3555
+    },
+    {
+      "epoch": 0.5839919528667912,
+      "grad_norm": 0.31069610623115584,
+      "learning_rate": 8.743404340595352e-06,
+      "loss": 0.5252,
+      "step": 3556
+    },
+    {
+      "epoch": 0.5841561800751339,
+      "grad_norm": 0.3453463165092735,
+      "learning_rate": 8.74324821876831e-06,
+      "loss": 0.5388,
+      "step": 3557
+    },
+    {
+      "epoch": 0.5843204072834767,
+      "grad_norm": 0.3017781549374568,
+      "learning_rate": 8.743092050855413e-06,
+      "loss": 0.5183,
+      "step": 3558
+    },
+    {
+      "epoch": 0.5844846344918194,
+      "grad_norm": 0.40474828712089234,
+      "learning_rate": 8.742935836858363e-06,
+      "loss": 0.5571,
+      "step": 3559
+    },
+    {
+      "epoch": 0.5846488617001622,
+      "grad_norm": 0.2922900982199023,
+      "learning_rate": 8.742779576778857e-06,
+      "loss": 0.5438,
+      "step": 3560
+    },
+    {
+      "epoch": 0.5848130889085049,
+      "grad_norm": 0.3191290542154945,
+      "learning_rate": 8.742623270618588e-06,
+      "loss": 0.4971,
+      "step": 3561
+    },
+    {
+      "epoch": 0.5849773161168477,
+      "grad_norm": 0.30540507514538917,
+      "learning_rate": 8.74246691837926e-06,
+      "loss": 0.5271,
+      "step": 3562
+    },
+    {
+      "epoch": 0.5851415433251904,
+      "grad_norm": 0.31493701820865183,
+      "learning_rate": 8.742310520062563e-06,
+      "loss": 0.5414,
+      "step": 3563
+    },
+    {
+      "epoch": 0.5853057705335332,
+      "grad_norm": 0.31388160368167506,
+      "learning_rate": 8.742154075670202e-06,
+      "loss": 0.5328,
+      "step": 3564
+    },
+    {
+      "epoch": 0.5854699977418759,
+      "grad_norm": 0.31964141957864006,
+      "learning_rate": 8.741997585203874e-06,
+      "loss": 0.5393,
+      "step": 3565
+    },
+    {
+      "epoch": 0.5856342249502187,
+      "grad_norm": 0.32701350425663833,
+      "learning_rate": 8.741841048665279e-06,
+      "loss": 0.5322,
+      "step": 3566
+    },
+    {
+      "epoch": 0.5857984521585614,
+      "grad_norm": 0.3203015136579183,
+      "learning_rate": 8.741684466056116e-06,
+      "loss": 0.5055,
+      "step": 3567
+    },
+    {
+      "epoch": 0.5859626793669042,
+      "grad_norm": 0.3208442953143098,
+      "learning_rate": 8.741527837378086e-06,
+      "loss": 0.5459,
+      "step": 3568
+    },
+    {
+      "epoch": 0.5861269065752469,
+      "grad_norm": 0.37630098391133593,
+      "learning_rate": 8.74137116263289e-06,
+      "loss": 0.5428,
+      "step": 3569
+    },
+    {
+      "epoch": 0.5862911337835895,
+      "grad_norm": 0.299680115481499,
+      "learning_rate": 8.741214441822231e-06,
+      "loss": 0.541,
+      "step": 3570
+    },
+    {
+      "epoch": 0.5864553609919323,
+      "grad_norm": 0.3896245426246468,
+      "learning_rate": 8.741057674947812e-06,
+      "loss": 0.5301,
+      "step": 3571
+    },
+    {
+      "epoch": 0.586619588200275,
+      "grad_norm": 0.28228508118203716,
+      "learning_rate": 8.740900862011332e-06,
+      "loss": 0.5237,
+      "step": 3572
+    },
+    {
+      "epoch": 0.5867838154086178,
+      "grad_norm": 0.26998847124009234,
+      "learning_rate": 8.740744003014498e-06,
+      "loss": 0.5622,
+      "step": 3573
+    },
+    {
+      "epoch": 0.5869480426169605,
+      "grad_norm": 0.31146496471387847,
+      "learning_rate": 8.74058709795901e-06,
+      "loss": 0.5439,
+      "step": 3574
+    },
+    {
+      "epoch": 0.5871122698253033,
+      "grad_norm": 0.35384476669187864,
+      "learning_rate": 8.740430146846576e-06,
+      "loss": 0.5403,
+      "step": 3575
+    },
+    {
+      "epoch": 0.587276497033646,
+      "grad_norm": 0.3270939587667193,
+      "learning_rate": 8.740273149678897e-06,
+      "loss": 0.53,
+      "step": 3576
+    },
+    {
+      "epoch": 0.5874407242419888,
+      "grad_norm": 0.29676668095238884,
+      "learning_rate": 8.74011610645768e-06,
+      "loss": 0.5135,
+      "step": 3577
+    },
+    {
+      "epoch": 0.5876049514503315,
+      "grad_norm": 0.28217328906910305,
+      "learning_rate": 8.739959017184629e-06,
+      "loss": 0.5421,
+      "step": 3578
+    },
+    {
+      "epoch": 0.5877691786586743,
+      "grad_norm": 0.26291304157258816,
+      "learning_rate": 8.739801881861453e-06,
+      "loss": 0.5278,
+      "step": 3579
+    },
+    {
+      "epoch": 0.587933405867017,
+      "grad_norm": 0.3132802190163169,
+      "learning_rate": 8.739644700489856e-06,
+      "loss": 0.5206,
+      "step": 3580
+    },
+    {
+      "epoch": 0.5880976330753598,
+      "grad_norm": 0.3025007721387032,
+      "learning_rate": 8.73948747307155e-06,
+      "loss": 0.532,
+      "step": 3581
+    },
+    {
+      "epoch": 0.5882618602837025,
+      "grad_norm": 0.26167893751892235,
+      "learning_rate": 8.739330199608235e-06,
+      "loss": 0.5501,
+      "step": 3582
+    },
+    {
+      "epoch": 0.5884260874920453,
+      "grad_norm": 0.2802592450047528,
+      "learning_rate": 8.739172880101624e-06,
+      "loss": 0.5373,
+      "step": 3583
+    },
+    {
+      "epoch": 0.588590314700388,
+      "grad_norm": 0.3227658244340501,
+      "learning_rate": 8.739015514553425e-06,
+      "loss": 0.5564,
+      "step": 3584
+    },
+    {
+      "epoch": 0.5887545419087308,
+      "grad_norm": 0.31112797368771505,
+      "learning_rate": 8.738858102965348e-06,
+      "loss": 0.5329,
+      "step": 3585
+    },
+    {
+      "epoch": 0.5889187691170735,
+      "grad_norm": 0.36689048311245726,
+      "learning_rate": 8.7387006453391e-06,
+      "loss": 0.534,
+      "step": 3586
+    },
+    {
+      "epoch": 0.5890829963254162,
+      "grad_norm": 0.30855317467713334,
+      "learning_rate": 8.738543141676393e-06,
+      "loss": 0.5394,
+      "step": 3587
+    },
+    {
+      "epoch": 0.5892472235337589,
+      "grad_norm": 0.35880633381663496,
+      "learning_rate": 8.738385591978936e-06,
+      "loss": 0.5273,
+      "step": 3588
+    },
+    {
+      "epoch": 0.5894114507421017,
+      "grad_norm": 0.3353793135243243,
+      "learning_rate": 8.738227996248444e-06,
+      "loss": 0.5352,
+      "step": 3589
+    },
+    {
+      "epoch": 0.5895756779504444,
+      "grad_norm": 0.29390223003626537,
+      "learning_rate": 8.738070354486626e-06,
+      "loss": 0.553,
+      "step": 3590
+    },
+    {
+      "epoch": 0.5897399051587872,
+      "grad_norm": 0.27791226239243794,
+      "learning_rate": 8.737912666695192e-06,
+      "loss": 0.5389,
+      "step": 3591
+    },
+    {
+      "epoch": 0.5899041323671299,
+      "grad_norm": 0.27245500324821437,
+      "learning_rate": 8.73775493287586e-06,
+      "loss": 0.5266,
+      "step": 3592
+    },
+    {
+      "epoch": 0.5900683595754727,
+      "grad_norm": 0.8524051814576923,
+      "learning_rate": 8.737597153030338e-06,
+      "loss": 0.5444,
+      "step": 3593
+    },
+    {
+      "epoch": 0.5902325867838154,
+      "grad_norm": 0.3002304556760417,
+      "learning_rate": 8.73743932716034e-06,
+      "loss": 0.5337,
+      "step": 3594
+    },
+    {
+      "epoch": 0.5903968139921582,
+      "grad_norm": 0.30605452136769806,
+      "learning_rate": 8.737281455267585e-06,
+      "loss": 0.5386,
+      "step": 3595
+    },
+    {
+      "epoch": 0.5905610412005009,
+      "grad_norm": 0.34386930237760743,
+      "learning_rate": 8.737123537353783e-06,
+      "loss": 0.5194,
+      "step": 3596
+    },
+    {
+      "epoch": 0.5907252684088437,
+      "grad_norm": 0.3312859202395492,
+      "learning_rate": 8.73696557342065e-06,
+      "loss": 0.5592,
+      "step": 3597
+    },
+    {
+      "epoch": 0.5908894956171864,
+      "grad_norm": 0.3658728879871337,
+      "learning_rate": 8.736807563469905e-06,
+      "loss": 0.5572,
+      "step": 3598
+    },
+    {
+      "epoch": 0.5910537228255291,
+      "grad_norm": 0.2881831918233398,
+      "learning_rate": 8.736649507503257e-06,
+      "loss": 0.5371,
+      "step": 3599
+    },
+    {
+      "epoch": 0.5912179500338719,
+      "grad_norm": 0.36645575351131854,
+      "learning_rate": 8.736491405522431e-06,
+      "loss": 0.5454,
+      "step": 3600
+    },
+    {
+      "epoch": 0.5913821772422146,
+      "grad_norm": 0.3179909063540978,
+      "learning_rate": 8.73633325752914e-06,
+      "loss": 0.551,
+      "step": 3601
+    },
+    {
+      "epoch": 0.5915464044505574,
+      "grad_norm": 0.38250970177758387,
+      "learning_rate": 8.7361750635251e-06,
+      "loss": 0.5431,
+      "step": 3602
+    },
+    {
+      "epoch": 0.5917106316589001,
+      "grad_norm": 0.2894129002667864,
+      "learning_rate": 8.736016823512031e-06,
+      "loss": 0.5438,
+      "step": 3603
+    },
+    {
+      "epoch": 0.5918748588672428,
+      "grad_norm": 0.2894750736546933,
+      "learning_rate": 8.735858537491652e-06,
+      "loss": 0.5393,
+      "step": 3604
+    },
+    {
+      "epoch": 0.5920390860755855,
+      "grad_norm": 0.3680922889819331,
+      "learning_rate": 8.735700205465683e-06,
+      "loss": 0.5356,
+      "step": 3605
+    },
+    {
+      "epoch": 0.5922033132839283,
+      "grad_norm": 0.3025193777662114,
+      "learning_rate": 8.73554182743584e-06,
+      "loss": 0.5447,
+      "step": 3606
+    },
+    {
+      "epoch": 0.592367540492271,
+      "grad_norm": 0.3136810875159733,
+      "learning_rate": 8.735383403403849e-06,
+      "loss": 0.5303,
+      "step": 3607
+    },
+    {
+      "epoch": 0.5925317677006138,
+      "grad_norm": 0.38311420078716235,
+      "learning_rate": 8.735224933371423e-06,
+      "loss": 0.5266,
+      "step": 3608
+    },
+    {
+      "epoch": 0.5926959949089565,
+      "grad_norm": 0.28233496253123336,
+      "learning_rate": 8.73506641734029e-06,
+      "loss": 0.5449,
+      "step": 3609
+    },
+    {
+      "epoch": 0.5928602221172993,
+      "grad_norm": 0.3227273047099139,
+      "learning_rate": 8.734907855312168e-06,
+      "loss": 0.5411,
+      "step": 3610
+    },
+    {
+      "epoch": 0.593024449325642,
+      "grad_norm": 0.27668525367867175,
+      "learning_rate": 8.734749247288782e-06,
+      "loss": 0.5377,
+      "step": 3611
+    },
+    {
+      "epoch": 0.5931886765339848,
+      "grad_norm": 0.28813590933697675,
+      "learning_rate": 8.734590593271851e-06,
+      "loss": 0.5364,
+      "step": 3612
+    },
+    {
+      "epoch": 0.5933529037423275,
+      "grad_norm": 0.3242787415516203,
+      "learning_rate": 8.7344318932631e-06,
+      "loss": 0.5472,
+      "step": 3613
+    },
+    {
+      "epoch": 0.5935171309506703,
+      "grad_norm": 0.30987179115108054,
+      "learning_rate": 8.734273147264252e-06,
+      "loss": 0.5264,
+      "step": 3614
+    },
+    {
+      "epoch": 0.593681358159013,
+      "grad_norm": 0.28439528277004245,
+      "learning_rate": 8.734114355277033e-06,
+      "loss": 0.5309,
+      "step": 3615
+    },
+    {
+      "epoch": 0.5938455853673558,
+      "grad_norm": 0.309183288887301,
+      "learning_rate": 8.733955517303165e-06,
+      "loss": 0.5451,
+      "step": 3616
+    },
+    {
+      "epoch": 0.5940098125756985,
+      "grad_norm": 0.2947685618065334,
+      "learning_rate": 8.733796633344375e-06,
+      "loss": 0.5292,
+      "step": 3617
+    },
+    {
+      "epoch": 0.5941740397840413,
+      "grad_norm": 0.3046893076201151,
+      "learning_rate": 8.733637703402387e-06,
+      "loss": 0.5221,
+      "step": 3618
+    },
+    {
+      "epoch": 0.594338266992384,
+      "grad_norm": 0.32928639712865154,
+      "learning_rate": 8.733478727478931e-06,
+      "loss": 0.528,
+      "step": 3619
+    },
+    {
+      "epoch": 0.5945024942007266,
+      "grad_norm": 0.4061734180068937,
+      "learning_rate": 8.733319705575728e-06,
+      "loss": 0.5405,
+      "step": 3620
+    },
+    {
+      "epoch": 0.5946667214090694,
+      "grad_norm": 0.2870014695047263,
+      "learning_rate": 8.73316063769451e-06,
+      "loss": 0.5378,
+      "step": 3621
+    },
+    {
+      "epoch": 0.5948309486174121,
+      "grad_norm": 0.35349649731908983,
+      "learning_rate": 8.733001523837003e-06,
+      "loss": 0.5466,
+      "step": 3622
+    },
+    {
+      "epoch": 0.5949951758257549,
+      "grad_norm": 0.3189730804637844,
+      "learning_rate": 8.732842364004932e-06,
+      "loss": 0.5431,
+      "step": 3623
+    },
+    {
+      "epoch": 0.5951594030340976,
+      "grad_norm": 0.3457512242859419,
+      "learning_rate": 8.73268315820003e-06,
+      "loss": 0.553,
+      "step": 3624
+    },
+    {
+      "epoch": 0.5953236302424404,
+      "grad_norm": 0.40936712834610717,
+      "learning_rate": 8.732523906424025e-06,
+      "loss": 0.5194,
+      "step": 3625
+    },
+    {
+      "epoch": 0.5954878574507831,
+      "grad_norm": 0.332203955122217,
+      "learning_rate": 8.732364608678644e-06,
+      "loss": 0.5539,
+      "step": 3626
+    },
+    {
+      "epoch": 0.5956520846591259,
+      "grad_norm": 0.2911967123741656,
+      "learning_rate": 8.732205264965622e-06,
+      "loss": 0.5376,
+      "step": 3627
+    },
+    {
+      "epoch": 0.5958163118674686,
+      "grad_norm": 0.287762468996504,
+      "learning_rate": 8.732045875286685e-06,
+      "loss": 0.5344,
+      "step": 3628
+    },
+    {
+      "epoch": 0.5959805390758114,
+      "grad_norm": 0.34854621639857275,
+      "learning_rate": 8.731886439643566e-06,
+      "loss": 0.5423,
+      "step": 3629
+    },
+    {
+      "epoch": 0.5961447662841541,
+      "grad_norm": 0.3698751509326475,
+      "learning_rate": 8.731726958037998e-06,
+      "loss": 0.529,
+      "step": 3630
+    },
+    {
+      "epoch": 0.5963089934924969,
+      "grad_norm": 0.27934921610336444,
+      "learning_rate": 8.731567430471711e-06,
+      "loss": 0.5443,
+      "step": 3631
+    },
+    {
+      "epoch": 0.5964732207008396,
+      "grad_norm": 0.3193564893222012,
+      "learning_rate": 8.731407856946438e-06,
+      "loss": 0.5296,
+      "step": 3632
+    },
+    {
+      "epoch": 0.5966374479091824,
+      "grad_norm": 0.30388255767050054,
+      "learning_rate": 8.731248237463913e-06,
+      "loss": 0.5581,
+      "step": 3633
+    },
+    {
+      "epoch": 0.5968016751175251,
+      "grad_norm": 0.34394183718979676,
+      "learning_rate": 8.73108857202587e-06,
+      "loss": 0.5245,
+      "step": 3634
+    },
+    {
+      "epoch": 0.5969659023258679,
+      "grad_norm": 0.3368222367193206,
+      "learning_rate": 8.730928860634041e-06,
+      "loss": 0.5364,
+      "step": 3635
+    },
+    {
+      "epoch": 0.5971301295342106,
+      "grad_norm": 0.6386666188329989,
+      "learning_rate": 8.730769103290162e-06,
+      "loss": 0.5315,
+      "step": 3636
+    },
+    {
+      "epoch": 0.5972943567425533,
+      "grad_norm": 0.39752248221935366,
+      "learning_rate": 8.73060929999597e-06,
+      "loss": 0.5453,
+      "step": 3637
+    },
+    {
+      "epoch": 0.597458583950896,
+      "grad_norm": 0.32135827832074965,
+      "learning_rate": 8.730449450753197e-06,
+      "loss": 0.5373,
+      "step": 3638
+    },
+    {
+      "epoch": 0.5976228111592388,
+      "grad_norm": 0.2997430977323944,
+      "learning_rate": 8.73028955556358e-06,
+      "loss": 0.5314,
+      "step": 3639
+    },
+    {
+      "epoch": 0.5977870383675815,
+      "grad_norm": 0.3172118412359191,
+      "learning_rate": 8.730129614428858e-06,
+      "loss": 0.5356,
+      "step": 3640
+    },
+    {
+      "epoch": 0.5979512655759243,
+      "grad_norm": 0.2722040870504818,
+      "learning_rate": 8.729969627350766e-06,
+      "loss": 0.5503,
+      "step": 3641
+    },
+    {
+      "epoch": 0.598115492784267,
+      "grad_norm": 0.29389145308594455,
+      "learning_rate": 8.72980959433104e-06,
+      "loss": 0.5079,
+      "step": 3642
+    },
+    {
+      "epoch": 0.5982797199926098,
+      "grad_norm": 0.2885264804569022,
+      "learning_rate": 8.729649515371423e-06,
+      "loss": 0.556,
+      "step": 3643
+    },
+    {
+      "epoch": 0.5984439472009525,
+      "grad_norm": 0.3008793211955447,
+      "learning_rate": 8.729489390473649e-06,
+      "loss": 0.5529,
+      "step": 3644
+    },
+    {
+      "epoch": 0.5986081744092953,
+      "grad_norm": 0.31658137204584463,
+      "learning_rate": 8.729329219639462e-06,
+      "loss": 0.5335,
+      "step": 3645
+    },
+    {
+      "epoch": 0.598772401617638,
+      "grad_norm": 0.34838847739756235,
+      "learning_rate": 8.729169002870596e-06,
+      "loss": 0.5502,
+      "step": 3646
+    },
+    {
+      "epoch": 0.5989366288259808,
+      "grad_norm": 0.3756933229117239,
+      "learning_rate": 8.729008740168793e-06,
+      "loss": 0.5294,
+      "step": 3647
+    },
+    {
+      "epoch": 0.5991008560343235,
+      "grad_norm": 0.34519823363513086,
+      "learning_rate": 8.728848431535795e-06,
+      "loss": 0.5284,
+      "step": 3648
+    },
+    {
+      "epoch": 0.5992650832426663,
+      "grad_norm": 0.3167375361570294,
+      "learning_rate": 8.728688076973344e-06,
+      "loss": 0.5358,
+      "step": 3649
+    },
+    {
+      "epoch": 0.599429310451009,
+      "grad_norm": 0.3306748583479528,
+      "learning_rate": 8.728527676483178e-06,
+      "loss": 0.5293,
+      "step": 3650
+    },
+    {
+      "epoch": 0.5995935376593517,
+      "grad_norm": 0.3708640363221295,
+      "learning_rate": 8.728367230067043e-06,
+      "loss": 0.5427,
+      "step": 3651
+    },
+    {
+      "epoch": 0.5997577648676945,
+      "grad_norm": 0.3431078530708263,
+      "learning_rate": 8.728206737726678e-06,
+      "loss": 0.5445,
+      "step": 3652
+    },
+    {
+      "epoch": 0.5999219920760372,
+      "grad_norm": 0.32219217359231533,
+      "learning_rate": 8.728046199463829e-06,
+      "loss": 0.5324,
+      "step": 3653
+    },
+    {
+      "epoch": 0.6000862192843799,
+      "grad_norm": 0.32057168261411606,
+      "learning_rate": 8.727885615280237e-06,
+      "loss": 0.5547,
+      "step": 3654
+    },
+    {
+      "epoch": 0.6002504464927226,
+      "grad_norm": 0.2767525924131096,
+      "learning_rate": 8.72772498517765e-06,
+      "loss": 0.5452,
+      "step": 3655
+    },
+    {
+      "epoch": 0.6004146737010654,
+      "grad_norm": 0.326894122884887,
+      "learning_rate": 8.727564309157807e-06,
+      "loss": 0.5449,
+      "step": 3656
+    },
+    {
+      "epoch": 0.6005789009094081,
+      "grad_norm": 0.3048109582358382,
+      "learning_rate": 8.727403587222457e-06,
+      "loss": 0.541,
+      "step": 3657
+    },
+    {
+      "epoch": 0.6007431281177509,
+      "grad_norm": 0.3520073738268126,
+      "learning_rate": 8.727242819373347e-06,
+      "loss": 0.5617,
+      "step": 3658
+    },
+    {
+      "epoch": 0.6009073553260936,
+      "grad_norm": 0.3499362485696832,
+      "learning_rate": 8.72708200561222e-06,
+      "loss": 0.546,
+      "step": 3659
+    },
+    {
+      "epoch": 0.6010715825344364,
+      "grad_norm": 0.29907581210466505,
+      "learning_rate": 8.726921145940824e-06,
+      "loss": 0.5367,
+      "step": 3660
+    },
+    {
+      "epoch": 0.6012358097427791,
+      "grad_norm": 0.35970980950843484,
+      "learning_rate": 8.726760240360904e-06,
+      "loss": 0.5456,
+      "step": 3661
+    },
+    {
+      "epoch": 0.6014000369511219,
+      "grad_norm": 0.3035374882855592,
+      "learning_rate": 8.726599288874211e-06,
+      "loss": 0.4972,
+      "step": 3662
+    },
+    {
+      "epoch": 0.6015642641594646,
+      "grad_norm": 0.3077833132019388,
+      "learning_rate": 8.72643829148249e-06,
+      "loss": 0.5381,
+      "step": 3663
+    },
+    {
+      "epoch": 0.6017284913678074,
+      "grad_norm": 0.27870036599180303,
+      "learning_rate": 8.726277248187491e-06,
+      "loss": 0.5331,
+      "step": 3664
+    },
+    {
+      "epoch": 0.6018927185761501,
+      "grad_norm": 0.3980801030680251,
+      "learning_rate": 8.726116158990964e-06,
+      "loss": 0.5534,
+      "step": 3665
+    },
+    {
+      "epoch": 0.6020569457844929,
+      "grad_norm": 0.306821683865838,
+      "learning_rate": 8.725955023894657e-06,
+      "loss": 0.5404,
+      "step": 3666
+    },
+    {
+      "epoch": 0.6022211729928356,
+      "grad_norm": 0.30355098474208053,
+      "learning_rate": 8.725793842900319e-06,
+      "loss": 0.5446,
+      "step": 3667
+    },
+    {
+      "epoch": 0.6023854002011784,
+      "grad_norm": 0.32484342105558917,
+      "learning_rate": 8.725632616009704e-06,
+      "loss": 0.5136,
+      "step": 3668
+    },
+    {
+      "epoch": 0.6025496274095211,
+      "grad_norm": 0.35942539543304525,
+      "learning_rate": 8.725471343224562e-06,
+      "loss": 0.5522,
+      "step": 3669
+    },
+    {
+      "epoch": 0.6027138546178639,
+      "grad_norm": 0.35114276088233815,
+      "learning_rate": 8.725310024546642e-06,
+      "loss": 0.5439,
+      "step": 3670
+    },
+    {
+      "epoch": 0.6028780818262065,
+      "grad_norm": 0.33071150493599055,
+      "learning_rate": 8.7251486599777e-06,
+      "loss": 0.5394,
+      "step": 3671
+    },
+    {
+      "epoch": 0.6030423090345493,
+      "grad_norm": 0.2914014429336388,
+      "learning_rate": 8.724987249519485e-06,
+      "loss": 0.5283,
+      "step": 3672
+    },
+    {
+      "epoch": 0.603206536242892,
+      "grad_norm": 0.2987885928184587,
+      "learning_rate": 8.724825793173752e-06,
+      "loss": 0.5484,
+      "step": 3673
+    },
+    {
+      "epoch": 0.6033707634512347,
+      "grad_norm": 0.35620845704867954,
+      "learning_rate": 8.724664290942254e-06,
+      "loss": 0.5232,
+      "step": 3674
+    },
+    {
+      "epoch": 0.6035349906595775,
+      "grad_norm": 0.8502964437012199,
+      "learning_rate": 8.724502742826746e-06,
+      "loss": 0.5313,
+      "step": 3675
+    },
+    {
+      "epoch": 0.6036992178679202,
+      "grad_norm": 0.35180116837141656,
+      "learning_rate": 8.724341148828982e-06,
+      "loss": 0.5405,
+      "step": 3676
+    },
+    {
+      "epoch": 0.603863445076263,
+      "grad_norm": 0.3428233815348743,
+      "learning_rate": 8.724179508950717e-06,
+      "loss": 0.5556,
+      "step": 3677
+    },
+    {
+      "epoch": 0.6040276722846057,
+      "grad_norm": 0.35078504166460134,
+      "learning_rate": 8.724017823193706e-06,
+      "loss": 0.5383,
+      "step": 3678
+    },
+    {
+      "epoch": 0.6041918994929485,
+      "grad_norm": 0.7235745909791813,
+      "learning_rate": 8.723856091559704e-06,
+      "loss": 0.5448,
+      "step": 3679
+    },
+    {
+      "epoch": 0.6043561267012912,
+      "grad_norm": 0.6140340695491158,
+      "learning_rate": 8.723694314050472e-06,
+      "loss": 0.5273,
+      "step": 3680
+    },
+    {
+      "epoch": 0.604520353909634,
+      "grad_norm": 0.3498517666631241,
+      "learning_rate": 8.723532490667763e-06,
+      "loss": 0.5331,
+      "step": 3681
+    },
+    {
+      "epoch": 0.6046845811179767,
+      "grad_norm": 0.32938150991708076,
+      "learning_rate": 8.723370621413335e-06,
+      "loss": 0.5391,
+      "step": 3682
+    },
+    {
+      "epoch": 0.6048488083263195,
+      "grad_norm": 0.3697450699685494,
+      "learning_rate": 8.723208706288946e-06,
+      "loss": 0.5395,
+      "step": 3683
+    },
+    {
+      "epoch": 0.6050130355346622,
+      "grad_norm": 0.32905386542315357,
+      "learning_rate": 8.723046745296357e-06,
+      "loss": 0.5269,
+      "step": 3684
+    },
+    {
+      "epoch": 0.605177262743005,
+      "grad_norm": 0.30086869065685895,
+      "learning_rate": 8.722884738437327e-06,
+      "loss": 0.5128,
+      "step": 3685
+    },
+    {
+      "epoch": 0.6053414899513477,
+      "grad_norm": 0.3113050683890726,
+      "learning_rate": 8.722722685713612e-06,
+      "loss": 0.5231,
+      "step": 3686
+    },
+    {
+      "epoch": 0.6055057171596905,
+      "grad_norm": 0.36319048353923766,
+      "learning_rate": 8.722560587126975e-06,
+      "loss": 0.5269,
+      "step": 3687
+    },
+    {
+      "epoch": 0.6056699443680331,
+      "grad_norm": 0.37114379821384086,
+      "learning_rate": 8.722398442679174e-06,
+      "loss": 0.5407,
+      "step": 3688
+    },
+    {
+      "epoch": 0.6058341715763759,
+      "grad_norm": 0.3219171912615586,
+      "learning_rate": 8.722236252371974e-06,
+      "loss": 0.5359,
+      "step": 3689
+    },
+    {
+      "epoch": 0.6059983987847186,
+      "grad_norm": 0.34569270952073616,
+      "learning_rate": 8.722074016207131e-06,
+      "loss": 0.5413,
+      "step": 3690
+    },
+    {
+      "epoch": 0.6061626259930614,
+      "grad_norm": 0.34523412105687357,
+      "learning_rate": 8.721911734186412e-06,
+      "loss": 0.5198,
+      "step": 3691
+    },
+    {
+      "epoch": 0.6063268532014041,
+      "grad_norm": 0.35292293229945165,
+      "learning_rate": 8.721749406311578e-06,
+      "loss": 0.5664,
+      "step": 3692
+    },
+    {
+      "epoch": 0.6064910804097469,
+      "grad_norm": 0.39140784378450616,
+      "learning_rate": 8.721587032584391e-06,
+      "loss": 0.5174,
+      "step": 3693
+    },
+    {
+      "epoch": 0.6066553076180896,
+      "grad_norm": 0.29287997907577673,
+      "learning_rate": 8.721424613006616e-06,
+      "loss": 0.5242,
+      "step": 3694
+    },
+    {
+      "epoch": 0.6068195348264324,
+      "grad_norm": 0.32102245081740655,
+      "learning_rate": 8.721262147580016e-06,
+      "loss": 0.5078,
+      "step": 3695
+    },
+    {
+      "epoch": 0.6069837620347751,
+      "grad_norm": 0.329148098587207,
+      "learning_rate": 8.721099636306357e-06,
+      "loss": 0.5179,
+      "step": 3696
+    },
+    {
+      "epoch": 0.6071479892431179,
+      "grad_norm": 0.43195330263064485,
+      "learning_rate": 8.720937079187402e-06,
+      "loss": 0.5486,
+      "step": 3697
+    },
+    {
+      "epoch": 0.6073122164514606,
+      "grad_norm": 0.2719468750025907,
+      "learning_rate": 8.720774476224918e-06,
+      "loss": 0.519,
+      "step": 3698
+    },
+    {
+      "epoch": 0.6074764436598034,
+      "grad_norm": 0.335292603650919,
+      "learning_rate": 8.72061182742067e-06,
+      "loss": 0.5406,
+      "step": 3699
+    },
+    {
+      "epoch": 0.6076406708681461,
+      "grad_norm": 0.3269121070823391,
+      "learning_rate": 8.720449132776424e-06,
+      "loss": 0.523,
+      "step": 3700
+    },
+    {
+      "epoch": 0.6078048980764889,
+      "grad_norm": 0.35219972464970845,
+      "learning_rate": 8.72028639229395e-06,
+      "loss": 0.5228,
+      "step": 3701
+    },
+    {
+      "epoch": 0.6079691252848316,
+      "grad_norm": 0.3391824315436888,
+      "learning_rate": 8.720123605975012e-06,
+      "loss": 0.5416,
+      "step": 3702
+    },
+    {
+      "epoch": 0.6081333524931744,
+      "grad_norm": 0.2759373012282259,
+      "learning_rate": 8.71996077382138e-06,
+      "loss": 0.5291,
+      "step": 3703
+    },
+    {
+      "epoch": 0.6082975797015171,
+      "grad_norm": 0.29854298935572593,
+      "learning_rate": 8.719797895834823e-06,
+      "loss": 0.5513,
+      "step": 3704
+    },
+    {
+      "epoch": 0.6084618069098597,
+      "grad_norm": 0.29485850608627,
+      "learning_rate": 8.719634972017109e-06,
+      "loss": 0.5346,
+      "step": 3705
+    },
+    {
+      "epoch": 0.6086260341182025,
+      "grad_norm": 0.29127554479034234,
+      "learning_rate": 8.719472002370007e-06,
+      "loss": 0.5288,
+      "step": 3706
+    },
+    {
+      "epoch": 0.6087902613265452,
+      "grad_norm": 0.35738849044273974,
+      "learning_rate": 8.719308986895288e-06,
+      "loss": 0.5316,
+      "step": 3707
+    },
+    {
+      "epoch": 0.608954488534888,
+      "grad_norm": 0.32317977559171934,
+      "learning_rate": 8.719145925594722e-06,
+      "loss": 0.5285,
+      "step": 3708
+    },
+    {
+      "epoch": 0.6091187157432307,
+      "grad_norm": 0.323539384053774,
+      "learning_rate": 8.718982818470081e-06,
+      "loss": 0.5224,
+      "step": 3709
+    },
+    {
+      "epoch": 0.6092829429515735,
+      "grad_norm": 0.3306250256102783,
+      "learning_rate": 8.718819665523135e-06,
+      "loss": 0.5515,
+      "step": 3710
+    },
+    {
+      "epoch": 0.6094471701599162,
+      "grad_norm": 0.3057541843863994,
+      "learning_rate": 8.718656466755657e-06,
+      "loss": 0.5364,
+      "step": 3711
+    },
+    {
+      "epoch": 0.609611397368259,
+      "grad_norm": 0.32189324965811217,
+      "learning_rate": 8.718493222169417e-06,
+      "loss": 0.56,
+      "step": 3712
+    },
+    {
+      "epoch": 0.6097756245766017,
+      "grad_norm": 0.2930020470373635,
+      "learning_rate": 8.718329931766193e-06,
+      "loss": 0.5491,
+      "step": 3713
+    },
+    {
+      "epoch": 0.6099398517849445,
+      "grad_norm": 0.37017958751656427,
+      "learning_rate": 8.718166595547755e-06,
+      "loss": 0.5273,
+      "step": 3714
+    },
+    {
+      "epoch": 0.6101040789932872,
+      "grad_norm": 0.38151259250123476,
+      "learning_rate": 8.718003213515876e-06,
+      "loss": 0.5495,
+      "step": 3715
+    },
+    {
+      "epoch": 0.61026830620163,
+      "grad_norm": 0.2899120782134563,
+      "learning_rate": 8.717839785672334e-06,
+      "loss": 0.5387,
+      "step": 3716
+    },
+    {
+      "epoch": 0.6104325334099727,
+      "grad_norm": 0.3129274447525849,
+      "learning_rate": 8.7176763120189e-06,
+      "loss": 0.5236,
+      "step": 3717
+    },
+    {
+      "epoch": 0.6105967606183155,
+      "grad_norm": 0.3245883613657934,
+      "learning_rate": 8.717512792557355e-06,
+      "loss": 0.5265,
+      "step": 3718
+    },
+    {
+      "epoch": 0.6107609878266582,
+      "grad_norm": 0.3046206158653733,
+      "learning_rate": 8.71734922728947e-06,
+      "loss": 0.5493,
+      "step": 3719
+    },
+    {
+      "epoch": 0.610925215035001,
+      "grad_norm": 0.3414330525019781,
+      "learning_rate": 8.717185616217022e-06,
+      "loss": 0.5487,
+      "step": 3720
+    },
+    {
+      "epoch": 0.6110894422433437,
+      "grad_norm": 0.30157466649864123,
+      "learning_rate": 8.71702195934179e-06,
+      "loss": 0.5214,
+      "step": 3721
+    },
+    {
+      "epoch": 0.6112536694516864,
+      "grad_norm": 0.2939600687346926,
+      "learning_rate": 8.71685825666555e-06,
+      "loss": 0.534,
+      "step": 3722
+    },
+    {
+      "epoch": 0.6114178966600291,
+      "grad_norm": 0.3532357313126545,
+      "learning_rate": 8.716694508190081e-06,
+      "loss": 0.522,
+      "step": 3723
+    },
+    {
+      "epoch": 0.6115821238683719,
+      "grad_norm": 0.2888952148557769,
+      "learning_rate": 8.716530713917162e-06,
+      "loss": 0.5391,
+      "step": 3724
+    },
+    {
+      "epoch": 0.6117463510767146,
+      "grad_norm": 0.34438682455747893,
+      "learning_rate": 8.716366873848569e-06,
+      "loss": 0.5399,
+      "step": 3725
+    },
+    {
+      "epoch": 0.6119105782850573,
+      "grad_norm": 0.364981358716848,
+      "learning_rate": 8.716202987986084e-06,
+      "loss": 0.5346,
+      "step": 3726
+    },
+    {
+      "epoch": 0.6120748054934001,
+      "grad_norm": 0.4694648437859305,
+      "learning_rate": 8.716039056331487e-06,
+      "loss": 0.5249,
+      "step": 3727
+    },
+    {
+      "epoch": 0.6122390327017428,
+      "grad_norm": 0.3210375908279311,
+      "learning_rate": 8.715875078886557e-06,
+      "loss": 0.5237,
+      "step": 3728
+    },
+    {
+      "epoch": 0.6124032599100856,
+      "grad_norm": 0.34752632577085785,
+      "learning_rate": 8.715711055653077e-06,
+      "loss": 0.5405,
+      "step": 3729
+    },
+    {
+      "epoch": 0.6125674871184283,
+      "grad_norm": 0.40203728653194964,
+      "learning_rate": 8.715546986632826e-06,
+      "loss": 0.5461,
+      "step": 3730
+    },
+    {
+      "epoch": 0.6127317143267711,
+      "grad_norm": 0.3049524203199884,
+      "learning_rate": 8.715382871827587e-06,
+      "loss": 0.5383,
+      "step": 3731
+    },
+    {
+      "epoch": 0.6128959415351138,
+      "grad_norm": 0.43035534035034767,
+      "learning_rate": 8.715218711239143e-06,
+      "loss": 0.5377,
+      "step": 3732
+    },
+    {
+      "epoch": 0.6130601687434566,
+      "grad_norm": 0.304376981147912,
+      "learning_rate": 8.715054504869277e-06,
+      "loss": 0.5365,
+      "step": 3733
+    },
+    {
+      "epoch": 0.6132243959517993,
+      "grad_norm": 0.3352498186486189,
+      "learning_rate": 8.714890252719772e-06,
+      "loss": 0.5405,
+      "step": 3734
+    },
+    {
+      "epoch": 0.6133886231601421,
+      "grad_norm": 0.37635670121733,
+      "learning_rate": 8.714725954792413e-06,
+      "loss": 0.55,
+      "step": 3735
+    },
+    {
+      "epoch": 0.6135528503684848,
+      "grad_norm": 0.3193338992698321,
+      "learning_rate": 8.714561611088982e-06,
+      "loss": 0.5254,
+      "step": 3736
+    },
+    {
+      "epoch": 0.6137170775768276,
+      "grad_norm": 0.28862874206631534,
+      "learning_rate": 8.714397221611264e-06,
+      "loss": 0.5122,
+      "step": 3737
+    },
+    {
+      "epoch": 0.6138813047851703,
+      "grad_norm": 0.36269165943906717,
+      "learning_rate": 8.714232786361049e-06,
+      "loss": 0.551,
+      "step": 3738
+    },
+    {
+      "epoch": 0.614045531993513,
+      "grad_norm": 0.43221986090902803,
+      "learning_rate": 8.714068305340117e-06,
+      "loss": 0.5409,
+      "step": 3739
+    },
+    {
+      "epoch": 0.6142097592018557,
+      "grad_norm": 0.30739719342988214,
+      "learning_rate": 8.713903778550258e-06,
+      "loss": 0.5617,
+      "step": 3740
+    },
+    {
+      "epoch": 0.6143739864101985,
+      "grad_norm": 0.3447769130969769,
+      "learning_rate": 8.713739205993259e-06,
+      "loss": 0.5031,
+      "step": 3741
+    },
+    {
+      "epoch": 0.6145382136185412,
+      "grad_norm": 0.36911964009951104,
+      "learning_rate": 8.713574587670906e-06,
+      "loss": 0.5487,
+      "step": 3742
+    },
+    {
+      "epoch": 0.614702440826884,
+      "grad_norm": 0.2882172434812275,
+      "learning_rate": 8.713409923584986e-06,
+      "loss": 0.5187,
+      "step": 3743
+    },
+    {
+      "epoch": 0.6148666680352267,
+      "grad_norm": 0.4483918078459419,
+      "learning_rate": 8.71324521373729e-06,
+      "loss": 0.5568,
+      "step": 3744
+    },
+    {
+      "epoch": 0.6150308952435695,
+      "grad_norm": 0.3979105216794056,
+      "learning_rate": 8.713080458129606e-06,
+      "loss": 0.5288,
+      "step": 3745
+    },
+    {
+      "epoch": 0.6151951224519122,
+      "grad_norm": 0.32051012765081943,
+      "learning_rate": 8.712915656763723e-06,
+      "loss": 0.5374,
+      "step": 3746
+    },
+    {
+      "epoch": 0.615359349660255,
+      "grad_norm": 0.4134079464940972,
+      "learning_rate": 8.71275080964143e-06,
+      "loss": 0.5364,
+      "step": 3747
+    },
+    {
+      "epoch": 0.6155235768685977,
+      "grad_norm": 0.3025117236117526,
+      "learning_rate": 8.71258591676452e-06,
+      "loss": 0.5315,
+      "step": 3748
+    },
+    {
+      "epoch": 0.6156878040769405,
+      "grad_norm": 0.5045071173403394,
+      "learning_rate": 8.71242097813478e-06,
+      "loss": 0.517,
+      "step": 3749
+    },
+    {
+      "epoch": 0.6158520312852832,
+      "grad_norm": 0.36659414586697897,
+      "learning_rate": 8.712255993754007e-06,
+      "loss": 0.5323,
+      "step": 3750
+    },
+    {
+      "epoch": 0.616016258493626,
+      "grad_norm": 0.3413997188779241,
+      "learning_rate": 8.712090963623987e-06,
+      "loss": 0.5415,
+      "step": 3751
+    },
+    {
+      "epoch": 0.6161804857019687,
+      "grad_norm": 0.30485270030431405,
+      "learning_rate": 8.711925887746516e-06,
+      "loss": 0.5422,
+      "step": 3752
+    },
+    {
+      "epoch": 0.6163447129103115,
+      "grad_norm": 0.3781760980970153,
+      "learning_rate": 8.711760766123385e-06,
+      "loss": 0.5226,
+      "step": 3753
+    },
+    {
+      "epoch": 0.6165089401186542,
+      "grad_norm": 0.32080829914928155,
+      "learning_rate": 8.71159559875639e-06,
+      "loss": 0.5307,
+      "step": 3754
+    },
+    {
+      "epoch": 0.616673167326997,
+      "grad_norm": 0.2814514445391986,
+      "learning_rate": 8.711430385647321e-06,
+      "loss": 0.518,
+      "step": 3755
+    },
+    {
+      "epoch": 0.6168373945353396,
+      "grad_norm": 0.3194129509030034,
+      "learning_rate": 8.711265126797976e-06,
+      "loss": 0.5458,
+      "step": 3756
+    },
+    {
+      "epoch": 0.6170016217436823,
+      "grad_norm": 0.31273916859958956,
+      "learning_rate": 8.711099822210148e-06,
+      "loss": 0.529,
+      "step": 3757
+    },
+    {
+      "epoch": 0.6171658489520251,
+      "grad_norm": 0.4807127966621582,
+      "learning_rate": 8.710934471885632e-06,
+      "loss": 0.5691,
+      "step": 3758
+    },
+    {
+      "epoch": 0.6173300761603678,
+      "grad_norm": 0.35464260050427854,
+      "learning_rate": 8.710769075826227e-06,
+      "loss": 0.5455,
+      "step": 3759
+    },
+    {
+      "epoch": 0.6174943033687106,
+      "grad_norm": 0.33236503232750314,
+      "learning_rate": 8.710603634033725e-06,
+      "loss": 0.5364,
+      "step": 3760
+    },
+    {
+      "epoch": 0.6176585305770533,
+      "grad_norm": 0.3464239772480392,
+      "learning_rate": 8.710438146509925e-06,
+      "loss": 0.5267,
+      "step": 3761
+    },
+    {
+      "epoch": 0.6178227577853961,
+      "grad_norm": 0.3150829346496259,
+      "learning_rate": 8.710272613256623e-06,
+      "loss": 0.5238,
+      "step": 3762
+    },
+    {
+      "epoch": 0.6179869849937388,
+      "grad_norm": 0.33797103760143943,
+      "learning_rate": 8.710107034275621e-06,
+      "loss": 0.5101,
+      "step": 3763
+    },
+    {
+      "epoch": 0.6181512122020816,
+      "grad_norm": 0.3460563845624605,
+      "learning_rate": 8.709941409568712e-06,
+      "loss": 0.5402,
+      "step": 3764
+    },
+    {
+      "epoch": 0.6183154394104243,
+      "grad_norm": 0.31892852721514725,
+      "learning_rate": 8.709775739137698e-06,
+      "loss": 0.5443,
+      "step": 3765
+    },
+    {
+      "epoch": 0.6184796666187671,
+      "grad_norm": 0.33910830608309395,
+      "learning_rate": 8.709610022984379e-06,
+      "loss": 0.5288,
+      "step": 3766
+    },
+    {
+      "epoch": 0.6186438938271098,
+      "grad_norm": 0.5006130609576214,
+      "learning_rate": 8.709444261110551e-06,
+      "loss": 0.5698,
+      "step": 3767
+    },
+    {
+      "epoch": 0.6188081210354526,
+      "grad_norm": 0.34041891842889027,
+      "learning_rate": 8.70927845351802e-06,
+      "loss": 0.5306,
+      "step": 3768
+    },
+    {
+      "epoch": 0.6189723482437953,
+      "grad_norm": 0.32959334265653095,
+      "learning_rate": 8.70911260020858e-06,
+      "loss": 0.5027,
+      "step": 3769
+    },
+    {
+      "epoch": 0.6191365754521381,
+      "grad_norm": 0.33632145365800753,
+      "learning_rate": 8.708946701184038e-06,
+      "loss": 0.5622,
+      "step": 3770
+    },
+    {
+      "epoch": 0.6193008026604808,
+      "grad_norm": 0.3284090276692645,
+      "learning_rate": 8.708780756446193e-06,
+      "loss": 0.5258,
+      "step": 3771
+    },
+    {
+      "epoch": 0.6194650298688236,
+      "grad_norm": 0.27420040378984273,
+      "learning_rate": 8.70861476599685e-06,
+      "loss": 0.5217,
+      "step": 3772
+    },
+    {
+      "epoch": 0.6196292570771662,
+      "grad_norm": 0.37164897196678104,
+      "learning_rate": 8.708448729837807e-06,
+      "loss": 0.5001,
+      "step": 3773
+    },
+    {
+      "epoch": 0.619793484285509,
+      "grad_norm": 0.3198369016389374,
+      "learning_rate": 8.708282647970872e-06,
+      "loss": 0.5501,
+      "step": 3774
+    },
+    {
+      "epoch": 0.6199577114938517,
+      "grad_norm": 0.4734567044434302,
+      "learning_rate": 8.708116520397847e-06,
+      "loss": 0.5429,
+      "step": 3775
+    },
+    {
+      "epoch": 0.6201219387021945,
+      "grad_norm": 0.3675363210937101,
+      "learning_rate": 8.707950347120536e-06,
+      "loss": 0.5343,
+      "step": 3776
+    },
+    {
+      "epoch": 0.6202861659105372,
+      "grad_norm": 0.3130171423294823,
+      "learning_rate": 8.707784128140745e-06,
+      "loss": 0.5626,
+      "step": 3777
+    },
+    {
+      "epoch": 0.62045039311888,
+      "grad_norm": 0.33656999876962657,
+      "learning_rate": 8.707617863460276e-06,
+      "loss": 0.5397,
+      "step": 3778
+    },
+    {
+      "epoch": 0.6206146203272227,
+      "grad_norm": 0.378701611013499,
+      "learning_rate": 8.70745155308094e-06,
+      "loss": 0.5236,
+      "step": 3779
+    },
+    {
+      "epoch": 0.6207788475355654,
+      "grad_norm": 0.40263268075287956,
+      "learning_rate": 8.70728519700454e-06,
+      "loss": 0.5522,
+      "step": 3780
+    },
+    {
+      "epoch": 0.6209430747439082,
+      "grad_norm": 0.31609267720649964,
+      "learning_rate": 8.707118795232882e-06,
+      "loss": 0.5341,
+      "step": 3781
+    },
+    {
+      "epoch": 0.621107301952251,
+      "grad_norm": 0.33076045899366674,
+      "learning_rate": 8.706952347767776e-06,
+      "loss": 0.5141,
+      "step": 3782
+    },
+    {
+      "epoch": 0.6212715291605937,
+      "grad_norm": 0.3119390105211355,
+      "learning_rate": 8.706785854611027e-06,
+      "loss": 0.5395,
+      "step": 3783
+    },
+    {
+      "epoch": 0.6214357563689364,
+      "grad_norm": 0.32407867611523294,
+      "learning_rate": 8.706619315764446e-06,
+      "loss": 0.5663,
+      "step": 3784
+    },
+    {
+      "epoch": 0.6215999835772792,
+      "grad_norm": 0.3366621208808585,
+      "learning_rate": 8.70645273122984e-06,
+      "loss": 0.5303,
+      "step": 3785
+    },
+    {
+      "epoch": 0.6217642107856219,
+      "grad_norm": 0.334100898833225,
+      "learning_rate": 8.706286101009021e-06,
+      "loss": 0.5346,
+      "step": 3786
+    },
+    {
+      "epoch": 0.6219284379939647,
+      "grad_norm": 0.2938793811806174,
+      "learning_rate": 8.706119425103793e-06,
+      "loss": 0.5509,
+      "step": 3787
+    },
+    {
+      "epoch": 0.6220926652023074,
+      "grad_norm": 0.30688080144004615,
+      "learning_rate": 8.705952703515972e-06,
+      "loss": 0.5409,
+      "step": 3788
+    },
+    {
+      "epoch": 0.6222568924106502,
+      "grad_norm": 0.30865811818944405,
+      "learning_rate": 8.705785936247364e-06,
+      "loss": 0.5298,
+      "step": 3789
+    },
+    {
+      "epoch": 0.6224211196189928,
+      "grad_norm": 0.34419297569497387,
+      "learning_rate": 8.705619123299786e-06,
+      "loss": 0.547,
+      "step": 3790
+    },
+    {
+      "epoch": 0.6225853468273356,
+      "grad_norm": 0.36703178884640025,
+      "learning_rate": 8.705452264675045e-06,
+      "loss": 0.5413,
+      "step": 3791
+    },
+    {
+      "epoch": 0.6227495740356783,
+      "grad_norm": 0.3923267640978763,
+      "learning_rate": 8.705285360374955e-06,
+      "loss": 0.5323,
+      "step": 3792
+    },
+    {
+      "epoch": 0.6229138012440211,
+      "grad_norm": 0.33842940645185976,
+      "learning_rate": 8.705118410401329e-06,
+      "loss": 0.5302,
+      "step": 3793
+    },
+    {
+      "epoch": 0.6230780284523638,
+      "grad_norm": 0.8608072591706694,
+      "learning_rate": 8.704951414755978e-06,
+      "loss": 0.5415,
+      "step": 3794
+    },
+    {
+      "epoch": 0.6232422556607066,
+      "grad_norm": 0.2936882961967509,
+      "learning_rate": 8.704784373440719e-06,
+      "loss": 0.5353,
+      "step": 3795
+    },
+    {
+      "epoch": 0.6234064828690493,
+      "grad_norm": 0.289643466614854,
+      "learning_rate": 8.704617286457365e-06,
+      "loss": 0.5305,
+      "step": 3796
+    },
+    {
+      "epoch": 0.6235707100773921,
+      "grad_norm": 0.45597156884259565,
+      "learning_rate": 8.70445015380773e-06,
+      "loss": 0.5335,
+      "step": 3797
+    },
+    {
+      "epoch": 0.6237349372857348,
+      "grad_norm": 0.37801086646828497,
+      "learning_rate": 8.70428297549363e-06,
+      "loss": 0.5075,
+      "step": 3798
+    },
+    {
+      "epoch": 0.6238991644940776,
+      "grad_norm": 0.2812051698019951,
+      "learning_rate": 8.70411575151688e-06,
+      "loss": 0.5319,
+      "step": 3799
+    },
+    {
+      "epoch": 0.6240633917024203,
+      "grad_norm": 0.3145164176523809,
+      "learning_rate": 8.703948481879296e-06,
+      "loss": 0.5154,
+      "step": 3800
+    },
+    {
+      "epoch": 0.6242276189107631,
+      "grad_norm": 0.3293466939126531,
+      "learning_rate": 8.703781166582696e-06,
+      "loss": 0.5329,
+      "step": 3801
+    },
+    {
+      "epoch": 0.6243918461191058,
+      "grad_norm": 0.3241613646790027,
+      "learning_rate": 8.703613805628897e-06,
+      "loss": 0.5366,
+      "step": 3802
+    },
+    {
+      "epoch": 0.6245560733274486,
+      "grad_norm": 0.30109707531054103,
+      "learning_rate": 8.703446399019716e-06,
+      "loss": 0.5267,
+      "step": 3803
+    },
+    {
+      "epoch": 0.6247203005357913,
+      "grad_norm": 0.3159230311428853,
+      "learning_rate": 8.703278946756972e-06,
+      "loss": 0.5408,
+      "step": 3804
+    },
+    {
+      "epoch": 0.624884527744134,
+      "grad_norm": 0.32223834189608774,
+      "learning_rate": 8.703111448842482e-06,
+      "loss": 0.5587,
+      "step": 3805
+    },
+    {
+      "epoch": 0.6250487549524768,
+      "grad_norm": 90.67267353364257,
+      "learning_rate": 8.702943905278067e-06,
+      "loss": 0.528,
+      "step": 3806
+    },
+    {
+      "epoch": 0.6252129821608194,
+      "grad_norm": 0.43270448510011983,
+      "learning_rate": 8.702776316065547e-06,
+      "loss": 0.5133,
+      "step": 3807
+    },
+    {
+      "epoch": 0.6253772093691622,
+      "grad_norm": 0.6509586015003804,
+      "learning_rate": 8.70260868120674e-06,
+      "loss": 0.5407,
+      "step": 3808
+    },
+    {
+      "epoch": 0.6255414365775049,
+      "grad_norm": 1.4183420052563982,
+      "learning_rate": 8.702441000703468e-06,
+      "loss": 0.563,
+      "step": 3809
+    },
+    {
+      "epoch": 0.6257056637858477,
+      "grad_norm": 1.8921901405577501,
+      "learning_rate": 8.702273274557552e-06,
+      "loss": 0.5751,
+      "step": 3810
+    },
+    {
+      "epoch": 0.6258698909941904,
+      "grad_norm": 1.4172223283468877,
+      "learning_rate": 8.702105502770813e-06,
+      "loss": 0.545,
+      "step": 3811
+    },
+    {
+      "epoch": 0.6260341182025332,
+      "grad_norm": 1.3874445788703256,
+      "learning_rate": 8.701937685345076e-06,
+      "loss": 0.573,
+      "step": 3812
+    },
+    {
+      "epoch": 0.6261983454108759,
+      "grad_norm": 1.4763769040707089,
+      "learning_rate": 8.70176982228216e-06,
+      "loss": 0.5626,
+      "step": 3813
+    },
+    {
+      "epoch": 0.6263625726192187,
+      "grad_norm": 1.2332840709413573,
+      "learning_rate": 8.701601913583891e-06,
+      "loss": 0.5505,
+      "step": 3814
+    },
+    {
+      "epoch": 0.6265267998275614,
+      "grad_norm": 1.082496847586964,
+      "learning_rate": 8.70143395925209e-06,
+      "loss": 0.5308,
+      "step": 3815
+    },
+    {
+      "epoch": 0.6266910270359042,
+      "grad_norm": 1.3361911741976398,
+      "learning_rate": 8.701265959288584e-06,
+      "loss": 0.5584,
+      "step": 3816
+    },
+    {
+      "epoch": 0.6268552542442469,
+      "grad_norm": 0.9853921793041766,
+      "learning_rate": 8.701097913695196e-06,
+      "loss": 0.5561,
+      "step": 3817
+    },
+    {
+      "epoch": 0.6270194814525897,
+      "grad_norm": 0.6673756700635366,
+      "learning_rate": 8.70092982247375e-06,
+      "loss": 0.5216,
+      "step": 3818
+    },
+    {
+      "epoch": 0.6271837086609324,
+      "grad_norm": 0.7139398117801092,
+      "learning_rate": 8.700761685626074e-06,
+      "loss": 0.5466,
+      "step": 3819
+    },
+    {
+      "epoch": 0.6273479358692752,
+      "grad_norm": 0.6201194455263372,
+      "learning_rate": 8.700593503153993e-06,
+      "loss": 0.5595,
+      "step": 3820
+    },
+    {
+      "epoch": 0.6275121630776179,
+      "grad_norm": 0.6338582846012263,
+      "learning_rate": 8.700425275059334e-06,
+      "loss": 0.5416,
+      "step": 3821
+    },
+    {
+      "epoch": 0.6276763902859607,
+      "grad_norm": 0.620727987673737,
+      "learning_rate": 8.700257001343924e-06,
+      "loss": 0.5448,
+      "step": 3822
+    },
+    {
+      "epoch": 0.6278406174943034,
+      "grad_norm": 0.6572165151879532,
+      "learning_rate": 8.70008868200959e-06,
+      "loss": 0.5585,
+      "step": 3823
+    },
+    {
+      "epoch": 0.6280048447026461,
+      "grad_norm": 0.7069587357491588,
+      "learning_rate": 8.69992031705816e-06,
+      "loss": 0.5505,
+      "step": 3824
+    },
+    {
+      "epoch": 0.6281690719109888,
+      "grad_norm": 0.5887814051268679,
+      "learning_rate": 8.699751906491464e-06,
+      "loss": 0.5453,
+      "step": 3825
+    },
+    {
+      "epoch": 0.6283332991193316,
+      "grad_norm": 0.5949769393475107,
+      "learning_rate": 8.69958345031133e-06,
+      "loss": 0.5485,
+      "step": 3826
+    },
+    {
+      "epoch": 0.6284975263276743,
+      "grad_norm": 0.45559918621982776,
+      "learning_rate": 8.699414948519588e-06,
+      "loss": 0.5507,
+      "step": 3827
+    },
+    {
+      "epoch": 0.628661753536017,
+      "grad_norm": 0.5066535100919715,
+      "learning_rate": 8.699246401118067e-06,
+      "loss": 0.5456,
+      "step": 3828
+    },
+    {
+      "epoch": 0.6288259807443598,
+      "grad_norm": 0.603987704362724,
+      "learning_rate": 8.699077808108598e-06,
+      "loss": 0.5367,
+      "step": 3829
+    },
+    {
+      "epoch": 0.6289902079527026,
+      "grad_norm": 0.5753829717531779,
+      "learning_rate": 8.698909169493014e-06,
+      "loss": 0.5354,
+      "step": 3830
+    },
+    {
+      "epoch": 0.6291544351610453,
+      "grad_norm": 0.48583557088688634,
+      "learning_rate": 8.698740485273147e-06,
+      "loss": 0.5613,
+      "step": 3831
+    },
+    {
+      "epoch": 0.629318662369388,
+      "grad_norm": 0.4923545359065946,
+      "learning_rate": 8.698571755450826e-06,
+      "loss": 0.5264,
+      "step": 3832
+    },
+    {
+      "epoch": 0.6294828895777308,
+      "grad_norm": 0.5021171246250056,
+      "learning_rate": 8.698402980027884e-06,
+      "loss": 0.5536,
+      "step": 3833
+    },
+    {
+      "epoch": 0.6296471167860735,
+      "grad_norm": 0.6308522128061763,
+      "learning_rate": 8.698234159006155e-06,
+      "loss": 0.5574,
+      "step": 3834
+    },
+    {
+      "epoch": 0.6298113439944163,
+      "grad_norm": 0.5317056736086443,
+      "learning_rate": 8.698065292387474e-06,
+      "loss": 0.5362,
+      "step": 3835
+    },
+    {
+      "epoch": 0.629975571202759,
+      "grad_norm": 0.41775683407473374,
+      "learning_rate": 8.697896380173673e-06,
+      "loss": 0.5521,
+      "step": 3836
+    },
+    {
+      "epoch": 0.6301397984111018,
+      "grad_norm": 0.35903545755316413,
+      "learning_rate": 8.697727422366586e-06,
+      "loss": 0.56,
+      "step": 3837
+    },
+    {
+      "epoch": 0.6303040256194445,
+      "grad_norm": 0.38789311643013547,
+      "learning_rate": 8.69755841896805e-06,
+      "loss": 0.525,
+      "step": 3838
+    },
+    {
+      "epoch": 0.6304682528277873,
+      "grad_norm": 0.5183715290306259,
+      "learning_rate": 8.697389369979901e-06,
+      "loss": 0.5372,
+      "step": 3839
+    },
+    {
+      "epoch": 0.63063248003613,
+      "grad_norm": 0.4307370872715388,
+      "learning_rate": 8.697220275403972e-06,
+      "loss": 0.5309,
+      "step": 3840
+    },
+    {
+      "epoch": 0.6307967072444727,
+      "grad_norm": 0.3726029619618114,
+      "learning_rate": 8.697051135242103e-06,
+      "loss": 0.5386,
+      "step": 3841
+    },
+    {
+      "epoch": 0.6309609344528154,
+      "grad_norm": 0.4780072696531049,
+      "learning_rate": 8.696881949496127e-06,
+      "loss": 0.5471,
+      "step": 3842
+    },
+    {
+      "epoch": 0.6311251616611582,
+      "grad_norm": 0.34995870047191846,
+      "learning_rate": 8.696712718167884e-06,
+      "loss": 0.5402,
+      "step": 3843
+    },
+    {
+      "epoch": 0.6312893888695009,
+      "grad_norm": 0.411103673149213,
+      "learning_rate": 8.696543441259215e-06,
+      "loss": 0.5433,
+      "step": 3844
+    },
+    {
+      "epoch": 0.6314536160778437,
+      "grad_norm": 0.3587501781903258,
+      "learning_rate": 8.696374118771952e-06,
+      "loss": 0.5486,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6316178432861864,
+      "grad_norm": 0.33772335432480305,
+      "learning_rate": 8.69620475070794e-06,
+      "loss": 0.5315,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6317820704945292,
+      "grad_norm": 0.3390924628374558,
+      "learning_rate": 8.696035337069013e-06,
+      "loss": 0.5622,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6319462977028719,
+      "grad_norm": 0.4403930593852699,
+      "learning_rate": 8.695865877857015e-06,
+      "loss": 0.5013,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6321105249112147,
+      "grad_norm": 0.41806220799017596,
+      "learning_rate": 8.695696373073787e-06,
+      "loss": 0.5464,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6322747521195574,
+      "grad_norm": 0.4085362195583205,
+      "learning_rate": 8.695526822721167e-06,
+      "loss": 0.5217,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6324389793279002,
+      "grad_norm": 0.3689830215891882,
+      "learning_rate": 8.695357226800999e-06,
+      "loss": 0.5503,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6326032065362429,
+      "grad_norm": 0.3731211509083733,
+      "learning_rate": 8.695187585315122e-06,
+      "loss": 0.5305,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6327674337445857,
+      "grad_norm": 0.3781825465409617,
+      "learning_rate": 8.695017898265381e-06,
+      "loss": 0.536,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6329316609529284,
+      "grad_norm": 0.6074947040600762,
+      "learning_rate": 8.694848165653618e-06,
+      "loss": 0.5354,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6330958881612712,
+      "grad_norm": 0.33740040810301675,
+      "learning_rate": 8.694678387481678e-06,
+      "loss": 0.5216,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6332601153696139,
+      "grad_norm": 0.3720129842158853,
+      "learning_rate": 8.694508563751404e-06,
+      "loss": 0.5304,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6334243425779567,
+      "grad_norm": 0.34325179252543697,
+      "learning_rate": 8.694338694464639e-06,
+      "loss": 0.5401,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6335885697862993,
+      "grad_norm": 0.3983565060019772,
+      "learning_rate": 8.694168779623229e-06,
+      "loss": 0.5368,
+      "step": 3858
+    },
+    {
+      "epoch": 0.633752796994642,
+      "grad_norm": 0.4136266171320773,
+      "learning_rate": 8.69399881922902e-06,
+      "loss": 0.5129,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6339170242029848,
+      "grad_norm": 0.5371242412763911,
+      "learning_rate": 8.693828813283856e-06,
+      "loss": 0.5441,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6340812514113275,
+      "grad_norm": 0.3961247480341962,
+      "learning_rate": 8.693658761789587e-06,
+      "loss": 0.5676,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6342454786196703,
+      "grad_norm": 0.39617220623101806,
+      "learning_rate": 8.693488664748058e-06,
+      "loss": 0.5468,
+      "step": 3862
+    },
+    {
+      "epoch": 0.634409705828013,
+      "grad_norm": 0.40097095211622075,
+      "learning_rate": 8.693318522161114e-06,
+      "loss": 0.5169,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6345739330363558,
+      "grad_norm": 0.38894572895815444,
+      "learning_rate": 8.693148334030607e-06,
+      "loss": 0.5135,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6347381602446985,
+      "grad_norm": 0.3763156946705871,
+      "learning_rate": 8.692978100358384e-06,
+      "loss": 0.5529,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6349023874530413,
+      "grad_norm": 0.4886376067594294,
+      "learning_rate": 8.692807821146292e-06,
+      "loss": 0.5406,
+      "step": 3866
+    },
+    {
+      "epoch": 0.635066614661384,
+      "grad_norm": 0.35352227694749655,
+      "learning_rate": 8.692637496396181e-06,
+      "loss": 0.535,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6352308418697268,
+      "grad_norm": 0.4983938112065874,
+      "learning_rate": 8.692467126109904e-06,
+      "loss": 0.5408,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6353950690780695,
+      "grad_norm": 0.31829193416012663,
+      "learning_rate": 8.692296710289309e-06,
+      "loss": 0.534,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6355592962864123,
+      "grad_norm": 0.3525883829667812,
+      "learning_rate": 8.692126248936246e-06,
+      "loss": 0.5417,
+      "step": 3870
+    },
+    {
+      "epoch": 0.635723523494755,
+      "grad_norm": 0.36401514044093797,
+      "learning_rate": 8.691955742052567e-06,
+      "loss": 0.5371,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6358877507030978,
+      "grad_norm": 0.3969898970872966,
+      "learning_rate": 8.691785189640127e-06,
+      "loss": 0.5472,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6360519779114405,
+      "grad_norm": 0.33217592523461514,
+      "learning_rate": 8.691614591700774e-06,
+      "loss": 0.5308,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6362162051197833,
+      "grad_norm": 0.42246583922748293,
+      "learning_rate": 8.691443948236361e-06,
+      "loss": 0.5424,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6363804323281259,
+      "grad_norm": 0.3403711310191984,
+      "learning_rate": 8.691273259248745e-06,
+      "loss": 0.5514,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6365446595364687,
+      "grad_norm": 0.3248716492797271,
+      "learning_rate": 8.691102524739778e-06,
+      "loss": 0.5387,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6367088867448114,
+      "grad_norm": 0.4398054414005426,
+      "learning_rate": 8.690931744711313e-06,
+      "loss": 0.543,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6368731139531542,
+      "grad_norm": 0.42677973526179014,
+      "learning_rate": 8.690760919165206e-06,
+      "loss": 0.5536,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6370373411614969,
+      "grad_norm": 0.40237646554369005,
+      "learning_rate": 8.690590048103313e-06,
+      "loss": 0.5328,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6372015683698397,
+      "grad_norm": 0.2960670627284331,
+      "learning_rate": 8.690419131527489e-06,
+      "loss": 0.569,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6373657955781824,
+      "grad_norm": 0.4324403747553011,
+      "learning_rate": 8.69024816943959e-06,
+      "loss": 0.541,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6375300227865252,
+      "grad_norm": 0.3444517645451572,
+      "learning_rate": 8.690077161841473e-06,
+      "loss": 0.5449,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6376942499948679,
+      "grad_norm": 0.34750063901480577,
+      "learning_rate": 8.689906108734994e-06,
+      "loss": 0.5387,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6378584772032106,
+      "grad_norm": 0.30817723250833423,
+      "learning_rate": 8.689735010122015e-06,
+      "loss": 0.5284,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6380227044115534,
+      "grad_norm": 0.3530145511907926,
+      "learning_rate": 8.68956386600439e-06,
+      "loss": 0.5128,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6381869316198961,
+      "grad_norm": 0.40882371198815415,
+      "learning_rate": 8.68939267638398e-06,
+      "loss": 0.5395,
+      "step": 3886
+    },
+    {
+      "epoch": 0.6383511588282389,
+      "grad_norm": 0.36352754615313715,
+      "learning_rate": 8.689221441262645e-06,
+      "loss": 0.5224,
+      "step": 3887
+    },
+    {
+      "epoch": 0.6385153860365816,
+      "grad_norm": 0.2907383919549112,
+      "learning_rate": 8.689050160642242e-06,
+      "loss": 0.5336,
+      "step": 3888
+    },
+    {
+      "epoch": 0.6386796132449244,
+      "grad_norm": 0.34749893951758054,
+      "learning_rate": 8.688878834524634e-06,
+      "loss": 0.5389,
+      "step": 3889
+    },
+    {
+      "epoch": 0.6388438404532671,
+      "grad_norm": 0.40513911865191865,
+      "learning_rate": 8.688707462911679e-06,
+      "loss": 0.5158,
+      "step": 3890
+    },
+    {
+      "epoch": 0.6390080676616099,
+      "grad_norm": 0.3030479591332535,
+      "learning_rate": 8.688536045805241e-06,
+      "loss": 0.5539,
+      "step": 3891
+    },
+    {
+      "epoch": 0.6391722948699525,
+      "grad_norm": 0.33893944587583974,
+      "learning_rate": 8.688364583207181e-06,
+      "loss": 0.5337,
+      "step": 3892
+    },
+    {
+      "epoch": 0.6393365220782953,
+      "grad_norm": 0.3315789532550877,
+      "learning_rate": 8.68819307511936e-06,
+      "loss": 0.5402,
+      "step": 3893
+    },
+    {
+      "epoch": 0.639500749286638,
+      "grad_norm": 0.4070342888416627,
+      "learning_rate": 8.68802152154364e-06,
+      "loss": 0.5243,
+      "step": 3894
+    },
+    {
+      "epoch": 0.6396649764949808,
+      "grad_norm": 0.30937112791655846,
+      "learning_rate": 8.687849922481888e-06,
+      "loss": 0.5215,
+      "step": 3895
+    },
+    {
+      "epoch": 0.6398292037033235,
+      "grad_norm": 0.38064616880789337,
+      "learning_rate": 8.687678277935965e-06,
+      "loss": 0.5338,
+      "step": 3896
+    },
+    {
+      "epoch": 0.6399934309116663,
+      "grad_norm": 0.47604423697526016,
+      "learning_rate": 8.687506587907736e-06,
+      "loss": 0.5311,
+      "step": 3897
+    },
+    {
+      "epoch": 0.640157658120009,
+      "grad_norm": 0.410893019484912,
+      "learning_rate": 8.687334852399064e-06,
+      "loss": 0.5234,
+      "step": 3898
+    },
+    {
+      "epoch": 0.6403218853283518,
+      "grad_norm": 0.41715319002167833,
+      "learning_rate": 8.687163071411817e-06,
+      "loss": 0.5223,
+      "step": 3899
+    },
+    {
+      "epoch": 0.6404861125366945,
+      "grad_norm": 0.3342019449167628,
+      "learning_rate": 8.686991244947861e-06,
+      "loss": 0.5272,
+      "step": 3900
+    },
+    {
+      "epoch": 0.6406503397450373,
+      "grad_norm": 0.30869298822600383,
+      "learning_rate": 8.68681937300906e-06,
+      "loss": 0.5458,
+      "step": 3901
+    },
+    {
+      "epoch": 0.64081456695338,
+      "grad_norm": 0.4454509392864555,
+      "learning_rate": 8.68664745559728e-06,
+      "loss": 0.5255,
+      "step": 3902
+    },
+    {
+      "epoch": 0.6409787941617228,
+      "grad_norm": 0.33893845187399224,
+      "learning_rate": 8.686475492714389e-06,
+      "loss": 0.5355,
+      "step": 3903
+    },
+    {
+      "epoch": 0.6411430213700655,
+      "grad_norm": 0.36988459118997447,
+      "learning_rate": 8.686303484362257e-06,
+      "loss": 0.5415,
+      "step": 3904
+    },
+    {
+      "epoch": 0.6413072485784083,
+      "grad_norm": 0.3705192814061367,
+      "learning_rate": 8.686131430542749e-06,
+      "loss": 0.5456,
+      "step": 3905
+    },
+    {
+      "epoch": 0.641471475786751,
+      "grad_norm": 0.35173078787427914,
+      "learning_rate": 8.685959331257736e-06,
+      "loss": 0.5217,
+      "step": 3906
+    },
+    {
+      "epoch": 0.6416357029950938,
+      "grad_norm": 0.2957226125413411,
+      "learning_rate": 8.685787186509084e-06,
+      "loss": 0.533,
+      "step": 3907
+    },
+    {
+      "epoch": 0.6417999302034365,
+      "grad_norm": 0.7762910142753189,
+      "learning_rate": 8.685614996298667e-06,
+      "loss": 0.536,
+      "step": 3908
+    },
+    {
+      "epoch": 0.6419641574117791,
+      "grad_norm": 0.32793594523929287,
+      "learning_rate": 8.685442760628354e-06,
+      "loss": 0.5203,
+      "step": 3909
+    },
+    {
+      "epoch": 0.6421283846201219,
+      "grad_norm": 0.29157387769874193,
+      "learning_rate": 8.685270479500013e-06,
+      "loss": 0.5282,
+      "step": 3910
+    },
+    {
+      "epoch": 0.6422926118284646,
+      "grad_norm": 0.35421199782762086,
+      "learning_rate": 8.685098152915517e-06,
+      "loss": 0.5606,
+      "step": 3911
+    },
+    {
+      "epoch": 0.6424568390368074,
+      "grad_norm": 0.3948593371824962,
+      "learning_rate": 8.684925780876737e-06,
+      "loss": 0.5283,
+      "step": 3912
+    },
+    {
+      "epoch": 0.6426210662451501,
+      "grad_norm": 0.3382510623900333,
+      "learning_rate": 8.684753363385547e-06,
+      "loss": 0.5291,
+      "step": 3913
+    },
+    {
+      "epoch": 0.6427852934534929,
+      "grad_norm": 0.34658199700209985,
+      "learning_rate": 8.684580900443818e-06,
+      "loss": 0.5205,
+      "step": 3914
+    },
+    {
+      "epoch": 0.6429495206618356,
+      "grad_norm": 0.347988158080561,
+      "learning_rate": 8.684408392053423e-06,
+      "loss": 0.5284,
+      "step": 3915
+    },
+    {
+      "epoch": 0.6431137478701784,
+      "grad_norm": 0.3289779514425289,
+      "learning_rate": 8.684235838216237e-06,
+      "loss": 0.5491,
+      "step": 3916
+    },
+    {
+      "epoch": 0.6432779750785211,
+      "grad_norm": 0.417822963644472,
+      "learning_rate": 8.684063238934131e-06,
+      "loss": 0.5272,
+      "step": 3917
+    },
+    {
+      "epoch": 0.6434422022868639,
+      "grad_norm": 0.3819259777641144,
+      "learning_rate": 8.683890594208982e-06,
+      "loss": 0.5391,
+      "step": 3918
+    },
+    {
+      "epoch": 0.6436064294952066,
+      "grad_norm": 0.3245771331101448,
+      "learning_rate": 8.683717904042665e-06,
+      "loss": 0.5288,
+      "step": 3919
+    },
+    {
+      "epoch": 0.6437706567035494,
+      "grad_norm": 0.3232520982859475,
+      "learning_rate": 8.683545168437057e-06,
+      "loss": 0.546,
+      "step": 3920
+    },
+    {
+      "epoch": 0.6439348839118921,
+      "grad_norm": 0.3237295073427335,
+      "learning_rate": 8.683372387394031e-06,
+      "loss": 0.5479,
+      "step": 3921
+    },
+    {
+      "epoch": 0.6440991111202349,
+      "grad_norm": 0.3807149827338743,
+      "learning_rate": 8.683199560915464e-06,
+      "loss": 0.5288,
+      "step": 3922
+    },
+    {
+      "epoch": 0.6442633383285776,
+      "grad_norm": 0.3063169142380427,
+      "learning_rate": 8.683026689003236e-06,
+      "loss": 0.5211,
+      "step": 3923
+    },
+    {
+      "epoch": 0.6444275655369204,
+      "grad_norm": 0.43993293159188923,
+      "learning_rate": 8.682853771659222e-06,
+      "loss": 0.5373,
+      "step": 3924
+    },
+    {
+      "epoch": 0.6445917927452631,
+      "grad_norm": 0.3216930168200892,
+      "learning_rate": 8.6826808088853e-06,
+      "loss": 0.5395,
+      "step": 3925
+    },
+    {
+      "epoch": 0.6447560199536058,
+      "grad_norm": 0.45364275257594294,
+      "learning_rate": 8.68250780068335e-06,
+      "loss": 0.5586,
+      "step": 3926
+    },
+    {
+      "epoch": 0.6449202471619485,
+      "grad_norm": 0.34204207855157476,
+      "learning_rate": 8.682334747055251e-06,
+      "loss": 0.5261,
+      "step": 3927
+    },
+    {
+      "epoch": 0.6450844743702913,
+      "grad_norm": 0.47713121914216133,
+      "learning_rate": 8.682161648002881e-06,
+      "loss": 0.5333,
+      "step": 3928
+    },
+    {
+      "epoch": 0.645248701578634,
+      "grad_norm": 0.4187590920449625,
+      "learning_rate": 8.681988503528119e-06,
+      "loss": 0.5417,
+      "step": 3929
+    },
+    {
+      "epoch": 0.6454129287869768,
+      "grad_norm": 0.47497527294790776,
+      "learning_rate": 8.68181531363285e-06,
+      "loss": 0.5462,
+      "step": 3930
+    },
+    {
+      "epoch": 0.6455771559953195,
+      "grad_norm": 0.3535004979395706,
+      "learning_rate": 8.681642078318952e-06,
+      "loss": 0.546,
+      "step": 3931
+    },
+    {
+      "epoch": 0.6457413832036623,
+      "grad_norm": 0.359940360159505,
+      "learning_rate": 8.681468797588304e-06,
+      "loss": 0.5273,
+      "step": 3932
+    },
+    {
+      "epoch": 0.645905610412005,
+      "grad_norm": 0.3685906434372928,
+      "learning_rate": 8.681295471442793e-06,
+      "loss": 0.5349,
+      "step": 3933
+    },
+    {
+      "epoch": 0.6460698376203478,
+      "grad_norm": 0.4558508535400147,
+      "learning_rate": 8.6811220998843e-06,
+      "loss": 0.5183,
+      "step": 3934
+    },
+    {
+      "epoch": 0.6462340648286905,
+      "grad_norm": 0.44492975761566206,
+      "learning_rate": 8.680948682914706e-06,
+      "loss": 0.5257,
+      "step": 3935
+    },
+    {
+      "epoch": 0.6463982920370333,
+      "grad_norm": 0.3440786536061186,
+      "learning_rate": 8.680775220535897e-06,
+      "loss": 0.5169,
+      "step": 3936
+    },
+    {
+      "epoch": 0.646562519245376,
+      "grad_norm": 0.41451230027792896,
+      "learning_rate": 8.680601712749755e-06,
+      "loss": 0.536,
+      "step": 3937
+    },
+    {
+      "epoch": 0.6467267464537187,
+      "grad_norm": 0.34734955497252606,
+      "learning_rate": 8.680428159558167e-06,
+      "loss": 0.563,
+      "step": 3938
+    },
+    {
+      "epoch": 0.6468909736620615,
+      "grad_norm": 0.4474330702145904,
+      "learning_rate": 8.680254560963014e-06,
+      "loss": 0.5298,
+      "step": 3939
+    },
+    {
+      "epoch": 0.6470552008704042,
+      "grad_norm": 0.4773063602198883,
+      "learning_rate": 8.680080916966183e-06,
+      "loss": 0.5524,
+      "step": 3940
+    },
+    {
+      "epoch": 0.647219428078747,
+      "grad_norm": 0.3754853566580427,
+      "learning_rate": 8.679907227569562e-06,
+      "loss": 0.5538,
+      "step": 3941
+    },
+    {
+      "epoch": 0.6473836552870897,
+      "grad_norm": 0.3432933716810742,
+      "learning_rate": 8.679733492775035e-06,
+      "loss": 0.5241,
+      "step": 3942
+    },
+    {
+      "epoch": 0.6475478824954324,
+      "grad_norm": 0.4450073864802975,
+      "learning_rate": 8.679559712584492e-06,
+      "loss": 0.5283,
+      "step": 3943
+    },
+    {
+      "epoch": 0.6477121097037751,
+      "grad_norm": 0.3940289611879266,
+      "learning_rate": 8.679385886999818e-06,
+      "loss": 0.543,
+      "step": 3944
+    },
+    {
+      "epoch": 0.6478763369121179,
+      "grad_norm": 0.31865120983857287,
+      "learning_rate": 8.6792120160229e-06,
+      "loss": 0.5276,
+      "step": 3945
+    },
+    {
+      "epoch": 0.6480405641204606,
+      "grad_norm": 0.37514316444909035,
+      "learning_rate": 8.679038099655629e-06,
+      "loss": 0.5283,
+      "step": 3946
+    },
+    {
+      "epoch": 0.6482047913288034,
+      "grad_norm": 0.38267915121977825,
+      "learning_rate": 8.678864137899892e-06,
+      "loss": 0.5355,
+      "step": 3947
+    },
+    {
+      "epoch": 0.6483690185371461,
+      "grad_norm": 0.3566722105246468,
+      "learning_rate": 8.678690130757579e-06,
+      "loss": 0.5222,
+      "step": 3948
+    },
+    {
+      "epoch": 0.6485332457454889,
+      "grad_norm": 0.36829644006372725,
+      "learning_rate": 8.67851607823058e-06,
+      "loss": 0.5476,
+      "step": 3949
+    },
+    {
+      "epoch": 0.6486974729538316,
+      "grad_norm": 0.4250247596289241,
+      "learning_rate": 8.678341980320785e-06,
+      "loss": 0.5248,
+      "step": 3950
+    },
+    {
+      "epoch": 0.6488617001621744,
+      "grad_norm": 0.35853628855086594,
+      "learning_rate": 8.678167837030085e-06,
+      "loss": 0.5204,
+      "step": 3951
+    },
+    {
+      "epoch": 0.6490259273705171,
+      "grad_norm": 0.3992651417520165,
+      "learning_rate": 8.677993648360371e-06,
+      "loss": 0.5544,
+      "step": 3952
+    },
+    {
+      "epoch": 0.6491901545788599,
+      "grad_norm": 0.3157472097393151,
+      "learning_rate": 8.677819414313537e-06,
+      "loss": 0.52,
+      "step": 3953
+    },
+    {
+      "epoch": 0.6493543817872026,
+      "grad_norm": 0.35866185315676313,
+      "learning_rate": 8.677645134891472e-06,
+      "loss": 0.5252,
+      "step": 3954
+    },
+    {
+      "epoch": 0.6495186089955454,
+      "grad_norm": 0.3720242063110086,
+      "learning_rate": 8.677470810096072e-06,
+      "loss": 0.5543,
+      "step": 3955
+    },
+    {
+      "epoch": 0.6496828362038881,
+      "grad_norm": 0.33452108930498453,
+      "learning_rate": 8.677296439929228e-06,
+      "loss": 0.5281,
+      "step": 3956
+    },
+    {
+      "epoch": 0.6498470634122309,
+      "grad_norm": 0.36791969880247277,
+      "learning_rate": 8.677122024392837e-06,
+      "loss": 0.543,
+      "step": 3957
+    },
+    {
+      "epoch": 0.6500112906205736,
+      "grad_norm": 0.4421337612938926,
+      "learning_rate": 8.676947563488789e-06,
+      "loss": 0.5399,
+      "step": 3958
+    },
+    {
+      "epoch": 0.6501755178289164,
+      "grad_norm": 0.317527166513708,
+      "learning_rate": 8.67677305721898e-06,
+      "loss": 0.5314,
+      "step": 3959
+    },
+    {
+      "epoch": 0.650339745037259,
+      "grad_norm": 1.3753223596720887,
+      "learning_rate": 8.676598505585308e-06,
+      "loss": 0.5365,
+      "step": 3960
+    },
+    {
+      "epoch": 0.6505039722456017,
+      "grad_norm": 0.35311319928316154,
+      "learning_rate": 8.676423908589667e-06,
+      "loss": 0.5168,
+      "step": 3961
+    },
+    {
+      "epoch": 0.6506681994539445,
+      "grad_norm": 0.36668233413529516,
+      "learning_rate": 8.676249266233952e-06,
+      "loss": 0.5094,
+      "step": 3962
+    },
+    {
+      "epoch": 0.6508324266622872,
+      "grad_norm": 0.34926006262255677,
+      "learning_rate": 8.676074578520061e-06,
+      "loss": 0.5266,
+      "step": 3963
+    },
+    {
+      "epoch": 0.65099665387063,
+      "grad_norm": 0.3230850253272737,
+      "learning_rate": 8.675899845449892e-06,
+      "loss": 0.5222,
+      "step": 3964
+    },
+    {
+      "epoch": 0.6511608810789727,
+      "grad_norm": 0.3178205063481531,
+      "learning_rate": 8.675725067025343e-06,
+      "loss": 0.5334,
+      "step": 3965
+    },
+    {
+      "epoch": 0.6513251082873155,
+      "grad_norm": 0.3604238765156544,
+      "learning_rate": 8.67555024324831e-06,
+      "loss": 0.5571,
+      "step": 3966
+    },
+    {
+      "epoch": 0.6514893354956582,
+      "grad_norm": 0.4148764805152981,
+      "learning_rate": 8.675375374120695e-06,
+      "loss": 0.5082,
+      "step": 3967
+    },
+    {
+      "epoch": 0.651653562704001,
+      "grad_norm": 0.45696787695365004,
+      "learning_rate": 8.675200459644393e-06,
+      "loss": 0.5309,
+      "step": 3968
+    },
+    {
+      "epoch": 0.6518177899123437,
+      "grad_norm": 0.32008682200636956,
+      "learning_rate": 8.675025499821309e-06,
+      "loss": 0.5424,
+      "step": 3969
+    },
+    {
+      "epoch": 0.6519820171206865,
+      "grad_norm": 0.36765540872038655,
+      "learning_rate": 8.674850494653338e-06,
+      "loss": 0.5279,
+      "step": 3970
+    },
+    {
+      "epoch": 0.6521462443290292,
+      "grad_norm": 0.4053643104635402,
+      "learning_rate": 8.674675444142385e-06,
+      "loss": 0.5384,
+      "step": 3971
+    },
+    {
+      "epoch": 0.652310471537372,
+      "grad_norm": 0.39808760881236877,
+      "learning_rate": 8.674500348290349e-06,
+      "loss": 0.5518,
+      "step": 3972
+    },
+    {
+      "epoch": 0.6524746987457147,
+      "grad_norm": 0.44611158326975175,
+      "learning_rate": 8.674325207099131e-06,
+      "loss": 0.5523,
+      "step": 3973
+    },
+    {
+      "epoch": 0.6526389259540575,
+      "grad_norm": 0.3183684377023937,
+      "learning_rate": 8.674150020570635e-06,
+      "loss": 0.5302,
+      "step": 3974
+    },
+    {
+      "epoch": 0.6528031531624002,
+      "grad_norm": 0.34196266717367213,
+      "learning_rate": 8.673974788706762e-06,
+      "loss": 0.5159,
+      "step": 3975
+    },
+    {
+      "epoch": 0.652967380370743,
+      "grad_norm": 0.2900554832151464,
+      "learning_rate": 8.673799511509418e-06,
+      "loss": 0.5334,
+      "step": 3976
+    },
+    {
+      "epoch": 0.6531316075790856,
+      "grad_norm": 0.3386359848323148,
+      "learning_rate": 8.673624188980503e-06,
+      "loss": 0.5537,
+      "step": 3977
+    },
+    {
+      "epoch": 0.6532958347874284,
+      "grad_norm": 0.39546656599547636,
+      "learning_rate": 8.673448821121923e-06,
+      "loss": 0.5237,
+      "step": 3978
+    },
+    {
+      "epoch": 0.6534600619957711,
+      "grad_norm": 0.3317443593450704,
+      "learning_rate": 8.673273407935584e-06,
+      "loss": 0.5053,
+      "step": 3979
+    },
+    {
+      "epoch": 0.6536242892041139,
+      "grad_norm": 0.418188518330638,
+      "learning_rate": 8.67309794942339e-06,
+      "loss": 0.5492,
+      "step": 3980
+    },
+    {
+      "epoch": 0.6537885164124566,
+      "grad_norm": 0.369270618448298,
+      "learning_rate": 8.672922445587245e-06,
+      "loss": 0.5104,
+      "step": 3981
+    },
+    {
+      "epoch": 0.6539527436207994,
+      "grad_norm": 0.38664898700693945,
+      "learning_rate": 8.672746896429058e-06,
+      "loss": 0.5316,
+      "step": 3982
+    },
+    {
+      "epoch": 0.6541169708291421,
+      "grad_norm": 0.33549576162982075,
+      "learning_rate": 8.672571301950733e-06,
+      "loss": 0.5425,
+      "step": 3983
+    },
+    {
+      "epoch": 0.6542811980374849,
+      "grad_norm": 0.40001762129435364,
+      "learning_rate": 8.67239566215418e-06,
+      "loss": 0.5307,
+      "step": 3984
+    },
+    {
+      "epoch": 0.6544454252458276,
+      "grad_norm": 0.3913057314551714,
+      "learning_rate": 8.672219977041304e-06,
+      "loss": 0.5247,
+      "step": 3985
+    },
+    {
+      "epoch": 0.6546096524541704,
+      "grad_norm": 0.31883678483466016,
+      "learning_rate": 8.672044246614013e-06,
+      "loss": 0.5077,
+      "step": 3986
+    },
+    {
+      "epoch": 0.6547738796625131,
+      "grad_norm": 0.29753011492717113,
+      "learning_rate": 8.67186847087422e-06,
+      "loss": 0.5299,
+      "step": 3987
+    },
+    {
+      "epoch": 0.6549381068708559,
+      "grad_norm": 0.30891222511550015,
+      "learning_rate": 8.671692649823828e-06,
+      "loss": 0.5183,
+      "step": 3988
+    },
+    {
+      "epoch": 0.6551023340791986,
+      "grad_norm": 0.36057053254078036,
+      "learning_rate": 8.671516783464751e-06,
+      "loss": 0.5522,
+      "step": 3989
+    },
+    {
+      "epoch": 0.6552665612875413,
+      "grad_norm": 0.30740628581649376,
+      "learning_rate": 8.671340871798895e-06,
+      "loss": 0.5421,
+      "step": 3990
+    },
+    {
+      "epoch": 0.6554307884958841,
+      "grad_norm": 0.3390539225771164,
+      "learning_rate": 8.671164914828174e-06,
+      "loss": 0.5367,
+      "step": 3991
+    },
+    {
+      "epoch": 0.6555950157042268,
+      "grad_norm": 0.3027722746444447,
+      "learning_rate": 8.670988912554501e-06,
+      "loss": 0.5447,
+      "step": 3992
+    },
+    {
+      "epoch": 0.6557592429125696,
+      "grad_norm": 0.8625813728280358,
+      "learning_rate": 8.670812864979783e-06,
+      "loss": 0.5196,
+      "step": 3993
+    },
+    {
+      "epoch": 0.6559234701209122,
+      "grad_norm": 0.4270267910110388,
+      "learning_rate": 8.670636772105932e-06,
+      "loss": 0.5183,
+      "step": 3994
+    },
+    {
+      "epoch": 0.656087697329255,
+      "grad_norm": 0.30519084206610886,
+      "learning_rate": 8.670460633934864e-06,
+      "loss": 0.5346,
+      "step": 3995
+    },
+    {
+      "epoch": 0.6562519245375977,
+      "grad_norm": 0.3523099685252621,
+      "learning_rate": 8.67028445046849e-06,
+      "loss": 0.5239,
+      "step": 3996
+    },
+    {
+      "epoch": 0.6564161517459405,
+      "grad_norm": 0.31211800772194953,
+      "learning_rate": 8.670108221708725e-06,
+      "loss": 0.5328,
+      "step": 3997
+    },
+    {
+      "epoch": 0.6565803789542832,
+      "grad_norm": 0.4759892153046304,
+      "learning_rate": 8.669931947657481e-06,
+      "loss": 0.5338,
+      "step": 3998
+    },
+    {
+      "epoch": 0.656744606162626,
+      "grad_norm": 0.4637487119659366,
+      "learning_rate": 8.669755628316673e-06,
+      "loss": 0.5278,
+      "step": 3999
+    },
+    {
+      "epoch": 0.6569088333709687,
+      "grad_norm": 0.32994096949419766,
+      "learning_rate": 8.669579263688216e-06,
+      "loss": 0.5261,
+      "step": 4000
+    },
+    {
+      "epoch": 0.6570730605793115,
+      "grad_norm": 0.3356383750274373,
+      "learning_rate": 8.669402853774026e-06,
+      "loss": 0.5187,
+      "step": 4001
+    },
+    {
+      "epoch": 0.6572372877876542,
+      "grad_norm": 0.6717913507630792,
+      "learning_rate": 8.66922639857602e-06,
+      "loss": 0.5238,
+      "step": 4002
+    },
+    {
+      "epoch": 0.657401514995997,
+      "grad_norm": 0.3221690604002636,
+      "learning_rate": 8.669049898096114e-06,
+      "loss": 0.5559,
+      "step": 4003
+    },
+    {
+      "epoch": 0.6575657422043397,
+      "grad_norm": 0.3081158006782644,
+      "learning_rate": 8.668873352336221e-06,
+      "loss": 0.5057,
+      "step": 4004
+    },
+    {
+      "epoch": 0.6577299694126825,
+      "grad_norm": 0.32070325867119254,
+      "learning_rate": 8.668696761298266e-06,
+      "loss": 0.5211,
+      "step": 4005
+    },
+    {
+      "epoch": 0.6578941966210252,
+      "grad_norm": 0.31266643294309454,
+      "learning_rate": 8.66852012498416e-06,
+      "loss": 0.5343,
+      "step": 4006
+    },
+    {
+      "epoch": 0.658058423829368,
+      "grad_norm": 0.2863747219818306,
+      "learning_rate": 8.668343443395824e-06,
+      "loss": 0.5046,
+      "step": 4007
+    },
+    {
+      "epoch": 0.6582226510377107,
+      "grad_norm": 0.659439189835058,
+      "learning_rate": 8.668166716535179e-06,
+      "loss": 0.5289,
+      "step": 4008
+    },
+    {
+      "epoch": 0.6583868782460535,
+      "grad_norm": 0.34529192696537236,
+      "learning_rate": 8.66798994440414e-06,
+      "loss": 0.5471,
+      "step": 4009
+    },
+    {
+      "epoch": 0.6585511054543961,
+      "grad_norm": 0.2888681021457298,
+      "learning_rate": 8.667813127004631e-06,
+      "loss": 0.5263,
+      "step": 4010
+    },
+    {
+      "epoch": 0.6587153326627389,
+      "grad_norm": 0.32980455433855144,
+      "learning_rate": 8.667636264338571e-06,
+      "loss": 0.5192,
+      "step": 4011
+    },
+    {
+      "epoch": 0.6588795598710816,
+      "grad_norm": 0.3485846589730208,
+      "learning_rate": 8.66745935640788e-06,
+      "loss": 0.5181,
+      "step": 4012
+    },
+    {
+      "epoch": 0.6590437870794243,
+      "grad_norm": 0.31578509817687833,
+      "learning_rate": 8.667282403214481e-06,
+      "loss": 0.5237,
+      "step": 4013
+    },
+    {
+      "epoch": 0.6592080142877671,
+      "grad_norm": 0.3229530274985868,
+      "learning_rate": 8.667105404760295e-06,
+      "loss": 0.5371,
+      "step": 4014
+    },
+    {
+      "epoch": 0.6593722414961098,
+      "grad_norm": 0.30354742074792096,
+      "learning_rate": 8.666928361047245e-06,
+      "loss": 0.5191,
+      "step": 4015
+    },
+    {
+      "epoch": 0.6595364687044526,
+      "grad_norm": 0.3509968909891949,
+      "learning_rate": 8.666751272077251e-06,
+      "loss": 0.5442,
+      "step": 4016
+    },
+    {
+      "epoch": 0.6597006959127953,
+      "grad_norm": 0.3153934088828136,
+      "learning_rate": 8.66657413785224e-06,
+      "loss": 0.5269,
+      "step": 4017
+    },
+    {
+      "epoch": 0.6598649231211381,
+      "grad_norm": 0.3170905924730358,
+      "learning_rate": 8.666396958374135e-06,
+      "loss": 0.5015,
+      "step": 4018
+    },
+    {
+      "epoch": 0.6600291503294808,
+      "grad_norm": 0.3199674654985588,
+      "learning_rate": 8.66621973364486e-06,
+      "loss": 0.5137,
+      "step": 4019
+    },
+    {
+      "epoch": 0.6601933775378236,
+      "grad_norm": 0.3528630549071056,
+      "learning_rate": 8.666042463666338e-06,
+      "loss": 0.5343,
+      "step": 4020
+    },
+    {
+      "epoch": 0.6603576047461663,
+      "grad_norm": 0.4338982051253656,
+      "learning_rate": 8.665865148440497e-06,
+      "loss": 0.5273,
+      "step": 4021
+    },
+    {
+      "epoch": 0.6605218319545091,
+      "grad_norm": 0.4618783895698222,
+      "learning_rate": 8.665687787969262e-06,
+      "loss": 0.5149,
+      "step": 4022
+    },
+    {
+      "epoch": 0.6606860591628518,
+      "grad_norm": 0.3659376755669406,
+      "learning_rate": 8.66551038225456e-06,
+      "loss": 0.5292,
+      "step": 4023
+    },
+    {
+      "epoch": 0.6608502863711946,
+      "grad_norm": 0.4078323654587953,
+      "learning_rate": 8.665332931298317e-06,
+      "loss": 0.5056,
+      "step": 4024
+    },
+    {
+      "epoch": 0.6610145135795373,
+      "grad_norm": 0.3036588935862725,
+      "learning_rate": 8.66515543510246e-06,
+      "loss": 0.5361,
+      "step": 4025
+    },
+    {
+      "epoch": 0.6611787407878801,
+      "grad_norm": 0.33300491685531275,
+      "learning_rate": 8.664977893668914e-06,
+      "loss": 0.5145,
+      "step": 4026
+    },
+    {
+      "epoch": 0.6613429679962227,
+      "grad_norm": 0.3164346698371754,
+      "learning_rate": 8.664800306999613e-06,
+      "loss": 0.5422,
+      "step": 4027
+    },
+    {
+      "epoch": 0.6615071952045655,
+      "grad_norm": 0.5384011020648434,
+      "learning_rate": 8.664622675096482e-06,
+      "loss": 0.5228,
+      "step": 4028
+    },
+    {
+      "epoch": 0.6616714224129082,
+      "grad_norm": 0.36002664128745665,
+      "learning_rate": 8.664444997961454e-06,
+      "loss": 0.5283,
+      "step": 4029
+    },
+    {
+      "epoch": 0.661835649621251,
+      "grad_norm": 0.34992110750106104,
+      "learning_rate": 8.664267275596453e-06,
+      "loss": 0.5273,
+      "step": 4030
+    },
+    {
+      "epoch": 0.6619998768295937,
+      "grad_norm": 0.33045355618722133,
+      "learning_rate": 8.664089508003413e-06,
+      "loss": 0.5309,
+      "step": 4031
+    },
+    {
+      "epoch": 0.6621641040379365,
+      "grad_norm": 0.3008997087087797,
+      "learning_rate": 8.663911695184265e-06,
+      "loss": 0.5214,
+      "step": 4032
+    },
+    {
+      "epoch": 0.6623283312462792,
+      "grad_norm": 0.2974714182775004,
+      "learning_rate": 8.663733837140939e-06,
+      "loss": 0.5514,
+      "step": 4033
+    },
+    {
+      "epoch": 0.662492558454622,
+      "grad_norm": 0.32714662729303373,
+      "learning_rate": 8.663555933875366e-06,
+      "loss": 0.5474,
+      "step": 4034
+    },
+    {
+      "epoch": 0.6626567856629647,
+      "grad_norm": 0.3179914587930655,
+      "learning_rate": 8.663377985389478e-06,
+      "loss": 0.5253,
+      "step": 4035
+    },
+    {
+      "epoch": 0.6628210128713075,
+      "grad_norm": 0.34607268067750174,
+      "learning_rate": 8.663199991685212e-06,
+      "loss": 0.5385,
+      "step": 4036
+    },
+    {
+      "epoch": 0.6629852400796502,
+      "grad_norm": 0.38888766148880033,
+      "learning_rate": 8.663021952764496e-06,
+      "loss": 0.5179,
+      "step": 4037
+    },
+    {
+      "epoch": 0.663149467287993,
+      "grad_norm": 0.34136523068815844,
+      "learning_rate": 8.662843868629267e-06,
+      "loss": 0.5359,
+      "step": 4038
+    },
+    {
+      "epoch": 0.6633136944963357,
+      "grad_norm": 0.36005173204058966,
+      "learning_rate": 8.662665739281458e-06,
+      "loss": 0.5454,
+      "step": 4039
+    },
+    {
+      "epoch": 0.6634779217046785,
+      "grad_norm": 0.3518704195072764,
+      "learning_rate": 8.662487564723002e-06,
+      "loss": 0.5415,
+      "step": 4040
+    },
+    {
+      "epoch": 0.6636421489130212,
+      "grad_norm": 0.3015583584370122,
+      "learning_rate": 8.662309344955838e-06,
+      "loss": 0.5281,
+      "step": 4041
+    },
+    {
+      "epoch": 0.663806376121364,
+      "grad_norm": 0.2901950538717286,
+      "learning_rate": 8.662131079981897e-06,
+      "loss": 0.5156,
+      "step": 4042
+    },
+    {
+      "epoch": 0.6639706033297067,
+      "grad_norm": 0.3280991639172377,
+      "learning_rate": 8.661952769803119e-06,
+      "loss": 0.5402,
+      "step": 4043
+    },
+    {
+      "epoch": 0.6641348305380493,
+      "grad_norm": 0.310899012651114,
+      "learning_rate": 8.661774414421438e-06,
+      "loss": 0.5298,
+      "step": 4044
+    },
+    {
+      "epoch": 0.6642990577463921,
+      "grad_norm": 0.34032542764121404,
+      "learning_rate": 8.661596013838793e-06,
+      "loss": 0.5329,
+      "step": 4045
+    },
+    {
+      "epoch": 0.6644632849547348,
+      "grad_norm": 0.3048173603991897,
+      "learning_rate": 8.66141756805712e-06,
+      "loss": 0.5272,
+      "step": 4046
+    },
+    {
+      "epoch": 0.6646275121630776,
+      "grad_norm": 0.3861615964125312,
+      "learning_rate": 8.661239077078358e-06,
+      "loss": 0.5392,
+      "step": 4047
+    },
+    {
+      "epoch": 0.6647917393714203,
+      "grad_norm": 0.37328731816957406,
+      "learning_rate": 8.661060540904447e-06,
+      "loss": 0.5333,
+      "step": 4048
+    },
+    {
+      "epoch": 0.6649559665797631,
+      "grad_norm": 0.33058996782192235,
+      "learning_rate": 8.660881959537324e-06,
+      "loss": 0.533,
+      "step": 4049
+    },
+    {
+      "epoch": 0.6651201937881058,
+      "grad_norm": 0.3452711619202778,
+      "learning_rate": 8.66070333297893e-06,
+      "loss": 0.5183,
+      "step": 4050
+    },
+    {
+      "epoch": 0.6652844209964486,
+      "grad_norm": 0.3659269032483274,
+      "learning_rate": 8.660524661231202e-06,
+      "loss": 0.5348,
+      "step": 4051
+    },
+    {
+      "epoch": 0.6654486482047913,
+      "grad_norm": 0.31385515871837055,
+      "learning_rate": 8.660345944296083e-06,
+      "loss": 0.5471,
+      "step": 4052
+    },
+    {
+      "epoch": 0.6656128754131341,
+      "grad_norm": 0.3140491300896386,
+      "learning_rate": 8.660167182175515e-06,
+      "loss": 0.5315,
+      "step": 4053
+    },
+    {
+      "epoch": 0.6657771026214768,
+      "grad_norm": 0.30202244601958633,
+      "learning_rate": 8.659988374871436e-06,
+      "loss": 0.5386,
+      "step": 4054
+    },
+    {
+      "epoch": 0.6659413298298196,
+      "grad_norm": 0.3905089298298893,
+      "learning_rate": 8.659809522385794e-06,
+      "loss": 0.5208,
+      "step": 4055
+    },
+    {
+      "epoch": 0.6661055570381623,
+      "grad_norm": 0.30906324457562473,
+      "learning_rate": 8.659630624720525e-06,
+      "loss": 0.5518,
+      "step": 4056
+    },
+    {
+      "epoch": 0.6662697842465051,
+      "grad_norm": 0.4354062962749167,
+      "learning_rate": 8.659451681877577e-06,
+      "loss": 0.5515,
+      "step": 4057
+    },
+    {
+      "epoch": 0.6664340114548478,
+      "grad_norm": 0.29273028085860114,
+      "learning_rate": 8.65927269385889e-06,
+      "loss": 0.514,
+      "step": 4058
+    },
+    {
+      "epoch": 0.6665982386631906,
+      "grad_norm": 0.42766559142493105,
+      "learning_rate": 8.659093660666411e-06,
+      "loss": 0.5504,
+      "step": 4059
+    },
+    {
+      "epoch": 0.6667624658715333,
+      "grad_norm": 0.4006447442117922,
+      "learning_rate": 8.658914582302082e-06,
+      "loss": 0.5121,
+      "step": 4060
+    },
+    {
+      "epoch": 0.666926693079876,
+      "grad_norm": 0.3034528462862328,
+      "learning_rate": 8.658735458767848e-06,
+      "loss": 0.5228,
+      "step": 4061
+    },
+    {
+      "epoch": 0.6670909202882187,
+      "grad_norm": 0.4233032717301708,
+      "learning_rate": 8.658556290065655e-06,
+      "loss": 0.5336,
+      "step": 4062
+    },
+    {
+      "epoch": 0.6672551474965615,
+      "grad_norm": 0.45100471286674626,
+      "learning_rate": 8.65837707619745e-06,
+      "loss": 0.5082,
+      "step": 4063
+    },
+    {
+      "epoch": 0.6674193747049042,
+      "grad_norm": 1.0766468981490969,
+      "learning_rate": 8.658197817165181e-06,
+      "loss": 0.5224,
+      "step": 4064
+    },
+    {
+      "epoch": 0.667583601913247,
+      "grad_norm": 0.33427942116827647,
+      "learning_rate": 8.658018512970788e-06,
+      "loss": 0.5121,
+      "step": 4065
+    },
+    {
+      "epoch": 0.6677478291215897,
+      "grad_norm": 0.3459647160115218,
+      "learning_rate": 8.657839163616227e-06,
+      "loss": 0.5368,
+      "step": 4066
+    },
+    {
+      "epoch": 0.6679120563299324,
+      "grad_norm": 0.36718583835824414,
+      "learning_rate": 8.657659769103439e-06,
+      "loss": 0.5273,
+      "step": 4067
+    },
+    {
+      "epoch": 0.6680762835382752,
+      "grad_norm": 0.3013228326013261,
+      "learning_rate": 8.657480329434378e-06,
+      "loss": 0.561,
+      "step": 4068
+    },
+    {
+      "epoch": 0.6682405107466179,
+      "grad_norm": 0.31179364960946604,
+      "learning_rate": 8.657300844610988e-06,
+      "loss": 0.5286,
+      "step": 4069
+    },
+    {
+      "epoch": 0.6684047379549607,
+      "grad_norm": 0.37872253313407145,
+      "learning_rate": 8.657121314635221e-06,
+      "loss": 0.5431,
+      "step": 4070
+    },
+    {
+      "epoch": 0.6685689651633034,
+      "grad_norm": 0.33270200758353,
+      "learning_rate": 8.656941739509027e-06,
+      "loss": 0.5331,
+      "step": 4071
+    },
+    {
+      "epoch": 0.6687331923716462,
+      "grad_norm": 0.4963399150092201,
+      "learning_rate": 8.656762119234356e-06,
+      "loss": 0.5292,
+      "step": 4072
+    },
+    {
+      "epoch": 0.6688974195799889,
+      "grad_norm": 0.29276600299262023,
+      "learning_rate": 8.656582453813157e-06,
+      "loss": 0.5015,
+      "step": 4073
+    },
+    {
+      "epoch": 0.6690616467883317,
+      "grad_norm": 0.3227354735508351,
+      "learning_rate": 8.656402743247385e-06,
+      "loss": 0.5265,
+      "step": 4074
+    },
+    {
+      "epoch": 0.6692258739966744,
+      "grad_norm": 0.3580768588600473,
+      "learning_rate": 8.65622298753899e-06,
+      "loss": 0.5258,
+      "step": 4075
+    },
+    {
+      "epoch": 0.6693901012050172,
+      "grad_norm": 0.3428832595500862,
+      "learning_rate": 8.656043186689923e-06,
+      "loss": 0.5283,
+      "step": 4076
+    },
+    {
+      "epoch": 0.6695543284133599,
+      "grad_norm": 0.43434915808781277,
+      "learning_rate": 8.655863340702139e-06,
+      "loss": 0.5317,
+      "step": 4077
+    },
+    {
+      "epoch": 0.6697185556217026,
+      "grad_norm": 0.3118494196364416,
+      "learning_rate": 8.65568344957759e-06,
+      "loss": 0.5465,
+      "step": 4078
+    },
+    {
+      "epoch": 0.6698827828300453,
+      "grad_norm": 0.281140918903688,
+      "learning_rate": 8.65550351331823e-06,
+      "loss": 0.5173,
+      "step": 4079
+    },
+    {
+      "epoch": 0.6700470100383881,
+      "grad_norm": 0.31825315533295956,
+      "learning_rate": 8.655323531926013e-06,
+      "loss": 0.5469,
+      "step": 4080
+    },
+    {
+      "epoch": 0.6702112372467308,
+      "grad_norm": 0.45548632455563126,
+      "learning_rate": 8.655143505402893e-06,
+      "loss": 0.5542,
+      "step": 4081
+    },
+    {
+      "epoch": 0.6703754644550736,
+      "grad_norm": 0.3272018942287102,
+      "learning_rate": 8.654963433750829e-06,
+      "loss": 0.5121,
+      "step": 4082
+    },
+    {
+      "epoch": 0.6705396916634163,
+      "grad_norm": 0.29951992766048624,
+      "learning_rate": 8.654783316971773e-06,
+      "loss": 0.5263,
+      "step": 4083
+    },
+    {
+      "epoch": 0.6707039188717591,
+      "grad_norm": 0.3251815306475271,
+      "learning_rate": 8.654603155067682e-06,
+      "loss": 0.5348,
+      "step": 4084
+    },
+    {
+      "epoch": 0.6708681460801018,
+      "grad_norm": 0.3957221610064905,
+      "learning_rate": 8.654422948040515e-06,
+      "loss": 0.5044,
+      "step": 4085
+    },
+    {
+      "epoch": 0.6710323732884446,
+      "grad_norm": 0.30994219622654656,
+      "learning_rate": 8.654242695892224e-06,
+      "loss": 0.5088,
+      "step": 4086
+    },
+    {
+      "epoch": 0.6711966004967873,
+      "grad_norm": 0.3344417178331218,
+      "learning_rate": 8.654062398624772e-06,
+      "loss": 0.517,
+      "step": 4087
+    },
+    {
+      "epoch": 0.6713608277051301,
+      "grad_norm": 0.3001680261537345,
+      "learning_rate": 8.653882056240116e-06,
+      "loss": 0.5182,
+      "step": 4088
+    },
+    {
+      "epoch": 0.6715250549134728,
+      "grad_norm": 0.3372489161830878,
+      "learning_rate": 8.653701668740214e-06,
+      "loss": 0.524,
+      "step": 4089
+    },
+    {
+      "epoch": 0.6716892821218156,
+      "grad_norm": 0.32761295082448727,
+      "learning_rate": 8.653521236127023e-06,
+      "loss": 0.5537,
+      "step": 4090
+    },
+    {
+      "epoch": 0.6718535093301583,
+      "grad_norm": 0.45024104103317986,
+      "learning_rate": 8.653340758402508e-06,
+      "loss": 0.5146,
+      "step": 4091
+    },
+    {
+      "epoch": 0.672017736538501,
+      "grad_norm": 0.3375009611373258,
+      "learning_rate": 8.653160235568622e-06,
+      "loss": 0.515,
+      "step": 4092
+    },
+    {
+      "epoch": 0.6721819637468438,
+      "grad_norm": 0.30487466834246935,
+      "learning_rate": 8.652979667627333e-06,
+      "loss": 0.507,
+      "step": 4093
+    },
+    {
+      "epoch": 0.6723461909551866,
+      "grad_norm": 0.3390407856209974,
+      "learning_rate": 8.6527990545806e-06,
+      "loss": 0.5235,
+      "step": 4094
+    },
+    {
+      "epoch": 0.6725104181635292,
+      "grad_norm": 0.44647654142082505,
+      "learning_rate": 8.65261839643038e-06,
+      "loss": 0.5302,
+      "step": 4095
+    },
+    {
+      "epoch": 0.6726746453718719,
+      "grad_norm": 0.4409501511793635,
+      "learning_rate": 8.65243769317864e-06,
+      "loss": 0.5076,
+      "step": 4096
+    },
+    {
+      "epoch": 0.6728388725802147,
+      "grad_norm": 0.3545266642369537,
+      "learning_rate": 8.652256944827341e-06,
+      "loss": 0.5139,
+      "step": 4097
+    },
+    {
+      "epoch": 0.6730030997885574,
+      "grad_norm": 0.3919734598018783,
+      "learning_rate": 8.652076151378446e-06,
+      "loss": 0.5267,
+      "step": 4098
+    },
+    {
+      "epoch": 0.6731673269969002,
+      "grad_norm": 0.38540276913674354,
+      "learning_rate": 8.65189531283392e-06,
+      "loss": 0.5229,
+      "step": 4099
+    },
+    {
+      "epoch": 0.6733315542052429,
+      "grad_norm": 0.2910524183414479,
+      "learning_rate": 8.651714429195725e-06,
+      "loss": 0.542,
+      "step": 4100
+    },
+    {
+      "epoch": 0.6734957814135857,
+      "grad_norm": 0.43159118996051093,
+      "learning_rate": 8.651533500465828e-06,
+      "loss": 0.5326,
+      "step": 4101
+    },
+    {
+      "epoch": 0.6736600086219284,
+      "grad_norm": 0.33738359200446183,
+      "learning_rate": 8.651352526646191e-06,
+      "loss": 0.5403,
+      "step": 4102
+    },
+    {
+      "epoch": 0.6738242358302712,
+      "grad_norm": 0.3315680073734393,
+      "learning_rate": 8.651171507738783e-06,
+      "loss": 0.516,
+      "step": 4103
+    },
+    {
+      "epoch": 0.6739884630386139,
+      "grad_norm": 0.3589770475880881,
+      "learning_rate": 8.650990443745567e-06,
+      "loss": 0.5414,
+      "step": 4104
+    },
+    {
+      "epoch": 0.6741526902469567,
+      "grad_norm": 0.35588571798248414,
+      "learning_rate": 8.65080933466851e-06,
+      "loss": 0.5215,
+      "step": 4105
+    },
+    {
+      "epoch": 0.6743169174552994,
+      "grad_norm": 0.3033052472495628,
+      "learning_rate": 8.65062818050958e-06,
+      "loss": 0.5435,
+      "step": 4106
+    },
+    {
+      "epoch": 0.6744811446636422,
+      "grad_norm": 0.6316488676677522,
+      "learning_rate": 8.650446981270744e-06,
+      "loss": 0.5224,
+      "step": 4107
+    },
+    {
+      "epoch": 0.6746453718719849,
+      "grad_norm": 0.3855301938414688,
+      "learning_rate": 8.650265736953972e-06,
+      "loss": 0.54,
+      "step": 4108
+    },
+    {
+      "epoch": 0.6748095990803277,
+      "grad_norm": 0.32885699707547555,
+      "learning_rate": 8.65008444756123e-06,
+      "loss": 0.5318,
+      "step": 4109
+    },
+    {
+      "epoch": 0.6749738262886704,
+      "grad_norm": 0.35338886666359093,
+      "learning_rate": 8.649903113094487e-06,
+      "loss": 0.528,
+      "step": 4110
+    },
+    {
+      "epoch": 0.6751380534970132,
+      "grad_norm": 0.36097457735813854,
+      "learning_rate": 8.649721733555715e-06,
+      "loss": 0.5244,
+      "step": 4111
+    },
+    {
+      "epoch": 0.6753022807053558,
+      "grad_norm": 0.31060975134752294,
+      "learning_rate": 8.64954030894688e-06,
+      "loss": 0.5215,
+      "step": 4112
+    },
+    {
+      "epoch": 0.6754665079136986,
+      "grad_norm": 0.3011457205831512,
+      "learning_rate": 8.649358839269955e-06,
+      "loss": 0.508,
+      "step": 4113
+    },
+    {
+      "epoch": 0.6756307351220413,
+      "grad_norm": 0.3418984691313553,
+      "learning_rate": 8.649177324526913e-06,
+      "loss": 0.5235,
+      "step": 4114
+    },
+    {
+      "epoch": 0.675794962330384,
+      "grad_norm": 0.43582773294360444,
+      "learning_rate": 8.64899576471972e-06,
+      "loss": 0.5541,
+      "step": 4115
+    },
+    {
+      "epoch": 0.6759591895387268,
+      "grad_norm": 0.35149410264601055,
+      "learning_rate": 8.648814159850354e-06,
+      "loss": 0.5347,
+      "step": 4116
+    },
+    {
+      "epoch": 0.6761234167470695,
+      "grad_norm": 0.368562017515984,
+      "learning_rate": 8.648632509920781e-06,
+      "loss": 0.542,
+      "step": 4117
+    },
+    {
+      "epoch": 0.6762876439554123,
+      "grad_norm": 0.323205499569439,
+      "learning_rate": 8.64845081493298e-06,
+      "loss": 0.5271,
+      "step": 4118
+    },
+    {
+      "epoch": 0.676451871163755,
+      "grad_norm": 0.412816416007039,
+      "learning_rate": 8.64826907488892e-06,
+      "loss": 0.5231,
+      "step": 4119
+    },
+    {
+      "epoch": 0.6766160983720978,
+      "grad_norm": 0.2996965923663482,
+      "learning_rate": 8.648087289790578e-06,
+      "loss": 0.5389,
+      "step": 4120
+    },
+    {
+      "epoch": 0.6767803255804405,
+      "grad_norm": 0.3222559504790764,
+      "learning_rate": 8.647905459639926e-06,
+      "loss": 0.5224,
+      "step": 4121
+    },
+    {
+      "epoch": 0.6769445527887833,
+      "grad_norm": 0.30650977456731987,
+      "learning_rate": 8.647723584438939e-06,
+      "loss": 0.5262,
+      "step": 4122
+    },
+    {
+      "epoch": 0.677108779997126,
+      "grad_norm": 0.5112994720395476,
+      "learning_rate": 8.647541664189593e-06,
+      "loss": 0.5178,
+      "step": 4123
+    },
+    {
+      "epoch": 0.6772730072054688,
+      "grad_norm": 0.3522817205412864,
+      "learning_rate": 8.647359698893867e-06,
+      "loss": 0.5199,
+      "step": 4124
+    },
+    {
+      "epoch": 0.6774372344138115,
+      "grad_norm": 0.3024109450912323,
+      "learning_rate": 8.647177688553731e-06,
+      "loss": 0.5135,
+      "step": 4125
+    },
+    {
+      "epoch": 0.6776014616221543,
+      "grad_norm": 0.3174797317269863,
+      "learning_rate": 8.646995633171165e-06,
+      "loss": 0.5123,
+      "step": 4126
+    },
+    {
+      "epoch": 0.677765688830497,
+      "grad_norm": 0.3490783488508878,
+      "learning_rate": 8.646813532748147e-06,
+      "loss": 0.5354,
+      "step": 4127
+    },
+    {
+      "epoch": 0.6779299160388398,
+      "grad_norm": 0.44894277917191494,
+      "learning_rate": 8.646631387286655e-06,
+      "loss": 0.5048,
+      "step": 4128
+    },
+    {
+      "epoch": 0.6780941432471824,
+      "grad_norm": 1.0038385015849884,
+      "learning_rate": 8.646449196788664e-06,
+      "loss": 0.5346,
+      "step": 4129
+    },
+    {
+      "epoch": 0.6782583704555252,
+      "grad_norm": 0.28209868548502737,
+      "learning_rate": 8.646266961256158e-06,
+      "loss": 0.5092,
+      "step": 4130
+    },
+    {
+      "epoch": 0.6784225976638679,
+      "grad_norm": 0.32056520106324327,
+      "learning_rate": 8.646084680691112e-06,
+      "loss": 0.521,
+      "step": 4131
+    },
+    {
+      "epoch": 0.6785868248722107,
+      "grad_norm": 0.30892792961083615,
+      "learning_rate": 8.645902355095507e-06,
+      "loss": 0.5233,
+      "step": 4132
+    },
+    {
+      "epoch": 0.6787510520805534,
+      "grad_norm": 0.3004522844978485,
+      "learning_rate": 8.645719984471325e-06,
+      "loss": 0.5519,
+      "step": 4133
+    },
+    {
+      "epoch": 0.6789152792888962,
+      "grad_norm": 0.37531256935974994,
+      "learning_rate": 8.645537568820544e-06,
+      "loss": 0.522,
+      "step": 4134
+    },
+    {
+      "epoch": 0.6790795064972389,
+      "grad_norm": 0.4463922435591353,
+      "learning_rate": 8.645355108145146e-06,
+      "loss": 0.5076,
+      "step": 4135
+    },
+    {
+      "epoch": 0.6792437337055817,
+      "grad_norm": 0.2871378370372213,
+      "learning_rate": 8.645172602447113e-06,
+      "loss": 0.5084,
+      "step": 4136
+    },
+    {
+      "epoch": 0.6794079609139244,
+      "grad_norm": 0.39296031580485435,
+      "learning_rate": 8.644990051728428e-06,
+      "loss": 0.5329,
+      "step": 4137
+    },
+    {
+      "epoch": 0.6795721881222672,
+      "grad_norm": 0.3206991930440626,
+      "learning_rate": 8.644807455991071e-06,
+      "loss": 0.5254,
+      "step": 4138
+    },
+    {
+      "epoch": 0.6797364153306099,
+      "grad_norm": 0.312091021839642,
+      "learning_rate": 8.644624815237029e-06,
+      "loss": 0.5262,
+      "step": 4139
+    },
+    {
+      "epoch": 0.6799006425389527,
+      "grad_norm": 0.2924739456006572,
+      "learning_rate": 8.644442129468284e-06,
+      "loss": 0.5002,
+      "step": 4140
+    },
+    {
+      "epoch": 0.6800648697472954,
+      "grad_norm": 0.3458651635031552,
+      "learning_rate": 8.64425939868682e-06,
+      "loss": 0.5326,
+      "step": 4141
+    },
+    {
+      "epoch": 0.6802290969556382,
+      "grad_norm": 0.6053301737235616,
+      "learning_rate": 8.644076622894621e-06,
+      "loss": 0.5496,
+      "step": 4142
+    },
+    {
+      "epoch": 0.6803933241639809,
+      "grad_norm": 0.3242956814925467,
+      "learning_rate": 8.643893802093671e-06,
+      "loss": 0.5138,
+      "step": 4143
+    },
+    {
+      "epoch": 0.6805575513723237,
+      "grad_norm": 0.2991928853840647,
+      "learning_rate": 8.64371093628596e-06,
+      "loss": 0.5154,
+      "step": 4144
+    },
+    {
+      "epoch": 0.6807217785806664,
+      "grad_norm": 0.2870604350100677,
+      "learning_rate": 8.64352802547347e-06,
+      "loss": 0.5152,
+      "step": 4145
+    },
+    {
+      "epoch": 0.680886005789009,
+      "grad_norm": 0.2885442702491027,
+      "learning_rate": 8.64334506965819e-06,
+      "loss": 0.5375,
+      "step": 4146
+    },
+    {
+      "epoch": 0.6810502329973518,
+      "grad_norm": 0.36928363961336547,
+      "learning_rate": 8.643162068842105e-06,
+      "loss": 0.5292,
+      "step": 4147
+    },
+    {
+      "epoch": 0.6812144602056945,
+      "grad_norm": 0.3343679239764359,
+      "learning_rate": 8.642979023027203e-06,
+      "loss": 0.5281,
+      "step": 4148
+    },
+    {
+      "epoch": 0.6813786874140373,
+      "grad_norm": 0.3116320481253324,
+      "learning_rate": 8.642795932215472e-06,
+      "loss": 0.5394,
+      "step": 4149
+    },
+    {
+      "epoch": 0.68154291462238,
+      "grad_norm": 0.2847083185415691,
+      "learning_rate": 8.642612796408904e-06,
+      "loss": 0.5315,
+      "step": 4150
+    },
+    {
+      "epoch": 0.6817071418307228,
+      "grad_norm": 0.3789055568649402,
+      "learning_rate": 8.642429615609483e-06,
+      "loss": 0.5175,
+      "step": 4151
+    },
+    {
+      "epoch": 0.6818713690390655,
+      "grad_norm": 0.35801662699820525,
+      "learning_rate": 8.642246389819202e-06,
+      "loss": 0.5237,
+      "step": 4152
+    },
+    {
+      "epoch": 0.6820355962474083,
+      "grad_norm": 0.2999952779090304,
+      "learning_rate": 8.642063119040049e-06,
+      "loss": 0.5155,
+      "step": 4153
+    },
+    {
+      "epoch": 0.682199823455751,
+      "grad_norm": 0.4117495657601018,
+      "learning_rate": 8.641879803274016e-06,
+      "loss": 0.5248,
+      "step": 4154
+    },
+    {
+      "epoch": 0.6823640506640938,
+      "grad_norm": 0.38786034920878637,
+      "learning_rate": 8.641696442523093e-06,
+      "loss": 0.5205,
+      "step": 4155
+    },
+    {
+      "epoch": 0.6825282778724365,
+      "grad_norm": 0.3532378510900068,
+      "learning_rate": 8.641513036789273e-06,
+      "loss": 0.5291,
+      "step": 4156
+    },
+    {
+      "epoch": 0.6826925050807793,
+      "grad_norm": 0.2924513736687471,
+      "learning_rate": 8.641329586074545e-06,
+      "loss": 0.517,
+      "step": 4157
+    },
+    {
+      "epoch": 0.682856732289122,
+      "grad_norm": 0.3274374660126311,
+      "learning_rate": 8.641146090380903e-06,
+      "loss": 0.5282,
+      "step": 4158
+    },
+    {
+      "epoch": 0.6830209594974648,
+      "grad_norm": 0.35013113932458,
+      "learning_rate": 8.64096254971034e-06,
+      "loss": 0.5332,
+      "step": 4159
+    },
+    {
+      "epoch": 0.6831851867058075,
+      "grad_norm": 0.3457641571260403,
+      "learning_rate": 8.640778964064852e-06,
+      "loss": 0.5289,
+      "step": 4160
+    },
+    {
+      "epoch": 0.6833494139141503,
+      "grad_norm": 0.27620805752207256,
+      "learning_rate": 8.640595333446427e-06,
+      "loss": 0.5458,
+      "step": 4161
+    },
+    {
+      "epoch": 0.683513641122493,
+      "grad_norm": 0.33925143437931815,
+      "learning_rate": 8.640411657857066e-06,
+      "loss": 0.5537,
+      "step": 4162
+    },
+    {
+      "epoch": 0.6836778683308357,
+      "grad_norm": 0.35668433810410605,
+      "learning_rate": 8.64022793729876e-06,
+      "loss": 0.5283,
+      "step": 4163
+    },
+    {
+      "epoch": 0.6838420955391784,
+      "grad_norm": 0.28380844419386303,
+      "learning_rate": 8.640044171773503e-06,
+      "loss": 0.5207,
+      "step": 4164
+    },
+    {
+      "epoch": 0.6840063227475212,
+      "grad_norm": 0.36428431243227716,
+      "learning_rate": 8.639860361283295e-06,
+      "loss": 0.5371,
+      "step": 4165
+    },
+    {
+      "epoch": 0.6841705499558639,
+      "grad_norm": 0.3638178524463825,
+      "learning_rate": 8.63967650583013e-06,
+      "loss": 0.5396,
+      "step": 4166
+    },
+    {
+      "epoch": 0.6843347771642067,
+      "grad_norm": 0.34929302756951347,
+      "learning_rate": 8.639492605416005e-06,
+      "loss": 0.5368,
+      "step": 4167
+    },
+    {
+      "epoch": 0.6844990043725494,
+      "grad_norm": 0.3769355480143036,
+      "learning_rate": 8.639308660042918e-06,
+      "loss": 0.5357,
+      "step": 4168
+    },
+    {
+      "epoch": 0.6846632315808922,
+      "grad_norm": 0.2722486789490862,
+      "learning_rate": 8.639124669712867e-06,
+      "loss": 0.5201,
+      "step": 4169
+    },
+    {
+      "epoch": 0.6848274587892349,
+      "grad_norm": 0.30186430431604144,
+      "learning_rate": 8.63894063442785e-06,
+      "loss": 0.5302,
+      "step": 4170
+    },
+    {
+      "epoch": 0.6849916859975776,
+      "grad_norm": 0.2866787935647283,
+      "learning_rate": 8.638756554189863e-06,
+      "loss": 0.501,
+      "step": 4171
+    },
+    {
+      "epoch": 0.6851559132059204,
+      "grad_norm": 0.38690681773281876,
+      "learning_rate": 8.63857242900091e-06,
+      "loss": 0.526,
+      "step": 4172
+    },
+    {
+      "epoch": 0.6853201404142631,
+      "grad_norm": 0.3231383434901005,
+      "learning_rate": 8.638388258862987e-06,
+      "loss": 0.5325,
+      "step": 4173
+    },
+    {
+      "epoch": 0.6854843676226059,
+      "grad_norm": 0.43681580478235355,
+      "learning_rate": 8.638204043778097e-06,
+      "loss": 0.5387,
+      "step": 4174
+    },
+    {
+      "epoch": 0.6856485948309486,
+      "grad_norm": 0.3668910158632165,
+      "learning_rate": 8.63801978374824e-06,
+      "loss": 0.5356,
+      "step": 4175
+    },
+    {
+      "epoch": 0.6858128220392914,
+      "grad_norm": 0.3455507180236111,
+      "learning_rate": 8.637835478775417e-06,
+      "loss": 0.5044,
+      "step": 4176
+    },
+    {
+      "epoch": 0.6859770492476341,
+      "grad_norm": 0.3556586108295262,
+      "learning_rate": 8.637651128861629e-06,
+      "loss": 0.5325,
+      "step": 4177
+    },
+    {
+      "epoch": 0.6861412764559769,
+      "grad_norm": 0.29839684395296556,
+      "learning_rate": 8.637466734008879e-06,
+      "loss": 0.5468,
+      "step": 4178
+    },
+    {
+      "epoch": 0.6863055036643196,
+      "grad_norm": 0.3371453476125466,
+      "learning_rate": 8.637282294219168e-06,
+      "loss": 0.508,
+      "step": 4179
+    },
+    {
+      "epoch": 0.6864697308726623,
+      "grad_norm": 0.48442277629443187,
+      "learning_rate": 8.637097809494504e-06,
+      "loss": 0.5469,
+      "step": 4180
+    },
+    {
+      "epoch": 0.686633958081005,
+      "grad_norm": 0.3501409566376713,
+      "learning_rate": 8.636913279836884e-06,
+      "loss": 0.5394,
+      "step": 4181
+    },
+    {
+      "epoch": 0.6867981852893478,
+      "grad_norm": 0.3153330952861787,
+      "learning_rate": 8.636728705248319e-06,
+      "loss": 0.5287,
+      "step": 4182
+    },
+    {
+      "epoch": 0.6869624124976905,
+      "grad_norm": 0.30647381190495343,
+      "learning_rate": 8.636544085730808e-06,
+      "loss": 0.509,
+      "step": 4183
+    },
+    {
+      "epoch": 0.6871266397060333,
+      "grad_norm": 0.32085086991778444,
+      "learning_rate": 8.636359421286358e-06,
+      "loss": 0.5338,
+      "step": 4184
+    },
+    {
+      "epoch": 0.687290866914376,
+      "grad_norm": 0.29831140561366876,
+      "learning_rate": 8.636174711916977e-06,
+      "loss": 0.5275,
+      "step": 4185
+    },
+    {
+      "epoch": 0.6874550941227188,
+      "grad_norm": 0.3066929012205568,
+      "learning_rate": 8.635989957624669e-06,
+      "loss": 0.5218,
+      "step": 4186
+    },
+    {
+      "epoch": 0.6876193213310615,
+      "grad_norm": 0.33259903521685663,
+      "learning_rate": 8.635805158411438e-06,
+      "loss": 0.5006,
+      "step": 4187
+    },
+    {
+      "epoch": 0.6877835485394043,
+      "grad_norm": 0.2893515986506582,
+      "learning_rate": 8.635620314279297e-06,
+      "loss": 0.5177,
+      "step": 4188
+    },
+    {
+      "epoch": 0.687947775747747,
+      "grad_norm": 0.33727758187688006,
+      "learning_rate": 8.63543542523025e-06,
+      "loss": 0.5136,
+      "step": 4189
+    },
+    {
+      "epoch": 0.6881120029560898,
+      "grad_norm": 0.31013060657131736,
+      "learning_rate": 8.635250491266304e-06,
+      "loss": 0.5238,
+      "step": 4190
+    },
+    {
+      "epoch": 0.6882762301644325,
+      "grad_norm": 0.38573616889290285,
+      "learning_rate": 8.63506551238947e-06,
+      "loss": 0.5505,
+      "step": 4191
+    },
+    {
+      "epoch": 0.6884404573727753,
+      "grad_norm": 0.313359998521622,
+      "learning_rate": 8.634880488601756e-06,
+      "loss": 0.5277,
+      "step": 4192
+    },
+    {
+      "epoch": 0.688604684581118,
+      "grad_norm": 0.349773895723136,
+      "learning_rate": 8.634695419905173e-06,
+      "loss": 0.5349,
+      "step": 4193
+    },
+    {
+      "epoch": 0.6887689117894608,
+      "grad_norm": 0.44196567700319817,
+      "learning_rate": 8.634510306301728e-06,
+      "loss": 0.5137,
+      "step": 4194
+    },
+    {
+      "epoch": 0.6889331389978035,
+      "grad_norm": 0.4226507437950207,
+      "learning_rate": 8.634325147793434e-06,
+      "loss": 0.5528,
+      "step": 4195
+    },
+    {
+      "epoch": 0.6890973662061463,
+      "grad_norm": 0.293215751832506,
+      "learning_rate": 8.6341399443823e-06,
+      "loss": 0.5337,
+      "step": 4196
+    },
+    {
+      "epoch": 0.6892615934144889,
+      "grad_norm": 0.28090267043332035,
+      "learning_rate": 8.63395469607034e-06,
+      "loss": 0.5268,
+      "step": 4197
+    },
+    {
+      "epoch": 0.6894258206228316,
+      "grad_norm": 0.35600281186827526,
+      "learning_rate": 8.633769402859566e-06,
+      "loss": 0.5323,
+      "step": 4198
+    },
+    {
+      "epoch": 0.6895900478311744,
+      "grad_norm": 0.34343124065854136,
+      "learning_rate": 8.633584064751989e-06,
+      "loss": 0.5365,
+      "step": 4199
+    },
+    {
+      "epoch": 0.6897542750395171,
+      "grad_norm": 0.32470237327045537,
+      "learning_rate": 8.63339868174962e-06,
+      "loss": 0.5258,
+      "step": 4200
+    },
+    {
+      "epoch": 0.6899185022478599,
+      "grad_norm": 0.3868201193984014,
+      "learning_rate": 8.633213253854476e-06,
+      "loss": 0.5275,
+      "step": 4201
+    },
+    {
+      "epoch": 0.6900827294562026,
+      "grad_norm": 0.4445413871679448,
+      "learning_rate": 8.63302778106857e-06,
+      "loss": 0.5337,
+      "step": 4202
+    },
+    {
+      "epoch": 0.6902469566645454,
+      "grad_norm": 0.48265095727302576,
+      "learning_rate": 8.632842263393915e-06,
+      "loss": 0.523,
+      "step": 4203
+    },
+    {
+      "epoch": 0.6904111838728881,
+      "grad_norm": 0.47827943743664814,
+      "learning_rate": 8.632656700832527e-06,
+      "loss": 0.5416,
+      "step": 4204
+    },
+    {
+      "epoch": 0.6905754110812309,
+      "grad_norm": 0.31881673938269417,
+      "learning_rate": 8.63247109338642e-06,
+      "loss": 0.5456,
+      "step": 4205
+    },
+    {
+      "epoch": 0.6907396382895736,
+      "grad_norm": 0.36208368158027626,
+      "learning_rate": 8.632285441057614e-06,
+      "loss": 0.5353,
+      "step": 4206
+    },
+    {
+      "epoch": 0.6909038654979164,
+      "grad_norm": 0.3044211762704708,
+      "learning_rate": 8.632099743848121e-06,
+      "loss": 0.5347,
+      "step": 4207
+    },
+    {
+      "epoch": 0.6910680927062591,
+      "grad_norm": 0.43847455160193893,
+      "learning_rate": 8.631914001759958e-06,
+      "loss": 0.5257,
+      "step": 4208
+    },
+    {
+      "epoch": 0.6912323199146019,
+      "grad_norm": 0.3791569545248908,
+      "learning_rate": 8.631728214795145e-06,
+      "loss": 0.5228,
+      "step": 4209
+    },
+    {
+      "epoch": 0.6913965471229446,
+      "grad_norm": 0.31693174954761305,
+      "learning_rate": 8.6315423829557e-06,
+      "loss": 0.5511,
+      "step": 4210
+    },
+    {
+      "epoch": 0.6915607743312874,
+      "grad_norm": 0.30018018314139316,
+      "learning_rate": 8.631356506243637e-06,
+      "loss": 0.5272,
+      "step": 4211
+    },
+    {
+      "epoch": 0.6917250015396301,
+      "grad_norm": 0.35174861259968854,
+      "learning_rate": 8.63117058466098e-06,
+      "loss": 0.5475,
+      "step": 4212
+    },
+    {
+      "epoch": 0.6918892287479729,
+      "grad_norm": 0.307674247694834,
+      "learning_rate": 8.630984618209743e-06,
+      "loss": 0.4938,
+      "step": 4213
+    },
+    {
+      "epoch": 0.6920534559563155,
+      "grad_norm": 0.36394072159220425,
+      "learning_rate": 8.630798606891951e-06,
+      "loss": 0.5274,
+      "step": 4214
+    },
+    {
+      "epoch": 0.6922176831646583,
+      "grad_norm": 0.33159368131380446,
+      "learning_rate": 8.630612550709622e-06,
+      "loss": 0.5469,
+      "step": 4215
+    },
+    {
+      "epoch": 0.692381910373001,
+      "grad_norm": 0.3251253917776796,
+      "learning_rate": 8.630426449664776e-06,
+      "loss": 0.547,
+      "step": 4216
+    },
+    {
+      "epoch": 0.6925461375813438,
+      "grad_norm": 0.3349215700548793,
+      "learning_rate": 8.630240303759436e-06,
+      "loss": 0.5146,
+      "step": 4217
+    },
+    {
+      "epoch": 0.6927103647896865,
+      "grad_norm": 0.320126842055859,
+      "learning_rate": 8.630054112995621e-06,
+      "loss": 0.5317,
+      "step": 4218
+    },
+    {
+      "epoch": 0.6928745919980293,
+      "grad_norm": 0.3873589155863503,
+      "learning_rate": 8.629867877375356e-06,
+      "loss": 0.5117,
+      "step": 4219
+    },
+    {
+      "epoch": 0.693038819206372,
+      "grad_norm": 0.3407313212235653,
+      "learning_rate": 8.629681596900663e-06,
+      "loss": 0.5329,
+      "step": 4220
+    },
+    {
+      "epoch": 0.6932030464147148,
+      "grad_norm": 0.295578297140357,
+      "learning_rate": 8.629495271573565e-06,
+      "loss": 0.5033,
+      "step": 4221
+    },
+    {
+      "epoch": 0.6933672736230575,
+      "grad_norm": 0.42228340319703683,
+      "learning_rate": 8.629308901396083e-06,
+      "loss": 0.4983,
+      "step": 4222
+    },
+    {
+      "epoch": 0.6935315008314002,
+      "grad_norm": 0.4463823743156875,
+      "learning_rate": 8.629122486370245e-06,
+      "loss": 0.513,
+      "step": 4223
+    },
+    {
+      "epoch": 0.693695728039743,
+      "grad_norm": 0.31355091465461943,
+      "learning_rate": 8.628936026498075e-06,
+      "loss": 0.5326,
+      "step": 4224
+    },
+    {
+      "epoch": 0.6938599552480857,
+      "grad_norm": 0.33679477663537927,
+      "learning_rate": 8.628749521781598e-06,
+      "loss": 0.5448,
+      "step": 4225
+    },
+    {
+      "epoch": 0.6940241824564285,
+      "grad_norm": 0.35080468703776435,
+      "learning_rate": 8.628562972222838e-06,
+      "loss": 0.4956,
+      "step": 4226
+    },
+    {
+      "epoch": 0.6941884096647712,
+      "grad_norm": 0.29616783212095005,
+      "learning_rate": 8.628376377823823e-06,
+      "loss": 0.5487,
+      "step": 4227
+    },
+    {
+      "epoch": 0.694352636873114,
+      "grad_norm": 0.3512870033644539,
+      "learning_rate": 8.628189738586577e-06,
+      "loss": 0.5192,
+      "step": 4228
+    },
+    {
+      "epoch": 0.6945168640814567,
+      "grad_norm": 0.7499849953119065,
+      "learning_rate": 8.628003054513129e-06,
+      "loss": 0.5412,
+      "step": 4229
+    },
+    {
+      "epoch": 0.6946810912897995,
+      "grad_norm": 0.34197986132326763,
+      "learning_rate": 8.627816325605509e-06,
+      "loss": 0.541,
+      "step": 4230
+    },
+    {
+      "epoch": 0.6948453184981421,
+      "grad_norm": 0.35850041319838344,
+      "learning_rate": 8.627629551865741e-06,
+      "loss": 0.5424,
+      "step": 4231
+    },
+    {
+      "epoch": 0.6950095457064849,
+      "grad_norm": 0.38240384165529273,
+      "learning_rate": 8.627442733295855e-06,
+      "loss": 0.5417,
+      "step": 4232
+    },
+    {
+      "epoch": 0.6951737729148276,
+      "grad_norm": 0.3236064766935071,
+      "learning_rate": 8.62725586989788e-06,
+      "loss": 0.5352,
+      "step": 4233
+    },
+    {
+      "epoch": 0.6953380001231704,
+      "grad_norm": 0.3556556217013173,
+      "learning_rate": 8.627068961673844e-06,
+      "loss": 0.5187,
+      "step": 4234
+    },
+    {
+      "epoch": 0.6955022273315131,
+      "grad_norm": 0.3423798038428814,
+      "learning_rate": 8.62688200862578e-06,
+      "loss": 0.5366,
+      "step": 4235
+    },
+    {
+      "epoch": 0.6956664545398559,
+      "grad_norm": 0.28350851553753786,
+      "learning_rate": 8.626695010755719e-06,
+      "loss": 0.522,
+      "step": 4236
+    },
+    {
+      "epoch": 0.6958306817481986,
+      "grad_norm": 0.3257472077657691,
+      "learning_rate": 8.626507968065687e-06,
+      "loss": 0.5215,
+      "step": 4237
+    },
+    {
+      "epoch": 0.6959949089565414,
+      "grad_norm": 0.31021167794491405,
+      "learning_rate": 8.62632088055772e-06,
+      "loss": 0.5088,
+      "step": 4238
+    },
+    {
+      "epoch": 0.6961591361648841,
+      "grad_norm": 0.3166485566164785,
+      "learning_rate": 8.626133748233847e-06,
+      "loss": 0.5213,
+      "step": 4239
+    },
+    {
+      "epoch": 0.6963233633732269,
+      "grad_norm": 0.29145841027820457,
+      "learning_rate": 8.625946571096106e-06,
+      "loss": 0.4946,
+      "step": 4240
+    },
+    {
+      "epoch": 0.6964875905815696,
+      "grad_norm": 0.319921850998031,
+      "learning_rate": 8.625759349146521e-06,
+      "loss": 0.5386,
+      "step": 4241
+    },
+    {
+      "epoch": 0.6966518177899124,
+      "grad_norm": 0.308972089997359,
+      "learning_rate": 8.625572082387132e-06,
+      "loss": 0.5413,
+      "step": 4242
+    },
+    {
+      "epoch": 0.6968160449982551,
+      "grad_norm": 0.2835317133915667,
+      "learning_rate": 8.625384770819972e-06,
+      "loss": 0.534,
+      "step": 4243
+    },
+    {
+      "epoch": 0.6969802722065979,
+      "grad_norm": 0.4233385444079988,
+      "learning_rate": 8.625197414447073e-06,
+      "loss": 0.5229,
+      "step": 4244
+    },
+    {
+      "epoch": 0.6971444994149406,
+      "grad_norm": 0.4137599078113359,
+      "learning_rate": 8.625010013270474e-06,
+      "loss": 0.4999,
+      "step": 4245
+    },
+    {
+      "epoch": 0.6973087266232834,
+      "grad_norm": 0.333057995130589,
+      "learning_rate": 8.624822567292205e-06,
+      "loss": 0.5021,
+      "step": 4246
+    },
+    {
+      "epoch": 0.6974729538316261,
+      "grad_norm": 0.3373896893862306,
+      "learning_rate": 8.624635076514307e-06,
+      "loss": 0.544,
+      "step": 4247
+    },
+    {
+      "epoch": 0.6976371810399687,
+      "grad_norm": 0.3598055058817781,
+      "learning_rate": 8.624447540938813e-06,
+      "loss": 0.5192,
+      "step": 4248
+    },
+    {
+      "epoch": 0.6978014082483115,
+      "grad_norm": 0.33727018756814525,
+      "learning_rate": 8.62425996056776e-06,
+      "loss": 0.5226,
+      "step": 4249
+    },
+    {
+      "epoch": 0.6979656354566542,
+      "grad_norm": 0.30647169364146576,
+      "learning_rate": 8.624072335403188e-06,
+      "loss": 0.5102,
+      "step": 4250
+    },
+    {
+      "epoch": 0.698129862664997,
+      "grad_norm": 0.2931637111302099,
+      "learning_rate": 8.62388466544713e-06,
+      "loss": 0.5347,
+      "step": 4251
+    },
+    {
+      "epoch": 0.6982940898733397,
+      "grad_norm": 0.41161972319441537,
+      "learning_rate": 8.623696950701629e-06,
+      "loss": 0.498,
+      "step": 4252
+    },
+    {
+      "epoch": 0.6984583170816825,
+      "grad_norm": 0.34023286853888174,
+      "learning_rate": 8.623509191168722e-06,
+      "loss": 0.5267,
+      "step": 4253
+    },
+    {
+      "epoch": 0.6986225442900252,
+      "grad_norm": 0.4900155873432433,
+      "learning_rate": 8.623321386850449e-06,
+      "loss": 0.5373,
+      "step": 4254
+    },
+    {
+      "epoch": 0.698786771498368,
+      "grad_norm": 0.39734723306453845,
+      "learning_rate": 8.623133537748847e-06,
+      "loss": 0.5201,
+      "step": 4255
+    },
+    {
+      "epoch": 0.6989509987067107,
+      "grad_norm": 0.3420535819762841,
+      "learning_rate": 8.622945643865959e-06,
+      "loss": 0.535,
+      "step": 4256
+    },
+    {
+      "epoch": 0.6991152259150535,
+      "grad_norm": 0.34538252628053706,
+      "learning_rate": 8.622757705203825e-06,
+      "loss": 0.5332,
+      "step": 4257
+    },
+    {
+      "epoch": 0.6992794531233962,
+      "grad_norm": 0.30859870827439856,
+      "learning_rate": 8.622569721764487e-06,
+      "loss": 0.5099,
+      "step": 4258
+    },
+    {
+      "epoch": 0.699443680331739,
+      "grad_norm": 0.3291857838753656,
+      "learning_rate": 8.622381693549985e-06,
+      "loss": 0.5341,
+      "step": 4259
+    },
+    {
+      "epoch": 0.6996079075400817,
+      "grad_norm": 0.36292305907656036,
+      "learning_rate": 8.62219362056236e-06,
+      "loss": 0.5301,
+      "step": 4260
+    },
+    {
+      "epoch": 0.6997721347484245,
+      "grad_norm": 0.3350119610981238,
+      "learning_rate": 8.622005502803659e-06,
+      "loss": 0.5331,
+      "step": 4261
+    },
+    {
+      "epoch": 0.6999363619567672,
+      "grad_norm": 0.35928706753215134,
+      "learning_rate": 8.621817340275921e-06,
+      "loss": 0.5262,
+      "step": 4262
+    },
+    {
+      "epoch": 0.70010058916511,
+      "grad_norm": 0.29995114712883486,
+      "learning_rate": 8.621629132981194e-06,
+      "loss": 0.5107,
+      "step": 4263
+    },
+    {
+      "epoch": 0.7002648163734527,
+      "grad_norm": 0.32459709306920986,
+      "learning_rate": 8.621440880921519e-06,
+      "loss": 0.5436,
+      "step": 4264
+    },
+    {
+      "epoch": 0.7004290435817954,
+      "grad_norm": 0.27867869217003954,
+      "learning_rate": 8.621252584098938e-06,
+      "loss": 0.5332,
+      "step": 4265
+    },
+    {
+      "epoch": 0.7005932707901381,
+      "grad_norm": 0.34422747615332705,
+      "learning_rate": 8.621064242515503e-06,
+      "loss": 0.5191,
+      "step": 4266
+    },
+    {
+      "epoch": 0.7007574979984809,
+      "grad_norm": 0.29331677428471403,
+      "learning_rate": 8.620875856173253e-06,
+      "loss": 0.5079,
+      "step": 4267
+    },
+    {
+      "epoch": 0.7009217252068236,
+      "grad_norm": 0.3121609634949846,
+      "learning_rate": 8.620687425074238e-06,
+      "loss": 0.5136,
+      "step": 4268
+    },
+    {
+      "epoch": 0.7010859524151664,
+      "grad_norm": 0.3838693596047789,
+      "learning_rate": 8.620498949220502e-06,
+      "loss": 0.5175,
+      "step": 4269
+    },
+    {
+      "epoch": 0.7012501796235091,
+      "grad_norm": 0.3399582939068176,
+      "learning_rate": 8.620310428614094e-06,
+      "loss": 0.5303,
+      "step": 4270
+    },
+    {
+      "epoch": 0.7014144068318519,
+      "grad_norm": 0.3882307518798506,
+      "learning_rate": 8.620121863257062e-06,
+      "loss": 0.5358,
+      "step": 4271
+    },
+    {
+      "epoch": 0.7015786340401946,
+      "grad_norm": 0.327844178142628,
+      "learning_rate": 8.619933253151452e-06,
+      "loss": 0.5179,
+      "step": 4272
+    },
+    {
+      "epoch": 0.7017428612485374,
+      "grad_norm": 0.3059441881333392,
+      "learning_rate": 8.619744598299315e-06,
+      "loss": 0.5003,
+      "step": 4273
+    },
+    {
+      "epoch": 0.7019070884568801,
+      "grad_norm": 0.33289088821210483,
+      "learning_rate": 8.619555898702695e-06,
+      "loss": 0.5203,
+      "step": 4274
+    },
+    {
+      "epoch": 0.7020713156652229,
+      "grad_norm": 0.40764636675339916,
+      "learning_rate": 8.619367154363647e-06,
+      "loss": 0.5287,
+      "step": 4275
+    },
+    {
+      "epoch": 0.7022355428735656,
+      "grad_norm": 0.32231801015454054,
+      "learning_rate": 8.61917836528422e-06,
+      "loss": 0.5228,
+      "step": 4276
+    },
+    {
+      "epoch": 0.7023997700819083,
+      "grad_norm": 0.36811692912587884,
+      "learning_rate": 8.618989531466462e-06,
+      "loss": 0.5104,
+      "step": 4277
+    },
+    {
+      "epoch": 0.7025639972902511,
+      "grad_norm": 0.3432330040499813,
+      "learning_rate": 8.618800652912425e-06,
+      "loss": 0.5105,
+      "step": 4278
+    },
+    {
+      "epoch": 0.7027282244985938,
+      "grad_norm": 0.3163500424266403,
+      "learning_rate": 8.618611729624161e-06,
+      "loss": 0.536,
+      "step": 4279
+    },
+    {
+      "epoch": 0.7028924517069366,
+      "grad_norm": 0.3334439933008529,
+      "learning_rate": 8.61842276160372e-06,
+      "loss": 0.5218,
+      "step": 4280
+    },
+    {
+      "epoch": 0.7030566789152793,
+      "grad_norm": 0.3651496955106728,
+      "learning_rate": 8.618233748853159e-06,
+      "loss": 0.5291,
+      "step": 4281
+    },
+    {
+      "epoch": 0.703220906123622,
+      "grad_norm": 0.3909489697860476,
+      "learning_rate": 8.618044691374524e-06,
+      "loss": 0.5138,
+      "step": 4282
+    },
+    {
+      "epoch": 0.7033851333319647,
+      "grad_norm": 0.3432299333346993,
+      "learning_rate": 8.617855589169873e-06,
+      "loss": 0.5296,
+      "step": 4283
+    },
+    {
+      "epoch": 0.7035493605403075,
+      "grad_norm": 0.32106670987419533,
+      "learning_rate": 8.617666442241261e-06,
+      "loss": 0.5538,
+      "step": 4284
+    },
+    {
+      "epoch": 0.7037135877486502,
+      "grad_norm": 0.278987267758817,
+      "learning_rate": 8.617477250590737e-06,
+      "loss": 0.5351,
+      "step": 4285
+    },
+    {
+      "epoch": 0.703877814956993,
+      "grad_norm": 0.30910724383387306,
+      "learning_rate": 8.617288014220362e-06,
+      "loss": 0.5326,
+      "step": 4286
+    },
+    {
+      "epoch": 0.7040420421653357,
+      "grad_norm": 0.323362918561536,
+      "learning_rate": 8.617098733132187e-06,
+      "loss": 0.5348,
+      "step": 4287
+    },
+    {
+      "epoch": 0.7042062693736785,
+      "grad_norm": 0.3078079738940741,
+      "learning_rate": 8.616909407328268e-06,
+      "loss": 0.5458,
+      "step": 4288
+    },
+    {
+      "epoch": 0.7043704965820212,
+      "grad_norm": 0.2844380907240079,
+      "learning_rate": 8.616720036810664e-06,
+      "loss": 0.5177,
+      "step": 4289
+    },
+    {
+      "epoch": 0.704534723790364,
+      "grad_norm": 0.32375350229388816,
+      "learning_rate": 8.61653062158143e-06,
+      "loss": 0.5277,
+      "step": 4290
+    },
+    {
+      "epoch": 0.7046989509987067,
+      "grad_norm": 0.3375882355708296,
+      "learning_rate": 8.616341161642622e-06,
+      "loss": 0.5172,
+      "step": 4291
+    },
+    {
+      "epoch": 0.7048631782070495,
+      "grad_norm": 0.2865337715238847,
+      "learning_rate": 8.616151656996301e-06,
+      "loss": 0.5367,
+      "step": 4292
+    },
+    {
+      "epoch": 0.7050274054153922,
+      "grad_norm": 0.3144427358498577,
+      "learning_rate": 8.615962107644523e-06,
+      "loss": 0.4955,
+      "step": 4293
+    },
+    {
+      "epoch": 0.705191632623735,
+      "grad_norm": 0.3656931892003649,
+      "learning_rate": 8.615772513589345e-06,
+      "loss": 0.5221,
+      "step": 4294
+    },
+    {
+      "epoch": 0.7053558598320777,
+      "grad_norm": 0.2958582370426801,
+      "learning_rate": 8.61558287483283e-06,
+      "loss": 0.5537,
+      "step": 4295
+    },
+    {
+      "epoch": 0.7055200870404205,
+      "grad_norm": 0.3768598902611741,
+      "learning_rate": 8.615393191377035e-06,
+      "loss": 0.5235,
+      "step": 4296
+    },
+    {
+      "epoch": 0.7056843142487632,
+      "grad_norm": 0.3347727814313395,
+      "learning_rate": 8.615203463224021e-06,
+      "loss": 0.5174,
+      "step": 4297
+    },
+    {
+      "epoch": 0.705848541457106,
+      "grad_norm": 0.38916588034781613,
+      "learning_rate": 8.615013690375849e-06,
+      "loss": 0.539,
+      "step": 4298
+    },
+    {
+      "epoch": 0.7060127686654486,
+      "grad_norm": 0.27779637428046317,
+      "learning_rate": 8.61482387283458e-06,
+      "loss": 0.5066,
+      "step": 4299
+    },
+    {
+      "epoch": 0.7061769958737913,
+      "grad_norm": 0.5400534178669388,
+      "learning_rate": 8.614634010602274e-06,
+      "loss": 0.508,
+      "step": 4300
+    },
+    {
+      "epoch": 0.7063412230821341,
+      "grad_norm": 0.37039212228978896,
+      "learning_rate": 8.614444103680995e-06,
+      "loss": 0.5402,
+      "step": 4301
+    },
+    {
+      "epoch": 0.7065054502904768,
+      "grad_norm": 0.279075184659078,
+      "learning_rate": 8.614254152072805e-06,
+      "loss": 0.5422,
+      "step": 4302
+    },
+    {
+      "epoch": 0.7066696774988196,
+      "grad_norm": 0.32756207172905416,
+      "learning_rate": 8.614064155779767e-06,
+      "loss": 0.5286,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7068339047071623,
+      "grad_norm": 0.34922911866740997,
+      "learning_rate": 8.613874114803945e-06,
+      "loss": 0.5264,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7069981319155051,
+      "grad_norm": 0.31257689692201895,
+      "learning_rate": 8.613684029147401e-06,
+      "loss": 0.5178,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7071623591238478,
+      "grad_norm": 0.30307569125011224,
+      "learning_rate": 8.613493898812202e-06,
+      "loss": 0.5124,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7073265863321906,
+      "grad_norm": 0.36069449916053775,
+      "learning_rate": 8.613303723800413e-06,
+      "loss": 0.5282,
+      "step": 4307
+    },
+    {
+      "epoch": 0.7074908135405333,
+      "grad_norm": 0.3092916348181574,
+      "learning_rate": 8.6131135041141e-06,
+      "loss": 0.5305,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7076550407488761,
+      "grad_norm": 0.40891948696883457,
+      "learning_rate": 8.612923239755325e-06,
+      "loss": 0.5214,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7078192679572188,
+      "grad_norm": 0.31616164432034843,
+      "learning_rate": 8.612732930726157e-06,
+      "loss": 0.5194,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7079834951655616,
+      "grad_norm": 0.3000505637839565,
+      "learning_rate": 8.612542577028663e-06,
+      "loss": 0.524,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7081477223739043,
+      "grad_norm": 0.3129782682534076,
+      "learning_rate": 8.612352178664911e-06,
+      "loss": 0.5238,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7083119495822471,
+      "grad_norm": 0.3206955929245123,
+      "learning_rate": 8.612161735636968e-06,
+      "loss": 0.5261,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7084761767905898,
+      "grad_norm": 0.37716743758695925,
+      "learning_rate": 8.611971247946904e-06,
+      "loss": 0.4876,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7086404039989326,
+      "grad_norm": 0.40831633759201236,
+      "learning_rate": 8.611780715596786e-06,
+      "loss": 0.5295,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7088046312072752,
+      "grad_norm": 0.304318562490442,
+      "learning_rate": 8.611590138588685e-06,
+      "loss": 0.5323,
+      "step": 4316
+    },
+    {
+      "epoch": 0.708968858415618,
+      "grad_norm": 0.3975476450385827,
+      "learning_rate": 8.611399516924666e-06,
+      "loss": 0.5175,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7091330856239607,
+      "grad_norm": 0.30548294363985984,
+      "learning_rate": 8.611208850606806e-06,
+      "loss": 0.5389,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7092973128323035,
+      "grad_norm": 0.3777837973806069,
+      "learning_rate": 8.611018139637175e-06,
+      "loss": 0.5311,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7094615400406462,
+      "grad_norm": 0.2973347910181853,
+      "learning_rate": 8.61082738401784e-06,
+      "loss": 0.5402,
+      "step": 4320
+    },
+    {
+      "epoch": 0.709625767248989,
+      "grad_norm": 0.36469127314551625,
+      "learning_rate": 8.610636583750874e-06,
+      "loss": 0.5209,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7097899944573317,
+      "grad_norm": 0.28293959836732613,
+      "learning_rate": 8.610445738838352e-06,
+      "loss": 0.5153,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7099542216656745,
+      "grad_norm": 0.42306390379784664,
+      "learning_rate": 8.610254849282345e-06,
+      "loss": 0.5531,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7101184488740172,
+      "grad_norm": 0.28628622993663794,
+      "learning_rate": 8.610063915084926e-06,
+      "loss": 0.5414,
+      "step": 4324
+    },
+    {
+      "epoch": 0.71028267608236,
+      "grad_norm": 0.281324125785271,
+      "learning_rate": 8.609872936248168e-06,
+      "loss": 0.5226,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7104469032907027,
+      "grad_norm": 0.37477344881709673,
+      "learning_rate": 8.609681912774149e-06,
+      "loss": 0.542,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7106111304990455,
+      "grad_norm": 0.3206181721344887,
+      "learning_rate": 8.609490844664938e-06,
+      "loss": 0.5223,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7107753577073882,
+      "grad_norm": 0.3339242114148264,
+      "learning_rate": 8.609299731922616e-06,
+      "loss": 0.5106,
+      "step": 4328
+    },
+    {
+      "epoch": 0.710939584915731,
+      "grad_norm": 0.2983358499194098,
+      "learning_rate": 8.609108574549254e-06,
+      "loss": 0.5397,
+      "step": 4329
+    },
+    {
+      "epoch": 0.7111038121240737,
+      "grad_norm": 0.33380596620035463,
+      "learning_rate": 8.608917372546931e-06,
+      "loss": 0.5411,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7112680393324164,
+      "grad_norm": 0.2698264997063511,
+      "learning_rate": 8.608726125917721e-06,
+      "loss": 0.5073,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7114322665407592,
+      "grad_norm": 0.3947051740162191,
+      "learning_rate": 8.608534834663705e-06,
+      "loss": 0.5249,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7115964937491018,
+      "grad_norm": 0.2906445128721743,
+      "learning_rate": 8.608343498786958e-06,
+      "loss": 0.5403,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7117607209574446,
+      "grad_norm": 0.30410226624743664,
+      "learning_rate": 8.60815211828956e-06,
+      "loss": 0.5335,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7119249481657873,
+      "grad_norm": 0.28918039186839667,
+      "learning_rate": 8.607960693173585e-06,
+      "loss": 0.5053,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7120891753741301,
+      "grad_norm": 0.39784143895681123,
+      "learning_rate": 8.607769223441118e-06,
+      "loss": 0.5263,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7122534025824728,
+      "grad_norm": 0.3816380356849168,
+      "learning_rate": 8.607577709094234e-06,
+      "loss": 0.536,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7124176297908156,
+      "grad_norm": 0.48876799069675336,
+      "learning_rate": 8.607386150135016e-06,
+      "loss": 0.52,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7125818569991583,
+      "grad_norm": 0.30419187380004425,
+      "learning_rate": 8.607194546565541e-06,
+      "loss": 0.5172,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7127460842075011,
+      "grad_norm": 0.3046540779190067,
+      "learning_rate": 8.607002898387894e-06,
+      "loss": 0.5255,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7129103114158438,
+      "grad_norm": 0.3697888338261436,
+      "learning_rate": 8.606811205604155e-06,
+      "loss": 0.5025,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7130745386241866,
+      "grad_norm": 0.32176447793422863,
+      "learning_rate": 8.606619468216403e-06,
+      "loss": 0.4727,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7132387658325293,
+      "grad_norm": 0.372657494023084,
+      "learning_rate": 8.606427686226727e-06,
+      "loss": 0.5283,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7134029930408721,
+      "grad_norm": 0.3190889106654982,
+      "learning_rate": 8.606235859637206e-06,
+      "loss": 0.5326,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7135672202492148,
+      "grad_norm": 0.2928457979647746,
+      "learning_rate": 8.606043988449922e-06,
+      "loss": 0.5128,
+      "step": 4345
+    },
+    {
+      "epoch": 0.7137314474575576,
+      "grad_norm": 0.2706617064341812,
+      "learning_rate": 8.60585207266696e-06,
+      "loss": 0.5224,
+      "step": 4346
+    },
+    {
+      "epoch": 0.7138956746659003,
+      "grad_norm": 0.3205314934694761,
+      "learning_rate": 8.605660112290406e-06,
+      "loss": 0.5028,
+      "step": 4347
+    },
+    {
+      "epoch": 0.7140599018742431,
+      "grad_norm": 0.33291126764659723,
+      "learning_rate": 8.605468107322342e-06,
+      "loss": 0.5144,
+      "step": 4348
+    },
+    {
+      "epoch": 0.7142241290825858,
+      "grad_norm": 0.319633960734903,
+      "learning_rate": 8.605276057764857e-06,
+      "loss": 0.5165,
+      "step": 4349
+    },
+    {
+      "epoch": 0.7143883562909285,
+      "grad_norm": 0.4054703641916312,
+      "learning_rate": 8.605083963620034e-06,
+      "loss": 0.5242,
+      "step": 4350
+    },
+    {
+      "epoch": 0.7145525834992712,
+      "grad_norm": 0.3103957865806984,
+      "learning_rate": 8.604891824889961e-06,
+      "loss": 0.5218,
+      "step": 4351
+    },
+    {
+      "epoch": 0.714716810707614,
+      "grad_norm": 0.3180642837392691,
+      "learning_rate": 8.604699641576723e-06,
+      "loss": 0.528,
+      "step": 4352
+    },
+    {
+      "epoch": 0.7148810379159567,
+      "grad_norm": 0.3133787583026555,
+      "learning_rate": 8.604507413682409e-06,
+      "loss": 0.507,
+      "step": 4353
+    },
+    {
+      "epoch": 0.7150452651242994,
+      "grad_norm": 0.33877895632971905,
+      "learning_rate": 8.604315141209108e-06,
+      "loss": 0.5238,
+      "step": 4354
+    },
+    {
+      "epoch": 0.7152094923326422,
+      "grad_norm": 0.38840289489043733,
+      "learning_rate": 8.604122824158905e-06,
+      "loss": 0.5504,
+      "step": 4355
+    },
+    {
+      "epoch": 0.7153737195409849,
+      "grad_norm": 0.3955477841512037,
+      "learning_rate": 8.603930462533889e-06,
+      "loss": 0.4993,
+      "step": 4356
+    },
+    {
+      "epoch": 0.7155379467493277,
+      "grad_norm": 0.27883260117502756,
+      "learning_rate": 8.603738056336152e-06,
+      "loss": 0.5121,
+      "step": 4357
+    },
+    {
+      "epoch": 0.7157021739576704,
+      "grad_norm": 0.2850443117354454,
+      "learning_rate": 8.603545605567782e-06,
+      "loss": 0.525,
+      "step": 4358
+    },
+    {
+      "epoch": 0.7158664011660132,
+      "grad_norm": 0.43399050909543463,
+      "learning_rate": 8.603353110230868e-06,
+      "loss": 0.5265,
+      "step": 4359
+    },
+    {
+      "epoch": 0.7160306283743559,
+      "grad_norm": 0.31443791711799157,
+      "learning_rate": 8.603160570327504e-06,
+      "loss": 0.535,
+      "step": 4360
+    },
+    {
+      "epoch": 0.7161948555826987,
+      "grad_norm": 0.2939906689274626,
+      "learning_rate": 8.602967985859779e-06,
+      "loss": 0.5262,
+      "step": 4361
+    },
+    {
+      "epoch": 0.7163590827910414,
+      "grad_norm": 0.3551573997835369,
+      "learning_rate": 8.602775356829783e-06,
+      "loss": 0.5221,
+      "step": 4362
+    },
+    {
+      "epoch": 0.7165233099993842,
+      "grad_norm": 0.37242204799922163,
+      "learning_rate": 8.602582683239612e-06,
+      "loss": 0.5108,
+      "step": 4363
+    },
+    {
+      "epoch": 0.7166875372077269,
+      "grad_norm": 0.3301229591082884,
+      "learning_rate": 8.602389965091357e-06,
+      "loss": 0.5033,
+      "step": 4364
+    },
+    {
+      "epoch": 0.7168517644160697,
+      "grad_norm": 0.5596674893832613,
+      "learning_rate": 8.602197202387109e-06,
+      "loss": 0.5257,
+      "step": 4365
+    },
+    {
+      "epoch": 0.7170159916244124,
+      "grad_norm": 0.32003899380645795,
+      "learning_rate": 8.602004395128963e-06,
+      "loss": 0.5096,
+      "step": 4366
+    },
+    {
+      "epoch": 0.7171802188327551,
+      "grad_norm": 0.32101527283833253,
+      "learning_rate": 8.601811543319016e-06,
+      "loss": 0.5154,
+      "step": 4367
+    },
+    {
+      "epoch": 0.7173444460410978,
+      "grad_norm": 0.3582105517657474,
+      "learning_rate": 8.601618646959359e-06,
+      "loss": 0.5368,
+      "step": 4368
+    },
+    {
+      "epoch": 0.7175086732494406,
+      "grad_norm": 0.3834938893158892,
+      "learning_rate": 8.601425706052086e-06,
+      "loss": 0.5245,
+      "step": 4369
+    },
+    {
+      "epoch": 0.7176729004577833,
+      "grad_norm": 0.31214498157447496,
+      "learning_rate": 8.601232720599298e-06,
+      "loss": 0.5392,
+      "step": 4370
+    },
+    {
+      "epoch": 0.7178371276661261,
+      "grad_norm": 0.2968901286066325,
+      "learning_rate": 8.601039690603085e-06,
+      "loss": 0.5512,
+      "step": 4371
+    },
+    {
+      "epoch": 0.7180013548744688,
+      "grad_norm": 0.3226767453514702,
+      "learning_rate": 8.600846616065546e-06,
+      "loss": 0.5344,
+      "step": 4372
+    },
+    {
+      "epoch": 0.7181655820828116,
+      "grad_norm": 0.28203986767629385,
+      "learning_rate": 8.600653496988781e-06,
+      "loss": 0.5252,
+      "step": 4373
+    },
+    {
+      "epoch": 0.7183298092911543,
+      "grad_norm": 0.3263527602745717,
+      "learning_rate": 8.60046033337488e-06,
+      "loss": 0.5381,
+      "step": 4374
+    },
+    {
+      "epoch": 0.7184940364994971,
+      "grad_norm": 0.3830517652455408,
+      "learning_rate": 8.600267125225948e-06,
+      "loss": 0.5363,
+      "step": 4375
+    },
+    {
+      "epoch": 0.7186582637078398,
+      "grad_norm": 0.3124628134298113,
+      "learning_rate": 8.60007387254408e-06,
+      "loss": 0.5039,
+      "step": 4376
+    },
+    {
+      "epoch": 0.7188224909161826,
+      "grad_norm": 0.4344770371599259,
+      "learning_rate": 8.599880575331379e-06,
+      "loss": 0.5167,
+      "step": 4377
+    },
+    {
+      "epoch": 0.7189867181245253,
+      "grad_norm": 0.44022478012369026,
+      "learning_rate": 8.599687233589938e-06,
+      "loss": 0.5214,
+      "step": 4378
+    },
+    {
+      "epoch": 0.719150945332868,
+      "grad_norm": 0.3688156748206383,
+      "learning_rate": 8.599493847321862e-06,
+      "loss": 0.5206,
+      "step": 4379
+    },
+    {
+      "epoch": 0.7193151725412108,
+      "grad_norm": 0.32935829629851043,
+      "learning_rate": 8.599300416529249e-06,
+      "loss": 0.5141,
+      "step": 4380
+    },
+    {
+      "epoch": 0.7194793997495535,
+      "grad_norm": 0.29232395104565306,
+      "learning_rate": 8.599106941214199e-06,
+      "loss": 0.5259,
+      "step": 4381
+    },
+    {
+      "epoch": 0.7196436269578963,
+      "grad_norm": 0.3230601984030344,
+      "learning_rate": 8.598913421378815e-06,
+      "loss": 0.5157,
+      "step": 4382
+    },
+    {
+      "epoch": 0.7198078541662389,
+      "grad_norm": 0.3366778464764349,
+      "learning_rate": 8.5987198570252e-06,
+      "loss": 0.5163,
+      "step": 4383
+    },
+    {
+      "epoch": 0.7199720813745817,
+      "grad_norm": 0.2822882327510804,
+      "learning_rate": 8.598526248155453e-06,
+      "loss": 0.5302,
+      "step": 4384
+    },
+    {
+      "epoch": 0.7201363085829244,
+      "grad_norm": 0.34913151065251047,
+      "learning_rate": 8.598332594771678e-06,
+      "loss": 0.5239,
+      "step": 4385
+    },
+    {
+      "epoch": 0.7203005357912672,
+      "grad_norm": 0.30615999757517137,
+      "learning_rate": 8.598138896875982e-06,
+      "loss": 0.5333,
+      "step": 4386
+    },
+    {
+      "epoch": 0.7204647629996099,
+      "grad_norm": 0.3338081110150123,
+      "learning_rate": 8.597945154470462e-06,
+      "loss": 0.55,
+      "step": 4387
+    },
+    {
+      "epoch": 0.7206289902079527,
+      "grad_norm": 0.27003702556268694,
+      "learning_rate": 8.597751367557229e-06,
+      "loss": 0.5179,
+      "step": 4388
+    },
+    {
+      "epoch": 0.7207932174162954,
+      "grad_norm": 0.3166293407081746,
+      "learning_rate": 8.59755753613838e-06,
+      "loss": 0.5193,
+      "step": 4389
+    },
+    {
+      "epoch": 0.7209574446246382,
+      "grad_norm": 0.2970707814051773,
+      "learning_rate": 8.597363660216028e-06,
+      "loss": 0.5089,
+      "step": 4390
+    },
+    {
+      "epoch": 0.7211216718329809,
+      "grad_norm": 0.4042859510774983,
+      "learning_rate": 8.597169739792276e-06,
+      "loss": 0.5184,
+      "step": 4391
+    },
+    {
+      "epoch": 0.7212858990413237,
+      "grad_norm": 0.36079226884819154,
+      "learning_rate": 8.596975774869229e-06,
+      "loss": 0.5115,
+      "step": 4392
+    },
+    {
+      "epoch": 0.7214501262496664,
+      "grad_norm": 0.2960898042956307,
+      "learning_rate": 8.596781765448994e-06,
+      "loss": 0.5214,
+      "step": 4393
+    },
+    {
+      "epoch": 0.7216143534580092,
+      "grad_norm": 0.3067528269205476,
+      "learning_rate": 8.596587711533677e-06,
+      "loss": 0.5404,
+      "step": 4394
+    },
+    {
+      "epoch": 0.7217785806663519,
+      "grad_norm": 0.5567244165497772,
+      "learning_rate": 8.596393613125387e-06,
+      "loss": 0.5322,
+      "step": 4395
+    },
+    {
+      "epoch": 0.7219428078746947,
+      "grad_norm": 0.31549950117169867,
+      "learning_rate": 8.596199470226234e-06,
+      "loss": 0.5202,
+      "step": 4396
+    },
+    {
+      "epoch": 0.7221070350830374,
+      "grad_norm": 0.3558221536624943,
+      "learning_rate": 8.596005282838324e-06,
+      "loss": 0.5098,
+      "step": 4397
+    },
+    {
+      "epoch": 0.7222712622913802,
+      "grad_norm": 0.3104817339740221,
+      "learning_rate": 8.595811050963767e-06,
+      "loss": 0.5341,
+      "step": 4398
+    },
+    {
+      "epoch": 0.7224354894997229,
+      "grad_norm": 0.32786752772407324,
+      "learning_rate": 8.59561677460467e-06,
+      "loss": 0.5089,
+      "step": 4399
+    },
+    {
+      "epoch": 0.7225997167080656,
+      "grad_norm": 0.4047878987733635,
+      "learning_rate": 8.595422453763149e-06,
+      "loss": 0.5458,
+      "step": 4400
+    },
+    {
+      "epoch": 0.7227639439164083,
+      "grad_norm": 0.2858382446467845,
+      "learning_rate": 8.595228088441307e-06,
+      "loss": 0.5354,
+      "step": 4401
+    },
+    {
+      "epoch": 0.722928171124751,
+      "grad_norm": 0.32023432751026115,
+      "learning_rate": 8.595033678641261e-06,
+      "loss": 0.5254,
+      "step": 4402
+    },
+    {
+      "epoch": 0.7230923983330938,
+      "grad_norm": 0.3045500747617058,
+      "learning_rate": 8.594839224365119e-06,
+      "loss": 0.521,
+      "step": 4403
+    },
+    {
+      "epoch": 0.7232566255414365,
+      "grad_norm": 0.29098460712999963,
+      "learning_rate": 8.594644725614995e-06,
+      "loss": 0.5173,
+      "step": 4404
+    },
+    {
+      "epoch": 0.7234208527497793,
+      "grad_norm": 0.30702617159676615,
+      "learning_rate": 8.594450182393002e-06,
+      "loss": 0.5294,
+      "step": 4405
+    },
+    {
+      "epoch": 0.723585079958122,
+      "grad_norm": 0.3657684663544443,
+      "learning_rate": 8.59425559470125e-06,
+      "loss": 0.5244,
+      "step": 4406
+    },
+    {
+      "epoch": 0.7237493071664648,
+      "grad_norm": 0.33389501184795556,
+      "learning_rate": 8.594060962541855e-06,
+      "loss": 0.5379,
+      "step": 4407
+    },
+    {
+      "epoch": 0.7239135343748075,
+      "grad_norm": 0.3401852810246631,
+      "learning_rate": 8.593866285916928e-06,
+      "loss": 0.523,
+      "step": 4408
+    },
+    {
+      "epoch": 0.7240777615831503,
+      "grad_norm": 0.31282359408606597,
+      "learning_rate": 8.593671564828587e-06,
+      "loss": 0.5356,
+      "step": 4409
+    },
+    {
+      "epoch": 0.724241988791493,
+      "grad_norm": 0.2828881931783121,
+      "learning_rate": 8.593476799278945e-06,
+      "loss": 0.5176,
+      "step": 4410
+    },
+    {
+      "epoch": 0.7244062159998358,
+      "grad_norm": 0.28740377845431986,
+      "learning_rate": 8.593281989270117e-06,
+      "loss": 0.5182,
+      "step": 4411
+    },
+    {
+      "epoch": 0.7245704432081785,
+      "grad_norm": 0.32739805868398925,
+      "learning_rate": 8.59308713480422e-06,
+      "loss": 0.5234,
+      "step": 4412
+    },
+    {
+      "epoch": 0.7247346704165213,
+      "grad_norm": 0.31856973622465923,
+      "learning_rate": 8.59289223588337e-06,
+      "loss": 0.5102,
+      "step": 4413
+    },
+    {
+      "epoch": 0.724898897624864,
+      "grad_norm": 0.30433947412820167,
+      "learning_rate": 8.592697292509682e-06,
+      "loss": 0.5323,
+      "step": 4414
+    },
+    {
+      "epoch": 0.7250631248332068,
+      "grad_norm": 0.3093281266377262,
+      "learning_rate": 8.592502304685274e-06,
+      "loss": 0.4877,
+      "step": 4415
+    },
+    {
+      "epoch": 0.7252273520415495,
+      "grad_norm": 0.4192937983246652,
+      "learning_rate": 8.592307272412267e-06,
+      "loss": 0.5315,
+      "step": 4416
+    },
+    {
+      "epoch": 0.7253915792498922,
+      "grad_norm": 0.43350275798950666,
+      "learning_rate": 8.592112195692776e-06,
+      "loss": 0.5063,
+      "step": 4417
+    },
+    {
+      "epoch": 0.7255558064582349,
+      "grad_norm": 0.4678623945111824,
+      "learning_rate": 8.591917074528921e-06,
+      "loss": 0.5424,
+      "step": 4418
+    },
+    {
+      "epoch": 0.7257200336665777,
+      "grad_norm": 0.287898271867044,
+      "learning_rate": 8.59172190892282e-06,
+      "loss": 0.5038,
+      "step": 4419
+    },
+    {
+      "epoch": 0.7258842608749204,
+      "grad_norm": 0.29822897607011906,
+      "learning_rate": 8.591526698876592e-06,
+      "loss": 0.5165,
+      "step": 4420
+    },
+    {
+      "epoch": 0.7260484880832632,
+      "grad_norm": 0.3109623634429174,
+      "learning_rate": 8.591331444392361e-06,
+      "loss": 0.5296,
+      "step": 4421
+    },
+    {
+      "epoch": 0.7262127152916059,
+      "grad_norm": 0.3741682736438768,
+      "learning_rate": 8.591136145472244e-06,
+      "loss": 0.5166,
+      "step": 4422
+    },
+    {
+      "epoch": 0.7263769424999487,
+      "grad_norm": 0.3281684045665565,
+      "learning_rate": 8.590940802118363e-06,
+      "loss": 0.4932,
+      "step": 4423
+    },
+    {
+      "epoch": 0.7265411697082914,
+      "grad_norm": 0.2967006165698348,
+      "learning_rate": 8.59074541433284e-06,
+      "loss": 0.5341,
+      "step": 4424
+    },
+    {
+      "epoch": 0.7267053969166342,
+      "grad_norm": 0.41241388629963116,
+      "learning_rate": 8.590549982117798e-06,
+      "loss": 0.5022,
+      "step": 4425
+    },
+    {
+      "epoch": 0.7268696241249769,
+      "grad_norm": 0.3291196432580282,
+      "learning_rate": 8.590354505475357e-06,
+      "loss": 0.5139,
+      "step": 4426
+    },
+    {
+      "epoch": 0.7270338513333197,
+      "grad_norm": 0.32679147888924837,
+      "learning_rate": 8.590158984407644e-06,
+      "loss": 0.5137,
+      "step": 4427
+    },
+    {
+      "epoch": 0.7271980785416624,
+      "grad_norm": 0.28783281690844026,
+      "learning_rate": 8.589963418916778e-06,
+      "loss": 0.5296,
+      "step": 4428
+    },
+    {
+      "epoch": 0.7273623057500052,
+      "grad_norm": 0.31121651109770526,
+      "learning_rate": 8.589767809004886e-06,
+      "loss": 0.5407,
+      "step": 4429
+    },
+    {
+      "epoch": 0.7275265329583479,
+      "grad_norm": 0.3231129949957135,
+      "learning_rate": 8.589572154674093e-06,
+      "loss": 0.5187,
+      "step": 4430
+    },
+    {
+      "epoch": 0.7276907601666907,
+      "grad_norm": 0.35344471569362507,
+      "learning_rate": 8.589376455926521e-06,
+      "loss": 0.5165,
+      "step": 4431
+    },
+    {
+      "epoch": 0.7278549873750334,
+      "grad_norm": 0.30854070183351756,
+      "learning_rate": 8.589180712764298e-06,
+      "loss": 0.5286,
+      "step": 4432
+    },
+    {
+      "epoch": 0.7280192145833762,
+      "grad_norm": 0.3771827531386895,
+      "learning_rate": 8.58898492518955e-06,
+      "loss": 0.5381,
+      "step": 4433
+    },
+    {
+      "epoch": 0.7281834417917188,
+      "grad_norm": 0.3842022509196105,
+      "learning_rate": 8.588789093204402e-06,
+      "loss": 0.5166,
+      "step": 4434
+    },
+    {
+      "epoch": 0.7283476690000615,
+      "grad_norm": 0.3276679515164658,
+      "learning_rate": 8.58859321681098e-06,
+      "loss": 0.5236,
+      "step": 4435
+    },
+    {
+      "epoch": 0.7285118962084043,
+      "grad_norm": 0.32004257785016976,
+      "learning_rate": 8.588397296011416e-06,
+      "loss": 0.5196,
+      "step": 4436
+    },
+    {
+      "epoch": 0.728676123416747,
+      "grad_norm": 0.33663249793293937,
+      "learning_rate": 8.588201330807833e-06,
+      "loss": 0.5278,
+      "step": 4437
+    },
+    {
+      "epoch": 0.7288403506250898,
+      "grad_norm": 0.33869601572570573,
+      "learning_rate": 8.588005321202361e-06,
+      "loss": 0.5294,
+      "step": 4438
+    },
+    {
+      "epoch": 0.7290045778334325,
+      "grad_norm": 0.29466669461762524,
+      "learning_rate": 8.587809267197132e-06,
+      "loss": 0.5347,
+      "step": 4439
+    },
+    {
+      "epoch": 0.7291688050417753,
+      "grad_norm": 0.32128430605037284,
+      "learning_rate": 8.587613168794269e-06,
+      "loss": 0.5244,
+      "step": 4440
+    },
+    {
+      "epoch": 0.729333032250118,
+      "grad_norm": 0.32825756800992645,
+      "learning_rate": 8.587417025995909e-06,
+      "loss": 0.5339,
+      "step": 4441
+    },
+    {
+      "epoch": 0.7294972594584608,
+      "grad_norm": 0.3301350950702061,
+      "learning_rate": 8.587220838804176e-06,
+      "loss": 0.5109,
+      "step": 4442
+    },
+    {
+      "epoch": 0.7296614866668035,
+      "grad_norm": 0.28104238952110344,
+      "learning_rate": 8.587024607221203e-06,
+      "loss": 0.5046,
+      "step": 4443
+    },
+    {
+      "epoch": 0.7298257138751463,
+      "grad_norm": 0.37224012508447873,
+      "learning_rate": 8.586828331249123e-06,
+      "loss": 0.5169,
+      "step": 4444
+    },
+    {
+      "epoch": 0.729989941083489,
+      "grad_norm": 0.39243426907646256,
+      "learning_rate": 8.586632010890065e-06,
+      "loss": 0.4953,
+      "step": 4445
+    },
+    {
+      "epoch": 0.7301541682918318,
+      "grad_norm": 0.3699144774546227,
+      "learning_rate": 8.586435646146164e-06,
+      "loss": 0.5161,
+      "step": 4446
+    },
+    {
+      "epoch": 0.7303183955001745,
+      "grad_norm": 0.403502931249784,
+      "learning_rate": 8.586239237019552e-06,
+      "loss": 0.5144,
+      "step": 4447
+    },
+    {
+      "epoch": 0.7304826227085173,
+      "grad_norm": 0.33530858772743866,
+      "learning_rate": 8.586042783512361e-06,
+      "loss": 0.5065,
+      "step": 4448
+    },
+    {
+      "epoch": 0.73064684991686,
+      "grad_norm": 0.36473692233901733,
+      "learning_rate": 8.585846285626724e-06,
+      "loss": 0.5289,
+      "step": 4449
+    },
+    {
+      "epoch": 0.7308110771252028,
+      "grad_norm": 0.39727298026229163,
+      "learning_rate": 8.585649743364778e-06,
+      "loss": 0.5261,
+      "step": 4450
+    },
+    {
+      "epoch": 0.7309753043335454,
+      "grad_norm": 0.34627397363700524,
+      "learning_rate": 8.585453156728655e-06,
+      "loss": 0.4968,
+      "step": 4451
+    },
+    {
+      "epoch": 0.7311395315418882,
+      "grad_norm": 0.3024856634002474,
+      "learning_rate": 8.58525652572049e-06,
+      "loss": 0.5308,
+      "step": 4452
+    },
+    {
+      "epoch": 0.7313037587502309,
+      "grad_norm": 0.3052356152052224,
+      "learning_rate": 8.585059850342422e-06,
+      "loss": 0.4987,
+      "step": 4453
+    },
+    {
+      "epoch": 0.7314679859585737,
+      "grad_norm": 0.3868676594902889,
+      "learning_rate": 8.584863130596584e-06,
+      "loss": 0.5319,
+      "step": 4454
+    },
+    {
+      "epoch": 0.7316322131669164,
+      "grad_norm": 0.3114269144053163,
+      "learning_rate": 8.584666366485115e-06,
+      "loss": 0.5191,
+      "step": 4455
+    },
+    {
+      "epoch": 0.7317964403752591,
+      "grad_norm": 0.32173171845094,
+      "learning_rate": 8.584469558010148e-06,
+      "loss": 0.5146,
+      "step": 4456
+    },
+    {
+      "epoch": 0.7319606675836019,
+      "grad_norm": 0.32749605605918813,
+      "learning_rate": 8.584272705173824e-06,
+      "loss": 0.5223,
+      "step": 4457
+    },
+    {
+      "epoch": 0.7321248947919446,
+      "grad_norm": 0.35873045536107073,
+      "learning_rate": 8.58407580797828e-06,
+      "loss": 0.5307,
+      "step": 4458
+    },
+    {
+      "epoch": 0.7322891220002874,
+      "grad_norm": 0.33696054538783504,
+      "learning_rate": 8.583878866425656e-06,
+      "loss": 0.5112,
+      "step": 4459
+    },
+    {
+      "epoch": 0.7324533492086301,
+      "grad_norm": 0.322150007734754,
+      "learning_rate": 8.583681880518088e-06,
+      "loss": 0.5408,
+      "step": 4460
+    },
+    {
+      "epoch": 0.7326175764169729,
+      "grad_norm": 0.3577905586167644,
+      "learning_rate": 8.583484850257717e-06,
+      "loss": 0.5247,
+      "step": 4461
+    },
+    {
+      "epoch": 0.7327818036253156,
+      "grad_norm": 0.3890314954445434,
+      "learning_rate": 8.583287775646683e-06,
+      "loss": 0.5403,
+      "step": 4462
+    },
+    {
+      "epoch": 0.7329460308336584,
+      "grad_norm": 0.322614607637845,
+      "learning_rate": 8.583090656687126e-06,
+      "loss": 0.5116,
+      "step": 4463
+    },
+    {
+      "epoch": 0.7331102580420011,
+      "grad_norm": 0.37697488192045586,
+      "learning_rate": 8.582893493381187e-06,
+      "loss": 0.5333,
+      "step": 4464
+    },
+    {
+      "epoch": 0.7332744852503439,
+      "grad_norm": 0.37182964027775867,
+      "learning_rate": 8.58269628573101e-06,
+      "loss": 0.5285,
+      "step": 4465
+    },
+    {
+      "epoch": 0.7334387124586866,
+      "grad_norm": 0.31262978823905135,
+      "learning_rate": 8.582499033738732e-06,
+      "loss": 0.5342,
+      "step": 4466
+    },
+    {
+      "epoch": 0.7336029396670294,
+      "grad_norm": 0.29094147937733944,
+      "learning_rate": 8.582301737406498e-06,
+      "loss": 0.522,
+      "step": 4467
+    },
+    {
+      "epoch": 0.733767166875372,
+      "grad_norm": 0.3595441468442823,
+      "learning_rate": 8.582104396736453e-06,
+      "loss": 0.5367,
+      "step": 4468
+    },
+    {
+      "epoch": 0.7339313940837148,
+      "grad_norm": 0.32354047052265966,
+      "learning_rate": 8.581907011730735e-06,
+      "loss": 0.5146,
+      "step": 4469
+    },
+    {
+      "epoch": 0.7340956212920575,
+      "grad_norm": 0.3434044743236067,
+      "learning_rate": 8.581709582391492e-06,
+      "loss": 0.5308,
+      "step": 4470
+    },
+    {
+      "epoch": 0.7342598485004003,
+      "grad_norm": 0.2966580125366594,
+      "learning_rate": 8.581512108720868e-06,
+      "loss": 0.5352,
+      "step": 4471
+    },
+    {
+      "epoch": 0.734424075708743,
+      "grad_norm": 0.3034642435029345,
+      "learning_rate": 8.581314590721006e-06,
+      "loss": 0.529,
+      "step": 4472
+    },
+    {
+      "epoch": 0.7345883029170858,
+      "grad_norm": 0.3172501587579415,
+      "learning_rate": 8.581117028394052e-06,
+      "loss": 0.4962,
+      "step": 4473
+    },
+    {
+      "epoch": 0.7347525301254285,
+      "grad_norm": 0.3057272334680113,
+      "learning_rate": 8.580919421742153e-06,
+      "loss": 0.5362,
+      "step": 4474
+    },
+    {
+      "epoch": 0.7349167573337713,
+      "grad_norm": 0.27472219505829215,
+      "learning_rate": 8.580721770767452e-06,
+      "loss": 0.5451,
+      "step": 4475
+    },
+    {
+      "epoch": 0.735080984542114,
+      "grad_norm": 0.3101713531189586,
+      "learning_rate": 8.580524075472099e-06,
+      "loss": 0.5333,
+      "step": 4476
+    },
+    {
+      "epoch": 0.7352452117504568,
+      "grad_norm": 0.3294719891527552,
+      "learning_rate": 8.58032633585824e-06,
+      "loss": 0.5193,
+      "step": 4477
+    },
+    {
+      "epoch": 0.7354094389587995,
+      "grad_norm": 2.171350755473058,
+      "learning_rate": 8.580128551928021e-06,
+      "loss": 0.4927,
+      "step": 4478
+    },
+    {
+      "epoch": 0.7355736661671423,
+      "grad_norm": 0.3730561619746024,
+      "learning_rate": 8.579930723683592e-06,
+      "loss": 0.4969,
+      "step": 4479
+    },
+    {
+      "epoch": 0.735737893375485,
+      "grad_norm": 0.5268860411236578,
+      "learning_rate": 8.579732851127102e-06,
+      "loss": 0.5256,
+      "step": 4480
+    },
+    {
+      "epoch": 0.7359021205838278,
+      "grad_norm": 0.3387582635476446,
+      "learning_rate": 8.5795349342607e-06,
+      "loss": 0.5631,
+      "step": 4481
+    },
+    {
+      "epoch": 0.7360663477921705,
+      "grad_norm": 0.28861855998881153,
+      "learning_rate": 8.579336973086535e-06,
+      "loss": 0.5199,
+      "step": 4482
+    },
+    {
+      "epoch": 0.7362305750005133,
+      "grad_norm": 0.2942221387932248,
+      "learning_rate": 8.579138967606755e-06,
+      "loss": 0.5036,
+      "step": 4483
+    },
+    {
+      "epoch": 0.736394802208856,
+      "grad_norm": 1.0358662167933195,
+      "learning_rate": 8.578940917823514e-06,
+      "loss": 0.5111,
+      "step": 4484
+    },
+    {
+      "epoch": 0.7365590294171986,
+      "grad_norm": 0.37438050679631024,
+      "learning_rate": 8.578742823738961e-06,
+      "loss": 0.5173,
+      "step": 4485
+    },
+    {
+      "epoch": 0.7367232566255414,
+      "grad_norm": 0.3277582159694442,
+      "learning_rate": 8.578544685355248e-06,
+      "loss": 0.528,
+      "step": 4486
+    },
+    {
+      "epoch": 0.7368874838338841,
+      "grad_norm": 0.30091880038912744,
+      "learning_rate": 8.578346502674526e-06,
+      "loss": 0.5231,
+      "step": 4487
+    },
+    {
+      "epoch": 0.7370517110422269,
+      "grad_norm": 0.3074065546946728,
+      "learning_rate": 8.578148275698951e-06,
+      "loss": 0.5222,
+      "step": 4488
+    },
+    {
+      "epoch": 0.7372159382505696,
+      "grad_norm": 0.354491887670448,
+      "learning_rate": 8.577950004430672e-06,
+      "loss": 0.5327,
+      "step": 4489
+    },
+    {
+      "epoch": 0.7373801654589124,
+      "grad_norm": 0.29587883852898034,
+      "learning_rate": 8.577751688871842e-06,
+      "loss": 0.5161,
+      "step": 4490
+    },
+    {
+      "epoch": 0.7375443926672551,
+      "grad_norm": 0.3460943354301366,
+      "learning_rate": 8.577553329024618e-06,
+      "loss": 0.5146,
+      "step": 4491
+    },
+    {
+      "epoch": 0.7377086198755979,
+      "grad_norm": 0.2967950619579812,
+      "learning_rate": 8.577354924891155e-06,
+      "loss": 0.53,
+      "step": 4492
+    },
+    {
+      "epoch": 0.7378728470839406,
+      "grad_norm": 0.342330627577297,
+      "learning_rate": 8.577156476473603e-06,
+      "loss": 0.5101,
+      "step": 4493
+    },
+    {
+      "epoch": 0.7380370742922834,
+      "grad_norm": 0.2936759776996891,
+      "learning_rate": 8.576957983774123e-06,
+      "loss": 0.5217,
+      "step": 4494
+    },
+    {
+      "epoch": 0.7382013015006261,
+      "grad_norm": 0.3769017994359706,
+      "learning_rate": 8.576759446794865e-06,
+      "loss": 0.5267,
+      "step": 4495
+    },
+    {
+      "epoch": 0.7383655287089689,
+      "grad_norm": 0.3243258117086145,
+      "learning_rate": 8.57656086553799e-06,
+      "loss": 0.5333,
+      "step": 4496
+    },
+    {
+      "epoch": 0.7385297559173116,
+      "grad_norm": 0.288283887436931,
+      "learning_rate": 8.576362240005653e-06,
+      "loss": 0.5028,
+      "step": 4497
+    },
+    {
+      "epoch": 0.7386939831256544,
+      "grad_norm": 0.47367175940020345,
+      "learning_rate": 8.576163570200013e-06,
+      "loss": 0.5243,
+      "step": 4498
+    },
+    {
+      "epoch": 0.7388582103339971,
+      "grad_norm": 0.33129672486938455,
+      "learning_rate": 8.575964856123224e-06,
+      "loss": 0.5116,
+      "step": 4499
+    },
+    {
+      "epoch": 0.7390224375423399,
+      "grad_norm": 0.30974731627218893,
+      "learning_rate": 8.575766097777447e-06,
+      "loss": 0.5142,
+      "step": 4500
+    },
+    {
+      "epoch": 0.7391866647506826,
+      "grad_norm": 0.32922854453451555,
+      "learning_rate": 8.575567295164842e-06,
+      "loss": 0.5181,
+      "step": 4501
+    },
+    {
+      "epoch": 0.7393508919590253,
+      "grad_norm": 0.3561359117611993,
+      "learning_rate": 8.575368448287564e-06,
+      "loss": 0.5343,
+      "step": 4502
+    },
+    {
+      "epoch": 0.739515119167368,
+      "grad_norm": 0.30225794184268934,
+      "learning_rate": 8.575169557147775e-06,
+      "loss": 0.5407,
+      "step": 4503
+    },
+    {
+      "epoch": 0.7396793463757108,
+      "grad_norm": 0.3190486485860462,
+      "learning_rate": 8.574970621747636e-06,
+      "loss": 0.5242,
+      "step": 4504
+    },
+    {
+      "epoch": 0.7398435735840535,
+      "grad_norm": 1.4215271237322333,
+      "learning_rate": 8.57477164208931e-06,
+      "loss": 0.4993,
+      "step": 4505
+    },
+    {
+      "epoch": 0.7400078007923963,
+      "grad_norm": 0.303708837722683,
+      "learning_rate": 8.574572618174951e-06,
+      "loss": 0.5197,
+      "step": 4506
+    },
+    {
+      "epoch": 0.740172028000739,
+      "grad_norm": 0.3312313092554724,
+      "learning_rate": 8.574373550006724e-06,
+      "loss": 0.5142,
+      "step": 4507
+    },
+    {
+      "epoch": 0.7403362552090818,
+      "grad_norm": 0.3313425846200784,
+      "learning_rate": 8.574174437586794e-06,
+      "loss": 0.5084,
+      "step": 4508
+    },
+    {
+      "epoch": 0.7405004824174245,
+      "grad_norm": 0.34204198699638977,
+      "learning_rate": 8.573975280917321e-06,
+      "loss": 0.5148,
+      "step": 4509
+    },
+    {
+      "epoch": 0.7406647096257672,
+      "grad_norm": 0.2988610461822118,
+      "learning_rate": 8.573776080000466e-06,
+      "loss": 0.5205,
+      "step": 4510
+    },
+    {
+      "epoch": 0.74082893683411,
+      "grad_norm": 0.28717571034744444,
+      "learning_rate": 8.573576834838397e-06,
+      "loss": 0.528,
+      "step": 4511
+    },
+    {
+      "epoch": 0.7409931640424527,
+      "grad_norm": 0.317583520589486,
+      "learning_rate": 8.573377545433275e-06,
+      "loss": 0.5248,
+      "step": 4512
+    },
+    {
+      "epoch": 0.7411573912507955,
+      "grad_norm": 0.32758612071497484,
+      "learning_rate": 8.573178211787266e-06,
+      "loss": 0.5128,
+      "step": 4513
+    },
+    {
+      "epoch": 0.7413216184591382,
+      "grad_norm": 0.33237284591262994,
+      "learning_rate": 8.572978833902531e-06,
+      "loss": 0.5349,
+      "step": 4514
+    },
+    {
+      "epoch": 0.741485845667481,
+      "grad_norm": 0.42198291061829296,
+      "learning_rate": 8.572779411781242e-06,
+      "loss": 0.5221,
+      "step": 4515
+    },
+    {
+      "epoch": 0.7416500728758237,
+      "grad_norm": 0.31109245246405137,
+      "learning_rate": 8.57257994542556e-06,
+      "loss": 0.5072,
+      "step": 4516
+    },
+    {
+      "epoch": 0.7418143000841665,
+      "grad_norm": 0.2980861107722376,
+      "learning_rate": 8.572380434837653e-06,
+      "loss": 0.5141,
+      "step": 4517
+    },
+    {
+      "epoch": 0.7419785272925092,
+      "grad_norm": 0.30924696499077564,
+      "learning_rate": 8.572180880019688e-06,
+      "loss": 0.5352,
+      "step": 4518
+    },
+    {
+      "epoch": 0.7421427545008519,
+      "grad_norm": 0.339786312302817,
+      "learning_rate": 8.571981280973832e-06,
+      "loss": 0.5196,
+      "step": 4519
+    },
+    {
+      "epoch": 0.7423069817091946,
+      "grad_norm": 0.274055301165695,
+      "learning_rate": 8.571781637702254e-06,
+      "loss": 0.5105,
+      "step": 4520
+    },
+    {
+      "epoch": 0.7424712089175374,
+      "grad_norm": 0.34291339425641154,
+      "learning_rate": 8.571581950207121e-06,
+      "loss": 0.5071,
+      "step": 4521
+    },
+    {
+      "epoch": 0.7426354361258801,
+      "grad_norm": 0.2971810938064332,
+      "learning_rate": 8.571382218490602e-06,
+      "loss": 0.5295,
+      "step": 4522
+    },
+    {
+      "epoch": 0.7427996633342229,
+      "grad_norm": 0.27086010154061957,
+      "learning_rate": 8.571182442554865e-06,
+      "loss": 0.4966,
+      "step": 4523
+    },
+    {
+      "epoch": 0.7429638905425656,
+      "grad_norm": 0.3187861637022362,
+      "learning_rate": 8.570982622402082e-06,
+      "loss": 0.5025,
+      "step": 4524
+    },
+    {
+      "epoch": 0.7431281177509084,
+      "grad_norm": 0.304476818924289,
+      "learning_rate": 8.570782758034423e-06,
+      "loss": 0.5017,
+      "step": 4525
+    },
+    {
+      "epoch": 0.7432923449592511,
+      "grad_norm": 0.3673418689445969,
+      "learning_rate": 8.570582849454057e-06,
+      "loss": 0.521,
+      "step": 4526
+    },
+    {
+      "epoch": 0.7434565721675939,
+      "grad_norm": 0.31022619613586055,
+      "learning_rate": 8.570382896663158e-06,
+      "loss": 0.5205,
+      "step": 4527
+    },
+    {
+      "epoch": 0.7436207993759366,
+      "grad_norm": 0.29831412465170415,
+      "learning_rate": 8.570182899663896e-06,
+      "loss": 0.538,
+      "step": 4528
+    },
+    {
+      "epoch": 0.7437850265842794,
+      "grad_norm": 0.26560917646771326,
+      "learning_rate": 8.569982858458441e-06,
+      "loss": 0.4942,
+      "step": 4529
+    },
+    {
+      "epoch": 0.7439492537926221,
+      "grad_norm": 0.40343919917248544,
+      "learning_rate": 8.56978277304897e-06,
+      "loss": 0.5388,
+      "step": 4530
+    },
+    {
+      "epoch": 0.7441134810009649,
+      "grad_norm": 0.3538356979890078,
+      "learning_rate": 8.569582643437653e-06,
+      "loss": 0.5234,
+      "step": 4531
+    },
+    {
+      "epoch": 0.7442777082093076,
+      "grad_norm": 0.3502814342255026,
+      "learning_rate": 8.569382469626664e-06,
+      "loss": 0.5098,
+      "step": 4532
+    },
+    {
+      "epoch": 0.7444419354176504,
+      "grad_norm": 0.33687403724482856,
+      "learning_rate": 8.56918225161818e-06,
+      "loss": 0.5133,
+      "step": 4533
+    },
+    {
+      "epoch": 0.7446061626259931,
+      "grad_norm": 0.28226515273483066,
+      "learning_rate": 8.56898198941437e-06,
+      "loss": 0.51,
+      "step": 4534
+    },
+    {
+      "epoch": 0.7447703898343359,
+      "grad_norm": 0.30469490349467954,
+      "learning_rate": 8.568781683017414e-06,
+      "loss": 0.5227,
+      "step": 4535
+    },
+    {
+      "epoch": 0.7449346170426785,
+      "grad_norm": 0.35862422640377734,
+      "learning_rate": 8.568581332429486e-06,
+      "loss": 0.5168,
+      "step": 4536
+    },
+    {
+      "epoch": 0.7450988442510212,
+      "grad_norm": 0.2818516886467351,
+      "learning_rate": 8.568380937652761e-06,
+      "loss": 0.5075,
+      "step": 4537
+    },
+    {
+      "epoch": 0.745263071459364,
+      "grad_norm": 0.2918812274214278,
+      "learning_rate": 8.568180498689417e-06,
+      "loss": 0.5286,
+      "step": 4538
+    },
+    {
+      "epoch": 0.7454272986677067,
+      "grad_norm": 0.3570737420094653,
+      "learning_rate": 8.56798001554163e-06,
+      "loss": 0.5322,
+      "step": 4539
+    },
+    {
+      "epoch": 0.7455915258760495,
+      "grad_norm": 0.2910121417077147,
+      "learning_rate": 8.567779488211577e-06,
+      "loss": 0.5371,
+      "step": 4540
+    },
+    {
+      "epoch": 0.7457557530843922,
+      "grad_norm": 0.3164570259557555,
+      "learning_rate": 8.567578916701437e-06,
+      "loss": 0.5149,
+      "step": 4541
+    },
+    {
+      "epoch": 0.745919980292735,
+      "grad_norm": 0.31432087465115854,
+      "learning_rate": 8.567378301013388e-06,
+      "loss": 0.5174,
+      "step": 4542
+    },
+    {
+      "epoch": 0.7460842075010777,
+      "grad_norm": 0.3332922788821445,
+      "learning_rate": 8.56717764114961e-06,
+      "loss": 0.4993,
+      "step": 4543
+    },
+    {
+      "epoch": 0.7462484347094205,
+      "grad_norm": 0.2942101993605978,
+      "learning_rate": 8.56697693711228e-06,
+      "loss": 0.5338,
+      "step": 4544
+    },
+    {
+      "epoch": 0.7464126619177632,
+      "grad_norm": 0.3123029158168016,
+      "learning_rate": 8.566776188903579e-06,
+      "loss": 0.53,
+      "step": 4545
+    },
+    {
+      "epoch": 0.746576889126106,
+      "grad_norm": 0.33334606207740836,
+      "learning_rate": 8.566575396525688e-06,
+      "loss": 0.5401,
+      "step": 4546
+    },
+    {
+      "epoch": 0.7467411163344487,
+      "grad_norm": 0.3499456470338676,
+      "learning_rate": 8.566374559980787e-06,
+      "loss": 0.5053,
+      "step": 4547
+    },
+    {
+      "epoch": 0.7469053435427915,
+      "grad_norm": 0.5894023063619104,
+      "learning_rate": 8.566173679271057e-06,
+      "loss": 0.5044,
+      "step": 4548
+    },
+    {
+      "epoch": 0.7470695707511342,
+      "grad_norm": 0.3020765404041951,
+      "learning_rate": 8.565972754398682e-06,
+      "loss": 0.5099,
+      "step": 4549
+    },
+    {
+      "epoch": 0.747233797959477,
+      "grad_norm": 0.3171528202979881,
+      "learning_rate": 8.565771785365841e-06,
+      "loss": 0.5044,
+      "step": 4550
+    },
+    {
+      "epoch": 0.7473980251678197,
+      "grad_norm": 0.36820579002695975,
+      "learning_rate": 8.565570772174718e-06,
+      "loss": 0.5264,
+      "step": 4551
+    },
+    {
+      "epoch": 0.7475622523761625,
+      "grad_norm": 0.3652549638401894,
+      "learning_rate": 8.565369714827497e-06,
+      "loss": 0.517,
+      "step": 4552
+    },
+    {
+      "epoch": 0.7477264795845051,
+      "grad_norm": 0.28535511401718994,
+      "learning_rate": 8.565168613326362e-06,
+      "loss": 0.4884,
+      "step": 4553
+    },
+    {
+      "epoch": 0.7478907067928479,
+      "grad_norm": 0.33634032346122994,
+      "learning_rate": 8.564967467673494e-06,
+      "loss": 0.5246,
+      "step": 4554
+    },
+    {
+      "epoch": 0.7480549340011906,
+      "grad_norm": 0.28810968260779946,
+      "learning_rate": 8.564766277871081e-06,
+      "loss": 0.507,
+      "step": 4555
+    },
+    {
+      "epoch": 0.7482191612095334,
+      "grad_norm": 0.2992743113260256,
+      "learning_rate": 8.564565043921308e-06,
+      "loss": 0.5188,
+      "step": 4556
+    },
+    {
+      "epoch": 0.7483833884178761,
+      "grad_norm": 0.3558728320611078,
+      "learning_rate": 8.564363765826358e-06,
+      "loss": 0.5265,
+      "step": 4557
+    },
+    {
+      "epoch": 0.7485476156262189,
+      "grad_norm": 0.292448000228156,
+      "learning_rate": 8.564162443588421e-06,
+      "loss": 0.5252,
+      "step": 4558
+    },
+    {
+      "epoch": 0.7487118428345616,
+      "grad_norm": 0.3439831596581984,
+      "learning_rate": 8.56396107720968e-06,
+      "loss": 0.5187,
+      "step": 4559
+    },
+    {
+      "epoch": 0.7488760700429044,
+      "grad_norm": 0.40810472098558376,
+      "learning_rate": 8.563759666692323e-06,
+      "loss": 0.524,
+      "step": 4560
+    },
+    {
+      "epoch": 0.7490402972512471,
+      "grad_norm": 0.2845210850491724,
+      "learning_rate": 8.563558212038538e-06,
+      "loss": 0.5223,
+      "step": 4561
+    },
+    {
+      "epoch": 0.7492045244595898,
+      "grad_norm": 0.361454778137837,
+      "learning_rate": 8.563356713250513e-06,
+      "loss": 0.5032,
+      "step": 4562
+    },
+    {
+      "epoch": 0.7493687516679326,
+      "grad_norm": 0.3123393644294071,
+      "learning_rate": 8.563155170330436e-06,
+      "loss": 0.5003,
+      "step": 4563
+    },
+    {
+      "epoch": 0.7495329788762753,
+      "grad_norm": 0.2893188997077662,
+      "learning_rate": 8.562953583280497e-06,
+      "loss": 0.5123,
+      "step": 4564
+    },
+    {
+      "epoch": 0.7496972060846181,
+      "grad_norm": 0.325060447014177,
+      "learning_rate": 8.562751952102883e-06,
+      "loss": 0.5326,
+      "step": 4565
+    },
+    {
+      "epoch": 0.7498614332929608,
+      "grad_norm": 0.38334438203554844,
+      "learning_rate": 8.562550276799788e-06,
+      "loss": 0.5499,
+      "step": 4566
+    },
+    {
+      "epoch": 0.7500256605013036,
+      "grad_norm": 0.2895758675873873,
+      "learning_rate": 8.5623485573734e-06,
+      "loss": 0.5258,
+      "step": 4567
+    },
+    {
+      "epoch": 0.7501898877096463,
+      "grad_norm": 0.34262490892337305,
+      "learning_rate": 8.562146793825907e-06,
+      "loss": 0.5252,
+      "step": 4568
+    },
+    {
+      "epoch": 0.7503541149179891,
+      "grad_norm": 0.3077035016491017,
+      "learning_rate": 8.561944986159505e-06,
+      "loss": 0.5358,
+      "step": 4569
+    },
+    {
+      "epoch": 0.7505183421263317,
+      "grad_norm": 0.3032102839545749,
+      "learning_rate": 8.561743134376384e-06,
+      "loss": 0.5255,
+      "step": 4570
+    },
+    {
+      "epoch": 0.7506825693346745,
+      "grad_norm": 0.33693998727707863,
+      "learning_rate": 8.561541238478735e-06,
+      "loss": 0.5141,
+      "step": 4571
+    },
+    {
+      "epoch": 0.7508467965430172,
+      "grad_norm": 0.2780956966036318,
+      "learning_rate": 8.561339298468753e-06,
+      "loss": 0.5382,
+      "step": 4572
+    },
+    {
+      "epoch": 0.75101102375136,
+      "grad_norm": 0.33119142853483213,
+      "learning_rate": 8.56113731434863e-06,
+      "loss": 0.5032,
+      "step": 4573
+    },
+    {
+      "epoch": 0.7511752509597027,
+      "grad_norm": 0.2969192531504866,
+      "learning_rate": 8.560935286120562e-06,
+      "loss": 0.5012,
+      "step": 4574
+    },
+    {
+      "epoch": 0.7513394781680455,
+      "grad_norm": 0.30932229204461903,
+      "learning_rate": 8.560733213786741e-06,
+      "loss": 0.5228,
+      "step": 4575
+    },
+    {
+      "epoch": 0.7515037053763882,
+      "grad_norm": 0.6375810468219024,
+      "learning_rate": 8.56053109734936e-06,
+      "loss": 0.5347,
+      "step": 4576
+    },
+    {
+      "epoch": 0.751667932584731,
+      "grad_norm": 0.3602747428237164,
+      "learning_rate": 8.56032893681062e-06,
+      "loss": 0.5155,
+      "step": 4577
+    },
+    {
+      "epoch": 0.7518321597930737,
+      "grad_norm": 0.3344243202645859,
+      "learning_rate": 8.560126732172709e-06,
+      "loss": 0.5525,
+      "step": 4578
+    },
+    {
+      "epoch": 0.7519963870014165,
+      "grad_norm": 0.6821768214109283,
+      "learning_rate": 8.559924483437827e-06,
+      "loss": 0.5184,
+      "step": 4579
+    },
+    {
+      "epoch": 0.7521606142097592,
+      "grad_norm": 0.35181079249212044,
+      "learning_rate": 8.559722190608174e-06,
+      "loss": 0.5069,
+      "step": 4580
+    },
+    {
+      "epoch": 0.752324841418102,
+      "grad_norm": 0.32076296100754514,
+      "learning_rate": 8.55951985368594e-06,
+      "loss": 0.5245,
+      "step": 4581
+    },
+    {
+      "epoch": 0.7524890686264447,
+      "grad_norm": 0.34037129519099774,
+      "learning_rate": 8.55931747267333e-06,
+      "loss": 0.5204,
+      "step": 4582
+    },
+    {
+      "epoch": 0.7526532958347875,
+      "grad_norm": 0.35839103085610946,
+      "learning_rate": 8.559115047572537e-06,
+      "loss": 0.5028,
+      "step": 4583
+    },
+    {
+      "epoch": 0.7528175230431302,
+      "grad_norm": 0.32243397643030064,
+      "learning_rate": 8.55891257838576e-06,
+      "loss": 0.5202,
+      "step": 4584
+    },
+    {
+      "epoch": 0.752981750251473,
+      "grad_norm": 0.3207564345064311,
+      "learning_rate": 8.5587100651152e-06,
+      "loss": 0.51,
+      "step": 4585
+    },
+    {
+      "epoch": 0.7531459774598157,
+      "grad_norm": 0.28960495569847716,
+      "learning_rate": 8.558507507763055e-06,
+      "loss": 0.5065,
+      "step": 4586
+    },
+    {
+      "epoch": 0.7533102046681583,
+      "grad_norm": 0.32541735259990995,
+      "learning_rate": 8.558304906331525e-06,
+      "loss": 0.5082,
+      "step": 4587
+    },
+    {
+      "epoch": 0.7534744318765011,
+      "grad_norm": 0.4295942413931405,
+      "learning_rate": 8.558102260822812e-06,
+      "loss": 0.5032,
+      "step": 4588
+    },
+    {
+      "epoch": 0.7536386590848438,
+      "grad_norm": 0.3270938372241511,
+      "learning_rate": 8.557899571239115e-06,
+      "loss": 0.5189,
+      "step": 4589
+    },
+    {
+      "epoch": 0.7538028862931866,
+      "grad_norm": 0.4618734327758789,
+      "learning_rate": 8.557696837582636e-06,
+      "loss": 0.5126,
+      "step": 4590
+    },
+    {
+      "epoch": 0.7539671135015293,
+      "grad_norm": 0.2856325652395653,
+      "learning_rate": 8.557494059855579e-06,
+      "loss": 0.495,
+      "step": 4591
+    },
+    {
+      "epoch": 0.7541313407098721,
+      "grad_norm": 0.2922332368335058,
+      "learning_rate": 8.557291238060142e-06,
+      "loss": 0.5173,
+      "step": 4592
+    },
+    {
+      "epoch": 0.7542955679182148,
+      "grad_norm": 0.26960221791122096,
+      "learning_rate": 8.557088372198532e-06,
+      "loss": 0.5192,
+      "step": 4593
+    },
+    {
+      "epoch": 0.7544597951265576,
+      "grad_norm": 0.29754034652447925,
+      "learning_rate": 8.55688546227295e-06,
+      "loss": 0.5044,
+      "step": 4594
+    },
+    {
+      "epoch": 0.7546240223349003,
+      "grad_norm": 0.302344635843857,
+      "learning_rate": 8.556682508285601e-06,
+      "loss": 0.5439,
+      "step": 4595
+    },
+    {
+      "epoch": 0.7547882495432431,
+      "grad_norm": 0.3497725787114854,
+      "learning_rate": 8.556479510238688e-06,
+      "loss": 0.5342,
+      "step": 4596
+    },
+    {
+      "epoch": 0.7549524767515858,
+      "grad_norm": 0.29555475207440246,
+      "learning_rate": 8.556276468134418e-06,
+      "loss": 0.5399,
+      "step": 4597
+    },
+    {
+      "epoch": 0.7551167039599286,
+      "grad_norm": 0.2604097571549104,
+      "learning_rate": 8.556073381974992e-06,
+      "loss": 0.509,
+      "step": 4598
+    },
+    {
+      "epoch": 0.7552809311682713,
+      "grad_norm": 0.33188600308422034,
+      "learning_rate": 8.555870251762619e-06,
+      "loss": 0.5198,
+      "step": 4599
+    },
+    {
+      "epoch": 0.7554451583766141,
+      "grad_norm": 0.3258458849112466,
+      "learning_rate": 8.555667077499506e-06,
+      "loss": 0.5124,
+      "step": 4600
+    },
+    {
+      "epoch": 0.7556093855849568,
+      "grad_norm": 0.34142640335866914,
+      "learning_rate": 8.555463859187858e-06,
+      "loss": 0.5383,
+      "step": 4601
+    },
+    {
+      "epoch": 0.7557736127932996,
+      "grad_norm": 0.32818633170502004,
+      "learning_rate": 8.555260596829882e-06,
+      "loss": 0.533,
+      "step": 4602
+    },
+    {
+      "epoch": 0.7559378400016423,
+      "grad_norm": 0.3873086927551184,
+      "learning_rate": 8.555057290427787e-06,
+      "loss": 0.5217,
+      "step": 4603
+    },
+    {
+      "epoch": 0.756102067209985,
+      "grad_norm": 0.33476337551175733,
+      "learning_rate": 8.55485393998378e-06,
+      "loss": 0.5138,
+      "step": 4604
+    },
+    {
+      "epoch": 0.7562662944183277,
+      "grad_norm": 0.3318686735908207,
+      "learning_rate": 8.554650545500068e-06,
+      "loss": 0.5022,
+      "step": 4605
+    },
+    {
+      "epoch": 0.7564305216266705,
+      "grad_norm": 0.326403174472294,
+      "learning_rate": 8.554447106978865e-06,
+      "loss": 0.5214,
+      "step": 4606
+    },
+    {
+      "epoch": 0.7565947488350132,
+      "grad_norm": 0.39247588112022863,
+      "learning_rate": 8.554243624422373e-06,
+      "loss": 0.5156,
+      "step": 4607
+    },
+    {
+      "epoch": 0.756758976043356,
+      "grad_norm": 0.326663427806945,
+      "learning_rate": 8.55404009783281e-06,
+      "loss": 0.5398,
+      "step": 4608
+    },
+    {
+      "epoch": 0.7569232032516987,
+      "grad_norm": 0.3459223648779705,
+      "learning_rate": 8.553836527212381e-06,
+      "loss": 0.4895,
+      "step": 4609
+    },
+    {
+      "epoch": 0.7570874304600415,
+      "grad_norm": 0.2941823147267791,
+      "learning_rate": 8.5536329125633e-06,
+      "loss": 0.518,
+      "step": 4610
+    },
+    {
+      "epoch": 0.7572516576683842,
+      "grad_norm": 0.30618428789154545,
+      "learning_rate": 8.553429253887778e-06,
+      "loss": 0.5023,
+      "step": 4611
+    },
+    {
+      "epoch": 0.757415884876727,
+      "grad_norm": 0.31806576498609396,
+      "learning_rate": 8.553225551188025e-06,
+      "loss": 0.5371,
+      "step": 4612
+    },
+    {
+      "epoch": 0.7575801120850697,
+      "grad_norm": 0.35278860041835625,
+      "learning_rate": 8.553021804466254e-06,
+      "loss": 0.5085,
+      "step": 4613
+    },
+    {
+      "epoch": 0.7577443392934124,
+      "grad_norm": 0.3214722436543048,
+      "learning_rate": 8.55281801372468e-06,
+      "loss": 0.5145,
+      "step": 4614
+    },
+    {
+      "epoch": 0.7579085665017552,
+      "grad_norm": 0.32654098045991403,
+      "learning_rate": 8.552614178965514e-06,
+      "loss": 0.5254,
+      "step": 4615
+    },
+    {
+      "epoch": 0.758072793710098,
+      "grad_norm": 0.3281384004825425,
+      "learning_rate": 8.552410300190972e-06,
+      "loss": 0.5087,
+      "step": 4616
+    },
+    {
+      "epoch": 0.7582370209184407,
+      "grad_norm": 0.3279057768090118,
+      "learning_rate": 8.552206377403265e-06,
+      "loss": 0.5073,
+      "step": 4617
+    },
+    {
+      "epoch": 0.7584012481267834,
+      "grad_norm": 0.33926049968773087,
+      "learning_rate": 8.552002410604613e-06,
+      "loss": 0.525,
+      "step": 4618
+    },
+    {
+      "epoch": 0.7585654753351262,
+      "grad_norm": 0.2795274039873359,
+      "learning_rate": 8.551798399797226e-06,
+      "loss": 0.5183,
+      "step": 4619
+    },
+    {
+      "epoch": 0.7587297025434689,
+      "grad_norm": 0.3206834642081961,
+      "learning_rate": 8.551594344983322e-06,
+      "loss": 0.5422,
+      "step": 4620
+    },
+    {
+      "epoch": 0.7588939297518116,
+      "grad_norm": 0.3058743058502909,
+      "learning_rate": 8.551390246165118e-06,
+      "loss": 0.4945,
+      "step": 4621
+    },
+    {
+      "epoch": 0.7590581569601543,
+      "grad_norm": 0.299835948991702,
+      "learning_rate": 8.551186103344828e-06,
+      "loss": 0.5012,
+      "step": 4622
+    },
+    {
+      "epoch": 0.7592223841684971,
+      "grad_norm": 0.2883728461560533,
+      "learning_rate": 8.550981916524673e-06,
+      "loss": 0.5257,
+      "step": 4623
+    },
+    {
+      "epoch": 0.7593866113768398,
+      "grad_norm": 0.2793782805803841,
+      "learning_rate": 8.550777685706869e-06,
+      "loss": 0.5489,
+      "step": 4624
+    },
+    {
+      "epoch": 0.7595508385851826,
+      "grad_norm": 0.2638335021783196,
+      "learning_rate": 8.550573410893633e-06,
+      "loss": 0.5093,
+      "step": 4625
+    },
+    {
+      "epoch": 0.7597150657935253,
+      "grad_norm": 0.3190328077240095,
+      "learning_rate": 8.550369092087185e-06,
+      "loss": 0.525,
+      "step": 4626
+    },
+    {
+      "epoch": 0.7598792930018681,
+      "grad_norm": 0.3627572414267673,
+      "learning_rate": 8.550164729289743e-06,
+      "loss": 0.5023,
+      "step": 4627
+    },
+    {
+      "epoch": 0.7600435202102108,
+      "grad_norm": 0.3706630695321476,
+      "learning_rate": 8.549960322503529e-06,
+      "loss": 0.5316,
+      "step": 4628
+    },
+    {
+      "epoch": 0.7602077474185536,
+      "grad_norm": 0.33057504555303346,
+      "learning_rate": 8.54975587173076e-06,
+      "loss": 0.5226,
+      "step": 4629
+    },
+    {
+      "epoch": 0.7603719746268963,
+      "grad_norm": 0.30309061540235327,
+      "learning_rate": 8.549551376973658e-06,
+      "loss": 0.5269,
+      "step": 4630
+    },
+    {
+      "epoch": 0.7605362018352391,
+      "grad_norm": 0.31184472191583695,
+      "learning_rate": 8.549346838234442e-06,
+      "loss": 0.5076,
+      "step": 4631
+    },
+    {
+      "epoch": 0.7607004290435818,
+      "grad_norm": 0.29095482111017995,
+      "learning_rate": 8.549142255515338e-06,
+      "loss": 0.4765,
+      "step": 4632
+    },
+    {
+      "epoch": 0.7608646562519246,
+      "grad_norm": 0.2922099969230343,
+      "learning_rate": 8.548937628818564e-06,
+      "loss": 0.5138,
+      "step": 4633
+    },
+    {
+      "epoch": 0.7610288834602673,
+      "grad_norm": 0.3400598354912872,
+      "learning_rate": 8.548732958146344e-06,
+      "loss": 0.5167,
+      "step": 4634
+    },
+    {
+      "epoch": 0.7611931106686101,
+      "grad_norm": 0.3381885179424978,
+      "learning_rate": 8.5485282435009e-06,
+      "loss": 0.5004,
+      "step": 4635
+    },
+    {
+      "epoch": 0.7613573378769528,
+      "grad_norm": 0.315306221932674,
+      "learning_rate": 8.548323484884457e-06,
+      "loss": 0.5548,
+      "step": 4636
+    },
+    {
+      "epoch": 0.7615215650852956,
+      "grad_norm": 0.3346000660174875,
+      "learning_rate": 8.548118682299237e-06,
+      "loss": 0.5271,
+      "step": 4637
+    },
+    {
+      "epoch": 0.7616857922936382,
+      "grad_norm": 0.4439712033523865,
+      "learning_rate": 8.547913835747465e-06,
+      "loss": 0.5002,
+      "step": 4638
+    },
+    {
+      "epoch": 0.761850019501981,
+      "grad_norm": 0.31153812213147325,
+      "learning_rate": 8.547708945231369e-06,
+      "loss": 0.52,
+      "step": 4639
+    },
+    {
+      "epoch": 0.7620142467103237,
+      "grad_norm": 0.340893810594224,
+      "learning_rate": 8.54750401075317e-06,
+      "loss": 0.518,
+      "step": 4640
+    },
+    {
+      "epoch": 0.7621784739186664,
+      "grad_norm": 0.2955501959425951,
+      "learning_rate": 8.547299032315092e-06,
+      "loss": 0.5409,
+      "step": 4641
+    },
+    {
+      "epoch": 0.7623427011270092,
+      "grad_norm": 0.28838767397812765,
+      "learning_rate": 8.547094009919367e-06,
+      "loss": 0.4977,
+      "step": 4642
+    },
+    {
+      "epoch": 0.7625069283353519,
+      "grad_norm": 0.36043013999166806,
+      "learning_rate": 8.546888943568222e-06,
+      "loss": 0.5148,
+      "step": 4643
+    },
+    {
+      "epoch": 0.7626711555436947,
+      "grad_norm": 0.2841530706289285,
+      "learning_rate": 8.546683833263877e-06,
+      "loss": 0.5261,
+      "step": 4644
+    },
+    {
+      "epoch": 0.7628353827520374,
+      "grad_norm": 0.3517544924162603,
+      "learning_rate": 8.546478679008567e-06,
+      "loss": 0.5239,
+      "step": 4645
+    },
+    {
+      "epoch": 0.7629996099603802,
+      "grad_norm": 0.3226642978400009,
+      "learning_rate": 8.546273480804516e-06,
+      "loss": 0.5214,
+      "step": 4646
+    },
+    {
+      "epoch": 0.7631638371687229,
+      "grad_norm": 0.3350012172177672,
+      "learning_rate": 8.546068238653956e-06,
+      "loss": 0.5265,
+      "step": 4647
+    },
+    {
+      "epoch": 0.7633280643770657,
+      "grad_norm": 0.3365204860158755,
+      "learning_rate": 8.54586295255911e-06,
+      "loss": 0.5224,
+      "step": 4648
+    },
+    {
+      "epoch": 0.7634922915854084,
+      "grad_norm": 0.30134315958920094,
+      "learning_rate": 8.545657622522215e-06,
+      "loss": 0.543,
+      "step": 4649
+    },
+    {
+      "epoch": 0.7636565187937512,
+      "grad_norm": 0.33019521783542843,
+      "learning_rate": 8.545452248545498e-06,
+      "loss": 0.5183,
+      "step": 4650
+    },
+    {
+      "epoch": 0.7638207460020939,
+      "grad_norm": 0.37232931232505206,
+      "learning_rate": 8.545246830631188e-06,
+      "loss": 0.5175,
+      "step": 4651
+    },
+    {
+      "epoch": 0.7639849732104367,
+      "grad_norm": 0.32383868334851396,
+      "learning_rate": 8.545041368781517e-06,
+      "loss": 0.5364,
+      "step": 4652
+    },
+    {
+      "epoch": 0.7641492004187794,
+      "grad_norm": 0.44489749436817166,
+      "learning_rate": 8.544835862998718e-06,
+      "loss": 0.5201,
+      "step": 4653
+    },
+    {
+      "epoch": 0.7643134276271222,
+      "grad_norm": 0.3077556378016318,
+      "learning_rate": 8.544630313285022e-06,
+      "loss": 0.5124,
+      "step": 4654
+    },
+    {
+      "epoch": 0.7644776548354648,
+      "grad_norm": 0.4603259411059813,
+      "learning_rate": 8.544424719642661e-06,
+      "loss": 0.5183,
+      "step": 4655
+    },
+    {
+      "epoch": 0.7646418820438076,
+      "grad_norm": 0.2832805233868871,
+      "learning_rate": 8.54421908207387e-06,
+      "loss": 0.53,
+      "step": 4656
+    },
+    {
+      "epoch": 0.7648061092521503,
+      "grad_norm": 0.29170501882490907,
+      "learning_rate": 8.54401340058088e-06,
+      "loss": 0.5308,
+      "step": 4657
+    },
+    {
+      "epoch": 0.7649703364604931,
+      "grad_norm": 0.2950625968091386,
+      "learning_rate": 8.543807675165924e-06,
+      "loss": 0.5321,
+      "step": 4658
+    },
+    {
+      "epoch": 0.7651345636688358,
+      "grad_norm": 0.3632685280310422,
+      "learning_rate": 8.543601905831239e-06,
+      "loss": 0.5116,
+      "step": 4659
+    },
+    {
+      "epoch": 0.7652987908771786,
+      "grad_norm": 0.3393323236915578,
+      "learning_rate": 8.54339609257906e-06,
+      "loss": 0.5123,
+      "step": 4660
+    },
+    {
+      "epoch": 0.7654630180855213,
+      "grad_norm": 0.4729462883437287,
+      "learning_rate": 8.543190235411619e-06,
+      "loss": 0.519,
+      "step": 4661
+    },
+    {
+      "epoch": 0.7656272452938641,
+      "grad_norm": 0.3293233461617647,
+      "learning_rate": 8.542984334331155e-06,
+      "loss": 0.5236,
+      "step": 4662
+    },
+    {
+      "epoch": 0.7657914725022068,
+      "grad_norm": 0.36363391812587026,
+      "learning_rate": 8.542778389339906e-06,
+      "loss": 0.5094,
+      "step": 4663
+    },
+    {
+      "epoch": 0.7659556997105496,
+      "grad_norm": 0.30508096976913734,
+      "learning_rate": 8.542572400440103e-06,
+      "loss": 0.5097,
+      "step": 4664
+    },
+    {
+      "epoch": 0.7661199269188923,
+      "grad_norm": 0.31527353025695465,
+      "learning_rate": 8.542366367633988e-06,
+      "loss": 0.5233,
+      "step": 4665
+    },
+    {
+      "epoch": 0.766284154127235,
+      "grad_norm": 0.3002322251825842,
+      "learning_rate": 8.542160290923796e-06,
+      "loss": 0.4995,
+      "step": 4666
+    },
+    {
+      "epoch": 0.7664483813355778,
+      "grad_norm": 0.35868121348697857,
+      "learning_rate": 8.541954170311768e-06,
+      "loss": 0.4835,
+      "step": 4667
+    },
+    {
+      "epoch": 0.7666126085439205,
+      "grad_norm": 0.2816569076089125,
+      "learning_rate": 8.541748005800139e-06,
+      "loss": 0.5175,
+      "step": 4668
+    },
+    {
+      "epoch": 0.7667768357522633,
+      "grad_norm": 0.2794544666938337,
+      "learning_rate": 8.54154179739115e-06,
+      "loss": 0.5097,
+      "step": 4669
+    },
+    {
+      "epoch": 0.766941062960606,
+      "grad_norm": 0.29808345975734646,
+      "learning_rate": 8.541335545087043e-06,
+      "loss": 0.4996,
+      "step": 4670
+    },
+    {
+      "epoch": 0.7671052901689488,
+      "grad_norm": 0.28395721446353817,
+      "learning_rate": 8.541129248890053e-06,
+      "loss": 0.5332,
+      "step": 4671
+    },
+    {
+      "epoch": 0.7672695173772914,
+      "grad_norm": 0.3514233085142723,
+      "learning_rate": 8.540922908802425e-06,
+      "loss": 0.4993,
+      "step": 4672
+    },
+    {
+      "epoch": 0.7674337445856342,
+      "grad_norm": 0.41429693067925427,
+      "learning_rate": 8.540716524826398e-06,
+      "loss": 0.5151,
+      "step": 4673
+    },
+    {
+      "epoch": 0.7675979717939769,
+      "grad_norm": 0.3520524597657039,
+      "learning_rate": 8.540510096964215e-06,
+      "loss": 0.5223,
+      "step": 4674
+    },
+    {
+      "epoch": 0.7677621990023197,
+      "grad_norm": 0.3328068751348965,
+      "learning_rate": 8.540303625218115e-06,
+      "loss": 0.5111,
+      "step": 4675
+    },
+    {
+      "epoch": 0.7679264262106624,
+      "grad_norm": 0.31064697105471334,
+      "learning_rate": 8.540097109590346e-06,
+      "loss": 0.5042,
+      "step": 4676
+    },
+    {
+      "epoch": 0.7680906534190052,
+      "grad_norm": 0.39821289781714075,
+      "learning_rate": 8.539890550083144e-06,
+      "loss": 0.5076,
+      "step": 4677
+    },
+    {
+      "epoch": 0.7682548806273479,
+      "grad_norm": 0.3141655423218781,
+      "learning_rate": 8.539683946698758e-06,
+      "loss": 0.5134,
+      "step": 4678
+    },
+    {
+      "epoch": 0.7684191078356907,
+      "grad_norm": 0.46016224300055375,
+      "learning_rate": 8.539477299439429e-06,
+      "loss": 0.5196,
+      "step": 4679
+    },
+    {
+      "epoch": 0.7685833350440334,
+      "grad_norm": 0.4735631016426876,
+      "learning_rate": 8.539270608307402e-06,
+      "loss": 0.5306,
+      "step": 4680
+    },
+    {
+      "epoch": 0.7687475622523762,
+      "grad_norm": 0.2952547172783303,
+      "learning_rate": 8.539063873304922e-06,
+      "loss": 0.4949,
+      "step": 4681
+    },
+    {
+      "epoch": 0.7689117894607189,
+      "grad_norm": 0.3079803366854746,
+      "learning_rate": 8.538857094434234e-06,
+      "loss": 0.5281,
+      "step": 4682
+    },
+    {
+      "epoch": 0.7690760166690617,
+      "grad_norm": 0.3159192970003273,
+      "learning_rate": 8.538650271697586e-06,
+      "loss": 0.49,
+      "step": 4683
+    },
+    {
+      "epoch": 0.7692402438774044,
+      "grad_norm": 0.31282817662971113,
+      "learning_rate": 8.538443405097223e-06,
+      "loss": 0.5307,
+      "step": 4684
+    },
+    {
+      "epoch": 0.7694044710857472,
+      "grad_norm": 0.32242372926036034,
+      "learning_rate": 8.538236494635389e-06,
+      "loss": 0.5396,
+      "step": 4685
+    },
+    {
+      "epoch": 0.7695686982940899,
+      "grad_norm": 0.48535860275289644,
+      "learning_rate": 8.538029540314334e-06,
+      "loss": 0.5167,
+      "step": 4686
+    },
+    {
+      "epoch": 0.7697329255024327,
+      "grad_norm": 0.30109361527460454,
+      "learning_rate": 8.537822542136306e-06,
+      "loss": 0.5317,
+      "step": 4687
+    },
+    {
+      "epoch": 0.7698971527107754,
+      "grad_norm": 0.4440781157107436,
+      "learning_rate": 8.537615500103553e-06,
+      "loss": 0.5044,
+      "step": 4688
+    },
+    {
+      "epoch": 0.770061379919118,
+      "grad_norm": 0.3322199954679528,
+      "learning_rate": 8.537408414218323e-06,
+      "loss": 0.5235,
+      "step": 4689
+    },
+    {
+      "epoch": 0.7702256071274608,
+      "grad_norm": 0.2736769475032565,
+      "learning_rate": 8.537201284482864e-06,
+      "loss": 0.5071,
+      "step": 4690
+    },
+    {
+      "epoch": 0.7703898343358035,
+      "grad_norm": 0.3251526213165742,
+      "learning_rate": 8.536994110899428e-06,
+      "loss": 0.4937,
+      "step": 4691
+    },
+    {
+      "epoch": 0.7705540615441463,
+      "grad_norm": 0.3412682423457841,
+      "learning_rate": 8.536786893470264e-06,
+      "loss": 0.5148,
+      "step": 4692
+    },
+    {
+      "epoch": 0.770718288752489,
+      "grad_norm": 0.3158500123302273,
+      "learning_rate": 8.536579632197622e-06,
+      "loss": 0.4904,
+      "step": 4693
+    },
+    {
+      "epoch": 0.7708825159608318,
+      "grad_norm": 0.38296860314735637,
+      "learning_rate": 8.536372327083755e-06,
+      "loss": 0.5299,
+      "step": 4694
+    },
+    {
+      "epoch": 0.7710467431691745,
+      "grad_norm": 0.31222738840079106,
+      "learning_rate": 8.536164978130913e-06,
+      "loss": 0.535,
+      "step": 4695
+    },
+    {
+      "epoch": 0.7712109703775173,
+      "grad_norm": 0.843025114686952,
+      "learning_rate": 8.535957585341349e-06,
+      "loss": 0.5245,
+      "step": 4696
+    },
+    {
+      "epoch": 0.77137519758586,
+      "grad_norm": 0.32381269833214565,
+      "learning_rate": 8.535750148717312e-06,
+      "loss": 0.5388,
+      "step": 4697
+    },
+    {
+      "epoch": 0.7715394247942028,
+      "grad_norm": 0.41423456394537234,
+      "learning_rate": 8.53554266826106e-06,
+      "loss": 0.5049,
+      "step": 4698
+    },
+    {
+      "epoch": 0.7717036520025455,
+      "grad_norm": 0.2918748934972015,
+      "learning_rate": 8.535335143974844e-06,
+      "loss": 0.5151,
+      "step": 4699
+    },
+    {
+      "epoch": 0.7718678792108883,
+      "grad_norm": 0.2785197474123093,
+      "learning_rate": 8.535127575860917e-06,
+      "loss": 0.523,
+      "step": 4700
+    },
+    {
+      "epoch": 0.772032106419231,
+      "grad_norm": 0.4162412664492259,
+      "learning_rate": 8.534919963921536e-06,
+      "loss": 0.5231,
+      "step": 4701
+    },
+    {
+      "epoch": 0.7721963336275738,
+      "grad_norm": 0.2769221462291233,
+      "learning_rate": 8.534712308158954e-06,
+      "loss": 0.5286,
+      "step": 4702
+    },
+    {
+      "epoch": 0.7723605608359165,
+      "grad_norm": 0.30009896167616434,
+      "learning_rate": 8.534504608575426e-06,
+      "loss": 0.5155,
+      "step": 4703
+    },
+    {
+      "epoch": 0.7725247880442593,
+      "grad_norm": 0.28533441325104786,
+      "learning_rate": 8.53429686517321e-06,
+      "loss": 0.4966,
+      "step": 4704
+    },
+    {
+      "epoch": 0.772689015252602,
+      "grad_norm": 1.2238805912739166,
+      "learning_rate": 8.534089077954558e-06,
+      "loss": 0.5149,
+      "step": 4705
+    },
+    {
+      "epoch": 0.7728532424609447,
+      "grad_norm": 0.29952221829630815,
+      "learning_rate": 8.533881246921732e-06,
+      "loss": 0.5181,
+      "step": 4706
+    },
+    {
+      "epoch": 0.7730174696692874,
+      "grad_norm": 0.320317196375448,
+      "learning_rate": 8.533673372076987e-06,
+      "loss": 0.492,
+      "step": 4707
+    },
+    {
+      "epoch": 0.7731816968776302,
+      "grad_norm": 0.3663326715952758,
+      "learning_rate": 8.53346545342258e-06,
+      "loss": 0.5393,
+      "step": 4708
+    },
+    {
+      "epoch": 0.7733459240859729,
+      "grad_norm": 0.3036228367230328,
+      "learning_rate": 8.533257490960768e-06,
+      "loss": 0.5426,
+      "step": 4709
+    },
+    {
+      "epoch": 0.7735101512943157,
+      "grad_norm": 0.2865700530075475,
+      "learning_rate": 8.533049484693813e-06,
+      "loss": 0.5117,
+      "step": 4710
+    },
+    {
+      "epoch": 0.7736743785026584,
+      "grad_norm": 0.31601286020177577,
+      "learning_rate": 8.532841434623974e-06,
+      "loss": 0.533,
+      "step": 4711
+    },
+    {
+      "epoch": 0.7738386057110012,
+      "grad_norm": 0.3364380677143077,
+      "learning_rate": 8.532633340753507e-06,
+      "loss": 0.53,
+      "step": 4712
+    },
+    {
+      "epoch": 0.7740028329193439,
+      "grad_norm": 0.35339859314853217,
+      "learning_rate": 8.532425203084675e-06,
+      "loss": 0.5104,
+      "step": 4713
+    },
+    {
+      "epoch": 0.7741670601276867,
+      "grad_norm": 0.5951590155877503,
+      "learning_rate": 8.532217021619738e-06,
+      "loss": 0.5445,
+      "step": 4714
+    },
+    {
+      "epoch": 0.7743312873360294,
+      "grad_norm": 0.29684524677217816,
+      "learning_rate": 8.532008796360957e-06,
+      "loss": 0.515,
+      "step": 4715
+    },
+    {
+      "epoch": 0.7744955145443722,
+      "grad_norm": 0.3610512910167456,
+      "learning_rate": 8.531800527310594e-06,
+      "loss": 0.4889,
+      "step": 4716
+    },
+    {
+      "epoch": 0.7746597417527149,
+      "grad_norm": 0.2866312562125442,
+      "learning_rate": 8.53159221447091e-06,
+      "loss": 0.5028,
+      "step": 4717
+    },
+    {
+      "epoch": 0.7748239689610577,
+      "grad_norm": 0.2883691284842144,
+      "learning_rate": 8.531383857844169e-06,
+      "loss": 0.5257,
+      "step": 4718
+    },
+    {
+      "epoch": 0.7749881961694004,
+      "grad_norm": 0.32116600915112437,
+      "learning_rate": 8.53117545743263e-06,
+      "loss": 0.511,
+      "step": 4719
+    },
+    {
+      "epoch": 0.7751524233777431,
+      "grad_norm": 0.36579551244838965,
+      "learning_rate": 8.530967013238562e-06,
+      "loss": 0.5279,
+      "step": 4720
+    },
+    {
+      "epoch": 0.7753166505860859,
+      "grad_norm": 0.335274958914863,
+      "learning_rate": 8.530758525264226e-06,
+      "loss": 0.5219,
+      "step": 4721
+    },
+    {
+      "epoch": 0.7754808777944286,
+      "grad_norm": 0.35908456979911946,
+      "learning_rate": 8.530549993511886e-06,
+      "loss": 0.5122,
+      "step": 4722
+    },
+    {
+      "epoch": 0.7756451050027713,
+      "grad_norm": 0.3303814149214648,
+      "learning_rate": 8.53034141798381e-06,
+      "loss": 0.5188,
+      "step": 4723
+    },
+    {
+      "epoch": 0.775809332211114,
+      "grad_norm": 0.552579852685583,
+      "learning_rate": 8.530132798682258e-06,
+      "loss": 0.5499,
+      "step": 4724
+    },
+    {
+      "epoch": 0.7759735594194568,
+      "grad_norm": 0.6038296709572685,
+      "learning_rate": 8.529924135609499e-06,
+      "loss": 0.5315,
+      "step": 4725
+    },
+    {
+      "epoch": 0.7761377866277995,
+      "grad_norm": 0.3259739996544789,
+      "learning_rate": 8.5297154287678e-06,
+      "loss": 0.4987,
+      "step": 4726
+    },
+    {
+      "epoch": 0.7763020138361423,
+      "grad_norm": 0.2899897486092834,
+      "learning_rate": 8.529506678159426e-06,
+      "loss": 0.5278,
+      "step": 4727
+    },
+    {
+      "epoch": 0.776466241044485,
+      "grad_norm": 0.2904572716926567,
+      "learning_rate": 8.529297883786645e-06,
+      "loss": 0.501,
+      "step": 4728
+    },
+    {
+      "epoch": 0.7766304682528278,
+      "grad_norm": 0.31236889089317593,
+      "learning_rate": 8.529089045651726e-06,
+      "loss": 0.4995,
+      "step": 4729
+    },
+    {
+      "epoch": 0.7767946954611705,
+      "grad_norm": 0.38956907875588986,
+      "learning_rate": 8.528880163756935e-06,
+      "loss": 0.5115,
+      "step": 4730
+    },
+    {
+      "epoch": 0.7769589226695133,
+      "grad_norm": 0.34404235566924035,
+      "learning_rate": 8.52867123810454e-06,
+      "loss": 0.5264,
+      "step": 4731
+    },
+    {
+      "epoch": 0.777123149877856,
+      "grad_norm": 0.3443712963310854,
+      "learning_rate": 8.528462268696812e-06,
+      "loss": 0.512,
+      "step": 4732
+    },
+    {
+      "epoch": 0.7772873770861988,
+      "grad_norm": 0.28851683888501506,
+      "learning_rate": 8.528253255536022e-06,
+      "loss": 0.504,
+      "step": 4733
+    },
+    {
+      "epoch": 0.7774516042945415,
+      "grad_norm": 0.3181508096343241,
+      "learning_rate": 8.528044198624438e-06,
+      "loss": 0.5244,
+      "step": 4734
+    },
+    {
+      "epoch": 0.7776158315028843,
+      "grad_norm": 0.3867330975953575,
+      "learning_rate": 8.527835097964331e-06,
+      "loss": 0.5271,
+      "step": 4735
+    },
+    {
+      "epoch": 0.777780058711227,
+      "grad_norm": 0.3066079602945166,
+      "learning_rate": 8.527625953557972e-06,
+      "loss": 0.4989,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7779442859195698,
+      "grad_norm": 0.3851892733779904,
+      "learning_rate": 8.527416765407633e-06,
+      "loss": 0.5198,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7781085131279125,
+      "grad_norm": 0.31990920459193956,
+      "learning_rate": 8.527207533515583e-06,
+      "loss": 0.5015,
+      "step": 4738
+    },
+    {
+      "epoch": 0.7782727403362553,
+      "grad_norm": 0.2720719422478482,
+      "learning_rate": 8.5269982578841e-06,
+      "loss": 0.4928,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7784369675445979,
+      "grad_norm": 0.7093090861871565,
+      "learning_rate": 8.526788938515451e-06,
+      "loss": 0.511,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7786011947529407,
+      "grad_norm": 0.28949201274533776,
+      "learning_rate": 8.526579575411914e-06,
+      "loss": 0.5134,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7787654219612834,
+      "grad_norm": 0.41192958031795,
+      "learning_rate": 8.526370168575762e-06,
+      "loss": 0.5001,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7789296491696261,
+      "grad_norm": 0.3955354645563975,
+      "learning_rate": 8.526160718009267e-06,
+      "loss": 0.4977,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7790938763779689,
+      "grad_norm": 0.27548066782513797,
+      "learning_rate": 8.525951223714705e-06,
+      "loss": 0.5406,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7792581035863116,
+      "grad_norm": 0.34792018231763794,
+      "learning_rate": 8.525741685694353e-06,
+      "loss": 0.5035,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7794223307946544,
+      "grad_norm": 0.2558824257167838,
+      "learning_rate": 8.525532103950485e-06,
+      "loss": 0.5181,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7795865580029971,
+      "grad_norm": 0.37147364997022847,
+      "learning_rate": 8.52532247848538e-06,
+      "loss": 0.5063,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7797507852113399,
+      "grad_norm": 0.26862453917752677,
+      "learning_rate": 8.525112809301308e-06,
+      "loss": 0.5384,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7799150124196826,
+      "grad_norm": 0.31769694408212135,
+      "learning_rate": 8.524903096400554e-06,
+      "loss": 0.5187,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7800792396280254,
+      "grad_norm": 0.298163868790542,
+      "learning_rate": 8.524693339785392e-06,
+      "loss": 0.507,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7802434668363681,
+      "grad_norm": 0.2933975935484392,
+      "learning_rate": 8.524483539458099e-06,
+      "loss": 0.5265,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7804076940447109,
+      "grad_norm": 0.3179903130943498,
+      "learning_rate": 8.524273695420957e-06,
+      "loss": 0.5126,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7805719212530536,
+      "grad_norm": 0.26716291057570807,
+      "learning_rate": 8.524063807676241e-06,
+      "loss": 0.5204,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7807361484613964,
+      "grad_norm": 0.31106935340054614,
+      "learning_rate": 8.523853876226236e-06,
+      "loss": 0.5321,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7809003756697391,
+      "grad_norm": 0.31817775903788903,
+      "learning_rate": 8.523643901073217e-06,
+      "loss": 0.5051,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7810646028780819,
+      "grad_norm": 0.3039687088510633,
+      "learning_rate": 8.523433882219467e-06,
+      "loss": 0.4994,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7812288300864245,
+      "grad_norm": 0.2839010503466107,
+      "learning_rate": 8.523223819667267e-06,
+      "loss": 0.5216,
+      "step": 4757
+    },
+    {
+      "epoch": 0.7813930572947673,
+      "grad_norm": 0.4120389386069269,
+      "learning_rate": 8.523013713418897e-06,
+      "loss": 0.481,
+      "step": 4758
+    },
+    {
+      "epoch": 0.78155728450311,
+      "grad_norm": 0.3062612421267718,
+      "learning_rate": 8.522803563476641e-06,
+      "loss": 0.4983,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7817215117114528,
+      "grad_norm": 0.28296796391139384,
+      "learning_rate": 8.52259336984278e-06,
+      "loss": 0.5013,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7818857389197955,
+      "grad_norm": 0.3046360962097441,
+      "learning_rate": 8.522383132519597e-06,
+      "loss": 0.53,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7820499661281383,
+      "grad_norm": 0.3142566255762791,
+      "learning_rate": 8.522172851509375e-06,
+      "loss": 0.5181,
+      "step": 4762
+    },
+    {
+      "epoch": 0.782214193336481,
+      "grad_norm": 0.33770779393011047,
+      "learning_rate": 8.5219625268144e-06,
+      "loss": 0.5003,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7823784205448238,
+      "grad_norm": 0.35718367865642603,
+      "learning_rate": 8.521752158436954e-06,
+      "loss": 0.4931,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7825426477531665,
+      "grad_norm": 0.3941435431475404,
+      "learning_rate": 8.521541746379323e-06,
+      "loss": 0.5216,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7827068749615093,
+      "grad_norm": 0.2865143701939515,
+      "learning_rate": 8.521331290643791e-06,
+      "loss": 0.5063,
+      "step": 4766
+    },
+    {
+      "epoch": 0.782871102169852,
+      "grad_norm": 0.28411967137007504,
+      "learning_rate": 8.521120791232646e-06,
+      "loss": 0.5222,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7830353293781948,
+      "grad_norm": 0.3053542974959439,
+      "learning_rate": 8.520910248148174e-06,
+      "loss": 0.5365,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7831995565865375,
+      "grad_norm": 0.3912844265585773,
+      "learning_rate": 8.52069966139266e-06,
+      "loss": 0.5257,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7833637837948803,
+      "grad_norm": 0.34825980844701676,
+      "learning_rate": 8.52048903096839e-06,
+      "loss": 0.5452,
+      "step": 4770
+    },
+    {
+      "epoch": 0.783528011003223,
+      "grad_norm": 0.2853114021603045,
+      "learning_rate": 8.520278356877654e-06,
+      "loss": 0.5073,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7836922382115658,
+      "grad_norm": 0.3703869120538064,
+      "learning_rate": 8.52006763912274e-06,
+      "loss": 0.4969,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7838564654199084,
+      "grad_norm": 0.27611234565761733,
+      "learning_rate": 8.519856877705937e-06,
+      "loss": 0.5344,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7840206926282511,
+      "grad_norm": 0.28744438714816317,
+      "learning_rate": 8.519646072629533e-06,
+      "loss": 0.4983,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7841849198365939,
+      "grad_norm": 0.3552014651411184,
+      "learning_rate": 8.519435223895817e-06,
+      "loss": 0.5018,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7843491470449366,
+      "grad_norm": 0.2757339930350272,
+      "learning_rate": 8.519224331507081e-06,
+      "loss": 0.5222,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7845133742532794,
+      "grad_norm": 0.3417843165815299,
+      "learning_rate": 8.519013395465614e-06,
+      "loss": 0.5093,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7846776014616221,
+      "grad_norm": 0.3300629702593418,
+      "learning_rate": 8.518802415773707e-06,
+      "loss": 0.5405,
+      "step": 4778
+    },
+    {
+      "epoch": 0.7848418286699649,
+      "grad_norm": 0.3143676247665393,
+      "learning_rate": 8.518591392433653e-06,
+      "loss": 0.5189,
+      "step": 4779
+    },
+    {
+      "epoch": 0.7850060558783076,
+      "grad_norm": 0.33319228445074417,
+      "learning_rate": 8.518380325447741e-06,
+      "loss": 0.5412,
+      "step": 4780
+    },
+    {
+      "epoch": 0.7851702830866504,
+      "grad_norm": 0.5097647567354313,
+      "learning_rate": 8.518169214818265e-06,
+      "loss": 0.4978,
+      "step": 4781
+    },
+    {
+      "epoch": 0.7853345102949931,
+      "grad_norm": 0.31219081932027304,
+      "learning_rate": 8.51795806054752e-06,
+      "loss": 0.5069,
+      "step": 4782
+    },
+    {
+      "epoch": 0.7854987375033359,
+      "grad_norm": 0.3588349702401945,
+      "learning_rate": 8.517746862637797e-06,
+      "loss": 0.5079,
+      "step": 4783
+    },
+    {
+      "epoch": 0.7856629647116786,
+      "grad_norm": 0.3102582288830573,
+      "learning_rate": 8.517535621091388e-06,
+      "loss": 0.5205,
+      "step": 4784
+    },
+    {
+      "epoch": 0.7858271919200214,
+      "grad_norm": 0.2918198592895266,
+      "learning_rate": 8.517324335910591e-06,
+      "loss": 0.5156,
+      "step": 4785
+    },
+    {
+      "epoch": 0.7859914191283641,
+      "grad_norm": 0.33430126386274583,
+      "learning_rate": 8.5171130070977e-06,
+      "loss": 0.5051,
+      "step": 4786
+    },
+    {
+      "epoch": 0.7861556463367069,
+      "grad_norm": 0.3254535703108297,
+      "learning_rate": 8.516901634655008e-06,
+      "loss": 0.5253,
+      "step": 4787
+    },
+    {
+      "epoch": 0.7863198735450496,
+      "grad_norm": 0.3484996970943442,
+      "learning_rate": 8.516690218584811e-06,
+      "loss": 0.5093,
+      "step": 4788
+    },
+    {
+      "epoch": 0.7864841007533924,
+      "grad_norm": 0.263519147067759,
+      "learning_rate": 8.51647875888941e-06,
+      "loss": 0.5106,
+      "step": 4789
+    },
+    {
+      "epoch": 0.786648327961735,
+      "grad_norm": 0.45546638774831777,
+      "learning_rate": 8.516267255571094e-06,
+      "loss": 0.5176,
+      "step": 4790
+    },
+    {
+      "epoch": 0.7868125551700778,
+      "grad_norm": 0.3449889074977041,
+      "learning_rate": 8.516055708632166e-06,
+      "loss": 0.5301,
+      "step": 4791
+    },
+    {
+      "epoch": 0.7869767823784205,
+      "grad_norm": 0.3698943802374998,
+      "learning_rate": 8.515844118074923e-06,
+      "loss": 0.5229,
+      "step": 4792
+    },
+    {
+      "epoch": 0.7871410095867633,
+      "grad_norm": 0.33462868726553185,
+      "learning_rate": 8.51563248390166e-06,
+      "loss": 0.5161,
+      "step": 4793
+    },
+    {
+      "epoch": 0.787305236795106,
+      "grad_norm": 0.3123568050312096,
+      "learning_rate": 8.515420806114677e-06,
+      "loss": 0.5109,
+      "step": 4794
+    },
+    {
+      "epoch": 0.7874694640034487,
+      "grad_norm": 0.34422885124330654,
+      "learning_rate": 8.515209084716275e-06,
+      "loss": 0.4919,
+      "step": 4795
+    },
+    {
+      "epoch": 0.7876336912117915,
+      "grad_norm": 0.3889682489845154,
+      "learning_rate": 8.514997319708751e-06,
+      "loss": 0.4851,
+      "step": 4796
+    },
+    {
+      "epoch": 0.7877979184201342,
+      "grad_norm": 0.36985116050350514,
+      "learning_rate": 8.514785511094408e-06,
+      "loss": 0.5292,
+      "step": 4797
+    },
+    {
+      "epoch": 0.787962145628477,
+      "grad_norm": 0.32305288714276936,
+      "learning_rate": 8.514573658875541e-06,
+      "loss": 0.4837,
+      "step": 4798
+    },
+    {
+      "epoch": 0.7881263728368197,
+      "grad_norm": 0.41138982582893036,
+      "learning_rate": 8.514361763054456e-06,
+      "loss": 0.5034,
+      "step": 4799
+    },
+    {
+      "epoch": 0.7882906000451625,
+      "grad_norm": 0.35851481494533577,
+      "learning_rate": 8.514149823633453e-06,
+      "loss": 0.53,
+      "step": 4800
+    },
+    {
+      "epoch": 0.7884548272535052,
+      "grad_norm": 0.3239257397621457,
+      "learning_rate": 8.513937840614832e-06,
+      "loss": 0.5206,
+      "step": 4801
+    },
+    {
+      "epoch": 0.788619054461848,
+      "grad_norm": 0.49352899282790785,
+      "learning_rate": 8.513725814000898e-06,
+      "loss": 0.5201,
+      "step": 4802
+    },
+    {
+      "epoch": 0.7887832816701907,
+      "grad_norm": 0.29476675219482074,
+      "learning_rate": 8.513513743793954e-06,
+      "loss": 0.4968,
+      "step": 4803
+    },
+    {
+      "epoch": 0.7889475088785335,
+      "grad_norm": 0.3763905524883607,
+      "learning_rate": 8.513301629996299e-06,
+      "loss": 0.4996,
+      "step": 4804
+    },
+    {
+      "epoch": 0.7891117360868762,
+      "grad_norm": 0.5222654404707635,
+      "learning_rate": 8.513089472610242e-06,
+      "loss": 0.5233,
+      "step": 4805
+    },
+    {
+      "epoch": 0.789275963295219,
+      "grad_norm": 0.33072041061279667,
+      "learning_rate": 8.512877271638084e-06,
+      "loss": 0.4858,
+      "step": 4806
+    },
+    {
+      "epoch": 0.7894401905035616,
+      "grad_norm": 0.28791095300849756,
+      "learning_rate": 8.51266502708213e-06,
+      "loss": 0.5171,
+      "step": 4807
+    },
+    {
+      "epoch": 0.7896044177119044,
+      "grad_norm": 0.29139790932119547,
+      "learning_rate": 8.512452738944686e-06,
+      "loss": 0.5295,
+      "step": 4808
+    },
+    {
+      "epoch": 0.7897686449202471,
+      "grad_norm": 0.2964080244353324,
+      "learning_rate": 8.51224040722806e-06,
+      "loss": 0.4971,
+      "step": 4809
+    },
+    {
+      "epoch": 0.7899328721285899,
+      "grad_norm": 0.32197530233101723,
+      "learning_rate": 8.512028031934554e-06,
+      "loss": 0.4932,
+      "step": 4810
+    },
+    {
+      "epoch": 0.7900970993369326,
+      "grad_norm": 0.2843080931858888,
+      "learning_rate": 8.511815613066475e-06,
+      "loss": 0.5005,
+      "step": 4811
+    },
+    {
+      "epoch": 0.7902613265452754,
+      "grad_norm": 0.535477832250208,
+      "learning_rate": 8.511603150626132e-06,
+      "loss": 0.5186,
+      "step": 4812
+    },
+    {
+      "epoch": 0.7904255537536181,
+      "grad_norm": 0.29692623047749966,
+      "learning_rate": 8.511390644615833e-06,
+      "loss": 0.5124,
+      "step": 4813
+    },
+    {
+      "epoch": 0.7905897809619609,
+      "grad_norm": 0.8690043401486638,
+      "learning_rate": 8.511178095037885e-06,
+      "loss": 0.5288,
+      "step": 4814
+    },
+    {
+      "epoch": 0.7907540081703036,
+      "grad_norm": 0.3356252094797372,
+      "learning_rate": 8.510965501894595e-06,
+      "loss": 0.5202,
+      "step": 4815
+    },
+    {
+      "epoch": 0.7909182353786464,
+      "grad_norm": 0.2574155639032546,
+      "learning_rate": 8.510752865188275e-06,
+      "loss": 0.4958,
+      "step": 4816
+    },
+    {
+      "epoch": 0.7910824625869891,
+      "grad_norm": 0.26080904931135435,
+      "learning_rate": 8.510540184921233e-06,
+      "loss": 0.4926,
+      "step": 4817
+    },
+    {
+      "epoch": 0.7912466897953319,
+      "grad_norm": 0.3428360730324517,
+      "learning_rate": 8.510327461095777e-06,
+      "loss": 0.5051,
+      "step": 4818
+    },
+    {
+      "epoch": 0.7914109170036746,
+      "grad_norm": 0.4252710145129881,
+      "learning_rate": 8.510114693714219e-06,
+      "loss": 0.5369,
+      "step": 4819
+    },
+    {
+      "epoch": 0.7915751442120174,
+      "grad_norm": 0.29726020666126923,
+      "learning_rate": 8.509901882778872e-06,
+      "loss": 0.4963,
+      "step": 4820
+    },
+    {
+      "epoch": 0.7917393714203601,
+      "grad_norm": 0.34172269458973087,
+      "learning_rate": 8.509689028292045e-06,
+      "loss": 0.539,
+      "step": 4821
+    },
+    {
+      "epoch": 0.7919035986287029,
+      "grad_norm": 0.30840633192280165,
+      "learning_rate": 8.509476130256049e-06,
+      "loss": 0.4956,
+      "step": 4822
+    },
+    {
+      "epoch": 0.7920678258370456,
+      "grad_norm": 0.27131424112280017,
+      "learning_rate": 8.509263188673198e-06,
+      "loss": 0.5096,
+      "step": 4823
+    },
+    {
+      "epoch": 0.7922320530453882,
+      "grad_norm": 0.3743847578318213,
+      "learning_rate": 8.509050203545803e-06,
+      "loss": 0.5065,
+      "step": 4824
+    },
+    {
+      "epoch": 0.792396280253731,
+      "grad_norm": 0.3795053796958781,
+      "learning_rate": 8.50883717487618e-06,
+      "loss": 0.5179,
+      "step": 4825
+    },
+    {
+      "epoch": 0.7925605074620737,
+      "grad_norm": 0.30026452426018335,
+      "learning_rate": 8.508624102666642e-06,
+      "loss": 0.5137,
+      "step": 4826
+    },
+    {
+      "epoch": 0.7927247346704165,
+      "grad_norm": 0.31326093789198417,
+      "learning_rate": 8.508410986919499e-06,
+      "loss": 0.5146,
+      "step": 4827
+    },
+    {
+      "epoch": 0.7928889618787592,
+      "grad_norm": 0.3620854669484693,
+      "learning_rate": 8.508197827637073e-06,
+      "loss": 0.5227,
+      "step": 4828
+    },
+    {
+      "epoch": 0.793053189087102,
+      "grad_norm": 0.3014049980906719,
+      "learning_rate": 8.507984624821672e-06,
+      "loss": 0.5094,
+      "step": 4829
+    },
+    {
+      "epoch": 0.7932174162954447,
+      "grad_norm": 0.3847631333967564,
+      "learning_rate": 8.507771378475614e-06,
+      "loss": 0.5218,
+      "step": 4830
+    },
+    {
+      "epoch": 0.7933816435037875,
+      "grad_norm": 0.2873710411856193,
+      "learning_rate": 8.507558088601218e-06,
+      "loss": 0.4793,
+      "step": 4831
+    },
+    {
+      "epoch": 0.7935458707121302,
+      "grad_norm": 0.2906589330262389,
+      "learning_rate": 8.507344755200797e-06,
+      "loss": 0.5164,
+      "step": 4832
+    },
+    {
+      "epoch": 0.793710097920473,
+      "grad_norm": 0.2843661884302468,
+      "learning_rate": 8.507131378276671e-06,
+      "loss": 0.5149,
+      "step": 4833
+    },
+    {
+      "epoch": 0.7938743251288157,
+      "grad_norm": 0.3119387145742007,
+      "learning_rate": 8.506917957831153e-06,
+      "loss": 0.4946,
+      "step": 4834
+    },
+    {
+      "epoch": 0.7940385523371585,
+      "grad_norm": 1.3155836455625542,
+      "learning_rate": 8.506704493866567e-06,
+      "loss": 0.5234,
+      "step": 4835
+    },
+    {
+      "epoch": 0.7942027795455012,
+      "grad_norm": 0.33312679287940616,
+      "learning_rate": 8.506490986385225e-06,
+      "loss": 0.503,
+      "step": 4836
+    },
+    {
+      "epoch": 0.794367006753844,
+      "grad_norm": 0.2994563085360433,
+      "learning_rate": 8.506277435389452e-06,
+      "loss": 0.5368,
+      "step": 4837
+    },
+    {
+      "epoch": 0.7945312339621867,
+      "grad_norm": 0.37734204973357055,
+      "learning_rate": 8.506063840881562e-06,
+      "loss": 0.5047,
+      "step": 4838
+    },
+    {
+      "epoch": 0.7946954611705295,
+      "grad_norm": 0.3317028490234112,
+      "learning_rate": 8.505850202863878e-06,
+      "loss": 0.5231,
+      "step": 4839
+    },
+    {
+      "epoch": 0.7948596883788722,
+      "grad_norm": 0.3180236812745663,
+      "learning_rate": 8.50563652133872e-06,
+      "loss": 0.498,
+      "step": 4840
+    },
+    {
+      "epoch": 0.7950239155872149,
+      "grad_norm": 0.31286873557820605,
+      "learning_rate": 8.505422796308408e-06,
+      "loss": 0.5154,
+      "step": 4841
+    },
+    {
+      "epoch": 0.7951881427955576,
+      "grad_norm": 0.3322648618516919,
+      "learning_rate": 8.505209027775263e-06,
+      "loss": 0.5144,
+      "step": 4842
+    },
+    {
+      "epoch": 0.7953523700039004,
+      "grad_norm": 0.3246666933401929,
+      "learning_rate": 8.504995215741608e-06,
+      "loss": 0.509,
+      "step": 4843
+    },
+    {
+      "epoch": 0.7955165972122431,
+      "grad_norm": 0.29849518101027295,
+      "learning_rate": 8.504781360209765e-06,
+      "loss": 0.5007,
+      "step": 4844
+    },
+    {
+      "epoch": 0.7956808244205859,
+      "grad_norm": 0.5686947232248678,
+      "learning_rate": 8.504567461182056e-06,
+      "loss": 0.5184,
+      "step": 4845
+    },
+    {
+      "epoch": 0.7958450516289286,
+      "grad_norm": 0.3001488106983265,
+      "learning_rate": 8.504353518660804e-06,
+      "loss": 0.5114,
+      "step": 4846
+    },
+    {
+      "epoch": 0.7960092788372714,
+      "grad_norm": 0.2899684109515469,
+      "learning_rate": 8.504139532648333e-06,
+      "loss": 0.5319,
+      "step": 4847
+    },
+    {
+      "epoch": 0.7961735060456141,
+      "grad_norm": 0.34811202789673956,
+      "learning_rate": 8.503925503146968e-06,
+      "loss": 0.5048,
+      "step": 4848
+    },
+    {
+      "epoch": 0.7963377332539568,
+      "grad_norm": 0.28213312593019935,
+      "learning_rate": 8.503711430159031e-06,
+      "loss": 0.5225,
+      "step": 4849
+    },
+    {
+      "epoch": 0.7965019604622996,
+      "grad_norm": 0.31306875930476336,
+      "learning_rate": 8.50349731368685e-06,
+      "loss": 0.5086,
+      "step": 4850
+    },
+    {
+      "epoch": 0.7966661876706423,
+      "grad_norm": 0.2767900885059024,
+      "learning_rate": 8.50328315373275e-06,
+      "loss": 0.5246,
+      "step": 4851
+    },
+    {
+      "epoch": 0.7968304148789851,
+      "grad_norm": 0.3163293048450247,
+      "learning_rate": 8.503068950299054e-06,
+      "loss": 0.5234,
+      "step": 4852
+    },
+    {
+      "epoch": 0.7969946420873278,
+      "grad_norm": 0.337058184039474,
+      "learning_rate": 8.502854703388094e-06,
+      "loss": 0.5123,
+      "step": 4853
+    },
+    {
+      "epoch": 0.7971588692956706,
+      "grad_norm": 0.5143054872690204,
+      "learning_rate": 8.502640413002191e-06,
+      "loss": 0.5047,
+      "step": 4854
+    },
+    {
+      "epoch": 0.7973230965040133,
+      "grad_norm": 0.4630950984648018,
+      "learning_rate": 8.502426079143675e-06,
+      "loss": 0.5085,
+      "step": 4855
+    },
+    {
+      "epoch": 0.7974873237123561,
+      "grad_norm": 0.329732576829339,
+      "learning_rate": 8.502211701814876e-06,
+      "loss": 0.5182,
+      "step": 4856
+    },
+    {
+      "epoch": 0.7976515509206988,
+      "grad_norm": 0.3156590174293186,
+      "learning_rate": 8.501997281018118e-06,
+      "loss": 0.5087,
+      "step": 4857
+    },
+    {
+      "epoch": 0.7978157781290415,
+      "grad_norm": 0.3275346397983897,
+      "learning_rate": 8.501782816755732e-06,
+      "loss": 0.5282,
+      "step": 4858
+    },
+    {
+      "epoch": 0.7979800053373842,
+      "grad_norm": 0.3255908453673845,
+      "learning_rate": 8.50156830903005e-06,
+      "loss": 0.534,
+      "step": 4859
+    },
+    {
+      "epoch": 0.798144232545727,
+      "grad_norm": 0.41658664353399205,
+      "learning_rate": 8.501353757843398e-06,
+      "loss": 0.5289,
+      "step": 4860
+    },
+    {
+      "epoch": 0.7983084597540697,
+      "grad_norm": 0.33432186201094094,
+      "learning_rate": 8.501139163198106e-06,
+      "loss": 0.4954,
+      "step": 4861
+    },
+    {
+      "epoch": 0.7984726869624125,
+      "grad_norm": 0.3844040400667963,
+      "learning_rate": 8.500924525096508e-06,
+      "loss": 0.5358,
+      "step": 4862
+    },
+    {
+      "epoch": 0.7986369141707552,
+      "grad_norm": 0.28735140226035877,
+      "learning_rate": 8.500709843540931e-06,
+      "loss": 0.5387,
+      "step": 4863
+    },
+    {
+      "epoch": 0.798801141379098,
+      "grad_norm": 0.4156195104655946,
+      "learning_rate": 8.500495118533712e-06,
+      "loss": 0.5146,
+      "step": 4864
+    },
+    {
+      "epoch": 0.7989653685874407,
+      "grad_norm": 0.4438432023377831,
+      "learning_rate": 8.500280350077176e-06,
+      "loss": 0.5063,
+      "step": 4865
+    },
+    {
+      "epoch": 0.7991295957957835,
+      "grad_norm": 0.313254971038212,
+      "learning_rate": 8.500065538173663e-06,
+      "loss": 0.5168,
+      "step": 4866
+    },
+    {
+      "epoch": 0.7992938230041262,
+      "grad_norm": 0.31318752813737577,
+      "learning_rate": 8.499850682825502e-06,
+      "loss": 0.5017,
+      "step": 4867
+    },
+    {
+      "epoch": 0.799458050212469,
+      "grad_norm": 0.29010297933150886,
+      "learning_rate": 8.499635784035026e-06,
+      "loss": 0.5076,
+      "step": 4868
+    },
+    {
+      "epoch": 0.7996222774208117,
+      "grad_norm": 0.34097058759666493,
+      "learning_rate": 8.499420841804572e-06,
+      "loss": 0.5075,
+      "step": 4869
+    },
+    {
+      "epoch": 0.7997865046291545,
+      "grad_norm": 0.34264820615195757,
+      "learning_rate": 8.499205856136473e-06,
+      "loss": 0.5353,
+      "step": 4870
+    },
+    {
+      "epoch": 0.7999507318374972,
+      "grad_norm": 0.31214009809327814,
+      "learning_rate": 8.498990827033062e-06,
+      "loss": 0.5176,
+      "step": 4871
+    },
+    {
+      "epoch": 0.80011495904584,
+      "grad_norm": 0.2919647666804401,
+      "learning_rate": 8.498775754496677e-06,
+      "loss": 0.5113,
+      "step": 4872
+    },
+    {
+      "epoch": 0.8002791862541827,
+      "grad_norm": 0.29216191333093305,
+      "learning_rate": 8.498560638529654e-06,
+      "loss": 0.4954,
+      "step": 4873
+    },
+    {
+      "epoch": 0.8004434134625255,
+      "grad_norm": 0.3648942224661276,
+      "learning_rate": 8.498345479134327e-06,
+      "loss": 0.5205,
+      "step": 4874
+    },
+    {
+      "epoch": 0.8006076406708681,
+      "grad_norm": 0.30774809099478256,
+      "learning_rate": 8.498130276313035e-06,
+      "loss": 0.5478,
+      "step": 4875
+    },
+    {
+      "epoch": 0.8007718678792108,
+      "grad_norm": 0.3454301234007663,
+      "learning_rate": 8.497915030068113e-06,
+      "loss": 0.4924,
+      "step": 4876
+    },
+    {
+      "epoch": 0.8009360950875536,
+      "grad_norm": 0.3290857865234024,
+      "learning_rate": 8.497699740401901e-06,
+      "loss": 0.5202,
+      "step": 4877
+    },
+    {
+      "epoch": 0.8011003222958963,
+      "grad_norm": 0.3546434922240655,
+      "learning_rate": 8.497484407316737e-06,
+      "loss": 0.4984,
+      "step": 4878
+    },
+    {
+      "epoch": 0.8012645495042391,
+      "grad_norm": 0.33323543843021725,
+      "learning_rate": 8.49726903081496e-06,
+      "loss": 0.5027,
+      "step": 4879
+    },
+    {
+      "epoch": 0.8014287767125818,
+      "grad_norm": 0.26386927946032285,
+      "learning_rate": 8.497053610898908e-06,
+      "loss": 0.5337,
+      "step": 4880
+    },
+    {
+      "epoch": 0.8015930039209246,
+      "grad_norm": 0.40317599119391767,
+      "learning_rate": 8.49683814757092e-06,
+      "loss": 0.5328,
+      "step": 4881
+    },
+    {
+      "epoch": 0.8017572311292673,
+      "grad_norm": 0.29720970041618555,
+      "learning_rate": 8.49662264083334e-06,
+      "loss": 0.5172,
+      "step": 4882
+    },
+    {
+      "epoch": 0.8019214583376101,
+      "grad_norm": 0.30483337934954274,
+      "learning_rate": 8.496407090688505e-06,
+      "loss": 0.5347,
+      "step": 4883
+    },
+    {
+      "epoch": 0.8020856855459528,
+      "grad_norm": 0.3390354026182259,
+      "learning_rate": 8.496191497138757e-06,
+      "loss": 0.5032,
+      "step": 4884
+    },
+    {
+      "epoch": 0.8022499127542956,
+      "grad_norm": 0.46548930989663456,
+      "learning_rate": 8.495975860186437e-06,
+      "loss": 0.5125,
+      "step": 4885
+    },
+    {
+      "epoch": 0.8024141399626383,
+      "grad_norm": 0.3909156993177809,
+      "learning_rate": 8.495760179833888e-06,
+      "loss": 0.5321,
+      "step": 4886
+    },
+    {
+      "epoch": 0.8025783671709811,
+      "grad_norm": 0.30815593000960556,
+      "learning_rate": 8.495544456083453e-06,
+      "loss": 0.5152,
+      "step": 4887
+    },
+    {
+      "epoch": 0.8027425943793238,
+      "grad_norm": 0.43576154208490675,
+      "learning_rate": 8.495328688937473e-06,
+      "loss": 0.5118,
+      "step": 4888
+    },
+    {
+      "epoch": 0.8029068215876666,
+      "grad_norm": 0.3362982778923814,
+      "learning_rate": 8.495112878398292e-06,
+      "loss": 0.5199,
+      "step": 4889
+    },
+    {
+      "epoch": 0.8030710487960093,
+      "grad_norm": 0.35648097356314234,
+      "learning_rate": 8.494897024468255e-06,
+      "loss": 0.5203,
+      "step": 4890
+    },
+    {
+      "epoch": 0.8032352760043521,
+      "grad_norm": 0.40099724052233265,
+      "learning_rate": 8.494681127149706e-06,
+      "loss": 0.5142,
+      "step": 4891
+    },
+    {
+      "epoch": 0.8033995032126947,
+      "grad_norm": 0.38363888188790995,
+      "learning_rate": 8.49446518644499e-06,
+      "loss": 0.5151,
+      "step": 4892
+    },
+    {
+      "epoch": 0.8035637304210375,
+      "grad_norm": 0.32707620079419586,
+      "learning_rate": 8.494249202356452e-06,
+      "loss": 0.5143,
+      "step": 4893
+    },
+    {
+      "epoch": 0.8037279576293802,
+      "grad_norm": 0.28527511913999304,
+      "learning_rate": 8.494033174886438e-06,
+      "loss": 0.5169,
+      "step": 4894
+    },
+    {
+      "epoch": 0.803892184837723,
+      "grad_norm": 0.2832345525798682,
+      "learning_rate": 8.493817104037294e-06,
+      "loss": 0.5078,
+      "step": 4895
+    },
+    {
+      "epoch": 0.8040564120460657,
+      "grad_norm": 1.4074803625802637,
+      "learning_rate": 8.493600989811366e-06,
+      "loss": 0.5151,
+      "step": 4896
+    },
+    {
+      "epoch": 0.8042206392544085,
+      "grad_norm": 0.31620078421037934,
+      "learning_rate": 8.493384832211003e-06,
+      "loss": 0.5338,
+      "step": 4897
+    },
+    {
+      "epoch": 0.8043848664627512,
+      "grad_norm": 0.36798937850075064,
+      "learning_rate": 8.49316863123855e-06,
+      "loss": 0.5047,
+      "step": 4898
+    },
+    {
+      "epoch": 0.804549093671094,
+      "grad_norm": 0.3155938826874778,
+      "learning_rate": 8.49295238689636e-06,
+      "loss": 0.5316,
+      "step": 4899
+    },
+    {
+      "epoch": 0.8047133208794367,
+      "grad_norm": 0.3090401650499103,
+      "learning_rate": 8.492736099186776e-06,
+      "loss": 0.4977,
+      "step": 4900
+    },
+    {
+      "epoch": 0.8048775480877794,
+      "grad_norm": 0.4837767806205548,
+      "learning_rate": 8.492519768112152e-06,
+      "loss": 0.5188,
+      "step": 4901
+    },
+    {
+      "epoch": 0.8050417752961222,
+      "grad_norm": 0.2831082962206799,
+      "learning_rate": 8.492303393674834e-06,
+      "loss": 0.5166,
+      "step": 4902
+    },
+    {
+      "epoch": 0.805206002504465,
+      "grad_norm": 0.3364271799232852,
+      "learning_rate": 8.492086975877173e-06,
+      "loss": 0.5032,
+      "step": 4903
+    },
+    {
+      "epoch": 0.8053702297128077,
+      "grad_norm": 0.3039287965588309,
+      "learning_rate": 8.49187051472152e-06,
+      "loss": 0.4997,
+      "step": 4904
+    },
+    {
+      "epoch": 0.8055344569211504,
+      "grad_norm": 0.34438901266088634,
+      "learning_rate": 8.491654010210224e-06,
+      "loss": 0.5224,
+      "step": 4905
+    },
+    {
+      "epoch": 0.8056986841294932,
+      "grad_norm": 0.410477956559512,
+      "learning_rate": 8.49143746234564e-06,
+      "loss": 0.521,
+      "step": 4906
+    },
+    {
+      "epoch": 0.8058629113378359,
+      "grad_norm": 0.33178147394701,
+      "learning_rate": 8.491220871130119e-06,
+      "loss": 0.5141,
+      "step": 4907
+    },
+    {
+      "epoch": 0.8060271385461787,
+      "grad_norm": 0.31226511617932345,
+      "learning_rate": 8.49100423656601e-06,
+      "loss": 0.5098,
+      "step": 4908
+    },
+    {
+      "epoch": 0.8061913657545213,
+      "grad_norm": 0.3314940198037555,
+      "learning_rate": 8.49078755865567e-06,
+      "loss": 0.4966,
+      "step": 4909
+    },
+    {
+      "epoch": 0.8063555929628641,
+      "grad_norm": 0.4213880727332896,
+      "learning_rate": 8.490570837401452e-06,
+      "loss": 0.5326,
+      "step": 4910
+    },
+    {
+      "epoch": 0.8065198201712068,
+      "grad_norm": 0.25770417899130643,
+      "learning_rate": 8.490354072805707e-06,
+      "loss": 0.5258,
+      "step": 4911
+    },
+    {
+      "epoch": 0.8066840473795496,
+      "grad_norm": 0.29938701046613064,
+      "learning_rate": 8.49013726487079e-06,
+      "loss": 0.5288,
+      "step": 4912
+    },
+    {
+      "epoch": 0.8068482745878923,
+      "grad_norm": 0.3284235847108112,
+      "learning_rate": 8.489920413599059e-06,
+      "loss": 0.5331,
+      "step": 4913
+    },
+    {
+      "epoch": 0.8070125017962351,
+      "grad_norm": 0.3092253263939411,
+      "learning_rate": 8.489703518992865e-06,
+      "loss": 0.5127,
+      "step": 4914
+    },
+    {
+      "epoch": 0.8071767290045778,
+      "grad_norm": 0.31053867213043934,
+      "learning_rate": 8.489486581054565e-06,
+      "loss": 0.5181,
+      "step": 4915
+    },
+    {
+      "epoch": 0.8073409562129206,
+      "grad_norm": 0.2781424039531393,
+      "learning_rate": 8.489269599786516e-06,
+      "loss": 0.4907,
+      "step": 4916
+    },
+    {
+      "epoch": 0.8075051834212633,
+      "grad_norm": 0.29524804072816085,
+      "learning_rate": 8.489052575191074e-06,
+      "loss": 0.5155,
+      "step": 4917
+    },
+    {
+      "epoch": 0.8076694106296061,
+      "grad_norm": 0.4577844332755319,
+      "learning_rate": 8.488835507270597e-06,
+      "loss": 0.5027,
+      "step": 4918
+    },
+    {
+      "epoch": 0.8078336378379488,
+      "grad_norm": 0.30584168281729424,
+      "learning_rate": 8.488618396027442e-06,
+      "loss": 0.521,
+      "step": 4919
+    },
+    {
+      "epoch": 0.8079978650462916,
+      "grad_norm": 0.3880806237530359,
+      "learning_rate": 8.488401241463966e-06,
+      "loss": 0.5352,
+      "step": 4920
+    },
+    {
+      "epoch": 0.8081620922546343,
+      "grad_norm": 0.2955963334280968,
+      "learning_rate": 8.48818404358253e-06,
+      "loss": 0.5111,
+      "step": 4921
+    },
+    {
+      "epoch": 0.8083263194629771,
+      "grad_norm": 0.3345446561322464,
+      "learning_rate": 8.48796680238549e-06,
+      "loss": 0.522,
+      "step": 4922
+    },
+    {
+      "epoch": 0.8084905466713198,
+      "grad_norm": 0.6481791245070249,
+      "learning_rate": 8.487749517875208e-06,
+      "loss": 0.5065,
+      "step": 4923
+    },
+    {
+      "epoch": 0.8086547738796626,
+      "grad_norm": 0.30909560781645284,
+      "learning_rate": 8.487532190054043e-06,
+      "loss": 0.5039,
+      "step": 4924
+    },
+    {
+      "epoch": 0.8088190010880053,
+      "grad_norm": 0.34142963673856785,
+      "learning_rate": 8.487314818924353e-06,
+      "loss": 0.5092,
+      "step": 4925
+    },
+    {
+      "epoch": 0.808983228296348,
+      "grad_norm": 0.4889539171715669,
+      "learning_rate": 8.487097404488502e-06,
+      "loss": 0.502,
+      "step": 4926
+    },
+    {
+      "epoch": 0.8091474555046907,
+      "grad_norm": 0.5818910886026137,
+      "learning_rate": 8.486879946748852e-06,
+      "loss": 0.4925,
+      "step": 4927
+    },
+    {
+      "epoch": 0.8093116827130334,
+      "grad_norm": 0.2932001298911658,
+      "learning_rate": 8.486662445707762e-06,
+      "loss": 0.5166,
+      "step": 4928
+    },
+    {
+      "epoch": 0.8094759099213762,
+      "grad_norm": 0.29941350737145134,
+      "learning_rate": 8.486444901367594e-06,
+      "loss": 0.5076,
+      "step": 4929
+    },
+    {
+      "epoch": 0.8096401371297189,
+      "grad_norm": 0.36587344576528646,
+      "learning_rate": 8.486227313730716e-06,
+      "loss": 0.4944,
+      "step": 4930
+    },
+    {
+      "epoch": 0.8098043643380617,
+      "grad_norm": 0.3008403149194141,
+      "learning_rate": 8.486009682799484e-06,
+      "loss": 0.5398,
+      "step": 4931
+    },
+    {
+      "epoch": 0.8099685915464044,
+      "grad_norm": 0.32000631815715636,
+      "learning_rate": 8.485792008576269e-06,
+      "loss": 0.5143,
+      "step": 4932
+    },
+    {
+      "epoch": 0.8101328187547472,
+      "grad_norm": 0.4554555718255998,
+      "learning_rate": 8.485574291063427e-06,
+      "loss": 0.5106,
+      "step": 4933
+    },
+    {
+      "epoch": 0.8102970459630899,
+      "grad_norm": 0.3396815075083648,
+      "learning_rate": 8.48535653026333e-06,
+      "loss": 0.5248,
+      "step": 4934
+    },
+    {
+      "epoch": 0.8104612731714327,
+      "grad_norm": 0.29132372652667043,
+      "learning_rate": 8.485138726178337e-06,
+      "loss": 0.5149,
+      "step": 4935
+    },
+    {
+      "epoch": 0.8106255003797754,
+      "grad_norm": 0.33433360634314313,
+      "learning_rate": 8.484920878810818e-06,
+      "loss": 0.5275,
+      "step": 4936
+    },
+    {
+      "epoch": 0.8107897275881182,
+      "grad_norm": 0.3804802811332157,
+      "learning_rate": 8.484702988163138e-06,
+      "loss": 0.5372,
+      "step": 4937
+    },
+    {
+      "epoch": 0.8109539547964609,
+      "grad_norm": 0.29590088600146536,
+      "learning_rate": 8.484485054237663e-06,
+      "loss": 0.5194,
+      "step": 4938
+    },
+    {
+      "epoch": 0.8111181820048037,
+      "grad_norm": 0.30898279514885435,
+      "learning_rate": 8.484267077036761e-06,
+      "loss": 0.5067,
+      "step": 4939
+    },
+    {
+      "epoch": 0.8112824092131464,
+      "grad_norm": 0.3169244925099661,
+      "learning_rate": 8.484049056562796e-06,
+      "loss": 0.5271,
+      "step": 4940
+    },
+    {
+      "epoch": 0.8114466364214892,
+      "grad_norm": 0.2927706493730989,
+      "learning_rate": 8.483830992818141e-06,
+      "loss": 0.5497,
+      "step": 4941
+    },
+    {
+      "epoch": 0.8116108636298319,
+      "grad_norm": 0.3785507710524582,
+      "learning_rate": 8.483612885805161e-06,
+      "loss": 0.495,
+      "step": 4942
+    },
+    {
+      "epoch": 0.8117750908381746,
+      "grad_norm": 0.2886097711844155,
+      "learning_rate": 8.483394735526226e-06,
+      "loss": 0.5161,
+      "step": 4943
+    },
+    {
+      "epoch": 0.8119393180465173,
+      "grad_norm": 0.3721913462493364,
+      "learning_rate": 8.483176541983706e-06,
+      "loss": 0.5196,
+      "step": 4944
+    },
+    {
+      "epoch": 0.8121035452548601,
+      "grad_norm": 0.30708790688915927,
+      "learning_rate": 8.482958305179967e-06,
+      "loss": 0.516,
+      "step": 4945
+    },
+    {
+      "epoch": 0.8122677724632028,
+      "grad_norm": 0.3761360983371278,
+      "learning_rate": 8.482740025117385e-06,
+      "loss": 0.5023,
+      "step": 4946
+    },
+    {
+      "epoch": 0.8124319996715456,
+      "grad_norm": 0.3213489583168228,
+      "learning_rate": 8.482521701798326e-06,
+      "loss": 0.5272,
+      "step": 4947
+    },
+    {
+      "epoch": 0.8125962268798883,
+      "grad_norm": 0.3341586201507787,
+      "learning_rate": 8.482303335225164e-06,
+      "loss": 0.5461,
+      "step": 4948
+    },
+    {
+      "epoch": 0.812760454088231,
+      "grad_norm": 0.29858786103982876,
+      "learning_rate": 8.48208492540027e-06,
+      "loss": 0.5105,
+      "step": 4949
+    },
+    {
+      "epoch": 0.8129246812965738,
+      "grad_norm": 0.32960934334045244,
+      "learning_rate": 8.481866472326015e-06,
+      "loss": 0.4996,
+      "step": 4950
+    },
+    {
+      "epoch": 0.8130889085049166,
+      "grad_norm": 0.2762505138441524,
+      "learning_rate": 8.481647976004773e-06,
+      "loss": 0.4963,
+      "step": 4951
+    },
+    {
+      "epoch": 0.8132531357132593,
+      "grad_norm": 0.3157606414855119,
+      "learning_rate": 8.481429436438916e-06,
+      "loss": 0.5259,
+      "step": 4952
+    },
+    {
+      "epoch": 0.813417362921602,
+      "grad_norm": 0.31727752076093524,
+      "learning_rate": 8.481210853630819e-06,
+      "loss": 0.4996,
+      "step": 4953
+    },
+    {
+      "epoch": 0.8135815901299448,
+      "grad_norm": 0.3480649578971305,
+      "learning_rate": 8.480992227582854e-06,
+      "loss": 0.5198,
+      "step": 4954
+    },
+    {
+      "epoch": 0.8137458173382875,
+      "grad_norm": 0.6437859463406534,
+      "learning_rate": 8.480773558297396e-06,
+      "loss": 0.5263,
+      "step": 4955
+    },
+    {
+      "epoch": 0.8139100445466303,
+      "grad_norm": 0.5095893397019333,
+      "learning_rate": 8.480554845776823e-06,
+      "loss": 0.5211,
+      "step": 4956
+    },
+    {
+      "epoch": 0.814074271754973,
+      "grad_norm": 0.3655560668561918,
+      "learning_rate": 8.480336090023506e-06,
+      "loss": 0.4943,
+      "step": 4957
+    },
+    {
+      "epoch": 0.8142384989633158,
+      "grad_norm": 0.29143515911184514,
+      "learning_rate": 8.480117291039825e-06,
+      "loss": 0.5029,
+      "step": 4958
+    },
+    {
+      "epoch": 0.8144027261716585,
+      "grad_norm": 0.3366487975682201,
+      "learning_rate": 8.479898448828154e-06,
+      "loss": 0.5425,
+      "step": 4959
+    },
+    {
+      "epoch": 0.8145669533800012,
+      "grad_norm": 0.3642127674595192,
+      "learning_rate": 8.479679563390868e-06,
+      "loss": 0.5197,
+      "step": 4960
+    },
+    {
+      "epoch": 0.8147311805883439,
+      "grad_norm": 0.2796960538183455,
+      "learning_rate": 8.479460634730347e-06,
+      "loss": 0.5093,
+      "step": 4961
+    },
+    {
+      "epoch": 0.8148954077966867,
+      "grad_norm": 0.3228600225851728,
+      "learning_rate": 8.47924166284897e-06,
+      "loss": 0.5218,
+      "step": 4962
+    },
+    {
+      "epoch": 0.8150596350050294,
+      "grad_norm": 0.4189371272712448,
+      "learning_rate": 8.479022647749112e-06,
+      "loss": 0.536,
+      "step": 4963
+    },
+    {
+      "epoch": 0.8152238622133722,
+      "grad_norm": 0.3709891099591292,
+      "learning_rate": 8.478803589433154e-06,
+      "loss": 0.4873,
+      "step": 4964
+    },
+    {
+      "epoch": 0.8153880894217149,
+      "grad_norm": 0.2950453878548158,
+      "learning_rate": 8.478584487903475e-06,
+      "loss": 0.5154,
+      "step": 4965
+    },
+    {
+      "epoch": 0.8155523166300577,
+      "grad_norm": 0.35668420853045724,
+      "learning_rate": 8.478365343162452e-06,
+      "loss": 0.5237,
+      "step": 4966
+    },
+    {
+      "epoch": 0.8157165438384004,
+      "grad_norm": 0.5139188172905095,
+      "learning_rate": 8.478146155212469e-06,
+      "loss": 0.5194,
+      "step": 4967
+    },
+    {
+      "epoch": 0.8158807710467432,
+      "grad_norm": 0.3158036394353838,
+      "learning_rate": 8.477926924055905e-06,
+      "loss": 0.5001,
+      "step": 4968
+    },
+    {
+      "epoch": 0.8160449982550859,
+      "grad_norm": 0.35621905036856805,
+      "learning_rate": 8.477707649695139e-06,
+      "loss": 0.5315,
+      "step": 4969
+    },
+    {
+      "epoch": 0.8162092254634287,
+      "grad_norm": 0.4811902163500481,
+      "learning_rate": 8.477488332132554e-06,
+      "loss": 0.4975,
+      "step": 4970
+    },
+    {
+      "epoch": 0.8163734526717714,
+      "grad_norm": 0.2924221106772756,
+      "learning_rate": 8.477268971370535e-06,
+      "loss": 0.5045,
+      "step": 4971
+    },
+    {
+      "epoch": 0.8165376798801142,
+      "grad_norm": 0.3111977388086835,
+      "learning_rate": 8.47704956741146e-06,
+      "loss": 0.5046,
+      "step": 4972
+    },
+    {
+      "epoch": 0.8167019070884569,
+      "grad_norm": 0.5014288256969606,
+      "learning_rate": 8.476830120257715e-06,
+      "loss": 0.522,
+      "step": 4973
+    },
+    {
+      "epoch": 0.8168661342967997,
+      "grad_norm": 0.30421993520314305,
+      "learning_rate": 8.47661062991168e-06,
+      "loss": 0.5172,
+      "step": 4974
+    },
+    {
+      "epoch": 0.8170303615051424,
+      "grad_norm": 0.3200743075995799,
+      "learning_rate": 8.476391096375744e-06,
+      "loss": 0.4989,
+      "step": 4975
+    },
+    {
+      "epoch": 0.8171945887134852,
+      "grad_norm": 0.3458207012656067,
+      "learning_rate": 8.476171519652288e-06,
+      "loss": 0.5056,
+      "step": 4976
+    },
+    {
+      "epoch": 0.8173588159218278,
+      "grad_norm": 0.2890057786017369,
+      "learning_rate": 8.475951899743695e-06,
+      "loss": 0.5013,
+      "step": 4977
+    },
+    {
+      "epoch": 0.8175230431301705,
+      "grad_norm": 0.3296551060524736,
+      "learning_rate": 8.475732236652356e-06,
+      "loss": 0.51,
+      "step": 4978
+    },
+    {
+      "epoch": 0.8176872703385133,
+      "grad_norm": 0.3430824710511646,
+      "learning_rate": 8.47551253038065e-06,
+      "loss": 0.4904,
+      "step": 4979
+    },
+    {
+      "epoch": 0.817851497546856,
+      "grad_norm": 0.2904594269875365,
+      "learning_rate": 8.475292780930968e-06,
+      "loss": 0.5092,
+      "step": 4980
+    },
+    {
+      "epoch": 0.8180157247551988,
+      "grad_norm": 0.30751543008238613,
+      "learning_rate": 8.475072988305696e-06,
+      "loss": 0.523,
+      "step": 4981
+    },
+    {
+      "epoch": 0.8181799519635415,
+      "grad_norm": 0.47299931672604006,
+      "learning_rate": 8.474853152507219e-06,
+      "loss": 0.4972,
+      "step": 4982
+    },
+    {
+      "epoch": 0.8183441791718843,
+      "grad_norm": 0.3891091435955476,
+      "learning_rate": 8.474633273537927e-06,
+      "loss": 0.5199,
+      "step": 4983
+    },
+    {
+      "epoch": 0.818508406380227,
+      "grad_norm": 0.3165875924441726,
+      "learning_rate": 8.474413351400206e-06,
+      "loss": 0.5225,
+      "step": 4984
+    },
+    {
+      "epoch": 0.8186726335885698,
+      "grad_norm": 0.3692974632226017,
+      "learning_rate": 8.474193386096447e-06,
+      "loss": 0.5208,
+      "step": 4985
+    },
+    {
+      "epoch": 0.8188368607969125,
+      "grad_norm": 0.7791784938729102,
+      "learning_rate": 8.473973377629036e-06,
+      "loss": 0.5032,
+      "step": 4986
+    },
+    {
+      "epoch": 0.8190010880052553,
+      "grad_norm": 0.4044083364461391,
+      "learning_rate": 8.473753326000367e-06,
+      "loss": 0.5233,
+      "step": 4987
+    },
+    {
+      "epoch": 0.819165315213598,
+      "grad_norm": 0.4033049001473696,
+      "learning_rate": 8.473533231212827e-06,
+      "loss": 0.5099,
+      "step": 4988
+    },
+    {
+      "epoch": 0.8193295424219408,
+      "grad_norm": 0.3596693458396118,
+      "learning_rate": 8.473313093268805e-06,
+      "loss": 0.5225,
+      "step": 4989
+    },
+    {
+      "epoch": 0.8194937696302835,
+      "grad_norm": 0.41076914940960707,
+      "learning_rate": 8.473092912170692e-06,
+      "loss": 0.5124,
+      "step": 4990
+    },
+    {
+      "epoch": 0.8196579968386263,
+      "grad_norm": 0.3401356269581785,
+      "learning_rate": 8.472872687920884e-06,
+      "loss": 0.506,
+      "step": 4991
+    },
+    {
+      "epoch": 0.819822224046969,
+      "grad_norm": 0.47125293670893087,
+      "learning_rate": 8.472652420521768e-06,
+      "loss": 0.5247,
+      "step": 4992
+    },
+    {
+      "epoch": 0.8199864512553118,
+      "grad_norm": 0.28918564354635806,
+      "learning_rate": 8.472432109975739e-06,
+      "loss": 0.5196,
+      "step": 4993
+    },
+    {
+      "epoch": 0.8201506784636544,
+      "grad_norm": 0.36219635279432766,
+      "learning_rate": 8.47221175628519e-06,
+      "loss": 0.5192,
+      "step": 4994
+    },
+    {
+      "epoch": 0.8203149056719972,
+      "grad_norm": 0.34955506029391137,
+      "learning_rate": 8.47199135945251e-06,
+      "loss": 0.51,
+      "step": 4995
+    },
+    {
+      "epoch": 0.8204791328803399,
+      "grad_norm": 0.45748560279916733,
+      "learning_rate": 8.471770919480099e-06,
+      "loss": 0.4931,
+      "step": 4996
+    },
+    {
+      "epoch": 0.8206433600886827,
+      "grad_norm": 0.9575608714826659,
+      "learning_rate": 8.471550436370348e-06,
+      "loss": 0.5599,
+      "step": 4997
+    },
+    {
+      "epoch": 0.8208075872970254,
+      "grad_norm": 0.2853618278425937,
+      "learning_rate": 8.47132991012565e-06,
+      "loss": 0.5248,
+      "step": 4998
+    },
+    {
+      "epoch": 0.8209718145053682,
+      "grad_norm": 0.3093195551711788,
+      "learning_rate": 8.471109340748404e-06,
+      "loss": 0.5099,
+      "step": 4999
+    },
+    {
+      "epoch": 0.8211360417137109,
+      "grad_norm": 0.319549610457436,
+      "learning_rate": 8.470888728241e-06,
+      "loss": 0.5118,
+      "step": 5000
+    },
+    {
+      "epoch": 0.8213002689220537,
+      "grad_norm": 0.27962509644036526,
+      "learning_rate": 8.47066807260584e-06,
+      "loss": 0.5468,
+      "step": 5001
+    },
+    {
+      "epoch": 0.8214644961303964,
+      "grad_norm": 0.2948514500194613,
+      "learning_rate": 8.470447373845318e-06,
+      "loss": 0.5016,
+      "step": 5002
+    },
+    {
+      "epoch": 0.8216287233387392,
+      "grad_norm": 0.34033167071439674,
+      "learning_rate": 8.470226631961833e-06,
+      "loss": 0.5065,
+      "step": 5003
+    },
+    {
+      "epoch": 0.8217929505470819,
+      "grad_norm": 0.2879423318987142,
+      "learning_rate": 8.470005846957777e-06,
+      "loss": 0.5033,
+      "step": 5004
+    },
+    {
+      "epoch": 0.8219571777554247,
+      "grad_norm": 0.29707502090044696,
+      "learning_rate": 8.469785018835555e-06,
+      "loss": 0.4962,
+      "step": 5005
+    },
+    {
+      "epoch": 0.8221214049637674,
+      "grad_norm": 0.32249529594575727,
+      "learning_rate": 8.46956414759756e-06,
+      "loss": 0.5189,
+      "step": 5006
+    },
+    {
+      "epoch": 0.8222856321721101,
+      "grad_norm": 0.44102538660174917,
+      "learning_rate": 8.469343233246193e-06,
+      "loss": 0.5341,
+      "step": 5007
+    },
+    {
+      "epoch": 0.8224498593804529,
+      "grad_norm": 0.36244736256653914,
+      "learning_rate": 8.469122275783853e-06,
+      "loss": 0.5041,
+      "step": 5008
+    },
+    {
+      "epoch": 0.8226140865887956,
+      "grad_norm": 0.43767800046863825,
+      "learning_rate": 8.46890127521294e-06,
+      "loss": 0.5344,
+      "step": 5009
+    },
+    {
+      "epoch": 0.8227783137971384,
+      "grad_norm": 0.7840909278204317,
+      "learning_rate": 8.468680231535856e-06,
+      "loss": 0.5084,
+      "step": 5010
+    },
+    {
+      "epoch": 0.822942541005481,
+      "grad_norm": 0.39755023280098944,
+      "learning_rate": 8.468459144754998e-06,
+      "loss": 0.5206,
+      "step": 5011
+    },
+    {
+      "epoch": 0.8231067682138238,
+      "grad_norm": 0.3169144920240181,
+      "learning_rate": 8.468238014872769e-06,
+      "loss": 0.5044,
+      "step": 5012
+    },
+    {
+      "epoch": 0.8232709954221665,
+      "grad_norm": 0.32759349000574406,
+      "learning_rate": 8.468016841891572e-06,
+      "loss": 0.5082,
+      "step": 5013
+    },
+    {
+      "epoch": 0.8234352226305093,
+      "grad_norm": 0.3040957401155583,
+      "learning_rate": 8.467795625813808e-06,
+      "loss": 0.5149,
+      "step": 5014
+    },
+    {
+      "epoch": 0.823599449838852,
+      "grad_norm": 0.5909867028738016,
+      "learning_rate": 8.46757436664188e-06,
+      "loss": 0.5066,
+      "step": 5015
+    },
+    {
+      "epoch": 0.8237636770471948,
+      "grad_norm": 0.3331879110597187,
+      "learning_rate": 8.46735306437819e-06,
+      "loss": 0.5252,
+      "step": 5016
+    },
+    {
+      "epoch": 0.8239279042555375,
+      "grad_norm": 0.28786272922468276,
+      "learning_rate": 8.467131719025143e-06,
+      "loss": 0.5219,
+      "step": 5017
+    },
+    {
+      "epoch": 0.8240921314638803,
+      "grad_norm": 0.37116912026207094,
+      "learning_rate": 8.466910330585142e-06,
+      "loss": 0.5163,
+      "step": 5018
+    },
+    {
+      "epoch": 0.824256358672223,
+      "grad_norm": 0.41794447544277624,
+      "learning_rate": 8.466688899060593e-06,
+      "loss": 0.5218,
+      "step": 5019
+    },
+    {
+      "epoch": 0.8244205858805658,
+      "grad_norm": 0.3212900078391455,
+      "learning_rate": 8.466467424453898e-06,
+      "loss": 0.5098,
+      "step": 5020
+    },
+    {
+      "epoch": 0.8245848130889085,
+      "grad_norm": 0.3470167865008464,
+      "learning_rate": 8.466245906767464e-06,
+      "loss": 0.4971,
+      "step": 5021
+    },
+    {
+      "epoch": 0.8247490402972513,
+      "grad_norm": 0.3205638094446894,
+      "learning_rate": 8.4660243460037e-06,
+      "loss": 0.5236,
+      "step": 5022
+    },
+    {
+      "epoch": 0.824913267505594,
+      "grad_norm": 0.31070966884605533,
+      "learning_rate": 8.465802742165007e-06,
+      "loss": 0.4918,
+      "step": 5023
+    },
+    {
+      "epoch": 0.8250774947139368,
+      "grad_norm": 0.3138870801849995,
+      "learning_rate": 8.465581095253795e-06,
+      "loss": 0.5128,
+      "step": 5024
+    },
+    {
+      "epoch": 0.8252417219222795,
+      "grad_norm": 0.45582958216423314,
+      "learning_rate": 8.465359405272471e-06,
+      "loss": 0.5094,
+      "step": 5025
+    },
+    {
+      "epoch": 0.8254059491306223,
+      "grad_norm": 0.32263746700503765,
+      "learning_rate": 8.465137672223444e-06,
+      "loss": 0.5176,
+      "step": 5026
+    },
+    {
+      "epoch": 0.825570176338965,
+      "grad_norm": 0.36571660406047724,
+      "learning_rate": 8.464915896109118e-06,
+      "loss": 0.5176,
+      "step": 5027
+    },
+    {
+      "epoch": 0.8257344035473076,
+      "grad_norm": 0.37715049533934497,
+      "learning_rate": 8.464694076931907e-06,
+      "loss": 0.5048,
+      "step": 5028
+    },
+    {
+      "epoch": 0.8258986307556504,
+      "grad_norm": 0.3525169971936548,
+      "learning_rate": 8.464472214694216e-06,
+      "loss": 0.4956,
+      "step": 5029
+    },
+    {
+      "epoch": 0.8260628579639931,
+      "grad_norm": 0.34935023750783545,
+      "learning_rate": 8.464250309398457e-06,
+      "loss": 0.5043,
+      "step": 5030
+    },
+    {
+      "epoch": 0.8262270851723359,
+      "grad_norm": 0.3003527457744416,
+      "learning_rate": 8.464028361047037e-06,
+      "loss": 0.5098,
+      "step": 5031
+    },
+    {
+      "epoch": 0.8263913123806786,
+      "grad_norm": 0.3211925859594905,
+      "learning_rate": 8.463806369642373e-06,
+      "loss": 0.5151,
+      "step": 5032
+    },
+    {
+      "epoch": 0.8265555395890214,
+      "grad_norm": 0.2881298613603624,
+      "learning_rate": 8.46358433518687e-06,
+      "loss": 0.522,
+      "step": 5033
+    },
+    {
+      "epoch": 0.8267197667973641,
+      "grad_norm": 0.3088604066956432,
+      "learning_rate": 8.46336225768294e-06,
+      "loss": 0.5042,
+      "step": 5034
+    },
+    {
+      "epoch": 0.8268839940057069,
+      "grad_norm": 0.3207947009742772,
+      "learning_rate": 8.463140137132997e-06,
+      "loss": 0.4924,
+      "step": 5035
+    },
+    {
+      "epoch": 0.8270482212140496,
+      "grad_norm": 0.30583600515085474,
+      "learning_rate": 8.462917973539454e-06,
+      "loss": 0.5144,
+      "step": 5036
+    },
+    {
+      "epoch": 0.8272124484223924,
+      "grad_norm": 0.3273592428080974,
+      "learning_rate": 8.462695766904724e-06,
+      "loss": 0.5046,
+      "step": 5037
+    },
+    {
+      "epoch": 0.8273766756307351,
+      "grad_norm": 0.24219539916900257,
+      "learning_rate": 8.462473517231217e-06,
+      "loss": 0.4942,
+      "step": 5038
+    },
+    {
+      "epoch": 0.8275409028390779,
+      "grad_norm": 0.30598944645398374,
+      "learning_rate": 8.462251224521349e-06,
+      "loss": 0.5089,
+      "step": 5039
+    },
+    {
+      "epoch": 0.8277051300474206,
+      "grad_norm": 0.2890067799934394,
+      "learning_rate": 8.462028888777536e-06,
+      "loss": 0.5153,
+      "step": 5040
+    },
+    {
+      "epoch": 0.8278693572557634,
+      "grad_norm": 0.31820577059903815,
+      "learning_rate": 8.461806510002189e-06,
+      "loss": 0.4943,
+      "step": 5041
+    },
+    {
+      "epoch": 0.8280335844641061,
+      "grad_norm": 0.3160940376849428,
+      "learning_rate": 8.461584088197726e-06,
+      "loss": 0.5148,
+      "step": 5042
+    },
+    {
+      "epoch": 0.8281978116724489,
+      "grad_norm": 0.30044612758537664,
+      "learning_rate": 8.461361623366564e-06,
+      "loss": 0.498,
+      "step": 5043
+    },
+    {
+      "epoch": 0.8283620388807916,
+      "grad_norm": 0.26764067211180775,
+      "learning_rate": 8.461139115511116e-06,
+      "loss": 0.4952,
+      "step": 5044
+    },
+    {
+      "epoch": 0.8285262660891343,
+      "grad_norm": 0.31280182664520434,
+      "learning_rate": 8.4609165646338e-06,
+      "loss": 0.5031,
+      "step": 5045
+    },
+    {
+      "epoch": 0.828690493297477,
+      "grad_norm": 1.0443570125444503,
+      "learning_rate": 8.460693970737033e-06,
+      "loss": 0.5144,
+      "step": 5046
+    },
+    {
+      "epoch": 0.8288547205058198,
+      "grad_norm": 0.31610380504111363,
+      "learning_rate": 8.460471333823232e-06,
+      "loss": 0.5046,
+      "step": 5047
+    },
+    {
+      "epoch": 0.8290189477141625,
+      "grad_norm": 0.2824804649298298,
+      "learning_rate": 8.460248653894818e-06,
+      "loss": 0.5478,
+      "step": 5048
+    },
+    {
+      "epoch": 0.8291831749225053,
+      "grad_norm": 0.345837308750286,
+      "learning_rate": 8.460025930954206e-06,
+      "loss": 0.511,
+      "step": 5049
+    },
+    {
+      "epoch": 0.829347402130848,
+      "grad_norm": 0.3016406355661234,
+      "learning_rate": 8.459803165003815e-06,
+      "loss": 0.5182,
+      "step": 5050
+    },
+    {
+      "epoch": 0.8295116293391908,
+      "grad_norm": 0.3562213213518067,
+      "learning_rate": 8.459580356046067e-06,
+      "loss": 0.5082,
+      "step": 5051
+    },
+    {
+      "epoch": 0.8296758565475335,
+      "grad_norm": 0.3025481909002121,
+      "learning_rate": 8.459357504083381e-06,
+      "loss": 0.5105,
+      "step": 5052
+    },
+    {
+      "epoch": 0.8298400837558763,
+      "grad_norm": 0.33732616817813776,
+      "learning_rate": 8.459134609118175e-06,
+      "loss": 0.5031,
+      "step": 5053
+    },
+    {
+      "epoch": 0.830004310964219,
+      "grad_norm": 0.330670296404074,
+      "learning_rate": 8.458911671152874e-06,
+      "loss": 0.491,
+      "step": 5054
+    },
+    {
+      "epoch": 0.8301685381725618,
+      "grad_norm": 0.31696394533350386,
+      "learning_rate": 8.458688690189897e-06,
+      "loss": 0.5151,
+      "step": 5055
+    },
+    {
+      "epoch": 0.8303327653809045,
+      "grad_norm": 0.35470601501695254,
+      "learning_rate": 8.458465666231665e-06,
+      "loss": 0.5185,
+      "step": 5056
+    },
+    {
+      "epoch": 0.8304969925892473,
+      "grad_norm": 0.2745766999857397,
+      "learning_rate": 8.4582425992806e-06,
+      "loss": 0.5128,
+      "step": 5057
+    },
+    {
+      "epoch": 0.83066121979759,
+      "grad_norm": 0.2910543164432107,
+      "learning_rate": 8.458019489339129e-06,
+      "loss": 0.5187,
+      "step": 5058
+    },
+    {
+      "epoch": 0.8308254470059327,
+      "grad_norm": 0.3428610977610601,
+      "learning_rate": 8.457796336409672e-06,
+      "loss": 0.5007,
+      "step": 5059
+    },
+    {
+      "epoch": 0.8309896742142755,
+      "grad_norm": 0.27773717516347307,
+      "learning_rate": 8.45757314049465e-06,
+      "loss": 0.4994,
+      "step": 5060
+    },
+    {
+      "epoch": 0.8311539014226182,
+      "grad_norm": 0.28802939482408896,
+      "learning_rate": 8.457349901596492e-06,
+      "loss": 0.5205,
+      "step": 5061
+    },
+    {
+      "epoch": 0.8313181286309609,
+      "grad_norm": 0.26938147503957355,
+      "learning_rate": 8.45712661971762e-06,
+      "loss": 0.4928,
+      "step": 5062
+    },
+    {
+      "epoch": 0.8314823558393036,
+      "grad_norm": 0.27567436454998057,
+      "learning_rate": 8.456903294860462e-06,
+      "loss": 0.506,
+      "step": 5063
+    },
+    {
+      "epoch": 0.8316465830476464,
+      "grad_norm": 0.2871225248215651,
+      "learning_rate": 8.456679927027438e-06,
+      "loss": 0.5058,
+      "step": 5064
+    },
+    {
+      "epoch": 0.8318108102559891,
+      "grad_norm": 0.3209137982397029,
+      "learning_rate": 8.45645651622098e-06,
+      "loss": 0.5263,
+      "step": 5065
+    },
+    {
+      "epoch": 0.8319750374643319,
+      "grad_norm": 0.2785837470184095,
+      "learning_rate": 8.456233062443508e-06,
+      "loss": 0.51,
+      "step": 5066
+    },
+    {
+      "epoch": 0.8321392646726746,
+      "grad_norm": 0.29409807614294214,
+      "learning_rate": 8.456009565697455e-06,
+      "loss": 0.5561,
+      "step": 5067
+    },
+    {
+      "epoch": 0.8323034918810174,
+      "grad_norm": 0.28712716258345833,
+      "learning_rate": 8.455786025985244e-06,
+      "loss": 0.5115,
+      "step": 5068
+    },
+    {
+      "epoch": 0.8324677190893601,
+      "grad_norm": 0.284167141367049,
+      "learning_rate": 8.455562443309308e-06,
+      "loss": 0.5049,
+      "step": 5069
+    },
+    {
+      "epoch": 0.8326319462977029,
+      "grad_norm": 0.29551746610441004,
+      "learning_rate": 8.455338817672069e-06,
+      "loss": 0.4867,
+      "step": 5070
+    },
+    {
+      "epoch": 0.8327961735060456,
+      "grad_norm": 0.27646057535916885,
+      "learning_rate": 8.455115149075961e-06,
+      "loss": 0.5164,
+      "step": 5071
+    },
+    {
+      "epoch": 0.8329604007143884,
+      "grad_norm": 0.3010201736022869,
+      "learning_rate": 8.45489143752341e-06,
+      "loss": 0.5152,
+      "step": 5072
+    },
+    {
+      "epoch": 0.8331246279227311,
+      "grad_norm": 0.2622201228558617,
+      "learning_rate": 8.454667683016847e-06,
+      "loss": 0.4946,
+      "step": 5073
+    },
+    {
+      "epoch": 0.8332888551310739,
+      "grad_norm": 0.3100844241068556,
+      "learning_rate": 8.454443885558702e-06,
+      "loss": 0.5018,
+      "step": 5074
+    },
+    {
+      "epoch": 0.8334530823394166,
+      "grad_norm": 0.2609866015590737,
+      "learning_rate": 8.454220045151407e-06,
+      "loss": 0.5066,
+      "step": 5075
+    },
+    {
+      "epoch": 0.8336173095477594,
+      "grad_norm": 0.443640168900632,
+      "learning_rate": 8.45399616179739e-06,
+      "loss": 0.4883,
+      "step": 5076
+    },
+    {
+      "epoch": 0.8337815367561021,
+      "grad_norm": 0.3412324134114467,
+      "learning_rate": 8.453772235499085e-06,
+      "loss": 0.5173,
+      "step": 5077
+    },
+    {
+      "epoch": 0.8339457639644449,
+      "grad_norm": 0.31007454373042503,
+      "learning_rate": 8.453548266258924e-06,
+      "loss": 0.5246,
+      "step": 5078
+    },
+    {
+      "epoch": 0.8341099911727875,
+      "grad_norm": 0.3047252056284528,
+      "learning_rate": 8.45332425407934e-06,
+      "loss": 0.521,
+      "step": 5079
+    },
+    {
+      "epoch": 0.8342742183811303,
+      "grad_norm": 0.3046193883490871,
+      "learning_rate": 8.453100198962764e-06,
+      "loss": 0.5287,
+      "step": 5080
+    },
+    {
+      "epoch": 0.834438445589473,
+      "grad_norm": 0.2786460052072658,
+      "learning_rate": 8.45287610091163e-06,
+      "loss": 0.5062,
+      "step": 5081
+    },
+    {
+      "epoch": 0.8346026727978157,
+      "grad_norm": 0.3078651792888881,
+      "learning_rate": 8.452651959928374e-06,
+      "loss": 0.5048,
+      "step": 5082
+    },
+    {
+      "epoch": 0.8347669000061585,
+      "grad_norm": 0.2877663514705134,
+      "learning_rate": 8.452427776015428e-06,
+      "loss": 0.5403,
+      "step": 5083
+    },
+    {
+      "epoch": 0.8349311272145012,
+      "grad_norm": 0.275191748660642,
+      "learning_rate": 8.452203549175226e-06,
+      "loss": 0.5112,
+      "step": 5084
+    },
+    {
+      "epoch": 0.835095354422844,
+      "grad_norm": 0.29317499292240806,
+      "learning_rate": 8.451979279410207e-06,
+      "loss": 0.5207,
+      "step": 5085
+    },
+    {
+      "epoch": 0.8352595816311867,
+      "grad_norm": 0.28347051510346993,
+      "learning_rate": 8.451754966722804e-06,
+      "loss": 0.4999,
+      "step": 5086
+    },
+    {
+      "epoch": 0.8354238088395295,
+      "grad_norm": 0.3021476025483518,
+      "learning_rate": 8.451530611115456e-06,
+      "loss": 0.5157,
+      "step": 5087
+    },
+    {
+      "epoch": 0.8355880360478722,
+      "grad_norm": 0.32453355242197585,
+      "learning_rate": 8.451306212590595e-06,
+      "loss": 0.4964,
+      "step": 5088
+    },
+    {
+      "epoch": 0.835752263256215,
+      "grad_norm": 0.2786831375180388,
+      "learning_rate": 8.451081771150663e-06,
+      "loss": 0.5112,
+      "step": 5089
+    },
+    {
+      "epoch": 0.8359164904645577,
+      "grad_norm": 0.3315162275707279,
+      "learning_rate": 8.450857286798095e-06,
+      "loss": 0.5169,
+      "step": 5090
+    },
+    {
+      "epoch": 0.8360807176729005,
+      "grad_norm": 0.43424131015762185,
+      "learning_rate": 8.450632759535329e-06,
+      "loss": 0.4969,
+      "step": 5091
+    },
+    {
+      "epoch": 0.8362449448812432,
+      "grad_norm": 0.34337420336609564,
+      "learning_rate": 8.450408189364805e-06,
+      "loss": 0.522,
+      "step": 5092
+    },
+    {
+      "epoch": 0.836409172089586,
+      "grad_norm": 0.27471053663336875,
+      "learning_rate": 8.450183576288962e-06,
+      "loss": 0.5141,
+      "step": 5093
+    },
+    {
+      "epoch": 0.8365733992979287,
+      "grad_norm": 0.416085438847666,
+      "learning_rate": 8.449958920310237e-06,
+      "loss": 0.5061,
+      "step": 5094
+    },
+    {
+      "epoch": 0.8367376265062715,
+      "grad_norm": 0.32310592092724405,
+      "learning_rate": 8.449734221431073e-06,
+      "loss": 0.5235,
+      "step": 5095
+    },
+    {
+      "epoch": 0.8369018537146141,
+      "grad_norm": 0.3186821997518407,
+      "learning_rate": 8.449509479653911e-06,
+      "loss": 0.4872,
+      "step": 5096
+    },
+    {
+      "epoch": 0.8370660809229569,
+      "grad_norm": 0.2843584333397972,
+      "learning_rate": 8.449284694981187e-06,
+      "loss": 0.5301,
+      "step": 5097
+    },
+    {
+      "epoch": 0.8372303081312996,
+      "grad_norm": 0.3032449654190625,
+      "learning_rate": 8.449059867415348e-06,
+      "loss": 0.5181,
+      "step": 5098
+    },
+    {
+      "epoch": 0.8373945353396424,
+      "grad_norm": 0.33532208528088053,
+      "learning_rate": 8.448834996958833e-06,
+      "loss": 0.5166,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8375587625479851,
+      "grad_norm": 0.3208069728155115,
+      "learning_rate": 8.448610083614085e-06,
+      "loss": 0.5185,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8377229897563279,
+      "grad_norm": 0.2805736230950408,
+      "learning_rate": 8.448385127383546e-06,
+      "loss": 0.5269,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8378872169646706,
+      "grad_norm": 0.28498221869287943,
+      "learning_rate": 8.448160128269659e-06,
+      "loss": 0.5032,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8380514441730134,
+      "grad_norm": 0.3092573990050545,
+      "learning_rate": 8.44793508627487e-06,
+      "loss": 0.5213,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8382156713813561,
+      "grad_norm": 0.2936327617561994,
+      "learning_rate": 8.447710001401622e-06,
+      "loss": 0.4974,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8383798985896989,
+      "grad_norm": 0.3591294236741196,
+      "learning_rate": 8.447484873652358e-06,
+      "loss": 0.5023,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8385441257980416,
+      "grad_norm": 0.3441918054279239,
+      "learning_rate": 8.447259703029525e-06,
+      "loss": 0.5147,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8387083530063844,
+      "grad_norm": 0.3430452362801982,
+      "learning_rate": 8.447034489535569e-06,
+      "loss": 0.5152,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8388725802147271,
+      "grad_norm": 0.31263270317206515,
+      "learning_rate": 8.446809233172934e-06,
+      "loss": 0.524,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8390368074230699,
+      "grad_norm": 0.28639111012314905,
+      "learning_rate": 8.446583933944067e-06,
+      "loss": 0.519,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8392010346314126,
+      "grad_norm": 0.3446313615810392,
+      "learning_rate": 8.446358591851417e-06,
+      "loss": 0.5128,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8393652618397553,
+      "grad_norm": 0.5348002296317965,
+      "learning_rate": 8.446133206897429e-06,
+      "loss": 0.4922,
+      "step": 5111
+    },
+    {
+      "epoch": 0.8395294890480981,
+      "grad_norm": 0.5051111534121978,
+      "learning_rate": 8.445907779084553e-06,
+      "loss": 0.5418,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8396937162564407,
+      "grad_norm": 0.30406474946147793,
+      "learning_rate": 8.445682308415235e-06,
+      "loss": 0.4946,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8398579434647835,
+      "grad_norm": 0.3380086818662066,
+      "learning_rate": 8.445456794891925e-06,
+      "loss": 0.5255,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8400221706731262,
+      "grad_norm": 0.3438745728098006,
+      "learning_rate": 8.445231238517073e-06,
+      "loss": 0.5423,
+      "step": 5115
+    },
+    {
+      "epoch": 0.840186397881469,
+      "grad_norm": 0.40381447460269726,
+      "learning_rate": 8.44500563929313e-06,
+      "loss": 0.5137,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8403506250898117,
+      "grad_norm": 0.40787796480711397,
+      "learning_rate": 8.444779997222541e-06,
+      "loss": 0.5064,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8405148522981545,
+      "grad_norm": 0.36288902748934776,
+      "learning_rate": 8.444554312307763e-06,
+      "loss": 0.5082,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8406790795064972,
+      "grad_norm": 0.30047433656216793,
+      "learning_rate": 8.444328584551243e-06,
+      "loss": 0.512,
+      "step": 5119
+    },
+    {
+      "epoch": 0.84084330671484,
+      "grad_norm": 0.3750624604659782,
+      "learning_rate": 8.444102813955435e-06,
+      "loss": 0.5186,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8410075339231827,
+      "grad_norm": 0.34600270284100215,
+      "learning_rate": 8.443877000522788e-06,
+      "loss": 0.4973,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8411717611315255,
+      "grad_norm": 0.3139983714728081,
+      "learning_rate": 8.443651144255756e-06,
+      "loss": 0.5256,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8413359883398682,
+      "grad_norm": 0.3346488739033981,
+      "learning_rate": 8.443425245156795e-06,
+      "loss": 0.5163,
+      "step": 5123
+    },
+    {
+      "epoch": 0.841500215548211,
+      "grad_norm": 0.2948288518207391,
+      "learning_rate": 8.443199303228355e-06,
+      "loss": 0.5234,
+      "step": 5124
+    },
+    {
+      "epoch": 0.8416644427565537,
+      "grad_norm": 0.33203493477375534,
+      "learning_rate": 8.44297331847289e-06,
+      "loss": 0.5105,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8418286699648965,
+      "grad_norm": 0.30379991266832673,
+      "learning_rate": 8.442747290892856e-06,
+      "loss": 0.5144,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8419928971732392,
+      "grad_norm": 0.28029415277526454,
+      "learning_rate": 8.44252122049071e-06,
+      "loss": 0.508,
+      "step": 5127
+    },
+    {
+      "epoch": 0.842157124381582,
+      "grad_norm": 0.49847696844628037,
+      "learning_rate": 8.4422951072689e-06,
+      "loss": 0.5056,
+      "step": 5128
+    },
+    {
+      "epoch": 0.8423213515899247,
+      "grad_norm": 0.5120725889382312,
+      "learning_rate": 8.44206895122989e-06,
+      "loss": 0.5298,
+      "step": 5129
+    },
+    {
+      "epoch": 0.8424855787982674,
+      "grad_norm": 0.2882998585510167,
+      "learning_rate": 8.44184275237613e-06,
+      "loss": 0.5023,
+      "step": 5130
+    },
+    {
+      "epoch": 0.8426498060066101,
+      "grad_norm": 0.3236394026994081,
+      "learning_rate": 8.441616510710082e-06,
+      "loss": 0.5098,
+      "step": 5131
+    },
+    {
+      "epoch": 0.8428140332149529,
+      "grad_norm": 0.3631167452727695,
+      "learning_rate": 8.441390226234199e-06,
+      "loss": 0.5206,
+      "step": 5132
+    },
+    {
+      "epoch": 0.8429782604232956,
+      "grad_norm": 0.3183503409285738,
+      "learning_rate": 8.441163898950941e-06,
+      "loss": 0.5229,
+      "step": 5133
+    },
+    {
+      "epoch": 0.8431424876316383,
+      "grad_norm": 0.30255751019464133,
+      "learning_rate": 8.440937528862766e-06,
+      "loss": 0.5029,
+      "step": 5134
+    },
+    {
+      "epoch": 0.8433067148399811,
+      "grad_norm": 0.3335196583075735,
+      "learning_rate": 8.440711115972131e-06,
+      "loss": 0.504,
+      "step": 5135
+    },
+    {
+      "epoch": 0.8434709420483238,
+      "grad_norm": 0.32001255656113226,
+      "learning_rate": 8.440484660281496e-06,
+      "loss": 0.495,
+      "step": 5136
+    },
+    {
+      "epoch": 0.8436351692566666,
+      "grad_norm": 0.36895012575981045,
+      "learning_rate": 8.440258161793321e-06,
+      "loss": 0.5073,
+      "step": 5137
+    },
+    {
+      "epoch": 0.8437993964650093,
+      "grad_norm": 0.2755984017532756,
+      "learning_rate": 8.440031620510068e-06,
+      "loss": 0.5289,
+      "step": 5138
+    },
+    {
+      "epoch": 0.8439636236733521,
+      "grad_norm": 0.3473947843063845,
+      "learning_rate": 8.439805036434191e-06,
+      "loss": 0.5169,
+      "step": 5139
+    },
+    {
+      "epoch": 0.8441278508816948,
+      "grad_norm": 0.30833567846849247,
+      "learning_rate": 8.439578409568158e-06,
+      "loss": 0.5049,
+      "step": 5140
+    },
+    {
+      "epoch": 0.8442920780900376,
+      "grad_norm": 0.2868463061560813,
+      "learning_rate": 8.439351739914427e-06,
+      "loss": 0.5024,
+      "step": 5141
+    },
+    {
+      "epoch": 0.8444563052983803,
+      "grad_norm": 0.27930508177210783,
+      "learning_rate": 8.439125027475459e-06,
+      "loss": 0.5055,
+      "step": 5142
+    },
+    {
+      "epoch": 0.8446205325067231,
+      "grad_norm": 0.38485774036475545,
+      "learning_rate": 8.438898272253719e-06,
+      "loss": 0.511,
+      "step": 5143
+    },
+    {
+      "epoch": 0.8447847597150658,
+      "grad_norm": 0.30246859023528366,
+      "learning_rate": 8.438671474251667e-06,
+      "loss": 0.5309,
+      "step": 5144
+    },
+    {
+      "epoch": 0.8449489869234086,
+      "grad_norm": 0.324011948208211,
+      "learning_rate": 8.43844463347177e-06,
+      "loss": 0.506,
+      "step": 5145
+    },
+    {
+      "epoch": 0.8451132141317512,
+      "grad_norm": 0.3072522424422487,
+      "learning_rate": 8.438217749916488e-06,
+      "loss": 0.5089,
+      "step": 5146
+    },
+    {
+      "epoch": 0.845277441340094,
+      "grad_norm": 0.44271916061839944,
+      "learning_rate": 8.437990823588285e-06,
+      "loss": 0.5399,
+      "step": 5147
+    },
+    {
+      "epoch": 0.8454416685484367,
+      "grad_norm": 0.4356296662687189,
+      "learning_rate": 8.43776385448963e-06,
+      "loss": 0.5228,
+      "step": 5148
+    },
+    {
+      "epoch": 0.8456058957567795,
+      "grad_norm": 0.2668006225000898,
+      "learning_rate": 8.437536842622982e-06,
+      "loss": 0.5014,
+      "step": 5149
+    },
+    {
+      "epoch": 0.8457701229651222,
+      "grad_norm": 0.26564245328819,
+      "learning_rate": 8.437309787990813e-06,
+      "loss": 0.5096,
+      "step": 5150
+    },
+    {
+      "epoch": 0.845934350173465,
+      "grad_norm": 0.2983680257581951,
+      "learning_rate": 8.437082690595584e-06,
+      "loss": 0.5136,
+      "step": 5151
+    },
+    {
+      "epoch": 0.8460985773818077,
+      "grad_norm": 0.40578399435243784,
+      "learning_rate": 8.436855550439765e-06,
+      "loss": 0.4978,
+      "step": 5152
+    },
+    {
+      "epoch": 0.8462628045901505,
+      "grad_norm": 0.44301373104727876,
+      "learning_rate": 8.43662836752582e-06,
+      "loss": 0.4988,
+      "step": 5153
+    },
+    {
+      "epoch": 0.8464270317984932,
+      "grad_norm": 0.39247545992276484,
+      "learning_rate": 8.436401141856218e-06,
+      "loss": 0.5338,
+      "step": 5154
+    },
+    {
+      "epoch": 0.846591259006836,
+      "grad_norm": 0.2681976914938843,
+      "learning_rate": 8.436173873433428e-06,
+      "loss": 0.5075,
+      "step": 5155
+    },
+    {
+      "epoch": 0.8467554862151787,
+      "grad_norm": 0.3673054253397242,
+      "learning_rate": 8.435946562259917e-06,
+      "loss": 0.4936,
+      "step": 5156
+    },
+    {
+      "epoch": 0.8469197134235215,
+      "grad_norm": 0.3534936525961469,
+      "learning_rate": 8.435719208338153e-06,
+      "loss": 0.5091,
+      "step": 5157
+    },
+    {
+      "epoch": 0.8470839406318642,
+      "grad_norm": 0.3079026562799358,
+      "learning_rate": 8.435491811670605e-06,
+      "loss": 0.5126,
+      "step": 5158
+    },
+    {
+      "epoch": 0.847248167840207,
+      "grad_norm": 0.37457392062869027,
+      "learning_rate": 8.435264372259745e-06,
+      "loss": 0.511,
+      "step": 5159
+    },
+    {
+      "epoch": 0.8474123950485497,
+      "grad_norm": 0.31250952877477833,
+      "learning_rate": 8.435036890108042e-06,
+      "loss": 0.5055,
+      "step": 5160
+    },
+    {
+      "epoch": 0.8475766222568925,
+      "grad_norm": 0.34748176886976945,
+      "learning_rate": 8.434809365217968e-06,
+      "loss": 0.5216,
+      "step": 5161
+    },
+    {
+      "epoch": 0.8477408494652352,
+      "grad_norm": 0.4337562987176165,
+      "learning_rate": 8.434581797591992e-06,
+      "loss": 0.5325,
+      "step": 5162
+    },
+    {
+      "epoch": 0.8479050766735778,
+      "grad_norm": 0.30289333346415104,
+      "learning_rate": 8.434354187232587e-06,
+      "loss": 0.5138,
+      "step": 5163
+    },
+    {
+      "epoch": 0.8480693038819206,
+      "grad_norm": 0.28790463141112965,
+      "learning_rate": 8.434126534142223e-06,
+      "loss": 0.5058,
+      "step": 5164
+    },
+    {
+      "epoch": 0.8482335310902633,
+      "grad_norm": 0.28706340640753114,
+      "learning_rate": 8.433898838323375e-06,
+      "loss": 0.5226,
+      "step": 5165
+    },
+    {
+      "epoch": 0.8483977582986061,
+      "grad_norm": 0.32648676816606786,
+      "learning_rate": 8.433671099778517e-06,
+      "loss": 0.5049,
+      "step": 5166
+    },
+    {
+      "epoch": 0.8485619855069488,
+      "grad_norm": 0.3414586375904832,
+      "learning_rate": 8.43344331851012e-06,
+      "loss": 0.4967,
+      "step": 5167
+    },
+    {
+      "epoch": 0.8487262127152916,
+      "grad_norm": 0.3822562701157267,
+      "learning_rate": 8.433215494520657e-06,
+      "loss": 0.4928,
+      "step": 5168
+    },
+    {
+      "epoch": 0.8488904399236343,
+      "grad_norm": 0.30473631955247477,
+      "learning_rate": 8.432987627812606e-06,
+      "loss": 0.5051,
+      "step": 5169
+    },
+    {
+      "epoch": 0.8490546671319771,
+      "grad_norm": 0.4222486533363501,
+      "learning_rate": 8.432759718388437e-06,
+      "loss": 0.5026,
+      "step": 5170
+    },
+    {
+      "epoch": 0.8492188943403198,
+      "grad_norm": 0.435952572835449,
+      "learning_rate": 8.43253176625063e-06,
+      "loss": 0.5014,
+      "step": 5171
+    },
+    {
+      "epoch": 0.8493831215486626,
+      "grad_norm": 0.35308923943760095,
+      "learning_rate": 8.432303771401659e-06,
+      "loss": 0.5239,
+      "step": 5172
+    },
+    {
+      "epoch": 0.8495473487570053,
+      "grad_norm": 0.3010694457237855,
+      "learning_rate": 8.432075733844e-06,
+      "loss": 0.5149,
+      "step": 5173
+    },
+    {
+      "epoch": 0.8497115759653481,
+      "grad_norm": 0.32228750187138333,
+      "learning_rate": 8.43184765358013e-06,
+      "loss": 0.5011,
+      "step": 5174
+    },
+    {
+      "epoch": 0.8498758031736908,
+      "grad_norm": 0.34908233140622946,
+      "learning_rate": 8.431619530612525e-06,
+      "loss": 0.5016,
+      "step": 5175
+    },
+    {
+      "epoch": 0.8500400303820336,
+      "grad_norm": 0.3444780549516735,
+      "learning_rate": 8.431391364943665e-06,
+      "loss": 0.4882,
+      "step": 5176
+    },
+    {
+      "epoch": 0.8502042575903763,
+      "grad_norm": 0.36189762611863124,
+      "learning_rate": 8.431163156576028e-06,
+      "loss": 0.5351,
+      "step": 5177
+    },
+    {
+      "epoch": 0.8503684847987191,
+      "grad_norm": 0.31406121921663654,
+      "learning_rate": 8.430934905512087e-06,
+      "loss": 0.5083,
+      "step": 5178
+    },
+    {
+      "epoch": 0.8505327120070618,
+      "grad_norm": 0.33779577449560827,
+      "learning_rate": 8.43070661175433e-06,
+      "loss": 0.5108,
+      "step": 5179
+    },
+    {
+      "epoch": 0.8506969392154045,
+      "grad_norm": 0.31857681875078253,
+      "learning_rate": 8.430478275305228e-06,
+      "loss": 0.5219,
+      "step": 5180
+    },
+    {
+      "epoch": 0.8508611664237472,
+      "grad_norm": 0.33539172162870345,
+      "learning_rate": 8.430249896167269e-06,
+      "loss": 0.5095,
+      "step": 5181
+    },
+    {
+      "epoch": 0.85102539363209,
+      "grad_norm": 0.3775500012400237,
+      "learning_rate": 8.430021474342928e-06,
+      "loss": 0.5084,
+      "step": 5182
+    },
+    {
+      "epoch": 0.8511896208404327,
+      "grad_norm": 0.3059924527973876,
+      "learning_rate": 8.429793009834685e-06,
+      "loss": 0.51,
+      "step": 5183
+    },
+    {
+      "epoch": 0.8513538480487755,
+      "grad_norm": 0.3871570428014681,
+      "learning_rate": 8.429564502645026e-06,
+      "loss": 0.527,
+      "step": 5184
+    },
+    {
+      "epoch": 0.8515180752571182,
+      "grad_norm": 0.3048202893164031,
+      "learning_rate": 8.429335952776428e-06,
+      "loss": 0.5057,
+      "step": 5185
+    },
+    {
+      "epoch": 0.851682302465461,
+      "grad_norm": 0.3039632803794553,
+      "learning_rate": 8.429107360231377e-06,
+      "loss": 0.4986,
+      "step": 5186
+    },
+    {
+      "epoch": 0.8518465296738037,
+      "grad_norm": 0.5060909240230959,
+      "learning_rate": 8.428878725012354e-06,
+      "loss": 0.5102,
+      "step": 5187
+    },
+    {
+      "epoch": 0.8520107568821464,
+      "grad_norm": 0.3212393034431408,
+      "learning_rate": 8.428650047121843e-06,
+      "loss": 0.5046,
+      "step": 5188
+    },
+    {
+      "epoch": 0.8521749840904892,
+      "grad_norm": 0.27915733547122695,
+      "learning_rate": 8.428421326562328e-06,
+      "loss": 0.5198,
+      "step": 5189
+    },
+    {
+      "epoch": 0.852339211298832,
+      "grad_norm": 0.28708430427289955,
+      "learning_rate": 8.42819256333629e-06,
+      "loss": 0.5145,
+      "step": 5190
+    },
+    {
+      "epoch": 0.8525034385071747,
+      "grad_norm": 0.30120292472904436,
+      "learning_rate": 8.427963757446218e-06,
+      "loss": 0.5093,
+      "step": 5191
+    },
+    {
+      "epoch": 0.8526676657155174,
+      "grad_norm": 0.35635732289785643,
+      "learning_rate": 8.427734908894594e-06,
+      "loss": 0.5047,
+      "step": 5192
+    },
+    {
+      "epoch": 0.8528318929238602,
+      "grad_norm": 0.28154755868785863,
+      "learning_rate": 8.427506017683905e-06,
+      "loss": 0.4954,
+      "step": 5193
+    },
+    {
+      "epoch": 0.8529961201322029,
+      "grad_norm": 0.31810663307518133,
+      "learning_rate": 8.427277083816636e-06,
+      "loss": 0.5163,
+      "step": 5194
+    },
+    {
+      "epoch": 0.8531603473405457,
+      "grad_norm": 0.4674936251067023,
+      "learning_rate": 8.427048107295275e-06,
+      "loss": 0.5103,
+      "step": 5195
+    },
+    {
+      "epoch": 0.8533245745488884,
+      "grad_norm": 0.3264717854724029,
+      "learning_rate": 8.426819088122307e-06,
+      "loss": 0.5157,
+      "step": 5196
+    },
+    {
+      "epoch": 0.8534888017572311,
+      "grad_norm": 0.34619937788013283,
+      "learning_rate": 8.42659002630022e-06,
+      "loss": 0.5148,
+      "step": 5197
+    },
+    {
+      "epoch": 0.8536530289655738,
+      "grad_norm": 0.30500362594914693,
+      "learning_rate": 8.426360921831503e-06,
+      "loss": 0.4945,
+      "step": 5198
+    },
+    {
+      "epoch": 0.8538172561739166,
+      "grad_norm": 0.35639512512192423,
+      "learning_rate": 8.426131774718641e-06,
+      "loss": 0.5236,
+      "step": 5199
+    },
+    {
+      "epoch": 0.8539814833822593,
+      "grad_norm": 0.32749143917043566,
+      "learning_rate": 8.425902584964129e-06,
+      "loss": 0.516,
+      "step": 5200
+    },
+    {
+      "epoch": 0.8541457105906021,
+      "grad_norm": 0.355371873683865,
+      "learning_rate": 8.425673352570448e-06,
+      "loss": 0.5243,
+      "step": 5201
+    },
+    {
+      "epoch": 0.8543099377989448,
+      "grad_norm": 0.3431264317700648,
+      "learning_rate": 8.425444077540094e-06,
+      "loss": 0.5081,
+      "step": 5202
+    },
+    {
+      "epoch": 0.8544741650072876,
+      "grad_norm": 0.46083194627277885,
+      "learning_rate": 8.425214759875558e-06,
+      "loss": 0.5187,
+      "step": 5203
+    },
+    {
+      "epoch": 0.8546383922156303,
+      "grad_norm": 0.3628692906065567,
+      "learning_rate": 8.424985399579323e-06,
+      "loss": 0.4859,
+      "step": 5204
+    },
+    {
+      "epoch": 0.8548026194239731,
+      "grad_norm": 0.3008814972088375,
+      "learning_rate": 8.424755996653889e-06,
+      "loss": 0.4949,
+      "step": 5205
+    },
+    {
+      "epoch": 0.8549668466323158,
+      "grad_norm": 0.33859496251143534,
+      "learning_rate": 8.424526551101741e-06,
+      "loss": 0.5117,
+      "step": 5206
+    },
+    {
+      "epoch": 0.8551310738406586,
+      "grad_norm": 0.3178485647764328,
+      "learning_rate": 8.424297062925375e-06,
+      "loss": 0.5097,
+      "step": 5207
+    },
+    {
+      "epoch": 0.8552953010490013,
+      "grad_norm": 0.310539171285264,
+      "learning_rate": 8.42406753212728e-06,
+      "loss": 0.5085,
+      "step": 5208
+    },
+    {
+      "epoch": 0.8554595282573441,
+      "grad_norm": 0.3222299566904057,
+      "learning_rate": 8.423837958709952e-06,
+      "loss": 0.5099,
+      "step": 5209
+    },
+    {
+      "epoch": 0.8556237554656868,
+      "grad_norm": 0.3062954314262607,
+      "learning_rate": 8.423608342675883e-06,
+      "loss": 0.4782,
+      "step": 5210
+    },
+    {
+      "epoch": 0.8557879826740296,
+      "grad_norm": 0.3855658097772442,
+      "learning_rate": 8.423378684027568e-06,
+      "loss": 0.525,
+      "step": 5211
+    },
+    {
+      "epoch": 0.8559522098823723,
+      "grad_norm": 0.4804967081262836,
+      "learning_rate": 8.4231489827675e-06,
+      "loss": 0.5052,
+      "step": 5212
+    },
+    {
+      "epoch": 0.856116437090715,
+      "grad_norm": 0.3271596419138035,
+      "learning_rate": 8.422919238898173e-06,
+      "loss": 0.4708,
+      "step": 5213
+    },
+    {
+      "epoch": 0.8562806642990577,
+      "grad_norm": 0.31481541934023216,
+      "learning_rate": 8.422689452422084e-06,
+      "loss": 0.5051,
+      "step": 5214
+    },
+    {
+      "epoch": 0.8564448915074004,
+      "grad_norm": 0.3099269647615455,
+      "learning_rate": 8.42245962334173e-06,
+      "loss": 0.4819,
+      "step": 5215
+    },
+    {
+      "epoch": 0.8566091187157432,
+      "grad_norm": 0.3438652387948608,
+      "learning_rate": 8.422229751659602e-06,
+      "loss": 0.4947,
+      "step": 5216
+    },
+    {
+      "epoch": 0.8567733459240859,
+      "grad_norm": 0.3459932973707239,
+      "learning_rate": 8.421999837378202e-06,
+      "loss": 0.53,
+      "step": 5217
+    },
+    {
+      "epoch": 0.8569375731324287,
+      "grad_norm": 0.27299285950949304,
+      "learning_rate": 8.421769880500025e-06,
+      "loss": 0.5281,
+      "step": 5218
+    },
+    {
+      "epoch": 0.8571018003407714,
+      "grad_norm": 0.27523955025295577,
+      "learning_rate": 8.421539881027568e-06,
+      "loss": 0.4841,
+      "step": 5219
+    },
+    {
+      "epoch": 0.8572660275491142,
+      "grad_norm": 0.34338702265752363,
+      "learning_rate": 8.42130983896333e-06,
+      "loss": 0.5255,
+      "step": 5220
+    },
+    {
+      "epoch": 0.8574302547574569,
+      "grad_norm": 0.3695909557488081,
+      "learning_rate": 8.421079754309808e-06,
+      "loss": 0.5017,
+      "step": 5221
+    },
+    {
+      "epoch": 0.8575944819657997,
+      "grad_norm": 0.26262217495334517,
+      "learning_rate": 8.420849627069504e-06,
+      "loss": 0.5295,
+      "step": 5222
+    },
+    {
+      "epoch": 0.8577587091741424,
+      "grad_norm": 0.3373161131890121,
+      "learning_rate": 8.420619457244915e-06,
+      "loss": 0.5172,
+      "step": 5223
+    },
+    {
+      "epoch": 0.8579229363824852,
+      "grad_norm": 0.4176117431308755,
+      "learning_rate": 8.42038924483854e-06,
+      "loss": 0.5118,
+      "step": 5224
+    },
+    {
+      "epoch": 0.8580871635908279,
+      "grad_norm": 0.4681792027400935,
+      "learning_rate": 8.420158989852881e-06,
+      "loss": 0.5068,
+      "step": 5225
+    },
+    {
+      "epoch": 0.8582513907991707,
+      "grad_norm": 0.2916379992376806,
+      "learning_rate": 8.41992869229044e-06,
+      "loss": 0.4982,
+      "step": 5226
+    },
+    {
+      "epoch": 0.8584156180075134,
+      "grad_norm": 0.27631425561265743,
+      "learning_rate": 8.419698352153715e-06,
+      "loss": 0.492,
+      "step": 5227
+    },
+    {
+      "epoch": 0.8585798452158562,
+      "grad_norm": 0.2795686993363181,
+      "learning_rate": 8.41946796944521e-06,
+      "loss": 0.5012,
+      "step": 5228
+    },
+    {
+      "epoch": 0.8587440724241989,
+      "grad_norm": 0.3963207725944165,
+      "learning_rate": 8.419237544167427e-06,
+      "loss": 0.4981,
+      "step": 5229
+    },
+    {
+      "epoch": 0.8589082996325417,
+      "grad_norm": 0.3328663936121557,
+      "learning_rate": 8.419007076322869e-06,
+      "loss": 0.5219,
+      "step": 5230
+    },
+    {
+      "epoch": 0.8590725268408843,
+      "grad_norm": 0.3493391789677639,
+      "learning_rate": 8.418776565914036e-06,
+      "loss": 0.5091,
+      "step": 5231
+    },
+    {
+      "epoch": 0.8592367540492271,
+      "grad_norm": 0.29098994266551437,
+      "learning_rate": 8.418546012943436e-06,
+      "loss": 0.5031,
+      "step": 5232
+    },
+    {
+      "epoch": 0.8594009812575698,
+      "grad_norm": 0.37668531359547847,
+      "learning_rate": 8.41831541741357e-06,
+      "loss": 0.5135,
+      "step": 5233
+    },
+    {
+      "epoch": 0.8595652084659126,
+      "grad_norm": 0.29737232046529355,
+      "learning_rate": 8.418084779326944e-06,
+      "loss": 0.5127,
+      "step": 5234
+    },
+    {
+      "epoch": 0.8597294356742553,
+      "grad_norm": 0.3572696957219688,
+      "learning_rate": 8.417854098686062e-06,
+      "loss": 0.5112,
+      "step": 5235
+    },
+    {
+      "epoch": 0.859893662882598,
+      "grad_norm": 0.29116505908127643,
+      "learning_rate": 8.41762337549343e-06,
+      "loss": 0.5032,
+      "step": 5236
+    },
+    {
+      "epoch": 0.8600578900909408,
+      "grad_norm": 0.3227422175765217,
+      "learning_rate": 8.417392609751553e-06,
+      "loss": 0.5192,
+      "step": 5237
+    },
+    {
+      "epoch": 0.8602221172992836,
+      "grad_norm": 0.3052186718364593,
+      "learning_rate": 8.417161801462939e-06,
+      "loss": 0.5033,
+      "step": 5238
+    },
+    {
+      "epoch": 0.8603863445076263,
+      "grad_norm": 0.2792044493744665,
+      "learning_rate": 8.416930950630094e-06,
+      "loss": 0.5113,
+      "step": 5239
+    },
+    {
+      "epoch": 0.860550571715969,
+      "grad_norm": 0.28137907424947856,
+      "learning_rate": 8.416700057255524e-06,
+      "loss": 0.485,
+      "step": 5240
+    },
+    {
+      "epoch": 0.8607147989243118,
+      "grad_norm": 0.3848989139038607,
+      "learning_rate": 8.416469121341739e-06,
+      "loss": 0.5146,
+      "step": 5241
+    },
+    {
+      "epoch": 0.8608790261326545,
+      "grad_norm": 0.3647020656817828,
+      "learning_rate": 8.416238142891246e-06,
+      "loss": 0.5123,
+      "step": 5242
+    },
+    {
+      "epoch": 0.8610432533409973,
+      "grad_norm": 0.30122311372081473,
+      "learning_rate": 8.416007121906553e-06,
+      "loss": 0.5191,
+      "step": 5243
+    },
+    {
+      "epoch": 0.86120748054934,
+      "grad_norm": 0.3073748720310311,
+      "learning_rate": 8.41577605839017e-06,
+      "loss": 0.5236,
+      "step": 5244
+    },
+    {
+      "epoch": 0.8613717077576828,
+      "grad_norm": 0.2905245125656662,
+      "learning_rate": 8.415544952344607e-06,
+      "loss": 0.4822,
+      "step": 5245
+    },
+    {
+      "epoch": 0.8615359349660255,
+      "grad_norm": 0.30468902373544227,
+      "learning_rate": 8.415313803772374e-06,
+      "loss": 0.4926,
+      "step": 5246
+    },
+    {
+      "epoch": 0.8617001621743683,
+      "grad_norm": 0.31458419828369694,
+      "learning_rate": 8.415082612675979e-06,
+      "loss": 0.5063,
+      "step": 5247
+    },
+    {
+      "epoch": 0.8618643893827109,
+      "grad_norm": 0.30256781957016937,
+      "learning_rate": 8.414851379057936e-06,
+      "loss": 0.5143,
+      "step": 5248
+    },
+    {
+      "epoch": 0.8620286165910537,
+      "grad_norm": 0.3512098795701238,
+      "learning_rate": 8.414620102920755e-06,
+      "loss": 0.5257,
+      "step": 5249
+    },
+    {
+      "epoch": 0.8621928437993964,
+      "grad_norm": 0.33016065512395715,
+      "learning_rate": 8.414388784266948e-06,
+      "loss": 0.5003,
+      "step": 5250
+    },
+    {
+      "epoch": 0.8623570710077392,
+      "grad_norm": 0.349765377162814,
+      "learning_rate": 8.41415742309903e-06,
+      "loss": 0.5098,
+      "step": 5251
+    },
+    {
+      "epoch": 0.8625212982160819,
+      "grad_norm": 0.3232602041674713,
+      "learning_rate": 8.413926019419508e-06,
+      "loss": 0.5323,
+      "step": 5252
+    },
+    {
+      "epoch": 0.8626855254244247,
+      "grad_norm": 0.33022614357718966,
+      "learning_rate": 8.4136945732309e-06,
+      "loss": 0.504,
+      "step": 5253
+    },
+    {
+      "epoch": 0.8628497526327674,
+      "grad_norm": 0.40419642120733285,
+      "learning_rate": 8.413463084535718e-06,
+      "loss": 0.5134,
+      "step": 5254
+    },
+    {
+      "epoch": 0.8630139798411102,
+      "grad_norm": 0.2947040903999055,
+      "learning_rate": 8.413231553336478e-06,
+      "loss": 0.52,
+      "step": 5255
+    },
+    {
+      "epoch": 0.8631782070494529,
+      "grad_norm": 0.3128924063031786,
+      "learning_rate": 8.412999979635692e-06,
+      "loss": 0.5085,
+      "step": 5256
+    },
+    {
+      "epoch": 0.8633424342577957,
+      "grad_norm": 0.34440429823573077,
+      "learning_rate": 8.412768363435875e-06,
+      "loss": 0.4934,
+      "step": 5257
+    },
+    {
+      "epoch": 0.8635066614661384,
+      "grad_norm": 0.3223141456692689,
+      "learning_rate": 8.412536704739547e-06,
+      "loss": 0.506,
+      "step": 5258
+    },
+    {
+      "epoch": 0.8636708886744812,
+      "grad_norm": 0.3620705564585329,
+      "learning_rate": 8.41230500354922e-06,
+      "loss": 0.5231,
+      "step": 5259
+    },
+    {
+      "epoch": 0.8638351158828239,
+      "grad_norm": 0.40002625330809766,
+      "learning_rate": 8.41207325986741e-06,
+      "loss": 0.5201,
+      "step": 5260
+    },
+    {
+      "epoch": 0.8639993430911667,
+      "grad_norm": 0.3658458873894559,
+      "learning_rate": 8.411841473696637e-06,
+      "loss": 0.5202,
+      "step": 5261
+    },
+    {
+      "epoch": 0.8641635702995094,
+      "grad_norm": 0.4622292620162486,
+      "learning_rate": 8.411609645039415e-06,
+      "loss": 0.4848,
+      "step": 5262
+    },
+    {
+      "epoch": 0.8643277975078522,
+      "grad_norm": 0.34047389699532343,
+      "learning_rate": 8.411377773898267e-06,
+      "loss": 0.5357,
+      "step": 5263
+    },
+    {
+      "epoch": 0.8644920247161949,
+      "grad_norm": 0.33676241125712014,
+      "learning_rate": 8.411145860275706e-06,
+      "loss": 0.4958,
+      "step": 5264
+    },
+    {
+      "epoch": 0.8646562519245375,
+      "grad_norm": 0.31135847431054403,
+      "learning_rate": 8.410913904174252e-06,
+      "loss": 0.5062,
+      "step": 5265
+    },
+    {
+      "epoch": 0.8648204791328803,
+      "grad_norm": 0.39107972410690034,
+      "learning_rate": 8.410681905596426e-06,
+      "loss": 0.5082,
+      "step": 5266
+    },
+    {
+      "epoch": 0.864984706341223,
+      "grad_norm": 0.3025022980375748,
+      "learning_rate": 8.410449864544748e-06,
+      "loss": 0.5117,
+      "step": 5267
+    },
+    {
+      "epoch": 0.8651489335495658,
+      "grad_norm": 0.27122718992549955,
+      "learning_rate": 8.410217781021736e-06,
+      "loss": 0.5079,
+      "step": 5268
+    },
+    {
+      "epoch": 0.8653131607579085,
+      "grad_norm": 0.33819644861136894,
+      "learning_rate": 8.409985655029912e-06,
+      "loss": 0.5044,
+      "step": 5269
+    },
+    {
+      "epoch": 0.8654773879662513,
+      "grad_norm": 0.3125457178886905,
+      "learning_rate": 8.409753486571795e-06,
+      "loss": 0.4664,
+      "step": 5270
+    },
+    {
+      "epoch": 0.865641615174594,
+      "grad_norm": 0.3283184967907039,
+      "learning_rate": 8.409521275649912e-06,
+      "loss": 0.5255,
+      "step": 5271
+    },
+    {
+      "epoch": 0.8658058423829368,
+      "grad_norm": 0.35440138371082225,
+      "learning_rate": 8.40928902226678e-06,
+      "loss": 0.5087,
+      "step": 5272
+    },
+    {
+      "epoch": 0.8659700695912795,
+      "grad_norm": 1.3917466044548144,
+      "learning_rate": 8.409056726424922e-06,
+      "loss": 0.4998,
+      "step": 5273
+    },
+    {
+      "epoch": 0.8661342967996223,
+      "grad_norm": 0.2904222163965924,
+      "learning_rate": 8.408824388126863e-06,
+      "loss": 0.4879,
+      "step": 5274
+    },
+    {
+      "epoch": 0.866298524007965,
+      "grad_norm": 0.3338819728127975,
+      "learning_rate": 8.408592007375125e-06,
+      "loss": 0.5197,
+      "step": 5275
+    },
+    {
+      "epoch": 0.8664627512163078,
+      "grad_norm": 0.3069337778776626,
+      "learning_rate": 8.408359584172234e-06,
+      "loss": 0.4761,
+      "step": 5276
+    },
+    {
+      "epoch": 0.8666269784246505,
+      "grad_norm": 0.3198664615484879,
+      "learning_rate": 8.40812711852071e-06,
+      "loss": 0.5159,
+      "step": 5277
+    },
+    {
+      "epoch": 0.8667912056329933,
+      "grad_norm": 0.2928415185108605,
+      "learning_rate": 8.407894610423082e-06,
+      "loss": 0.4881,
+      "step": 5278
+    },
+    {
+      "epoch": 0.866955432841336,
+      "grad_norm": 0.2942661306167567,
+      "learning_rate": 8.407662059881872e-06,
+      "loss": 0.5326,
+      "step": 5279
+    },
+    {
+      "epoch": 0.8671196600496788,
+      "grad_norm": 0.33386555344083646,
+      "learning_rate": 8.407429466899608e-06,
+      "loss": 0.5147,
+      "step": 5280
+    },
+    {
+      "epoch": 0.8672838872580215,
+      "grad_norm": 0.35711923571285464,
+      "learning_rate": 8.407196831478817e-06,
+      "loss": 0.5014,
+      "step": 5281
+    },
+    {
+      "epoch": 0.8674481144663642,
+      "grad_norm": 0.3719107496609669,
+      "learning_rate": 8.406964153622023e-06,
+      "loss": 0.5145,
+      "step": 5282
+    },
+    {
+      "epoch": 0.8676123416747069,
+      "grad_norm": 0.336945720581192,
+      "learning_rate": 8.406731433331756e-06,
+      "loss": 0.5096,
+      "step": 5283
+    },
+    {
+      "epoch": 0.8677765688830497,
+      "grad_norm": 0.3087115536368896,
+      "learning_rate": 8.40649867061054e-06,
+      "loss": 0.5227,
+      "step": 5284
+    },
+    {
+      "epoch": 0.8679407960913924,
+      "grad_norm": 0.47804182837248027,
+      "learning_rate": 8.406265865460905e-06,
+      "loss": 0.5085,
+      "step": 5285
+    },
+    {
+      "epoch": 0.8681050232997352,
+      "grad_norm": 0.3378849373597841,
+      "learning_rate": 8.406033017885381e-06,
+      "loss": 0.4883,
+      "step": 5286
+    },
+    {
+      "epoch": 0.8682692505080779,
+      "grad_norm": 0.4126831812504061,
+      "learning_rate": 8.405800127886493e-06,
+      "loss": 0.5181,
+      "step": 5287
+    },
+    {
+      "epoch": 0.8684334777164207,
+      "grad_norm": 0.27899762274396317,
+      "learning_rate": 8.405567195466775e-06,
+      "loss": 0.5227,
+      "step": 5288
+    },
+    {
+      "epoch": 0.8685977049247634,
+      "grad_norm": 0.33191781981524426,
+      "learning_rate": 8.405334220628754e-06,
+      "loss": 0.4971,
+      "step": 5289
+    },
+    {
+      "epoch": 0.8687619321331062,
+      "grad_norm": 0.29662629916763467,
+      "learning_rate": 8.405101203374962e-06,
+      "loss": 0.5116,
+      "step": 5290
+    },
+    {
+      "epoch": 0.8689261593414489,
+      "grad_norm": 0.32905380917085053,
+      "learning_rate": 8.404868143707927e-06,
+      "loss": 0.5111,
+      "step": 5291
+    },
+    {
+      "epoch": 0.8690903865497916,
+      "grad_norm": 0.3228454402336175,
+      "learning_rate": 8.404635041630184e-06,
+      "loss": 0.533,
+      "step": 5292
+    },
+    {
+      "epoch": 0.8692546137581344,
+      "grad_norm": 0.3159360456674922,
+      "learning_rate": 8.404401897144262e-06,
+      "loss": 0.5142,
+      "step": 5293
+    },
+    {
+      "epoch": 0.8694188409664771,
+      "grad_norm": 0.31946115452138973,
+      "learning_rate": 8.404168710252692e-06,
+      "loss": 0.5095,
+      "step": 5294
+    },
+    {
+      "epoch": 0.8695830681748199,
+      "grad_norm": 0.7962735316090845,
+      "learning_rate": 8.403935480958011e-06,
+      "loss": 0.5373,
+      "step": 5295
+    },
+    {
+      "epoch": 0.8697472953831626,
+      "grad_norm": 1.2910079032447723,
+      "learning_rate": 8.40370220926275e-06,
+      "loss": 0.4944,
+      "step": 5296
+    },
+    {
+      "epoch": 0.8699115225915054,
+      "grad_norm": 0.3078680638961182,
+      "learning_rate": 8.40346889516944e-06,
+      "loss": 0.4875,
+      "step": 5297
+    },
+    {
+      "epoch": 0.8700757497998481,
+      "grad_norm": 0.35891347383172,
+      "learning_rate": 8.40323553868062e-06,
+      "loss": 0.5185,
+      "step": 5298
+    },
+    {
+      "epoch": 0.8702399770081908,
+      "grad_norm": 0.5653737358384439,
+      "learning_rate": 8.40300213979882e-06,
+      "loss": 0.51,
+      "step": 5299
+    },
+    {
+      "epoch": 0.8704042042165335,
+      "grad_norm": 0.32168689984896887,
+      "learning_rate": 8.402768698526577e-06,
+      "loss": 0.5194,
+      "step": 5300
+    },
+    {
+      "epoch": 0.8705684314248763,
+      "grad_norm": 0.3381898859178391,
+      "learning_rate": 8.402535214866426e-06,
+      "loss": 0.5038,
+      "step": 5301
+    },
+    {
+      "epoch": 0.870732658633219,
+      "grad_norm": 0.34810336533021025,
+      "learning_rate": 8.402301688820903e-06,
+      "loss": 0.5198,
+      "step": 5302
+    },
+    {
+      "epoch": 0.8708968858415618,
+      "grad_norm": 0.3556606225523487,
+      "learning_rate": 8.402068120392545e-06,
+      "loss": 0.51,
+      "step": 5303
+    },
+    {
+      "epoch": 0.8710611130499045,
+      "grad_norm": 0.2772688917845348,
+      "learning_rate": 8.401834509583889e-06,
+      "loss": 0.5163,
+      "step": 5304
+    },
+    {
+      "epoch": 0.8712253402582473,
+      "grad_norm": 0.3070635849318607,
+      "learning_rate": 8.40160085639747e-06,
+      "loss": 0.5107,
+      "step": 5305
+    },
+    {
+      "epoch": 0.87138956746659,
+      "grad_norm": 0.35390726184331905,
+      "learning_rate": 8.401367160835826e-06,
+      "loss": 0.5043,
+      "step": 5306
+    },
+    {
+      "epoch": 0.8715537946749328,
+      "grad_norm": 0.33187000412444934,
+      "learning_rate": 8.401133422901497e-06,
+      "loss": 0.5116,
+      "step": 5307
+    },
+    {
+      "epoch": 0.8717180218832755,
+      "grad_norm": 0.40659711524377584,
+      "learning_rate": 8.400899642597022e-06,
+      "loss": 0.5036,
+      "step": 5308
+    },
+    {
+      "epoch": 0.8718822490916183,
+      "grad_norm": 0.3851914077124121,
+      "learning_rate": 8.400665819924938e-06,
+      "loss": 0.5056,
+      "step": 5309
+    },
+    {
+      "epoch": 0.872046476299961,
+      "grad_norm": 0.32665644167454533,
+      "learning_rate": 8.400431954887785e-06,
+      "loss": 0.513,
+      "step": 5310
+    },
+    {
+      "epoch": 0.8722107035083038,
+      "grad_norm": 0.41519115858739336,
+      "learning_rate": 8.400198047488105e-06,
+      "loss": 0.5219,
+      "step": 5311
+    },
+    {
+      "epoch": 0.8723749307166465,
+      "grad_norm": 0.3552816056869143,
+      "learning_rate": 8.399964097728436e-06,
+      "loss": 0.516,
+      "step": 5312
+    },
+    {
+      "epoch": 0.8725391579249893,
+      "grad_norm": 0.2861691798039096,
+      "learning_rate": 8.39973010561132e-06,
+      "loss": 0.4861,
+      "step": 5313
+    },
+    {
+      "epoch": 0.872703385133332,
+      "grad_norm": 0.32524768948820126,
+      "learning_rate": 8.399496071139298e-06,
+      "loss": 0.4899,
+      "step": 5314
+    },
+    {
+      "epoch": 0.8728676123416748,
+      "grad_norm": 0.34057732140952096,
+      "learning_rate": 8.39926199431491e-06,
+      "loss": 0.508,
+      "step": 5315
+    },
+    {
+      "epoch": 0.8730318395500174,
+      "grad_norm": 0.4133044624728016,
+      "learning_rate": 8.399027875140703e-06,
+      "loss": 0.4852,
+      "step": 5316
+    },
+    {
+      "epoch": 0.8731960667583601,
+      "grad_norm": 0.3126250586763464,
+      "learning_rate": 8.398793713619218e-06,
+      "loss": 0.4831,
+      "step": 5317
+    },
+    {
+      "epoch": 0.8733602939667029,
+      "grad_norm": 0.35913370774447195,
+      "learning_rate": 8.398559509752995e-06,
+      "loss": 0.5068,
+      "step": 5318
+    },
+    {
+      "epoch": 0.8735245211750456,
+      "grad_norm": 0.3425673312783436,
+      "learning_rate": 8.398325263544582e-06,
+      "loss": 0.5124,
+      "step": 5319
+    },
+    {
+      "epoch": 0.8736887483833884,
+      "grad_norm": 0.31258714176169194,
+      "learning_rate": 8.39809097499652e-06,
+      "loss": 0.5134,
+      "step": 5320
+    },
+    {
+      "epoch": 0.8738529755917311,
+      "grad_norm": 0.30534161830143747,
+      "learning_rate": 8.397856644111356e-06,
+      "loss": 0.5052,
+      "step": 5321
+    },
+    {
+      "epoch": 0.8740172028000739,
+      "grad_norm": 0.3471453883343114,
+      "learning_rate": 8.397622270891632e-06,
+      "loss": 0.475,
+      "step": 5322
+    },
+    {
+      "epoch": 0.8741814300084166,
+      "grad_norm": 0.30377985143302766,
+      "learning_rate": 8.397387855339896e-06,
+      "loss": 0.4997,
+      "step": 5323
+    },
+    {
+      "epoch": 0.8743456572167594,
+      "grad_norm": 0.32085464444851813,
+      "learning_rate": 8.397153397458694e-06,
+      "loss": 0.5079,
+      "step": 5324
+    },
+    {
+      "epoch": 0.8745098844251021,
+      "grad_norm": 0.3914308128149654,
+      "learning_rate": 8.396918897250571e-06,
+      "loss": 0.5051,
+      "step": 5325
+    },
+    {
+      "epoch": 0.8746741116334449,
+      "grad_norm": 0.3968388154681836,
+      "learning_rate": 8.396684354718076e-06,
+      "loss": 0.4986,
+      "step": 5326
+    },
+    {
+      "epoch": 0.8748383388417876,
+      "grad_norm": 0.2997877334792993,
+      "learning_rate": 8.396449769863754e-06,
+      "loss": 0.5121,
+      "step": 5327
+    },
+    {
+      "epoch": 0.8750025660501304,
+      "grad_norm": 0.4294864886076223,
+      "learning_rate": 8.396215142690154e-06,
+      "loss": 0.5012,
+      "step": 5328
+    },
+    {
+      "epoch": 0.8751667932584731,
+      "grad_norm": 0.31195660001293835,
+      "learning_rate": 8.395980473199826e-06,
+      "loss": 0.4977,
+      "step": 5329
+    },
+    {
+      "epoch": 0.8753310204668159,
+      "grad_norm": 0.3503078869614596,
+      "learning_rate": 8.395745761395314e-06,
+      "loss": 0.515,
+      "step": 5330
+    },
+    {
+      "epoch": 0.8754952476751586,
+      "grad_norm": 0.35642303207734,
+      "learning_rate": 8.395511007279172e-06,
+      "loss": 0.4936,
+      "step": 5331
+    },
+    {
+      "epoch": 0.8756594748835014,
+      "grad_norm": 0.37437379291643125,
+      "learning_rate": 8.395276210853946e-06,
+      "loss": 0.5064,
+      "step": 5332
+    },
+    {
+      "epoch": 0.875823702091844,
+      "grad_norm": 0.32548744308565203,
+      "learning_rate": 8.39504137212219e-06,
+      "loss": 0.5251,
+      "step": 5333
+    },
+    {
+      "epoch": 0.8759879293001868,
+      "grad_norm": 0.3368536214240922,
+      "learning_rate": 8.394806491086453e-06,
+      "loss": 0.4841,
+      "step": 5334
+    },
+    {
+      "epoch": 0.8761521565085295,
+      "grad_norm": 0.32379612932847374,
+      "learning_rate": 8.394571567749283e-06,
+      "loss": 0.519,
+      "step": 5335
+    },
+    {
+      "epoch": 0.8763163837168723,
+      "grad_norm": 0.254907543345491,
+      "learning_rate": 8.394336602113235e-06,
+      "loss": 0.4934,
+      "step": 5336
+    },
+    {
+      "epoch": 0.876480610925215,
+      "grad_norm": 0.292578050848311,
+      "learning_rate": 8.39410159418086e-06,
+      "loss": 0.5038,
+      "step": 5337
+    },
+    {
+      "epoch": 0.8766448381335578,
+      "grad_norm": 0.33068895234319823,
+      "learning_rate": 8.393866543954713e-06,
+      "loss": 0.4894,
+      "step": 5338
+    },
+    {
+      "epoch": 0.8768090653419005,
+      "grad_norm": 0.3022738707416738,
+      "learning_rate": 8.393631451437341e-06,
+      "loss": 0.495,
+      "step": 5339
+    },
+    {
+      "epoch": 0.8769732925502433,
+      "grad_norm": 0.3309924116679411,
+      "learning_rate": 8.393396316631301e-06,
+      "loss": 0.4907,
+      "step": 5340
+    },
+    {
+      "epoch": 0.877137519758586,
+      "grad_norm": 0.3072778428978678,
+      "learning_rate": 8.393161139539147e-06,
+      "loss": 0.5083,
+      "step": 5341
+    },
+    {
+      "epoch": 0.8773017469669288,
+      "grad_norm": 0.4056601024135022,
+      "learning_rate": 8.392925920163433e-06,
+      "loss": 0.494,
+      "step": 5342
+    },
+    {
+      "epoch": 0.8774659741752715,
+      "grad_norm": 0.3292865631147971,
+      "learning_rate": 8.392690658506713e-06,
+      "loss": 0.477,
+      "step": 5343
+    },
+    {
+      "epoch": 0.8776302013836143,
+      "grad_norm": 0.3173888394810557,
+      "learning_rate": 8.392455354571542e-06,
+      "loss": 0.5042,
+      "step": 5344
+    },
+    {
+      "epoch": 0.877794428591957,
+      "grad_norm": 0.41650203746899644,
+      "learning_rate": 8.392220008360478e-06,
+      "loss": 0.5128,
+      "step": 5345
+    },
+    {
+      "epoch": 0.8779586558002997,
+      "grad_norm": 0.3128633850645757,
+      "learning_rate": 8.391984619876073e-06,
+      "loss": 0.5117,
+      "step": 5346
+    },
+    {
+      "epoch": 0.8781228830086425,
+      "grad_norm": 0.3739198066157396,
+      "learning_rate": 8.391749189120889e-06,
+      "loss": 0.4983,
+      "step": 5347
+    },
+    {
+      "epoch": 0.8782871102169852,
+      "grad_norm": 0.3270565163589913,
+      "learning_rate": 8.391513716097476e-06,
+      "loss": 0.5094,
+      "step": 5348
+    },
+    {
+      "epoch": 0.878451337425328,
+      "grad_norm": 0.36481461717522407,
+      "learning_rate": 8.391278200808398e-06,
+      "loss": 0.4884,
+      "step": 5349
+    },
+    {
+      "epoch": 0.8786155646336706,
+      "grad_norm": 0.3380444447356424,
+      "learning_rate": 8.39104264325621e-06,
+      "loss": 0.4973,
+      "step": 5350
+    },
+    {
+      "epoch": 0.8787797918420134,
+      "grad_norm": 0.3226705104713618,
+      "learning_rate": 8.390807043443468e-06,
+      "loss": 0.5139,
+      "step": 5351
+    },
+    {
+      "epoch": 0.8789440190503561,
+      "grad_norm": 0.3562909994144863,
+      "learning_rate": 8.390571401372737e-06,
+      "loss": 0.5125,
+      "step": 5352
+    },
+    {
+      "epoch": 0.8791082462586989,
+      "grad_norm": 0.3465108100828412,
+      "learning_rate": 8.39033571704657e-06,
+      "loss": 0.5038,
+      "step": 5353
+    },
+    {
+      "epoch": 0.8792724734670416,
+      "grad_norm": 0.46935542963470755,
+      "learning_rate": 8.390099990467531e-06,
+      "loss": 0.5063,
+      "step": 5354
+    },
+    {
+      "epoch": 0.8794367006753844,
+      "grad_norm": 0.34137317758615526,
+      "learning_rate": 8.389864221638179e-06,
+      "loss": 0.514,
+      "step": 5355
+    },
+    {
+      "epoch": 0.8796009278837271,
+      "grad_norm": 0.29825938926351886,
+      "learning_rate": 8.389628410561074e-06,
+      "loss": 0.5368,
+      "step": 5356
+    },
+    {
+      "epoch": 0.8797651550920699,
+      "grad_norm": 0.32769587710392506,
+      "learning_rate": 8.389392557238777e-06,
+      "loss": 0.5395,
+      "step": 5357
+    },
+    {
+      "epoch": 0.8799293823004126,
+      "grad_norm": 0.3103698388540546,
+      "learning_rate": 8.389156661673851e-06,
+      "loss": 0.4951,
+      "step": 5358
+    },
+    {
+      "epoch": 0.8800936095087554,
+      "grad_norm": 0.31391936442953045,
+      "learning_rate": 8.388920723868858e-06,
+      "loss": 0.5139,
+      "step": 5359
+    },
+    {
+      "epoch": 0.8802578367170981,
+      "grad_norm": 0.30702069935676596,
+      "learning_rate": 8.388684743826358e-06,
+      "loss": 0.4914,
+      "step": 5360
+    },
+    {
+      "epoch": 0.8804220639254409,
+      "grad_norm": 0.287494636924883,
+      "learning_rate": 8.388448721548916e-06,
+      "loss": 0.5036,
+      "step": 5361
+    },
+    {
+      "epoch": 0.8805862911337836,
+      "grad_norm": 0.3079894511670955,
+      "learning_rate": 8.388212657039097e-06,
+      "loss": 0.4936,
+      "step": 5362
+    },
+    {
+      "epoch": 0.8807505183421264,
+      "grad_norm": 0.3360880269155661,
+      "learning_rate": 8.387976550299462e-06,
+      "loss": 0.5035,
+      "step": 5363
+    },
+    {
+      "epoch": 0.8809147455504691,
+      "grad_norm": 0.5102037605942567,
+      "learning_rate": 8.387740401332574e-06,
+      "loss": 0.5013,
+      "step": 5364
+    },
+    {
+      "epoch": 0.8810789727588119,
+      "grad_norm": 0.3313606109020479,
+      "learning_rate": 8.387504210141003e-06,
+      "loss": 0.4879,
+      "step": 5365
+    },
+    {
+      "epoch": 0.8812431999671546,
+      "grad_norm": 0.3999436482907233,
+      "learning_rate": 8.387267976727312e-06,
+      "loss": 0.5208,
+      "step": 5366
+    },
+    {
+      "epoch": 0.8814074271754972,
+      "grad_norm": 0.3326458033025108,
+      "learning_rate": 8.387031701094066e-06,
+      "loss": 0.5074,
+      "step": 5367
+    },
+    {
+      "epoch": 0.88157165438384,
+      "grad_norm": 0.34094614319822397,
+      "learning_rate": 8.386795383243828e-06,
+      "loss": 0.5159,
+      "step": 5368
+    },
+    {
+      "epoch": 0.8817358815921827,
+      "grad_norm": 0.3706442446081152,
+      "learning_rate": 8.386559023179172e-06,
+      "loss": 0.4928,
+      "step": 5369
+    },
+    {
+      "epoch": 0.8819001088005255,
+      "grad_norm": 0.35070846954591484,
+      "learning_rate": 8.38632262090266e-06,
+      "loss": 0.5259,
+      "step": 5370
+    },
+    {
+      "epoch": 0.8820643360088682,
+      "grad_norm": 0.3703793978322178,
+      "learning_rate": 8.386086176416859e-06,
+      "loss": 0.521,
+      "step": 5371
+    },
+    {
+      "epoch": 0.882228563217211,
+      "grad_norm": 0.3957430663808368,
+      "learning_rate": 8.38584968972434e-06,
+      "loss": 0.5108,
+      "step": 5372
+    },
+    {
+      "epoch": 0.8823927904255537,
+      "grad_norm": 0.3077741513865519,
+      "learning_rate": 8.385613160827672e-06,
+      "loss": 0.5007,
+      "step": 5373
+    },
+    {
+      "epoch": 0.8825570176338965,
+      "grad_norm": 0.32405883564522314,
+      "learning_rate": 8.385376589729419e-06,
+      "loss": 0.4945,
+      "step": 5374
+    },
+    {
+      "epoch": 0.8827212448422392,
+      "grad_norm": 0.32194303637237365,
+      "learning_rate": 8.385139976432155e-06,
+      "loss": 0.5216,
+      "step": 5375
+    },
+    {
+      "epoch": 0.882885472050582,
+      "grad_norm": 0.34449583253230837,
+      "learning_rate": 8.384903320938449e-06,
+      "loss": 0.5024,
+      "step": 5376
+    },
+    {
+      "epoch": 0.8830496992589247,
+      "grad_norm": 0.3035329723939267,
+      "learning_rate": 8.38466662325087e-06,
+      "loss": 0.5021,
+      "step": 5377
+    },
+    {
+      "epoch": 0.8832139264672675,
+      "grad_norm": 0.34319045438113116,
+      "learning_rate": 8.384429883371989e-06,
+      "loss": 0.5162,
+      "step": 5378
+    },
+    {
+      "epoch": 0.8833781536756102,
+      "grad_norm": 0.36034984090698774,
+      "learning_rate": 8.384193101304377e-06,
+      "loss": 0.5028,
+      "step": 5379
+    },
+    {
+      "epoch": 0.883542380883953,
+      "grad_norm": 0.33601207342080214,
+      "learning_rate": 8.383956277050609e-06,
+      "loss": 0.4913,
+      "step": 5380
+    },
+    {
+      "epoch": 0.8837066080922957,
+      "grad_norm": 0.394729287186722,
+      "learning_rate": 8.383719410613254e-06,
+      "loss": 0.5089,
+      "step": 5381
+    },
+    {
+      "epoch": 0.8838708353006385,
+      "grad_norm": 0.33148220436620357,
+      "learning_rate": 8.383482501994884e-06,
+      "loss": 0.4929,
+      "step": 5382
+    },
+    {
+      "epoch": 0.8840350625089812,
+      "grad_norm": 0.4144935425174578,
+      "learning_rate": 8.383245551198074e-06,
+      "loss": 0.5209,
+      "step": 5383
+    },
+    {
+      "epoch": 0.8841992897173239,
+      "grad_norm": 0.29606041865345,
+      "learning_rate": 8.383008558225395e-06,
+      "loss": 0.4981,
+      "step": 5384
+    },
+    {
+      "epoch": 0.8843635169256666,
+      "grad_norm": 0.31134244297032126,
+      "learning_rate": 8.382771523079424e-06,
+      "loss": 0.4992,
+      "step": 5385
+    },
+    {
+      "epoch": 0.8845277441340094,
+      "grad_norm": 0.3370958117903781,
+      "learning_rate": 8.382534445762735e-06,
+      "loss": 0.4937,
+      "step": 5386
+    },
+    {
+      "epoch": 0.8846919713423521,
+      "grad_norm": 0.3001698439254947,
+      "learning_rate": 8.3822973262779e-06,
+      "loss": 0.5329,
+      "step": 5387
+    },
+    {
+      "epoch": 0.8848561985506949,
+      "grad_norm": 0.30084477565026285,
+      "learning_rate": 8.382060164627499e-06,
+      "loss": 0.5059,
+      "step": 5388
+    },
+    {
+      "epoch": 0.8850204257590376,
+      "grad_norm": 0.3346426570065145,
+      "learning_rate": 8.381822960814102e-06,
+      "loss": 0.5013,
+      "step": 5389
+    },
+    {
+      "epoch": 0.8851846529673804,
+      "grad_norm": 0.29458072214294884,
+      "learning_rate": 8.381585714840291e-06,
+      "loss": 0.505,
+      "step": 5390
+    },
+    {
+      "epoch": 0.8853488801757231,
+      "grad_norm": 0.3378818988362336,
+      "learning_rate": 8.38134842670864e-06,
+      "loss": 0.5187,
+      "step": 5391
+    },
+    {
+      "epoch": 0.8855131073840659,
+      "grad_norm": 0.35910179479581733,
+      "learning_rate": 8.381111096421725e-06,
+      "loss": 0.5095,
+      "step": 5392
+    },
+    {
+      "epoch": 0.8856773345924086,
+      "grad_norm": 0.31216769263690824,
+      "learning_rate": 8.380873723982126e-06,
+      "loss": 0.5111,
+      "step": 5393
+    },
+    {
+      "epoch": 0.8858415618007514,
+      "grad_norm": 0.4289924367030972,
+      "learning_rate": 8.380636309392419e-06,
+      "loss": 0.4956,
+      "step": 5394
+    },
+    {
+      "epoch": 0.8860057890090941,
+      "grad_norm": 0.4956517822724945,
+      "learning_rate": 8.380398852655184e-06,
+      "loss": 0.4938,
+      "step": 5395
+    },
+    {
+      "epoch": 0.8861700162174369,
+      "grad_norm": 0.28831824779869475,
+      "learning_rate": 8.380161353773e-06,
+      "loss": 0.5125,
+      "step": 5396
+    },
+    {
+      "epoch": 0.8863342434257796,
+      "grad_norm": 0.3097895844490996,
+      "learning_rate": 8.379923812748447e-06,
+      "loss": 0.513,
+      "step": 5397
+    },
+    {
+      "epoch": 0.8864984706341223,
+      "grad_norm": 0.3296849530356679,
+      "learning_rate": 8.379686229584103e-06,
+      "loss": 0.4992,
+      "step": 5398
+    },
+    {
+      "epoch": 0.8866626978424651,
+      "grad_norm": 0.40568229779401993,
+      "learning_rate": 8.37944860428255e-06,
+      "loss": 0.5006,
+      "step": 5399
+    },
+    {
+      "epoch": 0.8868269250508078,
+      "grad_norm": 0.34325966725882706,
+      "learning_rate": 8.379210936846368e-06,
+      "loss": 0.4764,
+      "step": 5400
+    },
+    {
+      "epoch": 0.8869911522591505,
+      "grad_norm": 0.4758915504276449,
+      "learning_rate": 8.378973227278139e-06,
+      "loss": 0.5055,
+      "step": 5401
+    },
+    {
+      "epoch": 0.8871553794674932,
+      "grad_norm": 0.297381855982727,
+      "learning_rate": 8.378735475580444e-06,
+      "loss": 0.498,
+      "step": 5402
+    },
+    {
+      "epoch": 0.887319606675836,
+      "grad_norm": 0.4933418017370094,
+      "learning_rate": 8.378497681755865e-06,
+      "loss": 0.4995,
+      "step": 5403
+    },
+    {
+      "epoch": 0.8874838338841787,
+      "grad_norm": 0.3983723166968519,
+      "learning_rate": 8.378259845806986e-06,
+      "loss": 0.5365,
+      "step": 5404
+    },
+    {
+      "epoch": 0.8876480610925215,
+      "grad_norm": 0.3143111854125081,
+      "learning_rate": 8.378021967736388e-06,
+      "loss": 0.5197,
+      "step": 5405
+    },
+    {
+      "epoch": 0.8878122883008642,
+      "grad_norm": 0.3801638714521817,
+      "learning_rate": 8.377784047546657e-06,
+      "loss": 0.4944,
+      "step": 5406
+    },
+    {
+      "epoch": 0.887976515509207,
+      "grad_norm": 0.34407817181065525,
+      "learning_rate": 8.377546085240376e-06,
+      "loss": 0.4994,
+      "step": 5407
+    },
+    {
+      "epoch": 0.8881407427175497,
+      "grad_norm": 0.3133043465281556,
+      "learning_rate": 8.37730808082013e-06,
+      "loss": 0.4759,
+      "step": 5408
+    },
+    {
+      "epoch": 0.8883049699258925,
+      "grad_norm": 0.2965789041034004,
+      "learning_rate": 8.377070034288505e-06,
+      "loss": 0.5046,
+      "step": 5409
+    },
+    {
+      "epoch": 0.8884691971342352,
+      "grad_norm": 0.326618198478491,
+      "learning_rate": 8.376831945648081e-06,
+      "loss": 0.5124,
+      "step": 5410
+    },
+    {
+      "epoch": 0.888633424342578,
+      "grad_norm": 0.3053859757388855,
+      "learning_rate": 8.37659381490145e-06,
+      "loss": 0.4893,
+      "step": 5411
+    },
+    {
+      "epoch": 0.8887976515509207,
+      "grad_norm": 0.2954348959691266,
+      "learning_rate": 8.376355642051196e-06,
+      "loss": 0.5098,
+      "step": 5412
+    },
+    {
+      "epoch": 0.8889618787592635,
+      "grad_norm": 0.3135215141770848,
+      "learning_rate": 8.376117427099907e-06,
+      "loss": 0.4979,
+      "step": 5413
+    },
+    {
+      "epoch": 0.8891261059676062,
+      "grad_norm": 0.3271102608994896,
+      "learning_rate": 8.375879170050167e-06,
+      "loss": 0.5178,
+      "step": 5414
+    },
+    {
+      "epoch": 0.889290333175949,
+      "grad_norm": 0.38229922813402745,
+      "learning_rate": 8.375640870904568e-06,
+      "loss": 0.5232,
+      "step": 5415
+    },
+    {
+      "epoch": 0.8894545603842917,
+      "grad_norm": 0.3797870401260684,
+      "learning_rate": 8.375402529665694e-06,
+      "loss": 0.4986,
+      "step": 5416
+    },
+    {
+      "epoch": 0.8896187875926345,
+      "grad_norm": 0.31200316783816984,
+      "learning_rate": 8.375164146336137e-06,
+      "loss": 0.5241,
+      "step": 5417
+    },
+    {
+      "epoch": 0.8897830148009771,
+      "grad_norm": 0.3064704900631936,
+      "learning_rate": 8.374925720918485e-06,
+      "loss": 0.4955,
+      "step": 5418
+    },
+    {
+      "epoch": 0.8899472420093198,
+      "grad_norm": 0.3058769827785828,
+      "learning_rate": 8.374687253415326e-06,
+      "loss": 0.4883,
+      "step": 5419
+    },
+    {
+      "epoch": 0.8901114692176626,
+      "grad_norm": 0.2843411368389142,
+      "learning_rate": 8.374448743829252e-06,
+      "loss": 0.5217,
+      "step": 5420
+    },
+    {
+      "epoch": 0.8902756964260053,
+      "grad_norm": 0.3728611557769457,
+      "learning_rate": 8.37421019216285e-06,
+      "loss": 0.4938,
+      "step": 5421
+    },
+    {
+      "epoch": 0.8904399236343481,
+      "grad_norm": 0.3011377477606396,
+      "learning_rate": 8.373971598418717e-06,
+      "loss": 0.5077,
+      "step": 5422
+    },
+    {
+      "epoch": 0.8906041508426908,
+      "grad_norm": 0.36206134781906424,
+      "learning_rate": 8.373732962599441e-06,
+      "loss": 0.494,
+      "step": 5423
+    },
+    {
+      "epoch": 0.8907683780510336,
+      "grad_norm": 0.3908084739505783,
+      "learning_rate": 8.373494284707613e-06,
+      "loss": 0.5201,
+      "step": 5424
+    },
+    {
+      "epoch": 0.8909326052593763,
+      "grad_norm": 0.2749425323407986,
+      "learning_rate": 8.373255564745824e-06,
+      "loss": 0.4962,
+      "step": 5425
+    },
+    {
+      "epoch": 0.8910968324677191,
+      "grad_norm": 0.3303930882629094,
+      "learning_rate": 8.373016802716673e-06,
+      "loss": 0.5098,
+      "step": 5426
+    },
+    {
+      "epoch": 0.8912610596760618,
+      "grad_norm": 0.2998241441933591,
+      "learning_rate": 8.372777998622745e-06,
+      "loss": 0.4869,
+      "step": 5427
+    },
+    {
+      "epoch": 0.8914252868844046,
+      "grad_norm": 0.29451950284439987,
+      "learning_rate": 8.37253915246664e-06,
+      "loss": 0.5264,
+      "step": 5428
+    },
+    {
+      "epoch": 0.8915895140927473,
+      "grad_norm": 0.3063418491635172,
+      "learning_rate": 8.37230026425095e-06,
+      "loss": 0.5038,
+      "step": 5429
+    },
+    {
+      "epoch": 0.8917537413010901,
+      "grad_norm": 0.2996944732030016,
+      "learning_rate": 8.372061333978266e-06,
+      "loss": 0.5014,
+      "step": 5430
+    },
+    {
+      "epoch": 0.8919179685094328,
+      "grad_norm": 0.3424116018767127,
+      "learning_rate": 8.37182236165119e-06,
+      "loss": 0.5089,
+      "step": 5431
+    },
+    {
+      "epoch": 0.8920821957177756,
+      "grad_norm": 0.29079573719915974,
+      "learning_rate": 8.371583347272314e-06,
+      "loss": 0.5219,
+      "step": 5432
+    },
+    {
+      "epoch": 0.8922464229261183,
+      "grad_norm": 0.3304927425490326,
+      "learning_rate": 8.37134429084423e-06,
+      "loss": 0.5338,
+      "step": 5433
+    },
+    {
+      "epoch": 0.8924106501344611,
+      "grad_norm": 0.4902664749685111,
+      "learning_rate": 8.371105192369541e-06,
+      "loss": 0.5078,
+      "step": 5434
+    },
+    {
+      "epoch": 0.8925748773428037,
+      "grad_norm": 0.39179044375461103,
+      "learning_rate": 8.37086605185084e-06,
+      "loss": 0.514,
+      "step": 5435
+    },
+    {
+      "epoch": 0.8927391045511465,
+      "grad_norm": 0.3111137817750461,
+      "learning_rate": 8.370626869290725e-06,
+      "loss": 0.4722,
+      "step": 5436
+    },
+    {
+      "epoch": 0.8929033317594892,
+      "grad_norm": 0.30758120367970715,
+      "learning_rate": 8.370387644691796e-06,
+      "loss": 0.495,
+      "step": 5437
+    },
+    {
+      "epoch": 0.893067558967832,
+      "grad_norm": 0.3982671596775974,
+      "learning_rate": 8.370148378056647e-06,
+      "loss": 0.5242,
+      "step": 5438
+    },
+    {
+      "epoch": 0.8932317861761747,
+      "grad_norm": 0.27220767951031977,
+      "learning_rate": 8.369909069387879e-06,
+      "loss": 0.5019,
+      "step": 5439
+    },
+    {
+      "epoch": 0.8933960133845175,
+      "grad_norm": 0.3798596024471178,
+      "learning_rate": 8.369669718688093e-06,
+      "loss": 0.5175,
+      "step": 5440
+    },
+    {
+      "epoch": 0.8935602405928602,
+      "grad_norm": 0.46454027331217806,
+      "learning_rate": 8.369430325959884e-06,
+      "loss": 0.5056,
+      "step": 5441
+    },
+    {
+      "epoch": 0.893724467801203,
+      "grad_norm": 0.2694505235037501,
+      "learning_rate": 8.369190891205858e-06,
+      "loss": 0.5006,
+      "step": 5442
+    },
+    {
+      "epoch": 0.8938886950095457,
+      "grad_norm": 0.2990622026346461,
+      "learning_rate": 8.36895141442861e-06,
+      "loss": 0.5154,
+      "step": 5443
+    },
+    {
+      "epoch": 0.8940529222178885,
+      "grad_norm": 0.30742433459438284,
+      "learning_rate": 8.368711895630743e-06,
+      "loss": 0.5201,
+      "step": 5444
+    },
+    {
+      "epoch": 0.8942171494262312,
+      "grad_norm": 0.2877974512432656,
+      "learning_rate": 8.36847233481486e-06,
+      "loss": 0.4951,
+      "step": 5445
+    },
+    {
+      "epoch": 0.894381376634574,
+      "grad_norm": 0.3799216470015632,
+      "learning_rate": 8.368232731983559e-06,
+      "loss": 0.524,
+      "step": 5446
+    },
+    {
+      "epoch": 0.8945456038429167,
+      "grad_norm": 0.28985046386333174,
+      "learning_rate": 8.367993087139446e-06,
+      "loss": 0.4789,
+      "step": 5447
+    },
+    {
+      "epoch": 0.8947098310512595,
+      "grad_norm": 0.2770590893279945,
+      "learning_rate": 8.367753400285122e-06,
+      "loss": 0.5151,
+      "step": 5448
+    },
+    {
+      "epoch": 0.8948740582596022,
+      "grad_norm": 0.3125926454125177,
+      "learning_rate": 8.367513671423191e-06,
+      "loss": 0.5164,
+      "step": 5449
+    },
+    {
+      "epoch": 0.895038285467945,
+      "grad_norm": 0.2789030307629401,
+      "learning_rate": 8.367273900556256e-06,
+      "loss": 0.4886,
+      "step": 5450
+    },
+    {
+      "epoch": 0.8952025126762877,
+      "grad_norm": 0.3147777575769946,
+      "learning_rate": 8.367034087686924e-06,
+      "loss": 0.5288,
+      "step": 5451
+    },
+    {
+      "epoch": 0.8953667398846303,
+      "grad_norm": 0.3087500424646868,
+      "learning_rate": 8.366794232817795e-06,
+      "loss": 0.493,
+      "step": 5452
+    },
+    {
+      "epoch": 0.8955309670929731,
+      "grad_norm": 0.2880004076961026,
+      "learning_rate": 8.366554335951474e-06,
+      "loss": 0.4967,
+      "step": 5453
+    },
+    {
+      "epoch": 0.8956951943013158,
+      "grad_norm": 0.26914699822687915,
+      "learning_rate": 8.366314397090572e-06,
+      "loss": 0.4957,
+      "step": 5454
+    },
+    {
+      "epoch": 0.8958594215096586,
+      "grad_norm": 0.853277733033621,
+      "learning_rate": 8.36607441623769e-06,
+      "loss": 0.512,
+      "step": 5455
+    },
+    {
+      "epoch": 0.8960236487180013,
+      "grad_norm": 0.28789407556256835,
+      "learning_rate": 8.365834393395438e-06,
+      "loss": 0.5066,
+      "step": 5456
+    },
+    {
+      "epoch": 0.8961878759263441,
+      "grad_norm": 0.2684557559113861,
+      "learning_rate": 8.36559432856642e-06,
+      "loss": 0.5182,
+      "step": 5457
+    },
+    {
+      "epoch": 0.8963521031346868,
+      "grad_norm": 0.3644698809817585,
+      "learning_rate": 8.365354221753245e-06,
+      "loss": 0.5137,
+      "step": 5458
+    },
+    {
+      "epoch": 0.8965163303430296,
+      "grad_norm": 0.27994902706665303,
+      "learning_rate": 8.36511407295852e-06,
+      "loss": 0.4943,
+      "step": 5459
+    },
+    {
+      "epoch": 0.8966805575513723,
+      "grad_norm": 0.35579239769018195,
+      "learning_rate": 8.364873882184851e-06,
+      "loss": 0.4925,
+      "step": 5460
+    },
+    {
+      "epoch": 0.8968447847597151,
+      "grad_norm": 0.35041238641861094,
+      "learning_rate": 8.364633649434853e-06,
+      "loss": 0.5022,
+      "step": 5461
+    },
+    {
+      "epoch": 0.8970090119680578,
+      "grad_norm": 0.3580120860072093,
+      "learning_rate": 8.364393374711128e-06,
+      "loss": 0.5104,
+      "step": 5462
+    },
+    {
+      "epoch": 0.8971732391764006,
+      "grad_norm": 0.27954324217496795,
+      "learning_rate": 8.364153058016292e-06,
+      "loss": 0.5193,
+      "step": 5463
+    },
+    {
+      "epoch": 0.8973374663847433,
+      "grad_norm": 0.3101665856935753,
+      "learning_rate": 8.363912699352949e-06,
+      "loss": 0.5095,
+      "step": 5464
+    },
+    {
+      "epoch": 0.8975016935930861,
+      "grad_norm": 0.28512021161180856,
+      "learning_rate": 8.363672298723714e-06,
+      "loss": 0.4987,
+      "step": 5465
+    },
+    {
+      "epoch": 0.8976659208014288,
+      "grad_norm": 0.35211494571368435,
+      "learning_rate": 8.363431856131196e-06,
+      "loss": 0.4955,
+      "step": 5466
+    },
+    {
+      "epoch": 0.8978301480097716,
+      "grad_norm": 0.32744803108370507,
+      "learning_rate": 8.363191371578006e-06,
+      "loss": 0.5022,
+      "step": 5467
+    },
+    {
+      "epoch": 0.8979943752181143,
+      "grad_norm": 0.3249957296286912,
+      "learning_rate": 8.36295084506676e-06,
+      "loss": 0.4945,
+      "step": 5468
+    },
+    {
+      "epoch": 0.898158602426457,
+      "grad_norm": 0.28285293262085576,
+      "learning_rate": 8.362710276600065e-06,
+      "loss": 0.5094,
+      "step": 5469
+    },
+    {
+      "epoch": 0.8983228296347997,
+      "grad_norm": 0.27983567395929104,
+      "learning_rate": 8.362469666180536e-06,
+      "loss": 0.4874,
+      "step": 5470
+    },
+    {
+      "epoch": 0.8984870568431425,
+      "grad_norm": 0.397188968187713,
+      "learning_rate": 8.362229013810786e-06,
+      "loss": 0.5007,
+      "step": 5471
+    },
+    {
+      "epoch": 0.8986512840514852,
+      "grad_norm": 0.29935214259494414,
+      "learning_rate": 8.361988319493429e-06,
+      "loss": 0.4832,
+      "step": 5472
+    },
+    {
+      "epoch": 0.898815511259828,
+      "grad_norm": 0.35287779460582697,
+      "learning_rate": 8.36174758323108e-06,
+      "loss": 0.5194,
+      "step": 5473
+    },
+    {
+      "epoch": 0.8989797384681707,
+      "grad_norm": 0.3208882613202515,
+      "learning_rate": 8.361506805026352e-06,
+      "loss": 0.4964,
+      "step": 5474
+    },
+    {
+      "epoch": 0.8991439656765134,
+      "grad_norm": 0.3071412681372098,
+      "learning_rate": 8.361265984881862e-06,
+      "loss": 0.4833,
+      "step": 5475
+    },
+    {
+      "epoch": 0.8993081928848562,
+      "grad_norm": 0.2924027689988717,
+      "learning_rate": 8.361025122800223e-06,
+      "loss": 0.5363,
+      "step": 5476
+    },
+    {
+      "epoch": 0.8994724200931989,
+      "grad_norm": 0.2815585011141773,
+      "learning_rate": 8.360784218784054e-06,
+      "loss": 0.502,
+      "step": 5477
+    },
+    {
+      "epoch": 0.8996366473015417,
+      "grad_norm": 0.361696972613681,
+      "learning_rate": 8.360543272835968e-06,
+      "loss": 0.5054,
+      "step": 5478
+    },
+    {
+      "epoch": 0.8998008745098844,
+      "grad_norm": 0.2999832471190714,
+      "learning_rate": 8.360302284958586e-06,
+      "loss": 0.4908,
+      "step": 5479
+    },
+    {
+      "epoch": 0.8999651017182272,
+      "grad_norm": 0.4261624509261548,
+      "learning_rate": 8.360061255154521e-06,
+      "loss": 0.4898,
+      "step": 5480
+    },
+    {
+      "epoch": 0.9001293289265699,
+      "grad_norm": 0.2855915803003546,
+      "learning_rate": 8.359820183426395e-06,
+      "loss": 0.5164,
+      "step": 5481
+    },
+    {
+      "epoch": 0.9002935561349127,
+      "grad_norm": 0.29396206548228376,
+      "learning_rate": 8.359579069776822e-06,
+      "loss": 0.5047,
+      "step": 5482
+    },
+    {
+      "epoch": 0.9004577833432554,
+      "grad_norm": 0.3239709685752758,
+      "learning_rate": 8.359337914208424e-06,
+      "loss": 0.4881,
+      "step": 5483
+    },
+    {
+      "epoch": 0.9006220105515982,
+      "grad_norm": 0.3189529432048531,
+      "learning_rate": 8.35909671672382e-06,
+      "loss": 0.5118,
+      "step": 5484
+    },
+    {
+      "epoch": 0.9007862377599409,
+      "grad_norm": 0.3015118477260375,
+      "learning_rate": 8.358855477325628e-06,
+      "loss": 0.5252,
+      "step": 5485
+    },
+    {
+      "epoch": 0.9009504649682836,
+      "grad_norm": 0.3220454001782521,
+      "learning_rate": 8.35861419601647e-06,
+      "loss": 0.5245,
+      "step": 5486
+    },
+    {
+      "epoch": 0.9011146921766263,
+      "grad_norm": 0.29201083259315386,
+      "learning_rate": 8.358372872798964e-06,
+      "loss": 0.5238,
+      "step": 5487
+    },
+    {
+      "epoch": 0.9012789193849691,
+      "grad_norm": 0.3027181699860984,
+      "learning_rate": 8.358131507675735e-06,
+      "loss": 0.4992,
+      "step": 5488
+    },
+    {
+      "epoch": 0.9014431465933118,
+      "grad_norm": 0.3017338774888951,
+      "learning_rate": 8.357890100649397e-06,
+      "loss": 0.5002,
+      "step": 5489
+    },
+    {
+      "epoch": 0.9016073738016546,
+      "grad_norm": 0.2942002751508648,
+      "learning_rate": 8.357648651722582e-06,
+      "loss": 0.503,
+      "step": 5490
+    },
+    {
+      "epoch": 0.9017716010099973,
+      "grad_norm": 1.0776946000844136,
+      "learning_rate": 8.357407160897904e-06,
+      "loss": 0.5231,
+      "step": 5491
+    },
+    {
+      "epoch": 0.9019358282183401,
+      "grad_norm": 0.2969948419282714,
+      "learning_rate": 8.357165628177992e-06,
+      "loss": 0.5096,
+      "step": 5492
+    },
+    {
+      "epoch": 0.9021000554266828,
+      "grad_norm": 0.3085209220722856,
+      "learning_rate": 8.356924053565463e-06,
+      "loss": 0.4987,
+      "step": 5493
+    },
+    {
+      "epoch": 0.9022642826350256,
+      "grad_norm": 0.3732207561485276,
+      "learning_rate": 8.356682437062946e-06,
+      "loss": 0.5137,
+      "step": 5494
+    },
+    {
+      "epoch": 0.9024285098433683,
+      "grad_norm": 0.3006585094430494,
+      "learning_rate": 8.356440778673063e-06,
+      "loss": 0.4939,
+      "step": 5495
+    },
+    {
+      "epoch": 0.9025927370517111,
+      "grad_norm": 0.33128076755712377,
+      "learning_rate": 8.356199078398437e-06,
+      "loss": 0.4933,
+      "step": 5496
+    },
+    {
+      "epoch": 0.9027569642600538,
+      "grad_norm": 0.2931680063313428,
+      "learning_rate": 8.355957336241697e-06,
+      "loss": 0.4963,
+      "step": 5497
+    },
+    {
+      "epoch": 0.9029211914683966,
+      "grad_norm": 0.43709410205234245,
+      "learning_rate": 8.355715552205467e-06,
+      "loss": 0.5172,
+      "step": 5498
+    },
+    {
+      "epoch": 0.9030854186767393,
+      "grad_norm": 0.5518303131475227,
+      "learning_rate": 8.355473726292373e-06,
+      "loss": 0.5146,
+      "step": 5499
+    },
+    {
+      "epoch": 0.903249645885082,
+      "grad_norm": 0.35951046327730357,
+      "learning_rate": 8.35523185850504e-06,
+      "loss": 0.5053,
+      "step": 5500
+    },
+    {
+      "epoch": 0.9034138730934248,
+      "grad_norm": 0.31330180644732103,
+      "learning_rate": 8.354989948846096e-06,
+      "loss": 0.5169,
+      "step": 5501
+    },
+    {
+      "epoch": 0.9035781003017676,
+      "grad_norm": 0.2881362333682155,
+      "learning_rate": 8.354747997318168e-06,
+      "loss": 0.49,
+      "step": 5502
+    },
+    {
+      "epoch": 0.9037423275101102,
+      "grad_norm": 0.360101900220015,
+      "learning_rate": 8.354506003923884e-06,
+      "loss": 0.5072,
+      "step": 5503
+    },
+    {
+      "epoch": 0.9039065547184529,
+      "grad_norm": 0.31153369530916664,
+      "learning_rate": 8.354263968665873e-06,
+      "loss": 0.4925,
+      "step": 5504
+    },
+    {
+      "epoch": 0.9040707819267957,
+      "grad_norm": 0.37296961938079903,
+      "learning_rate": 8.354021891546764e-06,
+      "loss": 0.5135,
+      "step": 5505
+    },
+    {
+      "epoch": 0.9042350091351384,
+      "grad_norm": 0.9154141358636969,
+      "learning_rate": 8.353779772569184e-06,
+      "loss": 0.5015,
+      "step": 5506
+    },
+    {
+      "epoch": 0.9043992363434812,
+      "grad_norm": 0.296837819666743,
+      "learning_rate": 8.353537611735765e-06,
+      "loss": 0.5172,
+      "step": 5507
+    },
+    {
+      "epoch": 0.9045634635518239,
+      "grad_norm": 0.5177522705968512,
+      "learning_rate": 8.353295409049137e-06,
+      "loss": 0.5139,
+      "step": 5508
+    },
+    {
+      "epoch": 0.9047276907601667,
+      "grad_norm": 0.5204967918982784,
+      "learning_rate": 8.353053164511928e-06,
+      "loss": 0.5154,
+      "step": 5509
+    },
+    {
+      "epoch": 0.9048919179685094,
+      "grad_norm": 0.27139361175624743,
+      "learning_rate": 8.352810878126771e-06,
+      "loss": 0.4869,
+      "step": 5510
+    },
+    {
+      "epoch": 0.9050561451768522,
+      "grad_norm": 0.349204517447571,
+      "learning_rate": 8.352568549896298e-06,
+      "loss": 0.4964,
+      "step": 5511
+    },
+    {
+      "epoch": 0.9052203723851949,
+      "grad_norm": 0.3148885616058754,
+      "learning_rate": 8.352326179823139e-06,
+      "loss": 0.4921,
+      "step": 5512
+    },
+    {
+      "epoch": 0.9053845995935377,
+      "grad_norm": 0.5633992988398933,
+      "learning_rate": 8.352083767909929e-06,
+      "loss": 0.5095,
+      "step": 5513
+    },
+    {
+      "epoch": 0.9055488268018804,
+      "grad_norm": 0.3212531165331761,
+      "learning_rate": 8.351841314159298e-06,
+      "loss": 0.5034,
+      "step": 5514
+    },
+    {
+      "epoch": 0.9057130540102232,
+      "grad_norm": 0.45538311186140923,
+      "learning_rate": 8.351598818573881e-06,
+      "loss": 0.4843,
+      "step": 5515
+    },
+    {
+      "epoch": 0.9058772812185659,
+      "grad_norm": 0.2766324918567951,
+      "learning_rate": 8.351356281156313e-06,
+      "loss": 0.5163,
+      "step": 5516
+    },
+    {
+      "epoch": 0.9060415084269087,
+      "grad_norm": 0.36399910518378154,
+      "learning_rate": 8.351113701909225e-06,
+      "loss": 0.5107,
+      "step": 5517
+    },
+    {
+      "epoch": 0.9062057356352514,
+      "grad_norm": 0.2757312212678587,
+      "learning_rate": 8.350871080835253e-06,
+      "loss": 0.4934,
+      "step": 5518
+    },
+    {
+      "epoch": 0.9063699628435941,
+      "grad_norm": 0.2858011061130112,
+      "learning_rate": 8.350628417937031e-06,
+      "loss": 0.5012,
+      "step": 5519
+    },
+    {
+      "epoch": 0.9065341900519368,
+      "grad_norm": 0.4340611250040619,
+      "learning_rate": 8.350385713217198e-06,
+      "loss": 0.5228,
+      "step": 5520
+    },
+    {
+      "epoch": 0.9066984172602796,
+      "grad_norm": 0.3087231221133075,
+      "learning_rate": 8.350142966678389e-06,
+      "loss": 0.4983,
+      "step": 5521
+    },
+    {
+      "epoch": 0.9068626444686223,
+      "grad_norm": 0.3000003148489532,
+      "learning_rate": 8.349900178323235e-06,
+      "loss": 0.493,
+      "step": 5522
+    },
+    {
+      "epoch": 0.907026871676965,
+      "grad_norm": 0.32176413494489486,
+      "learning_rate": 8.349657348154382e-06,
+      "loss": 0.5153,
+      "step": 5523
+    },
+    {
+      "epoch": 0.9071910988853078,
+      "grad_norm": 0.33979379386785885,
+      "learning_rate": 8.34941447617446e-06,
+      "loss": 0.5057,
+      "step": 5524
+    },
+    {
+      "epoch": 0.9073553260936505,
+      "grad_norm": 0.2904116241815765,
+      "learning_rate": 8.349171562386111e-06,
+      "loss": 0.5207,
+      "step": 5525
+    },
+    {
+      "epoch": 0.9075195533019933,
+      "grad_norm": 0.3719685326435547,
+      "learning_rate": 8.348928606791971e-06,
+      "loss": 0.483,
+      "step": 5526
+    },
+    {
+      "epoch": 0.907683780510336,
+      "grad_norm": 0.42582862430775315,
+      "learning_rate": 8.348685609394678e-06,
+      "loss": 0.4974,
+      "step": 5527
+    },
+    {
+      "epoch": 0.9078480077186788,
+      "grad_norm": 0.37620185257746835,
+      "learning_rate": 8.348442570196875e-06,
+      "loss": 0.5158,
+      "step": 5528
+    },
+    {
+      "epoch": 0.9080122349270215,
+      "grad_norm": 0.34756089762014936,
+      "learning_rate": 8.348199489201198e-06,
+      "loss": 0.4865,
+      "step": 5529
+    },
+    {
+      "epoch": 0.9081764621353643,
+      "grad_norm": 0.3427940825007974,
+      "learning_rate": 8.34795636641029e-06,
+      "loss": 0.5037,
+      "step": 5530
+    },
+    {
+      "epoch": 0.908340689343707,
+      "grad_norm": 0.34160888168112385,
+      "learning_rate": 8.347713201826788e-06,
+      "loss": 0.5134,
+      "step": 5531
+    },
+    {
+      "epoch": 0.9085049165520498,
+      "grad_norm": 0.30878978550613373,
+      "learning_rate": 8.347469995453336e-06,
+      "loss": 0.5287,
+      "step": 5532
+    },
+    {
+      "epoch": 0.9086691437603925,
+      "grad_norm": 0.3362932220872595,
+      "learning_rate": 8.347226747292575e-06,
+      "loss": 0.5171,
+      "step": 5533
+    },
+    {
+      "epoch": 0.9088333709687353,
+      "grad_norm": 0.6424167138285161,
+      "learning_rate": 8.346983457347146e-06,
+      "loss": 0.513,
+      "step": 5534
+    },
+    {
+      "epoch": 0.908997598177078,
+      "grad_norm": 0.32394177404105495,
+      "learning_rate": 8.346740125619689e-06,
+      "loss": 0.4576,
+      "step": 5535
+    },
+    {
+      "epoch": 0.9091618253854207,
+      "grad_norm": 0.34000863262833253,
+      "learning_rate": 8.346496752112854e-06,
+      "loss": 0.5224,
+      "step": 5536
+    },
+    {
+      "epoch": 0.9093260525937634,
+      "grad_norm": 0.3247810713170286,
+      "learning_rate": 8.346253336829277e-06,
+      "loss": 0.4893,
+      "step": 5537
+    },
+    {
+      "epoch": 0.9094902798021062,
+      "grad_norm": 0.2939508899368689,
+      "learning_rate": 8.346009879771605e-06,
+      "loss": 0.51,
+      "step": 5538
+    },
+    {
+      "epoch": 0.9096545070104489,
+      "grad_norm": 0.32148698533969494,
+      "learning_rate": 8.345766380942483e-06,
+      "loss": 0.5037,
+      "step": 5539
+    },
+    {
+      "epoch": 0.9098187342187917,
+      "grad_norm": 0.28369865166116354,
+      "learning_rate": 8.345522840344553e-06,
+      "loss": 0.4952,
+      "step": 5540
+    },
+    {
+      "epoch": 0.9099829614271344,
+      "grad_norm": 0.5346924787068149,
+      "learning_rate": 8.345279257980461e-06,
+      "loss": 0.5126,
+      "step": 5541
+    },
+    {
+      "epoch": 0.9101471886354772,
+      "grad_norm": 0.32168574288406493,
+      "learning_rate": 8.345035633852855e-06,
+      "loss": 0.5086,
+      "step": 5542
+    },
+    {
+      "epoch": 0.9103114158438199,
+      "grad_norm": 0.3307362293505546,
+      "learning_rate": 8.344791967964377e-06,
+      "loss": 0.5107,
+      "step": 5543
+    },
+    {
+      "epoch": 0.9104756430521627,
+      "grad_norm": 0.3365560145974131,
+      "learning_rate": 8.344548260317678e-06,
+      "loss": 0.488,
+      "step": 5544
+    },
+    {
+      "epoch": 0.9106398702605054,
+      "grad_norm": 0.3480169970867232,
+      "learning_rate": 8.3443045109154e-06,
+      "loss": 0.4893,
+      "step": 5545
+    },
+    {
+      "epoch": 0.9108040974688482,
+      "grad_norm": 0.299993591213752,
+      "learning_rate": 8.344060719760193e-06,
+      "loss": 0.4908,
+      "step": 5546
+    },
+    {
+      "epoch": 0.9109683246771909,
+      "grad_norm": 0.2999257482434843,
+      "learning_rate": 8.343816886854707e-06,
+      "loss": 0.4876,
+      "step": 5547
+    },
+    {
+      "epoch": 0.9111325518855337,
+      "grad_norm": 0.3924513462074901,
+      "learning_rate": 8.343573012201584e-06,
+      "loss": 0.5141,
+      "step": 5548
+    },
+    {
+      "epoch": 0.9112967790938764,
+      "grad_norm": 0.3399115729754476,
+      "learning_rate": 8.34332909580348e-06,
+      "loss": 0.5082,
+      "step": 5549
+    },
+    {
+      "epoch": 0.9114610063022192,
+      "grad_norm": 0.3194858626705532,
+      "learning_rate": 8.343085137663037e-06,
+      "loss": 0.4766,
+      "step": 5550
+    },
+    {
+      "epoch": 0.9116252335105619,
+      "grad_norm": 0.47154597792158837,
+      "learning_rate": 8.342841137782912e-06,
+      "loss": 0.5031,
+      "step": 5551
+    },
+    {
+      "epoch": 0.9117894607189047,
+      "grad_norm": 0.4575805099261349,
+      "learning_rate": 8.342597096165748e-06,
+      "loss": 0.51,
+      "step": 5552
+    },
+    {
+      "epoch": 0.9119536879272473,
+      "grad_norm": 0.4071492592305457,
+      "learning_rate": 8.342353012814202e-06,
+      "loss": 0.4831,
+      "step": 5553
+    },
+    {
+      "epoch": 0.91211791513559,
+      "grad_norm": 0.3173333243154672,
+      "learning_rate": 8.34210888773092e-06,
+      "loss": 0.4953,
+      "step": 5554
+    },
+    {
+      "epoch": 0.9122821423439328,
+      "grad_norm": 0.3570551254398815,
+      "learning_rate": 8.341864720918558e-06,
+      "loss": 0.5008,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9124463695522755,
+      "grad_norm": 0.3179412588931241,
+      "learning_rate": 8.341620512379762e-06,
+      "loss": 0.5091,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9126105967606183,
+      "grad_norm": 0.3702640163571449,
+      "learning_rate": 8.341376262117189e-06,
+      "loss": 0.4887,
+      "step": 5557
+    },
+    {
+      "epoch": 0.912774823968961,
+      "grad_norm": 0.33515234301086044,
+      "learning_rate": 8.341131970133491e-06,
+      "loss": 0.5023,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9129390511773038,
+      "grad_norm": 0.29906964223061433,
+      "learning_rate": 8.34088763643132e-06,
+      "loss": 0.4972,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9131032783856465,
+      "grad_norm": 0.3263066136291013,
+      "learning_rate": 8.340643261013328e-06,
+      "loss": 0.4997,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9132675055939893,
+      "grad_norm": 0.27989015681060403,
+      "learning_rate": 8.340398843882175e-06,
+      "loss": 0.5007,
+      "step": 5561
+    },
+    {
+      "epoch": 0.913431732802332,
+      "grad_norm": 0.4458218715249904,
+      "learning_rate": 8.340154385040511e-06,
+      "loss": 0.4915,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9135959600106748,
+      "grad_norm": 0.30774662912355444,
+      "learning_rate": 8.339909884490991e-06,
+      "loss": 0.4783,
+      "step": 5563
+    },
+    {
+      "epoch": 0.9137601872190175,
+      "grad_norm": 0.2993000222057823,
+      "learning_rate": 8.339665342236275e-06,
+      "loss": 0.5053,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9139244144273603,
+      "grad_norm": 0.3966672210293373,
+      "learning_rate": 8.339420758279012e-06,
+      "loss": 0.5183,
+      "step": 5565
+    },
+    {
+      "epoch": 0.914088641635703,
+      "grad_norm": 0.30596909940584427,
+      "learning_rate": 8.339176132621864e-06,
+      "loss": 0.494,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9142528688440458,
+      "grad_norm": 0.35348430779350265,
+      "learning_rate": 8.338931465267485e-06,
+      "loss": 0.5163,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9144170960523885,
+      "grad_norm": 0.27033879796104393,
+      "learning_rate": 8.338686756218534e-06,
+      "loss": 0.5094,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9145813232607313,
+      "grad_norm": 0.3913309376275306,
+      "learning_rate": 8.338442005477667e-06,
+      "loss": 0.5038,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9147455504690739,
+      "grad_norm": 0.28689614433794935,
+      "learning_rate": 8.338197213047544e-06,
+      "loss": 0.5067,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9149097776774167,
+      "grad_norm": 0.4050157925870621,
+      "learning_rate": 8.337952378930823e-06,
+      "loss": 0.4983,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9150740048857594,
+      "grad_norm": 0.2865568953025675,
+      "learning_rate": 8.337707503130163e-06,
+      "loss": 0.4963,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9152382320941022,
+      "grad_norm": 0.29733174921046923,
+      "learning_rate": 8.337462585648224e-06,
+      "loss": 0.5153,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9154024593024449,
+      "grad_norm": 0.3180051252198621,
+      "learning_rate": 8.337217626487665e-06,
+      "loss": 0.5151,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9155666865107877,
+      "grad_norm": 0.40773067814807856,
+      "learning_rate": 8.336972625651149e-06,
+      "loss": 0.508,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9157309137191304,
+      "grad_norm": 0.3232392270623946,
+      "learning_rate": 8.336727583141335e-06,
+      "loss": 0.5015,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9158951409274732,
+      "grad_norm": 0.5050168357579791,
+      "learning_rate": 8.336482498960883e-06,
+      "loss": 0.5041,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9160593681358159,
+      "grad_norm": 0.36104906234395595,
+      "learning_rate": 8.336237373112456e-06,
+      "loss": 0.4821,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9162235953441586,
+      "grad_norm": 0.3076149792819011,
+      "learning_rate": 8.33599220559872e-06,
+      "loss": 0.5098,
+      "step": 5579
+    },
+    {
+      "epoch": 0.9163878225525014,
+      "grad_norm": 0.6724914596205441,
+      "learning_rate": 8.335746996422332e-06,
+      "loss": 0.504,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9165520497608441,
+      "grad_norm": 0.42408137374830124,
+      "learning_rate": 8.335501745585959e-06,
+      "loss": 0.504,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9167162769691869,
+      "grad_norm": 0.41731179466139035,
+      "learning_rate": 8.335256453092263e-06,
+      "loss": 0.5103,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9168805041775296,
+      "grad_norm": 0.47722812707169454,
+      "learning_rate": 8.335011118943908e-06,
+      "loss": 0.4966,
+      "step": 5583
+    },
+    {
+      "epoch": 0.9170447313858724,
+      "grad_norm": 0.4352879061716675,
+      "learning_rate": 8.33476574314356e-06,
+      "loss": 0.4979,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9172089585942151,
+      "grad_norm": 0.341870163144751,
+      "learning_rate": 8.334520325693881e-06,
+      "loss": 0.5084,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9173731858025579,
+      "grad_norm": 0.5324312522979684,
+      "learning_rate": 8.334274866597541e-06,
+      "loss": 0.4798,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9175374130109005,
+      "grad_norm": 0.39344851183289753,
+      "learning_rate": 8.334029365857202e-06,
+      "loss": 0.4974,
+      "step": 5587
+    },
+    {
+      "epoch": 0.9177016402192433,
+      "grad_norm": 0.46426361762009455,
+      "learning_rate": 8.333783823475533e-06,
+      "loss": 0.5035,
+      "step": 5588
+    },
+    {
+      "epoch": 0.917865867427586,
+      "grad_norm": 0.4075244707525329,
+      "learning_rate": 8.333538239455199e-06,
+      "loss": 0.48,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9180300946359288,
+      "grad_norm": 0.36870409810649557,
+      "learning_rate": 8.333292613798868e-06,
+      "loss": 0.505,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9181943218442715,
+      "grad_norm": 0.34786470966404165,
+      "learning_rate": 8.333046946509209e-06,
+      "loss": 0.4902,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9183585490526143,
+      "grad_norm": 0.2921529860594467,
+      "learning_rate": 8.332801237588886e-06,
+      "loss": 0.4824,
+      "step": 5592
+    },
+    {
+      "epoch": 0.918522776260957,
+      "grad_norm": 0.6192703487224838,
+      "learning_rate": 8.332555487040574e-06,
+      "loss": 0.4948,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9186870034692998,
+      "grad_norm": 0.3317052036236938,
+      "learning_rate": 8.332309694866937e-06,
+      "loss": 0.5057,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9188512306776425,
+      "grad_norm": 0.38647318368047534,
+      "learning_rate": 8.332063861070646e-06,
+      "loss": 0.5057,
+      "step": 5595
+    },
+    {
+      "epoch": 0.9190154578859853,
+      "grad_norm": 0.3435710297493405,
+      "learning_rate": 8.331817985654374e-06,
+      "loss": 0.4992,
+      "step": 5596
+    },
+    {
+      "epoch": 0.919179685094328,
+      "grad_norm": 0.34725817180505014,
+      "learning_rate": 8.331572068620785e-06,
+      "loss": 0.4929,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9193439123026708,
+      "grad_norm": 0.3235167989235117,
+      "learning_rate": 8.331326109972556e-06,
+      "loss": 0.5095,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9195081395110135,
+      "grad_norm": 0.39005161313201875,
+      "learning_rate": 8.331080109712355e-06,
+      "loss": 0.4961,
+      "step": 5599
+    },
+    {
+      "epoch": 0.9196723667193563,
+      "grad_norm": 0.3454792276254286,
+      "learning_rate": 8.330834067842853e-06,
+      "loss": 0.5285,
+      "step": 5600
+    },
+    {
+      "epoch": 0.919836593927699,
+      "grad_norm": 1.0221632544154688,
+      "learning_rate": 8.330587984366726e-06,
+      "loss": 0.4986,
+      "step": 5601
+    },
+    {
+      "epoch": 0.9200008211360418,
+      "grad_norm": 0.2994832026081828,
+      "learning_rate": 8.330341859286645e-06,
+      "loss": 0.4923,
+      "step": 5602
+    },
+    {
+      "epoch": 0.9201650483443845,
+      "grad_norm": 0.4052390833737057,
+      "learning_rate": 8.330095692605283e-06,
+      "loss": 0.5122,
+      "step": 5603
+    },
+    {
+      "epoch": 0.9203292755527271,
+      "grad_norm": 0.28648054829708214,
+      "learning_rate": 8.329849484325313e-06,
+      "loss": 0.5106,
+      "step": 5604
+    },
+    {
+      "epoch": 0.9204935027610699,
+      "grad_norm": 0.3063542399597326,
+      "learning_rate": 8.32960323444941e-06,
+      "loss": 0.5021,
+      "step": 5605
+    },
+    {
+      "epoch": 0.9206577299694126,
+      "grad_norm": 0.29985373636301893,
+      "learning_rate": 8.329356942980245e-06,
+      "loss": 0.5086,
+      "step": 5606
+    },
+    {
+      "epoch": 0.9208219571777554,
+      "grad_norm": 0.35166974021523667,
+      "learning_rate": 8.329110609920499e-06,
+      "loss": 0.5017,
+      "step": 5607
+    },
+    {
+      "epoch": 0.9209861843860981,
+      "grad_norm": 0.3991541658144142,
+      "learning_rate": 8.328864235272845e-06,
+      "loss": 0.5246,
+      "step": 5608
+    },
+    {
+      "epoch": 0.9211504115944409,
+      "grad_norm": 0.3259044403152944,
+      "learning_rate": 8.328617819039955e-06,
+      "loss": 0.4865,
+      "step": 5609
+    },
+    {
+      "epoch": 0.9213146388027836,
+      "grad_norm": 0.30778544588909174,
+      "learning_rate": 8.328371361224512e-06,
+      "loss": 0.4978,
+      "step": 5610
+    },
+    {
+      "epoch": 0.9214788660111264,
+      "grad_norm": 0.5084722341526301,
+      "learning_rate": 8.328124861829188e-06,
+      "loss": 0.5107,
+      "step": 5611
+    },
+    {
+      "epoch": 0.9216430932194691,
+      "grad_norm": 0.3117471003397797,
+      "learning_rate": 8.327878320856662e-06,
+      "loss": 0.4869,
+      "step": 5612
+    },
+    {
+      "epoch": 0.9218073204278119,
+      "grad_norm": 0.34543706062762425,
+      "learning_rate": 8.32763173830961e-06,
+      "loss": 0.515,
+      "step": 5613
+    },
+    {
+      "epoch": 0.9219715476361546,
+      "grad_norm": 0.334776273461246,
+      "learning_rate": 8.327385114190714e-06,
+      "loss": 0.4968,
+      "step": 5614
+    },
+    {
+      "epoch": 0.9221357748444974,
+      "grad_norm": 0.32328953533067045,
+      "learning_rate": 8.327138448502649e-06,
+      "loss": 0.4855,
+      "step": 5615
+    },
+    {
+      "epoch": 0.9223000020528401,
+      "grad_norm": 0.3302809998928379,
+      "learning_rate": 8.326891741248094e-06,
+      "loss": 0.5031,
+      "step": 5616
+    },
+    {
+      "epoch": 0.9224642292611829,
+      "grad_norm": 0.3127553399557638,
+      "learning_rate": 8.32664499242973e-06,
+      "loss": 0.4988,
+      "step": 5617
+    },
+    {
+      "epoch": 0.9226284564695256,
+      "grad_norm": 0.4586672526902129,
+      "learning_rate": 8.326398202050236e-06,
+      "loss": 0.4693,
+      "step": 5618
+    },
+    {
+      "epoch": 0.9227926836778684,
+      "grad_norm": 0.3313144730744136,
+      "learning_rate": 8.326151370112294e-06,
+      "loss": 0.5186,
+      "step": 5619
+    },
+    {
+      "epoch": 0.9229569108862111,
+      "grad_norm": 0.3316385292266702,
+      "learning_rate": 8.325904496618583e-06,
+      "loss": 0.5033,
+      "step": 5620
+    },
+    {
+      "epoch": 0.9231211380945538,
+      "grad_norm": 0.29228573535246216,
+      "learning_rate": 8.325657581571784e-06,
+      "loss": 0.5004,
+      "step": 5621
+    },
+    {
+      "epoch": 0.9232853653028965,
+      "grad_norm": 0.4226834910779713,
+      "learning_rate": 8.325410624974582e-06,
+      "loss": 0.493,
+      "step": 5622
+    },
+    {
+      "epoch": 0.9234495925112393,
+      "grad_norm": 0.3458030476801539,
+      "learning_rate": 8.325163626829656e-06,
+      "loss": 0.5041,
+      "step": 5623
+    },
+    {
+      "epoch": 0.923613819719582,
+      "grad_norm": 0.4242642019903402,
+      "learning_rate": 8.324916587139689e-06,
+      "loss": 0.5033,
+      "step": 5624
+    },
+    {
+      "epoch": 0.9237780469279248,
+      "grad_norm": 0.3100029512474492,
+      "learning_rate": 8.324669505907365e-06,
+      "loss": 0.5062,
+      "step": 5625
+    },
+    {
+      "epoch": 0.9239422741362675,
+      "grad_norm": 0.3185639380577177,
+      "learning_rate": 8.324422383135368e-06,
+      "loss": 0.4924,
+      "step": 5626
+    },
+    {
+      "epoch": 0.9241065013446103,
+      "grad_norm": 0.4050609577398203,
+      "learning_rate": 8.32417521882638e-06,
+      "loss": 0.4899,
+      "step": 5627
+    },
+    {
+      "epoch": 0.924270728552953,
+      "grad_norm": 0.31533095420755064,
+      "learning_rate": 8.323928012983087e-06,
+      "loss": 0.4909,
+      "step": 5628
+    },
+    {
+      "epoch": 0.9244349557612958,
+      "grad_norm": 0.2842521067234476,
+      "learning_rate": 8.323680765608173e-06,
+      "loss": 0.4873,
+      "step": 5629
+    },
+    {
+      "epoch": 0.9245991829696385,
+      "grad_norm": 0.3258802618679284,
+      "learning_rate": 8.323433476704325e-06,
+      "loss": 0.5076,
+      "step": 5630
+    },
+    {
+      "epoch": 0.9247634101779812,
+      "grad_norm": 0.299409442711691,
+      "learning_rate": 8.323186146274228e-06,
+      "loss": 0.4988,
+      "step": 5631
+    },
+    {
+      "epoch": 0.924927637386324,
+      "grad_norm": 0.3156360159559121,
+      "learning_rate": 8.322938774320568e-06,
+      "loss": 0.4989,
+      "step": 5632
+    },
+    {
+      "epoch": 0.9250918645946667,
+      "grad_norm": 0.2938555458815616,
+      "learning_rate": 8.32269136084603e-06,
+      "loss": 0.5171,
+      "step": 5633
+    },
+    {
+      "epoch": 0.9252560918030095,
+      "grad_norm": 0.40501414098775324,
+      "learning_rate": 8.322443905853303e-06,
+      "loss": 0.526,
+      "step": 5634
+    },
+    {
+      "epoch": 0.9254203190113522,
+      "grad_norm": 0.3931817021892204,
+      "learning_rate": 8.322196409345074e-06,
+      "loss": 0.4976,
+      "step": 5635
+    },
+    {
+      "epoch": 0.925584546219695,
+      "grad_norm": 0.4093911838085258,
+      "learning_rate": 8.321948871324033e-06,
+      "loss": 0.5009,
+      "step": 5636
+    },
+    {
+      "epoch": 0.9257487734280377,
+      "grad_norm": 0.3601362068180492,
+      "learning_rate": 8.321701291792867e-06,
+      "loss": 0.5002,
+      "step": 5637
+    },
+    {
+      "epoch": 0.9259130006363804,
+      "grad_norm": 0.8262370739905388,
+      "learning_rate": 8.321453670754264e-06,
+      "loss": 0.4923,
+      "step": 5638
+    },
+    {
+      "epoch": 0.9260772278447231,
+      "grad_norm": 1.323739662827969,
+      "learning_rate": 8.321206008210914e-06,
+      "loss": 0.5188,
+      "step": 5639
+    },
+    {
+      "epoch": 0.9262414550530659,
+      "grad_norm": 0.2963276769726195,
+      "learning_rate": 8.320958304165509e-06,
+      "loss": 0.5017,
+      "step": 5640
+    },
+    {
+      "epoch": 0.9264056822614086,
+      "grad_norm": 0.35472151961006554,
+      "learning_rate": 8.320710558620736e-06,
+      "loss": 0.5008,
+      "step": 5641
+    },
+    {
+      "epoch": 0.9265699094697514,
+      "grad_norm": 0.47096437311419886,
+      "learning_rate": 8.320462771579287e-06,
+      "loss": 0.5187,
+      "step": 5642
+    },
+    {
+      "epoch": 0.9267341366780941,
+      "grad_norm": 0.3116469536361895,
+      "learning_rate": 8.320214943043856e-06,
+      "loss": 0.4931,
+      "step": 5643
+    },
+    {
+      "epoch": 0.9268983638864369,
+      "grad_norm": 0.34571874140096176,
+      "learning_rate": 8.31996707301713e-06,
+      "loss": 0.5038,
+      "step": 5644
+    },
+    {
+      "epoch": 0.9270625910947796,
+      "grad_norm": 0.3000670197452908,
+      "learning_rate": 8.319719161501803e-06,
+      "loss": 0.5081,
+      "step": 5645
+    },
+    {
+      "epoch": 0.9272268183031224,
+      "grad_norm": 0.5315823597866184,
+      "learning_rate": 8.319471208500568e-06,
+      "loss": 0.5268,
+      "step": 5646
+    },
+    {
+      "epoch": 0.9273910455114651,
+      "grad_norm": 0.3194174964969078,
+      "learning_rate": 8.319223214016118e-06,
+      "loss": 0.4853,
+      "step": 5647
+    },
+    {
+      "epoch": 0.9275552727198079,
+      "grad_norm": 0.2909221729139333,
+      "learning_rate": 8.318975178051146e-06,
+      "loss": 0.4761,
+      "step": 5648
+    },
+    {
+      "epoch": 0.9277194999281506,
+      "grad_norm": 0.3266591840424489,
+      "learning_rate": 8.318727100608347e-06,
+      "loss": 0.4968,
+      "step": 5649
+    },
+    {
+      "epoch": 0.9278837271364934,
+      "grad_norm": 0.3951634535641033,
+      "learning_rate": 8.318478981690415e-06,
+      "loss": 0.4913,
+      "step": 5650
+    },
+    {
+      "epoch": 0.9280479543448361,
+      "grad_norm": 0.5327103196781549,
+      "learning_rate": 8.318230821300044e-06,
+      "loss": 0.5045,
+      "step": 5651
+    },
+    {
+      "epoch": 0.9282121815531789,
+      "grad_norm": 0.3387983201320291,
+      "learning_rate": 8.317982619439927e-06,
+      "loss": 0.496,
+      "step": 5652
+    },
+    {
+      "epoch": 0.9283764087615216,
+      "grad_norm": 0.3534333758409748,
+      "learning_rate": 8.317734376112766e-06,
+      "loss": 0.4819,
+      "step": 5653
+    },
+    {
+      "epoch": 0.9285406359698644,
+      "grad_norm": 0.45450659603234983,
+      "learning_rate": 8.317486091321253e-06,
+      "loss": 0.4933,
+      "step": 5654
+    },
+    {
+      "epoch": 0.928704863178207,
+      "grad_norm": 0.49711559052900334,
+      "learning_rate": 8.317237765068083e-06,
+      "loss": 0.5089,
+      "step": 5655
+    },
+    {
+      "epoch": 0.9288690903865497,
+      "grad_norm": 0.33609533448723666,
+      "learning_rate": 8.316989397355956e-06,
+      "loss": 0.5134,
+      "step": 5656
+    },
+    {
+      "epoch": 0.9290333175948925,
+      "grad_norm": 0.4638457318413463,
+      "learning_rate": 8.31674098818757e-06,
+      "loss": 0.4927,
+      "step": 5657
+    },
+    {
+      "epoch": 0.9291975448032352,
+      "grad_norm": 0.3490282675574865,
+      "learning_rate": 8.31649253756562e-06,
+      "loss": 0.4798,
+      "step": 5658
+    },
+    {
+      "epoch": 0.929361772011578,
+      "grad_norm": 0.37016474674914635,
+      "learning_rate": 8.316244045492809e-06,
+      "loss": 0.5155,
+      "step": 5659
+    },
+    {
+      "epoch": 0.9295259992199207,
+      "grad_norm": 0.37075033799182244,
+      "learning_rate": 8.31599551197183e-06,
+      "loss": 0.5202,
+      "step": 5660
+    },
+    {
+      "epoch": 0.9296902264282635,
+      "grad_norm": 0.3424732929894227,
+      "learning_rate": 8.315746937005386e-06,
+      "loss": 0.52,
+      "step": 5661
+    },
+    {
+      "epoch": 0.9298544536366062,
+      "grad_norm": 0.304676215982256,
+      "learning_rate": 8.315498320596177e-06,
+      "loss": 0.4924,
+      "step": 5662
+    },
+    {
+      "epoch": 0.930018680844949,
+      "grad_norm": 0.3090516157537145,
+      "learning_rate": 8.315249662746901e-06,
+      "loss": 0.5067,
+      "step": 5663
+    },
+    {
+      "epoch": 0.9301829080532917,
+      "grad_norm": 0.35434684092220853,
+      "learning_rate": 8.315000963460261e-06,
+      "loss": 0.5053,
+      "step": 5664
+    },
+    {
+      "epoch": 0.9303471352616345,
+      "grad_norm": 0.3235687017825308,
+      "learning_rate": 8.314752222738956e-06,
+      "loss": 0.5004,
+      "step": 5665
+    },
+    {
+      "epoch": 0.9305113624699772,
+      "grad_norm": 0.323592417972458,
+      "learning_rate": 8.31450344058569e-06,
+      "loss": 0.5178,
+      "step": 5666
+    },
+    {
+      "epoch": 0.93067558967832,
+      "grad_norm": 0.3269011607386471,
+      "learning_rate": 8.314254617003163e-06,
+      "loss": 0.5111,
+      "step": 5667
+    },
+    {
+      "epoch": 0.9308398168866627,
+      "grad_norm": 0.44163641425139216,
+      "learning_rate": 8.31400575199408e-06,
+      "loss": 0.5087,
+      "step": 5668
+    },
+    {
+      "epoch": 0.9310040440950055,
+      "grad_norm": 0.544952712160666,
+      "learning_rate": 8.313756845561139e-06,
+      "loss": 0.5166,
+      "step": 5669
+    },
+    {
+      "epoch": 0.9311682713033482,
+      "grad_norm": 0.31794568803873646,
+      "learning_rate": 8.313507897707047e-06,
+      "loss": 0.5212,
+      "step": 5670
+    },
+    {
+      "epoch": 0.931332498511691,
+      "grad_norm": 0.39736597639619714,
+      "learning_rate": 8.313258908434507e-06,
+      "loss": 0.509,
+      "step": 5671
+    },
+    {
+      "epoch": 0.9314967257200336,
+      "grad_norm": 0.30840079244447743,
+      "learning_rate": 8.313009877746226e-06,
+      "loss": 0.4957,
+      "step": 5672
+    },
+    {
+      "epoch": 0.9316609529283764,
+      "grad_norm": 0.34739304809928856,
+      "learning_rate": 8.312760805644905e-06,
+      "loss": 0.4971,
+      "step": 5673
+    },
+    {
+      "epoch": 0.9318251801367191,
+      "grad_norm": 0.335988944225995,
+      "learning_rate": 8.312511692133251e-06,
+      "loss": 0.5009,
+      "step": 5674
+    },
+    {
+      "epoch": 0.9319894073450619,
+      "grad_norm": 0.3893509807380632,
+      "learning_rate": 8.312262537213966e-06,
+      "loss": 0.5073,
+      "step": 5675
+    },
+    {
+      "epoch": 0.9321536345534046,
+      "grad_norm": 0.33293820028332555,
+      "learning_rate": 8.312013340889763e-06,
+      "loss": 0.4899,
+      "step": 5676
+    },
+    {
+      "epoch": 0.9323178617617474,
+      "grad_norm": 0.4913326187606124,
+      "learning_rate": 8.311764103163342e-06,
+      "loss": 0.5175,
+      "step": 5677
+    },
+    {
+      "epoch": 0.9324820889700901,
+      "grad_norm": 0.34344484845465234,
+      "learning_rate": 8.311514824037414e-06,
+      "loss": 0.4871,
+      "step": 5678
+    },
+    {
+      "epoch": 0.9326463161784329,
+      "grad_norm": 0.31321528574452057,
+      "learning_rate": 8.311265503514684e-06,
+      "loss": 0.5137,
+      "step": 5679
+    },
+    {
+      "epoch": 0.9328105433867756,
+      "grad_norm": 0.28952544103066263,
+      "learning_rate": 8.311016141597862e-06,
+      "loss": 0.4889,
+      "step": 5680
+    },
+    {
+      "epoch": 0.9329747705951184,
+      "grad_norm": 0.3679457217525709,
+      "learning_rate": 8.310766738289653e-06,
+      "loss": 0.49,
+      "step": 5681
+    },
+    {
+      "epoch": 0.9331389978034611,
+      "grad_norm": 2.1742606439642436,
+      "learning_rate": 8.31051729359277e-06,
+      "loss": 0.5173,
+      "step": 5682
+    },
+    {
+      "epoch": 0.9333032250118038,
+      "grad_norm": 0.3115240268193473,
+      "learning_rate": 8.310267807509918e-06,
+      "loss": 0.5,
+      "step": 5683
+    },
+    {
+      "epoch": 0.9334674522201466,
+      "grad_norm": 0.4170080800009593,
+      "learning_rate": 8.310018280043811e-06,
+      "loss": 0.53,
+      "step": 5684
+    },
+    {
+      "epoch": 0.9336316794284893,
+      "grad_norm": 0.33607485508133855,
+      "learning_rate": 8.309768711197156e-06,
+      "loss": 0.5004,
+      "step": 5685
+    },
+    {
+      "epoch": 0.9337959066368321,
+      "grad_norm": 0.34227008895253036,
+      "learning_rate": 8.309519100972664e-06,
+      "loss": 0.514,
+      "step": 5686
+    },
+    {
+      "epoch": 0.9339601338451748,
+      "grad_norm": 0.32230259816889123,
+      "learning_rate": 8.30926944937305e-06,
+      "loss": 0.4983,
+      "step": 5687
+    },
+    {
+      "epoch": 0.9341243610535176,
+      "grad_norm": 0.31003054600860314,
+      "learning_rate": 8.309019756401016e-06,
+      "loss": 0.4976,
+      "step": 5688
+    },
+    {
+      "epoch": 0.9342885882618602,
+      "grad_norm": 0.33146582662481516,
+      "learning_rate": 8.308770022059285e-06,
+      "loss": 0.5141,
+      "step": 5689
+    },
+    {
+      "epoch": 0.934452815470203,
+      "grad_norm": 0.29520921099226544,
+      "learning_rate": 8.30852024635056e-06,
+      "loss": 0.5035,
+      "step": 5690
+    },
+    {
+      "epoch": 0.9346170426785457,
+      "grad_norm": 0.2898500067975092,
+      "learning_rate": 8.308270429277562e-06,
+      "loss": 0.491,
+      "step": 5691
+    },
+    {
+      "epoch": 0.9347812698868885,
+      "grad_norm": 0.514172034927915,
+      "learning_rate": 8.308020570842998e-06,
+      "loss": 0.5117,
+      "step": 5692
+    },
+    {
+      "epoch": 0.9349454970952312,
+      "grad_norm": 0.32118782445662386,
+      "learning_rate": 8.307770671049586e-06,
+      "loss": 0.5002,
+      "step": 5693
+    },
+    {
+      "epoch": 0.935109724303574,
+      "grad_norm": 0.28434844389192415,
+      "learning_rate": 8.307520729900037e-06,
+      "loss": 0.5111,
+      "step": 5694
+    },
+    {
+      "epoch": 0.9352739515119167,
+      "grad_norm": 0.3150859794798156,
+      "learning_rate": 8.307270747397067e-06,
+      "loss": 0.5157,
+      "step": 5695
+    },
+    {
+      "epoch": 0.9354381787202595,
+      "grad_norm": 0.337073001654949,
+      "learning_rate": 8.30702072354339e-06,
+      "loss": 0.4861,
+      "step": 5696
+    },
+    {
+      "epoch": 0.9356024059286022,
+      "grad_norm": 0.3450124523657821,
+      "learning_rate": 8.306770658341723e-06,
+      "loss": 0.4864,
+      "step": 5697
+    },
+    {
+      "epoch": 0.935766633136945,
+      "grad_norm": 0.3124292867346078,
+      "learning_rate": 8.306520551794781e-06,
+      "loss": 0.5167,
+      "step": 5698
+    },
+    {
+      "epoch": 0.9359308603452877,
+      "grad_norm": 0.35958452699114296,
+      "learning_rate": 8.30627040390528e-06,
+      "loss": 0.4719,
+      "step": 5699
+    },
+    {
+      "epoch": 0.9360950875536305,
+      "grad_norm": 0.39144058349399646,
+      "learning_rate": 8.306020214675938e-06,
+      "loss": 0.4831,
+      "step": 5700
+    },
+    {
+      "epoch": 0.9362593147619732,
+      "grad_norm": 0.32711266635221053,
+      "learning_rate": 8.305769984109473e-06,
+      "loss": 0.5025,
+      "step": 5701
+    },
+    {
+      "epoch": 0.936423541970316,
+      "grad_norm": 0.32745055724297306,
+      "learning_rate": 8.3055197122086e-06,
+      "loss": 0.5015,
+      "step": 5702
+    },
+    {
+      "epoch": 0.9365877691786587,
+      "grad_norm": 0.2867180288316144,
+      "learning_rate": 8.30526939897604e-06,
+      "loss": 0.5111,
+      "step": 5703
+    },
+    {
+      "epoch": 0.9367519963870015,
+      "grad_norm": 0.295602851080418,
+      "learning_rate": 8.30501904441451e-06,
+      "loss": 0.4967,
+      "step": 5704
+    },
+    {
+      "epoch": 0.9369162235953442,
+      "grad_norm": 0.4263974761885831,
+      "learning_rate": 8.30476864852673e-06,
+      "loss": 0.4912,
+      "step": 5705
+    },
+    {
+      "epoch": 0.9370804508036868,
+      "grad_norm": 0.33887320596973486,
+      "learning_rate": 8.304518211315417e-06,
+      "loss": 0.4882,
+      "step": 5706
+    },
+    {
+      "epoch": 0.9372446780120296,
+      "grad_norm": 0.3675882652000894,
+      "learning_rate": 8.304267732783296e-06,
+      "loss": 0.4911,
+      "step": 5707
+    },
+    {
+      "epoch": 0.9374089052203723,
+      "grad_norm": 0.44201220610549874,
+      "learning_rate": 8.304017212933082e-06,
+      "loss": 0.5259,
+      "step": 5708
+    },
+    {
+      "epoch": 0.9375731324287151,
+      "grad_norm": 0.4404115118052987,
+      "learning_rate": 8.303766651767501e-06,
+      "loss": 0.5017,
+      "step": 5709
+    },
+    {
+      "epoch": 0.9377373596370578,
+      "grad_norm": 0.3166832436259222,
+      "learning_rate": 8.30351604928927e-06,
+      "loss": 0.498,
+      "step": 5710
+    },
+    {
+      "epoch": 0.9379015868454006,
+      "grad_norm": 0.3220410348280446,
+      "learning_rate": 8.303265405501113e-06,
+      "loss": 0.4816,
+      "step": 5711
+    },
+    {
+      "epoch": 0.9380658140537433,
+      "grad_norm": 0.49047811853744966,
+      "learning_rate": 8.303014720405753e-06,
+      "loss": 0.513,
+      "step": 5712
+    },
+    {
+      "epoch": 0.9382300412620861,
+      "grad_norm": 0.33189355575641755,
+      "learning_rate": 8.302763994005908e-06,
+      "loss": 0.5082,
+      "step": 5713
+    },
+    {
+      "epoch": 0.9383942684704288,
+      "grad_norm": 0.3758239947437896,
+      "learning_rate": 8.30251322630431e-06,
+      "loss": 0.5046,
+      "step": 5714
+    },
+    {
+      "epoch": 0.9385584956787716,
+      "grad_norm": 0.33422695693160437,
+      "learning_rate": 8.302262417303673e-06,
+      "loss": 0.4978,
+      "step": 5715
+    },
+    {
+      "epoch": 0.9387227228871143,
+      "grad_norm": 0.3216246403940893,
+      "learning_rate": 8.302011567006726e-06,
+      "loss": 0.4972,
+      "step": 5716
+    },
+    {
+      "epoch": 0.9388869500954571,
+      "grad_norm": 0.3375123906537756,
+      "learning_rate": 8.301760675416193e-06,
+      "loss": 0.5055,
+      "step": 5717
+    },
+    {
+      "epoch": 0.9390511773037998,
+      "grad_norm": 0.3245098375695823,
+      "learning_rate": 8.301509742534797e-06,
+      "loss": 0.4923,
+      "step": 5718
+    },
+    {
+      "epoch": 0.9392154045121426,
+      "grad_norm": 0.300498447635204,
+      "learning_rate": 8.301258768365269e-06,
+      "loss": 0.5084,
+      "step": 5719
+    },
+    {
+      "epoch": 0.9393796317204853,
+      "grad_norm": 0.2916667872950895,
+      "learning_rate": 8.301007752910327e-06,
+      "loss": 0.4728,
+      "step": 5720
+    },
+    {
+      "epoch": 0.9395438589288281,
+      "grad_norm": 0.29636711926991816,
+      "learning_rate": 8.300756696172703e-06,
+      "loss": 0.5363,
+      "step": 5721
+    },
+    {
+      "epoch": 0.9397080861371708,
+      "grad_norm": 0.36656193950960636,
+      "learning_rate": 8.300505598155121e-06,
+      "loss": 0.5059,
+      "step": 5722
+    },
+    {
+      "epoch": 0.9398723133455135,
+      "grad_norm": 0.3305934017377256,
+      "learning_rate": 8.30025445886031e-06,
+      "loss": 0.4864,
+      "step": 5723
+    },
+    {
+      "epoch": 0.9400365405538562,
+      "grad_norm": 0.2750640832912153,
+      "learning_rate": 8.300003278290996e-06,
+      "loss": 0.4961,
+      "step": 5724
+    },
+    {
+      "epoch": 0.940200767762199,
+      "grad_norm": 0.2876805407058977,
+      "learning_rate": 8.299752056449908e-06,
+      "loss": 0.4967,
+      "step": 5725
+    },
+    {
+      "epoch": 0.9403649949705417,
+      "grad_norm": 0.35466903617984374,
+      "learning_rate": 8.299500793339775e-06,
+      "loss": 0.4968,
+      "step": 5726
+    },
+    {
+      "epoch": 0.9405292221788845,
+      "grad_norm": 0.340814806284325,
+      "learning_rate": 8.299249488963322e-06,
+      "loss": 0.4908,
+      "step": 5727
+    },
+    {
+      "epoch": 0.9406934493872272,
+      "grad_norm": 0.3322242158557963,
+      "learning_rate": 8.298998143323286e-06,
+      "loss": 0.5069,
+      "step": 5728
+    },
+    {
+      "epoch": 0.94085767659557,
+      "grad_norm": 0.2999082292928264,
+      "learning_rate": 8.298746756422389e-06,
+      "loss": 0.4959,
+      "step": 5729
+    },
+    {
+      "epoch": 0.9410219038039127,
+      "grad_norm": 0.31878824304806774,
+      "learning_rate": 8.298495328263367e-06,
+      "loss": 0.5125,
+      "step": 5730
+    },
+    {
+      "epoch": 0.9411861310122555,
+      "grad_norm": 0.36476084540179676,
+      "learning_rate": 8.298243858848947e-06,
+      "loss": 0.4991,
+      "step": 5731
+    },
+    {
+      "epoch": 0.9413503582205982,
+      "grad_norm": 0.2634703515079286,
+      "learning_rate": 8.297992348181862e-06,
+      "loss": 0.5131,
+      "step": 5732
+    },
+    {
+      "epoch": 0.941514585428941,
+      "grad_norm": 0.2739454213919614,
+      "learning_rate": 8.297740796264845e-06,
+      "loss": 0.5084,
+      "step": 5733
+    },
+    {
+      "epoch": 0.9416788126372837,
+      "grad_norm": 0.32154129766877954,
+      "learning_rate": 8.297489203100623e-06,
+      "loss": 0.49,
+      "step": 5734
+    },
+    {
+      "epoch": 0.9418430398456265,
+      "grad_norm": 0.2974470814025031,
+      "learning_rate": 8.297237568691936e-06,
+      "loss": 0.4825,
+      "step": 5735
+    },
+    {
+      "epoch": 0.9420072670539692,
+      "grad_norm": 0.3876497388425138,
+      "learning_rate": 8.29698589304151e-06,
+      "loss": 0.523,
+      "step": 5736
+    },
+    {
+      "epoch": 0.942171494262312,
+      "grad_norm": 0.28791684142287477,
+      "learning_rate": 8.296734176152083e-06,
+      "loss": 0.4948,
+      "step": 5737
+    },
+    {
+      "epoch": 0.9423357214706547,
+      "grad_norm": 0.26299740923860576,
+      "learning_rate": 8.296482418026387e-06,
+      "loss": 0.5154,
+      "step": 5738
+    },
+    {
+      "epoch": 0.9424999486789974,
+      "grad_norm": 0.3163157616250974,
+      "learning_rate": 8.296230618667156e-06,
+      "loss": 0.4981,
+      "step": 5739
+    },
+    {
+      "epoch": 0.9426641758873401,
+      "grad_norm": 0.3183861930352154,
+      "learning_rate": 8.295978778077128e-06,
+      "loss": 0.4857,
+      "step": 5740
+    },
+    {
+      "epoch": 0.9428284030956828,
+      "grad_norm": 0.32525277310277495,
+      "learning_rate": 8.295726896259033e-06,
+      "loss": 0.5191,
+      "step": 5741
+    },
+    {
+      "epoch": 0.9429926303040256,
+      "grad_norm": 0.3938453035546823,
+      "learning_rate": 8.29547497321561e-06,
+      "loss": 0.49,
+      "step": 5742
+    },
+    {
+      "epoch": 0.9431568575123683,
+      "grad_norm": 0.2898850112140784,
+      "learning_rate": 8.295223008949595e-06,
+      "loss": 0.4781,
+      "step": 5743
+    },
+    {
+      "epoch": 0.9433210847207111,
+      "grad_norm": 0.3301700391413957,
+      "learning_rate": 8.294971003463724e-06,
+      "loss": 0.4957,
+      "step": 5744
+    },
+    {
+      "epoch": 0.9434853119290538,
+      "grad_norm": 0.2804840088969759,
+      "learning_rate": 8.294718956760736e-06,
+      "loss": 0.486,
+      "step": 5745
+    },
+    {
+      "epoch": 0.9436495391373966,
+      "grad_norm": 0.30848614862924606,
+      "learning_rate": 8.294466868843363e-06,
+      "loss": 0.4949,
+      "step": 5746
+    },
+    {
+      "epoch": 0.9438137663457393,
+      "grad_norm": 0.2778952083425503,
+      "learning_rate": 8.294214739714348e-06,
+      "loss": 0.4947,
+      "step": 5747
+    },
+    {
+      "epoch": 0.9439779935540821,
+      "grad_norm": 0.2912022792909345,
+      "learning_rate": 8.293962569376428e-06,
+      "loss": 0.5038,
+      "step": 5748
+    },
+    {
+      "epoch": 0.9441422207624248,
+      "grad_norm": 0.31372201934122734,
+      "learning_rate": 8.293710357832344e-06,
+      "loss": 0.5014,
+      "step": 5749
+    },
+    {
+      "epoch": 0.9443064479707676,
+      "grad_norm": 0.4827497389908659,
+      "learning_rate": 8.29345810508483e-06,
+      "loss": 0.4936,
+      "step": 5750
+    },
+    {
+      "epoch": 0.9444706751791103,
+      "grad_norm": 0.2978087052404848,
+      "learning_rate": 8.29320581113663e-06,
+      "loss": 0.5146,
+      "step": 5751
+    },
+    {
+      "epoch": 0.9446349023874531,
+      "grad_norm": 0.29384689410191,
+      "learning_rate": 8.292953475990481e-06,
+      "loss": 0.518,
+      "step": 5752
+    },
+    {
+      "epoch": 0.9447991295957958,
+      "grad_norm": 0.3121394187485199,
+      "learning_rate": 8.292701099649129e-06,
+      "loss": 0.4906,
+      "step": 5753
+    },
+    {
+      "epoch": 0.9449633568041386,
+      "grad_norm": 0.2965066015713293,
+      "learning_rate": 8.292448682115309e-06,
+      "loss": 0.5148,
+      "step": 5754
+    },
+    {
+      "epoch": 0.9451275840124813,
+      "grad_norm": 0.3857468121891203,
+      "learning_rate": 8.292196223391766e-06,
+      "loss": 0.4877,
+      "step": 5755
+    },
+    {
+      "epoch": 0.9452918112208241,
+      "grad_norm": 1.0082520752566388,
+      "learning_rate": 8.29194372348124e-06,
+      "loss": 0.4954,
+      "step": 5756
+    },
+    {
+      "epoch": 0.9454560384291667,
+      "grad_norm": 0.2987206706456987,
+      "learning_rate": 8.291691182386476e-06,
+      "loss": 0.5066,
+      "step": 5757
+    },
+    {
+      "epoch": 0.9456202656375094,
+      "grad_norm": 0.43792593819559156,
+      "learning_rate": 8.291438600110214e-06,
+      "loss": 0.5064,
+      "step": 5758
+    },
+    {
+      "epoch": 0.9457844928458522,
+      "grad_norm": 0.3036885285849899,
+      "learning_rate": 8.291185976655199e-06,
+      "loss": 0.4887,
+      "step": 5759
+    },
+    {
+      "epoch": 0.945948720054195,
+      "grad_norm": 0.39977706995255674,
+      "learning_rate": 8.290933312024174e-06,
+      "loss": 0.5063,
+      "step": 5760
+    },
+    {
+      "epoch": 0.9461129472625377,
+      "grad_norm": 0.30397770713157646,
+      "learning_rate": 8.290680606219883e-06,
+      "loss": 0.5182,
+      "step": 5761
+    },
+    {
+      "epoch": 0.9462771744708804,
+      "grad_norm": 0.29067054573184586,
+      "learning_rate": 8.290427859245072e-06,
+      "loss": 0.4953,
+      "step": 5762
+    },
+    {
+      "epoch": 0.9464414016792232,
+      "grad_norm": 0.2807311725585339,
+      "learning_rate": 8.290175071102486e-06,
+      "loss": 0.5128,
+      "step": 5763
+    },
+    {
+      "epoch": 0.9466056288875659,
+      "grad_norm": 0.35856042580615943,
+      "learning_rate": 8.289922241794869e-06,
+      "loss": 0.5088,
+      "step": 5764
+    },
+    {
+      "epoch": 0.9467698560959087,
+      "grad_norm": 0.36928919153283474,
+      "learning_rate": 8.289669371324966e-06,
+      "loss": 0.5089,
+      "step": 5765
+    },
+    {
+      "epoch": 0.9469340833042514,
+      "grad_norm": 0.6615780615800806,
+      "learning_rate": 8.289416459695527e-06,
+      "loss": 0.5067,
+      "step": 5766
+    },
+    {
+      "epoch": 0.9470983105125942,
+      "grad_norm": 0.28510177778702833,
+      "learning_rate": 8.289163506909297e-06,
+      "loss": 0.482,
+      "step": 5767
+    },
+    {
+      "epoch": 0.9472625377209369,
+      "grad_norm": 0.30217945823341685,
+      "learning_rate": 8.288910512969021e-06,
+      "loss": 0.5077,
+      "step": 5768
+    },
+    {
+      "epoch": 0.9474267649292797,
+      "grad_norm": 0.278858026357913,
+      "learning_rate": 8.288657477877452e-06,
+      "loss": 0.4868,
+      "step": 5769
+    },
+    {
+      "epoch": 0.9475909921376224,
+      "grad_norm": 0.3424671698610464,
+      "learning_rate": 8.288404401637332e-06,
+      "loss": 0.5011,
+      "step": 5770
+    },
+    {
+      "epoch": 0.9477552193459652,
+      "grad_norm": 0.29888762243279215,
+      "learning_rate": 8.288151284251417e-06,
+      "loss": 0.5014,
+      "step": 5771
+    },
+    {
+      "epoch": 0.9479194465543079,
+      "grad_norm": 0.3767666704556441,
+      "learning_rate": 8.28789812572245e-06,
+      "loss": 0.4968,
+      "step": 5772
+    },
+    {
+      "epoch": 0.9480836737626507,
+      "grad_norm": 0.3007862052726042,
+      "learning_rate": 8.287644926053182e-06,
+      "loss": 0.5111,
+      "step": 5773
+    },
+    {
+      "epoch": 0.9482479009709933,
+      "grad_norm": 0.36206526827161356,
+      "learning_rate": 8.287391685246363e-06,
+      "loss": 0.5185,
+      "step": 5774
+    },
+    {
+      "epoch": 0.9484121281793361,
+      "grad_norm": 0.29609444550203556,
+      "learning_rate": 8.287138403304746e-06,
+      "loss": 0.4892,
+      "step": 5775
+    },
+    {
+      "epoch": 0.9485763553876788,
+      "grad_norm": 0.2949042867632768,
+      "learning_rate": 8.286885080231079e-06,
+      "loss": 0.5162,
+      "step": 5776
+    },
+    {
+      "epoch": 0.9487405825960216,
+      "grad_norm": 0.3347524409023741,
+      "learning_rate": 8.286631716028112e-06,
+      "loss": 0.4835,
+      "step": 5777
+    },
+    {
+      "epoch": 0.9489048098043643,
+      "grad_norm": 0.2867731768433254,
+      "learning_rate": 8.286378310698603e-06,
+      "loss": 0.5115,
+      "step": 5778
+    },
+    {
+      "epoch": 0.9490690370127071,
+      "grad_norm": 0.3303083901828599,
+      "learning_rate": 8.286124864245298e-06,
+      "loss": 0.4904,
+      "step": 5779
+    },
+    {
+      "epoch": 0.9492332642210498,
+      "grad_norm": 0.30811639311477274,
+      "learning_rate": 8.285871376670953e-06,
+      "loss": 0.4935,
+      "step": 5780
+    },
+    {
+      "epoch": 0.9493974914293926,
+      "grad_norm": 0.3030673745127514,
+      "learning_rate": 8.285617847978318e-06,
+      "loss": 0.5014,
+      "step": 5781
+    },
+    {
+      "epoch": 0.9495617186377353,
+      "grad_norm": 0.3174059059837985,
+      "learning_rate": 8.285364278170152e-06,
+      "loss": 0.5008,
+      "step": 5782
+    },
+    {
+      "epoch": 0.9497259458460781,
+      "grad_norm": 0.32230661304949565,
+      "learning_rate": 8.285110667249202e-06,
+      "loss": 0.4981,
+      "step": 5783
+    },
+    {
+      "epoch": 0.9498901730544208,
+      "grad_norm": 0.2954900836401079,
+      "learning_rate": 8.284857015218228e-06,
+      "loss": 0.5152,
+      "step": 5784
+    },
+    {
+      "epoch": 0.9500544002627636,
+      "grad_norm": 0.38341331561161734,
+      "learning_rate": 8.284603322079982e-06,
+      "loss": 0.516,
+      "step": 5785
+    },
+    {
+      "epoch": 0.9502186274711063,
+      "grad_norm": 0.288180037663649,
+      "learning_rate": 8.284349587837222e-06,
+      "loss": 0.4971,
+      "step": 5786
+    },
+    {
+      "epoch": 0.950382854679449,
+      "grad_norm": 0.2639278006502859,
+      "learning_rate": 8.284095812492701e-06,
+      "loss": 0.4973,
+      "step": 5787
+    },
+    {
+      "epoch": 0.9505470818877918,
+      "grad_norm": 0.36465823278362897,
+      "learning_rate": 8.283841996049176e-06,
+      "loss": 0.5043,
+      "step": 5788
+    },
+    {
+      "epoch": 0.9507113090961345,
+      "grad_norm": 0.40869908171798774,
+      "learning_rate": 8.283588138509406e-06,
+      "loss": 0.497,
+      "step": 5789
+    },
+    {
+      "epoch": 0.9508755363044773,
+      "grad_norm": 0.2948081143710643,
+      "learning_rate": 8.283334239876145e-06,
+      "loss": 0.5012,
+      "step": 5790
+    },
+    {
+      "epoch": 0.9510397635128199,
+      "grad_norm": 0.30131869054135385,
+      "learning_rate": 8.283080300152151e-06,
+      "loss": 0.5096,
+      "step": 5791
+    },
+    {
+      "epoch": 0.9512039907211627,
+      "grad_norm": 0.2893470713777381,
+      "learning_rate": 8.282826319340185e-06,
+      "loss": 0.4803,
+      "step": 5792
+    },
+    {
+      "epoch": 0.9513682179295054,
+      "grad_norm": 0.2924402635990607,
+      "learning_rate": 8.282572297443002e-06,
+      "loss": 0.5045,
+      "step": 5793
+    },
+    {
+      "epoch": 0.9515324451378482,
+      "grad_norm": 0.30826834323579505,
+      "learning_rate": 8.282318234463361e-06,
+      "loss": 0.4942,
+      "step": 5794
+    },
+    {
+      "epoch": 0.9516966723461909,
+      "grad_norm": 0.2575049688426361,
+      "learning_rate": 8.282064130404025e-06,
+      "loss": 0.5073,
+      "step": 5795
+    },
+    {
+      "epoch": 0.9518608995545337,
+      "grad_norm": 0.2616141640965354,
+      "learning_rate": 8.281809985267752e-06,
+      "loss": 0.5015,
+      "step": 5796
+    },
+    {
+      "epoch": 0.9520251267628764,
+      "grad_norm": 0.3188316003808188,
+      "learning_rate": 8.2815557990573e-06,
+      "loss": 0.519,
+      "step": 5797
+    },
+    {
+      "epoch": 0.9521893539712192,
+      "grad_norm": 0.2517430104882772,
+      "learning_rate": 8.281301571775431e-06,
+      "loss": 0.4858,
+      "step": 5798
+    },
+    {
+      "epoch": 0.9523535811795619,
+      "grad_norm": 0.31257213248047166,
+      "learning_rate": 8.28104730342491e-06,
+      "loss": 0.4955,
+      "step": 5799
+    },
+    {
+      "epoch": 0.9525178083879047,
+      "grad_norm": 0.3406386793193978,
+      "learning_rate": 8.280792994008492e-06,
+      "loss": 0.4923,
+      "step": 5800
+    },
+    {
+      "epoch": 0.9526820355962474,
+      "grad_norm": 0.2913983953202928,
+      "learning_rate": 8.280538643528944e-06,
+      "loss": 0.5039,
+      "step": 5801
+    },
+    {
+      "epoch": 0.9528462628045902,
+      "grad_norm": 0.28772973551798076,
+      "learning_rate": 8.280284251989024e-06,
+      "loss": 0.506,
+      "step": 5802
+    },
+    {
+      "epoch": 0.9530104900129329,
+      "grad_norm": 0.30481361757707226,
+      "learning_rate": 8.280029819391499e-06,
+      "loss": 0.5027,
+      "step": 5803
+    },
+    {
+      "epoch": 0.9531747172212757,
+      "grad_norm": 0.31008301473548017,
+      "learning_rate": 8.279775345739133e-06,
+      "loss": 0.4902,
+      "step": 5804
+    },
+    {
+      "epoch": 0.9533389444296184,
+      "grad_norm": 0.31681938852800007,
+      "learning_rate": 8.279520831034688e-06,
+      "loss": 0.4983,
+      "step": 5805
+    },
+    {
+      "epoch": 0.9535031716379612,
+      "grad_norm": 0.36114831329488967,
+      "learning_rate": 8.279266275280926e-06,
+      "loss": 0.5168,
+      "step": 5806
+    },
+    {
+      "epoch": 0.9536673988463039,
+      "grad_norm": 0.35047545093351945,
+      "learning_rate": 8.279011678480614e-06,
+      "loss": 0.5074,
+      "step": 5807
+    },
+    {
+      "epoch": 0.9538316260546466,
+      "grad_norm": 0.3312605893409433,
+      "learning_rate": 8.27875704063652e-06,
+      "loss": 0.495,
+      "step": 5808
+    },
+    {
+      "epoch": 0.9539958532629893,
+      "grad_norm": 0.5613178656403057,
+      "learning_rate": 8.278502361751403e-06,
+      "loss": 0.5045,
+      "step": 5809
+    },
+    {
+      "epoch": 0.954160080471332,
+      "grad_norm": 0.28894635191610946,
+      "learning_rate": 8.278247641828035e-06,
+      "loss": 0.4889,
+      "step": 5810
+    },
+    {
+      "epoch": 0.9543243076796748,
+      "grad_norm": 0.33802860483632036,
+      "learning_rate": 8.27799288086918e-06,
+      "loss": 0.4807,
+      "step": 5811
+    },
+    {
+      "epoch": 0.9544885348880175,
+      "grad_norm": 0.3051254564753109,
+      "learning_rate": 8.277738078877606e-06,
+      "loss": 0.4955,
+      "step": 5812
+    },
+    {
+      "epoch": 0.9546527620963603,
+      "grad_norm": 0.31460451693320945,
+      "learning_rate": 8.277483235856079e-06,
+      "loss": 0.4888,
+      "step": 5813
+    },
+    {
+      "epoch": 0.954816989304703,
+      "grad_norm": 0.3011782131412546,
+      "learning_rate": 8.277228351807367e-06,
+      "loss": 0.4981,
+      "step": 5814
+    },
+    {
+      "epoch": 0.9549812165130458,
+      "grad_norm": 0.3790567372061715,
+      "learning_rate": 8.276973426734238e-06,
+      "loss": 0.5033,
+      "step": 5815
+    },
+    {
+      "epoch": 0.9551454437213885,
+      "grad_norm": 0.29233332389067523,
+      "learning_rate": 8.276718460639464e-06,
+      "loss": 0.5025,
+      "step": 5816
+    },
+    {
+      "epoch": 0.9553096709297313,
+      "grad_norm": 0.28425808137423847,
+      "learning_rate": 8.276463453525809e-06,
+      "loss": 0.4782,
+      "step": 5817
+    },
+    {
+      "epoch": 0.955473898138074,
+      "grad_norm": 0.3215019093899259,
+      "learning_rate": 8.276208405396048e-06,
+      "loss": 0.5106,
+      "step": 5818
+    },
+    {
+      "epoch": 0.9556381253464168,
+      "grad_norm": 0.38444114269586827,
+      "learning_rate": 8.275953316252946e-06,
+      "loss": 0.495,
+      "step": 5819
+    },
+    {
+      "epoch": 0.9558023525547595,
+      "grad_norm": 0.2708150920738918,
+      "learning_rate": 8.275698186099278e-06,
+      "loss": 0.4782,
+      "step": 5820
+    },
+    {
+      "epoch": 0.9559665797631023,
+      "grad_norm": 0.574284020674942,
+      "learning_rate": 8.27544301493781e-06,
+      "loss": 0.5015,
+      "step": 5821
+    },
+    {
+      "epoch": 0.956130806971445,
+      "grad_norm": 0.3558660647373787,
+      "learning_rate": 8.27518780277132e-06,
+      "loss": 0.4873,
+      "step": 5822
+    },
+    {
+      "epoch": 0.9562950341797878,
+      "grad_norm": 0.2853300694877287,
+      "learning_rate": 8.274932549602575e-06,
+      "loss": 0.5086,
+      "step": 5823
+    },
+    {
+      "epoch": 0.9564592613881305,
+      "grad_norm": 0.291071201624435,
+      "learning_rate": 8.274677255434348e-06,
+      "loss": 0.5011,
+      "step": 5824
+    },
+    {
+      "epoch": 0.9566234885964732,
+      "grad_norm": 0.3247243607708094,
+      "learning_rate": 8.274421920269412e-06,
+      "loss": 0.5057,
+      "step": 5825
+    },
+    {
+      "epoch": 0.9567877158048159,
+      "grad_norm": 0.26119747354567163,
+      "learning_rate": 8.274166544110541e-06,
+      "loss": 0.4932,
+      "step": 5826
+    },
+    {
+      "epoch": 0.9569519430131587,
+      "grad_norm": 0.3785889300892869,
+      "learning_rate": 8.273911126960507e-06,
+      "loss": 0.4927,
+      "step": 5827
+    },
+    {
+      "epoch": 0.9571161702215014,
+      "grad_norm": 0.38237933514300165,
+      "learning_rate": 8.273655668822086e-06,
+      "loss": 0.4997,
+      "step": 5828
+    },
+    {
+      "epoch": 0.9572803974298442,
+      "grad_norm": 0.3519938306470608,
+      "learning_rate": 8.273400169698051e-06,
+      "loss": 0.4969,
+      "step": 5829
+    },
+    {
+      "epoch": 0.9574446246381869,
+      "grad_norm": 0.2765721380041462,
+      "learning_rate": 8.27314462959118e-06,
+      "loss": 0.4796,
+      "step": 5830
+    },
+    {
+      "epoch": 0.9576088518465297,
+      "grad_norm": 0.3239066959022419,
+      "learning_rate": 8.272889048504244e-06,
+      "loss": 0.4818,
+      "step": 5831
+    },
+    {
+      "epoch": 0.9577730790548724,
+      "grad_norm": 0.28322233184270795,
+      "learning_rate": 8.272633426440021e-06,
+      "loss": 0.4928,
+      "step": 5832
+    },
+    {
+      "epoch": 0.9579373062632152,
+      "grad_norm": 0.2633683826750771,
+      "learning_rate": 8.272377763401287e-06,
+      "loss": 0.4911,
+      "step": 5833
+    },
+    {
+      "epoch": 0.9581015334715579,
+      "grad_norm": 0.29772403363255917,
+      "learning_rate": 8.27212205939082e-06,
+      "loss": 0.4798,
+      "step": 5834
+    },
+    {
+      "epoch": 0.9582657606799007,
+      "grad_norm": 0.3413359910040477,
+      "learning_rate": 8.271866314411395e-06,
+      "loss": 0.5078,
+      "step": 5835
+    },
+    {
+      "epoch": 0.9584299878882434,
+      "grad_norm": 0.34028950114887185,
+      "learning_rate": 8.271610528465792e-06,
+      "loss": 0.5045,
+      "step": 5836
+    },
+    {
+      "epoch": 0.9585942150965862,
+      "grad_norm": 0.27175870163911775,
+      "learning_rate": 8.271354701556786e-06,
+      "loss": 0.5135,
+      "step": 5837
+    },
+    {
+      "epoch": 0.9587584423049289,
+      "grad_norm": 0.3075850133116066,
+      "learning_rate": 8.27109883368716e-06,
+      "loss": 0.4803,
+      "step": 5838
+    },
+    {
+      "epoch": 0.9589226695132717,
+      "grad_norm": 0.3351562770791254,
+      "learning_rate": 8.270842924859688e-06,
+      "loss": 0.4922,
+      "step": 5839
+    },
+    {
+      "epoch": 0.9590868967216144,
+      "grad_norm": 0.2860224701466842,
+      "learning_rate": 8.270586975077154e-06,
+      "loss": 0.5001,
+      "step": 5840
+    },
+    {
+      "epoch": 0.9592511239299572,
+      "grad_norm": 0.2810690644684735,
+      "learning_rate": 8.270330984342334e-06,
+      "loss": 0.517,
+      "step": 5841
+    },
+    {
+      "epoch": 0.9594153511382998,
+      "grad_norm": 0.3647064774101022,
+      "learning_rate": 8.27007495265801e-06,
+      "loss": 0.5175,
+      "step": 5842
+    },
+    {
+      "epoch": 0.9595795783466425,
+      "grad_norm": 0.29633901351977443,
+      "learning_rate": 8.269818880026963e-06,
+      "loss": 0.5043,
+      "step": 5843
+    },
+    {
+      "epoch": 0.9597438055549853,
+      "grad_norm": 0.31863430795181835,
+      "learning_rate": 8.269562766451974e-06,
+      "loss": 0.5044,
+      "step": 5844
+    },
+    {
+      "epoch": 0.959908032763328,
+      "grad_norm": 0.288021139963858,
+      "learning_rate": 8.269306611935826e-06,
+      "loss": 0.486,
+      "step": 5845
+    },
+    {
+      "epoch": 0.9600722599716708,
+      "grad_norm": 0.40002756696662767,
+      "learning_rate": 8.269050416481298e-06,
+      "loss": 0.4917,
+      "step": 5846
+    },
+    {
+      "epoch": 0.9602364871800135,
+      "grad_norm": 0.2847656625043193,
+      "learning_rate": 8.268794180091175e-06,
+      "loss": 0.5,
+      "step": 5847
+    },
+    {
+      "epoch": 0.9604007143883563,
+      "grad_norm": 0.2876921231037368,
+      "learning_rate": 8.268537902768239e-06,
+      "loss": 0.4946,
+      "step": 5848
+    },
+    {
+      "epoch": 0.960564941596699,
+      "grad_norm": 0.3600375948303123,
+      "learning_rate": 8.268281584515273e-06,
+      "loss": 0.5072,
+      "step": 5849
+    },
+    {
+      "epoch": 0.9607291688050418,
+      "grad_norm": 0.27348127589456944,
+      "learning_rate": 8.268025225335063e-06,
+      "loss": 0.5142,
+      "step": 5850
+    },
+    {
+      "epoch": 0.9608933960133845,
+      "grad_norm": 0.29609853935174174,
+      "learning_rate": 8.267768825230392e-06,
+      "loss": 0.4903,
+      "step": 5851
+    },
+    {
+      "epoch": 0.9610576232217273,
+      "grad_norm": 0.3485982841735851,
+      "learning_rate": 8.267512384204043e-06,
+      "loss": 0.5078,
+      "step": 5852
+    },
+    {
+      "epoch": 0.96122185043007,
+      "grad_norm": 0.31985530611185276,
+      "learning_rate": 8.267255902258804e-06,
+      "loss": 0.4809,
+      "step": 5853
+    },
+    {
+      "epoch": 0.9613860776384128,
+      "grad_norm": 0.2827682502321915,
+      "learning_rate": 8.266999379397458e-06,
+      "loss": 0.5171,
+      "step": 5854
+    },
+    {
+      "epoch": 0.9615503048467555,
+      "grad_norm": 0.3558347178567936,
+      "learning_rate": 8.266742815622794e-06,
+      "loss": 0.5205,
+      "step": 5855
+    },
+    {
+      "epoch": 0.9617145320550983,
+      "grad_norm": 0.31569356701732815,
+      "learning_rate": 8.266486210937595e-06,
+      "loss": 0.488,
+      "step": 5856
+    },
+    {
+      "epoch": 0.961878759263441,
+      "grad_norm": 0.280486740752955,
+      "learning_rate": 8.266229565344651e-06,
+      "loss": 0.5223,
+      "step": 5857
+    },
+    {
+      "epoch": 0.9620429864717838,
+      "grad_norm": 0.30616030690303003,
+      "learning_rate": 8.265972878846751e-06,
+      "loss": 0.495,
+      "step": 5858
+    },
+    {
+      "epoch": 0.9622072136801264,
+      "grad_norm": 0.2971057720358319,
+      "learning_rate": 8.265716151446677e-06,
+      "loss": 0.4886,
+      "step": 5859
+    },
+    {
+      "epoch": 0.9623714408884692,
+      "grad_norm": 0.4860466387593598,
+      "learning_rate": 8.26545938314722e-06,
+      "loss": 0.4974,
+      "step": 5860
+    },
+    {
+      "epoch": 0.9625356680968119,
+      "grad_norm": 0.31491956302389884,
+      "learning_rate": 8.265202573951172e-06,
+      "loss": 0.4915,
+      "step": 5861
+    },
+    {
+      "epoch": 0.9626998953051547,
+      "grad_norm": 0.5128107426900643,
+      "learning_rate": 8.26494572386132e-06,
+      "loss": 0.5052,
+      "step": 5862
+    },
+    {
+      "epoch": 0.9628641225134974,
+      "grad_norm": 0.28645208841438113,
+      "learning_rate": 8.264688832880453e-06,
+      "loss": 0.4772,
+      "step": 5863
+    },
+    {
+      "epoch": 0.9630283497218401,
+      "grad_norm": 0.3334911284628625,
+      "learning_rate": 8.264431901011358e-06,
+      "loss": 0.5136,
+      "step": 5864
+    },
+    {
+      "epoch": 0.9631925769301829,
+      "grad_norm": 0.272860391882876,
+      "learning_rate": 8.264174928256832e-06,
+      "loss": 0.5046,
+      "step": 5865
+    },
+    {
+      "epoch": 0.9633568041385256,
+      "grad_norm": 0.3606448704975928,
+      "learning_rate": 8.263917914619662e-06,
+      "loss": 0.4904,
+      "step": 5866
+    },
+    {
+      "epoch": 0.9635210313468684,
+      "grad_norm": 0.3258598539123226,
+      "learning_rate": 8.263660860102641e-06,
+      "loss": 0.4854,
+      "step": 5867
+    },
+    {
+      "epoch": 0.9636852585552111,
+      "grad_norm": 0.28589834527541785,
+      "learning_rate": 8.26340376470856e-06,
+      "loss": 0.5102,
+      "step": 5868
+    },
+    {
+      "epoch": 0.9638494857635539,
+      "grad_norm": 0.2781441103589388,
+      "learning_rate": 8.26314662844021e-06,
+      "loss": 0.5007,
+      "step": 5869
+    },
+    {
+      "epoch": 0.9640137129718966,
+      "grad_norm": 0.2598080441340957,
+      "learning_rate": 8.262889451300386e-06,
+      "loss": 0.4703,
+      "step": 5870
+    },
+    {
+      "epoch": 0.9641779401802394,
+      "grad_norm": 0.3118870916093809,
+      "learning_rate": 8.26263223329188e-06,
+      "loss": 0.4806,
+      "step": 5871
+    },
+    {
+      "epoch": 0.9643421673885821,
+      "grad_norm": 0.28664634469887024,
+      "learning_rate": 8.262374974417486e-06,
+      "loss": 0.5147,
+      "step": 5872
+    },
+    {
+      "epoch": 0.9645063945969249,
+      "grad_norm": 0.29088471375612124,
+      "learning_rate": 8.26211767468e-06,
+      "loss": 0.4965,
+      "step": 5873
+    },
+    {
+      "epoch": 0.9646706218052676,
+      "grad_norm": 0.2934428904263493,
+      "learning_rate": 8.261860334082212e-06,
+      "loss": 0.4904,
+      "step": 5874
+    },
+    {
+      "epoch": 0.9648348490136104,
+      "grad_norm": 0.3262898836978391,
+      "learning_rate": 8.26160295262692e-06,
+      "loss": 0.4977,
+      "step": 5875
+    },
+    {
+      "epoch": 0.964999076221953,
+      "grad_norm": 0.28309291470955317,
+      "learning_rate": 8.26134553031692e-06,
+      "loss": 0.4942,
+      "step": 5876
+    },
+    {
+      "epoch": 0.9651633034302958,
+      "grad_norm": 0.2727890246165504,
+      "learning_rate": 8.261088067155008e-06,
+      "loss": 0.4947,
+      "step": 5877
+    },
+    {
+      "epoch": 0.9653275306386385,
+      "grad_norm": 0.28695720943171166,
+      "learning_rate": 8.260830563143976e-06,
+      "loss": 0.5002,
+      "step": 5878
+    },
+    {
+      "epoch": 0.9654917578469813,
+      "grad_norm": 0.5454995618657891,
+      "learning_rate": 8.260573018286626e-06,
+      "loss": 0.4919,
+      "step": 5879
+    },
+    {
+      "epoch": 0.965655985055324,
+      "grad_norm": 0.3329320093126202,
+      "learning_rate": 8.260315432585754e-06,
+      "loss": 0.5237,
+      "step": 5880
+    },
+    {
+      "epoch": 0.9658202122636668,
+      "grad_norm": 0.3253192711895212,
+      "learning_rate": 8.260057806044155e-06,
+      "loss": 0.5048,
+      "step": 5881
+    },
+    {
+      "epoch": 0.9659844394720095,
+      "grad_norm": 0.30670573862884587,
+      "learning_rate": 8.259800138664628e-06,
+      "loss": 0.4797,
+      "step": 5882
+    },
+    {
+      "epoch": 0.9661486666803523,
+      "grad_norm": 0.30624801488520037,
+      "learning_rate": 8.259542430449975e-06,
+      "loss": 0.4847,
+      "step": 5883
+    },
+    {
+      "epoch": 0.966312893888695,
+      "grad_norm": 0.29449002020199405,
+      "learning_rate": 8.259284681402992e-06,
+      "loss": 0.5065,
+      "step": 5884
+    },
+    {
+      "epoch": 0.9664771210970378,
+      "grad_norm": 0.30474092018275056,
+      "learning_rate": 8.259026891526478e-06,
+      "loss": 0.4815,
+      "step": 5885
+    },
+    {
+      "epoch": 0.9666413483053805,
+      "grad_norm": 0.284568010057758,
+      "learning_rate": 8.258769060823232e-06,
+      "loss": 0.486,
+      "step": 5886
+    },
+    {
+      "epoch": 0.9668055755137233,
+      "grad_norm": 0.32753086529797554,
+      "learning_rate": 8.258511189296057e-06,
+      "loss": 0.483,
+      "step": 5887
+    },
+    {
+      "epoch": 0.966969802722066,
+      "grad_norm": 0.2446509694431007,
+      "learning_rate": 8.258253276947752e-06,
+      "loss": 0.4698,
+      "step": 5888
+    },
+    {
+      "epoch": 0.9671340299304088,
+      "grad_norm": 0.28462449616187135,
+      "learning_rate": 8.257995323781122e-06,
+      "loss": 0.4873,
+      "step": 5889
+    },
+    {
+      "epoch": 0.9672982571387515,
+      "grad_norm": 0.26788096167237824,
+      "learning_rate": 8.257737329798961e-06,
+      "loss": 0.5108,
+      "step": 5890
+    },
+    {
+      "epoch": 0.9674624843470943,
+      "grad_norm": 0.27341395472584296,
+      "learning_rate": 8.257479295004079e-06,
+      "loss": 0.4932,
+      "step": 5891
+    },
+    {
+      "epoch": 0.967626711555437,
+      "grad_norm": 0.30169532415899364,
+      "learning_rate": 8.257221219399272e-06,
+      "loss": 0.4656,
+      "step": 5892
+    },
+    {
+      "epoch": 0.9677909387637796,
+      "grad_norm": 2.138078420857035,
+      "learning_rate": 8.256963102987349e-06,
+      "loss": 0.5071,
+      "step": 5893
+    },
+    {
+      "epoch": 0.9679551659721224,
+      "grad_norm": 0.29300939357485084,
+      "learning_rate": 8.256704945771108e-06,
+      "loss": 0.4886,
+      "step": 5894
+    },
+    {
+      "epoch": 0.9681193931804651,
+      "grad_norm": 0.2819584830142459,
+      "learning_rate": 8.256446747753356e-06,
+      "loss": 0.4812,
+      "step": 5895
+    },
+    {
+      "epoch": 0.9682836203888079,
+      "grad_norm": 0.30045867899014606,
+      "learning_rate": 8.256188508936896e-06,
+      "loss": 0.4915,
+      "step": 5896
+    },
+    {
+      "epoch": 0.9684478475971506,
+      "grad_norm": 0.7846435366223166,
+      "learning_rate": 8.255930229324535e-06,
+      "loss": 0.5115,
+      "step": 5897
+    },
+    {
+      "epoch": 0.9686120748054934,
+      "grad_norm": 0.330147198674231,
+      "learning_rate": 8.255671908919075e-06,
+      "loss": 0.4656,
+      "step": 5898
+    },
+    {
+      "epoch": 0.9687763020138361,
+      "grad_norm": 0.2768259882158433,
+      "learning_rate": 8.255413547723323e-06,
+      "loss": 0.5064,
+      "step": 5899
+    },
+    {
+      "epoch": 0.9689405292221789,
+      "grad_norm": 0.3766778545533828,
+      "learning_rate": 8.255155145740084e-06,
+      "loss": 0.5089,
+      "step": 5900
+    },
+    {
+      "epoch": 0.9691047564305216,
+      "grad_norm": 0.4340928286077416,
+      "learning_rate": 8.254896702972167e-06,
+      "loss": 0.5171,
+      "step": 5901
+    },
+    {
+      "epoch": 0.9692689836388644,
+      "grad_norm": 0.32064135591269965,
+      "learning_rate": 8.254638219422378e-06,
+      "loss": 0.4786,
+      "step": 5902
+    },
+    {
+      "epoch": 0.9694332108472071,
+      "grad_norm": 0.35896251714917476,
+      "learning_rate": 8.254379695093523e-06,
+      "loss": 0.4763,
+      "step": 5903
+    },
+    {
+      "epoch": 0.9695974380555499,
+      "grad_norm": 0.28803051965097765,
+      "learning_rate": 8.25412112998841e-06,
+      "loss": 0.498,
+      "step": 5904
+    },
+    {
+      "epoch": 0.9697616652638926,
+      "grad_norm": 0.2843031109294082,
+      "learning_rate": 8.253862524109849e-06,
+      "loss": 0.5165,
+      "step": 5905
+    },
+    {
+      "epoch": 0.9699258924722354,
+      "grad_norm": 0.2703477212769139,
+      "learning_rate": 8.253603877460647e-06,
+      "loss": 0.5026,
+      "step": 5906
+    },
+    {
+      "epoch": 0.9700901196805781,
+      "grad_norm": 0.37540094540547747,
+      "learning_rate": 8.253345190043613e-06,
+      "loss": 0.4745,
+      "step": 5907
+    },
+    {
+      "epoch": 0.9702543468889209,
+      "grad_norm": 0.34686637153638844,
+      "learning_rate": 8.253086461861561e-06,
+      "loss": 0.5061,
+      "step": 5908
+    },
+    {
+      "epoch": 0.9704185740972635,
+      "grad_norm": 0.28355130217507657,
+      "learning_rate": 8.252827692917295e-06,
+      "loss": 0.4877,
+      "step": 5909
+    },
+    {
+      "epoch": 0.9705828013056063,
+      "grad_norm": 0.3208759472372884,
+      "learning_rate": 8.252568883213628e-06,
+      "loss": 0.5004,
+      "step": 5910
+    },
+    {
+      "epoch": 0.970747028513949,
+      "grad_norm": 0.27097365741813967,
+      "learning_rate": 8.25231003275337e-06,
+      "loss": 0.5167,
+      "step": 5911
+    },
+    {
+      "epoch": 0.9709112557222918,
+      "grad_norm": 0.26847747863284094,
+      "learning_rate": 8.252051141539335e-06,
+      "loss": 0.5085,
+      "step": 5912
+    },
+    {
+      "epoch": 0.9710754829306345,
+      "grad_norm": 0.2744476278316739,
+      "learning_rate": 8.251792209574333e-06,
+      "loss": 0.4946,
+      "step": 5913
+    },
+    {
+      "epoch": 0.9712397101389773,
+      "grad_norm": 0.2949039780636973,
+      "learning_rate": 8.251533236861175e-06,
+      "loss": 0.5,
+      "step": 5914
+    },
+    {
+      "epoch": 0.97140393734732,
+      "grad_norm": 0.2546736861392502,
+      "learning_rate": 8.251274223402676e-06,
+      "loss": 0.4998,
+      "step": 5915
+    },
+    {
+      "epoch": 0.9715681645556627,
+      "grad_norm": 0.33011871520716857,
+      "learning_rate": 8.251015169201649e-06,
+      "loss": 0.4979,
+      "step": 5916
+    },
+    {
+      "epoch": 0.9717323917640055,
+      "grad_norm": 0.36791096342909235,
+      "learning_rate": 8.250756074260903e-06,
+      "loss": 0.5164,
+      "step": 5917
+    },
+    {
+      "epoch": 0.9718966189723482,
+      "grad_norm": 0.2940151689618452,
+      "learning_rate": 8.25049693858326e-06,
+      "loss": 0.4945,
+      "step": 5918
+    },
+    {
+      "epoch": 0.972060846180691,
+      "grad_norm": 0.30790865804305934,
+      "learning_rate": 8.250237762171527e-06,
+      "loss": 0.4901,
+      "step": 5919
+    },
+    {
+      "epoch": 0.9722250733890337,
+      "grad_norm": 0.30633815270914644,
+      "learning_rate": 8.249978545028526e-06,
+      "loss": 0.4894,
+      "step": 5920
+    },
+    {
+      "epoch": 0.9723893005973765,
+      "grad_norm": 0.29364052405576624,
+      "learning_rate": 8.249719287157066e-06,
+      "loss": 0.5001,
+      "step": 5921
+    },
+    {
+      "epoch": 0.9725535278057192,
+      "grad_norm": 0.368512399902772,
+      "learning_rate": 8.249459988559965e-06,
+      "loss": 0.4937,
+      "step": 5922
+    },
+    {
+      "epoch": 0.972717755014062,
+      "grad_norm": 0.2963704816857771,
+      "learning_rate": 8.249200649240041e-06,
+      "loss": 0.4981,
+      "step": 5923
+    },
+    {
+      "epoch": 0.9728819822224047,
+      "grad_norm": 0.3420546608598498,
+      "learning_rate": 8.248941269200109e-06,
+      "loss": 0.4944,
+      "step": 5924
+    },
+    {
+      "epoch": 0.9730462094307475,
+      "grad_norm": 0.26229608477987965,
+      "learning_rate": 8.248681848442985e-06,
+      "loss": 0.5222,
+      "step": 5925
+    },
+    {
+      "epoch": 0.9732104366390901,
+      "grad_norm": 0.3123298058628242,
+      "learning_rate": 8.248422386971489e-06,
+      "loss": 0.4972,
+      "step": 5926
+    },
+    {
+      "epoch": 0.9733746638474329,
+      "grad_norm": 0.2899224102116319,
+      "learning_rate": 8.248162884788437e-06,
+      "loss": 0.5155,
+      "step": 5927
+    },
+    {
+      "epoch": 0.9735388910557756,
+      "grad_norm": 0.2642905460836601,
+      "learning_rate": 8.24790334189665e-06,
+      "loss": 0.5054,
+      "step": 5928
+    },
+    {
+      "epoch": 0.9737031182641184,
+      "grad_norm": 0.2811233637741677,
+      "learning_rate": 8.247643758298943e-06,
+      "loss": 0.5033,
+      "step": 5929
+    },
+    {
+      "epoch": 0.9738673454724611,
+      "grad_norm": 0.3921209165058693,
+      "learning_rate": 8.24738413399814e-06,
+      "loss": 0.4756,
+      "step": 5930
+    },
+    {
+      "epoch": 0.9740315726808039,
+      "grad_norm": 0.4300011495238529,
+      "learning_rate": 8.247124468997057e-06,
+      "loss": 0.4726,
+      "step": 5931
+    },
+    {
+      "epoch": 0.9741957998891466,
+      "grad_norm": 0.32118620284015836,
+      "learning_rate": 8.246864763298516e-06,
+      "loss": 0.5117,
+      "step": 5932
+    },
+    {
+      "epoch": 0.9743600270974894,
+      "grad_norm": 0.2675433845486991,
+      "learning_rate": 8.246605016905338e-06,
+      "loss": 0.4906,
+      "step": 5933
+    },
+    {
+      "epoch": 0.9745242543058321,
+      "grad_norm": 0.3604906603250053,
+      "learning_rate": 8.246345229820341e-06,
+      "loss": 0.4758,
+      "step": 5934
+    },
+    {
+      "epoch": 0.9746884815141749,
+      "grad_norm": 0.2552549150791266,
+      "learning_rate": 8.246085402046351e-06,
+      "loss": 0.4811,
+      "step": 5935
+    },
+    {
+      "epoch": 0.9748527087225176,
+      "grad_norm": 0.27213223431139516,
+      "learning_rate": 8.245825533586188e-06,
+      "loss": 0.5028,
+      "step": 5936
+    },
+    {
+      "epoch": 0.9750169359308604,
+      "grad_norm": 0.2923501485405211,
+      "learning_rate": 8.245565624442674e-06,
+      "loss": 0.4761,
+      "step": 5937
+    },
+    {
+      "epoch": 0.9751811631392031,
+      "grad_norm": 0.25637703091485053,
+      "learning_rate": 8.245305674618631e-06,
+      "loss": 0.4957,
+      "step": 5938
+    },
+    {
+      "epoch": 0.9753453903475459,
+      "grad_norm": 0.29585731786389724,
+      "learning_rate": 8.245045684116885e-06,
+      "loss": 0.5024,
+      "step": 5939
+    },
+    {
+      "epoch": 0.9755096175558886,
+      "grad_norm": 0.3097492958132811,
+      "learning_rate": 8.244785652940257e-06,
+      "loss": 0.4911,
+      "step": 5940
+    },
+    {
+      "epoch": 0.9756738447642314,
+      "grad_norm": 0.2796684436959236,
+      "learning_rate": 8.244525581091574e-06,
+      "loss": 0.4873,
+      "step": 5941
+    },
+    {
+      "epoch": 0.9758380719725741,
+      "grad_norm": 0.4901344389329459,
+      "learning_rate": 8.244265468573657e-06,
+      "loss": 0.4952,
+      "step": 5942
+    },
+    {
+      "epoch": 0.9760022991809167,
+      "grad_norm": 0.29449038922514387,
+      "learning_rate": 8.244005315389335e-06,
+      "loss": 0.5011,
+      "step": 5943
+    },
+    {
+      "epoch": 0.9761665263892595,
+      "grad_norm": 0.2500820874976813,
+      "learning_rate": 8.24374512154143e-06,
+      "loss": 0.4833,
+      "step": 5944
+    },
+    {
+      "epoch": 0.9763307535976022,
+      "grad_norm": 0.2831071561547265,
+      "learning_rate": 8.24348488703277e-06,
+      "loss": 0.5094,
+      "step": 5945
+    },
+    {
+      "epoch": 0.976494980805945,
+      "grad_norm": 0.322064842547227,
+      "learning_rate": 8.243224611866182e-06,
+      "loss": 0.5126,
+      "step": 5946
+    },
+    {
+      "epoch": 0.9766592080142877,
+      "grad_norm": 0.2683354525422334,
+      "learning_rate": 8.242964296044494e-06,
+      "loss": 0.4987,
+      "step": 5947
+    },
+    {
+      "epoch": 0.9768234352226305,
+      "grad_norm": 0.2992209909395012,
+      "learning_rate": 8.242703939570527e-06,
+      "loss": 0.5231,
+      "step": 5948
+    },
+    {
+      "epoch": 0.9769876624309732,
+      "grad_norm": 0.3050729492504926,
+      "learning_rate": 8.242443542447115e-06,
+      "loss": 0.5136,
+      "step": 5949
+    },
+    {
+      "epoch": 0.977151889639316,
+      "grad_norm": 0.29155035443597166,
+      "learning_rate": 8.242183104677083e-06,
+      "loss": 0.4942,
+      "step": 5950
+    },
+    {
+      "epoch": 0.9773161168476587,
+      "grad_norm": 0.25878300704709084,
+      "learning_rate": 8.24192262626326e-06,
+      "loss": 0.4816,
+      "step": 5951
+    },
+    {
+      "epoch": 0.9774803440560015,
+      "grad_norm": 0.3087053142553389,
+      "learning_rate": 8.241662107208478e-06,
+      "loss": 0.5206,
+      "step": 5952
+    },
+    {
+      "epoch": 0.9776445712643442,
+      "grad_norm": 0.4975144518325369,
+      "learning_rate": 8.241401547515563e-06,
+      "loss": 0.4931,
+      "step": 5953
+    },
+    {
+      "epoch": 0.977808798472687,
+      "grad_norm": 0.3519450335069974,
+      "learning_rate": 8.241140947187347e-06,
+      "loss": 0.5137,
+      "step": 5954
+    },
+    {
+      "epoch": 0.9779730256810297,
+      "grad_norm": 0.27425219345136576,
+      "learning_rate": 8.240880306226659e-06,
+      "loss": 0.4938,
+      "step": 5955
+    },
+    {
+      "epoch": 0.9781372528893725,
+      "grad_norm": 0.313446231366959,
+      "learning_rate": 8.24061962463633e-06,
+      "loss": 0.4833,
+      "step": 5956
+    },
+    {
+      "epoch": 0.9783014800977152,
+      "grad_norm": 0.31047102932394477,
+      "learning_rate": 8.240358902419192e-06,
+      "loss": 0.5167,
+      "step": 5957
+    },
+    {
+      "epoch": 0.978465707306058,
+      "grad_norm": 0.3358474346649635,
+      "learning_rate": 8.240098139578076e-06,
+      "loss": 0.4916,
+      "step": 5958
+    },
+    {
+      "epoch": 0.9786299345144007,
+      "grad_norm": 0.2916965129858983,
+      "learning_rate": 8.239837336115814e-06,
+      "loss": 0.5071,
+      "step": 5959
+    },
+    {
+      "epoch": 0.9787941617227434,
+      "grad_norm": 0.3787787228362414,
+      "learning_rate": 8.23957649203524e-06,
+      "loss": 0.5196,
+      "step": 5960
+    },
+    {
+      "epoch": 0.9789583889310861,
+      "grad_norm": 0.3064291953677091,
+      "learning_rate": 8.239315607339186e-06,
+      "loss": 0.4962,
+      "step": 5961
+    },
+    {
+      "epoch": 0.9791226161394289,
+      "grad_norm": 0.2839286416723755,
+      "learning_rate": 8.239054682030485e-06,
+      "loss": 0.4926,
+      "step": 5962
+    },
+    {
+      "epoch": 0.9792868433477716,
+      "grad_norm": 0.28085264039852126,
+      "learning_rate": 8.238793716111971e-06,
+      "loss": 0.5022,
+      "step": 5963
+    },
+    {
+      "epoch": 0.9794510705561144,
+      "grad_norm": 0.33066819531167013,
+      "learning_rate": 8.23853270958648e-06,
+      "loss": 0.4936,
+      "step": 5964
+    },
+    {
+      "epoch": 0.9796152977644571,
+      "grad_norm": 0.26488196400063624,
+      "learning_rate": 8.238271662456844e-06,
+      "loss": 0.4916,
+      "step": 5965
+    },
+    {
+      "epoch": 0.9797795249727999,
+      "grad_norm": 0.3473924170842142,
+      "learning_rate": 8.2380105747259e-06,
+      "loss": 0.5109,
+      "step": 5966
+    },
+    {
+      "epoch": 0.9799437521811426,
+      "grad_norm": 0.3409154975608138,
+      "learning_rate": 8.237749446396485e-06,
+      "loss": 0.4985,
+      "step": 5967
+    },
+    {
+      "epoch": 0.9801079793894854,
+      "grad_norm": 0.3418439073219856,
+      "learning_rate": 8.237488277471433e-06,
+      "loss": 0.4969,
+      "step": 5968
+    },
+    {
+      "epoch": 0.9802722065978281,
+      "grad_norm": 0.2569144853172575,
+      "learning_rate": 8.237227067953581e-06,
+      "loss": 0.4764,
+      "step": 5969
+    },
+    {
+      "epoch": 0.9804364338061708,
+      "grad_norm": 0.31028003969795925,
+      "learning_rate": 8.236965817845766e-06,
+      "loss": 0.4886,
+      "step": 5970
+    },
+    {
+      "epoch": 0.9806006610145136,
+      "grad_norm": 0.254519739291749,
+      "learning_rate": 8.236704527150826e-06,
+      "loss": 0.4989,
+      "step": 5971
+    },
+    {
+      "epoch": 0.9807648882228563,
+      "grad_norm": 0.4203026362129426,
+      "learning_rate": 8.236443195871597e-06,
+      "loss": 0.4991,
+      "step": 5972
+    },
+    {
+      "epoch": 0.9809291154311991,
+      "grad_norm": 0.26955631129226254,
+      "learning_rate": 8.23618182401092e-06,
+      "loss": 0.5171,
+      "step": 5973
+    },
+    {
+      "epoch": 0.9810933426395418,
+      "grad_norm": 0.2675493381061797,
+      "learning_rate": 8.235920411571632e-06,
+      "loss": 0.5036,
+      "step": 5974
+    },
+    {
+      "epoch": 0.9812575698478846,
+      "grad_norm": 0.33819000674277316,
+      "learning_rate": 8.235658958556573e-06,
+      "loss": 0.4782,
+      "step": 5975
+    },
+    {
+      "epoch": 0.9814217970562273,
+      "grad_norm": 0.2732732202700863,
+      "learning_rate": 8.235397464968581e-06,
+      "loss": 0.517,
+      "step": 5976
+    },
+    {
+      "epoch": 0.98158602426457,
+      "grad_norm": 0.2927136697546449,
+      "learning_rate": 8.235135930810499e-06,
+      "loss": 0.4975,
+      "step": 5977
+    },
+    {
+      "epoch": 0.9817502514729127,
+      "grad_norm": 0.29528840625993147,
+      "learning_rate": 8.234874356085165e-06,
+      "loss": 0.5094,
+      "step": 5978
+    },
+    {
+      "epoch": 0.9819144786812555,
+      "grad_norm": 0.28678695381696284,
+      "learning_rate": 8.234612740795422e-06,
+      "loss": 0.5042,
+      "step": 5979
+    },
+    {
+      "epoch": 0.9820787058895982,
+      "grad_norm": 0.2947556284863859,
+      "learning_rate": 8.23435108494411e-06,
+      "loss": 0.4788,
+      "step": 5980
+    },
+    {
+      "epoch": 0.982242933097941,
+      "grad_norm": 0.32944895594123025,
+      "learning_rate": 8.23408938853407e-06,
+      "loss": 0.486,
+      "step": 5981
+    },
+    {
+      "epoch": 0.9824071603062837,
+      "grad_norm": 0.2849669802030251,
+      "learning_rate": 8.233827651568146e-06,
+      "loss": 0.5017,
+      "step": 5982
+    },
+    {
+      "epoch": 0.9825713875146265,
+      "grad_norm": 0.34083531204598366,
+      "learning_rate": 8.23356587404918e-06,
+      "loss": 0.4901,
+      "step": 5983
+    },
+    {
+      "epoch": 0.9827356147229692,
+      "grad_norm": 0.44565048901596793,
+      "learning_rate": 8.233304055980015e-06,
+      "loss": 0.5058,
+      "step": 5984
+    },
+    {
+      "epoch": 0.982899841931312,
+      "grad_norm": 0.3147569511729347,
+      "learning_rate": 8.233042197363495e-06,
+      "loss": 0.5062,
+      "step": 5985
+    },
+    {
+      "epoch": 0.9830640691396547,
+      "grad_norm": 0.30713689506371955,
+      "learning_rate": 8.232780298202464e-06,
+      "loss": 0.4924,
+      "step": 5986
+    },
+    {
+      "epoch": 0.9832282963479975,
+      "grad_norm": 0.35891472000084595,
+      "learning_rate": 8.232518358499768e-06,
+      "loss": 0.4979,
+      "step": 5987
+    },
+    {
+      "epoch": 0.9833925235563402,
+      "grad_norm": 0.2610523577041328,
+      "learning_rate": 8.232256378258248e-06,
+      "loss": 0.4785,
+      "step": 5988
+    },
+    {
+      "epoch": 0.983556750764683,
+      "grad_norm": 0.29778668994780577,
+      "learning_rate": 8.231994357480754e-06,
+      "loss": 0.5078,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9837209779730257,
+      "grad_norm": 0.37794507997054505,
+      "learning_rate": 8.231732296170127e-06,
+      "loss": 0.4875,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9838852051813685,
+      "grad_norm": 0.3080732467400163,
+      "learning_rate": 8.231470194329218e-06,
+      "loss": 0.5156,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9840494323897112,
+      "grad_norm": 0.31909404424037935,
+      "learning_rate": 8.23120805196087e-06,
+      "loss": 0.4996,
+      "step": 5992
+    },
+    {
+      "epoch": 0.984213659598054,
+      "grad_norm": 0.3035152405076065,
+      "learning_rate": 8.230945869067931e-06,
+      "loss": 0.4752,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9843778868063966,
+      "grad_norm": 0.3463376211275316,
+      "learning_rate": 8.23068364565325e-06,
+      "loss": 0.4903,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9845421140147393,
+      "grad_norm": 0.30248850655697274,
+      "learning_rate": 8.230421381719674e-06,
+      "loss": 0.5164,
+      "step": 5995
+    },
+    {
+      "epoch": 0.9847063412230821,
+      "grad_norm": 0.27197297877835336,
+      "learning_rate": 8.230159077270053e-06,
+      "loss": 0.5114,
+      "step": 5996
+    },
+    {
+      "epoch": 0.9848705684314248,
+      "grad_norm": 0.39503979748818113,
+      "learning_rate": 8.229896732307233e-06,
+      "loss": 0.4937,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9850347956397676,
+      "grad_norm": 0.27231037109643863,
+      "learning_rate": 8.229634346834064e-06,
+      "loss": 0.4854,
+      "step": 5998
+    },
+    {
+      "epoch": 0.9851990228481103,
+      "grad_norm": 0.2743727883763561,
+      "learning_rate": 8.229371920853399e-06,
+      "loss": 0.4894,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9853632500564531,
+      "grad_norm": 0.3098154188220738,
+      "learning_rate": 8.229109454368082e-06,
+      "loss": 0.4931,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9855274772647958,
+      "grad_norm": 0.31105934974202104,
+      "learning_rate": 8.22884694738097e-06,
+      "loss": 0.5277,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9856917044731386,
+      "grad_norm": 0.30333660429551784,
+      "learning_rate": 8.22858439989491e-06,
+      "loss": 0.4962,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9858559316814813,
+      "grad_norm": 0.28194793432475335,
+      "learning_rate": 8.228321811912757e-06,
+      "loss": 0.4693,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9860201588898241,
+      "grad_norm": 0.27521067135499894,
+      "learning_rate": 8.22805918343736e-06,
+      "loss": 0.4978,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9861843860981668,
+      "grad_norm": 0.3516121931850756,
+      "learning_rate": 8.227796514471571e-06,
+      "loss": 0.4923,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9863486133065096,
+      "grad_norm": 0.26710061007145536,
+      "learning_rate": 8.227533805018245e-06,
+      "loss": 0.5155,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9865128405148523,
+      "grad_norm": 0.29224389368396897,
+      "learning_rate": 8.227271055080236e-06,
+      "loss": 0.4866,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9866770677231951,
+      "grad_norm": 0.3534119652282733,
+      "learning_rate": 8.227008264660396e-06,
+      "loss": 0.5215,
+      "step": 6008
+    },
+    {
+      "epoch": 0.9868412949315378,
+      "grad_norm": 0.2726025456910224,
+      "learning_rate": 8.226745433761578e-06,
+      "loss": 0.5074,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9870055221398806,
+      "grad_norm": 0.3227584179564337,
+      "learning_rate": 8.226482562386638e-06,
+      "loss": 0.476,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9871697493482232,
+      "grad_norm": 0.2584528878550159,
+      "learning_rate": 8.226219650538432e-06,
+      "loss": 0.5026,
+      "step": 6011
+    },
+    {
+      "epoch": 0.987333976556566,
+      "grad_norm": 0.2983429827813741,
+      "learning_rate": 8.225956698219814e-06,
+      "loss": 0.5065,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9874982037649087,
+      "grad_norm": 0.26958664034702273,
+      "learning_rate": 8.225693705433639e-06,
+      "loss": 0.5034,
+      "step": 6013
+    },
+    {
+      "epoch": 0.9876624309732515,
+      "grad_norm": 0.3463746350662079,
+      "learning_rate": 8.225430672182768e-06,
+      "loss": 0.4991,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9878266581815942,
+      "grad_norm": 0.48975988570215695,
+      "learning_rate": 8.225167598470052e-06,
+      "loss": 0.4868,
+      "step": 6015
+    },
+    {
+      "epoch": 0.987990885389937,
+      "grad_norm": 0.28846554880418807,
+      "learning_rate": 8.22490448429835e-06,
+      "loss": 0.4941,
+      "step": 6016
+    },
+    {
+      "epoch": 0.9881551125982797,
+      "grad_norm": 0.25948125490643514,
+      "learning_rate": 8.224641329670522e-06,
+      "loss": 0.4822,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9883193398066225,
+      "grad_norm": 0.4795604048369903,
+      "learning_rate": 8.224378134589426e-06,
+      "loss": 0.4985,
+      "step": 6018
+    },
+    {
+      "epoch": 0.9884835670149652,
+      "grad_norm": 0.27508472004944673,
+      "learning_rate": 8.224114899057917e-06,
+      "loss": 0.4824,
+      "step": 6019
+    },
+    {
+      "epoch": 0.988647794223308,
+      "grad_norm": 0.2941581038633356,
+      "learning_rate": 8.223851623078856e-06,
+      "loss": 0.4916,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9888120214316507,
+      "grad_norm": 0.2857124408871979,
+      "learning_rate": 8.223588306655105e-06,
+      "loss": 0.4978,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9889762486399934,
+      "grad_norm": 0.28161753210610235,
+      "learning_rate": 8.22332494978952e-06,
+      "loss": 0.4917,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9891404758483362,
+      "grad_norm": 0.3404370427089407,
+      "learning_rate": 8.223061552484962e-06,
+      "loss": 0.5161,
+      "step": 6023
+    },
+    {
+      "epoch": 0.989304703056679,
+      "grad_norm": 0.29723498104859414,
+      "learning_rate": 8.222798114744294e-06,
+      "loss": 0.4953,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9894689302650217,
+      "grad_norm": 0.25821568853288496,
+      "learning_rate": 8.222534636570375e-06,
+      "loss": 0.4889,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9896331574733644,
+      "grad_norm": 0.34018544169293335,
+      "learning_rate": 8.222271117966067e-06,
+      "loss": 0.4987,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9897973846817072,
+      "grad_norm": 0.4406773775755572,
+      "learning_rate": 8.222007558934234e-06,
+      "loss": 0.5073,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9899616118900498,
+      "grad_norm": 0.5805674365221384,
+      "learning_rate": 8.221743959477735e-06,
+      "loss": 0.4827,
+      "step": 6028
+    },
+    {
+      "epoch": 0.9901258390983926,
+      "grad_norm": 0.2996142802822292,
+      "learning_rate": 8.221480319599435e-06,
+      "loss": 0.5045,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9902900663067353,
+      "grad_norm": 0.29344698377790773,
+      "learning_rate": 8.221216639302199e-06,
+      "loss": 0.4996,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9904542935150781,
+      "grad_norm": 0.32978471469877835,
+      "learning_rate": 8.220952918588888e-06,
+      "loss": 0.4832,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9906185207234208,
+      "grad_norm": 0.348936338155547,
+      "learning_rate": 8.220689157462367e-06,
+      "loss": 0.4999,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9907827479317636,
+      "grad_norm": 0.2929668547648481,
+      "learning_rate": 8.220425355925503e-06,
+      "loss": 0.4946,
+      "step": 6033
+    },
+    {
+      "epoch": 0.9909469751401063,
+      "grad_norm": 0.28969490861086555,
+      "learning_rate": 8.220161513981157e-06,
+      "loss": 0.508,
+      "step": 6034
+    },
+    {
+      "epoch": 0.9911112023484491,
+      "grad_norm": 0.30362512496995825,
+      "learning_rate": 8.219897631632197e-06,
+      "loss": 0.5036,
+      "step": 6035
+    },
+    {
+      "epoch": 0.9912754295567918,
+      "grad_norm": 0.6155640604324444,
+      "learning_rate": 8.21963370888149e-06,
+      "loss": 0.4919,
+      "step": 6036
+    },
+    {
+      "epoch": 0.9914396567651346,
+      "grad_norm": 0.37171675217371414,
+      "learning_rate": 8.219369745731901e-06,
+      "loss": 0.5144,
+      "step": 6037
+    },
+    {
+      "epoch": 0.9916038839734773,
+      "grad_norm": 0.285698244962252,
+      "learning_rate": 8.219105742186297e-06,
+      "loss": 0.5015,
+      "step": 6038
+    },
+    {
+      "epoch": 0.9917681111818201,
+      "grad_norm": 0.35864388593535235,
+      "learning_rate": 8.218841698247545e-06,
+      "loss": 0.512,
+      "step": 6039
+    },
+    {
+      "epoch": 0.9919323383901628,
+      "grad_norm": 0.277329370636313,
+      "learning_rate": 8.218577613918514e-06,
+      "loss": 0.4877,
+      "step": 6040
+    },
+    {
+      "epoch": 0.9920965655985056,
+      "grad_norm": 0.27621896049336875,
+      "learning_rate": 8.21831348920207e-06,
+      "loss": 0.4903,
+      "step": 6041
+    },
+    {
+      "epoch": 0.9922607928068483,
+      "grad_norm": 0.3112911663410362,
+      "learning_rate": 8.218049324101086e-06,
+      "loss": 0.4834,
+      "step": 6042
+    },
+    {
+      "epoch": 0.9924250200151911,
+      "grad_norm": 0.3014841204227243,
+      "learning_rate": 8.217785118618426e-06,
+      "loss": 0.4922,
+      "step": 6043
+    },
+    {
+      "epoch": 0.9925892472235338,
+      "grad_norm": 0.2767958028017139,
+      "learning_rate": 8.217520872756962e-06,
+      "loss": 0.4973,
+      "step": 6044
+    },
+    {
+      "epoch": 0.9927534744318764,
+      "grad_norm": 0.2918501643305302,
+      "learning_rate": 8.217256586519567e-06,
+      "loss": 0.5088,
+      "step": 6045
+    },
+    {
+      "epoch": 0.9929177016402192,
+      "grad_norm": 0.32321579020808244,
+      "learning_rate": 8.216992259909105e-06,
+      "loss": 0.4937,
+      "step": 6046
+    },
+    {
+      "epoch": 0.993081928848562,
+      "grad_norm": 0.29135334956770775,
+      "learning_rate": 8.21672789292845e-06,
+      "loss": 0.4867,
+      "step": 6047
+    },
+    {
+      "epoch": 0.9932461560569047,
+      "grad_norm": 0.2755437262202141,
+      "learning_rate": 8.216463485580474e-06,
+      "loss": 0.5026,
+      "step": 6048
+    },
+    {
+      "epoch": 0.9934103832652474,
+      "grad_norm": 0.28258408765853643,
+      "learning_rate": 8.216199037868048e-06,
+      "loss": 0.4922,
+      "step": 6049
+    },
+    {
+      "epoch": 0.9935746104735902,
+      "grad_norm": 0.2856610733524225,
+      "learning_rate": 8.215934549794043e-06,
+      "loss": 0.4934,
+      "step": 6050
+    },
+    {
+      "epoch": 0.9937388376819329,
+      "grad_norm": 0.26939264608863184,
+      "learning_rate": 8.215670021361335e-06,
+      "loss": 0.5089,
+      "step": 6051
+    },
+    {
+      "epoch": 0.9939030648902757,
+      "grad_norm": 0.26228522266134563,
+      "learning_rate": 8.215405452572793e-06,
+      "loss": 0.5082,
+      "step": 6052
+    },
+    {
+      "epoch": 0.9940672920986184,
+      "grad_norm": 0.2865666730968439,
+      "learning_rate": 8.215140843431293e-06,
+      "loss": 0.4986,
+      "step": 6053
+    },
+    {
+      "epoch": 0.9942315193069612,
+      "grad_norm": 0.29112888158896566,
+      "learning_rate": 8.21487619393971e-06,
+      "loss": 0.4927,
+      "step": 6054
+    },
+    {
+      "epoch": 0.9943957465153039,
+      "grad_norm": 0.32408251237526065,
+      "learning_rate": 8.214611504100914e-06,
+      "loss": 0.4819,
+      "step": 6055
+    },
+    {
+      "epoch": 0.9945599737236467,
+      "grad_norm": 0.2980623923723772,
+      "learning_rate": 8.214346773917784e-06,
+      "loss": 0.4951,
+      "step": 6056
+    },
+    {
+      "epoch": 0.9947242009319894,
+      "grad_norm": 0.3298708840715456,
+      "learning_rate": 8.214082003393193e-06,
+      "loss": 0.4933,
+      "step": 6057
+    },
+    {
+      "epoch": 0.9948884281403322,
+      "grad_norm": 0.2886454071627483,
+      "learning_rate": 8.213817192530015e-06,
+      "loss": 0.5084,
+      "step": 6058
+    },
+    {
+      "epoch": 0.9950526553486749,
+      "grad_norm": 0.26521275064187566,
+      "learning_rate": 8.213552341331133e-06,
+      "loss": 0.5027,
+      "step": 6059
+    },
+    {
+      "epoch": 0.9952168825570177,
+      "grad_norm": 0.35831906260849744,
+      "learning_rate": 8.213287449799416e-06,
+      "loss": 0.5002,
+      "step": 6060
+    },
+    {
+      "epoch": 0.9953811097653604,
+      "grad_norm": 0.49904049416821655,
+      "learning_rate": 8.213022517937744e-06,
+      "loss": 0.4946,
+      "step": 6061
+    },
+    {
+      "epoch": 0.9955453369737031,
+      "grad_norm": 0.273090783481902,
+      "learning_rate": 8.212757545748994e-06,
+      "loss": 0.5072,
+      "step": 6062
+    },
+    {
+      "epoch": 0.9957095641820458,
+      "grad_norm": 0.3750463965518032,
+      "learning_rate": 8.212492533236046e-06,
+      "loss": 0.493,
+      "step": 6063
+    },
+    {
+      "epoch": 0.9958737913903886,
+      "grad_norm": 0.2788278001894012,
+      "learning_rate": 8.212227480401774e-06,
+      "loss": 0.5033,
+      "step": 6064
+    },
+    {
+      "epoch": 0.9960380185987313,
+      "grad_norm": 0.25147500943496537,
+      "learning_rate": 8.211962387249062e-06,
+      "loss": 0.5144,
+      "step": 6065
+    },
+    {
+      "epoch": 0.9962022458070741,
+      "grad_norm": 0.32790265255873924,
+      "learning_rate": 8.211697253780785e-06,
+      "loss": 0.5175,
+      "step": 6066
+    },
+    {
+      "epoch": 0.9963664730154168,
+      "grad_norm": 0.3162485512645619,
+      "learning_rate": 8.211432079999824e-06,
+      "loss": 0.4795,
+      "step": 6067
+    },
+    {
+      "epoch": 0.9965307002237596,
+      "grad_norm": 0.31059518886608145,
+      "learning_rate": 8.211166865909058e-06,
+      "loss": 0.4959,
+      "step": 6068
+    },
+    {
+      "epoch": 0.9966949274321023,
+      "grad_norm": 0.34734654159685363,
+      "learning_rate": 8.210901611511371e-06,
+      "loss": 0.4801,
+      "step": 6069
+    },
+    {
+      "epoch": 0.996859154640445,
+      "grad_norm": 0.2947683331373121,
+      "learning_rate": 8.21063631680964e-06,
+      "loss": 0.5083,
+      "step": 6070
+    },
+    {
+      "epoch": 0.9970233818487878,
+      "grad_norm": 0.315138018398247,
+      "learning_rate": 8.210370981806748e-06,
+      "loss": 0.5118,
+      "step": 6071
+    },
+    {
+      "epoch": 0.9971876090571306,
+      "grad_norm": 0.33194756505125095,
+      "learning_rate": 8.210105606505577e-06,
+      "loss": 0.4917,
+      "step": 6072
+    },
+    {
+      "epoch": 0.9973518362654733,
+      "grad_norm": 0.3193908762263436,
+      "learning_rate": 8.209840190909009e-06,
+      "loss": 0.4971,
+      "step": 6073
+    },
+    {
+      "epoch": 0.997516063473816,
+      "grad_norm": 0.2793014691048366,
+      "learning_rate": 8.209574735019925e-06,
+      "loss": 0.4966,
+      "step": 6074
+    },
+    {
+      "epoch": 0.9976802906821588,
+      "grad_norm": 0.2966439416928268,
+      "learning_rate": 8.20930923884121e-06,
+      "loss": 0.5136,
+      "step": 6075
+    },
+    {
+      "epoch": 0.9978445178905015,
+      "grad_norm": 0.33803139976143903,
+      "learning_rate": 8.209043702375749e-06,
+      "loss": 0.5076,
+      "step": 6076
+    },
+    {
+      "epoch": 0.9980087450988443,
+      "grad_norm": 0.3729413860774098,
+      "learning_rate": 8.208778125626423e-06,
+      "loss": 0.4941,
+      "step": 6077
+    },
+    {
+      "epoch": 0.998172972307187,
+      "grad_norm": 0.29849360362127925,
+      "learning_rate": 8.208512508596118e-06,
+      "loss": 0.5089,
+      "step": 6078
+    },
+    {
+      "epoch": 0.9983371995155297,
+      "grad_norm": 0.3708493783333557,
+      "learning_rate": 8.208246851287717e-06,
+      "loss": 0.5017,
+      "step": 6079
+    },
+    {
+      "epoch": 0.9985014267238724,
+      "grad_norm": 0.2648040375203547,
+      "learning_rate": 8.207981153704108e-06,
+      "loss": 0.5087,
+      "step": 6080
+    },
+    {
+      "epoch": 0.9986656539322152,
+      "grad_norm": 0.330039886934558,
+      "learning_rate": 8.207715415848176e-06,
+      "loss": 0.4689,
+      "step": 6081
+    },
+    {
+      "epoch": 0.9988298811405579,
+      "grad_norm": 0.30651609282300624,
+      "learning_rate": 8.207449637722806e-06,
+      "loss": 0.4954,
+      "step": 6082
+    },
+    {
+      "epoch": 0.9989941083489007,
+      "grad_norm": 0.3019876534538926,
+      "learning_rate": 8.207183819330884e-06,
+      "loss": 0.5002,
+      "step": 6083
+    },
+    {
+      "epoch": 0.9991583355572434,
+      "grad_norm": 0.2769342143142423,
+      "learning_rate": 8.206917960675301e-06,
+      "loss": 0.5214,
+      "step": 6084
+    },
+    {
+      "epoch": 0.9993225627655862,
+      "grad_norm": 0.3556437500448538,
+      "learning_rate": 8.20665206175894e-06,
+      "loss": 0.4916,
+      "step": 6085
+    },
+    {
+      "epoch": 0.9994867899739289,
+      "grad_norm": 0.32734441118412566,
+      "learning_rate": 8.206386122584692e-06,
+      "loss": 0.476,
+      "step": 6086
+    },
+    {
+      "epoch": 0.9996510171822717,
+      "grad_norm": 0.33157470358823943,
+      "learning_rate": 8.206120143155443e-06,
+      "loss": 0.5219,
+      "step": 6087
+    },
+    {
+      "epoch": 0.9998152443906144,
+      "grad_norm": 0.3207924796762015,
+      "learning_rate": 8.205854123474083e-06,
+      "loss": 0.4969,
+      "step": 6088
+    },
+    {
+      "epoch": 0.9999794715989572,
+      "grad_norm": 0.268248980958867,
+      "learning_rate": 8.205588063543502e-06,
+      "loss": 0.4925,
+      "step": 6089
+    },
+    {
+      "epoch": 1.0001436988072998,
+      "grad_norm": 0.2701398571042544,
+      "learning_rate": 8.205321963366588e-06,
+      "loss": 0.4835,
+      "step": 6090
+    },
+    {
+      "epoch": 1.0003079260156427,
+      "grad_norm": 0.29488775351168806,
+      "learning_rate": 8.205055822946233e-06,
+      "loss": 0.4957,
+      "step": 6091
+    },
+    {
+      "epoch": 1.0004721532239853,
+      "grad_norm": 0.3105970052078029,
+      "learning_rate": 8.204789642285324e-06,
+      "loss": 0.4934,
+      "step": 6092
+    },
+    {
+      "epoch": 1.0006363804323282,
+      "grad_norm": 0.3507102653544947,
+      "learning_rate": 8.204523421386757e-06,
+      "loss": 0.4722,
+      "step": 6093
+    },
+    {
+      "epoch": 1.0008006076406708,
+      "grad_norm": 0.30553103864347814,
+      "learning_rate": 8.20425716025342e-06,
+      "loss": 0.4892,
+      "step": 6094
+    },
+    {
+      "epoch": 1.0009648348490137,
+      "grad_norm": 0.3301092842380056,
+      "learning_rate": 8.203990858888206e-06,
+      "loss": 0.4748,
+      "step": 6095
+    },
+    {
+      "epoch": 1.0011290620573563,
+      "grad_norm": 0.3285339943190183,
+      "learning_rate": 8.203724517294007e-06,
+      "loss": 0.5096,
+      "step": 6096
+    },
+    {
+      "epoch": 1.0012932892656992,
+      "grad_norm": 0.4022453312728853,
+      "learning_rate": 8.203458135473716e-06,
+      "loss": 0.482,
+      "step": 6097
+    },
+    {
+      "epoch": 1.0014575164740418,
+      "grad_norm": 0.3226651889997907,
+      "learning_rate": 8.203191713430225e-06,
+      "loss": 0.4779,
+      "step": 6098
+    },
+    {
+      "epoch": 1.0016217436823847,
+      "grad_norm": 0.32939182814617934,
+      "learning_rate": 8.20292525116643e-06,
+      "loss": 0.4972,
+      "step": 6099
+    },
+    {
+      "epoch": 1.0017859708907273,
+      "grad_norm": 0.3599122148730225,
+      "learning_rate": 8.202658748685223e-06,
+      "loss": 0.4785,
+      "step": 6100
+    },
+    {
+      "epoch": 1.0019501980990702,
+      "grad_norm": 0.26857771384618684,
+      "learning_rate": 8.202392205989498e-06,
+      "loss": 0.4919,
+      "step": 6101
+    },
+    {
+      "epoch": 1.0021144253074128,
+      "grad_norm": 0.25228675296021535,
+      "learning_rate": 8.202125623082151e-06,
+      "loss": 0.4978,
+      "step": 6102
+    },
+    {
+      "epoch": 1.0022786525157557,
+      "grad_norm": 0.2786770686176153,
+      "learning_rate": 8.20185899996608e-06,
+      "loss": 0.5098,
+      "step": 6103
+    },
+    {
+      "epoch": 1.0024428797240983,
+      "grad_norm": 0.27270211292454544,
+      "learning_rate": 8.201592336644176e-06,
+      "loss": 0.5052,
+      "step": 6104
+    },
+    {
+      "epoch": 1.002607106932441,
+      "grad_norm": 0.287144611347033,
+      "learning_rate": 8.201325633119337e-06,
+      "loss": 0.5038,
+      "step": 6105
+    },
+    {
+      "epoch": 1.0027713341407838,
+      "grad_norm": 0.336224498870284,
+      "learning_rate": 8.201058889394461e-06,
+      "loss": 0.4697,
+      "step": 6106
+    },
+    {
+      "epoch": 1.0029355613491264,
+      "grad_norm": 0.39484648633479735,
+      "learning_rate": 8.200792105472442e-06,
+      "loss": 0.5035,
+      "step": 6107
+    },
+    {
+      "epoch": 1.0030997885574693,
+      "grad_norm": 0.3712949110335043,
+      "learning_rate": 8.200525281356183e-06,
+      "loss": 0.4759,
+      "step": 6108
+    },
+    {
+      "epoch": 1.003264015765812,
+      "grad_norm": 0.270083642715779,
+      "learning_rate": 8.200258417048578e-06,
+      "loss": 0.5062,
+      "step": 6109
+    },
+    {
+      "epoch": 1.0034282429741548,
+      "grad_norm": 0.35149759365651945,
+      "learning_rate": 8.199991512552525e-06,
+      "loss": 0.485,
+      "step": 6110
+    },
+    {
+      "epoch": 1.0035924701824974,
+      "grad_norm": 0.27952436185145424,
+      "learning_rate": 8.199724567870923e-06,
+      "loss": 0.4849,
+      "step": 6111
+    },
+    {
+      "epoch": 1.0037566973908403,
+      "grad_norm": 0.31188337577290126,
+      "learning_rate": 8.199457583006671e-06,
+      "loss": 0.5232,
+      "step": 6112
+    },
+    {
+      "epoch": 1.003920924599183,
+      "grad_norm": 0.36877171718415647,
+      "learning_rate": 8.199190557962673e-06,
+      "loss": 0.4813,
+      "step": 6113
+    },
+    {
+      "epoch": 1.0040851518075258,
+      "grad_norm": 0.29489431860375687,
+      "learning_rate": 8.198923492741825e-06,
+      "loss": 0.5065,
+      "step": 6114
+    },
+    {
+      "epoch": 1.0042493790158684,
+      "grad_norm": 0.2871256782320963,
+      "learning_rate": 8.198656387347028e-06,
+      "loss": 0.4967,
+      "step": 6115
+    },
+    {
+      "epoch": 1.0044136062242113,
+      "grad_norm": 0.3253782587546533,
+      "learning_rate": 8.198389241781185e-06,
+      "loss": 0.5157,
+      "step": 6116
+    },
+    {
+      "epoch": 1.004577833432554,
+      "grad_norm": 0.2866600487798448,
+      "learning_rate": 8.198122056047195e-06,
+      "loss": 0.5116,
+      "step": 6117
+    },
+    {
+      "epoch": 1.0047420606408968,
+      "grad_norm": 0.3204737775521936,
+      "learning_rate": 8.197854830147961e-06,
+      "loss": 0.4931,
+      "step": 6118
+    },
+    {
+      "epoch": 1.0049062878492394,
+      "grad_norm": 0.305333919369849,
+      "learning_rate": 8.197587564086386e-06,
+      "loss": 0.495,
+      "step": 6119
+    },
+    {
+      "epoch": 1.0050705150575823,
+      "grad_norm": 0.3240184222494894,
+      "learning_rate": 8.19732025786537e-06,
+      "loss": 0.4901,
+      "step": 6120
+    },
+    {
+      "epoch": 1.005234742265925,
+      "grad_norm": 0.2586365842115682,
+      "learning_rate": 8.197052911487823e-06,
+      "loss": 0.5028,
+      "step": 6121
+    },
+    {
+      "epoch": 1.0053989694742675,
+      "grad_norm": 0.7875093702083333,
+      "learning_rate": 8.19678552495664e-06,
+      "loss": 0.4758,
+      "step": 6122
+    },
+    {
+      "epoch": 1.0055631966826104,
+      "grad_norm": 0.2959337138717507,
+      "learning_rate": 8.19651809827473e-06,
+      "loss": 0.5094,
+      "step": 6123
+    },
+    {
+      "epoch": 1.005727423890953,
+      "grad_norm": 0.36942449603120836,
+      "learning_rate": 8.196250631444996e-06,
+      "loss": 0.497,
+      "step": 6124
+    },
+    {
+      "epoch": 1.005891651099296,
+      "grad_norm": 0.2877734449312031,
+      "learning_rate": 8.195983124470346e-06,
+      "loss": 0.4854,
+      "step": 6125
+    },
+    {
+      "epoch": 1.0060558783076385,
+      "grad_norm": 0.285412931493051,
+      "learning_rate": 8.19571557735368e-06,
+      "loss": 0.4869,
+      "step": 6126
+    },
+    {
+      "epoch": 1.0062201055159814,
+      "grad_norm": 0.29427375436401393,
+      "learning_rate": 8.195447990097908e-06,
+      "loss": 0.4873,
+      "step": 6127
+    },
+    {
+      "epoch": 1.006384332724324,
+      "grad_norm": 0.3929275658433873,
+      "learning_rate": 8.195180362705935e-06,
+      "loss": 0.4779,
+      "step": 6128
+    },
+    {
+      "epoch": 1.006548559932667,
+      "grad_norm": 0.34591623515733444,
+      "learning_rate": 8.194912695180668e-06,
+      "loss": 0.494,
+      "step": 6129
+    },
+    {
+      "epoch": 1.0067127871410095,
+      "grad_norm": 0.30622477678448884,
+      "learning_rate": 8.194644987525013e-06,
+      "loss": 0.4757,
+      "step": 6130
+    },
+    {
+      "epoch": 1.0068770143493524,
+      "grad_norm": 0.29008846351927825,
+      "learning_rate": 8.194377239741879e-06,
+      "loss": 0.4972,
+      "step": 6131
+    },
+    {
+      "epoch": 1.007041241557695,
+      "grad_norm": 0.2901644160895777,
+      "learning_rate": 8.194109451834172e-06,
+      "loss": 0.4962,
+      "step": 6132
+    },
+    {
+      "epoch": 1.007205468766038,
+      "grad_norm": 0.39246471775157815,
+      "learning_rate": 8.193841623804803e-06,
+      "loss": 0.5022,
+      "step": 6133
+    },
+    {
+      "epoch": 1.0073696959743805,
+      "grad_norm": 0.33816828096835855,
+      "learning_rate": 8.193573755656681e-06,
+      "loss": 0.4755,
+      "step": 6134
+    },
+    {
+      "epoch": 1.0075339231827234,
+      "grad_norm": 0.2882989726136812,
+      "learning_rate": 8.193305847392713e-06,
+      "loss": 0.4939,
+      "step": 6135
+    },
+    {
+      "epoch": 1.007698150391066,
+      "grad_norm": 0.8477706290899284,
+      "learning_rate": 8.193037899015809e-06,
+      "loss": 0.4873,
+      "step": 6136
+    },
+    {
+      "epoch": 1.0078623775994089,
+      "grad_norm": 0.4657233345201176,
+      "learning_rate": 8.19276991052888e-06,
+      "loss": 0.489,
+      "step": 6137
+    },
+    {
+      "epoch": 1.0080266048077515,
+      "grad_norm": 0.3198771301938268,
+      "learning_rate": 8.192501881934838e-06,
+      "loss": 0.5226,
+      "step": 6138
+    },
+    {
+      "epoch": 1.0081908320160942,
+      "grad_norm": 0.37111830598748036,
+      "learning_rate": 8.192233813236591e-06,
+      "loss": 0.4778,
+      "step": 6139
+    },
+    {
+      "epoch": 1.008355059224437,
+      "grad_norm": 0.29241114461127055,
+      "learning_rate": 8.191965704437053e-06,
+      "loss": 0.4817,
+      "step": 6140
+    },
+    {
+      "epoch": 1.0085192864327797,
+      "grad_norm": 0.3005976492011157,
+      "learning_rate": 8.191697555539135e-06,
+      "loss": 0.4732,
+      "step": 6141
+    },
+    {
+      "epoch": 1.0086835136411225,
+      "grad_norm": 0.2967960749197657,
+      "learning_rate": 8.19142936654575e-06,
+      "loss": 0.4941,
+      "step": 6142
+    },
+    {
+      "epoch": 1.0088477408494652,
+      "grad_norm": 0.28884899020507654,
+      "learning_rate": 8.191161137459809e-06,
+      "loss": 0.4748,
+      "step": 6143
+    },
+    {
+      "epoch": 1.009011968057808,
+      "grad_norm": 0.525761885715074,
+      "learning_rate": 8.190892868284228e-06,
+      "loss": 0.4712,
+      "step": 6144
+    },
+    {
+      "epoch": 1.0091761952661507,
+      "grad_norm": 0.31060904710486237,
+      "learning_rate": 8.190624559021916e-06,
+      "loss": 0.4727,
+      "step": 6145
+    },
+    {
+      "epoch": 1.0093404224744935,
+      "grad_norm": 0.24722628302120625,
+      "learning_rate": 8.190356209675793e-06,
+      "loss": 0.4859,
+      "step": 6146
+    },
+    {
+      "epoch": 1.0095046496828362,
+      "grad_norm": 0.3700887463419303,
+      "learning_rate": 8.19008782024877e-06,
+      "loss": 0.4817,
+      "step": 6147
+    },
+    {
+      "epoch": 1.009668876891179,
+      "grad_norm": 0.28728444716481255,
+      "learning_rate": 8.189819390743762e-06,
+      "loss": 0.4808,
+      "step": 6148
+    },
+    {
+      "epoch": 1.0098331040995217,
+      "grad_norm": 0.29078167293740487,
+      "learning_rate": 8.189550921163685e-06,
+      "loss": 0.4983,
+      "step": 6149
+    },
+    {
+      "epoch": 1.0099973313078645,
+      "grad_norm": 0.3636550521570456,
+      "learning_rate": 8.189282411511457e-06,
+      "loss": 0.471,
+      "step": 6150
+    },
+    {
+      "epoch": 1.0101615585162071,
+      "grad_norm": 1.2961617563754146,
+      "learning_rate": 8.18901386178999e-06,
+      "loss": 0.499,
+      "step": 6151
+    },
+    {
+      "epoch": 1.01032578572455,
+      "grad_norm": 0.31558961032524924,
+      "learning_rate": 8.188745272002206e-06,
+      "loss": 0.5054,
+      "step": 6152
+    },
+    {
+      "epoch": 1.0104900129328926,
+      "grad_norm": 0.3187404488115447,
+      "learning_rate": 8.188476642151016e-06,
+      "loss": 0.4903,
+      "step": 6153
+    },
+    {
+      "epoch": 1.0106542401412355,
+      "grad_norm": 0.4396331603761133,
+      "learning_rate": 8.188207972239343e-06,
+      "loss": 0.5123,
+      "step": 6154
+    },
+    {
+      "epoch": 1.0108184673495781,
+      "grad_norm": 0.3098010717106893,
+      "learning_rate": 8.187939262270101e-06,
+      "loss": 0.5034,
+      "step": 6155
+    },
+    {
+      "epoch": 1.0109826945579208,
+      "grad_norm": 0.5654184895396511,
+      "learning_rate": 8.18767051224621e-06,
+      "loss": 0.494,
+      "step": 6156
+    },
+    {
+      "epoch": 1.0111469217662636,
+      "grad_norm": 0.3515168501435642,
+      "learning_rate": 8.18740172217059e-06,
+      "loss": 0.4978,
+      "step": 6157
+    },
+    {
+      "epoch": 1.0113111489746063,
+      "grad_norm": 0.3442198924300222,
+      "learning_rate": 8.18713289204616e-06,
+      "loss": 0.4948,
+      "step": 6158
+    },
+    {
+      "epoch": 1.0114753761829491,
+      "grad_norm": 0.2753283941014209,
+      "learning_rate": 8.186864021875836e-06,
+      "loss": 0.5203,
+      "step": 6159
+    },
+    {
+      "epoch": 1.0116396033912918,
+      "grad_norm": 0.27913090907463,
+      "learning_rate": 8.186595111662544e-06,
+      "loss": 0.4882,
+      "step": 6160
+    },
+    {
+      "epoch": 1.0118038305996346,
+      "grad_norm": 0.2762978360605873,
+      "learning_rate": 8.186326161409202e-06,
+      "loss": 0.4912,
+      "step": 6161
+    },
+    {
+      "epoch": 1.0119680578079773,
+      "grad_norm": 0.286562286465793,
+      "learning_rate": 8.186057171118731e-06,
+      "loss": 0.4877,
+      "step": 6162
+    },
+    {
+      "epoch": 1.0121322850163201,
+      "grad_norm": 0.33537335818622976,
+      "learning_rate": 8.185788140794053e-06,
+      "loss": 0.4776,
+      "step": 6163
+    },
+    {
+      "epoch": 1.0122965122246628,
+      "grad_norm": 0.39338628713320906,
+      "learning_rate": 8.18551907043809e-06,
+      "loss": 0.5107,
+      "step": 6164
+    },
+    {
+      "epoch": 1.0124607394330056,
+      "grad_norm": 0.2926564624467248,
+      "learning_rate": 8.18524996005376e-06,
+      "loss": 0.5054,
+      "step": 6165
+    },
+    {
+      "epoch": 1.0126249666413483,
+      "grad_norm": 0.3033449050165952,
+      "learning_rate": 8.184980809643992e-06,
+      "loss": 0.5057,
+      "step": 6166
+    },
+    {
+      "epoch": 1.0127891938496911,
+      "grad_norm": 0.34123205757993946,
+      "learning_rate": 8.184711619211708e-06,
+      "loss": 0.4784,
+      "step": 6167
+    },
+    {
+      "epoch": 1.0129534210580338,
+      "grad_norm": 0.4648784639014738,
+      "learning_rate": 8.18444238875983e-06,
+      "loss": 0.5068,
+      "step": 6168
+    },
+    {
+      "epoch": 1.0131176482663766,
+      "grad_norm": 0.3461201369794482,
+      "learning_rate": 8.184173118291282e-06,
+      "loss": 0.4855,
+      "step": 6169
+    },
+    {
+      "epoch": 1.0132818754747193,
+      "grad_norm": 0.3150133852802901,
+      "learning_rate": 8.183903807808989e-06,
+      "loss": 0.4996,
+      "step": 6170
+    },
+    {
+      "epoch": 1.0134461026830621,
+      "grad_norm": 0.401520522556826,
+      "learning_rate": 8.183634457315875e-06,
+      "loss": 0.4812,
+      "step": 6171
+    },
+    {
+      "epoch": 1.0136103298914048,
+      "grad_norm": 0.3923158095446052,
+      "learning_rate": 8.18336506681487e-06,
+      "loss": 0.4672,
+      "step": 6172
+    },
+    {
+      "epoch": 1.0137745570997474,
+      "grad_norm": 0.3516625662622814,
+      "learning_rate": 8.183095636308895e-06,
+      "loss": 0.5008,
+      "step": 6173
+    },
+    {
+      "epoch": 1.0139387843080903,
+      "grad_norm": 0.3250018573957792,
+      "learning_rate": 8.182826165800877e-06,
+      "loss": 0.5109,
+      "step": 6174
+    },
+    {
+      "epoch": 1.014103011516433,
+      "grad_norm": 0.28774650547836594,
+      "learning_rate": 8.182556655293743e-06,
+      "loss": 0.477,
+      "step": 6175
+    },
+    {
+      "epoch": 1.0142672387247758,
+      "grad_norm": 0.3476177081030143,
+      "learning_rate": 8.182287104790421e-06,
+      "loss": 0.4912,
+      "step": 6176
+    },
+    {
+      "epoch": 1.0144314659331184,
+      "grad_norm": 0.41693100296127283,
+      "learning_rate": 8.182017514293839e-06,
+      "loss": 0.5074,
+      "step": 6177
+    },
+    {
+      "epoch": 1.0145956931414613,
+      "grad_norm": 0.3266041736928415,
+      "learning_rate": 8.181747883806924e-06,
+      "loss": 0.4978,
+      "step": 6178
+    },
+    {
+      "epoch": 1.014759920349804,
+      "grad_norm": 0.3506095610631528,
+      "learning_rate": 8.181478213332604e-06,
+      "loss": 0.4827,
+      "step": 6179
+    },
+    {
+      "epoch": 1.0149241475581467,
+      "grad_norm": 0.2756591884041051,
+      "learning_rate": 8.181208502873809e-06,
+      "loss": 0.4806,
+      "step": 6180
+    },
+    {
+      "epoch": 1.0150883747664894,
+      "grad_norm": 0.2498743413864132,
+      "learning_rate": 8.180938752433467e-06,
+      "loss": 0.4937,
+      "step": 6181
+    },
+    {
+      "epoch": 1.0152526019748322,
+      "grad_norm": 0.32540974513603127,
+      "learning_rate": 8.180668962014509e-06,
+      "loss": 0.5095,
+      "step": 6182
+    },
+    {
+      "epoch": 1.0154168291831749,
+      "grad_norm": 0.29822184379944644,
+      "learning_rate": 8.180399131619865e-06,
+      "loss": 0.4799,
+      "step": 6183
+    },
+    {
+      "epoch": 1.0155810563915177,
+      "grad_norm": 0.29922201319644454,
+      "learning_rate": 8.180129261252465e-06,
+      "loss": 0.4823,
+      "step": 6184
+    },
+    {
+      "epoch": 1.0157452835998604,
+      "grad_norm": 0.3408181411078289,
+      "learning_rate": 8.17985935091524e-06,
+      "loss": 0.508,
+      "step": 6185
+    },
+    {
+      "epoch": 1.0159095108082032,
+      "grad_norm": 0.3017704206443717,
+      "learning_rate": 8.179589400611124e-06,
+      "loss": 0.4762,
+      "step": 6186
+    },
+    {
+      "epoch": 1.0160737380165459,
+      "grad_norm": 0.28087313141512144,
+      "learning_rate": 8.179319410343046e-06,
+      "loss": 0.4864,
+      "step": 6187
+    },
+    {
+      "epoch": 1.0162379652248887,
+      "grad_norm": 0.343968512634284,
+      "learning_rate": 8.17904938011394e-06,
+      "loss": 0.518,
+      "step": 6188
+    },
+    {
+      "epoch": 1.0164021924332314,
+      "grad_norm": 0.35943975526034677,
+      "learning_rate": 8.178779309926736e-06,
+      "loss": 0.4914,
+      "step": 6189
+    },
+    {
+      "epoch": 1.016566419641574,
+      "grad_norm": 0.30449402360044453,
+      "learning_rate": 8.17850919978437e-06,
+      "loss": 0.5151,
+      "step": 6190
+    },
+    {
+      "epoch": 1.0167306468499169,
+      "grad_norm": 0.4024272798430009,
+      "learning_rate": 8.178239049689776e-06,
+      "loss": 0.5001,
+      "step": 6191
+    },
+    {
+      "epoch": 1.0168948740582595,
+      "grad_norm": 0.42775826191503635,
+      "learning_rate": 8.177968859645886e-06,
+      "loss": 0.4927,
+      "step": 6192
+    },
+    {
+      "epoch": 1.0170591012666024,
+      "grad_norm": 0.3054641612310703,
+      "learning_rate": 8.177698629655635e-06,
+      "loss": 0.4729,
+      "step": 6193
+    },
+    {
+      "epoch": 1.017223328474945,
+      "grad_norm": 0.3662068760964567,
+      "learning_rate": 8.177428359721959e-06,
+      "loss": 0.4779,
+      "step": 6194
+    },
+    {
+      "epoch": 1.0173875556832879,
+      "grad_norm": 0.30360510720936423,
+      "learning_rate": 8.177158049847793e-06,
+      "loss": 0.4873,
+      "step": 6195
+    },
+    {
+      "epoch": 1.0175517828916305,
+      "grad_norm": 0.3055845582263593,
+      "learning_rate": 8.176887700036074e-06,
+      "loss": 0.5044,
+      "step": 6196
+    },
+    {
+      "epoch": 1.0177160100999734,
+      "grad_norm": 0.30314495024705285,
+      "learning_rate": 8.176617310289734e-06,
+      "loss": 0.5077,
+      "step": 6197
+    },
+    {
+      "epoch": 1.017880237308316,
+      "grad_norm": 0.27451553538847134,
+      "learning_rate": 8.176346880611716e-06,
+      "loss": 0.4883,
+      "step": 6198
+    },
+    {
+      "epoch": 1.0180444645166589,
+      "grad_norm": 0.29349998728599647,
+      "learning_rate": 8.176076411004954e-06,
+      "loss": 0.4968,
+      "step": 6199
+    },
+    {
+      "epoch": 1.0182086917250015,
+      "grad_norm": 0.690790330399613,
+      "learning_rate": 8.175805901472382e-06,
+      "loss": 0.5199,
+      "step": 6200
+    },
+    {
+      "epoch": 1.0183729189333444,
+      "grad_norm": 0.39758600380845943,
+      "learning_rate": 8.175535352016944e-06,
+      "loss": 0.5017,
+      "step": 6201
+    },
+    {
+      "epoch": 1.018537146141687,
+      "grad_norm": 0.30757472221998394,
+      "learning_rate": 8.175264762641575e-06,
+      "loss": 0.4902,
+      "step": 6202
+    },
+    {
+      "epoch": 1.0187013733500299,
+      "grad_norm": 0.2635507175010998,
+      "learning_rate": 8.174994133349214e-06,
+      "loss": 0.4966,
+      "step": 6203
+    },
+    {
+      "epoch": 1.0188656005583725,
+      "grad_norm": 0.345398957493199,
+      "learning_rate": 8.174723464142802e-06,
+      "loss": 0.4957,
+      "step": 6204
+    },
+    {
+      "epoch": 1.0190298277667154,
+      "grad_norm": 0.27142952139291604,
+      "learning_rate": 8.174452755025279e-06,
+      "loss": 0.4826,
+      "step": 6205
+    },
+    {
+      "epoch": 1.019194054975058,
+      "grad_norm": 0.2915572983361061,
+      "learning_rate": 8.174182005999583e-06,
+      "loss": 0.4927,
+      "step": 6206
+    },
+    {
+      "epoch": 1.0193582821834006,
+      "grad_norm": 0.27599364042760827,
+      "learning_rate": 8.173911217068654e-06,
+      "loss": 0.5096,
+      "step": 6207
+    },
+    {
+      "epoch": 1.0195225093917435,
+      "grad_norm": 0.29414455181960725,
+      "learning_rate": 8.173640388235436e-06,
+      "loss": 0.5038,
+      "step": 6208
+    },
+    {
+      "epoch": 1.0196867366000861,
+      "grad_norm": 0.342711202223622,
+      "learning_rate": 8.173369519502868e-06,
+      "loss": 0.4912,
+      "step": 6209
+    },
+    {
+      "epoch": 1.019850963808429,
+      "grad_norm": 0.2670830751046088,
+      "learning_rate": 8.173098610873893e-06,
+      "loss": 0.4981,
+      "step": 6210
+    },
+    {
+      "epoch": 1.0200151910167716,
+      "grad_norm": 0.2999878589401095,
+      "learning_rate": 8.172827662351455e-06,
+      "loss": 0.4759,
+      "step": 6211
+    },
+    {
+      "epoch": 1.0201794182251145,
+      "grad_norm": 0.31873822598064333,
+      "learning_rate": 8.172556673938493e-06,
+      "loss": 0.4731,
+      "step": 6212
+    },
+    {
+      "epoch": 1.0203436454334571,
+      "grad_norm": 0.2742035415040891,
+      "learning_rate": 8.172285645637952e-06,
+      "loss": 0.4674,
+      "step": 6213
+    },
+    {
+      "epoch": 1.0205078726418,
+      "grad_norm": 0.31574870554300666,
+      "learning_rate": 8.172014577452778e-06,
+      "loss": 0.4775,
+      "step": 6214
+    },
+    {
+      "epoch": 1.0206720998501426,
+      "grad_norm": 0.31740606222826007,
+      "learning_rate": 8.17174346938591e-06,
+      "loss": 0.5123,
+      "step": 6215
+    },
+    {
+      "epoch": 1.0208363270584855,
+      "grad_norm": 0.27353324703910803,
+      "learning_rate": 8.171472321440297e-06,
+      "loss": 0.4837,
+      "step": 6216
+    },
+    {
+      "epoch": 1.0210005542668281,
+      "grad_norm": 0.2944184923332211,
+      "learning_rate": 8.171201133618883e-06,
+      "loss": 0.4803,
+      "step": 6217
+    },
+    {
+      "epoch": 1.021164781475171,
+      "grad_norm": 0.3013537117744429,
+      "learning_rate": 8.170929905924613e-06,
+      "loss": 0.4995,
+      "step": 6218
+    },
+    {
+      "epoch": 1.0213290086835136,
+      "grad_norm": 0.44433352396382897,
+      "learning_rate": 8.17065863836043e-06,
+      "loss": 0.4857,
+      "step": 6219
+    },
+    {
+      "epoch": 1.0214932358918565,
+      "grad_norm": 0.363216757817997,
+      "learning_rate": 8.170387330929286e-06,
+      "loss": 0.4831,
+      "step": 6220
+    },
+    {
+      "epoch": 1.0216574631001991,
+      "grad_norm": 0.2788261710336551,
+      "learning_rate": 8.170115983634123e-06,
+      "loss": 0.484,
+      "step": 6221
+    },
+    {
+      "epoch": 1.021821690308542,
+      "grad_norm": 0.31049045914006534,
+      "learning_rate": 8.169844596477889e-06,
+      "loss": 0.4982,
+      "step": 6222
+    },
+    {
+      "epoch": 1.0219859175168846,
+      "grad_norm": 0.31160712706194205,
+      "learning_rate": 8.169573169463534e-06,
+      "loss": 0.5045,
+      "step": 6223
+    },
+    {
+      "epoch": 1.0221501447252272,
+      "grad_norm": 0.26787311763186605,
+      "learning_rate": 8.169301702594002e-06,
+      "loss": 0.4912,
+      "step": 6224
+    },
+    {
+      "epoch": 1.02231437193357,
+      "grad_norm": 0.28728606908000814,
+      "learning_rate": 8.169030195872242e-06,
+      "loss": 0.5129,
+      "step": 6225
+    },
+    {
+      "epoch": 1.0224785991419127,
+      "grad_norm": 0.4256925386572884,
+      "learning_rate": 8.168758649301207e-06,
+      "loss": 0.4969,
+      "step": 6226
+    },
+    {
+      "epoch": 1.0226428263502556,
+      "grad_norm": 0.2743511180790035,
+      "learning_rate": 8.168487062883842e-06,
+      "loss": 0.4909,
+      "step": 6227
+    },
+    {
+      "epoch": 1.0228070535585982,
+      "grad_norm": 0.3205816659724719,
+      "learning_rate": 8.168215436623099e-06,
+      "loss": 0.5116,
+      "step": 6228
+    },
+    {
+      "epoch": 1.022971280766941,
+      "grad_norm": 0.2797160146405526,
+      "learning_rate": 8.167943770521928e-06,
+      "loss": 0.4889,
+      "step": 6229
+    },
+    {
+      "epoch": 1.0231355079752837,
+      "grad_norm": 0.3298401016695871,
+      "learning_rate": 8.167672064583277e-06,
+      "loss": 0.4948,
+      "step": 6230
+    },
+    {
+      "epoch": 1.0232997351836266,
+      "grad_norm": 0.2820165824015003,
+      "learning_rate": 8.1674003188101e-06,
+      "loss": 0.4816,
+      "step": 6231
+    },
+    {
+      "epoch": 1.0234639623919692,
+      "grad_norm": 0.7516619960945273,
+      "learning_rate": 8.167128533205348e-06,
+      "loss": 0.4978,
+      "step": 6232
+    },
+    {
+      "epoch": 1.023628189600312,
+      "grad_norm": 0.3242252762838541,
+      "learning_rate": 8.16685670777197e-06,
+      "loss": 0.4701,
+      "step": 6233
+    },
+    {
+      "epoch": 1.0237924168086547,
+      "grad_norm": 0.28078186626292395,
+      "learning_rate": 8.166584842512922e-06,
+      "loss": 0.5035,
+      "step": 6234
+    },
+    {
+      "epoch": 1.0239566440169976,
+      "grad_norm": 0.2892632624945348,
+      "learning_rate": 8.166312937431154e-06,
+      "loss": 0.482,
+      "step": 6235
+    },
+    {
+      "epoch": 1.0241208712253402,
+      "grad_norm": 0.30068090471091746,
+      "learning_rate": 8.166040992529623e-06,
+      "loss": 0.4935,
+      "step": 6236
+    },
+    {
+      "epoch": 1.024285098433683,
+      "grad_norm": 0.30986046486983354,
+      "learning_rate": 8.165769007811278e-06,
+      "loss": 0.503,
+      "step": 6237
+    },
+    {
+      "epoch": 1.0244493256420257,
+      "grad_norm": 0.2652243823608064,
+      "learning_rate": 8.165496983279075e-06,
+      "loss": 0.4862,
+      "step": 6238
+    },
+    {
+      "epoch": 1.0246135528503686,
+      "grad_norm": 0.38129905728032715,
+      "learning_rate": 8.165224918935968e-06,
+      "loss": 0.5033,
+      "step": 6239
+    },
+    {
+      "epoch": 1.0247777800587112,
+      "grad_norm": 0.3330210906119507,
+      "learning_rate": 8.164952814784913e-06,
+      "loss": 0.4925,
+      "step": 6240
+    },
+    {
+      "epoch": 1.0249420072670539,
+      "grad_norm": 0.6285421145833283,
+      "learning_rate": 8.164680670828864e-06,
+      "loss": 0.4926,
+      "step": 6241
+    },
+    {
+      "epoch": 1.0251062344753967,
+      "grad_norm": 0.28295504725253984,
+      "learning_rate": 8.164408487070778e-06,
+      "loss": 0.4815,
+      "step": 6242
+    },
+    {
+      "epoch": 1.0252704616837394,
+      "grad_norm": 0.34828429823959156,
+      "learning_rate": 8.164136263513609e-06,
+      "loss": 0.4895,
+      "step": 6243
+    },
+    {
+      "epoch": 1.0254346888920822,
+      "grad_norm": 0.33668507831640215,
+      "learning_rate": 8.163864000160318e-06,
+      "loss": 0.5016,
+      "step": 6244
+    },
+    {
+      "epoch": 1.0255989161004249,
+      "grad_norm": 0.3371227103810928,
+      "learning_rate": 8.163591697013857e-06,
+      "loss": 0.4818,
+      "step": 6245
+    },
+    {
+      "epoch": 1.0257631433087677,
+      "grad_norm": 0.37006258498862343,
+      "learning_rate": 8.163319354077188e-06,
+      "loss": 0.4906,
+      "step": 6246
+    },
+    {
+      "epoch": 1.0259273705171104,
+      "grad_norm": 0.3321561343583107,
+      "learning_rate": 8.163046971353263e-06,
+      "loss": 0.4972,
+      "step": 6247
+    },
+    {
+      "epoch": 1.0260915977254532,
+      "grad_norm": 0.33758474607503053,
+      "learning_rate": 8.162774548845047e-06,
+      "loss": 0.4886,
+      "step": 6248
+    },
+    {
+      "epoch": 1.0262558249337959,
+      "grad_norm": 0.5719844368152448,
+      "learning_rate": 8.162502086555494e-06,
+      "loss": 0.4776,
+      "step": 6249
+    },
+    {
+      "epoch": 1.0264200521421387,
+      "grad_norm": 0.3230093959766538,
+      "learning_rate": 8.162229584487566e-06,
+      "loss": 0.4895,
+      "step": 6250
+    },
+    {
+      "epoch": 1.0265842793504814,
+      "grad_norm": 0.27381092738906676,
+      "learning_rate": 8.16195704264422e-06,
+      "loss": 0.4846,
+      "step": 6251
+    },
+    {
+      "epoch": 1.0267485065588242,
+      "grad_norm": 0.3576355296217209,
+      "learning_rate": 8.16168446102842e-06,
+      "loss": 0.5213,
+      "step": 6252
+    },
+    {
+      "epoch": 1.0269127337671669,
+      "grad_norm": 0.5889585564078031,
+      "learning_rate": 8.161411839643121e-06,
+      "loss": 0.5044,
+      "step": 6253
+    },
+    {
+      "epoch": 1.0270769609755097,
+      "grad_norm": 0.2979480569514252,
+      "learning_rate": 8.161139178491291e-06,
+      "loss": 0.4715,
+      "step": 6254
+    },
+    {
+      "epoch": 1.0272411881838523,
+      "grad_norm": 0.2771883322556505,
+      "learning_rate": 8.160866477575885e-06,
+      "loss": 0.4982,
+      "step": 6255
+    },
+    {
+      "epoch": 1.0274054153921952,
+      "grad_norm": 0.27340407742016054,
+      "learning_rate": 8.160593736899869e-06,
+      "loss": 0.494,
+      "step": 6256
+    },
+    {
+      "epoch": 1.0275696426005378,
+      "grad_norm": 0.3568661465565456,
+      "learning_rate": 8.1603209564662e-06,
+      "loss": 0.5152,
+      "step": 6257
+    },
+    {
+      "epoch": 1.0277338698088805,
+      "grad_norm": 0.27665099872106325,
+      "learning_rate": 8.160048136277846e-06,
+      "loss": 0.5109,
+      "step": 6258
+    },
+    {
+      "epoch": 1.0278980970172233,
+      "grad_norm": 0.31119032990878753,
+      "learning_rate": 8.15977527633777e-06,
+      "loss": 0.5023,
+      "step": 6259
+    },
+    {
+      "epoch": 1.028062324225566,
+      "grad_norm": 0.316524222391938,
+      "learning_rate": 8.159502376648932e-06,
+      "loss": 0.4826,
+      "step": 6260
+    },
+    {
+      "epoch": 1.0282265514339088,
+      "grad_norm": 0.3633517365497554,
+      "learning_rate": 8.159229437214298e-06,
+      "loss": 0.4991,
+      "step": 6261
+    },
+    {
+      "epoch": 1.0283907786422515,
+      "grad_norm": 0.2734262319354803,
+      "learning_rate": 8.158956458036833e-06,
+      "loss": 0.4845,
+      "step": 6262
+    },
+    {
+      "epoch": 1.0285550058505943,
+      "grad_norm": 0.3625579511734066,
+      "learning_rate": 8.158683439119499e-06,
+      "loss": 0.4943,
+      "step": 6263
+    },
+    {
+      "epoch": 1.028719233058937,
+      "grad_norm": 0.4547336209639623,
+      "learning_rate": 8.158410380465264e-06,
+      "loss": 0.4927,
+      "step": 6264
+    },
+    {
+      "epoch": 1.0288834602672798,
+      "grad_norm": 0.2978449984905939,
+      "learning_rate": 8.158137282077095e-06,
+      "loss": 0.5134,
+      "step": 6265
+    },
+    {
+      "epoch": 1.0290476874756225,
+      "grad_norm": 0.3792401828910724,
+      "learning_rate": 8.157864143957952e-06,
+      "loss": 0.4817,
+      "step": 6266
+    },
+    {
+      "epoch": 1.0292119146839653,
+      "grad_norm": 0.2943182894634262,
+      "learning_rate": 8.157590966110808e-06,
+      "loss": 0.4708,
+      "step": 6267
+    },
+    {
+      "epoch": 1.029376141892308,
+      "grad_norm": 0.29181766258028285,
+      "learning_rate": 8.157317748538628e-06,
+      "loss": 0.499,
+      "step": 6268
+    },
+    {
+      "epoch": 1.0295403691006508,
+      "grad_norm": 0.35172680386374233,
+      "learning_rate": 8.157044491244378e-06,
+      "loss": 0.488,
+      "step": 6269
+    },
+    {
+      "epoch": 1.0297045963089935,
+      "grad_norm": 0.4369680201670247,
+      "learning_rate": 8.156771194231026e-06,
+      "loss": 0.4881,
+      "step": 6270
+    },
+    {
+      "epoch": 1.0298688235173363,
+      "grad_norm": 0.32279653310986156,
+      "learning_rate": 8.156497857501543e-06,
+      "loss": 0.511,
+      "step": 6271
+    },
+    {
+      "epoch": 1.030033050725679,
+      "grad_norm": 0.29216682293506124,
+      "learning_rate": 8.156224481058893e-06,
+      "loss": 0.5069,
+      "step": 6272
+    },
+    {
+      "epoch": 1.0301972779340218,
+      "grad_norm": 0.3398346029618168,
+      "learning_rate": 8.155951064906052e-06,
+      "loss": 0.5047,
+      "step": 6273
+    },
+    {
+      "epoch": 1.0303615051423645,
+      "grad_norm": 0.2996315228865986,
+      "learning_rate": 8.155677609045982e-06,
+      "loss": 0.4937,
+      "step": 6274
+    },
+    {
+      "epoch": 1.030525732350707,
+      "grad_norm": 0.36267652125254585,
+      "learning_rate": 8.155404113481658e-06,
+      "loss": 0.5147,
+      "step": 6275
+    },
+    {
+      "epoch": 1.03068995955905,
+      "grad_norm": 0.3672582502381371,
+      "learning_rate": 8.155130578216048e-06,
+      "loss": 0.5099,
+      "step": 6276
+    },
+    {
+      "epoch": 1.0308541867673926,
+      "grad_norm": 0.30163867575416514,
+      "learning_rate": 8.154857003252125e-06,
+      "loss": 0.5172,
+      "step": 6277
+    },
+    {
+      "epoch": 1.0310184139757355,
+      "grad_norm": 0.3176725462606026,
+      "learning_rate": 8.154583388592858e-06,
+      "loss": 0.5058,
+      "step": 6278
+    },
+    {
+      "epoch": 1.031182641184078,
+      "grad_norm": 0.28021080159548745,
+      "learning_rate": 8.154309734241219e-06,
+      "loss": 0.4831,
+      "step": 6279
+    },
+    {
+      "epoch": 1.031346868392421,
+      "grad_norm": 0.3046852140004648,
+      "learning_rate": 8.154036040200182e-06,
+      "loss": 0.4731,
+      "step": 6280
+    },
+    {
+      "epoch": 1.0315110956007636,
+      "grad_norm": 0.34338511193132853,
+      "learning_rate": 8.153762306472718e-06,
+      "loss": 0.4932,
+      "step": 6281
+    },
+    {
+      "epoch": 1.0316753228091065,
+      "grad_norm": 0.33288344314950746,
+      "learning_rate": 8.153488533061803e-06,
+      "loss": 0.4729,
+      "step": 6282
+    },
+    {
+      "epoch": 1.031839550017449,
+      "grad_norm": 0.28550101428400565,
+      "learning_rate": 8.153214719970404e-06,
+      "loss": 0.5082,
+      "step": 6283
+    },
+    {
+      "epoch": 1.032003777225792,
+      "grad_norm": 0.3258304073804269,
+      "learning_rate": 8.152940867201502e-06,
+      "loss": 0.4795,
+      "step": 6284
+    },
+    {
+      "epoch": 1.0321680044341346,
+      "grad_norm": 0.2961807049535856,
+      "learning_rate": 8.152666974758068e-06,
+      "loss": 0.4791,
+      "step": 6285
+    },
+    {
+      "epoch": 1.0323322316424774,
+      "grad_norm": 0.3435258210556305,
+      "learning_rate": 8.152393042643075e-06,
+      "loss": 0.5016,
+      "step": 6286
+    },
+    {
+      "epoch": 1.03249645885082,
+      "grad_norm": 0.3171183540701704,
+      "learning_rate": 8.1521190708595e-06,
+      "loss": 0.5012,
+      "step": 6287
+    },
+    {
+      "epoch": 1.032660686059163,
+      "grad_norm": 0.3163255037337862,
+      "learning_rate": 8.15184505941032e-06,
+      "loss": 0.4734,
+      "step": 6288
+    },
+    {
+      "epoch": 1.0328249132675056,
+      "grad_norm": 0.28360770056732126,
+      "learning_rate": 8.15157100829851e-06,
+      "loss": 0.5327,
+      "step": 6289
+    },
+    {
+      "epoch": 1.0329891404758484,
+      "grad_norm": 0.3295009314736953,
+      "learning_rate": 8.151296917527048e-06,
+      "loss": 0.4601,
+      "step": 6290
+    },
+    {
+      "epoch": 1.033153367684191,
+      "grad_norm": 0.3109452542130047,
+      "learning_rate": 8.151022787098904e-06,
+      "loss": 0.4886,
+      "step": 6291
+    },
+    {
+      "epoch": 1.0333175948925337,
+      "grad_norm": 0.3422775777883866,
+      "learning_rate": 8.150748617017064e-06,
+      "loss": 0.4804,
+      "step": 6292
+    },
+    {
+      "epoch": 1.0334818221008766,
+      "grad_norm": 0.32723287713230975,
+      "learning_rate": 8.150474407284502e-06,
+      "loss": 0.5031,
+      "step": 6293
+    },
+    {
+      "epoch": 1.0336460493092192,
+      "grad_norm": 0.3649000496586239,
+      "learning_rate": 8.150200157904194e-06,
+      "loss": 0.4876,
+      "step": 6294
+    },
+    {
+      "epoch": 1.033810276517562,
+      "grad_norm": 0.2602417938122742,
+      "learning_rate": 8.149925868879123e-06,
+      "loss": 0.4804,
+      "step": 6295
+    },
+    {
+      "epoch": 1.0339745037259047,
+      "grad_norm": 0.2759579393389584,
+      "learning_rate": 8.149651540212267e-06,
+      "loss": 0.4981,
+      "step": 6296
+    },
+    {
+      "epoch": 1.0341387309342476,
+      "grad_norm": 0.2666833973790393,
+      "learning_rate": 8.149377171906601e-06,
+      "loss": 0.4779,
+      "step": 6297
+    },
+    {
+      "epoch": 1.0343029581425902,
+      "grad_norm": 0.42156453734766997,
+      "learning_rate": 8.149102763965112e-06,
+      "loss": 0.5211,
+      "step": 6298
+    },
+    {
+      "epoch": 1.034467185350933,
+      "grad_norm": 0.2823997950925321,
+      "learning_rate": 8.148828316390776e-06,
+      "loss": 0.496,
+      "step": 6299
+    },
+    {
+      "epoch": 1.0346314125592757,
+      "grad_norm": 0.34982207290752043,
+      "learning_rate": 8.148553829186573e-06,
+      "loss": 0.4689,
+      "step": 6300
+    },
+    {
+      "epoch": 1.0347956397676186,
+      "grad_norm": 0.391208131768673,
+      "learning_rate": 8.148279302355487e-06,
+      "loss": 0.4973,
+      "step": 6301
+    },
+    {
+      "epoch": 1.0349598669759612,
+      "grad_norm": 0.28742279855847286,
+      "learning_rate": 8.148004735900498e-06,
+      "loss": 0.511,
+      "step": 6302
+    },
+    {
+      "epoch": 1.035124094184304,
+      "grad_norm": 0.30432930318759355,
+      "learning_rate": 8.147730129824588e-06,
+      "loss": 0.4845,
+      "step": 6303
+    },
+    {
+      "epoch": 1.0352883213926467,
+      "grad_norm": 0.2696241666956859,
+      "learning_rate": 8.14745548413074e-06,
+      "loss": 0.4746,
+      "step": 6304
+    },
+    {
+      "epoch": 1.0354525486009896,
+      "grad_norm": 0.3517664508179872,
+      "learning_rate": 8.147180798821937e-06,
+      "loss": 0.5045,
+      "step": 6305
+    },
+    {
+      "epoch": 1.0356167758093322,
+      "grad_norm": 0.4605530445297964,
+      "learning_rate": 8.146906073901163e-06,
+      "loss": 0.4981,
+      "step": 6306
+    },
+    {
+      "epoch": 1.035781003017675,
+      "grad_norm": 0.6338064223676042,
+      "learning_rate": 8.1466313093714e-06,
+      "loss": 0.4754,
+      "step": 6307
+    },
+    {
+      "epoch": 1.0359452302260177,
+      "grad_norm": 0.30453246058597266,
+      "learning_rate": 8.146356505235634e-06,
+      "loss": 0.484,
+      "step": 6308
+    },
+    {
+      "epoch": 1.0361094574343603,
+      "grad_norm": 0.3257172020653211,
+      "learning_rate": 8.146081661496848e-06,
+      "loss": 0.4794,
+      "step": 6309
+    },
+    {
+      "epoch": 1.0362736846427032,
+      "grad_norm": 0.34488299780916387,
+      "learning_rate": 8.145806778158027e-06,
+      "loss": 0.4687,
+      "step": 6310
+    },
+    {
+      "epoch": 1.0364379118510458,
+      "grad_norm": 0.3213369359566095,
+      "learning_rate": 8.14553185522216e-06,
+      "loss": 0.4761,
+      "step": 6311
+    },
+    {
+      "epoch": 1.0366021390593887,
+      "grad_norm": 0.32874328293494665,
+      "learning_rate": 8.145256892692229e-06,
+      "loss": 0.5035,
+      "step": 6312
+    },
+    {
+      "epoch": 1.0367663662677313,
+      "grad_norm": 0.2970117333301102,
+      "learning_rate": 8.14498189057122e-06,
+      "loss": 0.4938,
+      "step": 6313
+    },
+    {
+      "epoch": 1.0369305934760742,
+      "grad_norm": 0.38096806888632473,
+      "learning_rate": 8.144706848862123e-06,
+      "loss": 0.4961,
+      "step": 6314
+    },
+    {
+      "epoch": 1.0370948206844168,
+      "grad_norm": 0.28450513054266885,
+      "learning_rate": 8.144431767567925e-06,
+      "loss": 0.4956,
+      "step": 6315
+    },
+    {
+      "epoch": 1.0372590478927597,
+      "grad_norm": 0.33008924689826424,
+      "learning_rate": 8.14415664669161e-06,
+      "loss": 0.4889,
+      "step": 6316
+    },
+    {
+      "epoch": 1.0374232751011023,
+      "grad_norm": 0.31032871022062886,
+      "learning_rate": 8.14388148623617e-06,
+      "loss": 0.4774,
+      "step": 6317
+    },
+    {
+      "epoch": 1.0375875023094452,
+      "grad_norm": 0.3226996695818067,
+      "learning_rate": 8.143606286204592e-06,
+      "loss": 0.4722,
+      "step": 6318
+    },
+    {
+      "epoch": 1.0377517295177878,
+      "grad_norm": 0.28015504286106274,
+      "learning_rate": 8.143331046599863e-06,
+      "loss": 0.4974,
+      "step": 6319
+    },
+    {
+      "epoch": 1.0379159567261307,
+      "grad_norm": 0.3216228193671261,
+      "learning_rate": 8.143055767424978e-06,
+      "loss": 0.5035,
+      "step": 6320
+    },
+    {
+      "epoch": 1.0380801839344733,
+      "grad_norm": 0.32400022353714,
+      "learning_rate": 8.14278044868292e-06,
+      "loss": 0.4915,
+      "step": 6321
+    },
+    {
+      "epoch": 1.0382444111428162,
+      "grad_norm": 0.3889307559896388,
+      "learning_rate": 8.142505090376683e-06,
+      "loss": 0.5154,
+      "step": 6322
+    },
+    {
+      "epoch": 1.0384086383511588,
+      "grad_norm": 0.31872095687613694,
+      "learning_rate": 8.142229692509258e-06,
+      "loss": 0.4746,
+      "step": 6323
+    },
+    {
+      "epoch": 1.0385728655595017,
+      "grad_norm": 0.3825206261235205,
+      "learning_rate": 8.141954255083633e-06,
+      "loss": 0.5269,
+      "step": 6324
+    },
+    {
+      "epoch": 1.0387370927678443,
+      "grad_norm": 0.3205615672991652,
+      "learning_rate": 8.141678778102804e-06,
+      "loss": 0.4983,
+      "step": 6325
+    },
+    {
+      "epoch": 1.038901319976187,
+      "grad_norm": 0.4040202032633061,
+      "learning_rate": 8.141403261569759e-06,
+      "loss": 0.4822,
+      "step": 6326
+    },
+    {
+      "epoch": 1.0390655471845298,
+      "grad_norm": 0.37001803297905866,
+      "learning_rate": 8.141127705487492e-06,
+      "loss": 0.4972,
+      "step": 6327
+    },
+    {
+      "epoch": 1.0392297743928725,
+      "grad_norm": 0.33923188659094977,
+      "learning_rate": 8.140852109858997e-06,
+      "loss": 0.511,
+      "step": 6328
+    },
+    {
+      "epoch": 1.0393940016012153,
+      "grad_norm": 0.29837615842553383,
+      "learning_rate": 8.140576474687264e-06,
+      "loss": 0.4919,
+      "step": 6329
+    },
+    {
+      "epoch": 1.039558228809558,
+      "grad_norm": 0.2711498146698936,
+      "learning_rate": 8.140300799975289e-06,
+      "loss": 0.4858,
+      "step": 6330
+    },
+    {
+      "epoch": 1.0397224560179008,
+      "grad_norm": 0.323007121971415,
+      "learning_rate": 8.140025085726067e-06,
+      "loss": 0.5001,
+      "step": 6331
+    },
+    {
+      "epoch": 1.0398866832262434,
+      "grad_norm": 0.4009112749843193,
+      "learning_rate": 8.139749331942591e-06,
+      "loss": 0.4872,
+      "step": 6332
+    },
+    {
+      "epoch": 1.0400509104345863,
+      "grad_norm": 0.26720307728953957,
+      "learning_rate": 8.139473538627855e-06,
+      "loss": 0.4931,
+      "step": 6333
+    },
+    {
+      "epoch": 1.040215137642929,
+      "grad_norm": 0.29601146133600026,
+      "learning_rate": 8.139197705784857e-06,
+      "loss": 0.4857,
+      "step": 6334
+    },
+    {
+      "epoch": 1.0403793648512718,
+      "grad_norm": 0.3176755739657561,
+      "learning_rate": 8.13892183341659e-06,
+      "loss": 0.4981,
+      "step": 6335
+    },
+    {
+      "epoch": 1.0405435920596144,
+      "grad_norm": 0.31923052268403973,
+      "learning_rate": 8.138645921526053e-06,
+      "loss": 0.4985,
+      "step": 6336
+    },
+    {
+      "epoch": 1.0407078192679573,
+      "grad_norm": 0.27262402169127714,
+      "learning_rate": 8.138369970116242e-06,
+      "loss": 0.4889,
+      "step": 6337
+    },
+    {
+      "epoch": 1.0408720464763,
+      "grad_norm": 0.2981051983207571,
+      "learning_rate": 8.13809397919015e-06,
+      "loss": 0.4818,
+      "step": 6338
+    },
+    {
+      "epoch": 1.0410362736846428,
+      "grad_norm": 0.26663296072728127,
+      "learning_rate": 8.137817948750781e-06,
+      "loss": 0.4983,
+      "step": 6339
+    },
+    {
+      "epoch": 1.0412005008929854,
+      "grad_norm": 0.3322672618709718,
+      "learning_rate": 8.13754187880113e-06,
+      "loss": 0.4972,
+      "step": 6340
+    },
+    {
+      "epoch": 1.0413647281013283,
+      "grad_norm": 0.3164200711690184,
+      "learning_rate": 8.137265769344193e-06,
+      "loss": 0.4834,
+      "step": 6341
+    },
+    {
+      "epoch": 1.041528955309671,
+      "grad_norm": 0.29570874323950064,
+      "learning_rate": 8.136989620382973e-06,
+      "loss": 0.4938,
+      "step": 6342
+    },
+    {
+      "epoch": 1.0416931825180136,
+      "grad_norm": 0.3395825507914,
+      "learning_rate": 8.136713431920469e-06,
+      "loss": 0.4713,
+      "step": 6343
+    },
+    {
+      "epoch": 1.0418574097263564,
+      "grad_norm": 0.3596792089972651,
+      "learning_rate": 8.136437203959677e-06,
+      "loss": 0.4718,
+      "step": 6344
+    },
+    {
+      "epoch": 1.042021636934699,
+      "grad_norm": 0.2913957363979316,
+      "learning_rate": 8.1361609365036e-06,
+      "loss": 0.49,
+      "step": 6345
+    },
+    {
+      "epoch": 1.042185864143042,
+      "grad_norm": 0.30153632831779337,
+      "learning_rate": 8.135884629555236e-06,
+      "loss": 0.4872,
+      "step": 6346
+    },
+    {
+      "epoch": 1.0423500913513846,
+      "grad_norm": 0.3027949712850462,
+      "learning_rate": 8.135608283117589e-06,
+      "loss": 0.5091,
+      "step": 6347
+    },
+    {
+      "epoch": 1.0425143185597274,
+      "grad_norm": 0.3667060182197534,
+      "learning_rate": 8.135331897193659e-06,
+      "loss": 0.5129,
+      "step": 6348
+    },
+    {
+      "epoch": 1.04267854576807,
+      "grad_norm": 0.34954670908517377,
+      "learning_rate": 8.135055471786448e-06,
+      "loss": 0.4907,
+      "step": 6349
+    },
+    {
+      "epoch": 1.042842772976413,
+      "grad_norm": 0.5746944717000683,
+      "learning_rate": 8.134779006898958e-06,
+      "loss": 0.5251,
+      "step": 6350
+    },
+    {
+      "epoch": 1.0430070001847556,
+      "grad_norm": 0.43380942060730043,
+      "learning_rate": 8.134502502534192e-06,
+      "loss": 0.4859,
+      "step": 6351
+    },
+    {
+      "epoch": 1.0431712273930984,
+      "grad_norm": 0.3356139791441698,
+      "learning_rate": 8.134225958695153e-06,
+      "loss": 0.4771,
+      "step": 6352
+    },
+    {
+      "epoch": 1.043335454601441,
+      "grad_norm": 0.3385240491576104,
+      "learning_rate": 8.133949375384844e-06,
+      "loss": 0.4863,
+      "step": 6353
+    },
+    {
+      "epoch": 1.043499681809784,
+      "grad_norm": 0.3291980594155157,
+      "learning_rate": 8.13367275260627e-06,
+      "loss": 0.5003,
+      "step": 6354
+    },
+    {
+      "epoch": 1.0436639090181266,
+      "grad_norm": 0.3383124118965275,
+      "learning_rate": 8.133396090362435e-06,
+      "loss": 0.4945,
+      "step": 6355
+    },
+    {
+      "epoch": 1.0438281362264694,
+      "grad_norm": 0.5204865829726072,
+      "learning_rate": 8.133119388656344e-06,
+      "loss": 0.4808,
+      "step": 6356
+    },
+    {
+      "epoch": 1.043992363434812,
+      "grad_norm": 0.2809584481919922,
+      "learning_rate": 8.132842647491002e-06,
+      "loss": 0.4964,
+      "step": 6357
+    },
+    {
+      "epoch": 1.044156590643155,
+      "grad_norm": 0.3785537153231929,
+      "learning_rate": 8.132565866869414e-06,
+      "loss": 0.4898,
+      "step": 6358
+    },
+    {
+      "epoch": 1.0443208178514976,
+      "grad_norm": 0.2990071029129028,
+      "learning_rate": 8.132289046794584e-06,
+      "loss": 0.4954,
+      "step": 6359
+    },
+    {
+      "epoch": 1.0444850450598402,
+      "grad_norm": 0.38818656992221234,
+      "learning_rate": 8.132012187269526e-06,
+      "loss": 0.5099,
+      "step": 6360
+    },
+    {
+      "epoch": 1.044649272268183,
+      "grad_norm": 0.3580751562409733,
+      "learning_rate": 8.13173528829724e-06,
+      "loss": 0.4925,
+      "step": 6361
+    },
+    {
+      "epoch": 1.0448134994765257,
+      "grad_norm": 0.4102291661116885,
+      "learning_rate": 8.131458349880735e-06,
+      "loss": 0.4965,
+      "step": 6362
+    },
+    {
+      "epoch": 1.0449777266848685,
+      "grad_norm": 0.3294150987730053,
+      "learning_rate": 8.13118137202302e-06,
+      "loss": 0.5011,
+      "step": 6363
+    },
+    {
+      "epoch": 1.0451419538932112,
+      "grad_norm": 0.31692405906232507,
+      "learning_rate": 8.130904354727103e-06,
+      "loss": 0.4875,
+      "step": 6364
+    },
+    {
+      "epoch": 1.045306181101554,
+      "grad_norm": 0.2950992571838892,
+      "learning_rate": 8.130627297995991e-06,
+      "loss": 0.5003,
+      "step": 6365
+    },
+    {
+      "epoch": 1.0454704083098967,
+      "grad_norm": 0.45984806915384435,
+      "learning_rate": 8.130350201832697e-06,
+      "loss": 0.4856,
+      "step": 6366
+    },
+    {
+      "epoch": 1.0456346355182395,
+      "grad_norm": 0.4105295951420729,
+      "learning_rate": 8.130073066240226e-06,
+      "loss": 0.5028,
+      "step": 6367
+    },
+    {
+      "epoch": 1.0457988627265822,
+      "grad_norm": 0.36942653174057527,
+      "learning_rate": 8.129795891221592e-06,
+      "loss": 0.4869,
+      "step": 6368
+    },
+    {
+      "epoch": 1.045963089934925,
+      "grad_norm": 0.34164681035447897,
+      "learning_rate": 8.1295186767798e-06,
+      "loss": 0.501,
+      "step": 6369
+    },
+    {
+      "epoch": 1.0461273171432677,
+      "grad_norm": 0.6678599138574988,
+      "learning_rate": 8.12924142291787e-06,
+      "loss": 0.4662,
+      "step": 6370
+    },
+    {
+      "epoch": 1.0462915443516105,
+      "grad_norm": 0.5031081494107003,
+      "learning_rate": 8.128964129638802e-06,
+      "loss": 0.4945,
+      "step": 6371
+    },
+    {
+      "epoch": 1.0464557715599532,
+      "grad_norm": 0.2899303698213951,
+      "learning_rate": 8.128686796945615e-06,
+      "loss": 0.4913,
+      "step": 6372
+    },
+    {
+      "epoch": 1.046619998768296,
+      "grad_norm": 0.3061299471747233,
+      "learning_rate": 8.128409424841319e-06,
+      "loss": 0.5124,
+      "step": 6373
+    },
+    {
+      "epoch": 1.0467842259766387,
+      "grad_norm": 0.3186445589777642,
+      "learning_rate": 8.128132013328928e-06,
+      "loss": 0.5102,
+      "step": 6374
+    },
+    {
+      "epoch": 1.0469484531849815,
+      "grad_norm": 0.30979226756641537,
+      "learning_rate": 8.127854562411452e-06,
+      "loss": 0.4907,
+      "step": 6375
+    },
+    {
+      "epoch": 1.0471126803933242,
+      "grad_norm": 0.35666387304376157,
+      "learning_rate": 8.127577072091906e-06,
+      "loss": 0.4862,
+      "step": 6376
+    },
+    {
+      "epoch": 1.0472769076016668,
+      "grad_norm": 0.33999127955278113,
+      "learning_rate": 8.127299542373306e-06,
+      "loss": 0.5005,
+      "step": 6377
+    },
+    {
+      "epoch": 1.0474411348100097,
+      "grad_norm": 0.32517006313544056,
+      "learning_rate": 8.127021973258664e-06,
+      "loss": 0.5225,
+      "step": 6378
+    },
+    {
+      "epoch": 1.0476053620183523,
+      "grad_norm": 0.30386677497274767,
+      "learning_rate": 8.126744364750991e-06,
+      "loss": 0.5187,
+      "step": 6379
+    },
+    {
+      "epoch": 1.0477695892266952,
+      "grad_norm": 0.3028445243411251,
+      "learning_rate": 8.12646671685331e-06,
+      "loss": 0.4911,
+      "step": 6380
+    },
+    {
+      "epoch": 1.0479338164350378,
+      "grad_norm": 0.3348978408227218,
+      "learning_rate": 8.126189029568631e-06,
+      "loss": 0.5013,
+      "step": 6381
+    },
+    {
+      "epoch": 1.0480980436433807,
+      "grad_norm": 0.31603848477950536,
+      "learning_rate": 8.125911302899973e-06,
+      "loss": 0.4846,
+      "step": 6382
+    },
+    {
+      "epoch": 1.0482622708517233,
+      "grad_norm": 0.2986142070302935,
+      "learning_rate": 8.125633536850349e-06,
+      "loss": 0.4859,
+      "step": 6383
+    },
+    {
+      "epoch": 1.0484264980600662,
+      "grad_norm": 0.267336828835043,
+      "learning_rate": 8.125355731422778e-06,
+      "loss": 0.5089,
+      "step": 6384
+    },
+    {
+      "epoch": 1.0485907252684088,
+      "grad_norm": 0.26584039989022573,
+      "learning_rate": 8.125077886620277e-06,
+      "loss": 0.4855,
+      "step": 6385
+    },
+    {
+      "epoch": 1.0487549524767517,
+      "grad_norm": 0.2942213330070643,
+      "learning_rate": 8.124800002445864e-06,
+      "loss": 0.4914,
+      "step": 6386
+    },
+    {
+      "epoch": 1.0489191796850943,
+      "grad_norm": 0.34842284141271185,
+      "learning_rate": 8.124522078902556e-06,
+      "loss": 0.4798,
+      "step": 6387
+    },
+    {
+      "epoch": 1.0490834068934372,
+      "grad_norm": 0.28931993160219377,
+      "learning_rate": 8.124244115993372e-06,
+      "loss": 0.4917,
+      "step": 6388
+    },
+    {
+      "epoch": 1.0492476341017798,
+      "grad_norm": 0.2914300210726281,
+      "learning_rate": 8.123966113721331e-06,
+      "loss": 0.4918,
+      "step": 6389
+    },
+    {
+      "epoch": 1.0494118613101227,
+      "grad_norm": 0.3163077994522434,
+      "learning_rate": 8.123688072089455e-06,
+      "loss": 0.5009,
+      "step": 6390
+    },
+    {
+      "epoch": 1.0495760885184653,
+      "grad_norm": 0.2867811327270028,
+      "learning_rate": 8.123409991100758e-06,
+      "loss": 0.4716,
+      "step": 6391
+    },
+    {
+      "epoch": 1.0497403157268081,
+      "grad_norm": 0.30351396153490035,
+      "learning_rate": 8.123131870758266e-06,
+      "loss": 0.5001,
+      "step": 6392
+    },
+    {
+      "epoch": 1.0499045429351508,
+      "grad_norm": 0.26456114320785307,
+      "learning_rate": 8.122853711064997e-06,
+      "loss": 0.5115,
+      "step": 6393
+    },
+    {
+      "epoch": 1.0500687701434934,
+      "grad_norm": 0.3490191334110551,
+      "learning_rate": 8.12257551202397e-06,
+      "loss": 0.5033,
+      "step": 6394
+    },
+    {
+      "epoch": 1.0502329973518363,
+      "grad_norm": 0.32342546862186944,
+      "learning_rate": 8.122297273638212e-06,
+      "loss": 0.4858,
+      "step": 6395
+    },
+    {
+      "epoch": 1.050397224560179,
+      "grad_norm": 0.32261052856924055,
+      "learning_rate": 8.122018995910738e-06,
+      "loss": 0.4902,
+      "step": 6396
+    },
+    {
+      "epoch": 1.0505614517685218,
+      "grad_norm": 0.32048637806924013,
+      "learning_rate": 8.121740678844576e-06,
+      "loss": 0.4695,
+      "step": 6397
+    },
+    {
+      "epoch": 1.0507256789768644,
+      "grad_norm": 0.3235467650978821,
+      "learning_rate": 8.121462322442749e-06,
+      "loss": 0.5074,
+      "step": 6398
+    },
+    {
+      "epoch": 1.0508899061852073,
+      "grad_norm": 0.3222995364524542,
+      "learning_rate": 8.121183926708274e-06,
+      "loss": 0.5017,
+      "step": 6399
+    },
+    {
+      "epoch": 1.05105413339355,
+      "grad_norm": 0.3099918502023668,
+      "learning_rate": 8.120905491644181e-06,
+      "loss": 0.4903,
+      "step": 6400
+    },
+    {
+      "epoch": 1.0512183606018928,
+      "grad_norm": 0.2991011827463223,
+      "learning_rate": 8.120627017253492e-06,
+      "loss": 0.4863,
+      "step": 6401
+    },
+    {
+      "epoch": 1.0513825878102354,
+      "grad_norm": 0.3032089180759681,
+      "learning_rate": 8.12034850353923e-06,
+      "loss": 0.5018,
+      "step": 6402
+    },
+    {
+      "epoch": 1.0515468150185783,
+      "grad_norm": 0.30778202378615055,
+      "learning_rate": 8.120069950504426e-06,
+      "loss": 0.4971,
+      "step": 6403
+    },
+    {
+      "epoch": 1.051711042226921,
+      "grad_norm": 0.33858123870495077,
+      "learning_rate": 8.119791358152097e-06,
+      "loss": 0.5004,
+      "step": 6404
+    },
+    {
+      "epoch": 1.0518752694352638,
+      "grad_norm": 0.31106722488997146,
+      "learning_rate": 8.119512726485272e-06,
+      "loss": 0.4889,
+      "step": 6405
+    },
+    {
+      "epoch": 1.0520394966436064,
+      "grad_norm": 0.28103243070829426,
+      "learning_rate": 8.119234055506979e-06,
+      "loss": 0.4964,
+      "step": 6406
+    },
+    {
+      "epoch": 1.0522037238519493,
+      "grad_norm": 0.30513334246236656,
+      "learning_rate": 8.118955345220243e-06,
+      "loss": 0.5038,
+      "step": 6407
+    },
+    {
+      "epoch": 1.052367951060292,
+      "grad_norm": 0.2703130936598874,
+      "learning_rate": 8.11867659562809e-06,
+      "loss": 0.4753,
+      "step": 6408
+    },
+    {
+      "epoch": 1.0525321782686348,
+      "grad_norm": 0.34695632434730056,
+      "learning_rate": 8.118397806733549e-06,
+      "loss": 0.4753,
+      "step": 6409
+    },
+    {
+      "epoch": 1.0526964054769774,
+      "grad_norm": 0.35036872667227215,
+      "learning_rate": 8.11811897853965e-06,
+      "loss": 0.4835,
+      "step": 6410
+    },
+    {
+      "epoch": 1.05286063268532,
+      "grad_norm": 0.2752474487062583,
+      "learning_rate": 8.117840111049418e-06,
+      "loss": 0.4795,
+      "step": 6411
+    },
+    {
+      "epoch": 1.053024859893663,
+      "grad_norm": 0.3954558426656484,
+      "learning_rate": 8.117561204265881e-06,
+      "loss": 0.4736,
+      "step": 6412
+    },
+    {
+      "epoch": 1.0531890871020055,
+      "grad_norm": 0.3024301299880532,
+      "learning_rate": 8.117282258192073e-06,
+      "loss": 0.4901,
+      "step": 6413
+    },
+    {
+      "epoch": 1.0533533143103484,
+      "grad_norm": 0.36585136371914173,
+      "learning_rate": 8.117003272831018e-06,
+      "loss": 0.4861,
+      "step": 6414
+    },
+    {
+      "epoch": 1.053517541518691,
+      "grad_norm": 0.30042123148773786,
+      "learning_rate": 8.116724248185751e-06,
+      "loss": 0.5021,
+      "step": 6415
+    },
+    {
+      "epoch": 1.053681768727034,
+      "grad_norm": 0.3273045503653264,
+      "learning_rate": 8.1164451842593e-06,
+      "loss": 0.5005,
+      "step": 6416
+    },
+    {
+      "epoch": 1.0538459959353765,
+      "grad_norm": 0.3148844502941428,
+      "learning_rate": 8.116166081054698e-06,
+      "loss": 0.4799,
+      "step": 6417
+    },
+    {
+      "epoch": 1.0540102231437194,
+      "grad_norm": 0.30624673877227687,
+      "learning_rate": 8.11588693857497e-06,
+      "loss": 0.5072,
+      "step": 6418
+    },
+    {
+      "epoch": 1.054174450352062,
+      "grad_norm": 0.3166765249654235,
+      "learning_rate": 8.115607756823156e-06,
+      "loss": 0.4966,
+      "step": 6419
+    },
+    {
+      "epoch": 1.054338677560405,
+      "grad_norm": 0.2659242128969254,
+      "learning_rate": 8.115328535802283e-06,
+      "loss": 0.488,
+      "step": 6420
+    },
+    {
+      "epoch": 1.0545029047687475,
+      "grad_norm": 0.25926394594099467,
+      "learning_rate": 8.115049275515386e-06,
+      "loss": 0.4734,
+      "step": 6421
+    },
+    {
+      "epoch": 1.0546671319770904,
+      "grad_norm": 0.36846139712139503,
+      "learning_rate": 8.114769975965496e-06,
+      "loss": 0.52,
+      "step": 6422
+    },
+    {
+      "epoch": 1.054831359185433,
+      "grad_norm": 0.28318609317997545,
+      "learning_rate": 8.114490637155648e-06,
+      "loss": 0.4973,
+      "step": 6423
+    },
+    {
+      "epoch": 1.0549955863937759,
+      "grad_norm": 0.3365623874313242,
+      "learning_rate": 8.114211259088875e-06,
+      "loss": 0.4806,
+      "step": 6424
+    },
+    {
+      "epoch": 1.0551598136021185,
+      "grad_norm": 0.37523450137919945,
+      "learning_rate": 8.113931841768212e-06,
+      "loss": 0.4829,
+      "step": 6425
+    },
+    {
+      "epoch": 1.0553240408104614,
+      "grad_norm": 0.29403851701146794,
+      "learning_rate": 8.113652385196695e-06,
+      "loss": 0.4704,
+      "step": 6426
+    },
+    {
+      "epoch": 1.055488268018804,
+      "grad_norm": 0.2945693218797296,
+      "learning_rate": 8.113372889377357e-06,
+      "loss": 0.4648,
+      "step": 6427
+    },
+    {
+      "epoch": 1.0556524952271467,
+      "grad_norm": 0.3418559064297287,
+      "learning_rate": 8.113093354313234e-06,
+      "loss": 0.498,
+      "step": 6428
+    },
+    {
+      "epoch": 1.0558167224354895,
+      "grad_norm": 0.3366328880212634,
+      "learning_rate": 8.112813780007362e-06,
+      "loss": 0.4891,
+      "step": 6429
+    },
+    {
+      "epoch": 1.0559809496438322,
+      "grad_norm": 0.3358091705998052,
+      "learning_rate": 8.112534166462778e-06,
+      "loss": 0.4675,
+      "step": 6430
+    },
+    {
+      "epoch": 1.056145176852175,
+      "grad_norm": 0.31678677491645746,
+      "learning_rate": 8.112254513682519e-06,
+      "loss": 0.4944,
+      "step": 6431
+    },
+    {
+      "epoch": 1.0563094040605177,
+      "grad_norm": 0.2648603428506825,
+      "learning_rate": 8.111974821669623e-06,
+      "loss": 0.4974,
+      "step": 6432
+    },
+    {
+      "epoch": 1.0564736312688605,
+      "grad_norm": 0.2982763427295238,
+      "learning_rate": 8.111695090427125e-06,
+      "loss": 0.4994,
+      "step": 6433
+    },
+    {
+      "epoch": 1.0566378584772032,
+      "grad_norm": 0.3244793940093539,
+      "learning_rate": 8.111415319958066e-06,
+      "loss": 0.505,
+      "step": 6434
+    },
+    {
+      "epoch": 1.056802085685546,
+      "grad_norm": 0.33260643099417697,
+      "learning_rate": 8.111135510265483e-06,
+      "loss": 0.5006,
+      "step": 6435
+    },
+    {
+      "epoch": 1.0569663128938886,
+      "grad_norm": 0.2925676158677974,
+      "learning_rate": 8.110855661352416e-06,
+      "loss": 0.4819,
+      "step": 6436
+    },
+    {
+      "epoch": 1.0571305401022315,
+      "grad_norm": 0.3493362290759631,
+      "learning_rate": 8.110575773221903e-06,
+      "loss": 0.4902,
+      "step": 6437
+    },
+    {
+      "epoch": 1.0572947673105741,
+      "grad_norm": 0.3925359125596493,
+      "learning_rate": 8.110295845876985e-06,
+      "loss": 0.4894,
+      "step": 6438
+    },
+    {
+      "epoch": 1.057458994518917,
+      "grad_norm": 0.2901191149400785,
+      "learning_rate": 8.110015879320703e-06,
+      "loss": 0.4952,
+      "step": 6439
+    },
+    {
+      "epoch": 1.0576232217272596,
+      "grad_norm": 0.31226298482160286,
+      "learning_rate": 8.109735873556097e-06,
+      "loss": 0.4765,
+      "step": 6440
+    },
+    {
+      "epoch": 1.0577874489356025,
+      "grad_norm": 0.2608839794267408,
+      "learning_rate": 8.109455828586206e-06,
+      "loss": 0.4841,
+      "step": 6441
+    },
+    {
+      "epoch": 1.0579516761439451,
+      "grad_norm": 0.41179817747126496,
+      "learning_rate": 8.109175744414074e-06,
+      "loss": 0.4995,
+      "step": 6442
+    },
+    {
+      "epoch": 1.058115903352288,
+      "grad_norm": 0.27599730187609833,
+      "learning_rate": 8.108895621042743e-06,
+      "loss": 0.4825,
+      "step": 6443
+    },
+    {
+      "epoch": 1.0582801305606306,
+      "grad_norm": 0.38717551812691675,
+      "learning_rate": 8.108615458475256e-06,
+      "loss": 0.4693,
+      "step": 6444
+    },
+    {
+      "epoch": 1.0584443577689733,
+      "grad_norm": 0.309503198292017,
+      "learning_rate": 8.108335256714653e-06,
+      "loss": 0.5025,
+      "step": 6445
+    },
+    {
+      "epoch": 1.0586085849773161,
+      "grad_norm": 0.3527154897099522,
+      "learning_rate": 8.108055015763979e-06,
+      "loss": 0.4978,
+      "step": 6446
+    },
+    {
+      "epoch": 1.0587728121856588,
+      "grad_norm": 0.34243211145802427,
+      "learning_rate": 8.10777473562628e-06,
+      "loss": 0.4918,
+      "step": 6447
+    },
+    {
+      "epoch": 1.0589370393940016,
+      "grad_norm": 0.5315935088433437,
+      "learning_rate": 8.107494416304595e-06,
+      "loss": 0.5002,
+      "step": 6448
+    },
+    {
+      "epoch": 1.0591012666023443,
+      "grad_norm": 0.34448608785796825,
+      "learning_rate": 8.107214057801971e-06,
+      "loss": 0.4802,
+      "step": 6449
+    },
+    {
+      "epoch": 1.0592654938106871,
+      "grad_norm": 0.2679657284532223,
+      "learning_rate": 8.106933660121455e-06,
+      "loss": 0.4862,
+      "step": 6450
+    },
+    {
+      "epoch": 1.0594297210190298,
+      "grad_norm": 0.3044665667323601,
+      "learning_rate": 8.10665322326609e-06,
+      "loss": 0.4964,
+      "step": 6451
+    },
+    {
+      "epoch": 1.0595939482273726,
+      "grad_norm": 0.44302506304396316,
+      "learning_rate": 8.106372747238923e-06,
+      "loss": 0.5033,
+      "step": 6452
+    },
+    {
+      "epoch": 1.0597581754357153,
+      "grad_norm": 0.3007104836080406,
+      "learning_rate": 8.106092232043002e-06,
+      "loss": 0.5041,
+      "step": 6453
+    },
+    {
+      "epoch": 1.0599224026440581,
+      "grad_norm": 0.30669849052499437,
+      "learning_rate": 8.105811677681367e-06,
+      "loss": 0.4813,
+      "step": 6454
+    },
+    {
+      "epoch": 1.0600866298524008,
+      "grad_norm": 0.33585983656673835,
+      "learning_rate": 8.105531084157072e-06,
+      "loss": 0.5003,
+      "step": 6455
+    },
+    {
+      "epoch": 1.0602508570607436,
+      "grad_norm": 0.35016795972792997,
+      "learning_rate": 8.105250451473162e-06,
+      "loss": 0.4867,
+      "step": 6456
+    },
+    {
+      "epoch": 1.0604150842690863,
+      "grad_norm": 0.3019447943301932,
+      "learning_rate": 8.104969779632685e-06,
+      "loss": 0.4829,
+      "step": 6457
+    },
+    {
+      "epoch": 1.0605793114774291,
+      "grad_norm": 0.44307691839299485,
+      "learning_rate": 8.10468906863869e-06,
+      "loss": 0.4817,
+      "step": 6458
+    },
+    {
+      "epoch": 1.0607435386857718,
+      "grad_norm": 0.7749990416144014,
+      "learning_rate": 8.104408318494224e-06,
+      "loss": 0.4748,
+      "step": 6459
+    },
+    {
+      "epoch": 1.0609077658941146,
+      "grad_norm": 0.3369575113116707,
+      "learning_rate": 8.104127529202338e-06,
+      "loss": 0.4745,
+      "step": 6460
+    },
+    {
+      "epoch": 1.0610719931024573,
+      "grad_norm": 0.30228965862062723,
+      "learning_rate": 8.103846700766081e-06,
+      "loss": 0.5156,
+      "step": 6461
+    },
+    {
+      "epoch": 1.0612362203108,
+      "grad_norm": 0.28323225167550203,
+      "learning_rate": 8.103565833188503e-06,
+      "loss": 0.4891,
+      "step": 6462
+    },
+    {
+      "epoch": 1.0614004475191428,
+      "grad_norm": 0.35102381961251805,
+      "learning_rate": 8.103284926472654e-06,
+      "loss": 0.5034,
+      "step": 6463
+    },
+    {
+      "epoch": 1.0615646747274854,
+      "grad_norm": 0.31904575779002,
+      "learning_rate": 8.103003980621585e-06,
+      "loss": 0.4735,
+      "step": 6464
+    },
+    {
+      "epoch": 1.0617289019358283,
+      "grad_norm": 0.37108455696765513,
+      "learning_rate": 8.10272299563835e-06,
+      "loss": 0.4903,
+      "step": 6465
+    },
+    {
+      "epoch": 1.0618931291441709,
+      "grad_norm": 0.3043493567023591,
+      "learning_rate": 8.102441971525999e-06,
+      "loss": 0.4817,
+      "step": 6466
+    },
+    {
+      "epoch": 1.0620573563525137,
+      "grad_norm": 0.2969183688186469,
+      "learning_rate": 8.10216090828758e-06,
+      "loss": 0.4812,
+      "step": 6467
+    },
+    {
+      "epoch": 1.0622215835608564,
+      "grad_norm": 0.2940431258639217,
+      "learning_rate": 8.101879805926152e-06,
+      "loss": 0.468,
+      "step": 6468
+    },
+    {
+      "epoch": 1.0623858107691992,
+      "grad_norm": 0.3572806332667413,
+      "learning_rate": 8.101598664444765e-06,
+      "loss": 0.4983,
+      "step": 6469
+    },
+    {
+      "epoch": 1.0625500379775419,
+      "grad_norm": 0.32983325411802555,
+      "learning_rate": 8.101317483846475e-06,
+      "loss": 0.4785,
+      "step": 6470
+    },
+    {
+      "epoch": 1.0627142651858847,
+      "grad_norm": 0.2903993681330892,
+      "learning_rate": 8.101036264134332e-06,
+      "loss": 0.4691,
+      "step": 6471
+    },
+    {
+      "epoch": 1.0628784923942274,
+      "grad_norm": 0.2829052100129413,
+      "learning_rate": 8.100755005311392e-06,
+      "loss": 0.5018,
+      "step": 6472
+    },
+    {
+      "epoch": 1.0630427196025702,
+      "grad_norm": 0.3091537259100867,
+      "learning_rate": 8.10047370738071e-06,
+      "loss": 0.4959,
+      "step": 6473
+    },
+    {
+      "epoch": 1.0632069468109129,
+      "grad_norm": 0.3072357245938547,
+      "learning_rate": 8.100192370345343e-06,
+      "loss": 0.5023,
+      "step": 6474
+    },
+    {
+      "epoch": 1.0633711740192557,
+      "grad_norm": 0.29934615692093325,
+      "learning_rate": 8.099910994208345e-06,
+      "loss": 0.4843,
+      "step": 6475
+    },
+    {
+      "epoch": 1.0635354012275984,
+      "grad_norm": 0.29597394203379296,
+      "learning_rate": 8.09962957897277e-06,
+      "loss": 0.4746,
+      "step": 6476
+    },
+    {
+      "epoch": 1.0636996284359412,
+      "grad_norm": 0.28468282474338635,
+      "learning_rate": 8.099348124641676e-06,
+      "loss": 0.475,
+      "step": 6477
+    },
+    {
+      "epoch": 1.0638638556442839,
+      "grad_norm": 0.33136689956285587,
+      "learning_rate": 8.09906663121812e-06,
+      "loss": 0.4736,
+      "step": 6478
+    },
+    {
+      "epoch": 1.0640280828526265,
+      "grad_norm": 0.31234683664303364,
+      "learning_rate": 8.09878509870516e-06,
+      "loss": 0.4958,
+      "step": 6479
+    },
+    {
+      "epoch": 1.0641923100609694,
+      "grad_norm": 0.4152643175893618,
+      "learning_rate": 8.098503527105852e-06,
+      "loss": 0.4807,
+      "step": 6480
+    },
+    {
+      "epoch": 1.064356537269312,
+      "grad_norm": 0.35436166148095855,
+      "learning_rate": 8.098221916423257e-06,
+      "loss": 0.4767,
+      "step": 6481
+    },
+    {
+      "epoch": 1.0645207644776549,
+      "grad_norm": 0.2786918130841321,
+      "learning_rate": 8.097940266660431e-06,
+      "loss": 0.4951,
+      "step": 6482
+    },
+    {
+      "epoch": 1.0646849916859975,
+      "grad_norm": 0.2984696445751109,
+      "learning_rate": 8.097658577820436e-06,
+      "loss": 0.5136,
+      "step": 6483
+    },
+    {
+      "epoch": 1.0648492188943404,
+      "grad_norm": 0.2860516275936091,
+      "learning_rate": 8.097376849906326e-06,
+      "loss": 0.5039,
+      "step": 6484
+    },
+    {
+      "epoch": 1.065013446102683,
+      "grad_norm": 0.35011048090627467,
+      "learning_rate": 8.097095082921165e-06,
+      "loss": 0.4975,
+      "step": 6485
+    },
+    {
+      "epoch": 1.0651776733110259,
+      "grad_norm": 0.3063508579204193,
+      "learning_rate": 8.096813276868014e-06,
+      "loss": 0.4957,
+      "step": 6486
+    },
+    {
+      "epoch": 1.0653419005193685,
+      "grad_norm": 0.4053799879807205,
+      "learning_rate": 8.09653143174993e-06,
+      "loss": 0.5002,
+      "step": 6487
+    },
+    {
+      "epoch": 1.0655061277277114,
+      "grad_norm": 0.32329174991449955,
+      "learning_rate": 8.096249547569976e-06,
+      "loss": 0.4846,
+      "step": 6488
+    },
+    {
+      "epoch": 1.065670354936054,
+      "grad_norm": 0.24985790032644659,
+      "learning_rate": 8.095967624331216e-06,
+      "loss": 0.4756,
+      "step": 6489
+    },
+    {
+      "epoch": 1.0658345821443969,
+      "grad_norm": 0.2642583003677523,
+      "learning_rate": 8.095685662036706e-06,
+      "loss": 0.4837,
+      "step": 6490
+    },
+    {
+      "epoch": 1.0659988093527395,
+      "grad_norm": 0.286132146880121,
+      "learning_rate": 8.095403660689514e-06,
+      "loss": 0.4827,
+      "step": 6491
+    },
+    {
+      "epoch": 1.0661630365610824,
+      "grad_norm": 0.2745315651876201,
+      "learning_rate": 8.0951216202927e-06,
+      "loss": 0.482,
+      "step": 6492
+    },
+    {
+      "epoch": 1.066327263769425,
+      "grad_norm": 0.34287529332732586,
+      "learning_rate": 8.094839540849332e-06,
+      "loss": 0.4928,
+      "step": 6493
+    },
+    {
+      "epoch": 1.0664914909777679,
+      "grad_norm": 0.3333024161421935,
+      "learning_rate": 8.094557422362467e-06,
+      "loss": 0.5016,
+      "step": 6494
+    },
+    {
+      "epoch": 1.0666557181861105,
+      "grad_norm": 0.35575934337835535,
+      "learning_rate": 8.094275264835171e-06,
+      "loss": 0.4946,
+      "step": 6495
+    },
+    {
+      "epoch": 1.0668199453944531,
+      "grad_norm": 0.3056346750452761,
+      "learning_rate": 8.09399306827051e-06,
+      "loss": 0.4858,
+      "step": 6496
+    },
+    {
+      "epoch": 1.066984172602796,
+      "grad_norm": 0.31099343207150504,
+      "learning_rate": 8.093710832671548e-06,
+      "loss": 0.5108,
+      "step": 6497
+    },
+    {
+      "epoch": 1.0671483998111386,
+      "grad_norm": 0.322914818516821,
+      "learning_rate": 8.09342855804135e-06,
+      "loss": 0.4811,
+      "step": 6498
+    },
+    {
+      "epoch": 1.0673126270194815,
+      "grad_norm": 0.300902999390282,
+      "learning_rate": 8.093146244382983e-06,
+      "loss": 0.4938,
+      "step": 6499
+    },
+    {
+      "epoch": 1.0674768542278241,
+      "grad_norm": 0.31601742707388225,
+      "learning_rate": 8.092863891699512e-06,
+      "loss": 0.478,
+      "step": 6500
+    },
+    {
+      "epoch": 1.067641081436167,
+      "grad_norm": 0.3046928653259927,
+      "learning_rate": 8.092581499994007e-06,
+      "loss": 0.4958,
+      "step": 6501
+    },
+    {
+      "epoch": 1.0678053086445096,
+      "grad_norm": 0.3645549423936772,
+      "learning_rate": 8.09229906926953e-06,
+      "loss": 0.4993,
+      "step": 6502
+    },
+    {
+      "epoch": 1.0679695358528525,
+      "grad_norm": 0.2768289330042299,
+      "learning_rate": 8.092016599529151e-06,
+      "loss": 0.4973,
+      "step": 6503
+    },
+    {
+      "epoch": 1.0681337630611951,
+      "grad_norm": 0.36660395418493563,
+      "learning_rate": 8.091734090775939e-06,
+      "loss": 0.4891,
+      "step": 6504
+    },
+    {
+      "epoch": 1.068297990269538,
+      "grad_norm": 0.3270702435861176,
+      "learning_rate": 8.09145154301296e-06,
+      "loss": 0.4957,
+      "step": 6505
+    },
+    {
+      "epoch": 1.0684622174778806,
+      "grad_norm": 0.3498736633281296,
+      "learning_rate": 8.091168956243282e-06,
+      "loss": 0.4966,
+      "step": 6506
+    },
+    {
+      "epoch": 1.0686264446862235,
+      "grad_norm": 0.2997356142625861,
+      "learning_rate": 8.090886330469978e-06,
+      "loss": 0.5231,
+      "step": 6507
+    },
+    {
+      "epoch": 1.0687906718945661,
+      "grad_norm": 0.3697001884215529,
+      "learning_rate": 8.090603665696114e-06,
+      "loss": 0.4855,
+      "step": 6508
+    },
+    {
+      "epoch": 1.068954899102909,
+      "grad_norm": 0.5614008437965243,
+      "learning_rate": 8.090320961924763e-06,
+      "loss": 0.4989,
+      "step": 6509
+    },
+    {
+      "epoch": 1.0691191263112516,
+      "grad_norm": 0.3112432296277617,
+      "learning_rate": 8.090038219158993e-06,
+      "loss": 0.4942,
+      "step": 6510
+    },
+    {
+      "epoch": 1.0692833535195945,
+      "grad_norm": 0.3863364983192195,
+      "learning_rate": 8.089755437401877e-06,
+      "loss": 0.4669,
+      "step": 6511
+    },
+    {
+      "epoch": 1.069447580727937,
+      "grad_norm": 0.2590499894062906,
+      "learning_rate": 8.089472616656484e-06,
+      "loss": 0.5017,
+      "step": 6512
+    },
+    {
+      "epoch": 1.0696118079362797,
+      "grad_norm": 0.3294195425723214,
+      "learning_rate": 8.089189756925888e-06,
+      "loss": 0.4761,
+      "step": 6513
+    },
+    {
+      "epoch": 1.0697760351446226,
+      "grad_norm": 0.3477239348659754,
+      "learning_rate": 8.088906858213158e-06,
+      "loss": 0.5022,
+      "step": 6514
+    },
+    {
+      "epoch": 1.0699402623529652,
+      "grad_norm": 0.2957800807082142,
+      "learning_rate": 8.08862392052137e-06,
+      "loss": 0.4896,
+      "step": 6515
+    },
+    {
+      "epoch": 1.070104489561308,
+      "grad_norm": 0.277456686832057,
+      "learning_rate": 8.088340943853595e-06,
+      "loss": 0.4879,
+      "step": 6516
+    },
+    {
+      "epoch": 1.0702687167696507,
+      "grad_norm": 0.2896224449064534,
+      "learning_rate": 8.088057928212907e-06,
+      "loss": 0.4825,
+      "step": 6517
+    },
+    {
+      "epoch": 1.0704329439779936,
+      "grad_norm": 0.366817404524916,
+      "learning_rate": 8.08777487360238e-06,
+      "loss": 0.5051,
+      "step": 6518
+    },
+    {
+      "epoch": 1.0705971711863362,
+      "grad_norm": 0.3243539755109171,
+      "learning_rate": 8.087491780025088e-06,
+      "loss": 0.4889,
+      "step": 6519
+    },
+    {
+      "epoch": 1.070761398394679,
+      "grad_norm": 0.28142740942527517,
+      "learning_rate": 8.087208647484104e-06,
+      "loss": 0.4679,
+      "step": 6520
+    },
+    {
+      "epoch": 1.0709256256030217,
+      "grad_norm": 0.29184571231550355,
+      "learning_rate": 8.086925475982506e-06,
+      "loss": 0.4633,
+      "step": 6521
+    },
+    {
+      "epoch": 1.0710898528113646,
+      "grad_norm": 0.3149114520591637,
+      "learning_rate": 8.08664226552337e-06,
+      "loss": 0.4862,
+      "step": 6522
+    },
+    {
+      "epoch": 1.0712540800197072,
+      "grad_norm": 0.3523906143591502,
+      "learning_rate": 8.086359016109768e-06,
+      "loss": 0.4926,
+      "step": 6523
+    },
+    {
+      "epoch": 1.07141830722805,
+      "grad_norm": 0.2883075445690294,
+      "learning_rate": 8.08607572774478e-06,
+      "loss": 0.4749,
+      "step": 6524
+    },
+    {
+      "epoch": 1.0715825344363927,
+      "grad_norm": 0.31977848898120104,
+      "learning_rate": 8.08579240043148e-06,
+      "loss": 0.4897,
+      "step": 6525
+    },
+    {
+      "epoch": 1.0717467616447356,
+      "grad_norm": 0.27525095872824124,
+      "learning_rate": 8.085509034172947e-06,
+      "loss": 0.4904,
+      "step": 6526
+    },
+    {
+      "epoch": 1.0719109888530782,
+      "grad_norm": 0.31146255558943825,
+      "learning_rate": 8.085225628972259e-06,
+      "loss": 0.5011,
+      "step": 6527
+    },
+    {
+      "epoch": 1.072075216061421,
+      "grad_norm": 0.3464338330797061,
+      "learning_rate": 8.084942184832492e-06,
+      "loss": 0.4926,
+      "step": 6528
+    },
+    {
+      "epoch": 1.0722394432697637,
+      "grad_norm": 0.27288841993252994,
+      "learning_rate": 8.084658701756726e-06,
+      "loss": 0.4704,
+      "step": 6529
+    },
+    {
+      "epoch": 1.0724036704781064,
+      "grad_norm": 0.4372370882014702,
+      "learning_rate": 8.08437517974804e-06,
+      "loss": 0.4896,
+      "step": 6530
+    },
+    {
+      "epoch": 1.0725678976864492,
+      "grad_norm": 0.30634972416888445,
+      "learning_rate": 8.084091618809513e-06,
+      "loss": 0.4975,
+      "step": 6531
+    },
+    {
+      "epoch": 1.0727321248947919,
+      "grad_norm": 0.3221699368499719,
+      "learning_rate": 8.083808018944226e-06,
+      "loss": 0.4745,
+      "step": 6532
+    },
+    {
+      "epoch": 1.0728963521031347,
+      "grad_norm": 0.31599052239465103,
+      "learning_rate": 8.083524380155257e-06,
+      "loss": 0.4781,
+      "step": 6533
+    },
+    {
+      "epoch": 1.0730605793114774,
+      "grad_norm": 0.302902407696806,
+      "learning_rate": 8.083240702445687e-06,
+      "loss": 0.4798,
+      "step": 6534
+    },
+    {
+      "epoch": 1.0732248065198202,
+      "grad_norm": 0.29767747935360217,
+      "learning_rate": 8.082956985818598e-06,
+      "loss": 0.4879,
+      "step": 6535
+    },
+    {
+      "epoch": 1.0733890337281629,
+      "grad_norm": 0.30394027053429523,
+      "learning_rate": 8.08267323027707e-06,
+      "loss": 0.5003,
+      "step": 6536
+    },
+    {
+      "epoch": 1.0735532609365057,
+      "grad_norm": 0.2854278337518416,
+      "learning_rate": 8.082389435824187e-06,
+      "loss": 0.4773,
+      "step": 6537
+    },
+    {
+      "epoch": 1.0737174881448484,
+      "grad_norm": 0.3401948261966552,
+      "learning_rate": 8.08210560246303e-06,
+      "loss": 0.5256,
+      "step": 6538
+    },
+    {
+      "epoch": 1.0738817153531912,
+      "grad_norm": 0.287518456779464,
+      "learning_rate": 8.081821730196682e-06,
+      "loss": 0.4981,
+      "step": 6539
+    },
+    {
+      "epoch": 1.0740459425615339,
+      "grad_norm": 0.32190628608598426,
+      "learning_rate": 8.081537819028225e-06,
+      "loss": 0.4962,
+      "step": 6540
+    },
+    {
+      "epoch": 1.0742101697698767,
+      "grad_norm": 0.37003432367879086,
+      "learning_rate": 8.081253868960745e-06,
+      "loss": 0.4993,
+      "step": 6541
+    },
+    {
+      "epoch": 1.0743743969782193,
+      "grad_norm": 0.2904353068128155,
+      "learning_rate": 8.080969879997323e-06,
+      "loss": 0.4684,
+      "step": 6542
+    },
+    {
+      "epoch": 1.0745386241865622,
+      "grad_norm": 0.29501265903261076,
+      "learning_rate": 8.080685852141045e-06,
+      "loss": 0.4512,
+      "step": 6543
+    },
+    {
+      "epoch": 1.0747028513949048,
+      "grad_norm": 0.2885464571816661,
+      "learning_rate": 8.080401785394997e-06,
+      "loss": 0.5054,
+      "step": 6544
+    },
+    {
+      "epoch": 1.0748670786032477,
+      "grad_norm": 0.282323525694132,
+      "learning_rate": 8.08011767976226e-06,
+      "loss": 0.4795,
+      "step": 6545
+    },
+    {
+      "epoch": 1.0750313058115903,
+      "grad_norm": 0.31958791943190906,
+      "learning_rate": 8.079833535245927e-06,
+      "loss": 0.5158,
+      "step": 6546
+    },
+    {
+      "epoch": 1.075195533019933,
+      "grad_norm": 0.2982945132663225,
+      "learning_rate": 8.079549351849077e-06,
+      "loss": 0.4961,
+      "step": 6547
+    },
+    {
+      "epoch": 1.0753597602282758,
+      "grad_norm": 0.3665811507990705,
+      "learning_rate": 8.0792651295748e-06,
+      "loss": 0.488,
+      "step": 6548
+    },
+    {
+      "epoch": 1.0755239874366185,
+      "grad_norm": 0.324428934023983,
+      "learning_rate": 8.078980868426183e-06,
+      "loss": 0.5029,
+      "step": 6549
+    },
+    {
+      "epoch": 1.0756882146449613,
+      "grad_norm": 0.30815951965964145,
+      "learning_rate": 8.078696568406311e-06,
+      "loss": 0.4726,
+      "step": 6550
+    },
+    {
+      "epoch": 1.075852441853304,
+      "grad_norm": 0.6351629022687446,
+      "learning_rate": 8.078412229518273e-06,
+      "loss": 0.4868,
+      "step": 6551
+    },
+    {
+      "epoch": 1.0760166690616468,
+      "grad_norm": 0.27095936706763635,
+      "learning_rate": 8.07812785176516e-06,
+      "loss": 0.4846,
+      "step": 6552
+    },
+    {
+      "epoch": 1.0761808962699895,
+      "grad_norm": 0.3373418483243227,
+      "learning_rate": 8.077843435150056e-06,
+      "loss": 0.5021,
+      "step": 6553
+    },
+    {
+      "epoch": 1.0763451234783323,
+      "grad_norm": 0.2769493764982112,
+      "learning_rate": 8.077558979676052e-06,
+      "loss": 0.4995,
+      "step": 6554
+    },
+    {
+      "epoch": 1.076509350686675,
+      "grad_norm": 0.3391246782727593,
+      "learning_rate": 8.077274485346239e-06,
+      "loss": 0.4947,
+      "step": 6555
+    },
+    {
+      "epoch": 1.0766735778950178,
+      "grad_norm": 0.3194796445907453,
+      "learning_rate": 8.076989952163704e-06,
+      "loss": 0.501,
+      "step": 6556
+    },
+    {
+      "epoch": 1.0768378051033605,
+      "grad_norm": 0.2884927330411459,
+      "learning_rate": 8.07670538013154e-06,
+      "loss": 0.4765,
+      "step": 6557
+    },
+    {
+      "epoch": 1.0770020323117033,
+      "grad_norm": 0.27582021063538154,
+      "learning_rate": 8.076420769252837e-06,
+      "loss": 0.4917,
+      "step": 6558
+    },
+    {
+      "epoch": 1.077166259520046,
+      "grad_norm": 0.32878546008855475,
+      "learning_rate": 8.076136119530685e-06,
+      "loss": 0.4925,
+      "step": 6559
+    },
+    {
+      "epoch": 1.0773304867283888,
+      "grad_norm": 0.3719181139981429,
+      "learning_rate": 8.075851430968176e-06,
+      "loss": 0.4834,
+      "step": 6560
+    },
+    {
+      "epoch": 1.0774947139367315,
+      "grad_norm": 0.299817050063877,
+      "learning_rate": 8.075566703568402e-06,
+      "loss": 0.4817,
+      "step": 6561
+    },
+    {
+      "epoch": 1.0776589411450743,
+      "grad_norm": 0.27926753984629715,
+      "learning_rate": 8.075281937334456e-06,
+      "loss": 0.4684,
+      "step": 6562
+    },
+    {
+      "epoch": 1.077823168353417,
+      "grad_norm": 0.3141612633195604,
+      "learning_rate": 8.074997132269431e-06,
+      "loss": 0.4943,
+      "step": 6563
+    },
+    {
+      "epoch": 1.0779873955617596,
+      "grad_norm": 0.3227802156237344,
+      "learning_rate": 8.07471228837642e-06,
+      "loss": 0.4909,
+      "step": 6564
+    },
+    {
+      "epoch": 1.0781516227701025,
+      "grad_norm": 0.27393818022295974,
+      "learning_rate": 8.074427405658516e-06,
+      "loss": 0.479,
+      "step": 6565
+    },
+    {
+      "epoch": 1.078315849978445,
+      "grad_norm": 0.34096578885764833,
+      "learning_rate": 8.074142484118814e-06,
+      "loss": 0.4819,
+      "step": 6566
+    },
+    {
+      "epoch": 1.078480077186788,
+      "grad_norm": 0.3628942414376617,
+      "learning_rate": 8.073857523760407e-06,
+      "loss": 0.4822,
+      "step": 6567
+    },
+    {
+      "epoch": 1.0786443043951306,
+      "grad_norm": 0.3121047978071521,
+      "learning_rate": 8.073572524586392e-06,
+      "loss": 0.499,
+      "step": 6568
+    },
+    {
+      "epoch": 1.0788085316034735,
+      "grad_norm": 0.3710136396120407,
+      "learning_rate": 8.073287486599864e-06,
+      "loss": 0.5031,
+      "step": 6569
+    },
+    {
+      "epoch": 1.078972758811816,
+      "grad_norm": 0.27173145739016036,
+      "learning_rate": 8.073002409803917e-06,
+      "loss": 0.4897,
+      "step": 6570
+    },
+    {
+      "epoch": 1.079136986020159,
+      "grad_norm": 0.3607583744416979,
+      "learning_rate": 8.072717294201649e-06,
+      "loss": 0.498,
+      "step": 6571
+    },
+    {
+      "epoch": 1.0793012132285016,
+      "grad_norm": 0.33623588695032847,
+      "learning_rate": 8.072432139796157e-06,
+      "loss": 0.4989,
+      "step": 6572
+    },
+    {
+      "epoch": 1.0794654404368444,
+      "grad_norm": 0.3315744739698578,
+      "learning_rate": 8.072146946590536e-06,
+      "loss": 0.4796,
+      "step": 6573
+    },
+    {
+      "epoch": 1.079629667645187,
+      "grad_norm": 0.3222025399523138,
+      "learning_rate": 8.071861714587885e-06,
+      "loss": 0.4918,
+      "step": 6574
+    },
+    {
+      "epoch": 1.07979389485353,
+      "grad_norm": 0.31798138184786123,
+      "learning_rate": 8.071576443791302e-06,
+      "loss": 0.496,
+      "step": 6575
+    },
+    {
+      "epoch": 1.0799581220618726,
+      "grad_norm": 0.29852066241948144,
+      "learning_rate": 8.071291134203885e-06,
+      "loss": 0.4937,
+      "step": 6576
+    },
+    {
+      "epoch": 1.0801223492702154,
+      "grad_norm": 0.26278834220161956,
+      "learning_rate": 8.071005785828732e-06,
+      "loss": 0.4872,
+      "step": 6577
+    },
+    {
+      "epoch": 1.080286576478558,
+      "grad_norm": 0.2891753603047115,
+      "learning_rate": 8.070720398668944e-06,
+      "loss": 0.4777,
+      "step": 6578
+    },
+    {
+      "epoch": 1.080450803686901,
+      "grad_norm": 0.3858364653964431,
+      "learning_rate": 8.070434972727617e-06,
+      "loss": 0.5132,
+      "step": 6579
+    },
+    {
+      "epoch": 1.0806150308952436,
+      "grad_norm": 0.3764177313317822,
+      "learning_rate": 8.070149508007854e-06,
+      "loss": 0.5014,
+      "step": 6580
+    },
+    {
+      "epoch": 1.0807792581035862,
+      "grad_norm": 0.37878161708248925,
+      "learning_rate": 8.069864004512756e-06,
+      "loss": 0.4936,
+      "step": 6581
+    },
+    {
+      "epoch": 1.080943485311929,
+      "grad_norm": 0.278614098273255,
+      "learning_rate": 8.069578462245422e-06,
+      "loss": 0.4762,
+      "step": 6582
+    },
+    {
+      "epoch": 1.0811077125202717,
+      "grad_norm": 0.32093940639927887,
+      "learning_rate": 8.069292881208955e-06,
+      "loss": 0.4862,
+      "step": 6583
+    },
+    {
+      "epoch": 1.0812719397286146,
+      "grad_norm": 0.29447297248866366,
+      "learning_rate": 8.069007261406454e-06,
+      "loss": 0.4892,
+      "step": 6584
+    },
+    {
+      "epoch": 1.0814361669369572,
+      "grad_norm": 0.36869470719217107,
+      "learning_rate": 8.068721602841023e-06,
+      "loss": 0.5079,
+      "step": 6585
+    },
+    {
+      "epoch": 1.0816003941453,
+      "grad_norm": 0.28216039771339985,
+      "learning_rate": 8.068435905515764e-06,
+      "loss": 0.4853,
+      "step": 6586
+    },
+    {
+      "epoch": 1.0817646213536427,
+      "grad_norm": 0.3304883936698149,
+      "learning_rate": 8.068150169433781e-06,
+      "loss": 0.5068,
+      "step": 6587
+    },
+    {
+      "epoch": 1.0819288485619856,
+      "grad_norm": 0.33187478705588386,
+      "learning_rate": 8.067864394598177e-06,
+      "loss": 0.4933,
+      "step": 6588
+    },
+    {
+      "epoch": 1.0820930757703282,
+      "grad_norm": 0.3068115166683626,
+      "learning_rate": 8.067578581012054e-06,
+      "loss": 0.5052,
+      "step": 6589
+    },
+    {
+      "epoch": 1.082257302978671,
+      "grad_norm": 0.26991613220821475,
+      "learning_rate": 8.067292728678519e-06,
+      "loss": 0.5059,
+      "step": 6590
+    },
+    {
+      "epoch": 1.0824215301870137,
+      "grad_norm": 0.3177761633438965,
+      "learning_rate": 8.067006837600674e-06,
+      "loss": 0.4982,
+      "step": 6591
+    },
+    {
+      "epoch": 1.0825857573953566,
+      "grad_norm": 0.27743314130195196,
+      "learning_rate": 8.066720907781625e-06,
+      "loss": 0.473,
+      "step": 6592
+    },
+    {
+      "epoch": 1.0827499846036992,
+      "grad_norm": 0.27281498017793654,
+      "learning_rate": 8.066434939224478e-06,
+      "loss": 0.4999,
+      "step": 6593
+    },
+    {
+      "epoch": 1.082914211812042,
+      "grad_norm": 0.29010521029271646,
+      "learning_rate": 8.06614893193234e-06,
+      "loss": 0.4622,
+      "step": 6594
+    },
+    {
+      "epoch": 1.0830784390203847,
+      "grad_norm": 0.3875811075929337,
+      "learning_rate": 8.065862885908317e-06,
+      "loss": 0.5063,
+      "step": 6595
+    },
+    {
+      "epoch": 1.0832426662287276,
+      "grad_norm": 0.30672239793347644,
+      "learning_rate": 8.065576801155512e-06,
+      "loss": 0.4781,
+      "step": 6596
+    },
+    {
+      "epoch": 1.0834068934370702,
+      "grad_norm": 0.2790991854199382,
+      "learning_rate": 8.065290677677036e-06,
+      "loss": 0.46,
+      "step": 6597
+    },
+    {
+      "epoch": 1.0835711206454128,
+      "grad_norm": 0.37332831621147183,
+      "learning_rate": 8.065004515475994e-06,
+      "loss": 0.508,
+      "step": 6598
+    },
+    {
+      "epoch": 1.0837353478537557,
+      "grad_norm": 0.3434760999904214,
+      "learning_rate": 8.064718314555497e-06,
+      "loss": 0.4793,
+      "step": 6599
+    },
+    {
+      "epoch": 1.0838995750620983,
+      "grad_norm": 0.6233700972166917,
+      "learning_rate": 8.06443207491865e-06,
+      "loss": 0.4577,
+      "step": 6600
+    },
+    {
+      "epoch": 1.0840638022704412,
+      "grad_norm": 0.3384347060463355,
+      "learning_rate": 8.064145796568567e-06,
+      "loss": 0.4932,
+      "step": 6601
+    },
+    {
+      "epoch": 1.0842280294787838,
+      "grad_norm": 0.289153448823887,
+      "learning_rate": 8.063859479508352e-06,
+      "loss": 0.4912,
+      "step": 6602
+    },
+    {
+      "epoch": 1.0843922566871267,
+      "grad_norm": 0.26405473877925406,
+      "learning_rate": 8.063573123741117e-06,
+      "loss": 0.5015,
+      "step": 6603
+    },
+    {
+      "epoch": 1.0845564838954693,
+      "grad_norm": 0.31409076950725495,
+      "learning_rate": 8.063286729269971e-06,
+      "loss": 0.4959,
+      "step": 6604
+    },
+    {
+      "epoch": 1.0847207111038122,
+      "grad_norm": 0.30705620044601656,
+      "learning_rate": 8.063000296098026e-06,
+      "loss": 0.4956,
+      "step": 6605
+    },
+    {
+      "epoch": 1.0848849383121548,
+      "grad_norm": 0.31862613078857577,
+      "learning_rate": 8.062713824228393e-06,
+      "loss": 0.5036,
+      "step": 6606
+    },
+    {
+      "epoch": 1.0850491655204977,
+      "grad_norm": 0.7253417981589776,
+      "learning_rate": 8.062427313664183e-06,
+      "loss": 0.4844,
+      "step": 6607
+    },
+    {
+      "epoch": 1.0852133927288403,
+      "grad_norm": 0.7457709244399511,
+      "learning_rate": 8.062140764408505e-06,
+      "loss": 0.4859,
+      "step": 6608
+    },
+    {
+      "epoch": 1.0853776199371832,
+      "grad_norm": 0.300353254880477,
+      "learning_rate": 8.061854176464477e-06,
+      "loss": 0.4849,
+      "step": 6609
+    },
+    {
+      "epoch": 1.0855418471455258,
+      "grad_norm": 0.3253936123596342,
+      "learning_rate": 8.061567549835206e-06,
+      "loss": 0.4553,
+      "step": 6610
+    },
+    {
+      "epoch": 1.0857060743538687,
+      "grad_norm": 0.3120425180000209,
+      "learning_rate": 8.061280884523808e-06,
+      "loss": 0.4681,
+      "step": 6611
+    },
+    {
+      "epoch": 1.0858703015622113,
+      "grad_norm": 0.34523817729463896,
+      "learning_rate": 8.060994180533395e-06,
+      "loss": 0.4912,
+      "step": 6612
+    },
+    {
+      "epoch": 1.0860345287705542,
+      "grad_norm": 0.3159024449975521,
+      "learning_rate": 8.060707437867082e-06,
+      "loss": 0.5238,
+      "step": 6613
+    },
+    {
+      "epoch": 1.0861987559788968,
+      "grad_norm": 0.2708163187055613,
+      "learning_rate": 8.060420656527983e-06,
+      "loss": 0.4802,
+      "step": 6614
+    },
+    {
+      "epoch": 1.0863629831872395,
+      "grad_norm": 0.2717693408294249,
+      "learning_rate": 8.060133836519213e-06,
+      "loss": 0.5011,
+      "step": 6615
+    },
+    {
+      "epoch": 1.0865272103955823,
+      "grad_norm": 0.3046084952629126,
+      "learning_rate": 8.059846977843885e-06,
+      "loss": 0.4808,
+      "step": 6616
+    },
+    {
+      "epoch": 1.086691437603925,
+      "grad_norm": 0.3484788088356775,
+      "learning_rate": 8.059560080505119e-06,
+      "loss": 0.4709,
+      "step": 6617
+    },
+    {
+      "epoch": 1.0868556648122678,
+      "grad_norm": 0.4789898019185958,
+      "learning_rate": 8.059273144506029e-06,
+      "loss": 0.4782,
+      "step": 6618
+    },
+    {
+      "epoch": 1.0870198920206104,
+      "grad_norm": 0.2721149994086529,
+      "learning_rate": 8.058986169849727e-06,
+      "loss": 0.5058,
+      "step": 6619
+    },
+    {
+      "epoch": 1.0871841192289533,
+      "grad_norm": 0.5085546027446511,
+      "learning_rate": 8.058699156539336e-06,
+      "loss": 0.4936,
+      "step": 6620
+    },
+    {
+      "epoch": 1.087348346437296,
+      "grad_norm": 0.28291266607791543,
+      "learning_rate": 8.058412104577971e-06,
+      "loss": 0.4956,
+      "step": 6621
+    },
+    {
+      "epoch": 1.0875125736456388,
+      "grad_norm": 0.24750759560662303,
+      "learning_rate": 8.058125013968749e-06,
+      "loss": 0.4757,
+      "step": 6622
+    },
+    {
+      "epoch": 1.0876768008539814,
+      "grad_norm": 0.2834097176537607,
+      "learning_rate": 8.057837884714789e-06,
+      "loss": 0.4825,
+      "step": 6623
+    },
+    {
+      "epoch": 1.0878410280623243,
+      "grad_norm": 0.6745488072249444,
+      "learning_rate": 8.05755071681921e-06,
+      "loss": 0.5016,
+      "step": 6624
+    },
+    {
+      "epoch": 1.088005255270667,
+      "grad_norm": 0.26972117677597274,
+      "learning_rate": 8.057263510285128e-06,
+      "loss": 0.4905,
+      "step": 6625
+    },
+    {
+      "epoch": 1.0881694824790098,
+      "grad_norm": 0.38028687216798407,
+      "learning_rate": 8.056976265115665e-06,
+      "loss": 0.5123,
+      "step": 6626
+    },
+    {
+      "epoch": 1.0883337096873524,
+      "grad_norm": 0.32171686241552383,
+      "learning_rate": 8.056688981313942e-06,
+      "loss": 0.5023,
+      "step": 6627
+    },
+    {
+      "epoch": 1.0884979368956953,
+      "grad_norm": 0.35366407163708424,
+      "learning_rate": 8.056401658883075e-06,
+      "loss": 0.4843,
+      "step": 6628
+    },
+    {
+      "epoch": 1.088662164104038,
+      "grad_norm": 0.31785783871908235,
+      "learning_rate": 8.056114297826187e-06,
+      "loss": 0.464,
+      "step": 6629
+    },
+    {
+      "epoch": 1.0888263913123808,
+      "grad_norm": 0.37713493307433466,
+      "learning_rate": 8.055826898146401e-06,
+      "loss": 0.5046,
+      "step": 6630
+    },
+    {
+      "epoch": 1.0889906185207234,
+      "grad_norm": 0.30751278180262237,
+      "learning_rate": 8.055539459846836e-06,
+      "loss": 0.4861,
+      "step": 6631
+    },
+    {
+      "epoch": 1.089154845729066,
+      "grad_norm": 0.2679614159390897,
+      "learning_rate": 8.055251982930612e-06,
+      "loss": 0.5073,
+      "step": 6632
+    },
+    {
+      "epoch": 1.089319072937409,
+      "grad_norm": 0.3583802438166624,
+      "learning_rate": 8.054964467400856e-06,
+      "loss": 0.4893,
+      "step": 6633
+    },
+    {
+      "epoch": 1.0894833001457516,
+      "grad_norm": 0.281128336748366,
+      "learning_rate": 8.054676913260687e-06,
+      "loss": 0.4896,
+      "step": 6634
+    },
+    {
+      "epoch": 1.0896475273540944,
+      "grad_norm": 0.31164083615547156,
+      "learning_rate": 8.054389320513229e-06,
+      "loss": 0.4937,
+      "step": 6635
+    },
+    {
+      "epoch": 1.089811754562437,
+      "grad_norm": 0.2842462734493817,
+      "learning_rate": 8.054101689161607e-06,
+      "loss": 0.5003,
+      "step": 6636
+    },
+    {
+      "epoch": 1.08997598177078,
+      "grad_norm": 0.41851163082840925,
+      "learning_rate": 8.053814019208944e-06,
+      "loss": 0.4766,
+      "step": 6637
+    },
+    {
+      "epoch": 1.0901402089791226,
+      "grad_norm": 0.32067594574199665,
+      "learning_rate": 8.053526310658364e-06,
+      "loss": 0.4981,
+      "step": 6638
+    },
+    {
+      "epoch": 1.0903044361874654,
+      "grad_norm": 0.39601430427739676,
+      "learning_rate": 8.053238563512993e-06,
+      "loss": 0.4858,
+      "step": 6639
+    },
+    {
+      "epoch": 1.090468663395808,
+      "grad_norm": 0.31151104434543075,
+      "learning_rate": 8.052950777775953e-06,
+      "loss": 0.4751,
+      "step": 6640
+    },
+    {
+      "epoch": 1.090632890604151,
+      "grad_norm": 0.35616453883131216,
+      "learning_rate": 8.052662953450373e-06,
+      "loss": 0.4833,
+      "step": 6641
+    },
+    {
+      "epoch": 1.0907971178124936,
+      "grad_norm": 0.32908409486875606,
+      "learning_rate": 8.05237509053938e-06,
+      "loss": 0.5086,
+      "step": 6642
+    },
+    {
+      "epoch": 1.0909613450208364,
+      "grad_norm": 0.34134888377075057,
+      "learning_rate": 8.052087189046095e-06,
+      "loss": 0.5087,
+      "step": 6643
+    },
+    {
+      "epoch": 1.091125572229179,
+      "grad_norm": 0.47345364219699215,
+      "learning_rate": 8.05179924897365e-06,
+      "loss": 0.5156,
+      "step": 6644
+    },
+    {
+      "epoch": 1.091289799437522,
+      "grad_norm": 0.3445889661685117,
+      "learning_rate": 8.05151127032517e-06,
+      "loss": 0.4996,
+      "step": 6645
+    },
+    {
+      "epoch": 1.0914540266458646,
+      "grad_norm": 0.29551209314615595,
+      "learning_rate": 8.051223253103785e-06,
+      "loss": 0.4818,
+      "step": 6646
+    },
+    {
+      "epoch": 1.0916182538542074,
+      "grad_norm": 0.2947551790268186,
+      "learning_rate": 8.05093519731262e-06,
+      "loss": 0.4825,
+      "step": 6647
+    },
+    {
+      "epoch": 1.09178248106255,
+      "grad_norm": 0.3834351018766136,
+      "learning_rate": 8.050647102954806e-06,
+      "loss": 0.4972,
+      "step": 6648
+    },
+    {
+      "epoch": 1.0919467082708927,
+      "grad_norm": 0.3396270899894368,
+      "learning_rate": 8.050358970033471e-06,
+      "loss": 0.4774,
+      "step": 6649
+    },
+    {
+      "epoch": 1.0921109354792355,
+      "grad_norm": 0.2892429889400422,
+      "learning_rate": 8.050070798551745e-06,
+      "loss": 0.5017,
+      "step": 6650
+    },
+    {
+      "epoch": 1.0922751626875782,
+      "grad_norm": 0.2861359821069163,
+      "learning_rate": 8.049782588512757e-06,
+      "loss": 0.4789,
+      "step": 6651
+    },
+    {
+      "epoch": 1.092439389895921,
+      "grad_norm": 0.297574352808594,
+      "learning_rate": 8.049494339919636e-06,
+      "loss": 0.483,
+      "step": 6652
+    },
+    {
+      "epoch": 1.0926036171042637,
+      "grad_norm": 0.295904675266878,
+      "learning_rate": 8.049206052775515e-06,
+      "loss": 0.4759,
+      "step": 6653
+    },
+    {
+      "epoch": 1.0927678443126065,
+      "grad_norm": 0.27957006578379917,
+      "learning_rate": 8.048917727083526e-06,
+      "loss": 0.4742,
+      "step": 6654
+    },
+    {
+      "epoch": 1.0929320715209492,
+      "grad_norm": 0.33890350213174564,
+      "learning_rate": 8.048629362846796e-06,
+      "loss": 0.4757,
+      "step": 6655
+    },
+    {
+      "epoch": 1.093096298729292,
+      "grad_norm": 0.3047428157307131,
+      "learning_rate": 8.04834096006846e-06,
+      "loss": 0.4861,
+      "step": 6656
+    },
+    {
+      "epoch": 1.0932605259376347,
+      "grad_norm": 0.5228184394903771,
+      "learning_rate": 8.048052518751653e-06,
+      "loss": 0.4754,
+      "step": 6657
+    },
+    {
+      "epoch": 1.0934247531459775,
+      "grad_norm": 0.3112575437387302,
+      "learning_rate": 8.047764038899505e-06,
+      "loss": 0.4831,
+      "step": 6658
+    },
+    {
+      "epoch": 1.0935889803543202,
+      "grad_norm": 0.29681541702163144,
+      "learning_rate": 8.047475520515147e-06,
+      "loss": 0.4925,
+      "step": 6659
+    },
+    {
+      "epoch": 1.093753207562663,
+      "grad_norm": 0.280871176322509,
+      "learning_rate": 8.047186963601714e-06,
+      "loss": 0.4898,
+      "step": 6660
+    },
+    {
+      "epoch": 1.0939174347710057,
+      "grad_norm": 0.30043549760570193,
+      "learning_rate": 8.04689836816234e-06,
+      "loss": 0.4929,
+      "step": 6661
+    },
+    {
+      "epoch": 1.0940816619793485,
+      "grad_norm": 0.31300225911045476,
+      "learning_rate": 8.046609734200162e-06,
+      "loss": 0.5052,
+      "step": 6662
+    },
+    {
+      "epoch": 1.0942458891876912,
+      "grad_norm": 0.28856161920277196,
+      "learning_rate": 8.046321061718312e-06,
+      "loss": 0.4975,
+      "step": 6663
+    },
+    {
+      "epoch": 1.094410116396034,
+      "grad_norm": 0.26293407990784057,
+      "learning_rate": 8.046032350719928e-06,
+      "loss": 0.4676,
+      "step": 6664
+    },
+    {
+      "epoch": 1.0945743436043767,
+      "grad_norm": 0.4804798489350901,
+      "learning_rate": 8.04574360120814e-06,
+      "loss": 0.5155,
+      "step": 6665
+    },
+    {
+      "epoch": 1.0947385708127193,
+      "grad_norm": 0.29817475251799247,
+      "learning_rate": 8.045454813186092e-06,
+      "loss": 0.4926,
+      "step": 6666
+    },
+    {
+      "epoch": 1.0949027980210622,
+      "grad_norm": 0.2792147268849725,
+      "learning_rate": 8.045165986656914e-06,
+      "loss": 0.4694,
+      "step": 6667
+    },
+    {
+      "epoch": 1.0950670252294048,
+      "grad_norm": 0.26740986430834013,
+      "learning_rate": 8.044877121623747e-06,
+      "loss": 0.47,
+      "step": 6668
+    },
+    {
+      "epoch": 1.0952312524377477,
+      "grad_norm": 0.3175206189654057,
+      "learning_rate": 8.044588218089726e-06,
+      "loss": 0.5061,
+      "step": 6669
+    },
+    {
+      "epoch": 1.0953954796460903,
+      "grad_norm": 0.3521071238807635,
+      "learning_rate": 8.04429927605799e-06,
+      "loss": 0.498,
+      "step": 6670
+    },
+    {
+      "epoch": 1.0955597068544332,
+      "grad_norm": 0.30179167992615835,
+      "learning_rate": 8.044010295531676e-06,
+      "loss": 0.474,
+      "step": 6671
+    },
+    {
+      "epoch": 1.0957239340627758,
+      "grad_norm": 0.41781233652809985,
+      "learning_rate": 8.043721276513922e-06,
+      "loss": 0.4931,
+      "step": 6672
+    },
+    {
+      "epoch": 1.0958881612711187,
+      "grad_norm": 0.32796241643504886,
+      "learning_rate": 8.043432219007872e-06,
+      "loss": 0.4832,
+      "step": 6673
+    },
+    {
+      "epoch": 1.0960523884794613,
+      "grad_norm": 0.2986490602919634,
+      "learning_rate": 8.04314312301666e-06,
+      "loss": 0.4907,
+      "step": 6674
+    },
+    {
+      "epoch": 1.0962166156878042,
+      "grad_norm": 0.3823780932901472,
+      "learning_rate": 8.042853988543427e-06,
+      "loss": 0.4867,
+      "step": 6675
+    },
+    {
+      "epoch": 1.0963808428961468,
+      "grad_norm": 0.3210929957900738,
+      "learning_rate": 8.042564815591314e-06,
+      "loss": 0.4768,
+      "step": 6676
+    },
+    {
+      "epoch": 1.0965450701044896,
+      "grad_norm": 0.37904104957766954,
+      "learning_rate": 8.042275604163462e-06,
+      "loss": 0.4926,
+      "step": 6677
+    },
+    {
+      "epoch": 1.0967092973128323,
+      "grad_norm": 0.3349723807012331,
+      "learning_rate": 8.041986354263013e-06,
+      "loss": 0.5118,
+      "step": 6678
+    },
+    {
+      "epoch": 1.0968735245211751,
+      "grad_norm": 0.3571590062105453,
+      "learning_rate": 8.041697065893105e-06,
+      "loss": 0.4751,
+      "step": 6679
+    },
+    {
+      "epoch": 1.0970377517295178,
+      "grad_norm": 0.2921285386665655,
+      "learning_rate": 8.041407739056885e-06,
+      "loss": 0.4739,
+      "step": 6680
+    },
+    {
+      "epoch": 1.0972019789378606,
+      "grad_norm": 0.31244656526919706,
+      "learning_rate": 8.04111837375749e-06,
+      "loss": 0.4929,
+      "step": 6681
+    },
+    {
+      "epoch": 1.0973662061462033,
+      "grad_norm": 0.47267268926626577,
+      "learning_rate": 8.040828969998068e-06,
+      "loss": 0.4898,
+      "step": 6682
+    },
+    {
+      "epoch": 1.097530433354546,
+      "grad_norm": 0.2693517145732392,
+      "learning_rate": 8.04053952778176e-06,
+      "loss": 0.498,
+      "step": 6683
+    },
+    {
+      "epoch": 1.0976946605628888,
+      "grad_norm": 0.28351936482992657,
+      "learning_rate": 8.040250047111706e-06,
+      "loss": 0.4876,
+      "step": 6684
+    },
+    {
+      "epoch": 1.0978588877712314,
+      "grad_norm": 0.3594674569385834,
+      "learning_rate": 8.039960527991055e-06,
+      "loss": 0.4815,
+      "step": 6685
+    },
+    {
+      "epoch": 1.0980231149795743,
+      "grad_norm": 0.30259751217946457,
+      "learning_rate": 8.03967097042295e-06,
+      "loss": 0.5022,
+      "step": 6686
+    },
+    {
+      "epoch": 1.098187342187917,
+      "grad_norm": 0.38365337530932603,
+      "learning_rate": 8.039381374410536e-06,
+      "loss": 0.4949,
+      "step": 6687
+    },
+    {
+      "epoch": 1.0983515693962598,
+      "grad_norm": 0.30124925300942795,
+      "learning_rate": 8.039091739956959e-06,
+      "loss": 0.5001,
+      "step": 6688
+    },
+    {
+      "epoch": 1.0985157966046024,
+      "grad_norm": 0.2848395875961075,
+      "learning_rate": 8.03880206706536e-06,
+      "loss": 0.5062,
+      "step": 6689
+    },
+    {
+      "epoch": 1.0986800238129453,
+      "grad_norm": 0.4221184803884185,
+      "learning_rate": 8.038512355738892e-06,
+      "loss": 0.4814,
+      "step": 6690
+    },
+    {
+      "epoch": 1.098844251021288,
+      "grad_norm": 0.2897228472309949,
+      "learning_rate": 8.038222605980698e-06,
+      "loss": 0.4858,
+      "step": 6691
+    },
+    {
+      "epoch": 1.0990084782296308,
+      "grad_norm": 0.2999559771720914,
+      "learning_rate": 8.037932817793924e-06,
+      "loss": 0.4975,
+      "step": 6692
+    },
+    {
+      "epoch": 1.0991727054379734,
+      "grad_norm": 0.2796159071321173,
+      "learning_rate": 8.037642991181721e-06,
+      "loss": 0.4843,
+      "step": 6693
+    },
+    {
+      "epoch": 1.0993369326463163,
+      "grad_norm": 0.3118067130960556,
+      "learning_rate": 8.037353126147233e-06,
+      "loss": 0.4975,
+      "step": 6694
+    },
+    {
+      "epoch": 1.099501159854659,
+      "grad_norm": 0.3917793286041874,
+      "learning_rate": 8.03706322269361e-06,
+      "loss": 0.4984,
+      "step": 6695
+    },
+    {
+      "epoch": 1.0996653870630018,
+      "grad_norm": 0.2701048707532968,
+      "learning_rate": 8.036773280824e-06,
+      "loss": 0.4592,
+      "step": 6696
+    },
+    {
+      "epoch": 1.0998296142713444,
+      "grad_norm": 0.3657316253260058,
+      "learning_rate": 8.036483300541554e-06,
+      "loss": 0.4971,
+      "step": 6697
+    },
+    {
+      "epoch": 1.0999938414796873,
+      "grad_norm": 0.3368853629369273,
+      "learning_rate": 8.036193281849419e-06,
+      "loss": 0.478,
+      "step": 6698
+    },
+    {
+      "epoch": 1.10015806868803,
+      "grad_norm": 0.30968426218990286,
+      "learning_rate": 8.035903224750745e-06,
+      "loss": 0.5134,
+      "step": 6699
+    },
+    {
+      "epoch": 1.1003222958963725,
+      "grad_norm": 0.31875699537562413,
+      "learning_rate": 8.035613129248683e-06,
+      "loss": 0.4717,
+      "step": 6700
+    },
+    {
+      "epoch": 1.1004865231047154,
+      "grad_norm": 0.2941617351152997,
+      "learning_rate": 8.035322995346386e-06,
+      "loss": 0.4622,
+      "step": 6701
+    },
+    {
+      "epoch": 1.100650750313058,
+      "grad_norm": 0.2992618349245907,
+      "learning_rate": 8.035032823047001e-06,
+      "loss": 0.4923,
+      "step": 6702
+    },
+    {
+      "epoch": 1.100814977521401,
+      "grad_norm": 0.32832975541354775,
+      "learning_rate": 8.034742612353681e-06,
+      "loss": 0.4931,
+      "step": 6703
+    },
+    {
+      "epoch": 1.1009792047297435,
+      "grad_norm": 0.36302102867175484,
+      "learning_rate": 8.034452363269581e-06,
+      "loss": 0.4783,
+      "step": 6704
+    },
+    {
+      "epoch": 1.1011434319380864,
+      "grad_norm": 0.33724953597312485,
+      "learning_rate": 8.034162075797849e-06,
+      "loss": 0.4849,
+      "step": 6705
+    },
+    {
+      "epoch": 1.101307659146429,
+      "grad_norm": 0.26336764360879583,
+      "learning_rate": 8.033871749941642e-06,
+      "loss": 0.4979,
+      "step": 6706
+    },
+    {
+      "epoch": 1.101471886354772,
+      "grad_norm": 0.46284018414744904,
+      "learning_rate": 8.033581385704108e-06,
+      "loss": 0.4716,
+      "step": 6707
+    },
+    {
+      "epoch": 1.1016361135631145,
+      "grad_norm": 0.26767536775305845,
+      "learning_rate": 8.033290983088405e-06,
+      "loss": 0.4923,
+      "step": 6708
+    },
+    {
+      "epoch": 1.1018003407714574,
+      "grad_norm": 0.33545393017343467,
+      "learning_rate": 8.033000542097685e-06,
+      "loss": 0.4894,
+      "step": 6709
+    },
+    {
+      "epoch": 1.1019645679798,
+      "grad_norm": 0.28161470373946695,
+      "learning_rate": 8.032710062735103e-06,
+      "loss": 0.4919,
+      "step": 6710
+    },
+    {
+      "epoch": 1.1021287951881429,
+      "grad_norm": 0.29950638831768156,
+      "learning_rate": 8.032419545003815e-06,
+      "loss": 0.4772,
+      "step": 6711
+    },
+    {
+      "epoch": 1.1022930223964855,
+      "grad_norm": 0.36968967122204577,
+      "learning_rate": 8.032128988906975e-06,
+      "loss": 0.5008,
+      "step": 6712
+    },
+    {
+      "epoch": 1.1024572496048284,
+      "grad_norm": 0.3152702117734696,
+      "learning_rate": 8.03183839444774e-06,
+      "loss": 0.5156,
+      "step": 6713
+    },
+    {
+      "epoch": 1.102621476813171,
+      "grad_norm": 0.27945506823563687,
+      "learning_rate": 8.031547761629264e-06,
+      "loss": 0.4931,
+      "step": 6714
+    },
+    {
+      "epoch": 1.1027857040215139,
+      "grad_norm": 0.40158231798578,
+      "learning_rate": 8.031257090454704e-06,
+      "loss": 0.4854,
+      "step": 6715
+    },
+    {
+      "epoch": 1.1029499312298565,
+      "grad_norm": 0.3677471976180123,
+      "learning_rate": 8.030966380927218e-06,
+      "loss": 0.4895,
+      "step": 6716
+    },
+    {
+      "epoch": 1.1031141584381992,
+      "grad_norm": 0.32174666226612697,
+      "learning_rate": 8.030675633049964e-06,
+      "loss": 0.5074,
+      "step": 6717
+    },
+    {
+      "epoch": 1.103278385646542,
+      "grad_norm": 0.28039648079111684,
+      "learning_rate": 8.030384846826098e-06,
+      "loss": 0.4949,
+      "step": 6718
+    },
+    {
+      "epoch": 1.1034426128548847,
+      "grad_norm": 0.373231733104368,
+      "learning_rate": 8.03009402225878e-06,
+      "loss": 0.4826,
+      "step": 6719
+    },
+    {
+      "epoch": 1.1036068400632275,
+      "grad_norm": 0.2952935970118553,
+      "learning_rate": 8.029803159351167e-06,
+      "loss": 0.477,
+      "step": 6720
+    },
+    {
+      "epoch": 1.1037710672715701,
+      "grad_norm": 0.708722428069576,
+      "learning_rate": 8.029512258106419e-06,
+      "loss": 0.4978,
+      "step": 6721
+    },
+    {
+      "epoch": 1.103935294479913,
+      "grad_norm": 0.5013498552378116,
+      "learning_rate": 8.029221318527697e-06,
+      "loss": 0.5029,
+      "step": 6722
+    },
+    {
+      "epoch": 1.1040995216882556,
+      "grad_norm": 0.2879331351210175,
+      "learning_rate": 8.028930340618158e-06,
+      "loss": 0.4899,
+      "step": 6723
+    },
+    {
+      "epoch": 1.1042637488965985,
+      "grad_norm": 0.30600835982671126,
+      "learning_rate": 8.028639324380962e-06,
+      "loss": 0.4869,
+      "step": 6724
+    },
+    {
+      "epoch": 1.1044279761049411,
+      "grad_norm": 0.3288737516377353,
+      "learning_rate": 8.028348269819273e-06,
+      "loss": 0.4858,
+      "step": 6725
+    },
+    {
+      "epoch": 1.104592203313284,
+      "grad_norm": 0.3064727566685573,
+      "learning_rate": 8.02805717693625e-06,
+      "loss": 0.4765,
+      "step": 6726
+    },
+    {
+      "epoch": 1.1047564305216266,
+      "grad_norm": 0.3315491714374862,
+      "learning_rate": 8.027766045735054e-06,
+      "loss": 0.4951,
+      "step": 6727
+    },
+    {
+      "epoch": 1.1049206577299695,
+      "grad_norm": 0.32381105894241796,
+      "learning_rate": 8.02747487621885e-06,
+      "loss": 0.4908,
+      "step": 6728
+    },
+    {
+      "epoch": 1.1050848849383121,
+      "grad_norm": 0.29213412468763467,
+      "learning_rate": 8.027183668390795e-06,
+      "loss": 0.4844,
+      "step": 6729
+    },
+    {
+      "epoch": 1.105249112146655,
+      "grad_norm": 0.3109379295275468,
+      "learning_rate": 8.026892422254058e-06,
+      "loss": 0.4955,
+      "step": 6730
+    },
+    {
+      "epoch": 1.1054133393549976,
+      "grad_norm": 0.30931573698151227,
+      "learning_rate": 8.0266011378118e-06,
+      "loss": 0.4868,
+      "step": 6731
+    },
+    {
+      "epoch": 1.1055775665633405,
+      "grad_norm": 0.33109195922803386,
+      "learning_rate": 8.02630981506718e-06,
+      "loss": 0.4952,
+      "step": 6732
+    },
+    {
+      "epoch": 1.1057417937716831,
+      "grad_norm": 0.3782335185626781,
+      "learning_rate": 8.026018454023368e-06,
+      "loss": 0.4915,
+      "step": 6733
+    },
+    {
+      "epoch": 1.1059060209800258,
+      "grad_norm": 0.2978339135303565,
+      "learning_rate": 8.025727054683528e-06,
+      "loss": 0.5001,
+      "step": 6734
+    },
+    {
+      "epoch": 1.1060702481883686,
+      "grad_norm": 0.32795336859626906,
+      "learning_rate": 8.02543561705082e-06,
+      "loss": 0.488,
+      "step": 6735
+    },
+    {
+      "epoch": 1.1062344753967113,
+      "grad_norm": 0.4264225502151214,
+      "learning_rate": 8.025144141128416e-06,
+      "loss": 0.4796,
+      "step": 6736
+    },
+    {
+      "epoch": 1.1063987026050541,
+      "grad_norm": 0.35193668625045177,
+      "learning_rate": 8.024852626919476e-06,
+      "loss": 0.4659,
+      "step": 6737
+    },
+    {
+      "epoch": 1.1065629298133968,
+      "grad_norm": 0.5959409766547124,
+      "learning_rate": 8.02456107442717e-06,
+      "loss": 0.5001,
+      "step": 6738
+    },
+    {
+      "epoch": 1.1067271570217396,
+      "grad_norm": 0.28543948369974015,
+      "learning_rate": 8.024269483654663e-06,
+      "loss": 0.4819,
+      "step": 6739
+    },
+    {
+      "epoch": 1.1068913842300823,
+      "grad_norm": 0.35121396805542204,
+      "learning_rate": 8.02397785460512e-06,
+      "loss": 0.5072,
+      "step": 6740
+    },
+    {
+      "epoch": 1.1070556114384251,
+      "grad_norm": 0.34711974817765606,
+      "learning_rate": 8.023686187281715e-06,
+      "loss": 0.497,
+      "step": 6741
+    },
+    {
+      "epoch": 1.1072198386467678,
+      "grad_norm": 0.2914907052694788,
+      "learning_rate": 8.023394481687607e-06,
+      "loss": 0.4943,
+      "step": 6742
+    },
+    {
+      "epoch": 1.1073840658551106,
+      "grad_norm": 0.24210791352085734,
+      "learning_rate": 8.023102737825968e-06,
+      "loss": 0.4686,
+      "step": 6743
+    },
+    {
+      "epoch": 1.1075482930634533,
+      "grad_norm": 0.30686925835224993,
+      "learning_rate": 8.022810955699969e-06,
+      "loss": 0.4998,
+      "step": 6744
+    },
+    {
+      "epoch": 1.1077125202717961,
+      "grad_norm": 0.364074147936255,
+      "learning_rate": 8.022519135312778e-06,
+      "loss": 0.4706,
+      "step": 6745
+    },
+    {
+      "epoch": 1.1078767474801388,
+      "grad_norm": 0.3451141324499041,
+      "learning_rate": 8.02222727666756e-06,
+      "loss": 0.4977,
+      "step": 6746
+    },
+    {
+      "epoch": 1.1080409746884816,
+      "grad_norm": 0.4224844755920305,
+      "learning_rate": 8.02193537976749e-06,
+      "loss": 0.4832,
+      "step": 6747
+    },
+    {
+      "epoch": 1.1082052018968243,
+      "grad_norm": 0.3336599123558074,
+      "learning_rate": 8.021643444615738e-06,
+      "loss": 0.4962,
+      "step": 6748
+    },
+    {
+      "epoch": 1.1083694291051671,
+      "grad_norm": 0.29675735868901576,
+      "learning_rate": 8.021351471215474e-06,
+      "loss": 0.4864,
+      "step": 6749
+    },
+    {
+      "epoch": 1.1085336563135098,
+      "grad_norm": 0.37380530835569564,
+      "learning_rate": 8.021059459569865e-06,
+      "loss": 0.5131,
+      "step": 6750
+    },
+    {
+      "epoch": 1.1086978835218524,
+      "grad_norm": 0.3060505137161988,
+      "learning_rate": 8.020767409682087e-06,
+      "loss": 0.4857,
+      "step": 6751
+    },
+    {
+      "epoch": 1.1088621107301952,
+      "grad_norm": 0.3374431312187076,
+      "learning_rate": 8.020475321555313e-06,
+      "loss": 0.497,
+      "step": 6752
+    },
+    {
+      "epoch": 1.1090263379385379,
+      "grad_norm": 0.2895837020492894,
+      "learning_rate": 8.020183195192712e-06,
+      "loss": 0.4897,
+      "step": 6753
+    },
+    {
+      "epoch": 1.1091905651468807,
+      "grad_norm": 0.30210173515026933,
+      "learning_rate": 8.019891030597459e-06,
+      "loss": 0.4866,
+      "step": 6754
+    },
+    {
+      "epoch": 1.1093547923552234,
+      "grad_norm": 0.30843038965702,
+      "learning_rate": 8.019598827772726e-06,
+      "loss": 0.4946,
+      "step": 6755
+    },
+    {
+      "epoch": 1.1095190195635662,
+      "grad_norm": 0.35010649192439486,
+      "learning_rate": 8.019306586721687e-06,
+      "loss": 0.4874,
+      "step": 6756
+    },
+    {
+      "epoch": 1.1096832467719089,
+      "grad_norm": 0.33598334774441196,
+      "learning_rate": 8.019014307447516e-06,
+      "loss": 0.4906,
+      "step": 6757
+    },
+    {
+      "epoch": 1.1098474739802517,
+      "grad_norm": 0.3193771491741254,
+      "learning_rate": 8.018721989953385e-06,
+      "loss": 0.4726,
+      "step": 6758
+    },
+    {
+      "epoch": 1.1100117011885944,
+      "grad_norm": 0.30846186162965583,
+      "learning_rate": 8.018429634242472e-06,
+      "loss": 0.4822,
+      "step": 6759
+    },
+    {
+      "epoch": 1.1101759283969372,
+      "grad_norm": 0.2720305115460639,
+      "learning_rate": 8.018137240317953e-06,
+      "loss": 0.4925,
+      "step": 6760
+    },
+    {
+      "epoch": 1.1103401556052799,
+      "grad_norm": 0.3475860658133177,
+      "learning_rate": 8.017844808183002e-06,
+      "loss": 0.4853,
+      "step": 6761
+    },
+    {
+      "epoch": 1.1105043828136227,
+      "grad_norm": 0.2984241485105639,
+      "learning_rate": 8.017552337840797e-06,
+      "loss": 0.503,
+      "step": 6762
+    },
+    {
+      "epoch": 1.1106686100219654,
+      "grad_norm": 0.2736700880333033,
+      "learning_rate": 8.017259829294508e-06,
+      "loss": 0.4958,
+      "step": 6763
+    },
+    {
+      "epoch": 1.1108328372303082,
+      "grad_norm": 0.2947106459552869,
+      "learning_rate": 8.01696728254732e-06,
+      "loss": 0.4935,
+      "step": 6764
+    },
+    {
+      "epoch": 1.1109970644386509,
+      "grad_norm": 0.4868928828665441,
+      "learning_rate": 8.016674697602408e-06,
+      "loss": 0.4979,
+      "step": 6765
+    },
+    {
+      "epoch": 1.1111612916469937,
+      "grad_norm": 0.35004835932239026,
+      "learning_rate": 8.016382074462947e-06,
+      "loss": 0.4935,
+      "step": 6766
+    },
+    {
+      "epoch": 1.1113255188553364,
+      "grad_norm": 0.31130301478663125,
+      "learning_rate": 8.016089413132118e-06,
+      "loss": 0.5001,
+      "step": 6767
+    },
+    {
+      "epoch": 1.111489746063679,
+      "grad_norm": 0.3159981965485992,
+      "learning_rate": 8.0157967136131e-06,
+      "loss": 0.4817,
+      "step": 6768
+    },
+    {
+      "epoch": 1.1116539732720219,
+      "grad_norm": 0.3877824930328974,
+      "learning_rate": 8.015503975909066e-06,
+      "loss": 0.4923,
+      "step": 6769
+    },
+    {
+      "epoch": 1.1118182004803645,
+      "grad_norm": 0.29734466757560113,
+      "learning_rate": 8.015211200023204e-06,
+      "loss": 0.4786,
+      "step": 6770
+    },
+    {
+      "epoch": 1.1119824276887074,
+      "grad_norm": 0.5139772002706117,
+      "learning_rate": 8.014918385958688e-06,
+      "loss": 0.4958,
+      "step": 6771
+    },
+    {
+      "epoch": 1.11214665489705,
+      "grad_norm": 0.3794058659356544,
+      "learning_rate": 8.0146255337187e-06,
+      "loss": 0.4877,
+      "step": 6772
+    },
+    {
+      "epoch": 1.1123108821053929,
+      "grad_norm": 0.5224552555310031,
+      "learning_rate": 8.014332643306422e-06,
+      "loss": 0.4793,
+      "step": 6773
+    },
+    {
+      "epoch": 1.1124751093137355,
+      "grad_norm": 0.33283451191101004,
+      "learning_rate": 8.014039714725034e-06,
+      "loss": 0.4912,
+      "step": 6774
+    },
+    {
+      "epoch": 1.1126393365220784,
+      "grad_norm": 0.29502653757934966,
+      "learning_rate": 8.013746747977716e-06,
+      "loss": 0.4949,
+      "step": 6775
+    },
+    {
+      "epoch": 1.112803563730421,
+      "grad_norm": 0.3077537731044598,
+      "learning_rate": 8.013453743067653e-06,
+      "loss": 0.4841,
+      "step": 6776
+    },
+    {
+      "epoch": 1.1129677909387639,
+      "grad_norm": 0.3273063129273603,
+      "learning_rate": 8.013160699998025e-06,
+      "loss": 0.4913,
+      "step": 6777
+    },
+    {
+      "epoch": 1.1131320181471065,
+      "grad_norm": 0.3087636468252292,
+      "learning_rate": 8.012867618772014e-06,
+      "loss": 0.4965,
+      "step": 6778
+    },
+    {
+      "epoch": 1.1132962453554494,
+      "grad_norm": 0.39746359055685737,
+      "learning_rate": 8.012574499392805e-06,
+      "loss": 0.489,
+      "step": 6779
+    },
+    {
+      "epoch": 1.113460472563792,
+      "grad_norm": 0.37044284746318834,
+      "learning_rate": 8.012281341863583e-06,
+      "loss": 0.5052,
+      "step": 6780
+    },
+    {
+      "epoch": 1.1136246997721349,
+      "grad_norm": 0.3366605150154806,
+      "learning_rate": 8.011988146187527e-06,
+      "loss": 0.4752,
+      "step": 6781
+    },
+    {
+      "epoch": 1.1137889269804775,
+      "grad_norm": 0.25809968388187504,
+      "learning_rate": 8.011694912367826e-06,
+      "loss": 0.4877,
+      "step": 6782
+    },
+    {
+      "epoch": 1.1139531541888203,
+      "grad_norm": 0.326670691401543,
+      "learning_rate": 8.011401640407663e-06,
+      "loss": 0.4749,
+      "step": 6783
+    },
+    {
+      "epoch": 1.114117381397163,
+      "grad_norm": 0.28171208952437016,
+      "learning_rate": 8.011108330310224e-06,
+      "loss": 0.4985,
+      "step": 6784
+    },
+    {
+      "epoch": 1.1142816086055056,
+      "grad_norm": 0.301397371835579,
+      "learning_rate": 8.010814982078693e-06,
+      "loss": 0.4937,
+      "step": 6785
+    },
+    {
+      "epoch": 1.1144458358138485,
+      "grad_norm": 0.31271833861298604,
+      "learning_rate": 8.010521595716257e-06,
+      "loss": 0.5043,
+      "step": 6786
+    },
+    {
+      "epoch": 1.1146100630221911,
+      "grad_norm": 0.3061642359290164,
+      "learning_rate": 8.010228171226104e-06,
+      "loss": 0.4773,
+      "step": 6787
+    },
+    {
+      "epoch": 1.114774290230534,
+      "grad_norm": 0.34226686963469144,
+      "learning_rate": 8.009934708611418e-06,
+      "loss": 0.4814,
+      "step": 6788
+    },
+    {
+      "epoch": 1.1149385174388766,
+      "grad_norm": 0.2925460215378301,
+      "learning_rate": 8.00964120787539e-06,
+      "loss": 0.5067,
+      "step": 6789
+    },
+    {
+      "epoch": 1.1151027446472195,
+      "grad_norm": 0.5386012244349162,
+      "learning_rate": 8.009347669021203e-06,
+      "loss": 0.5001,
+      "step": 6790
+    },
+    {
+      "epoch": 1.1152669718555621,
+      "grad_norm": 0.3130942002159212,
+      "learning_rate": 8.009054092052048e-06,
+      "loss": 0.4763,
+      "step": 6791
+    },
+    {
+      "epoch": 1.115431199063905,
+      "grad_norm": 0.31529862231909345,
+      "learning_rate": 8.008760476971114e-06,
+      "loss": 0.5011,
+      "step": 6792
+    },
+    {
+      "epoch": 1.1155954262722476,
+      "grad_norm": 0.2964668287424008,
+      "learning_rate": 8.00846682378159e-06,
+      "loss": 0.482,
+      "step": 6793
+    },
+    {
+      "epoch": 1.1157596534805905,
+      "grad_norm": 0.47499097388877043,
+      "learning_rate": 8.008173132486663e-06,
+      "loss": 0.4886,
+      "step": 6794
+    },
+    {
+      "epoch": 1.1159238806889331,
+      "grad_norm": 0.2928459229882892,
+      "learning_rate": 8.007879403089523e-06,
+      "loss": 0.4676,
+      "step": 6795
+    },
+    {
+      "epoch": 1.116088107897276,
+      "grad_norm": 0.3433109964688705,
+      "learning_rate": 8.007585635593364e-06,
+      "loss": 0.4909,
+      "step": 6796
+    },
+    {
+      "epoch": 1.1162523351056186,
+      "grad_norm": 0.4728391511187168,
+      "learning_rate": 8.007291830001372e-06,
+      "loss": 0.4819,
+      "step": 6797
+    },
+    {
+      "epoch": 1.1164165623139615,
+      "grad_norm": 0.4378303059005657,
+      "learning_rate": 8.006997986316741e-06,
+      "loss": 0.5186,
+      "step": 6798
+    },
+    {
+      "epoch": 1.116580789522304,
+      "grad_norm": 0.31289617638300654,
+      "learning_rate": 8.006704104542661e-06,
+      "loss": 0.4973,
+      "step": 6799
+    },
+    {
+      "epoch": 1.116745016730647,
+      "grad_norm": 0.34202747263702443,
+      "learning_rate": 8.006410184682325e-06,
+      "loss": 0.47,
+      "step": 6800
+    },
+    {
+      "epoch": 1.1169092439389896,
+      "grad_norm": 0.2870020047280274,
+      "learning_rate": 8.006116226738924e-06,
+      "loss": 0.4959,
+      "step": 6801
+    },
+    {
+      "epoch": 1.1170734711473322,
+      "grad_norm": 0.35538818945126666,
+      "learning_rate": 8.00582223071565e-06,
+      "loss": 0.4819,
+      "step": 6802
+    },
+    {
+      "epoch": 1.117237698355675,
+      "grad_norm": 0.39746931662385276,
+      "learning_rate": 8.005528196615698e-06,
+      "loss": 0.4882,
+      "step": 6803
+    },
+    {
+      "epoch": 1.1174019255640177,
+      "grad_norm": 0.32267732981456915,
+      "learning_rate": 8.005234124442263e-06,
+      "loss": 0.4929,
+      "step": 6804
+    },
+    {
+      "epoch": 1.1175661527723606,
+      "grad_norm": 0.4033719198932124,
+      "learning_rate": 8.004940014198535e-06,
+      "loss": 0.5093,
+      "step": 6805
+    },
+    {
+      "epoch": 1.1177303799807032,
+      "grad_norm": 0.29773637858266944,
+      "learning_rate": 8.00464586588771e-06,
+      "loss": 0.4692,
+      "step": 6806
+    },
+    {
+      "epoch": 1.117894607189046,
+      "grad_norm": 0.34154499889118795,
+      "learning_rate": 8.004351679512983e-06,
+      "loss": 0.4915,
+      "step": 6807
+    },
+    {
+      "epoch": 1.1180588343973887,
+      "grad_norm": 0.3385191940718238,
+      "learning_rate": 8.004057455077549e-06,
+      "loss": 0.4972,
+      "step": 6808
+    },
+    {
+      "epoch": 1.1182230616057316,
+      "grad_norm": 0.31447094702417544,
+      "learning_rate": 8.003763192584602e-06,
+      "loss": 0.4847,
+      "step": 6809
+    },
+    {
+      "epoch": 1.1183872888140742,
+      "grad_norm": 0.3199578937694982,
+      "learning_rate": 8.003468892037342e-06,
+      "loss": 0.4921,
+      "step": 6810
+    },
+    {
+      "epoch": 1.118551516022417,
+      "grad_norm": 0.28372764438025105,
+      "learning_rate": 8.003174553438961e-06,
+      "loss": 0.4905,
+      "step": 6811
+    },
+    {
+      "epoch": 1.1187157432307597,
+      "grad_norm": 0.2801623722482946,
+      "learning_rate": 8.002880176792659e-06,
+      "loss": 0.4684,
+      "step": 6812
+    },
+    {
+      "epoch": 1.1188799704391026,
+      "grad_norm": 0.30831350090955995,
+      "learning_rate": 8.002585762101632e-06,
+      "loss": 0.4729,
+      "step": 6813
+    },
+    {
+      "epoch": 1.1190441976474452,
+      "grad_norm": 0.3297361277092859,
+      "learning_rate": 8.002291309369075e-06,
+      "loss": 0.4947,
+      "step": 6814
+    },
+    {
+      "epoch": 1.119208424855788,
+      "grad_norm": 0.28407434308522084,
+      "learning_rate": 8.001996818598192e-06,
+      "loss": 0.4891,
+      "step": 6815
+    },
+    {
+      "epoch": 1.1193726520641307,
+      "grad_norm": 0.3316328143968119,
+      "learning_rate": 8.001702289792178e-06,
+      "loss": 0.4746,
+      "step": 6816
+    },
+    {
+      "epoch": 1.1195368792724736,
+      "grad_norm": 0.30843174241476523,
+      "learning_rate": 8.001407722954228e-06,
+      "loss": 0.4867,
+      "step": 6817
+    },
+    {
+      "epoch": 1.1197011064808162,
+      "grad_norm": 0.30407152262582604,
+      "learning_rate": 8.00111311808755e-06,
+      "loss": 0.4799,
+      "step": 6818
+    },
+    {
+      "epoch": 1.1198653336891589,
+      "grad_norm": 0.30124797465704084,
+      "learning_rate": 8.000818475195335e-06,
+      "loss": 0.5165,
+      "step": 6819
+    },
+    {
+      "epoch": 1.1200295608975017,
+      "grad_norm": 0.27888439867394527,
+      "learning_rate": 8.00052379428079e-06,
+      "loss": 0.4949,
+      "step": 6820
+    },
+    {
+      "epoch": 1.1201937881058444,
+      "grad_norm": 0.2996848725896366,
+      "learning_rate": 8.00022907534711e-06,
+      "loss": 0.4691,
+      "step": 6821
+    },
+    {
+      "epoch": 1.1203580153141872,
+      "grad_norm": 0.49092113883894145,
+      "learning_rate": 7.999934318397499e-06,
+      "loss": 0.4864,
+      "step": 6822
+    },
+    {
+      "epoch": 1.1205222425225299,
+      "grad_norm": 0.32199904567550197,
+      "learning_rate": 7.99963952343516e-06,
+      "loss": 0.5065,
+      "step": 6823
+    },
+    {
+      "epoch": 1.1206864697308727,
+      "grad_norm": 0.3454881500764911,
+      "learning_rate": 7.99934469046329e-06,
+      "loss": 0.5026,
+      "step": 6824
+    },
+    {
+      "epoch": 1.1208506969392154,
+      "grad_norm": 0.34670731138934535,
+      "learning_rate": 7.999049819485094e-06,
+      "loss": 0.4836,
+      "step": 6825
+    },
+    {
+      "epoch": 1.1210149241475582,
+      "grad_norm": 0.40482327295583864,
+      "learning_rate": 7.998754910503777e-06,
+      "loss": 0.4726,
+      "step": 6826
+    },
+    {
+      "epoch": 1.1211791513559008,
+      "grad_norm": 0.2845574511224747,
+      "learning_rate": 7.998459963522537e-06,
+      "loss": 0.4893,
+      "step": 6827
+    },
+    {
+      "epoch": 1.1213433785642437,
+      "grad_norm": 0.2887523915174508,
+      "learning_rate": 7.998164978544581e-06,
+      "loss": 0.4788,
+      "step": 6828
+    },
+    {
+      "epoch": 1.1215076057725863,
+      "grad_norm": 0.28850627113746496,
+      "learning_rate": 7.99786995557311e-06,
+      "loss": 0.5,
+      "step": 6829
+    },
+    {
+      "epoch": 1.1216718329809292,
+      "grad_norm": 0.42010910356929393,
+      "learning_rate": 7.997574894611332e-06,
+      "loss": 0.4881,
+      "step": 6830
+    },
+    {
+      "epoch": 1.1218360601892718,
+      "grad_norm": 0.3032247075642661,
+      "learning_rate": 7.997279795662447e-06,
+      "loss": 0.4677,
+      "step": 6831
+    },
+    {
+      "epoch": 1.1220002873976145,
+      "grad_norm": 0.3286475855194463,
+      "learning_rate": 7.996984658729664e-06,
+      "loss": 0.5162,
+      "step": 6832
+    },
+    {
+      "epoch": 1.1221645146059573,
+      "grad_norm": 0.28999341860845684,
+      "learning_rate": 7.996689483816187e-06,
+      "loss": 0.4928,
+      "step": 6833
+    },
+    {
+      "epoch": 1.1223287418143002,
+      "grad_norm": 0.42759062227189226,
+      "learning_rate": 7.996394270925222e-06,
+      "loss": 0.486,
+      "step": 6834
+    },
+    {
+      "epoch": 1.1224929690226428,
+      "grad_norm": 0.3450545136947932,
+      "learning_rate": 7.996099020059975e-06,
+      "loss": 0.4977,
+      "step": 6835
+    },
+    {
+      "epoch": 1.1226571962309855,
+      "grad_norm": 0.3517990654176804,
+      "learning_rate": 7.995803731223652e-06,
+      "loss": 0.4919,
+      "step": 6836
+    },
+    {
+      "epoch": 1.1228214234393283,
+      "grad_norm": 0.30347220564861077,
+      "learning_rate": 7.995508404419462e-06,
+      "loss": 0.4698,
+      "step": 6837
+    },
+    {
+      "epoch": 1.122985650647671,
+      "grad_norm": 0.4887298383039076,
+      "learning_rate": 7.995213039650613e-06,
+      "loss": 0.469,
+      "step": 6838
+    },
+    {
+      "epoch": 1.1231498778560138,
+      "grad_norm": 0.3297509541396508,
+      "learning_rate": 7.994917636920311e-06,
+      "loss": 0.4922,
+      "step": 6839
+    },
+    {
+      "epoch": 1.1233141050643565,
+      "grad_norm": 0.36093345558286805,
+      "learning_rate": 7.994622196231764e-06,
+      "loss": 0.5015,
+      "step": 6840
+    },
+    {
+      "epoch": 1.1234783322726993,
+      "grad_norm": 0.2868360244813704,
+      "learning_rate": 7.994326717588181e-06,
+      "loss": 0.4784,
+      "step": 6841
+    },
+    {
+      "epoch": 1.123642559481042,
+      "grad_norm": 0.3860834145580953,
+      "learning_rate": 7.994031200992771e-06,
+      "loss": 0.4697,
+      "step": 6842
+    },
+    {
+      "epoch": 1.1238067866893848,
+      "grad_norm": 0.30524245277727996,
+      "learning_rate": 7.993735646448747e-06,
+      "loss": 0.4782,
+      "step": 6843
+    },
+    {
+      "epoch": 1.1239710138977275,
+      "grad_norm": 0.33538156594853824,
+      "learning_rate": 7.993440053959317e-06,
+      "loss": 0.4734,
+      "step": 6844
+    },
+    {
+      "epoch": 1.1241352411060703,
+      "grad_norm": 0.29676905678033594,
+      "learning_rate": 7.993144423527688e-06,
+      "loss": 0.4693,
+      "step": 6845
+    },
+    {
+      "epoch": 1.124299468314413,
+      "grad_norm": 0.36949417708823395,
+      "learning_rate": 7.992848755157078e-06,
+      "loss": 0.4808,
+      "step": 6846
+    },
+    {
+      "epoch": 1.1244636955227558,
+      "grad_norm": 0.28107760405872567,
+      "learning_rate": 7.99255304885069e-06,
+      "loss": 0.4771,
+      "step": 6847
+    },
+    {
+      "epoch": 1.1246279227310985,
+      "grad_norm": 0.3159370048129486,
+      "learning_rate": 7.992257304611742e-06,
+      "loss": 0.5189,
+      "step": 6848
+    },
+    {
+      "epoch": 1.124792149939441,
+      "grad_norm": 0.3667710699910642,
+      "learning_rate": 7.991961522443443e-06,
+      "loss": 0.4874,
+      "step": 6849
+    },
+    {
+      "epoch": 1.124956377147784,
+      "grad_norm": 0.4336645803076832,
+      "learning_rate": 7.991665702349006e-06,
+      "loss": 0.4809,
+      "step": 6850
+    },
+    {
+      "epoch": 1.1251206043561268,
+      "grad_norm": 0.3175651323732905,
+      "learning_rate": 7.991369844331644e-06,
+      "loss": 0.4693,
+      "step": 6851
+    },
+    {
+      "epoch": 1.1252848315644695,
+      "grad_norm": 0.31051693646801093,
+      "learning_rate": 7.991073948394571e-06,
+      "loss": 0.4779,
+      "step": 6852
+    },
+    {
+      "epoch": 1.125449058772812,
+      "grad_norm": 0.3380347830518644,
+      "learning_rate": 7.990778014541e-06,
+      "loss": 0.5012,
+      "step": 6853
+    },
+    {
+      "epoch": 1.125613285981155,
+      "grad_norm": 0.3200429128632671,
+      "learning_rate": 7.990482042774146e-06,
+      "loss": 0.4813,
+      "step": 6854
+    },
+    {
+      "epoch": 1.1257775131894976,
+      "grad_norm": 0.6531875458342193,
+      "learning_rate": 7.990186033097221e-06,
+      "loss": 0.4828,
+      "step": 6855
+    },
+    {
+      "epoch": 1.1259417403978405,
+      "grad_norm": 0.29262763987154605,
+      "learning_rate": 7.989889985513443e-06,
+      "loss": 0.4919,
+      "step": 6856
+    },
+    {
+      "epoch": 1.126105967606183,
+      "grad_norm": 0.34676785422500556,
+      "learning_rate": 7.989593900026025e-06,
+      "loss": 0.4872,
+      "step": 6857
+    },
+    {
+      "epoch": 1.126270194814526,
+      "grad_norm": 0.3232893839101977,
+      "learning_rate": 7.989297776638185e-06,
+      "loss": 0.4767,
+      "step": 6858
+    },
+    {
+      "epoch": 1.1264344220228686,
+      "grad_norm": 0.4125187177592716,
+      "learning_rate": 7.98900161535314e-06,
+      "loss": 0.4851,
+      "step": 6859
+    },
+    {
+      "epoch": 1.1265986492312114,
+      "grad_norm": 0.4946867385562897,
+      "learning_rate": 7.988705416174103e-06,
+      "loss": 0.5002,
+      "step": 6860
+    },
+    {
+      "epoch": 1.126762876439554,
+      "grad_norm": 0.3315793843980202,
+      "learning_rate": 7.988409179104291e-06,
+      "loss": 0.4835,
+      "step": 6861
+    },
+    {
+      "epoch": 1.126927103647897,
+      "grad_norm": 0.306718722386014,
+      "learning_rate": 7.988112904146926e-06,
+      "loss": 0.4587,
+      "step": 6862
+    },
+    {
+      "epoch": 1.1270913308562396,
+      "grad_norm": 0.31903874679389793,
+      "learning_rate": 7.987816591305222e-06,
+      "loss": 0.4702,
+      "step": 6863
+    },
+    {
+      "epoch": 1.1272555580645824,
+      "grad_norm": 0.2940358413813076,
+      "learning_rate": 7.987520240582398e-06,
+      "loss": 0.4732,
+      "step": 6864
+    },
+    {
+      "epoch": 1.127419785272925,
+      "grad_norm": 0.2965111002119381,
+      "learning_rate": 7.987223851981673e-06,
+      "loss": 0.4842,
+      "step": 6865
+    },
+    {
+      "epoch": 1.1275840124812677,
+      "grad_norm": 0.29814281040642143,
+      "learning_rate": 7.986927425506266e-06,
+      "loss": 0.5065,
+      "step": 6866
+    },
+    {
+      "epoch": 1.1277482396896106,
+      "grad_norm": 0.44997584875021623,
+      "learning_rate": 7.986630961159396e-06,
+      "loss": 0.4937,
+      "step": 6867
+    },
+    {
+      "epoch": 1.1279124668979534,
+      "grad_norm": 0.3079144456977793,
+      "learning_rate": 7.986334458944284e-06,
+      "loss": 0.4688,
+      "step": 6868
+    },
+    {
+      "epoch": 1.128076694106296,
+      "grad_norm": 0.28911938864641884,
+      "learning_rate": 7.986037918864149e-06,
+      "loss": 0.4743,
+      "step": 6869
+    },
+    {
+      "epoch": 1.1282409213146387,
+      "grad_norm": 0.3061321697174208,
+      "learning_rate": 7.985741340922214e-06,
+      "loss": 0.4876,
+      "step": 6870
+    },
+    {
+      "epoch": 1.1284051485229816,
+      "grad_norm": 0.31625410917552904,
+      "learning_rate": 7.985444725121698e-06,
+      "loss": 0.5166,
+      "step": 6871
+    },
+    {
+      "epoch": 1.1285693757313242,
+      "grad_norm": 0.36309716087794164,
+      "learning_rate": 7.985148071465822e-06,
+      "loss": 0.4754,
+      "step": 6872
+    },
+    {
+      "epoch": 1.128733602939667,
+      "grad_norm": 0.6592469202298972,
+      "learning_rate": 7.984851379957809e-06,
+      "loss": 0.4874,
+      "step": 6873
+    },
+    {
+      "epoch": 1.1288978301480097,
+      "grad_norm": 0.31436127933999225,
+      "learning_rate": 7.984554650600883e-06,
+      "loss": 0.4789,
+      "step": 6874
+    },
+    {
+      "epoch": 1.1290620573563526,
+      "grad_norm": 0.2779929040169858,
+      "learning_rate": 7.984257883398264e-06,
+      "loss": 0.4827,
+      "step": 6875
+    },
+    {
+      "epoch": 1.1292262845646952,
+      "grad_norm": 0.2747578546096707,
+      "learning_rate": 7.983961078353175e-06,
+      "loss": 0.5012,
+      "step": 6876
+    },
+    {
+      "epoch": 1.129390511773038,
+      "grad_norm": 0.31846168785030443,
+      "learning_rate": 7.983664235468845e-06,
+      "loss": 0.5102,
+      "step": 6877
+    },
+    {
+      "epoch": 1.1295547389813807,
+      "grad_norm": 0.771861999176361,
+      "learning_rate": 7.98336735474849e-06,
+      "loss": 0.4946,
+      "step": 6878
+    },
+    {
+      "epoch": 1.1297189661897236,
+      "grad_norm": 0.502259440692531,
+      "learning_rate": 7.98307043619534e-06,
+      "loss": 0.4755,
+      "step": 6879
+    },
+    {
+      "epoch": 1.1298831933980662,
+      "grad_norm": 0.35067185395810174,
+      "learning_rate": 7.982773479812616e-06,
+      "loss": 0.4899,
+      "step": 6880
+    },
+    {
+      "epoch": 1.130047420606409,
+      "grad_norm": 0.27701882831760466,
+      "learning_rate": 7.98247648560355e-06,
+      "loss": 0.5046,
+      "step": 6881
+    },
+    {
+      "epoch": 1.1302116478147517,
+      "grad_norm": 0.3261854870605588,
+      "learning_rate": 7.98217945357136e-06,
+      "loss": 0.4823,
+      "step": 6882
+    },
+    {
+      "epoch": 1.1303758750230943,
+      "grad_norm": 0.2753413581996859,
+      "learning_rate": 7.981882383719276e-06,
+      "loss": 0.4917,
+      "step": 6883
+    },
+    {
+      "epoch": 1.1305401022314372,
+      "grad_norm": 0.2713442153753001,
+      "learning_rate": 7.981585276050522e-06,
+      "loss": 0.4697,
+      "step": 6884
+    },
+    {
+      "epoch": 1.13070432943978,
+      "grad_norm": 0.3051670893385003,
+      "learning_rate": 7.981288130568328e-06,
+      "loss": 0.4753,
+      "step": 6885
+    },
+    {
+      "epoch": 1.1308685566481227,
+      "grad_norm": 0.3512406823319829,
+      "learning_rate": 7.980990947275918e-06,
+      "loss": 0.4971,
+      "step": 6886
+    },
+    {
+      "epoch": 1.1310327838564653,
+      "grad_norm": 0.48853751705585835,
+      "learning_rate": 7.980693726176525e-06,
+      "loss": 0.491,
+      "step": 6887
+    },
+    {
+      "epoch": 1.1311970110648082,
+      "grad_norm": 0.42627311188328976,
+      "learning_rate": 7.98039646727337e-06,
+      "loss": 0.4857,
+      "step": 6888
+    },
+    {
+      "epoch": 1.1313612382731508,
+      "grad_norm": 0.2639648626523676,
+      "learning_rate": 7.980099170569687e-06,
+      "loss": 0.4999,
+      "step": 6889
+    },
+    {
+      "epoch": 1.1315254654814937,
+      "grad_norm": 0.339761116269173,
+      "learning_rate": 7.979801836068703e-06,
+      "loss": 0.5077,
+      "step": 6890
+    },
+    {
+      "epoch": 1.1316896926898363,
+      "grad_norm": 0.3125885329528476,
+      "learning_rate": 7.979504463773647e-06,
+      "loss": 0.5013,
+      "step": 6891
+    },
+    {
+      "epoch": 1.1318539198981792,
+      "grad_norm": 0.35306176008613577,
+      "learning_rate": 7.979207053687749e-06,
+      "loss": 0.4881,
+      "step": 6892
+    },
+    {
+      "epoch": 1.1320181471065218,
+      "grad_norm": 0.29872445785240126,
+      "learning_rate": 7.97890960581424e-06,
+      "loss": 0.475,
+      "step": 6893
+    },
+    {
+      "epoch": 1.1321823743148647,
+      "grad_norm": 0.44540460265682047,
+      "learning_rate": 7.97861212015635e-06,
+      "loss": 0.4885,
+      "step": 6894
+    },
+    {
+      "epoch": 1.1323466015232073,
+      "grad_norm": 0.292544060291412,
+      "learning_rate": 7.978314596717308e-06,
+      "loss": 0.4914,
+      "step": 6895
+    },
+    {
+      "epoch": 1.1325108287315502,
+      "grad_norm": 0.3707873447225147,
+      "learning_rate": 7.97801703550035e-06,
+      "loss": 0.4705,
+      "step": 6896
+    },
+    {
+      "epoch": 1.1326750559398928,
+      "grad_norm": 0.31368514148862653,
+      "learning_rate": 7.977719436508702e-06,
+      "loss": 0.4815,
+      "step": 6897
+    },
+    {
+      "epoch": 1.1328392831482357,
+      "grad_norm": 0.3268045977778402,
+      "learning_rate": 7.977421799745602e-06,
+      "loss": 0.4857,
+      "step": 6898
+    },
+    {
+      "epoch": 1.1330035103565783,
+      "grad_norm": 0.328387273053832,
+      "learning_rate": 7.977124125214278e-06,
+      "loss": 0.4928,
+      "step": 6899
+    },
+    {
+      "epoch": 1.133167737564921,
+      "grad_norm": 0.8777662985504897,
+      "learning_rate": 7.976826412917966e-06,
+      "loss": 0.4823,
+      "step": 6900
+    },
+    {
+      "epoch": 1.1333319647732638,
+      "grad_norm": 0.30699456552514476,
+      "learning_rate": 7.9765286628599e-06,
+      "loss": 0.4806,
+      "step": 6901
+    },
+    {
+      "epoch": 1.1334961919816067,
+      "grad_norm": 0.3127291424152957,
+      "learning_rate": 7.976230875043309e-06,
+      "loss": 0.4816,
+      "step": 6902
+    },
+    {
+      "epoch": 1.1336604191899493,
+      "grad_norm": 0.3543717904664054,
+      "learning_rate": 7.975933049471433e-06,
+      "loss": 0.4662,
+      "step": 6903
+    },
+    {
+      "epoch": 1.133824646398292,
+      "grad_norm": 0.3399489359198452,
+      "learning_rate": 7.975635186147504e-06,
+      "loss": 0.5052,
+      "step": 6904
+    },
+    {
+      "epoch": 1.1339888736066348,
+      "grad_norm": 0.30131910741574397,
+      "learning_rate": 7.975337285074755e-06,
+      "loss": 0.5135,
+      "step": 6905
+    },
+    {
+      "epoch": 1.1341531008149774,
+      "grad_norm": 0.31272070176723527,
+      "learning_rate": 7.975039346256427e-06,
+      "loss": 0.5042,
+      "step": 6906
+    },
+    {
+      "epoch": 1.1343173280233203,
+      "grad_norm": 0.29675626380205483,
+      "learning_rate": 7.974741369695752e-06,
+      "loss": 0.4861,
+      "step": 6907
+    },
+    {
+      "epoch": 1.134481555231663,
+      "grad_norm": 0.37560474384127424,
+      "learning_rate": 7.974443355395965e-06,
+      "loss": 0.4789,
+      "step": 6908
+    },
+    {
+      "epoch": 1.1346457824400058,
+      "grad_norm": 0.28532039885356286,
+      "learning_rate": 7.974145303360305e-06,
+      "loss": 0.5013,
+      "step": 6909
+    },
+    {
+      "epoch": 1.1348100096483484,
+      "grad_norm": 0.9029060408738501,
+      "learning_rate": 7.97384721359201e-06,
+      "loss": 0.4825,
+      "step": 6910
+    },
+    {
+      "epoch": 1.1349742368566913,
+      "grad_norm": 0.3563699828853312,
+      "learning_rate": 7.973549086094317e-06,
+      "loss": 0.5041,
+      "step": 6911
+    },
+    {
+      "epoch": 1.135138464065034,
+      "grad_norm": 0.28934448513751987,
+      "learning_rate": 7.973250920870463e-06,
+      "loss": 0.4934,
+      "step": 6912
+    },
+    {
+      "epoch": 1.1353026912733768,
+      "grad_norm": 0.30062125993335753,
+      "learning_rate": 7.972952717923686e-06,
+      "loss": 0.464,
+      "step": 6913
+    },
+    {
+      "epoch": 1.1354669184817194,
+      "grad_norm": 0.319334843105809,
+      "learning_rate": 7.972654477257226e-06,
+      "loss": 0.4851,
+      "step": 6914
+    },
+    {
+      "epoch": 1.1356311456900623,
+      "grad_norm": 0.29289754121365025,
+      "learning_rate": 7.972356198874322e-06,
+      "loss": 0.4783,
+      "step": 6915
+    },
+    {
+      "epoch": 1.135795372898405,
+      "grad_norm": 0.5508218005907385,
+      "learning_rate": 7.972057882778214e-06,
+      "loss": 0.4843,
+      "step": 6916
+    },
+    {
+      "epoch": 1.1359596001067476,
+      "grad_norm": 0.3015173952359254,
+      "learning_rate": 7.97175952897214e-06,
+      "loss": 0.4723,
+      "step": 6917
+    },
+    {
+      "epoch": 1.1361238273150904,
+      "grad_norm": 0.29478196668659074,
+      "learning_rate": 7.971461137459344e-06,
+      "loss": 0.4849,
+      "step": 6918
+    },
+    {
+      "epoch": 1.1362880545234333,
+      "grad_norm": 0.29208606846186064,
+      "learning_rate": 7.971162708243062e-06,
+      "loss": 0.5051,
+      "step": 6919
+    },
+    {
+      "epoch": 1.136452281731776,
+      "grad_norm": 0.3201941594667582,
+      "learning_rate": 7.97086424132654e-06,
+      "loss": 0.4686,
+      "step": 6920
+    },
+    {
+      "epoch": 1.1366165089401186,
+      "grad_norm": 0.320605390702946,
+      "learning_rate": 7.970565736713015e-06,
+      "loss": 0.4832,
+      "step": 6921
+    },
+    {
+      "epoch": 1.1367807361484614,
+      "grad_norm": 0.4047041761384959,
+      "learning_rate": 7.970267194405732e-06,
+      "loss": 0.4722,
+      "step": 6922
+    },
+    {
+      "epoch": 1.136944963356804,
+      "grad_norm": 0.30560656069904657,
+      "learning_rate": 7.969968614407934e-06,
+      "loss": 0.5096,
+      "step": 6923
+    },
+    {
+      "epoch": 1.137109190565147,
+      "grad_norm": 0.2658716031775788,
+      "learning_rate": 7.969669996722862e-06,
+      "loss": 0.4676,
+      "step": 6924
+    },
+    {
+      "epoch": 1.1372734177734896,
+      "grad_norm": 0.3873747529118186,
+      "learning_rate": 7.96937134135376e-06,
+      "loss": 0.4907,
+      "step": 6925
+    },
+    {
+      "epoch": 1.1374376449818324,
+      "grad_norm": 0.31960115771207515,
+      "learning_rate": 7.969072648303874e-06,
+      "loss": 0.4802,
+      "step": 6926
+    },
+    {
+      "epoch": 1.137601872190175,
+      "grad_norm": 0.3590451481218717,
+      "learning_rate": 7.968773917576445e-06,
+      "loss": 0.4909,
+      "step": 6927
+    },
+    {
+      "epoch": 1.137766099398518,
+      "grad_norm": 0.35887748907508493,
+      "learning_rate": 7.968475149174718e-06,
+      "loss": 0.4828,
+      "step": 6928
+    },
+    {
+      "epoch": 1.1379303266068606,
+      "grad_norm": 0.343101086073062,
+      "learning_rate": 7.96817634310194e-06,
+      "loss": 0.4925,
+      "step": 6929
+    },
+    {
+      "epoch": 1.1380945538152034,
+      "grad_norm": 0.30393988234287916,
+      "learning_rate": 7.967877499361351e-06,
+      "loss": 0.4939,
+      "step": 6930
+    },
+    {
+      "epoch": 1.138258781023546,
+      "grad_norm": 0.5099165088420305,
+      "learning_rate": 7.967578617956203e-06,
+      "loss": 0.471,
+      "step": 6931
+    },
+    {
+      "epoch": 1.138423008231889,
+      "grad_norm": 0.40570362767769275,
+      "learning_rate": 7.96727969888974e-06,
+      "loss": 0.4982,
+      "step": 6932
+    },
+    {
+      "epoch": 1.1385872354402315,
+      "grad_norm": 0.2764636379439328,
+      "learning_rate": 7.966980742165207e-06,
+      "loss": 0.4871,
+      "step": 6933
+    },
+    {
+      "epoch": 1.1387514626485742,
+      "grad_norm": 0.31228463120141703,
+      "learning_rate": 7.966681747785852e-06,
+      "loss": 0.4833,
+      "step": 6934
+    },
+    {
+      "epoch": 1.138915689856917,
+      "grad_norm": 0.28009158703806936,
+      "learning_rate": 7.966382715754922e-06,
+      "loss": 0.5005,
+      "step": 6935
+    },
+    {
+      "epoch": 1.13907991706526,
+      "grad_norm": 0.30035288065758636,
+      "learning_rate": 7.966083646075666e-06,
+      "loss": 0.4989,
+      "step": 6936
+    },
+    {
+      "epoch": 1.1392441442736025,
+      "grad_norm": 0.32496233991867585,
+      "learning_rate": 7.96578453875133e-06,
+      "loss": 0.4964,
+      "step": 6937
+    },
+    {
+      "epoch": 1.1394083714819452,
+      "grad_norm": 0.27863619283479796,
+      "learning_rate": 7.965485393785167e-06,
+      "loss": 0.4853,
+      "step": 6938
+    },
+    {
+      "epoch": 1.139572598690288,
+      "grad_norm": 0.3023354281991656,
+      "learning_rate": 7.965186211180421e-06,
+      "loss": 0.477,
+      "step": 6939
+    },
+    {
+      "epoch": 1.1397368258986307,
+      "grad_norm": 0.3838169698056991,
+      "learning_rate": 7.964886990940344e-06,
+      "loss": 0.4988,
+      "step": 6940
+    },
+    {
+      "epoch": 1.1399010531069735,
+      "grad_norm": 0.3102097812889304,
+      "learning_rate": 7.964587733068186e-06,
+      "loss": 0.4733,
+      "step": 6941
+    },
+    {
+      "epoch": 1.1400652803153162,
+      "grad_norm": 0.38695000979047367,
+      "learning_rate": 7.964288437567195e-06,
+      "loss": 0.4877,
+      "step": 6942
+    },
+    {
+      "epoch": 1.140229507523659,
+      "grad_norm": 0.30817199886482777,
+      "learning_rate": 7.963989104440625e-06,
+      "loss": 0.4835,
+      "step": 6943
+    },
+    {
+      "epoch": 1.1403937347320017,
+      "grad_norm": 0.30619596819987116,
+      "learning_rate": 7.963689733691724e-06,
+      "loss": 0.4842,
+      "step": 6944
+    },
+    {
+      "epoch": 1.1405579619403445,
+      "grad_norm": 0.2729947507683516,
+      "learning_rate": 7.963390325323744e-06,
+      "loss": 0.4712,
+      "step": 6945
+    },
+    {
+      "epoch": 1.1407221891486872,
+      "grad_norm": 0.2844335128707053,
+      "learning_rate": 7.963090879339939e-06,
+      "loss": 0.5104,
+      "step": 6946
+    },
+    {
+      "epoch": 1.14088641635703,
+      "grad_norm": 0.35143865483515885,
+      "learning_rate": 7.962791395743559e-06,
+      "loss": 0.5016,
+      "step": 6947
+    },
+    {
+      "epoch": 1.1410506435653727,
+      "grad_norm": 0.29935758721236755,
+      "learning_rate": 7.962491874537856e-06,
+      "loss": 0.4927,
+      "step": 6948
+    },
+    {
+      "epoch": 1.1412148707737155,
+      "grad_norm": 0.38742203083071736,
+      "learning_rate": 7.962192315726086e-06,
+      "loss": 0.4872,
+      "step": 6949
+    },
+    {
+      "epoch": 1.1413790979820582,
+      "grad_norm": 0.26125003171023947,
+      "learning_rate": 7.961892719311504e-06,
+      "loss": 0.4861,
+      "step": 6950
+    },
+    {
+      "epoch": 1.1415433251904008,
+      "grad_norm": 0.2788446059911499,
+      "learning_rate": 7.961593085297357e-06,
+      "loss": 0.4806,
+      "step": 6951
+    },
+    {
+      "epoch": 1.1417075523987437,
+      "grad_norm": 0.37282944348130787,
+      "learning_rate": 7.961293413686906e-06,
+      "loss": 0.514,
+      "step": 6952
+    },
+    {
+      "epoch": 1.1418717796070865,
+      "grad_norm": 0.3248997890565748,
+      "learning_rate": 7.960993704483402e-06,
+      "loss": 0.4991,
+      "step": 6953
+    },
+    {
+      "epoch": 1.1420360068154292,
+      "grad_norm": 0.36815448389317,
+      "learning_rate": 7.960693957690101e-06,
+      "loss": 0.4934,
+      "step": 6954
+    },
+    {
+      "epoch": 1.1422002340237718,
+      "grad_norm": 0.43626717191432984,
+      "learning_rate": 7.960394173310259e-06,
+      "loss": 0.4954,
+      "step": 6955
+    },
+    {
+      "epoch": 1.1423644612321147,
+      "grad_norm": 0.34564209422226533,
+      "learning_rate": 7.96009435134713e-06,
+      "loss": 0.5072,
+      "step": 6956
+    },
+    {
+      "epoch": 1.1425286884404573,
+      "grad_norm": 0.3229667721598629,
+      "learning_rate": 7.959794491803975e-06,
+      "loss": 0.484,
+      "step": 6957
+    },
+    {
+      "epoch": 1.1426929156488002,
+      "grad_norm": 0.5355285268543933,
+      "learning_rate": 7.959494594684047e-06,
+      "loss": 0.5168,
+      "step": 6958
+    },
+    {
+      "epoch": 1.1428571428571428,
+      "grad_norm": 0.3110418435011149,
+      "learning_rate": 7.959194659990602e-06,
+      "loss": 0.4927,
+      "step": 6959
+    },
+    {
+      "epoch": 1.1430213700654857,
+      "grad_norm": 0.28513444305183067,
+      "learning_rate": 7.958894687726902e-06,
+      "loss": 0.4817,
+      "step": 6960
+    },
+    {
+      "epoch": 1.1431855972738283,
+      "grad_norm": 0.27728566842644153,
+      "learning_rate": 7.958594677896201e-06,
+      "loss": 0.4884,
+      "step": 6961
+    },
+    {
+      "epoch": 1.1433498244821712,
+      "grad_norm": 0.3150109157076501,
+      "learning_rate": 7.958294630501761e-06,
+      "loss": 0.5025,
+      "step": 6962
+    },
+    {
+      "epoch": 1.1435140516905138,
+      "grad_norm": 0.37821826587535073,
+      "learning_rate": 7.957994545546838e-06,
+      "loss": 0.4677,
+      "step": 6963
+    },
+    {
+      "epoch": 1.1436782788988566,
+      "grad_norm": 0.272360968366249,
+      "learning_rate": 7.95769442303469e-06,
+      "loss": 0.4825,
+      "step": 6964
+    },
+    {
+      "epoch": 1.1438425061071993,
+      "grad_norm": 0.29025116102764614,
+      "learning_rate": 7.957394262968581e-06,
+      "loss": 0.4693,
+      "step": 6965
+    },
+    {
+      "epoch": 1.1440067333155421,
+      "grad_norm": 0.3988120980456517,
+      "learning_rate": 7.957094065351767e-06,
+      "loss": 0.4878,
+      "step": 6966
+    },
+    {
+      "epoch": 1.1441709605238848,
+      "grad_norm": 0.32215514818082297,
+      "learning_rate": 7.956793830187512e-06,
+      "loss": 0.5049,
+      "step": 6967
+    },
+    {
+      "epoch": 1.1443351877322274,
+      "grad_norm": 0.3141652271654795,
+      "learning_rate": 7.956493557479074e-06,
+      "loss": 0.4791,
+      "step": 6968
+    },
+    {
+      "epoch": 1.1444994149405703,
+      "grad_norm": 0.29312116929724086,
+      "learning_rate": 7.956193247229714e-06,
+      "loss": 0.4669,
+      "step": 6969
+    },
+    {
+      "epoch": 1.1446636421489131,
+      "grad_norm": 0.3198331201930757,
+      "learning_rate": 7.955892899442697e-06,
+      "loss": 0.4892,
+      "step": 6970
+    },
+    {
+      "epoch": 1.1448278693572558,
+      "grad_norm": 0.8979760840640667,
+      "learning_rate": 7.955592514121281e-06,
+      "loss": 0.4767,
+      "step": 6971
+    },
+    {
+      "epoch": 1.1449920965655984,
+      "grad_norm": 0.2990306712439826,
+      "learning_rate": 7.955292091268733e-06,
+      "loss": 0.5102,
+      "step": 6972
+    },
+    {
+      "epoch": 1.1451563237739413,
+      "grad_norm": 0.3053581095580193,
+      "learning_rate": 7.954991630888311e-06,
+      "loss": 0.4723,
+      "step": 6973
+    },
+    {
+      "epoch": 1.145320550982284,
+      "grad_norm": 0.3150822655293769,
+      "learning_rate": 7.954691132983282e-06,
+      "loss": 0.4704,
+      "step": 6974
+    },
+    {
+      "epoch": 1.1454847781906268,
+      "grad_norm": 0.2882659024696099,
+      "learning_rate": 7.954390597556908e-06,
+      "loss": 0.4725,
+      "step": 6975
+    },
+    {
+      "epoch": 1.1456490053989694,
+      "grad_norm": 0.2797251128466923,
+      "learning_rate": 7.954090024612453e-06,
+      "loss": 0.4743,
+      "step": 6976
+    },
+    {
+      "epoch": 1.1458132326073123,
+      "grad_norm": 0.3642369480396053,
+      "learning_rate": 7.953789414153183e-06,
+      "loss": 0.5029,
+      "step": 6977
+    },
+    {
+      "epoch": 1.145977459815655,
+      "grad_norm": 0.29543854541561576,
+      "learning_rate": 7.953488766182361e-06,
+      "loss": 0.4801,
+      "step": 6978
+    },
+    {
+      "epoch": 1.1461416870239978,
+      "grad_norm": 0.3695440138181109,
+      "learning_rate": 7.953188080703254e-06,
+      "loss": 0.4773,
+      "step": 6979
+    },
+    {
+      "epoch": 1.1463059142323404,
+      "grad_norm": 0.35804494112721486,
+      "learning_rate": 7.952887357719125e-06,
+      "loss": 0.4892,
+      "step": 6980
+    },
+    {
+      "epoch": 1.1464701414406833,
+      "grad_norm": 0.32574825202652413,
+      "learning_rate": 7.952586597233245e-06,
+      "loss": 0.5225,
+      "step": 6981
+    },
+    {
+      "epoch": 1.146634368649026,
+      "grad_norm": 0.2549369661307935,
+      "learning_rate": 7.952285799248875e-06,
+      "loss": 0.4872,
+      "step": 6982
+    },
+    {
+      "epoch": 1.1467985958573688,
+      "grad_norm": 0.2836474018881919,
+      "learning_rate": 7.951984963769287e-06,
+      "loss": 0.4807,
+      "step": 6983
+    },
+    {
+      "epoch": 1.1469628230657114,
+      "grad_norm": 0.3010588829949219,
+      "learning_rate": 7.951684090797744e-06,
+      "loss": 0.4766,
+      "step": 6984
+    },
+    {
+      "epoch": 1.147127050274054,
+      "grad_norm": 0.2774477777434259,
+      "learning_rate": 7.951383180337516e-06,
+      "loss": 0.4914,
+      "step": 6985
+    },
+    {
+      "epoch": 1.147291277482397,
+      "grad_norm": 0.43537709750406217,
+      "learning_rate": 7.951082232391873e-06,
+      "loss": 0.4822,
+      "step": 6986
+    },
+    {
+      "epoch": 1.1474555046907398,
+      "grad_norm": 0.29585741210346284,
+      "learning_rate": 7.950781246964079e-06,
+      "loss": 0.47,
+      "step": 6987
+    },
+    {
+      "epoch": 1.1476197318990824,
+      "grad_norm": 0.3355195438180755,
+      "learning_rate": 7.950480224057406e-06,
+      "loss": 0.4849,
+      "step": 6988
+    },
+    {
+      "epoch": 1.147783959107425,
+      "grad_norm": 0.32705446324079807,
+      "learning_rate": 7.950179163675124e-06,
+      "loss": 0.4889,
+      "step": 6989
+    },
+    {
+      "epoch": 1.147948186315768,
+      "grad_norm": 0.2670710136847381,
+      "learning_rate": 7.9498780658205e-06,
+      "loss": 0.4794,
+      "step": 6990
+    },
+    {
+      "epoch": 1.1481124135241105,
+      "grad_norm": 0.30118225566962875,
+      "learning_rate": 7.949576930496808e-06,
+      "loss": 0.4883,
+      "step": 6991
+    },
+    {
+      "epoch": 1.1482766407324534,
+      "grad_norm": 0.36070352710146236,
+      "learning_rate": 7.949275757707316e-06,
+      "loss": 0.477,
+      "step": 6992
+    },
+    {
+      "epoch": 1.148440867940796,
+      "grad_norm": 0.3061406159550247,
+      "learning_rate": 7.948974547455297e-06,
+      "loss": 0.4931,
+      "step": 6993
+    },
+    {
+      "epoch": 1.1486050951491389,
+      "grad_norm": 0.3545081273612038,
+      "learning_rate": 7.94867329974402e-06,
+      "loss": 0.4847,
+      "step": 6994
+    },
+    {
+      "epoch": 1.1487693223574815,
+      "grad_norm": 0.3311782995765307,
+      "learning_rate": 7.948372014576756e-06,
+      "loss": 0.4959,
+      "step": 6995
+    },
+    {
+      "epoch": 1.1489335495658244,
+      "grad_norm": 0.3077919240460935,
+      "learning_rate": 7.94807069195678e-06,
+      "loss": 0.4757,
+      "step": 6996
+    },
+    {
+      "epoch": 1.149097776774167,
+      "grad_norm": 0.2870260528487584,
+      "learning_rate": 7.947769331887365e-06,
+      "loss": 0.4615,
+      "step": 6997
+    },
+    {
+      "epoch": 1.1492620039825099,
+      "grad_norm": 0.29569179787697997,
+      "learning_rate": 7.947467934371783e-06,
+      "loss": 0.4936,
+      "step": 6998
+    },
+    {
+      "epoch": 1.1494262311908525,
+      "grad_norm": 0.35220813871941586,
+      "learning_rate": 7.947166499413307e-06,
+      "loss": 0.4593,
+      "step": 6999
+    },
+    {
+      "epoch": 1.1495904583991954,
+      "grad_norm": 0.32263584184992183,
+      "learning_rate": 7.946865027015212e-06,
+      "loss": 0.4904,
+      "step": 7000
+    },
+    {
+      "epoch": 1.149754685607538,
+      "grad_norm": 0.5166828266471525,
+      "learning_rate": 7.94656351718077e-06,
+      "loss": 0.4904,
+      "step": 7001
+    },
+    {
+      "epoch": 1.1499189128158807,
+      "grad_norm": 0.2712023974328414,
+      "learning_rate": 7.946261969913257e-06,
+      "loss": 0.4818,
+      "step": 7002
+    },
+    {
+      "epoch": 1.1500831400242235,
+      "grad_norm": 0.26915717210756196,
+      "learning_rate": 7.94596038521595e-06,
+      "loss": 0.4819,
+      "step": 7003
+    },
+    {
+      "epoch": 1.1502473672325664,
+      "grad_norm": 0.30042354024540735,
+      "learning_rate": 7.945658763092124e-06,
+      "loss": 0.4732,
+      "step": 7004
+    },
+    {
+      "epoch": 1.150411594440909,
+      "grad_norm": 0.32655935806435304,
+      "learning_rate": 7.94535710354505e-06,
+      "loss": 0.4978,
+      "step": 7005
+    },
+    {
+      "epoch": 1.1505758216492517,
+      "grad_norm": 0.3199955481342031,
+      "learning_rate": 7.94505540657801e-06,
+      "loss": 0.4744,
+      "step": 7006
+    },
+    {
+      "epoch": 1.1507400488575945,
+      "grad_norm": 0.35223648349278697,
+      "learning_rate": 7.94475367219428e-06,
+      "loss": 0.4725,
+      "step": 7007
+    },
+    {
+      "epoch": 1.1509042760659371,
+      "grad_norm": 0.34834147750497035,
+      "learning_rate": 7.944451900397133e-06,
+      "loss": 0.4773,
+      "step": 7008
+    },
+    {
+      "epoch": 1.15106850327428,
+      "grad_norm": 0.3568296672581254,
+      "learning_rate": 7.944150091189854e-06,
+      "loss": 0.495,
+      "step": 7009
+    },
+    {
+      "epoch": 1.1512327304826226,
+      "grad_norm": 0.3000946461945242,
+      "learning_rate": 7.943848244575712e-06,
+      "loss": 0.4742,
+      "step": 7010
+    },
+    {
+      "epoch": 1.1513969576909655,
+      "grad_norm": 0.2721934419714511,
+      "learning_rate": 7.943546360557992e-06,
+      "loss": 0.496,
+      "step": 7011
+    },
+    {
+      "epoch": 1.1515611848993081,
+      "grad_norm": 0.31830130435810267,
+      "learning_rate": 7.94324443913997e-06,
+      "loss": 0.508,
+      "step": 7012
+    },
+    {
+      "epoch": 1.151725412107651,
+      "grad_norm": 0.3229175169530436,
+      "learning_rate": 7.942942480324925e-06,
+      "loss": 0.4816,
+      "step": 7013
+    },
+    {
+      "epoch": 1.1518896393159936,
+      "grad_norm": 0.4111817485404657,
+      "learning_rate": 7.942640484116138e-06,
+      "loss": 0.4684,
+      "step": 7014
+    },
+    {
+      "epoch": 1.1520538665243365,
+      "grad_norm": 0.2656930494883279,
+      "learning_rate": 7.942338450516888e-06,
+      "loss": 0.458,
+      "step": 7015
+    },
+    {
+      "epoch": 1.1522180937326791,
+      "grad_norm": 0.31851059556159345,
+      "learning_rate": 7.942036379530456e-06,
+      "loss": 0.4754,
+      "step": 7016
+    },
+    {
+      "epoch": 1.152382320941022,
+      "grad_norm": 0.27602728318038483,
+      "learning_rate": 7.941734271160122e-06,
+      "loss": 0.4817,
+      "step": 7017
+    },
+    {
+      "epoch": 1.1525465481493646,
+      "grad_norm": 0.29468735688655845,
+      "learning_rate": 7.941432125409168e-06,
+      "loss": 0.4742,
+      "step": 7018
+    },
+    {
+      "epoch": 1.1527107753577073,
+      "grad_norm": 0.5489676416342217,
+      "learning_rate": 7.941129942280876e-06,
+      "loss": 0.4855,
+      "step": 7019
+    },
+    {
+      "epoch": 1.1528750025660501,
+      "grad_norm": 0.3221574182624256,
+      "learning_rate": 7.940827721778525e-06,
+      "loss": 0.4695,
+      "step": 7020
+    },
+    {
+      "epoch": 1.153039229774393,
+      "grad_norm": 0.314407501924177,
+      "learning_rate": 7.940525463905401e-06,
+      "loss": 0.4909,
+      "step": 7021
+    },
+    {
+      "epoch": 1.1532034569827356,
+      "grad_norm": 0.3598478943499774,
+      "learning_rate": 7.940223168664785e-06,
+      "loss": 0.5016,
+      "step": 7022
+    },
+    {
+      "epoch": 1.1533676841910783,
+      "grad_norm": 0.39422712873691973,
+      "learning_rate": 7.93992083605996e-06,
+      "loss": 0.4817,
+      "step": 7023
+    },
+    {
+      "epoch": 1.1535319113994211,
+      "grad_norm": 0.36165863495626405,
+      "learning_rate": 7.939618466094213e-06,
+      "loss": 0.4803,
+      "step": 7024
+    },
+    {
+      "epoch": 1.1536961386077638,
+      "grad_norm": 0.31260829963726494,
+      "learning_rate": 7.939316058770823e-06,
+      "loss": 0.486,
+      "step": 7025
+    },
+    {
+      "epoch": 1.1538603658161066,
+      "grad_norm": 0.34725257054626907,
+      "learning_rate": 7.939013614093078e-06,
+      "loss": 0.4775,
+      "step": 7026
+    },
+    {
+      "epoch": 1.1540245930244493,
+      "grad_norm": 0.2669254990323276,
+      "learning_rate": 7.93871113206426e-06,
+      "loss": 0.4889,
+      "step": 7027
+    },
+    {
+      "epoch": 1.1541888202327921,
+      "grad_norm": 0.4093792540774185,
+      "learning_rate": 7.938408612687657e-06,
+      "loss": 0.491,
+      "step": 7028
+    },
+    {
+      "epoch": 1.1543530474411348,
+      "grad_norm": 0.3290675100373528,
+      "learning_rate": 7.938106055966554e-06,
+      "loss": 0.4954,
+      "step": 7029
+    },
+    {
+      "epoch": 1.1545172746494776,
+      "grad_norm": 0.3225318096562668,
+      "learning_rate": 7.937803461904236e-06,
+      "loss": 0.4846,
+      "step": 7030
+    },
+    {
+      "epoch": 1.1546815018578203,
+      "grad_norm": 0.3516853816464303,
+      "learning_rate": 7.93750083050399e-06,
+      "loss": 0.4874,
+      "step": 7031
+    },
+    {
+      "epoch": 1.1548457290661631,
+      "grad_norm": 0.32631152573393535,
+      "learning_rate": 7.937198161769102e-06,
+      "loss": 0.4735,
+      "step": 7032
+    },
+    {
+      "epoch": 1.1550099562745058,
+      "grad_norm": 0.31907307198476026,
+      "learning_rate": 7.936895455702861e-06,
+      "loss": 0.49,
+      "step": 7033
+    },
+    {
+      "epoch": 1.1551741834828486,
+      "grad_norm": 0.27136595054747875,
+      "learning_rate": 7.936592712308557e-06,
+      "loss": 0.4598,
+      "step": 7034
+    },
+    {
+      "epoch": 1.1553384106911913,
+      "grad_norm": 0.3763844729573229,
+      "learning_rate": 7.93628993158947e-06,
+      "loss": 0.4689,
+      "step": 7035
+    },
+    {
+      "epoch": 1.155502637899534,
+      "grad_norm": 0.3195562295488287,
+      "learning_rate": 7.935987113548896e-06,
+      "loss": 0.4878,
+      "step": 7036
+    },
+    {
+      "epoch": 1.1556668651078768,
+      "grad_norm": 0.2761377095445116,
+      "learning_rate": 7.93568425819012e-06,
+      "loss": 0.4706,
+      "step": 7037
+    },
+    {
+      "epoch": 1.1558310923162196,
+      "grad_norm": 0.5413992090888955,
+      "learning_rate": 7.935381365516435e-06,
+      "loss": 0.5024,
+      "step": 7038
+    },
+    {
+      "epoch": 1.1559953195245622,
+      "grad_norm": 0.5469978575561415,
+      "learning_rate": 7.935078435531127e-06,
+      "loss": 0.4571,
+      "step": 7039
+    },
+    {
+      "epoch": 1.1561595467329049,
+      "grad_norm": 0.2823534067099165,
+      "learning_rate": 7.934775468237486e-06,
+      "loss": 0.4698,
+      "step": 7040
+    },
+    {
+      "epoch": 1.1563237739412477,
+      "grad_norm": 0.44633643276642965,
+      "learning_rate": 7.934472463638807e-06,
+      "loss": 0.4956,
+      "step": 7041
+    },
+    {
+      "epoch": 1.1564880011495904,
+      "grad_norm": 0.376102625780165,
+      "learning_rate": 7.934169421738377e-06,
+      "loss": 0.4824,
+      "step": 7042
+    },
+    {
+      "epoch": 1.1566522283579332,
+      "grad_norm": 0.27588776388549247,
+      "learning_rate": 7.933866342539488e-06,
+      "loss": 0.5206,
+      "step": 7043
+    },
+    {
+      "epoch": 1.1568164555662759,
+      "grad_norm": 0.2878482112541905,
+      "learning_rate": 7.933563226045431e-06,
+      "loss": 0.4852,
+      "step": 7044
+    },
+    {
+      "epoch": 1.1569806827746187,
+      "grad_norm": 0.2697563816295159,
+      "learning_rate": 7.933260072259501e-06,
+      "loss": 0.499,
+      "step": 7045
+    },
+    {
+      "epoch": 1.1571449099829614,
+      "grad_norm": 0.28236168172776016,
+      "learning_rate": 7.932956881184988e-06,
+      "loss": 0.5005,
+      "step": 7046
+    },
+    {
+      "epoch": 1.1573091371913042,
+      "grad_norm": 0.3338941095186326,
+      "learning_rate": 7.932653652825185e-06,
+      "loss": 0.4796,
+      "step": 7047
+    },
+    {
+      "epoch": 1.1574733643996469,
+      "grad_norm": 0.3625356119489219,
+      "learning_rate": 7.932350387183387e-06,
+      "loss": 0.5081,
+      "step": 7048
+    },
+    {
+      "epoch": 1.1576375916079897,
+      "grad_norm": 0.27136370495062984,
+      "learning_rate": 7.932047084262887e-06,
+      "loss": 0.4879,
+      "step": 7049
+    },
+    {
+      "epoch": 1.1578018188163324,
+      "grad_norm": 0.3366519673880185,
+      "learning_rate": 7.931743744066978e-06,
+      "loss": 0.491,
+      "step": 7050
+    },
+    {
+      "epoch": 1.1579660460246752,
+      "grad_norm": 0.30178495916921333,
+      "learning_rate": 7.931440366598956e-06,
+      "loss": 0.4831,
+      "step": 7051
+    },
+    {
+      "epoch": 1.1581302732330179,
+      "grad_norm": 0.32500677131083183,
+      "learning_rate": 7.931136951862117e-06,
+      "loss": 0.4977,
+      "step": 7052
+    },
+    {
+      "epoch": 1.1582945004413605,
+      "grad_norm": 0.2751069112378041,
+      "learning_rate": 7.930833499859752e-06,
+      "loss": 0.4689,
+      "step": 7053
+    },
+    {
+      "epoch": 1.1584587276497034,
+      "grad_norm": 0.3078889240750525,
+      "learning_rate": 7.930530010595161e-06,
+      "loss": 0.4922,
+      "step": 7054
+    },
+    {
+      "epoch": 1.1586229548580462,
+      "grad_norm": 0.2956108438153107,
+      "learning_rate": 7.93022648407164e-06,
+      "loss": 0.4897,
+      "step": 7055
+    },
+    {
+      "epoch": 1.1587871820663889,
+      "grad_norm": 0.2927169524393216,
+      "learning_rate": 7.929922920292483e-06,
+      "loss": 0.4975,
+      "step": 7056
+    },
+    {
+      "epoch": 1.1589514092747315,
+      "grad_norm": 0.33526725775792277,
+      "learning_rate": 7.929619319260988e-06,
+      "loss": 0.4927,
+      "step": 7057
+    },
+    {
+      "epoch": 1.1591156364830744,
+      "grad_norm": 0.4759390949788766,
+      "learning_rate": 7.929315680980456e-06,
+      "loss": 0.4908,
+      "step": 7058
+    },
+    {
+      "epoch": 1.159279863691417,
+      "grad_norm": 0.32076151526244373,
+      "learning_rate": 7.929012005454178e-06,
+      "loss": 0.4795,
+      "step": 7059
+    },
+    {
+      "epoch": 1.1594440908997599,
+      "grad_norm": 0.31178290972430867,
+      "learning_rate": 7.928708292685458e-06,
+      "loss": 0.5054,
+      "step": 7060
+    },
+    {
+      "epoch": 1.1596083181081025,
+      "grad_norm": 0.3028328443170019,
+      "learning_rate": 7.928404542677592e-06,
+      "loss": 0.4939,
+      "step": 7061
+    },
+    {
+      "epoch": 1.1597725453164454,
+      "grad_norm": 0.8078948086337088,
+      "learning_rate": 7.92810075543388e-06,
+      "loss": 0.4848,
+      "step": 7062
+    },
+    {
+      "epoch": 1.159936772524788,
+      "grad_norm": 0.2892938136222857,
+      "learning_rate": 7.927796930957622e-06,
+      "loss": 0.4998,
+      "step": 7063
+    },
+    {
+      "epoch": 1.1601009997331309,
+      "grad_norm": 0.3019318848201967,
+      "learning_rate": 7.927493069252115e-06,
+      "loss": 0.4862,
+      "step": 7064
+    },
+    {
+      "epoch": 1.1602652269414735,
+      "grad_norm": 0.2908551244366857,
+      "learning_rate": 7.927189170320663e-06,
+      "loss": 0.4978,
+      "step": 7065
+    },
+    {
+      "epoch": 1.1604294541498164,
+      "grad_norm": 0.29280416653917135,
+      "learning_rate": 7.926885234166562e-06,
+      "loss": 0.4755,
+      "step": 7066
+    },
+    {
+      "epoch": 1.160593681358159,
+      "grad_norm": 0.269957864981838,
+      "learning_rate": 7.926581260793119e-06,
+      "loss": 0.4701,
+      "step": 7067
+    },
+    {
+      "epoch": 1.1607579085665019,
+      "grad_norm": 0.30097708194571493,
+      "learning_rate": 7.926277250203629e-06,
+      "loss": 0.4644,
+      "step": 7068
+    },
+    {
+      "epoch": 1.1609221357748445,
+      "grad_norm": 0.32757186478730316,
+      "learning_rate": 7.925973202401399e-06,
+      "loss": 0.4749,
+      "step": 7069
+    },
+    {
+      "epoch": 1.1610863629831871,
+      "grad_norm": 0.40123882700762675,
+      "learning_rate": 7.925669117389728e-06,
+      "loss": 0.4788,
+      "step": 7070
+    },
+    {
+      "epoch": 1.16125059019153,
+      "grad_norm": 0.3603074875810292,
+      "learning_rate": 7.92536499517192e-06,
+      "loss": 0.4812,
+      "step": 7071
+    },
+    {
+      "epoch": 1.1614148173998728,
+      "grad_norm": 0.3698279608261931,
+      "learning_rate": 7.92506083575128e-06,
+      "loss": 0.4665,
+      "step": 7072
+    },
+    {
+      "epoch": 1.1615790446082155,
+      "grad_norm": 0.4518716080031763,
+      "learning_rate": 7.924756639131108e-06,
+      "loss": 0.4997,
+      "step": 7073
+    },
+    {
+      "epoch": 1.1617432718165581,
+      "grad_norm": 0.33189232317990797,
+      "learning_rate": 7.92445240531471e-06,
+      "loss": 0.4534,
+      "step": 7074
+    },
+    {
+      "epoch": 1.161907499024901,
+      "grad_norm": 0.3044263355830535,
+      "learning_rate": 7.924148134305389e-06,
+      "loss": 0.5074,
+      "step": 7075
+    },
+    {
+      "epoch": 1.1620717262332436,
+      "grad_norm": 0.34185010495544116,
+      "learning_rate": 7.92384382610645e-06,
+      "loss": 0.4768,
+      "step": 7076
+    },
+    {
+      "epoch": 1.1622359534415865,
+      "grad_norm": 0.3453996765799976,
+      "learning_rate": 7.9235394807212e-06,
+      "loss": 0.4863,
+      "step": 7077
+    },
+    {
+      "epoch": 1.1624001806499291,
+      "grad_norm": 0.3128815977202802,
+      "learning_rate": 7.923235098152943e-06,
+      "loss": 0.5233,
+      "step": 7078
+    },
+    {
+      "epoch": 1.162564407858272,
+      "grad_norm": 0.5247087085994546,
+      "learning_rate": 7.922930678404983e-06,
+      "loss": 0.4742,
+      "step": 7079
+    },
+    {
+      "epoch": 1.1627286350666146,
+      "grad_norm": 0.39150724559453093,
+      "learning_rate": 7.922626221480629e-06,
+      "loss": 0.4781,
+      "step": 7080
+    },
+    {
+      "epoch": 1.1628928622749575,
+      "grad_norm": 0.33804533874168374,
+      "learning_rate": 7.922321727383187e-06,
+      "loss": 0.4862,
+      "step": 7081
+    },
+    {
+      "epoch": 1.1630570894833,
+      "grad_norm": 0.43511358014238194,
+      "learning_rate": 7.922017196115964e-06,
+      "loss": 0.4706,
+      "step": 7082
+    },
+    {
+      "epoch": 1.163221316691643,
+      "grad_norm": 0.32505034890033285,
+      "learning_rate": 7.921712627682266e-06,
+      "loss": 0.483,
+      "step": 7083
+    },
+    {
+      "epoch": 1.1633855438999856,
+      "grad_norm": 0.29173852633659864,
+      "learning_rate": 7.921408022085404e-06,
+      "loss": 0.4763,
+      "step": 7084
+    },
+    {
+      "epoch": 1.1635497711083285,
+      "grad_norm": 0.32413350522873113,
+      "learning_rate": 7.921103379328685e-06,
+      "loss": 0.5107,
+      "step": 7085
+    },
+    {
+      "epoch": 1.163713998316671,
+      "grad_norm": 0.29827489864002776,
+      "learning_rate": 7.920798699415416e-06,
+      "loss": 0.4904,
+      "step": 7086
+    },
+    {
+      "epoch": 1.1638782255250137,
+      "grad_norm": 0.32881461136004475,
+      "learning_rate": 7.92049398234891e-06,
+      "loss": 0.4878,
+      "step": 7087
+    },
+    {
+      "epoch": 1.1640424527333566,
+      "grad_norm": 0.26966455996545646,
+      "learning_rate": 7.92018922813247e-06,
+      "loss": 0.4843,
+      "step": 7088
+    },
+    {
+      "epoch": 1.1642066799416995,
+      "grad_norm": 0.3524190577600752,
+      "learning_rate": 7.919884436769413e-06,
+      "loss": 0.4807,
+      "step": 7089
+    },
+    {
+      "epoch": 1.164370907150042,
+      "grad_norm": 0.2873202615000851,
+      "learning_rate": 7.919579608263045e-06,
+      "loss": 0.4889,
+      "step": 7090
+    },
+    {
+      "epoch": 1.1645351343583847,
+      "grad_norm": 0.32072446570760915,
+      "learning_rate": 7.91927474261668e-06,
+      "loss": 0.4881,
+      "step": 7091
+    },
+    {
+      "epoch": 1.1646993615667276,
+      "grad_norm": 0.35024441319253874,
+      "learning_rate": 7.918969839833625e-06,
+      "loss": 0.4794,
+      "step": 7092
+    },
+    {
+      "epoch": 1.1648635887750702,
+      "grad_norm": 0.3299835820205014,
+      "learning_rate": 7.918664899917194e-06,
+      "loss": 0.4884,
+      "step": 7093
+    },
+    {
+      "epoch": 1.165027815983413,
+      "grad_norm": 0.29717281189054834,
+      "learning_rate": 7.9183599228707e-06,
+      "loss": 0.4886,
+      "step": 7094
+    },
+    {
+      "epoch": 1.1651920431917557,
+      "grad_norm": 0.3284713919170598,
+      "learning_rate": 7.918054908697453e-06,
+      "loss": 0.4818,
+      "step": 7095
+    },
+    {
+      "epoch": 1.1653562704000986,
+      "grad_norm": 0.2925217287720715,
+      "learning_rate": 7.917749857400766e-06,
+      "loss": 0.4869,
+      "step": 7096
+    },
+    {
+      "epoch": 1.1655204976084412,
+      "grad_norm": 0.2765372710402451,
+      "learning_rate": 7.917444768983954e-06,
+      "loss": 0.4661,
+      "step": 7097
+    },
+    {
+      "epoch": 1.165684724816784,
+      "grad_norm": 0.3101792926600381,
+      "learning_rate": 7.917139643450331e-06,
+      "loss": 0.4848,
+      "step": 7098
+    },
+    {
+      "epoch": 1.1658489520251267,
+      "grad_norm": 0.2848413701567469,
+      "learning_rate": 7.916834480803207e-06,
+      "loss": 0.4973,
+      "step": 7099
+    },
+    {
+      "epoch": 1.1660131792334696,
+      "grad_norm": 0.28047689897386036,
+      "learning_rate": 7.9165292810459e-06,
+      "loss": 0.4548,
+      "step": 7100
+    },
+    {
+      "epoch": 1.1661774064418122,
+      "grad_norm": 0.35063738100165215,
+      "learning_rate": 7.916224044181723e-06,
+      "loss": 0.4869,
+      "step": 7101
+    },
+    {
+      "epoch": 1.166341633650155,
+      "grad_norm": 0.2823049704902083,
+      "learning_rate": 7.915918770213992e-06,
+      "loss": 0.4783,
+      "step": 7102
+    },
+    {
+      "epoch": 1.1665058608584977,
+      "grad_norm": 0.29208609576357075,
+      "learning_rate": 7.915613459146022e-06,
+      "loss": 0.4933,
+      "step": 7103
+    },
+    {
+      "epoch": 1.1666700880668404,
+      "grad_norm": 0.2908603084533477,
+      "learning_rate": 7.915308110981129e-06,
+      "loss": 0.4762,
+      "step": 7104
+    },
+    {
+      "epoch": 1.1668343152751832,
+      "grad_norm": 0.3142549539312152,
+      "learning_rate": 7.915002725722632e-06,
+      "loss": 0.467,
+      "step": 7105
+    },
+    {
+      "epoch": 1.166998542483526,
+      "grad_norm": 0.297802378353576,
+      "learning_rate": 7.914697303373843e-06,
+      "loss": 0.4712,
+      "step": 7106
+    },
+    {
+      "epoch": 1.1671627696918687,
+      "grad_norm": 0.30827567036324066,
+      "learning_rate": 7.914391843938082e-06,
+      "loss": 0.4782,
+      "step": 7107
+    },
+    {
+      "epoch": 1.1673269969002114,
+      "grad_norm": 0.27376585280710897,
+      "learning_rate": 7.914086347418667e-06,
+      "loss": 0.4667,
+      "step": 7108
+    },
+    {
+      "epoch": 1.1674912241085542,
+      "grad_norm": 0.2912467864119177,
+      "learning_rate": 7.913780813818914e-06,
+      "loss": 0.4891,
+      "step": 7109
+    },
+    {
+      "epoch": 1.1676554513168969,
+      "grad_norm": 0.3755970122826635,
+      "learning_rate": 7.913475243142145e-06,
+      "loss": 0.4698,
+      "step": 7110
+    },
+    {
+      "epoch": 1.1678196785252397,
+      "grad_norm": 0.3126792391825246,
+      "learning_rate": 7.913169635391675e-06,
+      "loss": 0.4962,
+      "step": 7111
+    },
+    {
+      "epoch": 1.1679839057335824,
+      "grad_norm": 0.33128634368155074,
+      "learning_rate": 7.912863990570826e-06,
+      "loss": 0.4906,
+      "step": 7112
+    },
+    {
+      "epoch": 1.1681481329419252,
+      "grad_norm": 0.276069135824882,
+      "learning_rate": 7.912558308682914e-06,
+      "loss": 0.5039,
+      "step": 7113
+    },
+    {
+      "epoch": 1.1683123601502678,
+      "grad_norm": 0.30298873479148086,
+      "learning_rate": 7.912252589731262e-06,
+      "loss": 0.4697,
+      "step": 7114
+    },
+    {
+      "epoch": 1.1684765873586107,
+      "grad_norm": 0.34034916001374615,
+      "learning_rate": 7.91194683371919e-06,
+      "loss": 0.4949,
+      "step": 7115
+    },
+    {
+      "epoch": 1.1686408145669533,
+      "grad_norm": 0.30127604093427846,
+      "learning_rate": 7.911641040650019e-06,
+      "loss": 0.4656,
+      "step": 7116
+    },
+    {
+      "epoch": 1.1688050417752962,
+      "grad_norm": 0.30448083371606166,
+      "learning_rate": 7.911335210527068e-06,
+      "loss": 0.504,
+      "step": 7117
+    },
+    {
+      "epoch": 1.1689692689836388,
+      "grad_norm": 0.3490091808051166,
+      "learning_rate": 7.911029343353664e-06,
+      "loss": 0.4747,
+      "step": 7118
+    },
+    {
+      "epoch": 1.1691334961919817,
+      "grad_norm": 0.28283313823611345,
+      "learning_rate": 7.910723439133123e-06,
+      "loss": 0.4985,
+      "step": 7119
+    },
+    {
+      "epoch": 1.1692977234003243,
+      "grad_norm": 1.1831427012153144,
+      "learning_rate": 7.910417497868768e-06,
+      "loss": 0.5003,
+      "step": 7120
+    },
+    {
+      "epoch": 1.169461950608667,
+      "grad_norm": 0.37970954472659757,
+      "learning_rate": 7.910111519563926e-06,
+      "loss": 0.4743,
+      "step": 7121
+    },
+    {
+      "epoch": 1.1696261778170098,
+      "grad_norm": 0.3060323198171699,
+      "learning_rate": 7.909805504221917e-06,
+      "loss": 0.4727,
+      "step": 7122
+    },
+    {
+      "epoch": 1.1697904050253527,
+      "grad_norm": 0.3090375579006394,
+      "learning_rate": 7.909499451846065e-06,
+      "loss": 0.4854,
+      "step": 7123
+    },
+    {
+      "epoch": 1.1699546322336953,
+      "grad_norm": 0.35422086742882625,
+      "learning_rate": 7.909193362439696e-06,
+      "loss": 0.481,
+      "step": 7124
+    },
+    {
+      "epoch": 1.170118859442038,
+      "grad_norm": 0.2833349932119131,
+      "learning_rate": 7.908887236006131e-06,
+      "loss": 0.4609,
+      "step": 7125
+    },
+    {
+      "epoch": 1.1702830866503808,
+      "grad_norm": 0.3438027843268845,
+      "learning_rate": 7.908581072548698e-06,
+      "loss": 0.4978,
+      "step": 7126
+    },
+    {
+      "epoch": 1.1704473138587235,
+      "grad_norm": 0.3084865886495486,
+      "learning_rate": 7.908274872070718e-06,
+      "loss": 0.4879,
+      "step": 7127
+    },
+    {
+      "epoch": 1.1706115410670663,
+      "grad_norm": 0.2968696229849295,
+      "learning_rate": 7.907968634575524e-06,
+      "loss": 0.4669,
+      "step": 7128
+    },
+    {
+      "epoch": 1.170775768275409,
+      "grad_norm": 0.30218235608708677,
+      "learning_rate": 7.907662360066435e-06,
+      "loss": 0.4836,
+      "step": 7129
+    },
+    {
+      "epoch": 1.1709399954837518,
+      "grad_norm": 0.29267877493515143,
+      "learning_rate": 7.907356048546781e-06,
+      "loss": 0.4828,
+      "step": 7130
+    },
+    {
+      "epoch": 1.1711042226920945,
+      "grad_norm": 0.30254596069123213,
+      "learning_rate": 7.907049700019888e-06,
+      "loss": 0.4937,
+      "step": 7131
+    },
+    {
+      "epoch": 1.1712684499004373,
+      "grad_norm": 0.3029131007987612,
+      "learning_rate": 7.906743314489084e-06,
+      "loss": 0.4709,
+      "step": 7132
+    },
+    {
+      "epoch": 1.17143267710878,
+      "grad_norm": 0.38195767468163316,
+      "learning_rate": 7.906436891957694e-06,
+      "loss": 0.4835,
+      "step": 7133
+    },
+    {
+      "epoch": 1.1715969043171228,
+      "grad_norm": 0.32624775519999805,
+      "learning_rate": 7.906130432429048e-06,
+      "loss": 0.5137,
+      "step": 7134
+    },
+    {
+      "epoch": 1.1717611315254655,
+      "grad_norm": 0.2754028052145745,
+      "learning_rate": 7.905823935906474e-06,
+      "loss": 0.4959,
+      "step": 7135
+    },
+    {
+      "epoch": 1.1719253587338083,
+      "grad_norm": 0.4444728587958453,
+      "learning_rate": 7.905517402393304e-06,
+      "loss": 0.514,
+      "step": 7136
+    },
+    {
+      "epoch": 1.172089585942151,
+      "grad_norm": 0.2726245516199059,
+      "learning_rate": 7.905210831892863e-06,
+      "loss": 0.4567,
+      "step": 7137
+    },
+    {
+      "epoch": 1.1722538131504936,
+      "grad_norm": 0.28681242761135123,
+      "learning_rate": 7.904904224408481e-06,
+      "loss": 0.4986,
+      "step": 7138
+    },
+    {
+      "epoch": 1.1724180403588365,
+      "grad_norm": 0.2756319660402748,
+      "learning_rate": 7.904597579943488e-06,
+      "loss": 0.4842,
+      "step": 7139
+    },
+    {
+      "epoch": 1.1725822675671793,
+      "grad_norm": 0.33331624186474573,
+      "learning_rate": 7.904290898501218e-06,
+      "loss": 0.4678,
+      "step": 7140
+    },
+    {
+      "epoch": 1.172746494775522,
+      "grad_norm": 0.3325946939274747,
+      "learning_rate": 7.903984180084999e-06,
+      "loss": 0.4677,
+      "step": 7141
+    },
+    {
+      "epoch": 1.1729107219838646,
+      "grad_norm": 0.2699322093677336,
+      "learning_rate": 7.903677424698163e-06,
+      "loss": 0.473,
+      "step": 7142
+    },
+    {
+      "epoch": 1.1730749491922075,
+      "grad_norm": 0.28956151873217795,
+      "learning_rate": 7.903370632344042e-06,
+      "loss": 0.4828,
+      "step": 7143
+    },
+    {
+      "epoch": 1.17323917640055,
+      "grad_norm": 0.5323259173407964,
+      "learning_rate": 7.903063803025965e-06,
+      "loss": 0.4677,
+      "step": 7144
+    },
+    {
+      "epoch": 1.173403403608893,
+      "grad_norm": 0.38857627895908703,
+      "learning_rate": 7.902756936747268e-06,
+      "loss": 0.4758,
+      "step": 7145
+    },
+    {
+      "epoch": 1.1735676308172356,
+      "grad_norm": 0.3227428663071013,
+      "learning_rate": 7.902450033511284e-06,
+      "loss": 0.4831,
+      "step": 7146
+    },
+    {
+      "epoch": 1.1737318580255784,
+      "grad_norm": 0.3086496978455175,
+      "learning_rate": 7.902143093321344e-06,
+      "loss": 0.4857,
+      "step": 7147
+    },
+    {
+      "epoch": 1.173896085233921,
+      "grad_norm": 0.32069342974778725,
+      "learning_rate": 7.901836116180784e-06,
+      "loss": 0.4907,
+      "step": 7148
+    },
+    {
+      "epoch": 1.174060312442264,
+      "grad_norm": 0.4404613405259411,
+      "learning_rate": 7.901529102092935e-06,
+      "loss": 0.4866,
+      "step": 7149
+    },
+    {
+      "epoch": 1.1742245396506066,
+      "grad_norm": 0.28651128563078515,
+      "learning_rate": 7.901222051061133e-06,
+      "loss": 0.4649,
+      "step": 7150
+    },
+    {
+      "epoch": 1.1743887668589494,
+      "grad_norm": 2.9411089773533248,
+      "learning_rate": 7.900914963088717e-06,
+      "loss": 0.4908,
+      "step": 7151
+    },
+    {
+      "epoch": 1.174552994067292,
+      "grad_norm": 0.3004882084769234,
+      "learning_rate": 7.900607838179015e-06,
+      "loss": 0.4819,
+      "step": 7152
+    },
+    {
+      "epoch": 1.174717221275635,
+      "grad_norm": 0.2929426082232998,
+      "learning_rate": 7.900300676335366e-06,
+      "loss": 0.4785,
+      "step": 7153
+    },
+    {
+      "epoch": 1.1748814484839776,
+      "grad_norm": 0.3017982722752946,
+      "learning_rate": 7.899993477561107e-06,
+      "loss": 0.4783,
+      "step": 7154
+    },
+    {
+      "epoch": 1.1750456756923202,
+      "grad_norm": 0.37263447468744393,
+      "learning_rate": 7.899686241859574e-06,
+      "loss": 0.4787,
+      "step": 7155
+    },
+    {
+      "epoch": 1.175209902900663,
+      "grad_norm": 0.2718529502001876,
+      "learning_rate": 7.899378969234103e-06,
+      "loss": 0.4677,
+      "step": 7156
+    },
+    {
+      "epoch": 1.175374130109006,
+      "grad_norm": 0.32113420311098073,
+      "learning_rate": 7.899071659688032e-06,
+      "loss": 0.4871,
+      "step": 7157
+    },
+    {
+      "epoch": 1.1755383573173486,
+      "grad_norm": 0.30453037794151244,
+      "learning_rate": 7.8987643132247e-06,
+      "loss": 0.4866,
+      "step": 7158
+    },
+    {
+      "epoch": 1.1757025845256912,
+      "grad_norm": 0.37467717111887067,
+      "learning_rate": 7.898456929847442e-06,
+      "loss": 0.4978,
+      "step": 7159
+    },
+    {
+      "epoch": 1.175866811734034,
+      "grad_norm": 0.3034542609392963,
+      "learning_rate": 7.898149509559599e-06,
+      "loss": 0.4899,
+      "step": 7160
+    },
+    {
+      "epoch": 1.1760310389423767,
+      "grad_norm": 0.40653052696984354,
+      "learning_rate": 7.897842052364508e-06,
+      "loss": 0.4876,
+      "step": 7161
+    },
+    {
+      "epoch": 1.1761952661507196,
+      "grad_norm": 0.4270247739034037,
+      "learning_rate": 7.89753455826551e-06,
+      "loss": 0.4674,
+      "step": 7162
+    },
+    {
+      "epoch": 1.1763594933590622,
+      "grad_norm": 0.3277953459600014,
+      "learning_rate": 7.897227027265943e-06,
+      "loss": 0.4772,
+      "step": 7163
+    },
+    {
+      "epoch": 1.176523720567405,
+      "grad_norm": 0.4031180802456469,
+      "learning_rate": 7.89691945936915e-06,
+      "loss": 0.5218,
+      "step": 7164
+    },
+    {
+      "epoch": 1.1766879477757477,
+      "grad_norm": 0.3228656700450739,
+      "learning_rate": 7.89661185457847e-06,
+      "loss": 0.4886,
+      "step": 7165
+    },
+    {
+      "epoch": 1.1768521749840906,
+      "grad_norm": 0.30127464136306215,
+      "learning_rate": 7.89630421289724e-06,
+      "loss": 0.4974,
+      "step": 7166
+    },
+    {
+      "epoch": 1.1770164021924332,
+      "grad_norm": 0.3407442389191083,
+      "learning_rate": 7.895996534328806e-06,
+      "loss": 0.4672,
+      "step": 7167
+    },
+    {
+      "epoch": 1.177180629400776,
+      "grad_norm": 0.3611630703926426,
+      "learning_rate": 7.895688818876508e-06,
+      "loss": 0.5189,
+      "step": 7168
+    },
+    {
+      "epoch": 1.1773448566091187,
+      "grad_norm": 0.3194743450347194,
+      "learning_rate": 7.895381066543691e-06,
+      "loss": 0.5102,
+      "step": 7169
+    },
+    {
+      "epoch": 1.1775090838174616,
+      "grad_norm": 0.33776411509241444,
+      "learning_rate": 7.895073277333694e-06,
+      "loss": 0.4825,
+      "step": 7170
+    },
+    {
+      "epoch": 1.1776733110258042,
+      "grad_norm": 0.33471683930689616,
+      "learning_rate": 7.89476545124986e-06,
+      "loss": 0.4675,
+      "step": 7171
+    },
+    {
+      "epoch": 1.1778375382341468,
+      "grad_norm": 0.3248330468687593,
+      "learning_rate": 7.894457588295533e-06,
+      "loss": 0.4876,
+      "step": 7172
+    },
+    {
+      "epoch": 1.1780017654424897,
+      "grad_norm": 0.3094418395243431,
+      "learning_rate": 7.894149688474058e-06,
+      "loss": 0.4779,
+      "step": 7173
+    },
+    {
+      "epoch": 1.1781659926508325,
+      "grad_norm": 0.4121252840483226,
+      "learning_rate": 7.893841751788777e-06,
+      "loss": 0.5082,
+      "step": 7174
+    },
+    {
+      "epoch": 1.1783302198591752,
+      "grad_norm": 0.306690976540399,
+      "learning_rate": 7.893533778243037e-06,
+      "loss": 0.4679,
+      "step": 7175
+    },
+    {
+      "epoch": 1.1784944470675178,
+      "grad_norm": 0.3098485120429855,
+      "learning_rate": 7.89322576784018e-06,
+      "loss": 0.4656,
+      "step": 7176
+    },
+    {
+      "epoch": 1.1786586742758607,
+      "grad_norm": 0.3139158217304466,
+      "learning_rate": 7.892917720583553e-06,
+      "loss": 0.4817,
+      "step": 7177
+    },
+    {
+      "epoch": 1.1788229014842033,
+      "grad_norm": 0.27150948840561384,
+      "learning_rate": 7.892609636476502e-06,
+      "loss": 0.4734,
+      "step": 7178
+    },
+    {
+      "epoch": 1.1789871286925462,
+      "grad_norm": 0.3792047308148616,
+      "learning_rate": 7.892301515522371e-06,
+      "loss": 0.4892,
+      "step": 7179
+    },
+    {
+      "epoch": 1.1791513559008888,
+      "grad_norm": 0.2695161583725051,
+      "learning_rate": 7.89199335772451e-06,
+      "loss": 0.4469,
+      "step": 7180
+    },
+    {
+      "epoch": 1.1793155831092317,
+      "grad_norm": 0.29931825739146534,
+      "learning_rate": 7.891685163086262e-06,
+      "loss": 0.4822,
+      "step": 7181
+    },
+    {
+      "epoch": 1.1794798103175743,
+      "grad_norm": 0.2696831618893344,
+      "learning_rate": 7.891376931610977e-06,
+      "loss": 0.4963,
+      "step": 7182
+    },
+    {
+      "epoch": 1.1796440375259172,
+      "grad_norm": 0.3602896096956024,
+      "learning_rate": 7.891068663302003e-06,
+      "loss": 0.4941,
+      "step": 7183
+    },
+    {
+      "epoch": 1.1798082647342598,
+      "grad_norm": 0.3380357807441332,
+      "learning_rate": 7.890760358162686e-06,
+      "loss": 0.4828,
+      "step": 7184
+    },
+    {
+      "epoch": 1.1799724919426027,
+      "grad_norm": 0.3158752432413669,
+      "learning_rate": 7.890452016196373e-06,
+      "loss": 0.4987,
+      "step": 7185
+    },
+    {
+      "epoch": 1.1801367191509453,
+      "grad_norm": 0.54489675202446,
+      "learning_rate": 7.89014363740642e-06,
+      "loss": 0.4751,
+      "step": 7186
+    },
+    {
+      "epoch": 1.1803009463592882,
+      "grad_norm": 0.2946086684484438,
+      "learning_rate": 7.889835221796168e-06,
+      "loss": 0.4927,
+      "step": 7187
+    },
+    {
+      "epoch": 1.1804651735676308,
+      "grad_norm": 0.29319451335645197,
+      "learning_rate": 7.889526769368971e-06,
+      "loss": 0.4791,
+      "step": 7188
+    },
+    {
+      "epoch": 1.1806294007759734,
+      "grad_norm": 0.3168463859588959,
+      "learning_rate": 7.88921828012818e-06,
+      "loss": 0.4695,
+      "step": 7189
+    },
+    {
+      "epoch": 1.1807936279843163,
+      "grad_norm": 0.3594514406581799,
+      "learning_rate": 7.888909754077142e-06,
+      "loss": 0.4798,
+      "step": 7190
+    },
+    {
+      "epoch": 1.1809578551926592,
+      "grad_norm": 0.2990964083755579,
+      "learning_rate": 7.888601191219211e-06,
+      "loss": 0.4916,
+      "step": 7191
+    },
+    {
+      "epoch": 1.1811220824010018,
+      "grad_norm": 0.5104588170168158,
+      "learning_rate": 7.888292591557738e-06,
+      "loss": 0.4768,
+      "step": 7192
+    },
+    {
+      "epoch": 1.1812863096093444,
+      "grad_norm": 0.31064065418444264,
+      "learning_rate": 7.887983955096072e-06,
+      "loss": 0.4828,
+      "step": 7193
+    },
+    {
+      "epoch": 1.1814505368176873,
+      "grad_norm": 0.31641511809024947,
+      "learning_rate": 7.887675281837568e-06,
+      "loss": 0.4864,
+      "step": 7194
+    },
+    {
+      "epoch": 1.18161476402603,
+      "grad_norm": 0.3876844228637263,
+      "learning_rate": 7.887366571785577e-06,
+      "loss": 0.4788,
+      "step": 7195
+    },
+    {
+      "epoch": 1.1817789912343728,
+      "grad_norm": 0.28781277770735475,
+      "learning_rate": 7.887057824943451e-06,
+      "loss": 0.4813,
+      "step": 7196
+    },
+    {
+      "epoch": 1.1819432184427154,
+      "grad_norm": 0.2780428348147511,
+      "learning_rate": 7.886749041314546e-06,
+      "loss": 0.4975,
+      "step": 7197
+    },
+    {
+      "epoch": 1.1821074456510583,
+      "grad_norm": 0.3092106232254678,
+      "learning_rate": 7.886440220902214e-06,
+      "loss": 0.4926,
+      "step": 7198
+    },
+    {
+      "epoch": 1.182271672859401,
+      "grad_norm": 0.2840760200057259,
+      "learning_rate": 7.88613136370981e-06,
+      "loss": 0.4898,
+      "step": 7199
+    },
+    {
+      "epoch": 1.1824359000677438,
+      "grad_norm": 0.27574696423953743,
+      "learning_rate": 7.885822469740688e-06,
+      "loss": 0.4891,
+      "step": 7200
+    },
+    {
+      "epoch": 1.1826001272760864,
+      "grad_norm": 0.35965149805749896,
+      "learning_rate": 7.885513538998203e-06,
+      "loss": 0.4588,
+      "step": 7201
+    },
+    {
+      "epoch": 1.1827643544844293,
+      "grad_norm": 0.33303040766200065,
+      "learning_rate": 7.885204571485709e-06,
+      "loss": 0.4753,
+      "step": 7202
+    },
+    {
+      "epoch": 1.182928581692772,
+      "grad_norm": 0.2593780964338368,
+      "learning_rate": 7.884895567206563e-06,
+      "loss": 0.4835,
+      "step": 7203
+    },
+    {
+      "epoch": 1.1830928089011148,
+      "grad_norm": 0.35610378246540597,
+      "learning_rate": 7.884586526164121e-06,
+      "loss": 0.4746,
+      "step": 7204
+    },
+    {
+      "epoch": 1.1832570361094574,
+      "grad_norm": 0.28443270166004814,
+      "learning_rate": 7.88427744836174e-06,
+      "loss": 0.4791,
+      "step": 7205
+    },
+    {
+      "epoch": 1.1834212633178,
+      "grad_norm": 0.2831758991670496,
+      "learning_rate": 7.883968333802774e-06,
+      "loss": 0.4684,
+      "step": 7206
+    },
+    {
+      "epoch": 1.183585490526143,
+      "grad_norm": 0.35471014312117216,
+      "learning_rate": 7.883659182490585e-06,
+      "loss": 0.4668,
+      "step": 7207
+    },
+    {
+      "epoch": 1.1837497177344858,
+      "grad_norm": 0.29537181388674116,
+      "learning_rate": 7.883349994428527e-06,
+      "loss": 0.4802,
+      "step": 7208
+    },
+    {
+      "epoch": 1.1839139449428284,
+      "grad_norm": 0.3368787078344288,
+      "learning_rate": 7.88304076961996e-06,
+      "loss": 0.4802,
+      "step": 7209
+    },
+    {
+      "epoch": 1.184078172151171,
+      "grad_norm": 0.36423298240411317,
+      "learning_rate": 7.88273150806824e-06,
+      "loss": 0.4846,
+      "step": 7210
+    },
+    {
+      "epoch": 1.184242399359514,
+      "grad_norm": 0.45191874292943746,
+      "learning_rate": 7.88242220977673e-06,
+      "loss": 0.4971,
+      "step": 7211
+    },
+    {
+      "epoch": 1.1844066265678566,
+      "grad_norm": 0.31730056357201625,
+      "learning_rate": 7.882112874748787e-06,
+      "loss": 0.5007,
+      "step": 7212
+    },
+    {
+      "epoch": 1.1845708537761994,
+      "grad_norm": 0.276740020652526,
+      "learning_rate": 7.881803502987769e-06,
+      "loss": 0.4603,
+      "step": 7213
+    },
+    {
+      "epoch": 1.184735080984542,
+      "grad_norm": 0.3226551874841209,
+      "learning_rate": 7.881494094497038e-06,
+      "loss": 0.4973,
+      "step": 7214
+    },
+    {
+      "epoch": 1.184899308192885,
+      "grad_norm": 0.2698997195963249,
+      "learning_rate": 7.881184649279956e-06,
+      "loss": 0.4571,
+      "step": 7215
+    },
+    {
+      "epoch": 1.1850635354012276,
+      "grad_norm": 0.3116107257091327,
+      "learning_rate": 7.88087516733988e-06,
+      "loss": 0.4939,
+      "step": 7216
+    },
+    {
+      "epoch": 1.1852277626095704,
+      "grad_norm": 0.26601114390957037,
+      "learning_rate": 7.880565648680174e-06,
+      "loss": 0.4781,
+      "step": 7217
+    },
+    {
+      "epoch": 1.185391989817913,
+      "grad_norm": 0.2960317652813218,
+      "learning_rate": 7.880256093304199e-06,
+      "loss": 0.498,
+      "step": 7218
+    },
+    {
+      "epoch": 1.185556217026256,
+      "grad_norm": 0.2897931619508775,
+      "learning_rate": 7.879946501215317e-06,
+      "loss": 0.4928,
+      "step": 7219
+    },
+    {
+      "epoch": 1.1857204442345985,
+      "grad_norm": 0.3444469864217827,
+      "learning_rate": 7.87963687241689e-06,
+      "loss": 0.4992,
+      "step": 7220
+    },
+    {
+      "epoch": 1.1858846714429414,
+      "grad_norm": 0.32963841574885944,
+      "learning_rate": 7.879327206912283e-06,
+      "loss": 0.4832,
+      "step": 7221
+    },
+    {
+      "epoch": 1.186048898651284,
+      "grad_norm": 0.27874786931557877,
+      "learning_rate": 7.879017504704856e-06,
+      "loss": 0.4807,
+      "step": 7222
+    },
+    {
+      "epoch": 1.1862131258596267,
+      "grad_norm": 0.2954496205585582,
+      "learning_rate": 7.878707765797975e-06,
+      "loss": 0.4638,
+      "step": 7223
+    },
+    {
+      "epoch": 1.1863773530679695,
+      "grad_norm": 0.2997197947778376,
+      "learning_rate": 7.878397990195004e-06,
+      "loss": 0.4811,
+      "step": 7224
+    },
+    {
+      "epoch": 1.1865415802763124,
+      "grad_norm": 0.3608949284326675,
+      "learning_rate": 7.878088177899307e-06,
+      "loss": 0.4879,
+      "step": 7225
+    },
+    {
+      "epoch": 1.186705807484655,
+      "grad_norm": 0.31512186735330405,
+      "learning_rate": 7.877778328914248e-06,
+      "loss": 0.4778,
+      "step": 7226
+    },
+    {
+      "epoch": 1.1868700346929977,
+      "grad_norm": 0.5780317877360018,
+      "learning_rate": 7.877468443243195e-06,
+      "loss": 0.4951,
+      "step": 7227
+    },
+    {
+      "epoch": 1.1870342619013405,
+      "grad_norm": 0.30742943688087454,
+      "learning_rate": 7.877158520889509e-06,
+      "loss": 0.4756,
+      "step": 7228
+    },
+    {
+      "epoch": 1.1871984891096832,
+      "grad_norm": 0.30278431149830287,
+      "learning_rate": 7.87684856185656e-06,
+      "loss": 0.4714,
+      "step": 7229
+    },
+    {
+      "epoch": 1.187362716318026,
+      "grad_norm": 0.3118677206225516,
+      "learning_rate": 7.876538566147713e-06,
+      "loss": 0.4795,
+      "step": 7230
+    },
+    {
+      "epoch": 1.1875269435263687,
+      "grad_norm": 0.3327275128495535,
+      "learning_rate": 7.876228533766335e-06,
+      "loss": 0.485,
+      "step": 7231
+    },
+    {
+      "epoch": 1.1876911707347115,
+      "grad_norm": 0.280947840200177,
+      "learning_rate": 7.875918464715795e-06,
+      "loss": 0.4733,
+      "step": 7232
+    },
+    {
+      "epoch": 1.1878553979430542,
+      "grad_norm": 0.27631423496822494,
+      "learning_rate": 7.875608358999456e-06,
+      "loss": 0.4769,
+      "step": 7233
+    },
+    {
+      "epoch": 1.188019625151397,
+      "grad_norm": 0.3424215714956797,
+      "learning_rate": 7.875298216620692e-06,
+      "loss": 0.4745,
+      "step": 7234
+    },
+    {
+      "epoch": 1.1881838523597397,
+      "grad_norm": 0.3392612659281871,
+      "learning_rate": 7.874988037582868e-06,
+      "loss": 0.4821,
+      "step": 7235
+    },
+    {
+      "epoch": 1.1883480795680825,
+      "grad_norm": 0.30768246487911133,
+      "learning_rate": 7.874677821889352e-06,
+      "loss": 0.492,
+      "step": 7236
+    },
+    {
+      "epoch": 1.1885123067764252,
+      "grad_norm": 0.4768362934562472,
+      "learning_rate": 7.874367569543516e-06,
+      "loss": 0.4709,
+      "step": 7237
+    },
+    {
+      "epoch": 1.188676533984768,
+      "grad_norm": 0.39163590605246273,
+      "learning_rate": 7.874057280548727e-06,
+      "loss": 0.4987,
+      "step": 7238
+    },
+    {
+      "epoch": 1.1888407611931107,
+      "grad_norm": 0.48949358688016253,
+      "learning_rate": 7.873746954908358e-06,
+      "loss": 0.473,
+      "step": 7239
+    },
+    {
+      "epoch": 1.1890049884014533,
+      "grad_norm": 0.34669067573595047,
+      "learning_rate": 7.873436592625775e-06,
+      "loss": 0.4871,
+      "step": 7240
+    },
+    {
+      "epoch": 1.1891692156097962,
+      "grad_norm": 0.2881379597778867,
+      "learning_rate": 7.873126193704353e-06,
+      "loss": 0.4835,
+      "step": 7241
+    },
+    {
+      "epoch": 1.189333442818139,
+      "grad_norm": 0.28928702283893437,
+      "learning_rate": 7.872815758147463e-06,
+      "loss": 0.5041,
+      "step": 7242
+    },
+    {
+      "epoch": 1.1894976700264817,
+      "grad_norm": 0.2779046323729139,
+      "learning_rate": 7.872505285958475e-06,
+      "loss": 0.4861,
+      "step": 7243
+    },
+    {
+      "epoch": 1.1896618972348243,
+      "grad_norm": 0.32171186966240967,
+      "learning_rate": 7.872194777140761e-06,
+      "loss": 0.4708,
+      "step": 7244
+    },
+    {
+      "epoch": 1.1898261244431672,
+      "grad_norm": 0.33224800587019176,
+      "learning_rate": 7.871884231697693e-06,
+      "loss": 0.4752,
+      "step": 7245
+    },
+    {
+      "epoch": 1.1899903516515098,
+      "grad_norm": 0.39098922706360334,
+      "learning_rate": 7.871573649632646e-06,
+      "loss": 0.4722,
+      "step": 7246
+    },
+    {
+      "epoch": 1.1901545788598527,
+      "grad_norm": 0.260711638664843,
+      "learning_rate": 7.871263030948992e-06,
+      "loss": 0.4923,
+      "step": 7247
+    },
+    {
+      "epoch": 1.1903188060681953,
+      "grad_norm": 0.3056696630001395,
+      "learning_rate": 7.870952375650105e-06,
+      "loss": 0.4753,
+      "step": 7248
+    },
+    {
+      "epoch": 1.1904830332765381,
+      "grad_norm": 0.33626309268971577,
+      "learning_rate": 7.870641683739358e-06,
+      "loss": 0.4845,
+      "step": 7249
+    },
+    {
+      "epoch": 1.1906472604848808,
+      "grad_norm": 0.40524575271035757,
+      "learning_rate": 7.870330955220124e-06,
+      "loss": 0.476,
+      "step": 7250
+    },
+    {
+      "epoch": 1.1908114876932236,
+      "grad_norm": 0.30297205462394605,
+      "learning_rate": 7.870020190095783e-06,
+      "loss": 0.4827,
+      "step": 7251
+    },
+    {
+      "epoch": 1.1909757149015663,
+      "grad_norm": 0.2851543936766989,
+      "learning_rate": 7.869709388369705e-06,
+      "loss": 0.4808,
+      "step": 7252
+    },
+    {
+      "epoch": 1.1911399421099091,
+      "grad_norm": 0.27115976850848544,
+      "learning_rate": 7.869398550045268e-06,
+      "loss": 0.4858,
+      "step": 7253
+    },
+    {
+      "epoch": 1.1913041693182518,
+      "grad_norm": 0.30476809755351303,
+      "learning_rate": 7.86908767512585e-06,
+      "loss": 0.4876,
+      "step": 7254
+    },
+    {
+      "epoch": 1.1914683965265946,
+      "grad_norm": 0.3758066090495282,
+      "learning_rate": 7.868776763614824e-06,
+      "loss": 0.4615,
+      "step": 7255
+    },
+    {
+      "epoch": 1.1916326237349373,
+      "grad_norm": 0.2936697469025561,
+      "learning_rate": 7.868465815515568e-06,
+      "loss": 0.4939,
+      "step": 7256
+    },
+    {
+      "epoch": 1.19179685094328,
+      "grad_norm": 0.2704607304163568,
+      "learning_rate": 7.868154830831458e-06,
+      "loss": 0.4794,
+      "step": 7257
+    },
+    {
+      "epoch": 1.1919610781516228,
+      "grad_norm": 0.28176731888861567,
+      "learning_rate": 7.867843809565873e-06,
+      "loss": 0.4723,
+      "step": 7258
+    },
+    {
+      "epoch": 1.1921253053599656,
+      "grad_norm": 0.2760536908218363,
+      "learning_rate": 7.86753275172219e-06,
+      "loss": 0.4917,
+      "step": 7259
+    },
+    {
+      "epoch": 1.1922895325683083,
+      "grad_norm": 0.25631381946862203,
+      "learning_rate": 7.86722165730379e-06,
+      "loss": 0.4652,
+      "step": 7260
+    },
+    {
+      "epoch": 1.192453759776651,
+      "grad_norm": 0.27318642320256536,
+      "learning_rate": 7.866910526314049e-06,
+      "loss": 0.4479,
+      "step": 7261
+    },
+    {
+      "epoch": 1.1926179869849938,
+      "grad_norm": 0.28956412253272856,
+      "learning_rate": 7.866599358756347e-06,
+      "loss": 0.4779,
+      "step": 7262
+    },
+    {
+      "epoch": 1.1927822141933364,
+      "grad_norm": 0.30001643869690897,
+      "learning_rate": 7.866288154634064e-06,
+      "loss": 0.4873,
+      "step": 7263
+    },
+    {
+      "epoch": 1.1929464414016793,
+      "grad_norm": 0.2924444418420769,
+      "learning_rate": 7.86597691395058e-06,
+      "loss": 0.4865,
+      "step": 7264
+    },
+    {
+      "epoch": 1.193110668610022,
+      "grad_norm": 0.25822309694150847,
+      "learning_rate": 7.865665636709275e-06,
+      "loss": 0.4937,
+      "step": 7265
+    },
+    {
+      "epoch": 1.1932748958183648,
+      "grad_norm": 0.2582323654371515,
+      "learning_rate": 7.865354322913529e-06,
+      "loss": 0.4789,
+      "step": 7266
+    },
+    {
+      "epoch": 1.1934391230267074,
+      "grad_norm": 0.2560281240282641,
+      "learning_rate": 7.865042972566723e-06,
+      "loss": 0.4958,
+      "step": 7267
+    },
+    {
+      "epoch": 1.1936033502350503,
+      "grad_norm": 0.283105334712507,
+      "learning_rate": 7.86473158567224e-06,
+      "loss": 0.4857,
+      "step": 7268
+    },
+    {
+      "epoch": 1.193767577443393,
+      "grad_norm": 0.3074559301853208,
+      "learning_rate": 7.864420162233464e-06,
+      "loss": 0.4822,
+      "step": 7269
+    },
+    {
+      "epoch": 1.1939318046517358,
+      "grad_norm": 0.272089654585132,
+      "learning_rate": 7.864108702253773e-06,
+      "loss": 0.4826,
+      "step": 7270
+    },
+    {
+      "epoch": 1.1940960318600784,
+      "grad_norm": 0.34574826869825065,
+      "learning_rate": 7.863797205736552e-06,
+      "loss": 0.4898,
+      "step": 7271
+    },
+    {
+      "epoch": 1.1942602590684213,
+      "grad_norm": 0.3400107585169252,
+      "learning_rate": 7.863485672685183e-06,
+      "loss": 0.4679,
+      "step": 7272
+    },
+    {
+      "epoch": 1.194424486276764,
+      "grad_norm": 0.2794512133883541,
+      "learning_rate": 7.863174103103053e-06,
+      "loss": 0.4935,
+      "step": 7273
+    },
+    {
+      "epoch": 1.1945887134851065,
+      "grad_norm": 0.26107892204001254,
+      "learning_rate": 7.862862496993541e-06,
+      "loss": 0.4756,
+      "step": 7274
+    },
+    {
+      "epoch": 1.1947529406934494,
+      "grad_norm": 0.3010930525172073,
+      "learning_rate": 7.862550854360033e-06,
+      "loss": 0.4778,
+      "step": 7275
+    },
+    {
+      "epoch": 1.1949171679017923,
+      "grad_norm": 0.31706413409182155,
+      "learning_rate": 7.862239175205915e-06,
+      "loss": 0.4893,
+      "step": 7276
+    },
+    {
+      "epoch": 1.195081395110135,
+      "grad_norm": 0.6518163187078756,
+      "learning_rate": 7.861927459534572e-06,
+      "loss": 0.4701,
+      "step": 7277
+    },
+    {
+      "epoch": 1.1952456223184775,
+      "grad_norm": 0.2829791035958546,
+      "learning_rate": 7.86161570734939e-06,
+      "loss": 0.4871,
+      "step": 7278
+    },
+    {
+      "epoch": 1.1954098495268204,
+      "grad_norm": 0.2739474278242147,
+      "learning_rate": 7.861303918653752e-06,
+      "loss": 0.4755,
+      "step": 7279
+    },
+    {
+      "epoch": 1.195574076735163,
+      "grad_norm": 0.3436202620943284,
+      "learning_rate": 7.860992093451049e-06,
+      "loss": 0.4733,
+      "step": 7280
+    },
+    {
+      "epoch": 1.1957383039435059,
+      "grad_norm": 0.30182634928669677,
+      "learning_rate": 7.860680231744663e-06,
+      "loss": 0.5101,
+      "step": 7281
+    },
+    {
+      "epoch": 1.1959025311518485,
+      "grad_norm": 0.2987955559095261,
+      "learning_rate": 7.860368333537984e-06,
+      "loss": 0.5102,
+      "step": 7282
+    },
+    {
+      "epoch": 1.1960667583601914,
+      "grad_norm": 0.2696173051533773,
+      "learning_rate": 7.860056398834399e-06,
+      "loss": 0.4914,
+      "step": 7283
+    },
+    {
+      "epoch": 1.196230985568534,
+      "grad_norm": 0.3588548931620607,
+      "learning_rate": 7.859744427637295e-06,
+      "loss": 0.4714,
+      "step": 7284
+    },
+    {
+      "epoch": 1.1963952127768769,
+      "grad_norm": 0.2960177348685412,
+      "learning_rate": 7.85943241995006e-06,
+      "loss": 0.487,
+      "step": 7285
+    },
+    {
+      "epoch": 1.1965594399852195,
+      "grad_norm": 0.3321192088445762,
+      "learning_rate": 7.859120375776086e-06,
+      "loss": 0.4574,
+      "step": 7286
+    },
+    {
+      "epoch": 1.1967236671935624,
+      "grad_norm": 0.3047274250456088,
+      "learning_rate": 7.85880829511876e-06,
+      "loss": 0.4767,
+      "step": 7287
+    },
+    {
+      "epoch": 1.196887894401905,
+      "grad_norm": 0.2940277359499893,
+      "learning_rate": 7.85849617798147e-06,
+      "loss": 0.4727,
+      "step": 7288
+    },
+    {
+      "epoch": 1.1970521216102479,
+      "grad_norm": 0.31805435075245675,
+      "learning_rate": 7.858184024367606e-06,
+      "loss": 0.5016,
+      "step": 7289
+    },
+    {
+      "epoch": 1.1972163488185905,
+      "grad_norm": 0.2893430047178316,
+      "learning_rate": 7.857871834280562e-06,
+      "loss": 0.473,
+      "step": 7290
+    },
+    {
+      "epoch": 1.1973805760269332,
+      "grad_norm": 0.3214483934935793,
+      "learning_rate": 7.857559607723724e-06,
+      "loss": 0.4891,
+      "step": 7291
+    },
+    {
+      "epoch": 1.197544803235276,
+      "grad_norm": 0.2779040042258734,
+      "learning_rate": 7.857247344700485e-06,
+      "loss": 0.4857,
+      "step": 7292
+    },
+    {
+      "epoch": 1.1977090304436186,
+      "grad_norm": 0.29268673694745784,
+      "learning_rate": 7.85693504521424e-06,
+      "loss": 0.5038,
+      "step": 7293
+    },
+    {
+      "epoch": 1.1978732576519615,
+      "grad_norm": 0.2583632166117344,
+      "learning_rate": 7.856622709268375e-06,
+      "loss": 0.4702,
+      "step": 7294
+    },
+    {
+      "epoch": 1.1980374848603041,
+      "grad_norm": 0.3039569672751797,
+      "learning_rate": 7.856310336866284e-06,
+      "loss": 0.4916,
+      "step": 7295
+    },
+    {
+      "epoch": 1.198201712068647,
+      "grad_norm": 0.37623591673858825,
+      "learning_rate": 7.855997928011363e-06,
+      "loss": 0.4649,
+      "step": 7296
+    },
+    {
+      "epoch": 1.1983659392769896,
+      "grad_norm": 0.32792556540288026,
+      "learning_rate": 7.855685482707001e-06,
+      "loss": 0.5132,
+      "step": 7297
+    },
+    {
+      "epoch": 1.1985301664853325,
+      "grad_norm": 0.33159113984877747,
+      "learning_rate": 7.855373000956595e-06,
+      "loss": 0.4862,
+      "step": 7298
+    },
+    {
+      "epoch": 1.1986943936936751,
+      "grad_norm": 0.46686228289895976,
+      "learning_rate": 7.855060482763534e-06,
+      "loss": 0.5074,
+      "step": 7299
+    },
+    {
+      "epoch": 1.198858620902018,
+      "grad_norm": 0.29765525293464823,
+      "learning_rate": 7.854747928131219e-06,
+      "loss": 0.4779,
+      "step": 7300
+    },
+    {
+      "epoch": 1.1990228481103606,
+      "grad_norm": 0.31934509278801926,
+      "learning_rate": 7.854435337063037e-06,
+      "loss": 0.4852,
+      "step": 7301
+    },
+    {
+      "epoch": 1.1991870753187035,
+      "grad_norm": 0.424679981759697,
+      "learning_rate": 7.85412270956239e-06,
+      "loss": 0.4786,
+      "step": 7302
+    },
+    {
+      "epoch": 1.1993513025270461,
+      "grad_norm": 0.28126109328181276,
+      "learning_rate": 7.853810045632668e-06,
+      "loss": 0.4889,
+      "step": 7303
+    },
+    {
+      "epoch": 1.199515529735389,
+      "grad_norm": 0.27650129426773495,
+      "learning_rate": 7.853497345277272e-06,
+      "loss": 0.4622,
+      "step": 7304
+    },
+    {
+      "epoch": 1.1996797569437316,
+      "grad_norm": 0.3398188905283428,
+      "learning_rate": 7.853184608499593e-06,
+      "loss": 0.4689,
+      "step": 7305
+    },
+    {
+      "epoch": 1.1998439841520745,
+      "grad_norm": 0.298516041385809,
+      "learning_rate": 7.852871835303031e-06,
+      "loss": 0.4656,
+      "step": 7306
+    },
+    {
+      "epoch": 1.2000082113604171,
+      "grad_norm": 0.24499209339877884,
+      "learning_rate": 7.852559025690981e-06,
+      "loss": 0.4924,
+      "step": 7307
+    },
+    {
+      "epoch": 1.2001724385687598,
+      "grad_norm": 0.27863042979543257,
+      "learning_rate": 7.852246179666844e-06,
+      "loss": 0.4795,
+      "step": 7308
+    },
+    {
+      "epoch": 1.2003366657771026,
+      "grad_norm": 0.2756647160445179,
+      "learning_rate": 7.851933297234012e-06,
+      "loss": 0.4754,
+      "step": 7309
+    },
+    {
+      "epoch": 1.2005008929854453,
+      "grad_norm": 0.2831834434790002,
+      "learning_rate": 7.85162037839589e-06,
+      "loss": 0.4737,
+      "step": 7310
+    },
+    {
+      "epoch": 1.2006651201937881,
+      "grad_norm": 0.29162631607674616,
+      "learning_rate": 7.851307423155871e-06,
+      "loss": 0.4983,
+      "step": 7311
+    },
+    {
+      "epoch": 1.2008293474021308,
+      "grad_norm": 0.27672574832406416,
+      "learning_rate": 7.850994431517356e-06,
+      "loss": 0.4777,
+      "step": 7312
+    },
+    {
+      "epoch": 1.2009935746104736,
+      "grad_norm": 0.27869712581342154,
+      "learning_rate": 7.850681403483745e-06,
+      "loss": 0.506,
+      "step": 7313
+    },
+    {
+      "epoch": 1.2011578018188163,
+      "grad_norm": 0.2923731692650365,
+      "learning_rate": 7.850368339058438e-06,
+      "loss": 0.503,
+      "step": 7314
+    },
+    {
+      "epoch": 1.2013220290271591,
+      "grad_norm": 0.42420781724762624,
+      "learning_rate": 7.850055238244835e-06,
+      "loss": 0.472,
+      "step": 7315
+    },
+    {
+      "epoch": 1.2014862562355018,
+      "grad_norm": 0.2648715107534324,
+      "learning_rate": 7.849742101046333e-06,
+      "loss": 0.4963,
+      "step": 7316
+    },
+    {
+      "epoch": 1.2016504834438446,
+      "grad_norm": 0.2982734322965936,
+      "learning_rate": 7.849428927466338e-06,
+      "loss": 0.4765,
+      "step": 7317
+    },
+    {
+      "epoch": 1.2018147106521873,
+      "grad_norm": 0.34924463673787454,
+      "learning_rate": 7.849115717508252e-06,
+      "loss": 0.4945,
+      "step": 7318
+    },
+    {
+      "epoch": 1.2019789378605301,
+      "grad_norm": 0.3021561206760538,
+      "learning_rate": 7.84880247117547e-06,
+      "loss": 0.4746,
+      "step": 7319
+    },
+    {
+      "epoch": 1.2021431650688728,
+      "grad_norm": 0.4022430865737298,
+      "learning_rate": 7.848489188471401e-06,
+      "loss": 0.4897,
+      "step": 7320
+    },
+    {
+      "epoch": 1.2023073922772156,
+      "grad_norm": 0.28673004576717576,
+      "learning_rate": 7.848175869399444e-06,
+      "loss": 0.4809,
+      "step": 7321
+    },
+    {
+      "epoch": 1.2024716194855583,
+      "grad_norm": 0.2604773400575898,
+      "learning_rate": 7.847862513963003e-06,
+      "loss": 0.4712,
+      "step": 7322
+    },
+    {
+      "epoch": 1.2026358466939011,
+      "grad_norm": 0.44028520566728796,
+      "learning_rate": 7.847549122165481e-06,
+      "loss": 0.4975,
+      "step": 7323
+    },
+    {
+      "epoch": 1.2028000739022437,
+      "grad_norm": 0.30359081977857466,
+      "learning_rate": 7.847235694010283e-06,
+      "loss": 0.4754,
+      "step": 7324
+    },
+    {
+      "epoch": 1.2029643011105864,
+      "grad_norm": 0.2688194720858126,
+      "learning_rate": 7.846922229500812e-06,
+      "loss": 0.4809,
+      "step": 7325
+    },
+    {
+      "epoch": 1.2031285283189292,
+      "grad_norm": 0.27441461039552933,
+      "learning_rate": 7.846608728640471e-06,
+      "loss": 0.4881,
+      "step": 7326
+    },
+    {
+      "epoch": 1.2032927555272719,
+      "grad_norm": 0.36663350552704727,
+      "learning_rate": 7.846295191432668e-06,
+      "loss": 0.4883,
+      "step": 7327
+    },
+    {
+      "epoch": 1.2034569827356147,
+      "grad_norm": 0.30093440887610545,
+      "learning_rate": 7.845981617880808e-06,
+      "loss": 0.4791,
+      "step": 7328
+    },
+    {
+      "epoch": 1.2036212099439574,
+      "grad_norm": 0.30653428630193286,
+      "learning_rate": 7.845668007988292e-06,
+      "loss": 0.47,
+      "step": 7329
+    },
+    {
+      "epoch": 1.2037854371523002,
+      "grad_norm": 0.2898402343328568,
+      "learning_rate": 7.845354361758533e-06,
+      "loss": 0.4795,
+      "step": 7330
+    },
+    {
+      "epoch": 1.2039496643606429,
+      "grad_norm": 0.2890752297414382,
+      "learning_rate": 7.845040679194934e-06,
+      "loss": 0.4865,
+      "step": 7331
+    },
+    {
+      "epoch": 1.2041138915689857,
+      "grad_norm": 0.3624219698154435,
+      "learning_rate": 7.8447269603009e-06,
+      "loss": 0.475,
+      "step": 7332
+    },
+    {
+      "epoch": 1.2042781187773284,
+      "grad_norm": 0.296336578656,
+      "learning_rate": 7.844413205079842e-06,
+      "loss": 0.4954,
+      "step": 7333
+    },
+    {
+      "epoch": 1.2044423459856712,
+      "grad_norm": 0.35399910890682923,
+      "learning_rate": 7.844099413535167e-06,
+      "loss": 0.4947,
+      "step": 7334
+    },
+    {
+      "epoch": 1.2046065731940139,
+      "grad_norm": 0.33519400632511764,
+      "learning_rate": 7.843785585670279e-06,
+      "loss": 0.4739,
+      "step": 7335
+    },
+    {
+      "epoch": 1.2047708004023567,
+      "grad_norm": 0.28240279450470623,
+      "learning_rate": 7.843471721488593e-06,
+      "loss": 0.5025,
+      "step": 7336
+    },
+    {
+      "epoch": 1.2049350276106994,
+      "grad_norm": 0.25326684831156665,
+      "learning_rate": 7.843157820993515e-06,
+      "loss": 0.4706,
+      "step": 7337
+    },
+    {
+      "epoch": 1.2050992548190422,
+      "grad_norm": 0.29493194722350735,
+      "learning_rate": 7.84284388418845e-06,
+      "loss": 0.4717,
+      "step": 7338
+    },
+    {
+      "epoch": 1.2052634820273849,
+      "grad_norm": 0.298157102180501,
+      "learning_rate": 7.842529911076815e-06,
+      "loss": 0.5083,
+      "step": 7339
+    },
+    {
+      "epoch": 1.2054277092357277,
+      "grad_norm": 0.4151324217642813,
+      "learning_rate": 7.842215901662014e-06,
+      "loss": 0.4794,
+      "step": 7340
+    },
+    {
+      "epoch": 1.2055919364440704,
+      "grad_norm": 0.2826353275894072,
+      "learning_rate": 7.841901855947462e-06,
+      "loss": 0.4877,
+      "step": 7341
+    },
+    {
+      "epoch": 1.205756163652413,
+      "grad_norm": 0.28516782000663105,
+      "learning_rate": 7.841587773936568e-06,
+      "loss": 0.4773,
+      "step": 7342
+    },
+    {
+      "epoch": 1.2059203908607559,
+      "grad_norm": 0.3039940380114767,
+      "learning_rate": 7.841273655632741e-06,
+      "loss": 0.4945,
+      "step": 7343
+    },
+    {
+      "epoch": 1.2060846180690985,
+      "grad_norm": 0.40356339790866297,
+      "learning_rate": 7.840959501039397e-06,
+      "loss": 0.4859,
+      "step": 7344
+    },
+    {
+      "epoch": 1.2062488452774414,
+      "grad_norm": 0.5442637135737383,
+      "learning_rate": 7.840645310159945e-06,
+      "loss": 0.474,
+      "step": 7345
+    },
+    {
+      "epoch": 1.206413072485784,
+      "grad_norm": 0.3455945345487415,
+      "learning_rate": 7.840331082997799e-06,
+      "loss": 0.5056,
+      "step": 7346
+    },
+    {
+      "epoch": 1.2065772996941269,
+      "grad_norm": 0.3203479224628378,
+      "learning_rate": 7.840016819556369e-06,
+      "loss": 0.491,
+      "step": 7347
+    },
+    {
+      "epoch": 1.2067415269024695,
+      "grad_norm": 0.43912639657814323,
+      "learning_rate": 7.839702519839071e-06,
+      "loss": 0.4847,
+      "step": 7348
+    },
+    {
+      "epoch": 1.2069057541108124,
+      "grad_norm": 0.2913019474705546,
+      "learning_rate": 7.839388183849318e-06,
+      "loss": 0.4542,
+      "step": 7349
+    },
+    {
+      "epoch": 1.207069981319155,
+      "grad_norm": 0.2997673311165085,
+      "learning_rate": 7.839073811590524e-06,
+      "loss": 0.451,
+      "step": 7350
+    },
+    {
+      "epoch": 1.2072342085274979,
+      "grad_norm": 0.32497023477104164,
+      "learning_rate": 7.838759403066103e-06,
+      "loss": 0.4607,
+      "step": 7351
+    },
+    {
+      "epoch": 1.2073984357358405,
+      "grad_norm": 0.30168217147672366,
+      "learning_rate": 7.83844495827947e-06,
+      "loss": 0.4608,
+      "step": 7352
+    },
+    {
+      "epoch": 1.2075626629441834,
+      "grad_norm": 0.3171359606091761,
+      "learning_rate": 7.83813047723404e-06,
+      "loss": 0.4863,
+      "step": 7353
+    },
+    {
+      "epoch": 1.207726890152526,
+      "grad_norm": 0.3601274627022798,
+      "learning_rate": 7.83781595993323e-06,
+      "loss": 0.4696,
+      "step": 7354
+    },
+    {
+      "epoch": 1.2078911173608688,
+      "grad_norm": 0.3203665019807861,
+      "learning_rate": 7.837501406380452e-06,
+      "loss": 0.4717,
+      "step": 7355
+    },
+    {
+      "epoch": 1.2080553445692115,
+      "grad_norm": 0.29351737271594,
+      "learning_rate": 7.837186816579128e-06,
+      "loss": 0.4904,
+      "step": 7356
+    },
+    {
+      "epoch": 1.2082195717775543,
+      "grad_norm": 0.33220095890415596,
+      "learning_rate": 7.83687219053267e-06,
+      "loss": 0.4878,
+      "step": 7357
+    },
+    {
+      "epoch": 1.208383798985897,
+      "grad_norm": 0.2975976385334508,
+      "learning_rate": 7.836557528244497e-06,
+      "loss": 0.4774,
+      "step": 7358
+    },
+    {
+      "epoch": 1.2085480261942396,
+      "grad_norm": 0.277639522422605,
+      "learning_rate": 7.836242829718028e-06,
+      "loss": 0.4827,
+      "step": 7359
+    },
+    {
+      "epoch": 1.2087122534025825,
+      "grad_norm": 0.28230791338424405,
+      "learning_rate": 7.835928094956677e-06,
+      "loss": 0.4858,
+      "step": 7360
+    },
+    {
+      "epoch": 1.2088764806109251,
+      "grad_norm": 0.3070274198241027,
+      "learning_rate": 7.835613323963867e-06,
+      "loss": 0.4732,
+      "step": 7361
+    },
+    {
+      "epoch": 1.209040707819268,
+      "grad_norm": 0.3168714949165464,
+      "learning_rate": 7.835298516743014e-06,
+      "loss": 0.4867,
+      "step": 7362
+    },
+    {
+      "epoch": 1.2092049350276106,
+      "grad_norm": 0.3453218427128881,
+      "learning_rate": 7.834983673297537e-06,
+      "loss": 0.4556,
+      "step": 7363
+    },
+    {
+      "epoch": 1.2093691622359535,
+      "grad_norm": 0.3528769701781255,
+      "learning_rate": 7.834668793630856e-06,
+      "loss": 0.4985,
+      "step": 7364
+    },
+    {
+      "epoch": 1.2095333894442961,
+      "grad_norm": 0.2638158286045824,
+      "learning_rate": 7.834353877746391e-06,
+      "loss": 0.4675,
+      "step": 7365
+    },
+    {
+      "epoch": 1.209697616652639,
+      "grad_norm": 0.2788721286113114,
+      "learning_rate": 7.834038925647563e-06,
+      "loss": 0.4796,
+      "step": 7366
+    },
+    {
+      "epoch": 1.2098618438609816,
+      "grad_norm": 0.31074874505065736,
+      "learning_rate": 7.833723937337792e-06,
+      "loss": 0.4873,
+      "step": 7367
+    },
+    {
+      "epoch": 1.2100260710693245,
+      "grad_norm": 0.2616105603314999,
+      "learning_rate": 7.8334089128205e-06,
+      "loss": 0.4853,
+      "step": 7368
+    },
+    {
+      "epoch": 1.210190298277667,
+      "grad_norm": 0.25894275590866295,
+      "learning_rate": 7.833093852099104e-06,
+      "loss": 0.4763,
+      "step": 7369
+    },
+    {
+      "epoch": 1.21035452548601,
+      "grad_norm": 0.28972598700545293,
+      "learning_rate": 7.832778755177034e-06,
+      "loss": 0.4718,
+      "step": 7370
+    },
+    {
+      "epoch": 1.2105187526943526,
+      "grad_norm": 0.53511817324183,
+      "learning_rate": 7.832463622057705e-06,
+      "loss": 0.478,
+      "step": 7371
+    },
+    {
+      "epoch": 1.2106829799026955,
+      "grad_norm": 0.2908338880957059,
+      "learning_rate": 7.832148452744544e-06,
+      "loss": 0.4872,
+      "step": 7372
+    },
+    {
+      "epoch": 1.210847207111038,
+      "grad_norm": 0.31807390385130946,
+      "learning_rate": 7.83183324724097e-06,
+      "loss": 0.492,
+      "step": 7373
+    },
+    {
+      "epoch": 1.211011434319381,
+      "grad_norm": 0.296720479252511,
+      "learning_rate": 7.831518005550412e-06,
+      "loss": 0.4774,
+      "step": 7374
+    },
+    {
+      "epoch": 1.2111756615277236,
+      "grad_norm": 0.2966182688407565,
+      "learning_rate": 7.831202727676287e-06,
+      "loss": 0.4725,
+      "step": 7375
+    },
+    {
+      "epoch": 1.2113398887360662,
+      "grad_norm": 0.2888443450249788,
+      "learning_rate": 7.830887413622027e-06,
+      "loss": 0.4726,
+      "step": 7376
+    },
+    {
+      "epoch": 1.211504115944409,
+      "grad_norm": 0.27658756184306366,
+      "learning_rate": 7.830572063391049e-06,
+      "loss": 0.4877,
+      "step": 7377
+    },
+    {
+      "epoch": 1.2116683431527517,
+      "grad_norm": 0.3201453442232231,
+      "learning_rate": 7.830256676986785e-06,
+      "loss": 0.4801,
+      "step": 7378
+    },
+    {
+      "epoch": 1.2118325703610946,
+      "grad_norm": 0.30096927890993763,
+      "learning_rate": 7.829941254412654e-06,
+      "loss": 0.4828,
+      "step": 7379
+    },
+    {
+      "epoch": 1.2119967975694372,
+      "grad_norm": 0.3084680100763514,
+      "learning_rate": 7.829625795672085e-06,
+      "loss": 0.4981,
+      "step": 7380
+    },
+    {
+      "epoch": 1.21216102477778,
+      "grad_norm": 0.2678693496989439,
+      "learning_rate": 7.829310300768505e-06,
+      "loss": 0.4669,
+      "step": 7381
+    },
+    {
+      "epoch": 1.2123252519861227,
+      "grad_norm": 0.2697735354847363,
+      "learning_rate": 7.828994769705339e-06,
+      "loss": 0.4929,
+      "step": 7382
+    },
+    {
+      "epoch": 1.2124894791944656,
+      "grad_norm": 0.2560319566439462,
+      "learning_rate": 7.828679202486015e-06,
+      "loss": 0.4841,
+      "step": 7383
+    },
+    {
+      "epoch": 1.2126537064028082,
+      "grad_norm": 0.2836939551834787,
+      "learning_rate": 7.828363599113959e-06,
+      "loss": 0.4697,
+      "step": 7384
+    },
+    {
+      "epoch": 1.212817933611151,
+      "grad_norm": 0.31007946274580067,
+      "learning_rate": 7.828047959592601e-06,
+      "loss": 0.4882,
+      "step": 7385
+    },
+    {
+      "epoch": 1.2129821608194937,
+      "grad_norm": 0.29459524995287784,
+      "learning_rate": 7.827732283925366e-06,
+      "loss": 0.482,
+      "step": 7386
+    },
+    {
+      "epoch": 1.2131463880278366,
+      "grad_norm": 0.3793586584509073,
+      "learning_rate": 7.827416572115686e-06,
+      "loss": 0.4817,
+      "step": 7387
+    },
+    {
+      "epoch": 1.2133106152361792,
+      "grad_norm": 0.2644897980491223,
+      "learning_rate": 7.827100824166988e-06,
+      "loss": 0.4794,
+      "step": 7388
+    },
+    {
+      "epoch": 1.213474842444522,
+      "grad_norm": 0.3017168977924796,
+      "learning_rate": 7.826785040082702e-06,
+      "loss": 0.4818,
+      "step": 7389
+    },
+    {
+      "epoch": 1.2136390696528647,
+      "grad_norm": 0.305708816789366,
+      "learning_rate": 7.826469219866257e-06,
+      "loss": 0.4802,
+      "step": 7390
+    },
+    {
+      "epoch": 1.2138032968612076,
+      "grad_norm": 0.3059340241183472,
+      "learning_rate": 7.826153363521082e-06,
+      "loss": 0.5047,
+      "step": 7391
+    },
+    {
+      "epoch": 1.2139675240695502,
+      "grad_norm": 0.40313192821803495,
+      "learning_rate": 7.82583747105061e-06,
+      "loss": 0.497,
+      "step": 7392
+    },
+    {
+      "epoch": 1.2141317512778929,
+      "grad_norm": 0.3177682689036567,
+      "learning_rate": 7.82552154245827e-06,
+      "loss": 0.4912,
+      "step": 7393
+    },
+    {
+      "epoch": 1.2142959784862357,
+      "grad_norm": 0.33379735245100434,
+      "learning_rate": 7.825205577747495e-06,
+      "loss": 0.4787,
+      "step": 7394
+    },
+    {
+      "epoch": 1.2144602056945784,
+      "grad_norm": 0.32610810464507667,
+      "learning_rate": 7.824889576921718e-06,
+      "loss": 0.4641,
+      "step": 7395
+    },
+    {
+      "epoch": 1.2146244329029212,
+      "grad_norm": 0.2576373012534115,
+      "learning_rate": 7.824573539984367e-06,
+      "loss": 0.48,
+      "step": 7396
+    },
+    {
+      "epoch": 1.2147886601112639,
+      "grad_norm": 0.2962684067235093,
+      "learning_rate": 7.824257466938875e-06,
+      "loss": 0.4809,
+      "step": 7397
+    },
+    {
+      "epoch": 1.2149528873196067,
+      "grad_norm": 0.3373698620542774,
+      "learning_rate": 7.823941357788679e-06,
+      "loss": 0.4949,
+      "step": 7398
+    },
+    {
+      "epoch": 1.2151171145279493,
+      "grad_norm": 0.2714003069990936,
+      "learning_rate": 7.823625212537206e-06,
+      "loss": 0.4791,
+      "step": 7399
+    },
+    {
+      "epoch": 1.2152813417362922,
+      "grad_norm": 0.27004087404298466,
+      "learning_rate": 7.823309031187897e-06,
+      "loss": 0.4778,
+      "step": 7400
+    },
+    {
+      "epoch": 1.2154455689446348,
+      "grad_norm": 0.27836482022600734,
+      "learning_rate": 7.82299281374418e-06,
+      "loss": 0.4814,
+      "step": 7401
+    },
+    {
+      "epoch": 1.2156097961529777,
+      "grad_norm": 0.5709071029748667,
+      "learning_rate": 7.822676560209493e-06,
+      "loss": 0.4845,
+      "step": 7402
+    },
+    {
+      "epoch": 1.2157740233613203,
+      "grad_norm": 0.28200824916519507,
+      "learning_rate": 7.822360270587269e-06,
+      "loss": 0.4959,
+      "step": 7403
+    },
+    {
+      "epoch": 1.2159382505696632,
+      "grad_norm": 0.36164400269974106,
+      "learning_rate": 7.822043944880943e-06,
+      "loss": 0.4751,
+      "step": 7404
+    },
+    {
+      "epoch": 1.2161024777780058,
+      "grad_norm": 0.2920172636322637,
+      "learning_rate": 7.821727583093951e-06,
+      "loss": 0.4898,
+      "step": 7405
+    },
+    {
+      "epoch": 1.2162667049863487,
+      "grad_norm": 0.27259579242872134,
+      "learning_rate": 7.821411185229732e-06,
+      "loss": 0.4832,
+      "step": 7406
+    },
+    {
+      "epoch": 1.2164309321946913,
+      "grad_norm": 2.1964313211670277,
+      "learning_rate": 7.821094751291716e-06,
+      "loss": 0.4811,
+      "step": 7407
+    },
+    {
+      "epoch": 1.2165951594030342,
+      "grad_norm": 0.3163824618854596,
+      "learning_rate": 7.820778281283346e-06,
+      "loss": 0.4965,
+      "step": 7408
+    },
+    {
+      "epoch": 1.2167593866113768,
+      "grad_norm": 0.29345503086766606,
+      "learning_rate": 7.820461775208056e-06,
+      "loss": 0.4871,
+      "step": 7409
+    },
+    {
+      "epoch": 1.2169236138197195,
+      "grad_norm": 0.30196703868592945,
+      "learning_rate": 7.820145233069284e-06,
+      "loss": 0.4605,
+      "step": 7410
+    },
+    {
+      "epoch": 1.2170878410280623,
+      "grad_norm": 0.28311895504982654,
+      "learning_rate": 7.81982865487047e-06,
+      "loss": 0.4912,
+      "step": 7411
+    },
+    {
+      "epoch": 1.217252068236405,
+      "grad_norm": 0.30339328791642767,
+      "learning_rate": 7.819512040615047e-06,
+      "loss": 0.4609,
+      "step": 7412
+    },
+    {
+      "epoch": 1.2174162954447478,
+      "grad_norm": 0.2596969919404455,
+      "learning_rate": 7.819195390306459e-06,
+      "loss": 0.4753,
+      "step": 7413
+    },
+    {
+      "epoch": 1.2175805226530905,
+      "grad_norm": 0.322969273895278,
+      "learning_rate": 7.818878703948144e-06,
+      "loss": 0.4794,
+      "step": 7414
+    },
+    {
+      "epoch": 1.2177447498614333,
+      "grad_norm": 0.3519640745006145,
+      "learning_rate": 7.818561981543541e-06,
+      "loss": 0.4773,
+      "step": 7415
+    },
+    {
+      "epoch": 1.217908977069776,
+      "grad_norm": 0.34107543982912136,
+      "learning_rate": 7.81824522309609e-06,
+      "loss": 0.492,
+      "step": 7416
+    },
+    {
+      "epoch": 1.2180732042781188,
+      "grad_norm": 0.3281282771747284,
+      "learning_rate": 7.817928428609229e-06,
+      "loss": 0.4803,
+      "step": 7417
+    },
+    {
+      "epoch": 1.2182374314864615,
+      "grad_norm": 0.5308962228651787,
+      "learning_rate": 7.817611598086403e-06,
+      "loss": 0.4642,
+      "step": 7418
+    },
+    {
+      "epoch": 1.2184016586948043,
+      "grad_norm": 0.26900387954254784,
+      "learning_rate": 7.81729473153105e-06,
+      "loss": 0.4797,
+      "step": 7419
+    },
+    {
+      "epoch": 1.218565885903147,
+      "grad_norm": 0.28024744617753733,
+      "learning_rate": 7.816977828946612e-06,
+      "loss": 0.4846,
+      "step": 7420
+    },
+    {
+      "epoch": 1.2187301131114898,
+      "grad_norm": 0.2647335194613172,
+      "learning_rate": 7.816660890336532e-06,
+      "loss": 0.4755,
+      "step": 7421
+    },
+    {
+      "epoch": 1.2188943403198325,
+      "grad_norm": 0.2945285536442789,
+      "learning_rate": 7.816343915704252e-06,
+      "loss": 0.4746,
+      "step": 7422
+    },
+    {
+      "epoch": 1.2190585675281753,
+      "grad_norm": 0.2835965254364738,
+      "learning_rate": 7.816026905053214e-06,
+      "loss": 0.4782,
+      "step": 7423
+    },
+    {
+      "epoch": 1.219222794736518,
+      "grad_norm": 0.4352423901004788,
+      "learning_rate": 7.815709858386861e-06,
+      "loss": 0.481,
+      "step": 7424
+    },
+    {
+      "epoch": 1.2193870219448608,
+      "grad_norm": 0.2687199290132344,
+      "learning_rate": 7.815392775708639e-06,
+      "loss": 0.4907,
+      "step": 7425
+    },
+    {
+      "epoch": 1.2195512491532035,
+      "grad_norm": 0.26676255690653605,
+      "learning_rate": 7.815075657021986e-06,
+      "loss": 0.4787,
+      "step": 7426
+    },
+    {
+      "epoch": 1.219715476361546,
+      "grad_norm": 0.32625127264289355,
+      "learning_rate": 7.814758502330352e-06,
+      "loss": 0.4822,
+      "step": 7427
+    },
+    {
+      "epoch": 1.219879703569889,
+      "grad_norm": 0.2957996126267077,
+      "learning_rate": 7.814441311637179e-06,
+      "loss": 0.4793,
+      "step": 7428
+    },
+    {
+      "epoch": 1.2200439307782316,
+      "grad_norm": 0.3095382988258092,
+      "learning_rate": 7.814124084945911e-06,
+      "loss": 0.4866,
+      "step": 7429
+    },
+    {
+      "epoch": 1.2202081579865744,
+      "grad_norm": 0.2886374847171023,
+      "learning_rate": 7.813806822259996e-06,
+      "loss": 0.4791,
+      "step": 7430
+    },
+    {
+      "epoch": 1.220372385194917,
+      "grad_norm": 0.3268027407819441,
+      "learning_rate": 7.81348952358288e-06,
+      "loss": 0.4972,
+      "step": 7431
+    },
+    {
+      "epoch": 1.22053661240326,
+      "grad_norm": 0.3406074181446285,
+      "learning_rate": 7.813172188918005e-06,
+      "loss": 0.5041,
+      "step": 7432
+    },
+    {
+      "epoch": 1.2207008396116026,
+      "grad_norm": 0.4363014651036817,
+      "learning_rate": 7.81285481826882e-06,
+      "loss": 0.466,
+      "step": 7433
+    },
+    {
+      "epoch": 1.2208650668199454,
+      "grad_norm": 0.27869275838433155,
+      "learning_rate": 7.812537411638776e-06,
+      "loss": 0.4781,
+      "step": 7434
+    },
+    {
+      "epoch": 1.221029294028288,
+      "grad_norm": 0.28443886078659325,
+      "learning_rate": 7.812219969031313e-06,
+      "loss": 0.4733,
+      "step": 7435
+    },
+    {
+      "epoch": 1.221193521236631,
+      "grad_norm": 0.2876297976501121,
+      "learning_rate": 7.811902490449884e-06,
+      "loss": 0.4638,
+      "step": 7436
+    },
+    {
+      "epoch": 1.2213577484449736,
+      "grad_norm": 0.27526069765232986,
+      "learning_rate": 7.811584975897936e-06,
+      "loss": 0.4671,
+      "step": 7437
+    },
+    {
+      "epoch": 1.2215219756533164,
+      "grad_norm": 0.2595326036832628,
+      "learning_rate": 7.811267425378915e-06,
+      "loss": 0.4738,
+      "step": 7438
+    },
+    {
+      "epoch": 1.221686202861659,
+      "grad_norm": 0.47780546288542847,
+      "learning_rate": 7.810949838896273e-06,
+      "loss": 0.4714,
+      "step": 7439
+    },
+    {
+      "epoch": 1.221850430070002,
+      "grad_norm": 0.3353387377929998,
+      "learning_rate": 7.81063221645346e-06,
+      "loss": 0.4911,
+      "step": 7440
+    },
+    {
+      "epoch": 1.2220146572783446,
+      "grad_norm": 0.2867479853754556,
+      "learning_rate": 7.81031455805392e-06,
+      "loss": 0.4739,
+      "step": 7441
+    },
+    {
+      "epoch": 1.2221788844866874,
+      "grad_norm": 0.2873199750403933,
+      "learning_rate": 7.80999686370111e-06,
+      "loss": 0.4777,
+      "step": 7442
+    },
+    {
+      "epoch": 1.22234311169503,
+      "grad_norm": 0.2786622828477527,
+      "learning_rate": 7.809679133398477e-06,
+      "loss": 0.4834,
+      "step": 7443
+    },
+    {
+      "epoch": 1.2225073389033727,
+      "grad_norm": 0.26644504322703355,
+      "learning_rate": 7.809361367149472e-06,
+      "loss": 0.4781,
+      "step": 7444
+    },
+    {
+      "epoch": 1.2226715661117156,
+      "grad_norm": 0.32504077445550017,
+      "learning_rate": 7.809043564957546e-06,
+      "loss": 0.4964,
+      "step": 7445
+    },
+    {
+      "epoch": 1.2228357933200582,
+      "grad_norm": 0.3443886570485489,
+      "learning_rate": 7.808725726826152e-06,
+      "loss": 0.4831,
+      "step": 7446
+    },
+    {
+      "epoch": 1.223000020528401,
+      "grad_norm": 0.3078921394902557,
+      "learning_rate": 7.808407852758741e-06,
+      "loss": 0.486,
+      "step": 7447
+    },
+    {
+      "epoch": 1.2231642477367437,
+      "grad_norm": 0.3137312338134038,
+      "learning_rate": 7.808089942758765e-06,
+      "loss": 0.4817,
+      "step": 7448
+    },
+    {
+      "epoch": 1.2233284749450866,
+      "grad_norm": 0.29602008603252766,
+      "learning_rate": 7.80777199682968e-06,
+      "loss": 0.4811,
+      "step": 7449
+    },
+    {
+      "epoch": 1.2234927021534292,
+      "grad_norm": 0.33542775379730577,
+      "learning_rate": 7.807454014974935e-06,
+      "loss": 0.4842,
+      "step": 7450
+    },
+    {
+      "epoch": 1.223656929361772,
+      "grad_norm": 0.28587329682265583,
+      "learning_rate": 7.807135997197983e-06,
+      "loss": 0.4821,
+      "step": 7451
+    },
+    {
+      "epoch": 1.2238211565701147,
+      "grad_norm": 0.31582625696235567,
+      "learning_rate": 7.806817943502283e-06,
+      "loss": 0.4764,
+      "step": 7452
+    },
+    {
+      "epoch": 1.2239853837784576,
+      "grad_norm": 0.4508737429529132,
+      "learning_rate": 7.806499853891286e-06,
+      "loss": 0.4682,
+      "step": 7453
+    },
+    {
+      "epoch": 1.2241496109868002,
+      "grad_norm": 0.34199337453924816,
+      "learning_rate": 7.806181728368447e-06,
+      "loss": 0.4876,
+      "step": 7454
+    },
+    {
+      "epoch": 1.224313838195143,
+      "grad_norm": 0.29162050236465775,
+      "learning_rate": 7.805863566937222e-06,
+      "loss": 0.4816,
+      "step": 7455
+    },
+    {
+      "epoch": 1.2244780654034857,
+      "grad_norm": 0.25706162141912203,
+      "learning_rate": 7.805545369601068e-06,
+      "loss": 0.4749,
+      "step": 7456
+    },
+    {
+      "epoch": 1.2246422926118286,
+      "grad_norm": 0.30467782555406403,
+      "learning_rate": 7.805227136363438e-06,
+      "loss": 0.482,
+      "step": 7457
+    },
+    {
+      "epoch": 1.2248065198201712,
+      "grad_norm": 0.4620126635135206,
+      "learning_rate": 7.804908867227787e-06,
+      "loss": 0.4814,
+      "step": 7458
+    },
+    {
+      "epoch": 1.224970747028514,
+      "grad_norm": 0.6847698761546367,
+      "learning_rate": 7.804590562197577e-06,
+      "loss": 0.4663,
+      "step": 7459
+    },
+    {
+      "epoch": 1.2251349742368567,
+      "grad_norm": 0.3919829953997119,
+      "learning_rate": 7.80427222127626e-06,
+      "loss": 0.46,
+      "step": 7460
+    },
+    {
+      "epoch": 1.2252992014451993,
+      "grad_norm": 0.9675079986361146,
+      "learning_rate": 7.803953844467296e-06,
+      "loss": 0.4948,
+      "step": 7461
+    },
+    {
+      "epoch": 1.2254634286535422,
+      "grad_norm": 0.4048339109198711,
+      "learning_rate": 7.803635431774145e-06,
+      "loss": 0.4751,
+      "step": 7462
+    },
+    {
+      "epoch": 1.2256276558618848,
+      "grad_norm": 0.33359033850303504,
+      "learning_rate": 7.80331698320026e-06,
+      "loss": 0.4903,
+      "step": 7463
+    },
+    {
+      "epoch": 1.2257918830702277,
+      "grad_norm": 0.4551179514390609,
+      "learning_rate": 7.802998498749104e-06,
+      "loss": 0.4831,
+      "step": 7464
+    },
+    {
+      "epoch": 1.2259561102785703,
+      "grad_norm": 0.2962278157330623,
+      "learning_rate": 7.802679978424136e-06,
+      "loss": 0.4894,
+      "step": 7465
+    },
+    {
+      "epoch": 1.2261203374869132,
+      "grad_norm": 0.3567481952133428,
+      "learning_rate": 7.802361422228812e-06,
+      "loss": 0.4846,
+      "step": 7466
+    },
+    {
+      "epoch": 1.2262845646952558,
+      "grad_norm": 0.2693500702419807,
+      "learning_rate": 7.802042830166594e-06,
+      "loss": 0.4673,
+      "step": 7467
+    },
+    {
+      "epoch": 1.2264487919035987,
+      "grad_norm": 0.2679369801617365,
+      "learning_rate": 7.801724202240943e-06,
+      "loss": 0.4873,
+      "step": 7468
+    },
+    {
+      "epoch": 1.2266130191119413,
+      "grad_norm": 0.28647992777890036,
+      "learning_rate": 7.801405538455317e-06,
+      "loss": 0.4603,
+      "step": 7469
+    },
+    {
+      "epoch": 1.2267772463202842,
+      "grad_norm": 0.2850138121930835,
+      "learning_rate": 7.801086838813181e-06,
+      "loss": 0.4809,
+      "step": 7470
+    },
+    {
+      "epoch": 1.2269414735286268,
+      "grad_norm": 0.3343754442632032,
+      "learning_rate": 7.800768103317991e-06,
+      "loss": 0.4717,
+      "step": 7471
+    },
+    {
+      "epoch": 1.2271057007369697,
+      "grad_norm": 0.3380996961344888,
+      "learning_rate": 7.800449331973215e-06,
+      "loss": 0.4688,
+      "step": 7472
+    },
+    {
+      "epoch": 1.2272699279453123,
+      "grad_norm": 0.2828645657673943,
+      "learning_rate": 7.80013052478231e-06,
+      "loss": 0.4826,
+      "step": 7473
+    },
+    {
+      "epoch": 1.2274341551536552,
+      "grad_norm": 0.34488922487887297,
+      "learning_rate": 7.799811681748743e-06,
+      "loss": 0.4963,
+      "step": 7474
+    },
+    {
+      "epoch": 1.2275983823619978,
+      "grad_norm": 0.36977738285825107,
+      "learning_rate": 7.799492802875973e-06,
+      "loss": 0.4799,
+      "step": 7475
+    },
+    {
+      "epoch": 1.2277626095703407,
+      "grad_norm": 0.4107523883100698,
+      "learning_rate": 7.799173888167465e-06,
+      "loss": 0.4791,
+      "step": 7476
+    },
+    {
+      "epoch": 1.2279268367786833,
+      "grad_norm": 0.3065610046874601,
+      "learning_rate": 7.798854937626682e-06,
+      "loss": 0.4837,
+      "step": 7477
+    },
+    {
+      "epoch": 1.228091063987026,
+      "grad_norm": 0.3125906559459154,
+      "learning_rate": 7.79853595125709e-06,
+      "loss": 0.4933,
+      "step": 7478
+    },
+    {
+      "epoch": 1.2282552911953688,
+      "grad_norm": 0.2626573945619913,
+      "learning_rate": 7.79821692906215e-06,
+      "loss": 0.4714,
+      "step": 7479
+    },
+    {
+      "epoch": 1.2284195184037114,
+      "grad_norm": 0.3571791073570016,
+      "learning_rate": 7.797897871045332e-06,
+      "loss": 0.4705,
+      "step": 7480
+    },
+    {
+      "epoch": 1.2285837456120543,
+      "grad_norm": 0.34360557571526473,
+      "learning_rate": 7.797578777210096e-06,
+      "loss": 0.4663,
+      "step": 7481
+    },
+    {
+      "epoch": 1.228747972820397,
+      "grad_norm": 0.2963914639094682,
+      "learning_rate": 7.797259647559912e-06,
+      "loss": 0.4723,
+      "step": 7482
+    },
+    {
+      "epoch": 1.2289122000287398,
+      "grad_norm": 0.27699861933825076,
+      "learning_rate": 7.796940482098244e-06,
+      "loss": 0.4731,
+      "step": 7483
+    },
+    {
+      "epoch": 1.2290764272370824,
+      "grad_norm": 0.2969103632510052,
+      "learning_rate": 7.796621280828558e-06,
+      "loss": 0.4761,
+      "step": 7484
+    },
+    {
+      "epoch": 1.2292406544454253,
+      "grad_norm": 0.3065597940547537,
+      "learning_rate": 7.796302043754321e-06,
+      "loss": 0.4505,
+      "step": 7485
+    },
+    {
+      "epoch": 1.229404881653768,
+      "grad_norm": 0.26417503495718664,
+      "learning_rate": 7.795982770879e-06,
+      "loss": 0.4748,
+      "step": 7486
+    },
+    {
+      "epoch": 1.2295691088621108,
+      "grad_norm": 0.3398290499434709,
+      "learning_rate": 7.795663462206067e-06,
+      "loss": 0.4669,
+      "step": 7487
+    },
+    {
+      "epoch": 1.2297333360704534,
+      "grad_norm": 0.34742823434001574,
+      "learning_rate": 7.795344117738982e-06,
+      "loss": 0.4781,
+      "step": 7488
+    },
+    {
+      "epoch": 1.2298975632787963,
+      "grad_norm": 0.2945219892926102,
+      "learning_rate": 7.795024737481219e-06,
+      "loss": 0.4687,
+      "step": 7489
+    },
+    {
+      "epoch": 1.230061790487139,
+      "grad_norm": 0.2770494663562303,
+      "learning_rate": 7.794705321436248e-06,
+      "loss": 0.4587,
+      "step": 7490
+    },
+    {
+      "epoch": 1.2302260176954818,
+      "grad_norm": 0.2968149118488759,
+      "learning_rate": 7.794385869607532e-06,
+      "loss": 0.511,
+      "step": 7491
+    },
+    {
+      "epoch": 1.2303902449038244,
+      "grad_norm": 0.2915104293412883,
+      "learning_rate": 7.794066381998546e-06,
+      "loss": 0.4689,
+      "step": 7492
+    },
+    {
+      "epoch": 1.2305544721121673,
+      "grad_norm": 0.35989630798640004,
+      "learning_rate": 7.793746858612759e-06,
+      "loss": 0.4944,
+      "step": 7493
+    },
+    {
+      "epoch": 1.23071869932051,
+      "grad_norm": 0.3141717355163613,
+      "learning_rate": 7.79342729945364e-06,
+      "loss": 0.4933,
+      "step": 7494
+    },
+    {
+      "epoch": 1.2308829265288526,
+      "grad_norm": 0.3191568033697543,
+      "learning_rate": 7.793107704524659e-06,
+      "loss": 0.4767,
+      "step": 7495
+    },
+    {
+      "epoch": 1.2310471537371954,
+      "grad_norm": 0.3239132079874789,
+      "learning_rate": 7.792788073829289e-06,
+      "loss": 0.4672,
+      "step": 7496
+    },
+    {
+      "epoch": 1.231211380945538,
+      "grad_norm": 0.2635865983108721,
+      "learning_rate": 7.792468407371e-06,
+      "loss": 0.4478,
+      "step": 7497
+    },
+    {
+      "epoch": 1.231375608153881,
+      "grad_norm": 0.38292422598333736,
+      "learning_rate": 7.792148705153266e-06,
+      "loss": 0.4812,
+      "step": 7498
+    },
+    {
+      "epoch": 1.2315398353622236,
+      "grad_norm": 0.37145374460876307,
+      "learning_rate": 7.791828967179559e-06,
+      "loss": 0.4876,
+      "step": 7499
+    },
+    {
+      "epoch": 1.2317040625705664,
+      "grad_norm": 0.28915250611709403,
+      "learning_rate": 7.791509193453348e-06,
+      "loss": 0.4792,
+      "step": 7500
+    },
+    {
+      "epoch": 1.231868289778909,
+      "grad_norm": 0.2889110278668834,
+      "learning_rate": 7.79118938397811e-06,
+      "loss": 0.4905,
+      "step": 7501
+    },
+    {
+      "epoch": 1.232032516987252,
+      "grad_norm": 0.33366376283958143,
+      "learning_rate": 7.790869538757317e-06,
+      "loss": 0.5038,
+      "step": 7502
+    },
+    {
+      "epoch": 1.2321967441955946,
+      "grad_norm": 0.44289224941063765,
+      "learning_rate": 7.790549657794443e-06,
+      "loss": 0.4876,
+      "step": 7503
+    },
+    {
+      "epoch": 1.2323609714039374,
+      "grad_norm": 0.3136464858676508,
+      "learning_rate": 7.790229741092962e-06,
+      "loss": 0.4955,
+      "step": 7504
+    },
+    {
+      "epoch": 1.23252519861228,
+      "grad_norm": 0.2531177671121733,
+      "learning_rate": 7.78990978865635e-06,
+      "loss": 0.4962,
+      "step": 7505
+    },
+    {
+      "epoch": 1.232689425820623,
+      "grad_norm": 0.2978682970598931,
+      "learning_rate": 7.789589800488081e-06,
+      "loss": 0.5,
+      "step": 7506
+    },
+    {
+      "epoch": 1.2328536530289655,
+      "grad_norm": 0.2534835272168865,
+      "learning_rate": 7.789269776591631e-06,
+      "loss": 0.4681,
+      "step": 7507
+    },
+    {
+      "epoch": 1.2330178802373084,
+      "grad_norm": 0.35502693503125016,
+      "learning_rate": 7.788949716970472e-06,
+      "loss": 0.4868,
+      "step": 7508
+    },
+    {
+      "epoch": 1.233182107445651,
+      "grad_norm": 0.35494605281913755,
+      "learning_rate": 7.788629621628084e-06,
+      "loss": 0.4862,
+      "step": 7509
+    },
+    {
+      "epoch": 1.233346334653994,
+      "grad_norm": 0.2796519161526337,
+      "learning_rate": 7.788309490567945e-06,
+      "loss": 0.4809,
+      "step": 7510
+    },
+    {
+      "epoch": 1.2335105618623365,
+      "grad_norm": 0.35202315535595735,
+      "learning_rate": 7.787989323793527e-06,
+      "loss": 0.4927,
+      "step": 7511
+    },
+    {
+      "epoch": 1.2336747890706792,
+      "grad_norm": 0.27572529786791367,
+      "learning_rate": 7.787669121308312e-06,
+      "loss": 0.477,
+      "step": 7512
+    },
+    {
+      "epoch": 1.233839016279022,
+      "grad_norm": 0.29646983151833184,
+      "learning_rate": 7.787348883115774e-06,
+      "loss": 0.4731,
+      "step": 7513
+    },
+    {
+      "epoch": 1.2340032434873647,
+      "grad_norm": 0.32889863862531893,
+      "learning_rate": 7.787028609219394e-06,
+      "loss": 0.4951,
+      "step": 7514
+    },
+    {
+      "epoch": 1.2341674706957075,
+      "grad_norm": 0.2655153867934256,
+      "learning_rate": 7.78670829962265e-06,
+      "loss": 0.4506,
+      "step": 7515
+    },
+    {
+      "epoch": 1.2343316979040502,
+      "grad_norm": 0.2663148975867709,
+      "learning_rate": 7.786387954329018e-06,
+      "loss": 0.4742,
+      "step": 7516
+    },
+    {
+      "epoch": 1.234495925112393,
+      "grad_norm": 0.42508777196790876,
+      "learning_rate": 7.786067573341982e-06,
+      "loss": 0.4826,
+      "step": 7517
+    },
+    {
+      "epoch": 1.2346601523207357,
+      "grad_norm": 0.3041477860219649,
+      "learning_rate": 7.785747156665018e-06,
+      "loss": 0.4699,
+      "step": 7518
+    },
+    {
+      "epoch": 1.2348243795290785,
+      "grad_norm": 0.2897416257015565,
+      "learning_rate": 7.785426704301607e-06,
+      "loss": 0.4883,
+      "step": 7519
+    },
+    {
+      "epoch": 1.2349886067374212,
+      "grad_norm": 0.2720161031746429,
+      "learning_rate": 7.785106216255229e-06,
+      "loss": 0.4819,
+      "step": 7520
+    },
+    {
+      "epoch": 1.235152833945764,
+      "grad_norm": 0.3006375841709048,
+      "learning_rate": 7.784785692529365e-06,
+      "loss": 0.4817,
+      "step": 7521
+    },
+    {
+      "epoch": 1.2353170611541067,
+      "grad_norm": 0.31301572775303516,
+      "learning_rate": 7.784465133127498e-06,
+      "loss": 0.4642,
+      "step": 7522
+    },
+    {
+      "epoch": 1.2354812883624495,
+      "grad_norm": 0.3852031934863473,
+      "learning_rate": 7.784144538053108e-06,
+      "loss": 0.4686,
+      "step": 7523
+    },
+    {
+      "epoch": 1.2356455155707922,
+      "grad_norm": 0.2872787747457896,
+      "learning_rate": 7.783823907309676e-06,
+      "loss": 0.4778,
+      "step": 7524
+    },
+    {
+      "epoch": 1.235809742779135,
+      "grad_norm": 0.2864937598952176,
+      "learning_rate": 7.783503240900686e-06,
+      "loss": 0.4869,
+      "step": 7525
+    },
+    {
+      "epoch": 1.2359739699874777,
+      "grad_norm": 0.3873222568675411,
+      "learning_rate": 7.78318253882962e-06,
+      "loss": 0.4727,
+      "step": 7526
+    },
+    {
+      "epoch": 1.2361381971958205,
+      "grad_norm": 0.32706793356824965,
+      "learning_rate": 7.782861801099963e-06,
+      "loss": 0.4931,
+      "step": 7527
+    },
+    {
+      "epoch": 1.2363024244041632,
+      "grad_norm": 0.24977174505164065,
+      "learning_rate": 7.782541027715195e-06,
+      "loss": 0.4436,
+      "step": 7528
+    },
+    {
+      "epoch": 1.2364666516125058,
+      "grad_norm": 0.31267245459003645,
+      "learning_rate": 7.782220218678804e-06,
+      "loss": 0.4731,
+      "step": 7529
+    },
+    {
+      "epoch": 1.2366308788208487,
+      "grad_norm": 0.33351244334836483,
+      "learning_rate": 7.78189937399427e-06,
+      "loss": 0.4922,
+      "step": 7530
+    },
+    {
+      "epoch": 1.2367951060291913,
+      "grad_norm": 0.3155971525806248,
+      "learning_rate": 7.781578493665083e-06,
+      "loss": 0.5006,
+      "step": 7531
+    },
+    {
+      "epoch": 1.2369593332375342,
+      "grad_norm": 0.3519882716741187,
+      "learning_rate": 7.78125757769472e-06,
+      "loss": 0.4795,
+      "step": 7532
+    },
+    {
+      "epoch": 1.2371235604458768,
+      "grad_norm": 0.2979601040823767,
+      "learning_rate": 7.780936626086675e-06,
+      "loss": 0.4766,
+      "step": 7533
+    },
+    {
+      "epoch": 1.2372877876542197,
+      "grad_norm": 0.3063945709411937,
+      "learning_rate": 7.78061563884443e-06,
+      "loss": 0.4789,
+      "step": 7534
+    },
+    {
+      "epoch": 1.2374520148625623,
+      "grad_norm": 0.26722072179018536,
+      "learning_rate": 7.780294615971471e-06,
+      "loss": 0.4838,
+      "step": 7535
+    },
+    {
+      "epoch": 1.2376162420709051,
+      "grad_norm": 0.292009730274553,
+      "learning_rate": 7.779973557471285e-06,
+      "loss": 0.4987,
+      "step": 7536
+    },
+    {
+      "epoch": 1.2377804692792478,
+      "grad_norm": 0.2807020434939495,
+      "learning_rate": 7.77965246334736e-06,
+      "loss": 0.4715,
+      "step": 7537
+    },
+    {
+      "epoch": 1.2379446964875906,
+      "grad_norm": 0.3453670092832338,
+      "learning_rate": 7.77933133360318e-06,
+      "loss": 0.4902,
+      "step": 7538
+    },
+    {
+      "epoch": 1.2381089236959333,
+      "grad_norm": 0.31388854164530383,
+      "learning_rate": 7.779010168242236e-06,
+      "loss": 0.4856,
+      "step": 7539
+    },
+    {
+      "epoch": 1.2382731509042761,
+      "grad_norm": 0.2867020260405545,
+      "learning_rate": 7.778688967268017e-06,
+      "loss": 0.4847,
+      "step": 7540
+    },
+    {
+      "epoch": 1.2384373781126188,
+      "grad_norm": 0.3438963236599197,
+      "learning_rate": 7.77836773068401e-06,
+      "loss": 0.5006,
+      "step": 7541
+    },
+    {
+      "epoch": 1.2386016053209616,
+      "grad_norm": 0.4297254179891695,
+      "learning_rate": 7.778046458493703e-06,
+      "loss": 0.4747,
+      "step": 7542
+    },
+    {
+      "epoch": 1.2387658325293043,
+      "grad_norm": 0.3582137581158092,
+      "learning_rate": 7.777725150700587e-06,
+      "loss": 0.4832,
+      "step": 7543
+    },
+    {
+      "epoch": 1.2389300597376471,
+      "grad_norm": 0.29072289657454703,
+      "learning_rate": 7.777403807308148e-06,
+      "loss": 0.4966,
+      "step": 7544
+    },
+    {
+      "epoch": 1.2390942869459898,
+      "grad_norm": 0.3252671784530234,
+      "learning_rate": 7.777082428319884e-06,
+      "loss": 0.4866,
+      "step": 7545
+    },
+    {
+      "epoch": 1.2392585141543324,
+      "grad_norm": 0.2908401782457425,
+      "learning_rate": 7.776761013739277e-06,
+      "loss": 0.4755,
+      "step": 7546
+    },
+    {
+      "epoch": 1.2394227413626753,
+      "grad_norm": 0.28684276655151986,
+      "learning_rate": 7.776439563569825e-06,
+      "loss": 0.4761,
+      "step": 7547
+    },
+    {
+      "epoch": 1.239586968571018,
+      "grad_norm": 0.30531377495350137,
+      "learning_rate": 7.776118077815012e-06,
+      "loss": 0.4831,
+      "step": 7548
+    },
+    {
+      "epoch": 1.2397511957793608,
+      "grad_norm": 0.29723211564359786,
+      "learning_rate": 7.775796556478336e-06,
+      "loss": 0.4801,
+      "step": 7549
+    },
+    {
+      "epoch": 1.2399154229877034,
+      "grad_norm": 0.2936682730249869,
+      "learning_rate": 7.775474999563285e-06,
+      "loss": 0.4832,
+      "step": 7550
+    },
+    {
+      "epoch": 1.2400796501960463,
+      "grad_norm": 0.3018141932268422,
+      "learning_rate": 7.775153407073353e-06,
+      "loss": 0.4906,
+      "step": 7551
+    },
+    {
+      "epoch": 1.240243877404389,
+      "grad_norm": 0.2787629798347041,
+      "learning_rate": 7.774831779012033e-06,
+      "loss": 0.4896,
+      "step": 7552
+    },
+    {
+      "epoch": 1.2404081046127318,
+      "grad_norm": 0.34481809463993834,
+      "learning_rate": 7.774510115382818e-06,
+      "loss": 0.4659,
+      "step": 7553
+    },
+    {
+      "epoch": 1.2405723318210744,
+      "grad_norm": 0.2779282824687869,
+      "learning_rate": 7.774188416189201e-06,
+      "loss": 0.4546,
+      "step": 7554
+    },
+    {
+      "epoch": 1.2407365590294173,
+      "grad_norm": 0.2786407520838788,
+      "learning_rate": 7.773866681434676e-06,
+      "loss": 0.4798,
+      "step": 7555
+    },
+    {
+      "epoch": 1.24090078623776,
+      "grad_norm": 0.3285343236852004,
+      "learning_rate": 7.77354491112274e-06,
+      "loss": 0.4797,
+      "step": 7556
+    },
+    {
+      "epoch": 1.2410650134461028,
+      "grad_norm": 0.37506765944836495,
+      "learning_rate": 7.773223105256883e-06,
+      "loss": 0.4845,
+      "step": 7557
+    },
+    {
+      "epoch": 1.2412292406544454,
+      "grad_norm": 0.35685647641502605,
+      "learning_rate": 7.772901263840605e-06,
+      "loss": 0.4877,
+      "step": 7558
+    },
+    {
+      "epoch": 1.2413934678627883,
+      "grad_norm": 0.3358896496566247,
+      "learning_rate": 7.772579386877396e-06,
+      "loss": 0.49,
+      "step": 7559
+    },
+    {
+      "epoch": 1.241557695071131,
+      "grad_norm": 0.3514289943534612,
+      "learning_rate": 7.772257474370757e-06,
+      "loss": 0.4744,
+      "step": 7560
+    },
+    {
+      "epoch": 1.2417219222794738,
+      "grad_norm": 0.2937933570234482,
+      "learning_rate": 7.771935526324183e-06,
+      "loss": 0.4843,
+      "step": 7561
+    },
+    {
+      "epoch": 1.2418861494878164,
+      "grad_norm": 0.3393020982420418,
+      "learning_rate": 7.77161354274117e-06,
+      "loss": 0.4573,
+      "step": 7562
+    },
+    {
+      "epoch": 1.242050376696159,
+      "grad_norm": 0.3228901072188093,
+      "learning_rate": 7.771291523625214e-06,
+      "loss": 0.5017,
+      "step": 7563
+    },
+    {
+      "epoch": 1.242214603904502,
+      "grad_norm": 0.3065450686112153,
+      "learning_rate": 7.770969468979814e-06,
+      "loss": 0.4758,
+      "step": 7564
+    },
+    {
+      "epoch": 1.2423788311128445,
+      "grad_norm": 0.3106520824315073,
+      "learning_rate": 7.770647378808469e-06,
+      "loss": 0.4759,
+      "step": 7565
+    },
+    {
+      "epoch": 1.2425430583211874,
+      "grad_norm": 0.2937174607831136,
+      "learning_rate": 7.770325253114674e-06,
+      "loss": 0.4622,
+      "step": 7566
+    },
+    {
+      "epoch": 1.24270728552953,
+      "grad_norm": 0.3386310405929065,
+      "learning_rate": 7.770003091901928e-06,
+      "loss": 0.4851,
+      "step": 7567
+    },
+    {
+      "epoch": 1.2428715127378729,
+      "grad_norm": 0.34444528064525604,
+      "learning_rate": 7.769680895173733e-06,
+      "loss": 0.4878,
+      "step": 7568
+    },
+    {
+      "epoch": 1.2430357399462155,
+      "grad_norm": 0.27978430576428204,
+      "learning_rate": 7.769358662933584e-06,
+      "loss": 0.4847,
+      "step": 7569
+    },
+    {
+      "epoch": 1.2431999671545584,
+      "grad_norm": 0.31389488030986795,
+      "learning_rate": 7.769036395184987e-06,
+      "loss": 0.4745,
+      "step": 7570
+    },
+    {
+      "epoch": 1.243364194362901,
+      "grad_norm": 0.27170554485549553,
+      "learning_rate": 7.768714091931436e-06,
+      "loss": 0.4794,
+      "step": 7571
+    },
+    {
+      "epoch": 1.2435284215712439,
+      "grad_norm": 0.33136051639556463,
+      "learning_rate": 7.768391753176434e-06,
+      "loss": 0.4644,
+      "step": 7572
+    },
+    {
+      "epoch": 1.2436926487795865,
+      "grad_norm": 0.3093556203425737,
+      "learning_rate": 7.768069378923483e-06,
+      "loss": 0.4646,
+      "step": 7573
+    },
+    {
+      "epoch": 1.2438568759879294,
+      "grad_norm": 0.46973993506400613,
+      "learning_rate": 7.767746969176082e-06,
+      "loss": 0.4632,
+      "step": 7574
+    },
+    {
+      "epoch": 1.244021103196272,
+      "grad_norm": 0.26908039743071843,
+      "learning_rate": 7.767424523937735e-06,
+      "loss": 0.4705,
+      "step": 7575
+    },
+    {
+      "epoch": 1.2441853304046149,
+      "grad_norm": 0.37290345238525935,
+      "learning_rate": 7.767102043211942e-06,
+      "loss": 0.4944,
+      "step": 7576
+    },
+    {
+      "epoch": 1.2443495576129575,
+      "grad_norm": 0.43611265222170165,
+      "learning_rate": 7.766779527002208e-06,
+      "loss": 0.4909,
+      "step": 7577
+    },
+    {
+      "epoch": 1.2445137848213004,
+      "grad_norm": 0.3328354020728991,
+      "learning_rate": 7.766456975312032e-06,
+      "loss": 0.4696,
+      "step": 7578
+    },
+    {
+      "epoch": 1.244678012029643,
+      "grad_norm": 0.25915177451173665,
+      "learning_rate": 7.766134388144921e-06,
+      "loss": 0.4875,
+      "step": 7579
+    },
+    {
+      "epoch": 1.2448422392379856,
+      "grad_norm": 0.25832700528645586,
+      "learning_rate": 7.765811765504377e-06,
+      "loss": 0.4837,
+      "step": 7580
+    },
+    {
+      "epoch": 1.2450064664463285,
+      "grad_norm": 0.30829434869947336,
+      "learning_rate": 7.765489107393903e-06,
+      "loss": 0.4748,
+      "step": 7581
+    },
+    {
+      "epoch": 1.2451706936546711,
+      "grad_norm": 0.30836619497139556,
+      "learning_rate": 7.765166413817006e-06,
+      "loss": 0.486,
+      "step": 7582
+    },
+    {
+      "epoch": 1.245334920863014,
+      "grad_norm": 0.6948548750504677,
+      "learning_rate": 7.76484368477719e-06,
+      "loss": 0.4691,
+      "step": 7583
+    },
+    {
+      "epoch": 1.2454991480713566,
+      "grad_norm": 0.2700607921883031,
+      "learning_rate": 7.764520920277958e-06,
+      "loss": 0.473,
+      "step": 7584
+    },
+    {
+      "epoch": 1.2456633752796995,
+      "grad_norm": 0.3288894558422325,
+      "learning_rate": 7.764198120322816e-06,
+      "loss": 0.4831,
+      "step": 7585
+    },
+    {
+      "epoch": 1.2458276024880421,
+      "grad_norm": 0.3254518062997147,
+      "learning_rate": 7.763875284915272e-06,
+      "loss": 0.4723,
+      "step": 7586
+    },
+    {
+      "epoch": 1.245991829696385,
+      "grad_norm": 0.3169491475419649,
+      "learning_rate": 7.76355241405883e-06,
+      "loss": 0.5002,
+      "step": 7587
+    },
+    {
+      "epoch": 1.2461560569047276,
+      "grad_norm": 0.2896778965883232,
+      "learning_rate": 7.763229507757e-06,
+      "loss": 0.4784,
+      "step": 7588
+    },
+    {
+      "epoch": 1.2463202841130705,
+      "grad_norm": 0.6643183902372932,
+      "learning_rate": 7.762906566013287e-06,
+      "loss": 0.4641,
+      "step": 7589
+    },
+    {
+      "epoch": 1.2464845113214131,
+      "grad_norm": 0.4694379861367071,
+      "learning_rate": 7.762583588831197e-06,
+      "loss": 0.4951,
+      "step": 7590
+    },
+    {
+      "epoch": 1.246648738529756,
+      "grad_norm": 0.2714033681351259,
+      "learning_rate": 7.76226057621424e-06,
+      "loss": 0.4512,
+      "step": 7591
+    },
+    {
+      "epoch": 1.2468129657380986,
+      "grad_norm": 0.33738219660603047,
+      "learning_rate": 7.761937528165923e-06,
+      "loss": 0.4711,
+      "step": 7592
+    },
+    {
+      "epoch": 1.2469771929464415,
+      "grad_norm": 0.31785803759257947,
+      "learning_rate": 7.761614444689755e-06,
+      "loss": 0.4799,
+      "step": 7593
+    },
+    {
+      "epoch": 1.2471414201547841,
+      "grad_norm": 0.2944630094956895,
+      "learning_rate": 7.761291325789244e-06,
+      "loss": 0.4857,
+      "step": 7594
+    },
+    {
+      "epoch": 1.247305647363127,
+      "grad_norm": 0.2539993541834688,
+      "learning_rate": 7.760968171467903e-06,
+      "loss": 0.4545,
+      "step": 7595
+    },
+    {
+      "epoch": 1.2474698745714696,
+      "grad_norm": 0.31121532100643007,
+      "learning_rate": 7.760644981729238e-06,
+      "loss": 0.4742,
+      "step": 7596
+    },
+    {
+      "epoch": 1.2476341017798123,
+      "grad_norm": 0.41726789416464144,
+      "learning_rate": 7.76032175657676e-06,
+      "loss": 0.4569,
+      "step": 7597
+    },
+    {
+      "epoch": 1.2477983289881551,
+      "grad_norm": 0.2858322392854143,
+      "learning_rate": 7.759998496013981e-06,
+      "loss": 0.4628,
+      "step": 7598
+    },
+    {
+      "epoch": 1.2479625561964978,
+      "grad_norm": 0.29840578768448217,
+      "learning_rate": 7.759675200044411e-06,
+      "loss": 0.4865,
+      "step": 7599
+    },
+    {
+      "epoch": 1.2481267834048406,
+      "grad_norm": 0.429100208558146,
+      "learning_rate": 7.75935186867156e-06,
+      "loss": 0.4707,
+      "step": 7600
+    },
+    {
+      "epoch": 1.2482910106131833,
+      "grad_norm": 0.2838003969767729,
+      "learning_rate": 7.759028501898942e-06,
+      "loss": 0.4709,
+      "step": 7601
+    },
+    {
+      "epoch": 1.2484552378215261,
+      "grad_norm": 0.3557099389351078,
+      "learning_rate": 7.758705099730069e-06,
+      "loss": 0.4723,
+      "step": 7602
+    },
+    {
+      "epoch": 1.2486194650298688,
+      "grad_norm": 0.2870847693915563,
+      "learning_rate": 7.75838166216845e-06,
+      "loss": 0.4866,
+      "step": 7603
+    },
+    {
+      "epoch": 1.2487836922382116,
+      "grad_norm": 0.3188591937844555,
+      "learning_rate": 7.758058189217604e-06,
+      "loss": 0.4486,
+      "step": 7604
+    },
+    {
+      "epoch": 1.2489479194465543,
+      "grad_norm": 0.37056779120212635,
+      "learning_rate": 7.757734680881036e-06,
+      "loss": 0.4855,
+      "step": 7605
+    },
+    {
+      "epoch": 1.2491121466548971,
+      "grad_norm": 0.3639198034440745,
+      "learning_rate": 7.757411137162267e-06,
+      "loss": 0.4747,
+      "step": 7606
+    },
+    {
+      "epoch": 1.2492763738632398,
+      "grad_norm": 0.2787655583803941,
+      "learning_rate": 7.757087558064806e-06,
+      "loss": 0.4783,
+      "step": 7607
+    },
+    {
+      "epoch": 1.2494406010715826,
+      "grad_norm": 0.30779539880108575,
+      "learning_rate": 7.756763943592173e-06,
+      "loss": 0.4965,
+      "step": 7608
+    },
+    {
+      "epoch": 1.2496048282799253,
+      "grad_norm": 0.31669536490305483,
+      "learning_rate": 7.756440293747877e-06,
+      "loss": 0.4912,
+      "step": 7609
+    },
+    {
+      "epoch": 1.249769055488268,
+      "grad_norm": 0.308687391665335,
+      "learning_rate": 7.756116608535436e-06,
+      "loss": 0.4889,
+      "step": 7610
+    },
+    {
+      "epoch": 1.2499332826966107,
+      "grad_norm": 0.38901222664933416,
+      "learning_rate": 7.755792887958365e-06,
+      "loss": 0.4977,
+      "step": 7611
+    },
+    {
+      "epoch": 1.2500975099049536,
+      "grad_norm": 0.28324939107336244,
+      "learning_rate": 7.75546913202018e-06,
+      "loss": 0.4828,
+      "step": 7612
+    },
+    {
+      "epoch": 1.2502617371132962,
+      "grad_norm": 0.4111706793850341,
+      "learning_rate": 7.755145340724396e-06,
+      "loss": 0.4756,
+      "step": 7613
+    },
+    {
+      "epoch": 1.2504259643216389,
+      "grad_norm": 0.3162486727419526,
+      "learning_rate": 7.754821514074534e-06,
+      "loss": 0.4869,
+      "step": 7614
+    },
+    {
+      "epoch": 1.2505901915299817,
+      "grad_norm": 0.30369645916751725,
+      "learning_rate": 7.754497652074106e-06,
+      "loss": 0.498,
+      "step": 7615
+    },
+    {
+      "epoch": 1.2507544187383246,
+      "grad_norm": 0.2973675199028055,
+      "learning_rate": 7.754173754726631e-06,
+      "loss": 0.4724,
+      "step": 7616
+    },
+    {
+      "epoch": 1.2509186459466672,
+      "grad_norm": 0.3635026371907883,
+      "learning_rate": 7.75384982203563e-06,
+      "loss": 0.4735,
+      "step": 7617
+    },
+    {
+      "epoch": 1.2510828731550099,
+      "grad_norm": 0.3335327209309586,
+      "learning_rate": 7.753525854004618e-06,
+      "loss": 0.4845,
+      "step": 7618
+    },
+    {
+      "epoch": 1.2512471003633527,
+      "grad_norm": 0.27222796938938104,
+      "learning_rate": 7.753201850637111e-06,
+      "loss": 0.4745,
+      "step": 7619
+    },
+    {
+      "epoch": 1.2514113275716954,
+      "grad_norm": 0.25174104281048826,
+      "learning_rate": 7.752877811936634e-06,
+      "loss": 0.4695,
+      "step": 7620
+    },
+    {
+      "epoch": 1.2515755547800382,
+      "grad_norm": 0.37642279840391096,
+      "learning_rate": 7.752553737906702e-06,
+      "loss": 0.4846,
+      "step": 7621
+    },
+    {
+      "epoch": 1.2517397819883809,
+      "grad_norm": 0.28660532539995426,
+      "learning_rate": 7.752229628550837e-06,
+      "loss": 0.4891,
+      "step": 7622
+    },
+    {
+      "epoch": 1.2519040091967237,
+      "grad_norm": 0.5222307981778282,
+      "learning_rate": 7.75190548387256e-06,
+      "loss": 0.4841,
+      "step": 7623
+    },
+    {
+      "epoch": 1.2520682364050664,
+      "grad_norm": 0.2689431118979005,
+      "learning_rate": 7.751581303875387e-06,
+      "loss": 0.4654,
+      "step": 7624
+    },
+    {
+      "epoch": 1.2522324636134092,
+      "grad_norm": 0.26136425839250754,
+      "learning_rate": 7.751257088562843e-06,
+      "loss": 0.4892,
+      "step": 7625
+    },
+    {
+      "epoch": 1.2523966908217519,
+      "grad_norm": 0.3392486862232706,
+      "learning_rate": 7.75093283793845e-06,
+      "loss": 0.4785,
+      "step": 7626
+    },
+    {
+      "epoch": 1.2525609180300945,
+      "grad_norm": 0.28026937023228876,
+      "learning_rate": 7.750608552005726e-06,
+      "loss": 0.4753,
+      "step": 7627
+    },
+    {
+      "epoch": 1.2527251452384374,
+      "grad_norm": 0.3107954217489059,
+      "learning_rate": 7.750284230768194e-06,
+      "loss": 0.4883,
+      "step": 7628
+    },
+    {
+      "epoch": 1.2528893724467802,
+      "grad_norm": 0.3049446665147886,
+      "learning_rate": 7.749959874229381e-06,
+      "loss": 0.4796,
+      "step": 7629
+    },
+    {
+      "epoch": 1.2530535996551229,
+      "grad_norm": 0.3168276384106462,
+      "learning_rate": 7.749635482392802e-06,
+      "loss": 0.4955,
+      "step": 7630
+    },
+    {
+      "epoch": 1.2532178268634655,
+      "grad_norm": 0.3235340299163566,
+      "learning_rate": 7.749311055261989e-06,
+      "loss": 0.481,
+      "step": 7631
+    },
+    {
+      "epoch": 1.2533820540718084,
+      "grad_norm": 0.39371662529450346,
+      "learning_rate": 7.748986592840457e-06,
+      "loss": 0.4861,
+      "step": 7632
+    },
+    {
+      "epoch": 1.2535462812801512,
+      "grad_norm": 0.285427242350543,
+      "learning_rate": 7.748662095131736e-06,
+      "loss": 0.498,
+      "step": 7633
+    },
+    {
+      "epoch": 1.2537105084884939,
+      "grad_norm": 0.33650249587902675,
+      "learning_rate": 7.748337562139348e-06,
+      "loss": 0.4865,
+      "step": 7634
+    },
+    {
+      "epoch": 1.2538747356968365,
+      "grad_norm": 0.3027523754521188,
+      "learning_rate": 7.748012993866817e-06,
+      "loss": 0.4752,
+      "step": 7635
+    },
+    {
+      "epoch": 1.2540389629051794,
+      "grad_norm": 0.2806626536948512,
+      "learning_rate": 7.747688390317672e-06,
+      "loss": 0.4701,
+      "step": 7636
+    },
+    {
+      "epoch": 1.254203190113522,
+      "grad_norm": 0.2646422086313468,
+      "learning_rate": 7.747363751495434e-06,
+      "loss": 0.4891,
+      "step": 7637
+    },
+    {
+      "epoch": 1.2543674173218649,
+      "grad_norm": 0.273441689961552,
+      "learning_rate": 7.747039077403631e-06,
+      "loss": 0.4606,
+      "step": 7638
+    },
+    {
+      "epoch": 1.2545316445302075,
+      "grad_norm": 0.39997955334967183,
+      "learning_rate": 7.746714368045788e-06,
+      "loss": 0.4665,
+      "step": 7639
+    },
+    {
+      "epoch": 1.2546958717385504,
+      "grad_norm": 0.3156118656661959,
+      "learning_rate": 7.746389623425435e-06,
+      "loss": 0.4848,
+      "step": 7640
+    },
+    {
+      "epoch": 1.254860098946893,
+      "grad_norm": 0.2993686116712013,
+      "learning_rate": 7.746064843546096e-06,
+      "loss": 0.4766,
+      "step": 7641
+    },
+    {
+      "epoch": 1.2550243261552358,
+      "grad_norm": 0.35049350059175216,
+      "learning_rate": 7.745740028411296e-06,
+      "loss": 0.4811,
+      "step": 7642
+    },
+    {
+      "epoch": 1.2551885533635785,
+      "grad_norm": 0.33841545284685837,
+      "learning_rate": 7.74541517802457e-06,
+      "loss": 0.4875,
+      "step": 7643
+    },
+    {
+      "epoch": 1.2553527805719211,
+      "grad_norm": 0.29653348815931907,
+      "learning_rate": 7.745090292389438e-06,
+      "loss": 0.4834,
+      "step": 7644
+    },
+    {
+      "epoch": 1.255517007780264,
+      "grad_norm": 1.0619524130211613,
+      "learning_rate": 7.744765371509437e-06,
+      "loss": 0.4715,
+      "step": 7645
+    },
+    {
+      "epoch": 1.2556812349886068,
+      "grad_norm": 0.33283296033041115,
+      "learning_rate": 7.744440415388089e-06,
+      "loss": 0.4661,
+      "step": 7646
+    },
+    {
+      "epoch": 1.2558454621969495,
+      "grad_norm": 0.3195752649112358,
+      "learning_rate": 7.744115424028925e-06,
+      "loss": 0.4981,
+      "step": 7647
+    },
+    {
+      "epoch": 1.2560096894052921,
+      "grad_norm": 0.2546867337134892,
+      "learning_rate": 7.74379039743548e-06,
+      "loss": 0.4771,
+      "step": 7648
+    },
+    {
+      "epoch": 1.256173916613635,
+      "grad_norm": 0.30738186122930333,
+      "learning_rate": 7.743465335611276e-06,
+      "loss": 0.4666,
+      "step": 7649
+    },
+    {
+      "epoch": 1.2563381438219778,
+      "grad_norm": 0.27859639311481776,
+      "learning_rate": 7.74314023855985e-06,
+      "loss": 0.4712,
+      "step": 7650
+    },
+    {
+      "epoch": 1.2565023710303205,
+      "grad_norm": 0.3330781418753308,
+      "learning_rate": 7.742815106284728e-06,
+      "loss": 0.4659,
+      "step": 7651
+    },
+    {
+      "epoch": 1.2566665982386631,
+      "grad_norm": 0.32030910397101686,
+      "learning_rate": 7.742489938789444e-06,
+      "loss": 0.4489,
+      "step": 7652
+    },
+    {
+      "epoch": 1.256830825447006,
+      "grad_norm": 0.358633123705577,
+      "learning_rate": 7.74216473607753e-06,
+      "loss": 0.455,
+      "step": 7653
+    },
+    {
+      "epoch": 1.2569950526553486,
+      "grad_norm": 0.28410974154650376,
+      "learning_rate": 7.741839498152515e-06,
+      "loss": 0.4801,
+      "step": 7654
+    },
+    {
+      "epoch": 1.2571592798636915,
+      "grad_norm": 0.34427076281168334,
+      "learning_rate": 7.741514225017935e-06,
+      "loss": 0.487,
+      "step": 7655
+    },
+    {
+      "epoch": 1.257323507072034,
+      "grad_norm": 0.31974128073632185,
+      "learning_rate": 7.741188916677321e-06,
+      "loss": 0.4781,
+      "step": 7656
+    },
+    {
+      "epoch": 1.257487734280377,
+      "grad_norm": 0.2841276269841283,
+      "learning_rate": 7.740863573134208e-06,
+      "loss": 0.4656,
+      "step": 7657
+    },
+    {
+      "epoch": 1.2576519614887196,
+      "grad_norm": 0.29801297035583985,
+      "learning_rate": 7.740538194392126e-06,
+      "loss": 0.492,
+      "step": 7658
+    },
+    {
+      "epoch": 1.2578161886970625,
+      "grad_norm": 0.40026325549229824,
+      "learning_rate": 7.740212780454611e-06,
+      "loss": 0.4843,
+      "step": 7659
+    },
+    {
+      "epoch": 1.257980415905405,
+      "grad_norm": 0.3383340220937869,
+      "learning_rate": 7.739887331325199e-06,
+      "loss": 0.4584,
+      "step": 7660
+    },
+    {
+      "epoch": 1.2581446431137477,
+      "grad_norm": 0.3879046565559537,
+      "learning_rate": 7.73956184700742e-06,
+      "loss": 0.4712,
+      "step": 7661
+    },
+    {
+      "epoch": 1.2583088703220906,
+      "grad_norm": 0.3046193115336133,
+      "learning_rate": 7.739236327504814e-06,
+      "loss": 0.4757,
+      "step": 7662
+    },
+    {
+      "epoch": 1.2584730975304335,
+      "grad_norm": 0.27583639779518276,
+      "learning_rate": 7.738910772820915e-06,
+      "loss": 0.4625,
+      "step": 7663
+    },
+    {
+      "epoch": 1.258637324738776,
+      "grad_norm": 0.2743611344103936,
+      "learning_rate": 7.738585182959257e-06,
+      "loss": 0.4819,
+      "step": 7664
+    },
+    {
+      "epoch": 1.2588015519471187,
+      "grad_norm": 0.26396529926757517,
+      "learning_rate": 7.73825955792338e-06,
+      "loss": 0.4645,
+      "step": 7665
+    },
+    {
+      "epoch": 1.2589657791554616,
+      "grad_norm": 0.33484568225629724,
+      "learning_rate": 7.737933897716815e-06,
+      "loss": 0.483,
+      "step": 7666
+    },
+    {
+      "epoch": 1.2591300063638045,
+      "grad_norm": 0.2855404254518291,
+      "learning_rate": 7.737608202343104e-06,
+      "loss": 0.4797,
+      "step": 7667
+    },
+    {
+      "epoch": 1.259294233572147,
+      "grad_norm": 0.2678897122032849,
+      "learning_rate": 7.737282471805782e-06,
+      "loss": 0.4677,
+      "step": 7668
+    },
+    {
+      "epoch": 1.2594584607804897,
+      "grad_norm": 0.33378647809677103,
+      "learning_rate": 7.736956706108388e-06,
+      "loss": 0.4698,
+      "step": 7669
+    },
+    {
+      "epoch": 1.2596226879888326,
+      "grad_norm": 0.33762949882200577,
+      "learning_rate": 7.736630905254458e-06,
+      "loss": 0.4698,
+      "step": 7670
+    },
+    {
+      "epoch": 1.2597869151971752,
+      "grad_norm": 0.28816229123616627,
+      "learning_rate": 7.736305069247535e-06,
+      "loss": 0.4851,
+      "step": 7671
+    },
+    {
+      "epoch": 1.259951142405518,
+      "grad_norm": 0.3474911045704335,
+      "learning_rate": 7.735979198091151e-06,
+      "loss": 0.4729,
+      "step": 7672
+    },
+    {
+      "epoch": 1.2601153696138607,
+      "grad_norm": 0.31218960292473513,
+      "learning_rate": 7.735653291788851e-06,
+      "loss": 0.4663,
+      "step": 7673
+    },
+    {
+      "epoch": 1.2602795968222036,
+      "grad_norm": 0.2741703738720939,
+      "learning_rate": 7.735327350344173e-06,
+      "loss": 0.4788,
+      "step": 7674
+    },
+    {
+      "epoch": 1.2604438240305462,
+      "grad_norm": 0.3220346800939815,
+      "learning_rate": 7.735001373760658e-06,
+      "loss": 0.4785,
+      "step": 7675
+    },
+    {
+      "epoch": 1.260608051238889,
+      "grad_norm": 0.2687168544207819,
+      "learning_rate": 7.734675362041843e-06,
+      "loss": 0.4707,
+      "step": 7676
+    },
+    {
+      "epoch": 1.2607722784472317,
+      "grad_norm": 0.2882314916411854,
+      "learning_rate": 7.734349315191272e-06,
+      "loss": 0.4771,
+      "step": 7677
+    },
+    {
+      "epoch": 1.2609365056555744,
+      "grad_norm": 0.27978276123566304,
+      "learning_rate": 7.734023233212484e-06,
+      "loss": 0.4883,
+      "step": 7678
+    },
+    {
+      "epoch": 1.2611007328639172,
+      "grad_norm": 0.4947967595175015,
+      "learning_rate": 7.733697116109024e-06,
+      "loss": 0.4965,
+      "step": 7679
+    },
+    {
+      "epoch": 1.26126496007226,
+      "grad_norm": 0.3388665890482255,
+      "learning_rate": 7.73337096388443e-06,
+      "loss": 0.4687,
+      "step": 7680
+    },
+    {
+      "epoch": 1.2614291872806027,
+      "grad_norm": 0.34797616819323746,
+      "learning_rate": 7.733044776542248e-06,
+      "loss": 0.4794,
+      "step": 7681
+    },
+    {
+      "epoch": 1.2615934144889454,
+      "grad_norm": 0.3238567132030185,
+      "learning_rate": 7.732718554086017e-06,
+      "loss": 0.4868,
+      "step": 7682
+    },
+    {
+      "epoch": 1.2617576416972882,
+      "grad_norm": 0.2742114952715846,
+      "learning_rate": 7.732392296519283e-06,
+      "loss": 0.4742,
+      "step": 7683
+    },
+    {
+      "epoch": 1.261921868905631,
+      "grad_norm": 0.29131539497818637,
+      "learning_rate": 7.732066003845588e-06,
+      "loss": 0.4896,
+      "step": 7684
+    },
+    {
+      "epoch": 1.2620860961139737,
+      "grad_norm": 0.2622760343595224,
+      "learning_rate": 7.731739676068477e-06,
+      "loss": 0.4703,
+      "step": 7685
+    },
+    {
+      "epoch": 1.2622503233223163,
+      "grad_norm": 0.2930264769742631,
+      "learning_rate": 7.731413313191492e-06,
+      "loss": 0.4515,
+      "step": 7686
+    },
+    {
+      "epoch": 1.2624145505306592,
+      "grad_norm": 0.34566046733120337,
+      "learning_rate": 7.731086915218181e-06,
+      "loss": 0.4722,
+      "step": 7687
+    },
+    {
+      "epoch": 1.2625787777390018,
+      "grad_norm": 0.39056124524545094,
+      "learning_rate": 7.730760482152085e-06,
+      "loss": 0.4809,
+      "step": 7688
+    },
+    {
+      "epoch": 1.2627430049473447,
+      "grad_norm": 0.3505071210062579,
+      "learning_rate": 7.730434013996753e-06,
+      "loss": 0.4887,
+      "step": 7689
+    },
+    {
+      "epoch": 1.2629072321556873,
+      "grad_norm": 0.34205645309234606,
+      "learning_rate": 7.730107510755729e-06,
+      "loss": 0.4932,
+      "step": 7690
+    },
+    {
+      "epoch": 1.2630714593640302,
+      "grad_norm": 0.29161143835530734,
+      "learning_rate": 7.729780972432559e-06,
+      "loss": 0.4854,
+      "step": 7691
+    },
+    {
+      "epoch": 1.2632356865723728,
+      "grad_norm": 0.45514430334380673,
+      "learning_rate": 7.729454399030791e-06,
+      "loss": 0.4945,
+      "step": 7692
+    },
+    {
+      "epoch": 1.2633999137807157,
+      "grad_norm": 0.3141566231439699,
+      "learning_rate": 7.72912779055397e-06,
+      "loss": 0.4761,
+      "step": 7693
+    },
+    {
+      "epoch": 1.2635641409890583,
+      "grad_norm": 0.3559790524001432,
+      "learning_rate": 7.728801147005643e-06,
+      "loss": 0.4822,
+      "step": 7694
+    },
+    {
+      "epoch": 1.263728368197401,
+      "grad_norm": 0.3237951014226304,
+      "learning_rate": 7.728474468389361e-06,
+      "loss": 0.5001,
+      "step": 7695
+    },
+    {
+      "epoch": 1.2638925954057438,
+      "grad_norm": 0.3288019447855348,
+      "learning_rate": 7.72814775470867e-06,
+      "loss": 0.4917,
+      "step": 7696
+    },
+    {
+      "epoch": 1.2640568226140867,
+      "grad_norm": 0.431996411415779,
+      "learning_rate": 7.727821005967117e-06,
+      "loss": 0.4715,
+      "step": 7697
+    },
+    {
+      "epoch": 1.2642210498224293,
+      "grad_norm": 0.2891597431196658,
+      "learning_rate": 7.727494222168252e-06,
+      "loss": 0.4691,
+      "step": 7698
+    },
+    {
+      "epoch": 1.264385277030772,
+      "grad_norm": 0.37428064068458555,
+      "learning_rate": 7.727167403315625e-06,
+      "loss": 0.4835,
+      "step": 7699
+    },
+    {
+      "epoch": 1.2645495042391148,
+      "grad_norm": 0.3053769528132182,
+      "learning_rate": 7.726840549412784e-06,
+      "loss": 0.4808,
+      "step": 7700
+    },
+    {
+      "epoch": 1.2647137314474577,
+      "grad_norm": 0.4850739518874018,
+      "learning_rate": 7.726513660463282e-06,
+      "loss": 0.4627,
+      "step": 7701
+    },
+    {
+      "epoch": 1.2648779586558003,
+      "grad_norm": 0.2920598155904834,
+      "learning_rate": 7.726186736470666e-06,
+      "loss": 0.4685,
+      "step": 7702
+    },
+    {
+      "epoch": 1.265042185864143,
+      "grad_norm": 0.338510521081132,
+      "learning_rate": 7.725859777438487e-06,
+      "loss": 0.4743,
+      "step": 7703
+    },
+    {
+      "epoch": 1.2652064130724858,
+      "grad_norm": 0.4296571563713848,
+      "learning_rate": 7.725532783370298e-06,
+      "loss": 0.4923,
+      "step": 7704
+    },
+    {
+      "epoch": 1.2653706402808285,
+      "grad_norm": 0.30765920212094816,
+      "learning_rate": 7.725205754269648e-06,
+      "loss": 0.5024,
+      "step": 7705
+    },
+    {
+      "epoch": 1.2655348674891713,
+      "grad_norm": 0.2709899542946473,
+      "learning_rate": 7.724878690140093e-06,
+      "loss": 0.4717,
+      "step": 7706
+    },
+    {
+      "epoch": 1.265699094697514,
+      "grad_norm": 0.29725395484723255,
+      "learning_rate": 7.724551590985182e-06,
+      "loss": 0.4627,
+      "step": 7707
+    },
+    {
+      "epoch": 1.2658633219058568,
+      "grad_norm": 0.4053905691221062,
+      "learning_rate": 7.724224456808465e-06,
+      "loss": 0.4852,
+      "step": 7708
+    },
+    {
+      "epoch": 1.2660275491141995,
+      "grad_norm": 0.3116877740483153,
+      "learning_rate": 7.723897287613502e-06,
+      "loss": 0.4735,
+      "step": 7709
+    },
+    {
+      "epoch": 1.2661917763225423,
+      "grad_norm": 0.35449971632410593,
+      "learning_rate": 7.72357008340384e-06,
+      "loss": 0.4981,
+      "step": 7710
+    },
+    {
+      "epoch": 1.266356003530885,
+      "grad_norm": 0.3482399264073418,
+      "learning_rate": 7.723242844183038e-06,
+      "loss": 0.4552,
+      "step": 7711
+    },
+    {
+      "epoch": 1.2665202307392276,
+      "grad_norm": 0.437774132734577,
+      "learning_rate": 7.722915569954646e-06,
+      "loss": 0.4631,
+      "step": 7712
+    },
+    {
+      "epoch": 1.2666844579475705,
+      "grad_norm": 0.26839427210078487,
+      "learning_rate": 7.72258826072222e-06,
+      "loss": 0.495,
+      "step": 7713
+    },
+    {
+      "epoch": 1.2668486851559133,
+      "grad_norm": 0.31217056993105924,
+      "learning_rate": 7.722260916489313e-06,
+      "loss": 0.4844,
+      "step": 7714
+    },
+    {
+      "epoch": 1.267012912364256,
+      "grad_norm": 0.4914939353825111,
+      "learning_rate": 7.721933537259483e-06,
+      "loss": 0.4797,
+      "step": 7715
+    },
+    {
+      "epoch": 1.2671771395725986,
+      "grad_norm": 0.29182947550725524,
+      "learning_rate": 7.721606123036288e-06,
+      "loss": 0.4833,
+      "step": 7716
+    },
+    {
+      "epoch": 1.2673413667809414,
+      "grad_norm": 0.37793151937577985,
+      "learning_rate": 7.721278673823278e-06,
+      "loss": 0.4669,
+      "step": 7717
+    },
+    {
+      "epoch": 1.2675055939892843,
+      "grad_norm": 0.26048313020401365,
+      "learning_rate": 7.720951189624013e-06,
+      "loss": 0.4815,
+      "step": 7718
+    },
+    {
+      "epoch": 1.267669821197627,
+      "grad_norm": 0.3111343689668257,
+      "learning_rate": 7.720623670442048e-06,
+      "loss": 0.471,
+      "step": 7719
+    },
+    {
+      "epoch": 1.2678340484059696,
+      "grad_norm": 0.3135420875235267,
+      "learning_rate": 7.720296116280944e-06,
+      "loss": 0.4746,
+      "step": 7720
+    },
+    {
+      "epoch": 1.2679982756143124,
+      "grad_norm": 0.2976984908795594,
+      "learning_rate": 7.719968527144253e-06,
+      "loss": 0.4635,
+      "step": 7721
+    },
+    {
+      "epoch": 1.268162502822655,
+      "grad_norm": 0.2756385116643583,
+      "learning_rate": 7.719640903035538e-06,
+      "loss": 0.4821,
+      "step": 7722
+    },
+    {
+      "epoch": 1.268326730030998,
+      "grad_norm": 0.3310928364944098,
+      "learning_rate": 7.719313243958353e-06,
+      "loss": 0.5135,
+      "step": 7723
+    },
+    {
+      "epoch": 1.2684909572393406,
+      "grad_norm": 0.2898540745776658,
+      "learning_rate": 7.71898554991626e-06,
+      "loss": 0.4747,
+      "step": 7724
+    },
+    {
+      "epoch": 1.2686551844476834,
+      "grad_norm": 0.29846089427282313,
+      "learning_rate": 7.718657820912816e-06,
+      "loss": 0.4645,
+      "step": 7725
+    },
+    {
+      "epoch": 1.268819411656026,
+      "grad_norm": 0.4277166211567597,
+      "learning_rate": 7.718330056951582e-06,
+      "loss": 0.4626,
+      "step": 7726
+    },
+    {
+      "epoch": 1.268983638864369,
+      "grad_norm": 0.3386584451128976,
+      "learning_rate": 7.718002258036117e-06,
+      "loss": 0.45,
+      "step": 7727
+    },
+    {
+      "epoch": 1.2691478660727116,
+      "grad_norm": 0.31044845044724056,
+      "learning_rate": 7.717674424169983e-06,
+      "loss": 0.493,
+      "step": 7728
+    },
+    {
+      "epoch": 1.2693120932810542,
+      "grad_norm": 0.28929110658363827,
+      "learning_rate": 7.717346555356737e-06,
+      "loss": 0.4995,
+      "step": 7729
+    },
+    {
+      "epoch": 1.269476320489397,
+      "grad_norm": 0.2750277317962798,
+      "learning_rate": 7.717018651599942e-06,
+      "loss": 0.4523,
+      "step": 7730
+    },
+    {
+      "epoch": 1.26964054769774,
+      "grad_norm": 0.2646663022346015,
+      "learning_rate": 7.71669071290316e-06,
+      "loss": 0.4684,
+      "step": 7731
+    },
+    {
+      "epoch": 1.2698047749060826,
+      "grad_norm": 0.2722408985365008,
+      "learning_rate": 7.716362739269952e-06,
+      "loss": 0.4867,
+      "step": 7732
+    },
+    {
+      "epoch": 1.2699690021144252,
+      "grad_norm": 0.2813196075417855,
+      "learning_rate": 7.71603473070388e-06,
+      "loss": 0.5011,
+      "step": 7733
+    },
+    {
+      "epoch": 1.270133229322768,
+      "grad_norm": 0.2811195028154342,
+      "learning_rate": 7.715706687208507e-06,
+      "loss": 0.4883,
+      "step": 7734
+    },
+    {
+      "epoch": 1.270297456531111,
+      "grad_norm": 0.3121290061085945,
+      "learning_rate": 7.715378608787394e-06,
+      "loss": 0.5022,
+      "step": 7735
+    },
+    {
+      "epoch": 1.2704616837394536,
+      "grad_norm": 0.29219405088342026,
+      "learning_rate": 7.715050495444108e-06,
+      "loss": 0.4765,
+      "step": 7736
+    },
+    {
+      "epoch": 1.2706259109477962,
+      "grad_norm": 0.29988768120792714,
+      "learning_rate": 7.71472234718221e-06,
+      "loss": 0.4849,
+      "step": 7737
+    },
+    {
+      "epoch": 1.270790138156139,
+      "grad_norm": 0.30856320601214104,
+      "learning_rate": 7.714394164005264e-06,
+      "loss": 0.4753,
+      "step": 7738
+    },
+    {
+      "epoch": 1.2709543653644817,
+      "grad_norm": 0.29738267786753775,
+      "learning_rate": 7.714065945916834e-06,
+      "loss": 0.4698,
+      "step": 7739
+    },
+    {
+      "epoch": 1.2711185925728246,
+      "grad_norm": 0.3323745523516455,
+      "learning_rate": 7.713737692920488e-06,
+      "loss": 0.4704,
+      "step": 7740
+    },
+    {
+      "epoch": 1.2712828197811672,
+      "grad_norm": 0.2512031804860789,
+      "learning_rate": 7.713409405019786e-06,
+      "loss": 0.4737,
+      "step": 7741
+    },
+    {
+      "epoch": 1.27144704698951,
+      "grad_norm": 0.28820433890993685,
+      "learning_rate": 7.713081082218297e-06,
+      "loss": 0.4625,
+      "step": 7742
+    },
+    {
+      "epoch": 1.2716112741978527,
+      "grad_norm": 0.3977011594936401,
+      "learning_rate": 7.712752724519588e-06,
+      "loss": 0.4781,
+      "step": 7743
+    },
+    {
+      "epoch": 1.2717755014061956,
+      "grad_norm": 0.26281093424723,
+      "learning_rate": 7.712424331927221e-06,
+      "loss": 0.4773,
+      "step": 7744
+    },
+    {
+      "epoch": 1.2719397286145382,
+      "grad_norm": 0.3895924508235183,
+      "learning_rate": 7.712095904444767e-06,
+      "loss": 0.4865,
+      "step": 7745
+    },
+    {
+      "epoch": 1.2721039558228808,
+      "grad_norm": 0.6117971589272038,
+      "learning_rate": 7.71176744207579e-06,
+      "loss": 0.4621,
+      "step": 7746
+    },
+    {
+      "epoch": 1.2722681830312237,
+      "grad_norm": 0.2916185512611372,
+      "learning_rate": 7.71143894482386e-06,
+      "loss": 0.4791,
+      "step": 7747
+    },
+    {
+      "epoch": 1.2724324102395665,
+      "grad_norm": 0.3215675610815016,
+      "learning_rate": 7.711110412692543e-06,
+      "loss": 0.4743,
+      "step": 7748
+    },
+    {
+      "epoch": 1.2725966374479092,
+      "grad_norm": 0.3985246567089694,
+      "learning_rate": 7.710781845685406e-06,
+      "loss": 0.4826,
+      "step": 7749
+    },
+    {
+      "epoch": 1.2727608646562518,
+      "grad_norm": 0.3815067560152665,
+      "learning_rate": 7.710453243806021e-06,
+      "loss": 0.4847,
+      "step": 7750
+    },
+    {
+      "epoch": 1.2729250918645947,
+      "grad_norm": 0.2545415166282443,
+      "learning_rate": 7.710124607057954e-06,
+      "loss": 0.452,
+      "step": 7751
+    },
+    {
+      "epoch": 1.2730893190729375,
+      "grad_norm": 0.3512974958871999,
+      "learning_rate": 7.709795935444777e-06,
+      "loss": 0.4545,
+      "step": 7752
+    },
+    {
+      "epoch": 1.2732535462812802,
+      "grad_norm": 0.33763955275343127,
+      "learning_rate": 7.709467228970056e-06,
+      "loss": 0.4819,
+      "step": 7753
+    },
+    {
+      "epoch": 1.2734177734896228,
+      "grad_norm": 0.47180854525995025,
+      "learning_rate": 7.709138487637365e-06,
+      "loss": 0.4773,
+      "step": 7754
+    },
+    {
+      "epoch": 1.2735820006979657,
+      "grad_norm": 0.2883546400469497,
+      "learning_rate": 7.708809711450272e-06,
+      "loss": 0.4654,
+      "step": 7755
+    },
+    {
+      "epoch": 1.2737462279063083,
+      "grad_norm": 0.27563610891719326,
+      "learning_rate": 7.708480900412348e-06,
+      "loss": 0.4801,
+      "step": 7756
+    },
+    {
+      "epoch": 1.2739104551146512,
+      "grad_norm": 0.5716890214745179,
+      "learning_rate": 7.708152054527165e-06,
+      "loss": 0.4688,
+      "step": 7757
+    },
+    {
+      "epoch": 1.2740746823229938,
+      "grad_norm": 0.42195057385213214,
+      "learning_rate": 7.707823173798295e-06,
+      "loss": 0.4796,
+      "step": 7758
+    },
+    {
+      "epoch": 1.2742389095313367,
+      "grad_norm": 0.30176871840975483,
+      "learning_rate": 7.707494258229308e-06,
+      "loss": 0.4812,
+      "step": 7759
+    },
+    {
+      "epoch": 1.2744031367396793,
+      "grad_norm": 0.3565361831398607,
+      "learning_rate": 7.707165307823778e-06,
+      "loss": 0.4662,
+      "step": 7760
+    },
+    {
+      "epoch": 1.2745673639480222,
+      "grad_norm": 0.29370991851580663,
+      "learning_rate": 7.706836322585278e-06,
+      "loss": 0.485,
+      "step": 7761
+    },
+    {
+      "epoch": 1.2747315911563648,
+      "grad_norm": 0.3047484919897347,
+      "learning_rate": 7.70650730251738e-06,
+      "loss": 0.4779,
+      "step": 7762
+    },
+    {
+      "epoch": 1.2748958183647074,
+      "grad_norm": 0.28808269074598614,
+      "learning_rate": 7.706178247623659e-06,
+      "loss": 0.4763,
+      "step": 7763
+    },
+    {
+      "epoch": 1.2750600455730503,
+      "grad_norm": 0.27915183570391694,
+      "learning_rate": 7.705849157907686e-06,
+      "loss": 0.4792,
+      "step": 7764
+    },
+    {
+      "epoch": 1.2752242727813932,
+      "grad_norm": 0.32806606132369737,
+      "learning_rate": 7.705520033373038e-06,
+      "loss": 0.4698,
+      "step": 7765
+    },
+    {
+      "epoch": 1.2753884999897358,
+      "grad_norm": 0.3089376020988689,
+      "learning_rate": 7.70519087402329e-06,
+      "loss": 0.5008,
+      "step": 7766
+    },
+    {
+      "epoch": 1.2755527271980784,
+      "grad_norm": 0.2725056465017377,
+      "learning_rate": 7.704861679862013e-06,
+      "loss": 0.4721,
+      "step": 7767
+    },
+    {
+      "epoch": 1.2757169544064213,
+      "grad_norm": 0.28418412299316054,
+      "learning_rate": 7.704532450892785e-06,
+      "loss": 0.4714,
+      "step": 7768
+    },
+    {
+      "epoch": 1.2758811816147642,
+      "grad_norm": 0.3165971984097728,
+      "learning_rate": 7.704203187119183e-06,
+      "loss": 0.4929,
+      "step": 7769
+    },
+    {
+      "epoch": 1.2760454088231068,
+      "grad_norm": 0.36671004266675833,
+      "learning_rate": 7.703873888544782e-06,
+      "loss": 0.4695,
+      "step": 7770
+    },
+    {
+      "epoch": 1.2762096360314494,
+      "grad_norm": 0.2693544975307786,
+      "learning_rate": 7.703544555173158e-06,
+      "loss": 0.476,
+      "step": 7771
+    },
+    {
+      "epoch": 1.2763738632397923,
+      "grad_norm": 0.3594750493142773,
+      "learning_rate": 7.703215187007889e-06,
+      "loss": 0.4489,
+      "step": 7772
+    },
+    {
+      "epoch": 1.276538090448135,
+      "grad_norm": 0.3247458570973557,
+      "learning_rate": 7.70288578405255e-06,
+      "loss": 0.4897,
+      "step": 7773
+    },
+    {
+      "epoch": 1.2767023176564778,
+      "grad_norm": 0.31715952464757263,
+      "learning_rate": 7.702556346310721e-06,
+      "loss": 0.4879,
+      "step": 7774
+    },
+    {
+      "epoch": 1.2768665448648204,
+      "grad_norm": 0.3040453912621078,
+      "learning_rate": 7.70222687378598e-06,
+      "loss": 0.4747,
+      "step": 7775
+    },
+    {
+      "epoch": 1.2770307720731633,
+      "grad_norm": 0.35248094697874277,
+      "learning_rate": 7.701897366481903e-06,
+      "loss": 0.4936,
+      "step": 7776
+    },
+    {
+      "epoch": 1.277194999281506,
+      "grad_norm": 0.28411064669334896,
+      "learning_rate": 7.70156782440207e-06,
+      "loss": 0.4849,
+      "step": 7777
+    },
+    {
+      "epoch": 1.2773592264898488,
+      "grad_norm": 0.30851415055717885,
+      "learning_rate": 7.701238247550064e-06,
+      "loss": 0.4883,
+      "step": 7778
+    },
+    {
+      "epoch": 1.2775234536981914,
+      "grad_norm": 0.30100077430210254,
+      "learning_rate": 7.700908635929458e-06,
+      "loss": 0.4926,
+      "step": 7779
+    },
+    {
+      "epoch": 1.277687680906534,
+      "grad_norm": 0.3256247158390222,
+      "learning_rate": 7.700578989543835e-06,
+      "loss": 0.4769,
+      "step": 7780
+    },
+    {
+      "epoch": 1.277851908114877,
+      "grad_norm": 0.3831007490176088,
+      "learning_rate": 7.700249308396775e-06,
+      "loss": 0.481,
+      "step": 7781
+    },
+    {
+      "epoch": 1.2780161353232198,
+      "grad_norm": 0.2900015324255877,
+      "learning_rate": 7.699919592491862e-06,
+      "loss": 0.4581,
+      "step": 7782
+    },
+    {
+      "epoch": 1.2781803625315624,
+      "grad_norm": 0.39968965012855234,
+      "learning_rate": 7.699589841832671e-06,
+      "loss": 0.475,
+      "step": 7783
+    },
+    {
+      "epoch": 1.278344589739905,
+      "grad_norm": 0.2987916540815722,
+      "learning_rate": 7.699260056422787e-06,
+      "loss": 0.4799,
+      "step": 7784
+    },
+    {
+      "epoch": 1.278508816948248,
+      "grad_norm": 0.3031895511173132,
+      "learning_rate": 7.69893023626579e-06,
+      "loss": 0.4982,
+      "step": 7785
+    },
+    {
+      "epoch": 1.2786730441565908,
+      "grad_norm": 0.3036242316048294,
+      "learning_rate": 7.698600381365264e-06,
+      "loss": 0.4725,
+      "step": 7786
+    },
+    {
+      "epoch": 1.2788372713649334,
+      "grad_norm": 0.29859693726042974,
+      "learning_rate": 7.698270491724793e-06,
+      "loss": 0.4765,
+      "step": 7787
+    },
+    {
+      "epoch": 1.279001498573276,
+      "grad_norm": 0.2845879852561358,
+      "learning_rate": 7.697940567347956e-06,
+      "loss": 0.4864,
+      "step": 7788
+    },
+    {
+      "epoch": 1.279165725781619,
+      "grad_norm": 0.28163700973601685,
+      "learning_rate": 7.697610608238338e-06,
+      "loss": 0.4703,
+      "step": 7789
+    },
+    {
+      "epoch": 1.2793299529899615,
+      "grad_norm": 0.30036853676830083,
+      "learning_rate": 7.697280614399523e-06,
+      "loss": 0.4867,
+      "step": 7790
+    },
+    {
+      "epoch": 1.2794941801983044,
+      "grad_norm": 0.28760448854099513,
+      "learning_rate": 7.696950585835094e-06,
+      "loss": 0.4448,
+      "step": 7791
+    },
+    {
+      "epoch": 1.279658407406647,
+      "grad_norm": 0.33460690891254224,
+      "learning_rate": 7.696620522548638e-06,
+      "loss": 0.4686,
+      "step": 7792
+    },
+    {
+      "epoch": 1.27982263461499,
+      "grad_norm": 0.27290421235975976,
+      "learning_rate": 7.696290424543737e-06,
+      "loss": 0.4735,
+      "step": 7793
+    },
+    {
+      "epoch": 1.2799868618233325,
+      "grad_norm": 0.2896345427090911,
+      "learning_rate": 7.695960291823978e-06,
+      "loss": 0.4585,
+      "step": 7794
+    },
+    {
+      "epoch": 1.2801510890316754,
+      "grad_norm": 0.2884151664673828,
+      "learning_rate": 7.695630124392945e-06,
+      "loss": 0.459,
+      "step": 7795
+    },
+    {
+      "epoch": 1.280315316240018,
+      "grad_norm": 0.29439858268755925,
+      "learning_rate": 7.695299922254224e-06,
+      "loss": 0.4635,
+      "step": 7796
+    },
+    {
+      "epoch": 1.2804795434483607,
+      "grad_norm": 0.27699445691313457,
+      "learning_rate": 7.694969685411404e-06,
+      "loss": 0.484,
+      "step": 7797
+    },
+    {
+      "epoch": 1.2806437706567035,
+      "grad_norm": 0.24163729471658402,
+      "learning_rate": 7.694639413868068e-06,
+      "loss": 0.4605,
+      "step": 7798
+    },
+    {
+      "epoch": 1.2808079978650464,
+      "grad_norm": 0.2896967081115697,
+      "learning_rate": 7.694309107627806e-06,
+      "loss": 0.4856,
+      "step": 7799
+    },
+    {
+      "epoch": 1.280972225073389,
+      "grad_norm": 0.299584030087436,
+      "learning_rate": 7.693978766694204e-06,
+      "loss": 0.452,
+      "step": 7800
+    },
+    {
+      "epoch": 1.2811364522817317,
+      "grad_norm": 0.3385113952854699,
+      "learning_rate": 7.693648391070851e-06,
+      "loss": 0.4516,
+      "step": 7801
+    },
+    {
+      "epoch": 1.2813006794900745,
+      "grad_norm": 0.2772793930227556,
+      "learning_rate": 7.693317980761334e-06,
+      "loss": 0.4786,
+      "step": 7802
+    },
+    {
+      "epoch": 1.2814649066984174,
+      "grad_norm": 0.3340403620740822,
+      "learning_rate": 7.69298753576924e-06,
+      "loss": 0.485,
+      "step": 7803
+    },
+    {
+      "epoch": 1.28162913390676,
+      "grad_norm": 0.3315139362421689,
+      "learning_rate": 7.692657056098163e-06,
+      "loss": 0.4865,
+      "step": 7804
+    },
+    {
+      "epoch": 1.2817933611151027,
+      "grad_norm": 0.36606393979166196,
+      "learning_rate": 7.692326541751687e-06,
+      "loss": 0.4633,
+      "step": 7805
+    },
+    {
+      "epoch": 1.2819575883234455,
+      "grad_norm": 0.2958031648909456,
+      "learning_rate": 7.691995992733404e-06,
+      "loss": 0.4859,
+      "step": 7806
+    },
+    {
+      "epoch": 1.2821218155317882,
+      "grad_norm": 0.29278420815146317,
+      "learning_rate": 7.691665409046905e-06,
+      "loss": 0.4689,
+      "step": 7807
+    },
+    {
+      "epoch": 1.282286042740131,
+      "grad_norm": 0.26629433728377444,
+      "learning_rate": 7.69133479069578e-06,
+      "loss": 0.4652,
+      "step": 7808
+    },
+    {
+      "epoch": 1.2824502699484737,
+      "grad_norm": 0.3085504664475122,
+      "learning_rate": 7.691004137683617e-06,
+      "loss": 0.4806,
+      "step": 7809
+    },
+    {
+      "epoch": 1.2826144971568165,
+      "grad_norm": 0.31694594797826553,
+      "learning_rate": 7.69067345001401e-06,
+      "loss": 0.4751,
+      "step": 7810
+    },
+    {
+      "epoch": 1.2827787243651592,
+      "grad_norm": 0.31604221751382056,
+      "learning_rate": 7.690342727690553e-06,
+      "loss": 0.4662,
+      "step": 7811
+    },
+    {
+      "epoch": 1.282942951573502,
+      "grad_norm": 0.3576539317488584,
+      "learning_rate": 7.690011970716833e-06,
+      "loss": 0.483,
+      "step": 7812
+    },
+    {
+      "epoch": 1.2831071787818447,
+      "grad_norm": 0.26640826289893155,
+      "learning_rate": 7.689681179096443e-06,
+      "loss": 0.4585,
+      "step": 7813
+    },
+    {
+      "epoch": 1.2832714059901873,
+      "grad_norm": 0.24764426877232484,
+      "learning_rate": 7.68935035283298e-06,
+      "loss": 0.4626,
+      "step": 7814
+    },
+    {
+      "epoch": 1.2834356331985302,
+      "grad_norm": 0.31402171978360294,
+      "learning_rate": 7.689019491930033e-06,
+      "loss": 0.4765,
+      "step": 7815
+    },
+    {
+      "epoch": 1.283599860406873,
+      "grad_norm": 0.33472276789419364,
+      "learning_rate": 7.688688596391197e-06,
+      "loss": 0.4848,
+      "step": 7816
+    },
+    {
+      "epoch": 1.2837640876152157,
+      "grad_norm": 0.31467631849619093,
+      "learning_rate": 7.688357666220065e-06,
+      "loss": 0.4696,
+      "step": 7817
+    },
+    {
+      "epoch": 1.2839283148235583,
+      "grad_norm": 0.5620271778383376,
+      "learning_rate": 7.688026701420233e-06,
+      "loss": 0.4873,
+      "step": 7818
+    },
+    {
+      "epoch": 1.2840925420319012,
+      "grad_norm": 0.30644020342381995,
+      "learning_rate": 7.687695701995295e-06,
+      "loss": 0.5001,
+      "step": 7819
+    },
+    {
+      "epoch": 1.284256769240244,
+      "grad_norm": 0.31653718755727756,
+      "learning_rate": 7.687364667948842e-06,
+      "loss": 0.4556,
+      "step": 7820
+    },
+    {
+      "epoch": 1.2844209964485866,
+      "grad_norm": 0.35087942099619734,
+      "learning_rate": 7.687033599284475e-06,
+      "loss": 0.4497,
+      "step": 7821
+    },
+    {
+      "epoch": 1.2845852236569293,
+      "grad_norm": 0.28672157890856287,
+      "learning_rate": 7.686702496005788e-06,
+      "loss": 0.4981,
+      "step": 7822
+    },
+    {
+      "epoch": 1.2847494508652721,
+      "grad_norm": 0.3079510574279276,
+      "learning_rate": 7.686371358116374e-06,
+      "loss": 0.4902,
+      "step": 7823
+    },
+    {
+      "epoch": 1.2849136780736148,
+      "grad_norm": 0.44588293878026053,
+      "learning_rate": 7.686040185619835e-06,
+      "loss": 0.46,
+      "step": 7824
+    },
+    {
+      "epoch": 1.2850779052819576,
+      "grad_norm": 0.872447022198342,
+      "learning_rate": 7.685708978519764e-06,
+      "loss": 0.459,
+      "step": 7825
+    },
+    {
+      "epoch": 1.2852421324903003,
+      "grad_norm": 0.3303561048803215,
+      "learning_rate": 7.68537773681976e-06,
+      "loss": 0.4975,
+      "step": 7826
+    },
+    {
+      "epoch": 1.2854063596986431,
+      "grad_norm": 0.26456162856250737,
+      "learning_rate": 7.685046460523419e-06,
+      "loss": 0.4888,
+      "step": 7827
+    },
+    {
+      "epoch": 1.2855705869069858,
+      "grad_norm": 0.2816527615338236,
+      "learning_rate": 7.684715149634339e-06,
+      "loss": 0.4782,
+      "step": 7828
+    },
+    {
+      "epoch": 1.2857348141153286,
+      "grad_norm": 0.3147258571690476,
+      "learning_rate": 7.68438380415612e-06,
+      "loss": 0.4844,
+      "step": 7829
+    },
+    {
+      "epoch": 1.2858990413236713,
+      "grad_norm": 0.27745651039854247,
+      "learning_rate": 7.68405242409236e-06,
+      "loss": 0.4556,
+      "step": 7830
+    },
+    {
+      "epoch": 1.286063268532014,
+      "grad_norm": 0.30628636649669544,
+      "learning_rate": 7.683721009446657e-06,
+      "loss": 0.4868,
+      "step": 7831
+    },
+    {
+      "epoch": 1.2862274957403568,
+      "grad_norm": 0.42664379936326835,
+      "learning_rate": 7.683389560222612e-06,
+      "loss": 0.4784,
+      "step": 7832
+    },
+    {
+      "epoch": 1.2863917229486996,
+      "grad_norm": 0.2680221112365354,
+      "learning_rate": 7.683058076423825e-06,
+      "loss": 0.4547,
+      "step": 7833
+    },
+    {
+      "epoch": 1.2865559501570423,
+      "grad_norm": 0.28850192993663787,
+      "learning_rate": 7.682726558053896e-06,
+      "loss": 0.4816,
+      "step": 7834
+    },
+    {
+      "epoch": 1.286720177365385,
+      "grad_norm": 0.3826116769850467,
+      "learning_rate": 7.682395005116424e-06,
+      "loss": 0.4604,
+      "step": 7835
+    },
+    {
+      "epoch": 1.2868844045737278,
+      "grad_norm": 0.3269271472240874,
+      "learning_rate": 7.682063417615011e-06,
+      "loss": 0.4856,
+      "step": 7836
+    },
+    {
+      "epoch": 1.2870486317820706,
+      "grad_norm": 0.2903246693555103,
+      "learning_rate": 7.681731795553259e-06,
+      "loss": 0.4717,
+      "step": 7837
+    },
+    {
+      "epoch": 1.2872128589904133,
+      "grad_norm": 0.40635716934363103,
+      "learning_rate": 7.681400138934768e-06,
+      "loss": 0.4641,
+      "step": 7838
+    },
+    {
+      "epoch": 1.287377086198756,
+      "grad_norm": 2.2006250297078624,
+      "learning_rate": 7.681068447763143e-06,
+      "loss": 0.4772,
+      "step": 7839
+    },
+    {
+      "epoch": 1.2875413134070988,
+      "grad_norm": 0.25645959340248886,
+      "learning_rate": 7.680736722041985e-06,
+      "loss": 0.4555,
+      "step": 7840
+    },
+    {
+      "epoch": 1.2877055406154414,
+      "grad_norm": 0.29428794743469366,
+      "learning_rate": 7.680404961774898e-06,
+      "loss": 0.4554,
+      "step": 7841
+    },
+    {
+      "epoch": 1.2878697678237843,
+      "grad_norm": 0.504873026867374,
+      "learning_rate": 7.680073166965482e-06,
+      "loss": 0.4737,
+      "step": 7842
+    },
+    {
+      "epoch": 1.288033995032127,
+      "grad_norm": 0.3206722968962623,
+      "learning_rate": 7.679741337617344e-06,
+      "loss": 0.4555,
+      "step": 7843
+    },
+    {
+      "epoch": 1.2881982222404698,
+      "grad_norm": 0.31375416139386475,
+      "learning_rate": 7.679409473734085e-06,
+      "loss": 0.4731,
+      "step": 7844
+    },
+    {
+      "epoch": 1.2883624494488124,
+      "grad_norm": 0.32465719686723155,
+      "learning_rate": 7.67907757531931e-06,
+      "loss": 0.4608,
+      "step": 7845
+    },
+    {
+      "epoch": 1.2885266766571553,
+      "grad_norm": 0.2873096643272258,
+      "learning_rate": 7.678745642376627e-06,
+      "loss": 0.4706,
+      "step": 7846
+    },
+    {
+      "epoch": 1.288690903865498,
+      "grad_norm": 0.28816537852812213,
+      "learning_rate": 7.67841367490964e-06,
+      "loss": 0.4673,
+      "step": 7847
+    },
+    {
+      "epoch": 1.2888551310738405,
+      "grad_norm": 0.27675067172937057,
+      "learning_rate": 7.67808167292195e-06,
+      "loss": 0.4654,
+      "step": 7848
+    },
+    {
+      "epoch": 1.2890193582821834,
+      "grad_norm": 0.3456711175077695,
+      "learning_rate": 7.67774963641717e-06,
+      "loss": 0.4813,
+      "step": 7849
+    },
+    {
+      "epoch": 1.2891835854905263,
+      "grad_norm": 0.3354036100274437,
+      "learning_rate": 7.677417565398899e-06,
+      "loss": 0.4839,
+      "step": 7850
+    },
+    {
+      "epoch": 1.289347812698869,
+      "grad_norm": 0.6324881168177896,
+      "learning_rate": 7.677085459870749e-06,
+      "loss": 0.4587,
+      "step": 7851
+    },
+    {
+      "epoch": 1.2895120399072115,
+      "grad_norm": 0.3504917226050069,
+      "learning_rate": 7.676753319836324e-06,
+      "loss": 0.4733,
+      "step": 7852
+    },
+    {
+      "epoch": 1.2896762671155544,
+      "grad_norm": 0.44980333945469475,
+      "learning_rate": 7.676421145299233e-06,
+      "loss": 0.4676,
+      "step": 7853
+    },
+    {
+      "epoch": 1.2898404943238972,
+      "grad_norm": 0.5372650657765722,
+      "learning_rate": 7.676088936263084e-06,
+      "loss": 0.4819,
+      "step": 7854
+    },
+    {
+      "epoch": 1.2900047215322399,
+      "grad_norm": 0.3343313623537713,
+      "learning_rate": 7.675756692731483e-06,
+      "loss": 0.4794,
+      "step": 7855
+    },
+    {
+      "epoch": 1.2901689487405825,
+      "grad_norm": 0.3946859710241651,
+      "learning_rate": 7.67542441470804e-06,
+      "loss": 0.476,
+      "step": 7856
+    },
+    {
+      "epoch": 1.2903331759489254,
+      "grad_norm": 0.2813570879176587,
+      "learning_rate": 7.675092102196365e-06,
+      "loss": 0.4871,
+      "step": 7857
+    },
+    {
+      "epoch": 1.290497403157268,
+      "grad_norm": 0.3379571926965503,
+      "learning_rate": 7.674759755200064e-06,
+      "loss": 0.4802,
+      "step": 7858
+    },
+    {
+      "epoch": 1.2906616303656109,
+      "grad_norm": 0.44337435809416115,
+      "learning_rate": 7.67442737372275e-06,
+      "loss": 0.5086,
+      "step": 7859
+    },
+    {
+      "epoch": 1.2908258575739535,
+      "grad_norm": 0.2620657068472747,
+      "learning_rate": 7.674094957768031e-06,
+      "loss": 0.4507,
+      "step": 7860
+    },
+    {
+      "epoch": 1.2909900847822964,
+      "grad_norm": 0.39448810420551583,
+      "learning_rate": 7.673762507339517e-06,
+      "loss": 0.4757,
+      "step": 7861
+    },
+    {
+      "epoch": 1.291154311990639,
+      "grad_norm": 0.27602612398995907,
+      "learning_rate": 7.67343002244082e-06,
+      "loss": 0.4791,
+      "step": 7862
+    },
+    {
+      "epoch": 1.2913185391989819,
+      "grad_norm": 0.30145053047947384,
+      "learning_rate": 7.67309750307555e-06,
+      "loss": 0.4637,
+      "step": 7863
+    },
+    {
+      "epoch": 1.2914827664073245,
+      "grad_norm": 0.28617067481182934,
+      "learning_rate": 7.672764949247322e-06,
+      "loss": 0.4892,
+      "step": 7864
+    },
+    {
+      "epoch": 1.2916469936156671,
+      "grad_norm": 0.30999809485905655,
+      "learning_rate": 7.672432360959743e-06,
+      "loss": 0.5012,
+      "step": 7865
+    },
+    {
+      "epoch": 1.29181122082401,
+      "grad_norm": 0.32598640649592947,
+      "learning_rate": 7.672099738216427e-06,
+      "loss": 0.4727,
+      "step": 7866
+    },
+    {
+      "epoch": 1.2919754480323529,
+      "grad_norm": 0.4536033430129751,
+      "learning_rate": 7.671767081020988e-06,
+      "loss": 0.4617,
+      "step": 7867
+    },
+    {
+      "epoch": 1.2921396752406955,
+      "grad_norm": 0.2727844305112922,
+      "learning_rate": 7.671434389377038e-06,
+      "loss": 0.4555,
+      "step": 7868
+    },
+    {
+      "epoch": 1.2923039024490381,
+      "grad_norm": 0.3128387369899632,
+      "learning_rate": 7.67110166328819e-06,
+      "loss": 0.4667,
+      "step": 7869
+    },
+    {
+      "epoch": 1.292468129657381,
+      "grad_norm": 0.314249956724488,
+      "learning_rate": 7.670768902758058e-06,
+      "loss": 0.4856,
+      "step": 7870
+    },
+    {
+      "epoch": 1.2926323568657239,
+      "grad_norm": 0.4280519629179726,
+      "learning_rate": 7.670436107790254e-06,
+      "loss": 0.4689,
+      "step": 7871
+    },
+    {
+      "epoch": 1.2927965840740665,
+      "grad_norm": 0.289677597791496,
+      "learning_rate": 7.670103278388398e-06,
+      "loss": 0.4686,
+      "step": 7872
+    },
+    {
+      "epoch": 1.2929608112824091,
+      "grad_norm": 0.2946877863022075,
+      "learning_rate": 7.6697704145561e-06,
+      "loss": 0.5096,
+      "step": 7873
+    },
+    {
+      "epoch": 1.293125038490752,
+      "grad_norm": 0.2894735076390649,
+      "learning_rate": 7.669437516296976e-06,
+      "loss": 0.4614,
+      "step": 7874
+    },
+    {
+      "epoch": 1.2932892656990946,
+      "grad_norm": 0.3028986464007853,
+      "learning_rate": 7.669104583614642e-06,
+      "loss": 0.4541,
+      "step": 7875
+    },
+    {
+      "epoch": 1.2934534929074375,
+      "grad_norm": 0.25582922737852953,
+      "learning_rate": 7.668771616512716e-06,
+      "loss": 0.4725,
+      "step": 7876
+    },
+    {
+      "epoch": 1.2936177201157801,
+      "grad_norm": 0.33755488887573115,
+      "learning_rate": 7.668438614994812e-06,
+      "loss": 0.4787,
+      "step": 7877
+    },
+    {
+      "epoch": 1.293781947324123,
+      "grad_norm": 0.32217811133152874,
+      "learning_rate": 7.668105579064546e-06,
+      "loss": 0.4938,
+      "step": 7878
+    },
+    {
+      "epoch": 1.2939461745324656,
+      "grad_norm": 0.30142798654133546,
+      "learning_rate": 7.667772508725538e-06,
+      "loss": 0.4816,
+      "step": 7879
+    },
+    {
+      "epoch": 1.2941104017408085,
+      "grad_norm": 0.26134890830428037,
+      "learning_rate": 7.667439403981402e-06,
+      "loss": 0.4805,
+      "step": 7880
+    },
+    {
+      "epoch": 1.2942746289491511,
+      "grad_norm": 0.2859171992401748,
+      "learning_rate": 7.66710626483576e-06,
+      "loss": 0.4585,
+      "step": 7881
+    },
+    {
+      "epoch": 1.2944388561574938,
+      "grad_norm": 0.27490980788645614,
+      "learning_rate": 7.666773091292227e-06,
+      "loss": 0.4662,
+      "step": 7882
+    },
+    {
+      "epoch": 1.2946030833658366,
+      "grad_norm": 0.29612051837616454,
+      "learning_rate": 7.666439883354421e-06,
+      "loss": 0.4681,
+      "step": 7883
+    },
+    {
+      "epoch": 1.2947673105741795,
+      "grad_norm": 0.3558706635350218,
+      "learning_rate": 7.666106641025965e-06,
+      "loss": 0.4668,
+      "step": 7884
+    },
+    {
+      "epoch": 1.2949315377825221,
+      "grad_norm": 0.29081743807681926,
+      "learning_rate": 7.665773364310476e-06,
+      "loss": 0.4673,
+      "step": 7885
+    },
+    {
+      "epoch": 1.2950957649908648,
+      "grad_norm": 0.36587349094750693,
+      "learning_rate": 7.665440053211571e-06,
+      "loss": 0.472,
+      "step": 7886
+    },
+    {
+      "epoch": 1.2952599921992076,
+      "grad_norm": 0.3846805506898003,
+      "learning_rate": 7.665106707732875e-06,
+      "loss": 0.478,
+      "step": 7887
+    },
+    {
+      "epoch": 1.2954242194075505,
+      "grad_norm": 0.4763302799103845,
+      "learning_rate": 7.664773327878005e-06,
+      "loss": 0.4822,
+      "step": 7888
+    },
+    {
+      "epoch": 1.2955884466158931,
+      "grad_norm": 0.2659489183122764,
+      "learning_rate": 7.664439913650583e-06,
+      "loss": 0.4482,
+      "step": 7889
+    },
+    {
+      "epoch": 1.2957526738242358,
+      "grad_norm": 0.3211381900041531,
+      "learning_rate": 7.66410646505423e-06,
+      "loss": 0.4634,
+      "step": 7890
+    },
+    {
+      "epoch": 1.2959169010325786,
+      "grad_norm": 0.3289264312374959,
+      "learning_rate": 7.663772982092569e-06,
+      "loss": 0.4912,
+      "step": 7891
+    },
+    {
+      "epoch": 1.2960811282409213,
+      "grad_norm": 0.2874286485731477,
+      "learning_rate": 7.66343946476922e-06,
+      "loss": 0.4758,
+      "step": 7892
+    },
+    {
+      "epoch": 1.2962453554492641,
+      "grad_norm": 0.29145863615207435,
+      "learning_rate": 7.663105913087804e-06,
+      "loss": 0.4948,
+      "step": 7893
+    },
+    {
+      "epoch": 1.2964095826576068,
+      "grad_norm": 0.3151056218253248,
+      "learning_rate": 7.662772327051947e-06,
+      "loss": 0.4864,
+      "step": 7894
+    },
+    {
+      "epoch": 1.2965738098659496,
+      "grad_norm": 0.5383538039811437,
+      "learning_rate": 7.662438706665272e-06,
+      "loss": 0.4827,
+      "step": 7895
+    },
+    {
+      "epoch": 1.2967380370742922,
+      "grad_norm": 0.2904499074975075,
+      "learning_rate": 7.662105051931401e-06,
+      "loss": 0.4577,
+      "step": 7896
+    },
+    {
+      "epoch": 1.296902264282635,
+      "grad_norm": 0.3404388178106021,
+      "learning_rate": 7.661771362853958e-06,
+      "loss": 0.4973,
+      "step": 7897
+    },
+    {
+      "epoch": 1.2970664914909777,
+      "grad_norm": 0.28420348043237087,
+      "learning_rate": 7.661437639436565e-06,
+      "loss": 0.4746,
+      "step": 7898
+    },
+    {
+      "epoch": 1.2972307186993204,
+      "grad_norm": 0.31300648105957757,
+      "learning_rate": 7.661103881682851e-06,
+      "loss": 0.4718,
+      "step": 7899
+    },
+    {
+      "epoch": 1.2973949459076632,
+      "grad_norm": 0.2617389825483887,
+      "learning_rate": 7.660770089596437e-06,
+      "loss": 0.5036,
+      "step": 7900
+    },
+    {
+      "epoch": 1.297559173116006,
+      "grad_norm": 0.3558329374409586,
+      "learning_rate": 7.660436263180954e-06,
+      "loss": 0.4704,
+      "step": 7901
+    },
+    {
+      "epoch": 1.2977234003243487,
+      "grad_norm": 0.3296553098157969,
+      "learning_rate": 7.66010240244002e-06,
+      "loss": 0.4836,
+      "step": 7902
+    },
+    {
+      "epoch": 1.2978876275326914,
+      "grad_norm": 0.40087534912194606,
+      "learning_rate": 7.659768507377265e-06,
+      "loss": 0.4807,
+      "step": 7903
+    },
+    {
+      "epoch": 1.2980518547410342,
+      "grad_norm": 0.29088669339411083,
+      "learning_rate": 7.659434577996318e-06,
+      "loss": 0.4739,
+      "step": 7904
+    },
+    {
+      "epoch": 1.298216081949377,
+      "grad_norm": 0.2990246246306228,
+      "learning_rate": 7.659100614300798e-06,
+      "loss": 0.4529,
+      "step": 7905
+    },
+    {
+      "epoch": 1.2983803091577197,
+      "grad_norm": 0.29901740330002713,
+      "learning_rate": 7.658766616294343e-06,
+      "loss": 0.4725,
+      "step": 7906
+    },
+    {
+      "epoch": 1.2985445363660624,
+      "grad_norm": 0.30059464854419815,
+      "learning_rate": 7.65843258398057e-06,
+      "loss": 0.4856,
+      "step": 7907
+    },
+    {
+      "epoch": 1.2987087635744052,
+      "grad_norm": 0.6435372599150829,
+      "learning_rate": 7.658098517363115e-06,
+      "loss": 0.4717,
+      "step": 7908
+    },
+    {
+      "epoch": 1.2988729907827479,
+      "grad_norm": 0.3214157341766446,
+      "learning_rate": 7.657764416445601e-06,
+      "loss": 0.4679,
+      "step": 7909
+    },
+    {
+      "epoch": 1.2990372179910907,
+      "grad_norm": 0.324097294636231,
+      "learning_rate": 7.65743028123166e-06,
+      "loss": 0.458,
+      "step": 7910
+    },
+    {
+      "epoch": 1.2992014451994334,
+      "grad_norm": 0.3110100490491375,
+      "learning_rate": 7.657096111724917e-06,
+      "loss": 0.46,
+      "step": 7911
+    },
+    {
+      "epoch": 1.2993656724077762,
+      "grad_norm": 0.3124904806207106,
+      "learning_rate": 7.656761907929006e-06,
+      "loss": 0.4666,
+      "step": 7912
+    },
+    {
+      "epoch": 1.2995298996161189,
+      "grad_norm": 0.29953706182151885,
+      "learning_rate": 7.656427669847557e-06,
+      "loss": 0.5017,
+      "step": 7913
+    },
+    {
+      "epoch": 1.2996941268244617,
+      "grad_norm": 0.3084623975455468,
+      "learning_rate": 7.656093397484195e-06,
+      "loss": 0.4756,
+      "step": 7914
+    },
+    {
+      "epoch": 1.2998583540328044,
+      "grad_norm": 0.2995223537477973,
+      "learning_rate": 7.655759090842554e-06,
+      "loss": 0.4865,
+      "step": 7915
+    },
+    {
+      "epoch": 1.300022581241147,
+      "grad_norm": 0.34171160928676775,
+      "learning_rate": 7.655424749926265e-06,
+      "loss": 0.4924,
+      "step": 7916
+    },
+    {
+      "epoch": 1.3001868084494899,
+      "grad_norm": 0.37593951569139283,
+      "learning_rate": 7.655090374738958e-06,
+      "loss": 0.4758,
+      "step": 7917
+    },
+    {
+      "epoch": 1.3003510356578327,
+      "grad_norm": 0.27259654106749204,
+      "learning_rate": 7.654755965284266e-06,
+      "loss": 0.4855,
+      "step": 7918
+    },
+    {
+      "epoch": 1.3005152628661754,
+      "grad_norm": 0.26982803154783935,
+      "learning_rate": 7.65442152156582e-06,
+      "loss": 0.4638,
+      "step": 7919
+    },
+    {
+      "epoch": 1.300679490074518,
+      "grad_norm": 0.3358575968821194,
+      "learning_rate": 7.654087043587253e-06,
+      "loss": 0.4822,
+      "step": 7920
+    },
+    {
+      "epoch": 1.3008437172828609,
+      "grad_norm": 0.3025242716671943,
+      "learning_rate": 7.653752531352197e-06,
+      "loss": 0.4854,
+      "step": 7921
+    },
+    {
+      "epoch": 1.3010079444912037,
+      "grad_norm": 0.31977526421106844,
+      "learning_rate": 7.653417984864286e-06,
+      "loss": 0.4781,
+      "step": 7922
+    },
+    {
+      "epoch": 1.3011721716995464,
+      "grad_norm": 0.4255861102576953,
+      "learning_rate": 7.653083404127154e-06,
+      "loss": 0.4695,
+      "step": 7923
+    },
+    {
+      "epoch": 1.301336398907889,
+      "grad_norm": 0.32376807035451516,
+      "learning_rate": 7.652748789144432e-06,
+      "loss": 0.4724,
+      "step": 7924
+    },
+    {
+      "epoch": 1.3015006261162319,
+      "grad_norm": 0.2621931897315843,
+      "learning_rate": 7.652414139919758e-06,
+      "loss": 0.483,
+      "step": 7925
+    },
+    {
+      "epoch": 1.3016648533245745,
+      "grad_norm": 0.32369868128194706,
+      "learning_rate": 7.652079456456765e-06,
+      "loss": 0.4788,
+      "step": 7926
+    },
+    {
+      "epoch": 1.3018290805329173,
+      "grad_norm": 0.29915004458420025,
+      "learning_rate": 7.651744738759086e-06,
+      "loss": 0.4656,
+      "step": 7927
+    },
+    {
+      "epoch": 1.30199330774126,
+      "grad_norm": 0.3552626648407864,
+      "learning_rate": 7.65140998683036e-06,
+      "loss": 0.4942,
+      "step": 7928
+    },
+    {
+      "epoch": 1.3021575349496028,
+      "grad_norm": 0.28752600760195246,
+      "learning_rate": 7.65107520067422e-06,
+      "loss": 0.4736,
+      "step": 7929
+    },
+    {
+      "epoch": 1.3023217621579455,
+      "grad_norm": 0.3278934170298695,
+      "learning_rate": 7.650740380294304e-06,
+      "loss": 0.4824,
+      "step": 7930
+    },
+    {
+      "epoch": 1.3024859893662883,
+      "grad_norm": 0.29068672152221015,
+      "learning_rate": 7.650405525694247e-06,
+      "loss": 0.4952,
+      "step": 7931
+    },
+    {
+      "epoch": 1.302650216574631,
+      "grad_norm": 0.36659481274056716,
+      "learning_rate": 7.650070636877686e-06,
+      "loss": 0.4905,
+      "step": 7932
+    },
+    {
+      "epoch": 1.3028144437829736,
+      "grad_norm": 0.318759886416517,
+      "learning_rate": 7.64973571384826e-06,
+      "loss": 0.4993,
+      "step": 7933
+    },
+    {
+      "epoch": 1.3029786709913165,
+      "grad_norm": 0.4421733290213734,
+      "learning_rate": 7.649400756609603e-06,
+      "loss": 0.4704,
+      "step": 7934
+    },
+    {
+      "epoch": 1.3031428981996593,
+      "grad_norm": 0.4854735090770185,
+      "learning_rate": 7.649065765165356e-06,
+      "loss": 0.4819,
+      "step": 7935
+    },
+    {
+      "epoch": 1.303307125408002,
+      "grad_norm": 0.3538652860607975,
+      "learning_rate": 7.648730739519159e-06,
+      "loss": 0.4599,
+      "step": 7936
+    },
+    {
+      "epoch": 1.3034713526163446,
+      "grad_norm": 0.2711419157816641,
+      "learning_rate": 7.648395679674645e-06,
+      "loss": 0.4653,
+      "step": 7937
+    },
+    {
+      "epoch": 1.3036355798246875,
+      "grad_norm": 0.28791299089951694,
+      "learning_rate": 7.648060585635457e-06,
+      "loss": 0.467,
+      "step": 7938
+    },
+    {
+      "epoch": 1.3037998070330303,
+      "grad_norm": 0.3147023306380566,
+      "learning_rate": 7.647725457405235e-06,
+      "loss": 0.4785,
+      "step": 7939
+    },
+    {
+      "epoch": 1.303964034241373,
+      "grad_norm": 0.33694405249481874,
+      "learning_rate": 7.647390294987618e-06,
+      "loss": 0.4759,
+      "step": 7940
+    },
+    {
+      "epoch": 1.3041282614497156,
+      "grad_norm": 0.3704706312364689,
+      "learning_rate": 7.647055098386243e-06,
+      "loss": 0.4731,
+      "step": 7941
+    },
+    {
+      "epoch": 1.3042924886580585,
+      "grad_norm": 0.2981416710176911,
+      "learning_rate": 7.646719867604756e-06,
+      "loss": 0.4727,
+      "step": 7942
+    },
+    {
+      "epoch": 1.304456715866401,
+      "grad_norm": 0.4236048668542745,
+      "learning_rate": 7.646384602646794e-06,
+      "loss": 0.491,
+      "step": 7943
+    },
+    {
+      "epoch": 1.304620943074744,
+      "grad_norm": 0.25920675415941064,
+      "learning_rate": 7.646049303516001e-06,
+      "loss": 0.4647,
+      "step": 7944
+    },
+    {
+      "epoch": 1.3047851702830866,
+      "grad_norm": 0.37288961353740263,
+      "learning_rate": 7.645713970216015e-06,
+      "loss": 0.4729,
+      "step": 7945
+    },
+    {
+      "epoch": 1.3049493974914295,
+      "grad_norm": 0.30280530467800054,
+      "learning_rate": 7.645378602750481e-06,
+      "loss": 0.4951,
+      "step": 7946
+    },
+    {
+      "epoch": 1.305113624699772,
+      "grad_norm": 0.32637097512312524,
+      "learning_rate": 7.645043201123042e-06,
+      "loss": 0.4863,
+      "step": 7947
+    },
+    {
+      "epoch": 1.305277851908115,
+      "grad_norm": 0.46033399936430586,
+      "learning_rate": 7.64470776533734e-06,
+      "loss": 0.4814,
+      "step": 7948
+    },
+    {
+      "epoch": 1.3054420791164576,
+      "grad_norm": 0.31590207279548094,
+      "learning_rate": 7.644372295397015e-06,
+      "loss": 0.4821,
+      "step": 7949
+    },
+    {
+      "epoch": 1.3056063063248002,
+      "grad_norm": 0.31853698788592905,
+      "learning_rate": 7.644036791305715e-06,
+      "loss": 0.4544,
+      "step": 7950
+    },
+    {
+      "epoch": 1.305770533533143,
+      "grad_norm": 0.28684806650613004,
+      "learning_rate": 7.643701253067082e-06,
+      "loss": 0.4912,
+      "step": 7951
+    },
+    {
+      "epoch": 1.305934760741486,
+      "grad_norm": 0.43154634307997536,
+      "learning_rate": 7.64336568068476e-06,
+      "loss": 0.4867,
+      "step": 7952
+    },
+    {
+      "epoch": 1.3060989879498286,
+      "grad_norm": 0.3614722309268757,
+      "learning_rate": 7.643030074162395e-06,
+      "loss": 0.4644,
+      "step": 7953
+    },
+    {
+      "epoch": 1.3062632151581712,
+      "grad_norm": 0.2847346531183879,
+      "learning_rate": 7.64269443350363e-06,
+      "loss": 0.4463,
+      "step": 7954
+    },
+    {
+      "epoch": 1.306427442366514,
+      "grad_norm": 0.33721951652964166,
+      "learning_rate": 7.642358758712112e-06,
+      "loss": 0.4786,
+      "step": 7955
+    },
+    {
+      "epoch": 1.306591669574857,
+      "grad_norm": 0.29914397820153493,
+      "learning_rate": 7.642023049791485e-06,
+      "loss": 0.4622,
+      "step": 7956
+    },
+    {
+      "epoch": 1.3067558967831996,
+      "grad_norm": 0.2687785996804088,
+      "learning_rate": 7.641687306745399e-06,
+      "loss": 0.4709,
+      "step": 7957
+    },
+    {
+      "epoch": 1.3069201239915422,
+      "grad_norm": 0.31963062378427437,
+      "learning_rate": 7.641351529577494e-06,
+      "loss": 0.468,
+      "step": 7958
+    },
+    {
+      "epoch": 1.307084351199885,
+      "grad_norm": 0.545023022599308,
+      "learning_rate": 7.641015718291425e-06,
+      "loss": 0.4806,
+      "step": 7959
+    },
+    {
+      "epoch": 1.3072485784082277,
+      "grad_norm": 0.395362345706082,
+      "learning_rate": 7.640679872890832e-06,
+      "loss": 0.4628,
+      "step": 7960
+    },
+    {
+      "epoch": 1.3074128056165706,
+      "grad_norm": 0.466981966839366,
+      "learning_rate": 7.640343993379368e-06,
+      "loss": 0.4682,
+      "step": 7961
+    },
+    {
+      "epoch": 1.3075770328249132,
+      "grad_norm": 0.28939268969208437,
+      "learning_rate": 7.640008079760676e-06,
+      "loss": 0.4592,
+      "step": 7962
+    },
+    {
+      "epoch": 1.307741260033256,
+      "grad_norm": 0.33538113413961285,
+      "learning_rate": 7.639672132038407e-06,
+      "loss": 0.4806,
+      "step": 7963
+    },
+    {
+      "epoch": 1.3079054872415987,
+      "grad_norm": 0.32099795596270486,
+      "learning_rate": 7.639336150216211e-06,
+      "loss": 0.482,
+      "step": 7964
+    },
+    {
+      "epoch": 1.3080697144499416,
+      "grad_norm": 0.553802750895834,
+      "learning_rate": 7.639000134297735e-06,
+      "loss": 0.4876,
+      "step": 7965
+    },
+    {
+      "epoch": 1.3082339416582842,
+      "grad_norm": 0.34601641793654353,
+      "learning_rate": 7.638664084286629e-06,
+      "loss": 0.4836,
+      "step": 7966
+    },
+    {
+      "epoch": 1.3083981688666269,
+      "grad_norm": 0.2989365043791791,
+      "learning_rate": 7.638328000186545e-06,
+      "loss": 0.4683,
+      "step": 7967
+    },
+    {
+      "epoch": 1.3085623960749697,
+      "grad_norm": 0.3029078532588946,
+      "learning_rate": 7.63799188200113e-06,
+      "loss": 0.4701,
+      "step": 7968
+    },
+    {
+      "epoch": 1.3087266232833126,
+      "grad_norm": 0.4104382751629648,
+      "learning_rate": 7.637655729734036e-06,
+      "loss": 0.4682,
+      "step": 7969
+    },
+    {
+      "epoch": 1.3088908504916552,
+      "grad_norm": 0.3401069351664016,
+      "learning_rate": 7.637319543388913e-06,
+      "loss": 0.4816,
+      "step": 7970
+    },
+    {
+      "epoch": 1.3090550776999978,
+      "grad_norm": 0.36051234077198613,
+      "learning_rate": 7.63698332296941e-06,
+      "loss": 0.4704,
+      "step": 7971
+    },
+    {
+      "epoch": 1.3092193049083407,
+      "grad_norm": 0.2680635234976543,
+      "learning_rate": 7.636647068479188e-06,
+      "loss": 0.4862,
+      "step": 7972
+    },
+    {
+      "epoch": 1.3093835321166836,
+      "grad_norm": 0.35280649807851294,
+      "learning_rate": 7.636310779921889e-06,
+      "loss": 0.4844,
+      "step": 7973
+    },
+    {
+      "epoch": 1.3095477593250262,
+      "grad_norm": 0.3241893700421079,
+      "learning_rate": 7.63597445730117e-06,
+      "loss": 0.482,
+      "step": 7974
+    },
+    {
+      "epoch": 1.3097119865333688,
+      "grad_norm": 0.25658079785257215,
+      "learning_rate": 7.635638100620683e-06,
+      "loss": 0.4656,
+      "step": 7975
+    },
+    {
+      "epoch": 1.3098762137417117,
+      "grad_norm": 0.340790886688055,
+      "learning_rate": 7.63530170988408e-06,
+      "loss": 0.4806,
+      "step": 7976
+    },
+    {
+      "epoch": 1.3100404409500543,
+      "grad_norm": 0.29263625469920773,
+      "learning_rate": 7.634965285095018e-06,
+      "loss": 0.4735,
+      "step": 7977
+    },
+    {
+      "epoch": 1.3102046681583972,
+      "grad_norm": 0.31652479596424954,
+      "learning_rate": 7.634628826257148e-06,
+      "loss": 0.4759,
+      "step": 7978
+    },
+    {
+      "epoch": 1.3103688953667398,
+      "grad_norm": 0.3299859845201978,
+      "learning_rate": 7.634292333374123e-06,
+      "loss": 0.4708,
+      "step": 7979
+    },
+    {
+      "epoch": 1.3105331225750827,
+      "grad_norm": 0.272066769077068,
+      "learning_rate": 7.633955806449603e-06,
+      "loss": 0.4941,
+      "step": 7980
+    },
+    {
+      "epoch": 1.3106973497834253,
+      "grad_norm": 0.31916471197677193,
+      "learning_rate": 7.633619245487237e-06,
+      "loss": 0.4492,
+      "step": 7981
+    },
+    {
+      "epoch": 1.3108615769917682,
+      "grad_norm": 0.41008229525586465,
+      "learning_rate": 7.633282650490684e-06,
+      "loss": 0.4817,
+      "step": 7982
+    },
+    {
+      "epoch": 1.3110258042001108,
+      "grad_norm": 0.46510580333747115,
+      "learning_rate": 7.632946021463597e-06,
+      "loss": 0.4693,
+      "step": 7983
+    },
+    {
+      "epoch": 1.3111900314084535,
+      "grad_norm": 0.26202627385773686,
+      "learning_rate": 7.632609358409637e-06,
+      "loss": 0.4582,
+      "step": 7984
+    },
+    {
+      "epoch": 1.3113542586167963,
+      "grad_norm": 0.5539807098934775,
+      "learning_rate": 7.632272661332454e-06,
+      "loss": 0.4755,
+      "step": 7985
+    },
+    {
+      "epoch": 1.3115184858251392,
+      "grad_norm": 0.28471144410474203,
+      "learning_rate": 7.63193593023571e-06,
+      "loss": 0.4628,
+      "step": 7986
+    },
+    {
+      "epoch": 1.3116827130334818,
+      "grad_norm": 0.2806760601317828,
+      "learning_rate": 7.631599165123058e-06,
+      "loss": 0.4716,
+      "step": 7987
+    },
+    {
+      "epoch": 1.3118469402418245,
+      "grad_norm": 0.3691693460679819,
+      "learning_rate": 7.631262365998161e-06,
+      "loss": 0.4798,
+      "step": 7988
+    },
+    {
+      "epoch": 1.3120111674501673,
+      "grad_norm": 0.2821735353713559,
+      "learning_rate": 7.630925532864672e-06,
+      "loss": 0.4818,
+      "step": 7989
+    },
+    {
+      "epoch": 1.3121753946585102,
+      "grad_norm": 0.34444638448337844,
+      "learning_rate": 7.630588665726253e-06,
+      "loss": 0.4925,
+      "step": 7990
+    },
+    {
+      "epoch": 1.3123396218668528,
+      "grad_norm": 0.33577779366945215,
+      "learning_rate": 7.63025176458656e-06,
+      "loss": 0.4806,
+      "step": 7991
+    },
+    {
+      "epoch": 1.3125038490751955,
+      "grad_norm": 0.30897078274720075,
+      "learning_rate": 7.629914829449253e-06,
+      "loss": 0.4806,
+      "step": 7992
+    },
+    {
+      "epoch": 1.3126680762835383,
+      "grad_norm": 0.390373369768928,
+      "learning_rate": 7.629577860317991e-06,
+      "loss": 0.5017,
+      "step": 7993
+    },
+    {
+      "epoch": 1.312832303491881,
+      "grad_norm": 0.3872526842133121,
+      "learning_rate": 7.6292408571964354e-06,
+      "loss": 0.4827,
+      "step": 7994
+    },
+    {
+      "epoch": 1.3129965307002238,
+      "grad_norm": 0.32955819048590035,
+      "learning_rate": 7.6289038200882445e-06,
+      "loss": 0.4687,
+      "step": 7995
+    },
+    {
+      "epoch": 1.3131607579085665,
+      "grad_norm": 0.31306415443858815,
+      "learning_rate": 7.628566748997081e-06,
+      "loss": 0.4652,
+      "step": 7996
+    },
+    {
+      "epoch": 1.3133249851169093,
+      "grad_norm": 0.34561030406592913,
+      "learning_rate": 7.628229643926603e-06,
+      "loss": 0.475,
+      "step": 7997
+    },
+    {
+      "epoch": 1.313489212325252,
+      "grad_norm": 0.35034799088508667,
+      "learning_rate": 7.627892504880474e-06,
+      "loss": 0.4701,
+      "step": 7998
+    },
+    {
+      "epoch": 1.3136534395335948,
+      "grad_norm": 0.2711573129803233,
+      "learning_rate": 7.627555331862355e-06,
+      "loss": 0.4778,
+      "step": 7999
+    },
+    {
+      "epoch": 1.3138176667419375,
+      "grad_norm": 0.376753307536093,
+      "learning_rate": 7.627218124875908e-06,
+      "loss": 0.4861,
+      "step": 8000
+    },
+    {
+      "epoch": 1.31398189395028,
+      "grad_norm": 0.3478626411436554,
+      "learning_rate": 7.626880883924795e-06,
+      "loss": 0.4648,
+      "step": 8001
+    },
+    {
+      "epoch": 1.314146121158623,
+      "grad_norm": 0.29665293395553316,
+      "learning_rate": 7.62654360901268e-06,
+      "loss": 0.484,
+      "step": 8002
+    },
+    {
+      "epoch": 1.3143103483669658,
+      "grad_norm": 0.3763583038100201,
+      "learning_rate": 7.626206300143224e-06,
+      "loss": 0.4835,
+      "step": 8003
+    },
+    {
+      "epoch": 1.3144745755753084,
+      "grad_norm": 0.33950295810942993,
+      "learning_rate": 7.625868957320092e-06,
+      "loss": 0.4712,
+      "step": 8004
+    },
+    {
+      "epoch": 1.314638802783651,
+      "grad_norm": 0.2764303498324206,
+      "learning_rate": 7.62553158054695e-06,
+      "loss": 0.4672,
+      "step": 8005
+    },
+    {
+      "epoch": 1.314803029991994,
+      "grad_norm": 0.2943201514727324,
+      "learning_rate": 7.625194169827458e-06,
+      "loss": 0.488,
+      "step": 8006
+    },
+    {
+      "epoch": 1.3149672572003368,
+      "grad_norm": 0.3697002180213209,
+      "learning_rate": 7.6248567251652825e-06,
+      "loss": 0.4727,
+      "step": 8007
+    },
+    {
+      "epoch": 1.3151314844086794,
+      "grad_norm": 0.3757351595733627,
+      "learning_rate": 7.6245192465640885e-06,
+      "loss": 0.4865,
+      "step": 8008
+    },
+    {
+      "epoch": 1.315295711617022,
+      "grad_norm": 0.3738711600535506,
+      "learning_rate": 7.624181734027541e-06,
+      "loss": 0.4717,
+      "step": 8009
+    },
+    {
+      "epoch": 1.315459938825365,
+      "grad_norm": 0.28050608814768224,
+      "learning_rate": 7.623844187559308e-06,
+      "loss": 0.5125,
+      "step": 8010
+    },
+    {
+      "epoch": 1.3156241660337076,
+      "grad_norm": 0.2919003234877574,
+      "learning_rate": 7.623506607163052e-06,
+      "loss": 0.4539,
+      "step": 8011
+    },
+    {
+      "epoch": 1.3157883932420504,
+      "grad_norm": 0.3622981290018002,
+      "learning_rate": 7.6231689928424415e-06,
+      "loss": 0.4625,
+      "step": 8012
+    },
+    {
+      "epoch": 1.315952620450393,
+      "grad_norm": 0.3516883118340059,
+      "learning_rate": 7.622831344601143e-06,
+      "loss": 0.4845,
+      "step": 8013
+    },
+    {
+      "epoch": 1.316116847658736,
+      "grad_norm": 0.29660929220781496,
+      "learning_rate": 7.622493662442823e-06,
+      "loss": 0.4642,
+      "step": 8014
+    },
+    {
+      "epoch": 1.3162810748670786,
+      "grad_norm": 1.2368299556551743,
+      "learning_rate": 7.622155946371151e-06,
+      "loss": 0.4798,
+      "step": 8015
+    },
+    {
+      "epoch": 1.3164453020754214,
+      "grad_norm": 0.29717759444454805,
+      "learning_rate": 7.621818196389793e-06,
+      "loss": 0.4756,
+      "step": 8016
+    },
+    {
+      "epoch": 1.316609529283764,
+      "grad_norm": 0.3918627797621546,
+      "learning_rate": 7.621480412502418e-06,
+      "loss": 0.4788,
+      "step": 8017
+    },
+    {
+      "epoch": 1.3167737564921067,
+      "grad_norm": 0.29543561155814907,
+      "learning_rate": 7.621142594712694e-06,
+      "loss": 0.4981,
+      "step": 8018
+    },
+    {
+      "epoch": 1.3169379837004496,
+      "grad_norm": 0.37159328406059056,
+      "learning_rate": 7.620804743024291e-06,
+      "loss": 0.4948,
+      "step": 8019
+    },
+    {
+      "epoch": 1.3171022109087924,
+      "grad_norm": 0.2783048589916877,
+      "learning_rate": 7.620466857440879e-06,
+      "loss": 0.4717,
+      "step": 8020
+    },
+    {
+      "epoch": 1.317266438117135,
+      "grad_norm": 0.306567811369454,
+      "learning_rate": 7.620128937966125e-06,
+      "loss": 0.4724,
+      "step": 8021
+    },
+    {
+      "epoch": 1.3174306653254777,
+      "grad_norm": 0.29327978868608195,
+      "learning_rate": 7.619790984603702e-06,
+      "loss": 0.4633,
+      "step": 8022
+    },
+    {
+      "epoch": 1.3175948925338206,
+      "grad_norm": 0.3917006265643016,
+      "learning_rate": 7.61945299735728e-06,
+      "loss": 0.4604,
+      "step": 8023
+    },
+    {
+      "epoch": 1.3177591197421634,
+      "grad_norm": 0.2667837817112557,
+      "learning_rate": 7.619114976230528e-06,
+      "loss": 0.4541,
+      "step": 8024
+    },
+    {
+      "epoch": 1.317923346950506,
+      "grad_norm": 0.35010204950872126,
+      "learning_rate": 7.6187769212271194e-06,
+      "loss": 0.4935,
+      "step": 8025
+    },
+    {
+      "epoch": 1.3180875741588487,
+      "grad_norm": 0.29046284897377767,
+      "learning_rate": 7.618438832350725e-06,
+      "loss": 0.4803,
+      "step": 8026
+    },
+    {
+      "epoch": 1.3182518013671916,
+      "grad_norm": 0.33231245117304176,
+      "learning_rate": 7.618100709605017e-06,
+      "loss": 0.469,
+      "step": 8027
+    },
+    {
+      "epoch": 1.3184160285755342,
+      "grad_norm": 0.29343771355212916,
+      "learning_rate": 7.617762552993667e-06,
+      "loss": 0.4845,
+      "step": 8028
+    },
+    {
+      "epoch": 1.318580255783877,
+      "grad_norm": 0.3313127989483271,
+      "learning_rate": 7.617424362520349e-06,
+      "loss": 0.4435,
+      "step": 8029
+    },
+    {
+      "epoch": 1.3187444829922197,
+      "grad_norm": 0.28290016169336146,
+      "learning_rate": 7.617086138188733e-06,
+      "loss": 0.4678,
+      "step": 8030
+    },
+    {
+      "epoch": 1.3189087102005626,
+      "grad_norm": 0.38666467517347725,
+      "learning_rate": 7.616747880002497e-06,
+      "loss": 0.4674,
+      "step": 8031
+    },
+    {
+      "epoch": 1.3190729374089052,
+      "grad_norm": 0.356353790147301,
+      "learning_rate": 7.616409587965312e-06,
+      "loss": 0.4725,
+      "step": 8032
+    },
+    {
+      "epoch": 1.319237164617248,
+      "grad_norm": 0.40378474599862085,
+      "learning_rate": 7.616071262080853e-06,
+      "loss": 0.4639,
+      "step": 8033
+    },
+    {
+      "epoch": 1.3194013918255907,
+      "grad_norm": 0.3306098327327851,
+      "learning_rate": 7.6157329023527925e-06,
+      "loss": 0.4802,
+      "step": 8034
+    },
+    {
+      "epoch": 1.3195656190339333,
+      "grad_norm": 0.35127743283140145,
+      "learning_rate": 7.61539450878481e-06,
+      "loss": 0.4667,
+      "step": 8035
+    },
+    {
+      "epoch": 1.3197298462422762,
+      "grad_norm": 0.4309638643922434,
+      "learning_rate": 7.615056081380577e-06,
+      "loss": 0.4838,
+      "step": 8036
+    },
+    {
+      "epoch": 1.319894073450619,
+      "grad_norm": 0.3355719844565293,
+      "learning_rate": 7.6147176201437695e-06,
+      "loss": 0.4957,
+      "step": 8037
+    },
+    {
+      "epoch": 1.3200583006589617,
+      "grad_norm": 0.5136141703832557,
+      "learning_rate": 7.614379125078063e-06,
+      "loss": 0.4786,
+      "step": 8038
+    },
+    {
+      "epoch": 1.3202225278673043,
+      "grad_norm": 0.4938624681910936,
+      "learning_rate": 7.614040596187138e-06,
+      "loss": 0.4649,
+      "step": 8039
+    },
+    {
+      "epoch": 1.3203867550756472,
+      "grad_norm": 0.3426636485394977,
+      "learning_rate": 7.613702033474667e-06,
+      "loss": 0.4791,
+      "step": 8040
+    },
+    {
+      "epoch": 1.32055098228399,
+      "grad_norm": 0.29924421972371945,
+      "learning_rate": 7.613363436944328e-06,
+      "loss": 0.4634,
+      "step": 8041
+    },
+    {
+      "epoch": 1.3207152094923327,
+      "grad_norm": 0.37481981284715726,
+      "learning_rate": 7.613024806599799e-06,
+      "loss": 0.4718,
+      "step": 8042
+    },
+    {
+      "epoch": 1.3208794367006753,
+      "grad_norm": 0.2831538659040023,
+      "learning_rate": 7.612686142444757e-06,
+      "loss": 0.4685,
+      "step": 8043
+    },
+    {
+      "epoch": 1.3210436639090182,
+      "grad_norm": 0.29528741651177665,
+      "learning_rate": 7.612347444482883e-06,
+      "loss": 0.4519,
+      "step": 8044
+    },
+    {
+      "epoch": 1.3212078911173608,
+      "grad_norm": 0.34103817691768396,
+      "learning_rate": 7.612008712717853e-06,
+      "loss": 0.4884,
+      "step": 8045
+    },
+    {
+      "epoch": 1.3213721183257037,
+      "grad_norm": 0.32873295054672647,
+      "learning_rate": 7.611669947153346e-06,
+      "loss": 0.4951,
+      "step": 8046
+    },
+    {
+      "epoch": 1.3215363455340463,
+      "grad_norm": 0.3487948323682672,
+      "learning_rate": 7.611331147793042e-06,
+      "loss": 0.4863,
+      "step": 8047
+    },
+    {
+      "epoch": 1.3217005727423892,
+      "grad_norm": 0.3518588634886226,
+      "learning_rate": 7.610992314640621e-06,
+      "loss": 0.4878,
+      "step": 8048
+    },
+    {
+      "epoch": 1.3218647999507318,
+      "grad_norm": 0.3851542782176887,
+      "learning_rate": 7.610653447699763e-06,
+      "loss": 0.4799,
+      "step": 8049
+    },
+    {
+      "epoch": 1.3220290271590747,
+      "grad_norm": 0.30015345022801126,
+      "learning_rate": 7.610314546974146e-06,
+      "loss": 0.4742,
+      "step": 8050
+    },
+    {
+      "epoch": 1.3221932543674173,
+      "grad_norm": 0.3396771981317586,
+      "learning_rate": 7.6099756124674555e-06,
+      "loss": 0.4692,
+      "step": 8051
+    },
+    {
+      "epoch": 1.32235748157576,
+      "grad_norm": 0.37353944059171346,
+      "learning_rate": 7.6096366441833686e-06,
+      "loss": 0.4923,
+      "step": 8052
+    },
+    {
+      "epoch": 1.3225217087841028,
+      "grad_norm": 0.3611586443903467,
+      "learning_rate": 7.609297642125568e-06,
+      "loss": 0.4695,
+      "step": 8053
+    },
+    {
+      "epoch": 1.3226859359924457,
+      "grad_norm": 0.3391887708192555,
+      "learning_rate": 7.6089586062977375e-06,
+      "loss": 0.4752,
+      "step": 8054
+    },
+    {
+      "epoch": 1.3228501632007883,
+      "grad_norm": 0.3614387838925698,
+      "learning_rate": 7.608619536703557e-06,
+      "loss": 0.4739,
+      "step": 8055
+    },
+    {
+      "epoch": 1.323014390409131,
+      "grad_norm": 0.3365756410105476,
+      "learning_rate": 7.608280433346709e-06,
+      "loss": 0.4798,
+      "step": 8056
+    },
+    {
+      "epoch": 1.3231786176174738,
+      "grad_norm": 1.238861339205873,
+      "learning_rate": 7.607941296230878e-06,
+      "loss": 0.4753,
+      "step": 8057
+    },
+    {
+      "epoch": 1.3233428448258167,
+      "grad_norm": 0.3307128371330611,
+      "learning_rate": 7.6076021253597465e-06,
+      "loss": 0.4914,
+      "step": 8058
+    },
+    {
+      "epoch": 1.3235070720341593,
+      "grad_norm": 0.4128137716597536,
+      "learning_rate": 7.607262920736999e-06,
+      "loss": 0.4686,
+      "step": 8059
+    },
+    {
+      "epoch": 1.323671299242502,
+      "grad_norm": 0.32301645197658924,
+      "learning_rate": 7.606923682366318e-06,
+      "loss": 0.4168,
+      "step": 8060
+    },
+    {
+      "epoch": 1.3238355264508448,
+      "grad_norm": 0.2996180664900655,
+      "learning_rate": 7.60658441025139e-06,
+      "loss": 0.4777,
+      "step": 8061
+    },
+    {
+      "epoch": 1.3239997536591874,
+      "grad_norm": 0.3167191179377588,
+      "learning_rate": 7.606245104395898e-06,
+      "loss": 0.4795,
+      "step": 8062
+    },
+    {
+      "epoch": 1.3241639808675303,
+      "grad_norm": 0.32137763185269175,
+      "learning_rate": 7.605905764803528e-06,
+      "loss": 0.4715,
+      "step": 8063
+    },
+    {
+      "epoch": 1.324328208075873,
+      "grad_norm": 0.47425814179048575,
+      "learning_rate": 7.6055663914779665e-06,
+      "loss": 0.477,
+      "step": 8064
+    },
+    {
+      "epoch": 1.3244924352842158,
+      "grad_norm": 0.38687364650229183,
+      "learning_rate": 7.605226984422899e-06,
+      "loss": 0.47,
+      "step": 8065
+    },
+    {
+      "epoch": 1.3246566624925584,
+      "grad_norm": 0.29405401978587575,
+      "learning_rate": 7.60488754364201e-06,
+      "loss": 0.4682,
+      "step": 8066
+    },
+    {
+      "epoch": 1.3248208897009013,
+      "grad_norm": 0.3612032326513613,
+      "learning_rate": 7.604548069138988e-06,
+      "loss": 0.4877,
+      "step": 8067
+    },
+    {
+      "epoch": 1.324985116909244,
+      "grad_norm": 0.3739762267080125,
+      "learning_rate": 7.604208560917519e-06,
+      "loss": 0.4728,
+      "step": 8068
+    },
+    {
+      "epoch": 1.3251493441175866,
+      "grad_norm": 0.33058668071605135,
+      "learning_rate": 7.603869018981292e-06,
+      "loss": 0.4707,
+      "step": 8069
+    },
+    {
+      "epoch": 1.3253135713259294,
+      "grad_norm": 0.5793876431825817,
+      "learning_rate": 7.603529443333993e-06,
+      "loss": 0.491,
+      "step": 8070
+    },
+    {
+      "epoch": 1.3254777985342723,
+      "grad_norm": 0.2881885125717379,
+      "learning_rate": 7.603189833979311e-06,
+      "loss": 0.4674,
+      "step": 8071
+    },
+    {
+      "epoch": 1.325642025742615,
+      "grad_norm": 0.2923464384400088,
+      "learning_rate": 7.602850190920933e-06,
+      "loss": 0.4742,
+      "step": 8072
+    },
+    {
+      "epoch": 1.3258062529509576,
+      "grad_norm": 0.32004321913986356,
+      "learning_rate": 7.602510514162551e-06,
+      "loss": 0.4909,
+      "step": 8073
+    },
+    {
+      "epoch": 1.3259704801593004,
+      "grad_norm": 0.3242085527775499,
+      "learning_rate": 7.602170803707852e-06,
+      "loss": 0.4895,
+      "step": 8074
+    },
+    {
+      "epoch": 1.3261347073676433,
+      "grad_norm": 0.31932451799372075,
+      "learning_rate": 7.601831059560525e-06,
+      "loss": 0.4472,
+      "step": 8075
+    },
+    {
+      "epoch": 1.326298934575986,
+      "grad_norm": 0.43480385093250357,
+      "learning_rate": 7.6014912817242615e-06,
+      "loss": 0.5229,
+      "step": 8076
+    },
+    {
+      "epoch": 1.3264631617843285,
+      "grad_norm": 0.28575102479694664,
+      "learning_rate": 7.601151470202752e-06,
+      "loss": 0.4712,
+      "step": 8077
+    },
+    {
+      "epoch": 1.3266273889926714,
+      "grad_norm": 0.3413516866373773,
+      "learning_rate": 7.600811624999685e-06,
+      "loss": 0.4893,
+      "step": 8078
+    },
+    {
+      "epoch": 1.326791616201014,
+      "grad_norm": 0.4762227758293843,
+      "learning_rate": 7.600471746118754e-06,
+      "loss": 0.4593,
+      "step": 8079
+    },
+    {
+      "epoch": 1.326955843409357,
+      "grad_norm": 0.36642531096793896,
+      "learning_rate": 7.600131833563648e-06,
+      "loss": 0.4828,
+      "step": 8080
+    },
+    {
+      "epoch": 1.3271200706176995,
+      "grad_norm": 0.24844728932647275,
+      "learning_rate": 7.599791887338061e-06,
+      "loss": 0.4486,
+      "step": 8081
+    },
+    {
+      "epoch": 1.3272842978260424,
+      "grad_norm": 0.2761053375616054,
+      "learning_rate": 7.599451907445685e-06,
+      "loss": 0.4666,
+      "step": 8082
+    },
+    {
+      "epoch": 1.327448525034385,
+      "grad_norm": 0.35201886560818063,
+      "learning_rate": 7.599111893890211e-06,
+      "loss": 0.4635,
+      "step": 8083
+    },
+    {
+      "epoch": 1.327612752242728,
+      "grad_norm": 0.33488597740115933,
+      "learning_rate": 7.598771846675333e-06,
+      "loss": 0.4694,
+      "step": 8084
+    },
+    {
+      "epoch": 1.3277769794510705,
+      "grad_norm": 2.0433814286588077,
+      "learning_rate": 7.598431765804745e-06,
+      "loss": 0.472,
+      "step": 8085
+    },
+    {
+      "epoch": 1.3279412066594132,
+      "grad_norm": 0.28476359738830287,
+      "learning_rate": 7.598091651282138e-06,
+      "loss": 0.4753,
+      "step": 8086
+    },
+    {
+      "epoch": 1.328105433867756,
+      "grad_norm": 0.3091484184216086,
+      "learning_rate": 7.597751503111208e-06,
+      "loss": 0.4704,
+      "step": 8087
+    },
+    {
+      "epoch": 1.328269661076099,
+      "grad_norm": 0.2939492958850838,
+      "learning_rate": 7.597411321295649e-06,
+      "loss": 0.4881,
+      "step": 8088
+    },
+    {
+      "epoch": 1.3284338882844415,
+      "grad_norm": 0.2938888929723579,
+      "learning_rate": 7.597071105839155e-06,
+      "loss": 0.4922,
+      "step": 8089
+    },
+    {
+      "epoch": 1.3285981154927842,
+      "grad_norm": 0.30286546203247877,
+      "learning_rate": 7.596730856745423e-06,
+      "loss": 0.4762,
+      "step": 8090
+    },
+    {
+      "epoch": 1.328762342701127,
+      "grad_norm": 0.30359089033544046,
+      "learning_rate": 7.5963905740181465e-06,
+      "loss": 0.4806,
+      "step": 8091
+    },
+    {
+      "epoch": 1.3289265699094697,
+      "grad_norm": 0.3143376216387629,
+      "learning_rate": 7.5960502576610206e-06,
+      "loss": 0.4909,
+      "step": 8092
+    },
+    {
+      "epoch": 1.3290907971178125,
+      "grad_norm": 0.38876383498655837,
+      "learning_rate": 7.5957099076777445e-06,
+      "loss": 0.4568,
+      "step": 8093
+    },
+    {
+      "epoch": 1.3292550243261552,
+      "grad_norm": 0.36562505267623785,
+      "learning_rate": 7.595369524072013e-06,
+      "loss": 0.4806,
+      "step": 8094
+    },
+    {
+      "epoch": 1.329419251534498,
+      "grad_norm": 0.34457728082214156,
+      "learning_rate": 7.595029106847523e-06,
+      "loss": 0.4746,
+      "step": 8095
+    },
+    {
+      "epoch": 1.3295834787428407,
+      "grad_norm": 0.3324520354871399,
+      "learning_rate": 7.59468865600797e-06,
+      "loss": 0.4903,
+      "step": 8096
+    },
+    {
+      "epoch": 1.3297477059511835,
+      "grad_norm": 0.4155866763230899,
+      "learning_rate": 7.594348171557055e-06,
+      "loss": 0.4703,
+      "step": 8097
+    },
+    {
+      "epoch": 1.3299119331595262,
+      "grad_norm": 0.2969915001045424,
+      "learning_rate": 7.594007653498475e-06,
+      "loss": 0.4858,
+      "step": 8098
+    },
+    {
+      "epoch": 1.330076160367869,
+      "grad_norm": 0.30978801552869756,
+      "learning_rate": 7.593667101835927e-06,
+      "loss": 0.4635,
+      "step": 8099
+    },
+    {
+      "epoch": 1.3302403875762117,
+      "grad_norm": 0.34181601930545924,
+      "learning_rate": 7.593326516573111e-06,
+      "loss": 0.4701,
+      "step": 8100
+    },
+    {
+      "epoch": 1.3304046147845545,
+      "grad_norm": 0.42274058666425973,
+      "learning_rate": 7.592985897713724e-06,
+      "loss": 0.4826,
+      "step": 8101
+    },
+    {
+      "epoch": 1.3305688419928972,
+      "grad_norm": 0.492798755390736,
+      "learning_rate": 7.592645245261468e-06,
+      "loss": 0.4745,
+      "step": 8102
+    },
+    {
+      "epoch": 1.3307330692012398,
+      "grad_norm": 0.2959789659905488,
+      "learning_rate": 7.5923045592200425e-06,
+      "loss": 0.4788,
+      "step": 8103
+    },
+    {
+      "epoch": 1.3308972964095827,
+      "grad_norm": 0.2900045872919727,
+      "learning_rate": 7.591963839593147e-06,
+      "loss": 0.4897,
+      "step": 8104
+    },
+    {
+      "epoch": 1.3310615236179255,
+      "grad_norm": 0.28604574921354087,
+      "learning_rate": 7.591623086384482e-06,
+      "loss": 0.4801,
+      "step": 8105
+    },
+    {
+      "epoch": 1.3312257508262682,
+      "grad_norm": 0.2883666824407346,
+      "learning_rate": 7.5912822995977485e-06,
+      "loss": 0.4641,
+      "step": 8106
+    },
+    {
+      "epoch": 1.3313899780346108,
+      "grad_norm": 0.3693527994487577,
+      "learning_rate": 7.590941479236647e-06,
+      "loss": 0.497,
+      "step": 8107
+    },
+    {
+      "epoch": 1.3315542052429536,
+      "grad_norm": 0.26417888929256916,
+      "learning_rate": 7.590600625304882e-06,
+      "loss": 0.4961,
+      "step": 8108
+    },
+    {
+      "epoch": 1.3317184324512963,
+      "grad_norm": 0.37961262463328665,
+      "learning_rate": 7.590259737806151e-06,
+      "loss": 0.4758,
+      "step": 8109
+    },
+    {
+      "epoch": 1.3318826596596391,
+      "grad_norm": 0.32640878987499683,
+      "learning_rate": 7.5899188167441596e-06,
+      "loss": 0.4956,
+      "step": 8110
+    },
+    {
+      "epoch": 1.3320468868679818,
+      "grad_norm": 0.29588985784763766,
+      "learning_rate": 7.589577862122611e-06,
+      "loss": 0.4828,
+      "step": 8111
+    },
+    {
+      "epoch": 1.3322111140763246,
+      "grad_norm": 0.37197198970391443,
+      "learning_rate": 7.589236873945205e-06,
+      "loss": 0.4528,
+      "step": 8112
+    },
+    {
+      "epoch": 1.3323753412846673,
+      "grad_norm": 0.370013610959104,
+      "learning_rate": 7.588895852215649e-06,
+      "loss": 0.4599,
+      "step": 8113
+    },
+    {
+      "epoch": 1.3325395684930101,
+      "grad_norm": 0.37106657556119116,
+      "learning_rate": 7.588554796937643e-06,
+      "loss": 0.4644,
+      "step": 8114
+    },
+    {
+      "epoch": 1.3327037957013528,
+      "grad_norm": 0.3269685481961678,
+      "learning_rate": 7.588213708114895e-06,
+      "loss": 0.4845,
+      "step": 8115
+    },
+    {
+      "epoch": 1.3328680229096956,
+      "grad_norm": 0.2555741764001781,
+      "learning_rate": 7.587872585751108e-06,
+      "loss": 0.4721,
+      "step": 8116
+    },
+    {
+      "epoch": 1.3330322501180383,
+      "grad_norm": 0.3073360793567854,
+      "learning_rate": 7.587531429849986e-06,
+      "loss": 0.4881,
+      "step": 8117
+    },
+    {
+      "epoch": 1.3331964773263811,
+      "grad_norm": 0.48663839387542973,
+      "learning_rate": 7.587190240415235e-06,
+      "loss": 0.4736,
+      "step": 8118
+    },
+    {
+      "epoch": 1.3333607045347238,
+      "grad_norm": 0.37921822868911753,
+      "learning_rate": 7.58684901745056e-06,
+      "loss": 0.487,
+      "step": 8119
+    },
+    {
+      "epoch": 1.3335249317430664,
+      "grad_norm": 0.2710430476175914,
+      "learning_rate": 7.586507760959668e-06,
+      "loss": 0.4553,
+      "step": 8120
+    },
+    {
+      "epoch": 1.3336891589514093,
+      "grad_norm": 0.2823588509201183,
+      "learning_rate": 7.586166470946265e-06,
+      "loss": 0.4723,
+      "step": 8121
+    },
+    {
+      "epoch": 1.3338533861597521,
+      "grad_norm": 0.3588907283766576,
+      "learning_rate": 7.585825147414058e-06,
+      "loss": 0.4588,
+      "step": 8122
+    },
+    {
+      "epoch": 1.3340176133680948,
+      "grad_norm": 0.26534977430795154,
+      "learning_rate": 7.585483790366755e-06,
+      "loss": 0.4488,
+      "step": 8123
+    },
+    {
+      "epoch": 1.3341818405764374,
+      "grad_norm": 0.36041667654909365,
+      "learning_rate": 7.58514239980806e-06,
+      "loss": 0.4778,
+      "step": 8124
+    },
+    {
+      "epoch": 1.3343460677847803,
+      "grad_norm": 0.3025737628651881,
+      "learning_rate": 7.584800975741684e-06,
+      "loss": 0.486,
+      "step": 8125
+    },
+    {
+      "epoch": 1.334510294993123,
+      "grad_norm": 0.2634336667104377,
+      "learning_rate": 7.584459518171334e-06,
+      "loss": 0.4659,
+      "step": 8126
+    },
+    {
+      "epoch": 1.3346745222014658,
+      "grad_norm": 0.44720072697100516,
+      "learning_rate": 7.58411802710072e-06,
+      "loss": 0.4965,
+      "step": 8127
+    },
+    {
+      "epoch": 1.3348387494098084,
+      "grad_norm": 0.33297381374331897,
+      "learning_rate": 7.58377650253355e-06,
+      "loss": 0.5003,
+      "step": 8128
+    },
+    {
+      "epoch": 1.3350029766181513,
+      "grad_norm": 0.2732531330453581,
+      "learning_rate": 7.583434944473531e-06,
+      "loss": 0.4742,
+      "step": 8129
+    },
+    {
+      "epoch": 1.335167203826494,
+      "grad_norm": 0.314812564000472,
+      "learning_rate": 7.583093352924377e-06,
+      "loss": 0.4602,
+      "step": 8130
+    },
+    {
+      "epoch": 1.3353314310348368,
+      "grad_norm": 0.3506246269150991,
+      "learning_rate": 7.582751727889795e-06,
+      "loss": 0.4655,
+      "step": 8131
+    },
+    {
+      "epoch": 1.3354956582431794,
+      "grad_norm": 0.31631151263064144,
+      "learning_rate": 7.582410069373497e-06,
+      "loss": 0.4862,
+      "step": 8132
+    },
+    {
+      "epoch": 1.3356598854515223,
+      "grad_norm": 0.3378754290915035,
+      "learning_rate": 7.582068377379192e-06,
+      "loss": 0.4874,
+      "step": 8133
+    },
+    {
+      "epoch": 1.335824112659865,
+      "grad_norm": 0.2950314734983047,
+      "learning_rate": 7.581726651910592e-06,
+      "loss": 0.4602,
+      "step": 8134
+    },
+    {
+      "epoch": 1.3359883398682078,
+      "grad_norm": 0.37993331588181634,
+      "learning_rate": 7.58138489297141e-06,
+      "loss": 0.485,
+      "step": 8135
+    },
+    {
+      "epoch": 1.3361525670765504,
+      "grad_norm": 0.47312803234745243,
+      "learning_rate": 7.5810431005653555e-06,
+      "loss": 0.4602,
+      "step": 8136
+    },
+    {
+      "epoch": 1.336316794284893,
+      "grad_norm": 0.38400139770922437,
+      "learning_rate": 7.580701274696141e-06,
+      "loss": 0.473,
+      "step": 8137
+    },
+    {
+      "epoch": 1.3364810214932359,
+      "grad_norm": 0.2713798872467751,
+      "learning_rate": 7.58035941536748e-06,
+      "loss": 0.4735,
+      "step": 8138
+    },
+    {
+      "epoch": 1.3366452487015787,
+      "grad_norm": 0.28391273813649176,
+      "learning_rate": 7.580017522583085e-06,
+      "loss": 0.4917,
+      "step": 8139
+    },
+    {
+      "epoch": 1.3368094759099214,
+      "grad_norm": 0.29927465149936233,
+      "learning_rate": 7.57967559634667e-06,
+      "loss": 0.4704,
+      "step": 8140
+    },
+    {
+      "epoch": 1.336973703118264,
+      "grad_norm": 0.36460727628287165,
+      "learning_rate": 7.579333636661947e-06,
+      "loss": 0.4906,
+      "step": 8141
+    },
+    {
+      "epoch": 1.3371379303266069,
+      "grad_norm": 0.33819147075200523,
+      "learning_rate": 7.578991643532631e-06,
+      "loss": 0.4861,
+      "step": 8142
+    },
+    {
+      "epoch": 1.3373021575349495,
+      "grad_norm": 0.29856704467356415,
+      "learning_rate": 7.578649616962437e-06,
+      "loss": 0.4814,
+      "step": 8143
+    },
+    {
+      "epoch": 1.3374663847432924,
+      "grad_norm": 0.3243878064563001,
+      "learning_rate": 7.57830755695508e-06,
+      "loss": 0.474,
+      "step": 8144
+    },
+    {
+      "epoch": 1.337630611951635,
+      "grad_norm": 0.3260830025390945,
+      "learning_rate": 7.577965463514273e-06,
+      "loss": 0.4439,
+      "step": 8145
+    },
+    {
+      "epoch": 1.3377948391599779,
+      "grad_norm": 0.3561533434330576,
+      "learning_rate": 7.577623336643734e-06,
+      "loss": 0.4819,
+      "step": 8146
+    },
+    {
+      "epoch": 1.3379590663683205,
+      "grad_norm": 0.40282604646571896,
+      "learning_rate": 7.5772811763471765e-06,
+      "loss": 0.4731,
+      "step": 8147
+    },
+    {
+      "epoch": 1.3381232935766634,
+      "grad_norm": 0.4354568605464644,
+      "learning_rate": 7.576938982628319e-06,
+      "loss": 0.4791,
+      "step": 8148
+    },
+    {
+      "epoch": 1.338287520785006,
+      "grad_norm": 0.27024614355661286,
+      "learning_rate": 7.5765967554908766e-06,
+      "loss": 0.4534,
+      "step": 8149
+    },
+    {
+      "epoch": 1.3384517479933489,
+      "grad_norm": 0.2800449426831548,
+      "learning_rate": 7.576254494938565e-06,
+      "loss": 0.4638,
+      "step": 8150
+    },
+    {
+      "epoch": 1.3386159752016915,
+      "grad_norm": 0.31469804284271313,
+      "learning_rate": 7.5759122009751034e-06,
+      "loss": 0.4672,
+      "step": 8151
+    },
+    {
+      "epoch": 1.3387802024100344,
+      "grad_norm": 0.31981087249106815,
+      "learning_rate": 7.575569873604211e-06,
+      "loss": 0.4754,
+      "step": 8152
+    },
+    {
+      "epoch": 1.338944429618377,
+      "grad_norm": 0.32041798503279073,
+      "learning_rate": 7.575227512829601e-06,
+      "loss": 0.4542,
+      "step": 8153
+    },
+    {
+      "epoch": 1.3391086568267196,
+      "grad_norm": 0.30671216747447894,
+      "learning_rate": 7.574885118654997e-06,
+      "loss": 0.4687,
+      "step": 8154
+    },
+    {
+      "epoch": 1.3392728840350625,
+      "grad_norm": 0.2954637648278154,
+      "learning_rate": 7.574542691084114e-06,
+      "loss": 0.4976,
+      "step": 8155
+    },
+    {
+      "epoch": 1.3394371112434054,
+      "grad_norm": 0.3081145232331669,
+      "learning_rate": 7.574200230120672e-06,
+      "loss": 0.4592,
+      "step": 8156
+    },
+    {
+      "epoch": 1.339601338451748,
+      "grad_norm": 0.320559291908532,
+      "learning_rate": 7.573857735768392e-06,
+      "loss": 0.4742,
+      "step": 8157
+    },
+    {
+      "epoch": 1.3397655656600906,
+      "grad_norm": 0.3365256596562173,
+      "learning_rate": 7.573515208030992e-06,
+      "loss": 0.4914,
+      "step": 8158
+    },
+    {
+      "epoch": 1.3399297928684335,
+      "grad_norm": 0.2751284881552089,
+      "learning_rate": 7.5731726469121925e-06,
+      "loss": 0.489,
+      "step": 8159
+    },
+    {
+      "epoch": 1.3400940200767761,
+      "grad_norm": 0.37688645099946766,
+      "learning_rate": 7.572830052415716e-06,
+      "loss": 0.4933,
+      "step": 8160
+    },
+    {
+      "epoch": 1.340258247285119,
+      "grad_norm": 0.30999225372323935,
+      "learning_rate": 7.572487424545282e-06,
+      "loss": 0.4993,
+      "step": 8161
+    },
+    {
+      "epoch": 1.3404224744934616,
+      "grad_norm": 0.33954222243627585,
+      "learning_rate": 7.572144763304609e-06,
+      "loss": 0.4623,
+      "step": 8162
+    },
+    {
+      "epoch": 1.3405867017018045,
+      "grad_norm": 0.35287127368966964,
+      "learning_rate": 7.571802068697424e-06,
+      "loss": 0.4674,
+      "step": 8163
+    },
+    {
+      "epoch": 1.3407509289101471,
+      "grad_norm": 0.26149503644123145,
+      "learning_rate": 7.571459340727444e-06,
+      "loss": 0.4895,
+      "step": 8164
+    },
+    {
+      "epoch": 1.34091515611849,
+      "grad_norm": 0.340594769554577,
+      "learning_rate": 7.5711165793983955e-06,
+      "loss": 0.4668,
+      "step": 8165
+    },
+    {
+      "epoch": 1.3410793833268326,
+      "grad_norm": 0.35235304957633606,
+      "learning_rate": 7.570773784714e-06,
+      "loss": 0.4808,
+      "step": 8166
+    },
+    {
+      "epoch": 1.3412436105351755,
+      "grad_norm": 0.2674302744225442,
+      "learning_rate": 7.570430956677978e-06,
+      "loss": 0.4555,
+      "step": 8167
+    },
+    {
+      "epoch": 1.3414078377435181,
+      "grad_norm": 0.3802926912537419,
+      "learning_rate": 7.570088095294056e-06,
+      "loss": 0.4788,
+      "step": 8168
+    },
+    {
+      "epoch": 1.341572064951861,
+      "grad_norm": 0.3044697674369472,
+      "learning_rate": 7.569745200565956e-06,
+      "loss": 0.479,
+      "step": 8169
+    },
+    {
+      "epoch": 1.3417362921602036,
+      "grad_norm": 0.5955916581848941,
+      "learning_rate": 7.569402272497403e-06,
+      "loss": 0.4623,
+      "step": 8170
+    },
+    {
+      "epoch": 1.3419005193685463,
+      "grad_norm": 0.31407739496613557,
+      "learning_rate": 7.569059311092121e-06,
+      "loss": 0.4711,
+      "step": 8171
+    },
+    {
+      "epoch": 1.3420647465768891,
+      "grad_norm": 0.34767475507542017,
+      "learning_rate": 7.568716316353837e-06,
+      "loss": 0.5016,
+      "step": 8172
+    },
+    {
+      "epoch": 1.342228973785232,
+      "grad_norm": 0.38120882129971084,
+      "learning_rate": 7.568373288286274e-06,
+      "loss": 0.4619,
+      "step": 8173
+    },
+    {
+      "epoch": 1.3423932009935746,
+      "grad_norm": 0.33282327887772417,
+      "learning_rate": 7.568030226893158e-06,
+      "loss": 0.4911,
+      "step": 8174
+    },
+    {
+      "epoch": 1.3425574282019173,
+      "grad_norm": 0.2965917498704802,
+      "learning_rate": 7.567687132178216e-06,
+      "loss": 0.4884,
+      "step": 8175
+    },
+    {
+      "epoch": 1.3427216554102601,
+      "grad_norm": 0.3396648386159076,
+      "learning_rate": 7.567344004145172e-06,
+      "loss": 0.483,
+      "step": 8176
+    },
+    {
+      "epoch": 1.3428858826186028,
+      "grad_norm": 0.30333232393757004,
+      "learning_rate": 7.567000842797754e-06,
+      "loss": 0.4851,
+      "step": 8177
+    },
+    {
+      "epoch": 1.3430501098269456,
+      "grad_norm": 0.34618357769248426,
+      "learning_rate": 7.56665764813969e-06,
+      "loss": 0.4906,
+      "step": 8178
+    },
+    {
+      "epoch": 1.3432143370352883,
+      "grad_norm": 0.3789647075081895,
+      "learning_rate": 7.566314420174707e-06,
+      "loss": 0.4798,
+      "step": 8179
+    },
+    {
+      "epoch": 1.3433785642436311,
+      "grad_norm": 0.8248885820482107,
+      "learning_rate": 7.565971158906533e-06,
+      "loss": 0.4696,
+      "step": 8180
+    },
+    {
+      "epoch": 1.3435427914519738,
+      "grad_norm": 0.34738798654195413,
+      "learning_rate": 7.565627864338896e-06,
+      "loss": 0.4731,
+      "step": 8181
+    },
+    {
+      "epoch": 1.3437070186603166,
+      "grad_norm": 0.3001257042915786,
+      "learning_rate": 7.565284536475523e-06,
+      "loss": 0.4761,
+      "step": 8182
+    },
+    {
+      "epoch": 1.3438712458686592,
+      "grad_norm": 0.29395541996703284,
+      "learning_rate": 7.564941175320145e-06,
+      "loss": 0.4704,
+      "step": 8183
+    },
+    {
+      "epoch": 1.344035473077002,
+      "grad_norm": 0.3490565668842923,
+      "learning_rate": 7.564597780876489e-06,
+      "loss": 0.484,
+      "step": 8184
+    },
+    {
+      "epoch": 1.3441997002853447,
+      "grad_norm": 0.33784447165967335,
+      "learning_rate": 7.564254353148286e-06,
+      "loss": 0.4551,
+      "step": 8185
+    },
+    {
+      "epoch": 1.3443639274936876,
+      "grad_norm": 0.31079920003562245,
+      "learning_rate": 7.563910892139268e-06,
+      "loss": 0.4629,
+      "step": 8186
+    },
+    {
+      "epoch": 1.3445281547020302,
+      "grad_norm": 0.339855134296473,
+      "learning_rate": 7.563567397853162e-06,
+      "loss": 0.4723,
+      "step": 8187
+    },
+    {
+      "epoch": 1.3446923819103729,
+      "grad_norm": 0.31361048502666455,
+      "learning_rate": 7.5632238702937e-06,
+      "loss": 0.4626,
+      "step": 8188
+    },
+    {
+      "epoch": 1.3448566091187157,
+      "grad_norm": 0.30916236843543365,
+      "learning_rate": 7.562880309464612e-06,
+      "loss": 0.4885,
+      "step": 8189
+    },
+    {
+      "epoch": 1.3450208363270586,
+      "grad_norm": 0.42762086725224197,
+      "learning_rate": 7.562536715369632e-06,
+      "loss": 0.4798,
+      "step": 8190
+    },
+    {
+      "epoch": 1.3451850635354012,
+      "grad_norm": 0.29582310475373164,
+      "learning_rate": 7.562193088012489e-06,
+      "loss": 0.4757,
+      "step": 8191
+    },
+    {
+      "epoch": 1.3453492907437439,
+      "grad_norm": 0.2885250979215616,
+      "learning_rate": 7.561849427396916e-06,
+      "loss": 0.4677,
+      "step": 8192
+    },
+    {
+      "epoch": 1.3455135179520867,
+      "grad_norm": 0.3149369224702306,
+      "learning_rate": 7.561505733526646e-06,
+      "loss": 0.4801,
+      "step": 8193
+    },
+    {
+      "epoch": 1.3456777451604294,
+      "grad_norm": 0.3230766779512786,
+      "learning_rate": 7.561162006405413e-06,
+      "loss": 0.4705,
+      "step": 8194
+    },
+    {
+      "epoch": 1.3458419723687722,
+      "grad_norm": 0.2664636359519095,
+      "learning_rate": 7.560818246036948e-06,
+      "loss": 0.4632,
+      "step": 8195
+    },
+    {
+      "epoch": 1.3460061995771149,
+      "grad_norm": 0.33555822669646035,
+      "learning_rate": 7.560474452424984e-06,
+      "loss": 0.4661,
+      "step": 8196
+    },
+    {
+      "epoch": 1.3461704267854577,
+      "grad_norm": 0.6934805707048038,
+      "learning_rate": 7.560130625573259e-06,
+      "loss": 0.4862,
+      "step": 8197
+    },
+    {
+      "epoch": 1.3463346539938004,
+      "grad_norm": 0.2607196059610446,
+      "learning_rate": 7.559786765485503e-06,
+      "loss": 0.481,
+      "step": 8198
+    },
+    {
+      "epoch": 1.3464988812021432,
+      "grad_norm": 0.32406956402984705,
+      "learning_rate": 7.559442872165452e-06,
+      "loss": 0.4648,
+      "step": 8199
+    },
+    {
+      "epoch": 1.3466631084104859,
+      "grad_norm": 0.23869746457513058,
+      "learning_rate": 7.5590989456168425e-06,
+      "loss": 0.4656,
+      "step": 8200
+    },
+    {
+      "epoch": 1.3468273356188287,
+      "grad_norm": 0.3208825001499969,
+      "learning_rate": 7.558754985843408e-06,
+      "loss": 0.4713,
+      "step": 8201
+    },
+    {
+      "epoch": 1.3469915628271714,
+      "grad_norm": 0.2841932547090098,
+      "learning_rate": 7.558410992848886e-06,
+      "loss": 0.4806,
+      "step": 8202
+    },
+    {
+      "epoch": 1.3471557900355142,
+      "grad_norm": 0.3941615072914976,
+      "learning_rate": 7.55806696663701e-06,
+      "loss": 0.462,
+      "step": 8203
+    },
+    {
+      "epoch": 1.3473200172438569,
+      "grad_norm": 0.30906597354873233,
+      "learning_rate": 7.557722907211518e-06,
+      "loss": 0.4597,
+      "step": 8204
+    },
+    {
+      "epoch": 1.3474842444521995,
+      "grad_norm": 0.3261144107035512,
+      "learning_rate": 7.557378814576148e-06,
+      "loss": 0.4767,
+      "step": 8205
+    },
+    {
+      "epoch": 1.3476484716605424,
+      "grad_norm": 0.3231943091793371,
+      "learning_rate": 7.557034688734636e-06,
+      "loss": 0.4903,
+      "step": 8206
+    },
+    {
+      "epoch": 1.3478126988688852,
+      "grad_norm": 0.31891508433673865,
+      "learning_rate": 7.556690529690719e-06,
+      "loss": 0.4661,
+      "step": 8207
+    },
+    {
+      "epoch": 1.3479769260772279,
+      "grad_norm": 0.3561392861742652,
+      "learning_rate": 7.556346337448135e-06,
+      "loss": 0.468,
+      "step": 8208
+    },
+    {
+      "epoch": 1.3481411532855705,
+      "grad_norm": 0.29269560637612674,
+      "learning_rate": 7.556002112010623e-06,
+      "loss": 0.489,
+      "step": 8209
+    },
+    {
+      "epoch": 1.3483053804939134,
+      "grad_norm": 0.29961111960033693,
+      "learning_rate": 7.555657853381921e-06,
+      "loss": 0.4822,
+      "step": 8210
+    },
+    {
+      "epoch": 1.348469607702256,
+      "grad_norm": 0.37927679526581437,
+      "learning_rate": 7.55531356156577e-06,
+      "loss": 0.475,
+      "step": 8211
+    },
+    {
+      "epoch": 1.3486338349105988,
+      "grad_norm": 0.2841023304201933,
+      "learning_rate": 7.554969236565906e-06,
+      "loss": 0.4762,
+      "step": 8212
+    },
+    {
+      "epoch": 1.3487980621189415,
+      "grad_norm": 0.28326627175304997,
+      "learning_rate": 7.554624878386071e-06,
+      "loss": 0.4959,
+      "step": 8213
+    },
+    {
+      "epoch": 1.3489622893272843,
+      "grad_norm": 0.41227391057862817,
+      "learning_rate": 7.5542804870300035e-06,
+      "loss": 0.4711,
+      "step": 8214
+    },
+    {
+      "epoch": 1.349126516535627,
+      "grad_norm": 0.3915072018838373,
+      "learning_rate": 7.553936062501448e-06,
+      "loss": 0.481,
+      "step": 8215
+    },
+    {
+      "epoch": 1.3492907437439698,
+      "grad_norm": 0.3381486335554688,
+      "learning_rate": 7.55359160480414e-06,
+      "loss": 0.4789,
+      "step": 8216
+    },
+    {
+      "epoch": 1.3494549709523125,
+      "grad_norm": 0.27882630443105566,
+      "learning_rate": 7.553247113941822e-06,
+      "loss": 0.4697,
+      "step": 8217
+    },
+    {
+      "epoch": 1.3496191981606553,
+      "grad_norm": 0.31148174098022713,
+      "learning_rate": 7.552902589918237e-06,
+      "loss": 0.4677,
+      "step": 8218
+    },
+    {
+      "epoch": 1.349783425368998,
+      "grad_norm": 0.32014703747449885,
+      "learning_rate": 7.552558032737128e-06,
+      "loss": 0.4803,
+      "step": 8219
+    },
+    {
+      "epoch": 1.3499476525773408,
+      "grad_norm": 0.3466674862984574,
+      "learning_rate": 7.552213442402233e-06,
+      "loss": 0.4739,
+      "step": 8220
+    },
+    {
+      "epoch": 1.3501118797856835,
+      "grad_norm": 0.33280537484605616,
+      "learning_rate": 7.551868818917298e-06,
+      "loss": 0.4627,
+      "step": 8221
+    },
+    {
+      "epoch": 1.3502761069940261,
+      "grad_norm": 0.2603930956385385,
+      "learning_rate": 7.551524162286065e-06,
+      "loss": 0.4662,
+      "step": 8222
+    },
+    {
+      "epoch": 1.350440334202369,
+      "grad_norm": 0.2930172645612939,
+      "learning_rate": 7.551179472512278e-06,
+      "loss": 0.5013,
+      "step": 8223
+    },
+    {
+      "epoch": 1.3506045614107118,
+      "grad_norm": 0.4860934666600133,
+      "learning_rate": 7.5508347495996785e-06,
+      "loss": 0.4903,
+      "step": 8224
+    },
+    {
+      "epoch": 1.3507687886190545,
+      "grad_norm": 0.29072814081065385,
+      "learning_rate": 7.5504899935520135e-06,
+      "loss": 0.4827,
+      "step": 8225
+    },
+    {
+      "epoch": 1.350933015827397,
+      "grad_norm": 0.28925745501679684,
+      "learning_rate": 7.550145204373025e-06,
+      "loss": 0.467,
+      "step": 8226
+    },
+    {
+      "epoch": 1.35109724303574,
+      "grad_norm": 0.3215258109778846,
+      "learning_rate": 7.549800382066458e-06,
+      "loss": 0.4546,
+      "step": 8227
+    },
+    {
+      "epoch": 1.3512614702440826,
+      "grad_norm": 0.3339575673451271,
+      "learning_rate": 7.549455526636061e-06,
+      "loss": 0.4796,
+      "step": 8228
+    },
+    {
+      "epoch": 1.3514256974524255,
+      "grad_norm": 0.2885210154394203,
+      "learning_rate": 7.549110638085574e-06,
+      "loss": 0.449,
+      "step": 8229
+    },
+    {
+      "epoch": 1.351589924660768,
+      "grad_norm": 0.2797488004637664,
+      "learning_rate": 7.548765716418745e-06,
+      "loss": 0.4589,
+      "step": 8230
+    },
+    {
+      "epoch": 1.351754151869111,
+      "grad_norm": 0.42563485332320505,
+      "learning_rate": 7.5484207616393225e-06,
+      "loss": 0.4715,
+      "step": 8231
+    },
+    {
+      "epoch": 1.3519183790774536,
+      "grad_norm": 0.32789103541201964,
+      "learning_rate": 7.548075773751052e-06,
+      "loss": 0.4707,
+      "step": 8232
+    },
+    {
+      "epoch": 1.3520826062857965,
+      "grad_norm": 0.3307854208254755,
+      "learning_rate": 7.547730752757679e-06,
+      "loss": 0.4846,
+      "step": 8233
+    },
+    {
+      "epoch": 1.352246833494139,
+      "grad_norm": 0.33558511584102185,
+      "learning_rate": 7.547385698662949e-06,
+      "loss": 0.4713,
+      "step": 8234
+    },
+    {
+      "epoch": 1.352411060702482,
+      "grad_norm": 0.5575831635879408,
+      "learning_rate": 7.547040611470615e-06,
+      "loss": 0.4816,
+      "step": 8235
+    },
+    {
+      "epoch": 1.3525752879108246,
+      "grad_norm": 0.29420650994646813,
+      "learning_rate": 7.546695491184422e-06,
+      "loss": 0.4757,
+      "step": 8236
+    },
+    {
+      "epoch": 1.3527395151191675,
+      "grad_norm": 0.43134384746264715,
+      "learning_rate": 7.546350337808117e-06,
+      "loss": 0.458,
+      "step": 8237
+    },
+    {
+      "epoch": 1.35290374232751,
+      "grad_norm": 0.28010941089448865,
+      "learning_rate": 7.546005151345451e-06,
+      "loss": 0.461,
+      "step": 8238
+    },
+    {
+      "epoch": 1.3530679695358527,
+      "grad_norm": 0.4136198715585237,
+      "learning_rate": 7.545659931800171e-06,
+      "loss": 0.474,
+      "step": 8239
+    },
+    {
+      "epoch": 1.3532321967441956,
+      "grad_norm": 0.2773429745089072,
+      "learning_rate": 7.5453146791760295e-06,
+      "loss": 0.4636,
+      "step": 8240
+    },
+    {
+      "epoch": 1.3533964239525385,
+      "grad_norm": 0.4222897509697245,
+      "learning_rate": 7.544969393476774e-06,
+      "loss": 0.4792,
+      "step": 8241
+    },
+    {
+      "epoch": 1.353560651160881,
+      "grad_norm": 0.30400034421879774,
+      "learning_rate": 7.544624074706155e-06,
+      "loss": 0.482,
+      "step": 8242
+    },
+    {
+      "epoch": 1.3537248783692237,
+      "grad_norm": 0.3404611704418366,
+      "learning_rate": 7.544278722867922e-06,
+      "loss": 0.4801,
+      "step": 8243
+    },
+    {
+      "epoch": 1.3538891055775666,
+      "grad_norm": 0.3795342780211706,
+      "learning_rate": 7.543933337965828e-06,
+      "loss": 0.4357,
+      "step": 8244
+    },
+    {
+      "epoch": 1.3540533327859092,
+      "grad_norm": 0.2967089858210644,
+      "learning_rate": 7.543587920003622e-06,
+      "loss": 0.4621,
+      "step": 8245
+    },
+    {
+      "epoch": 1.354217559994252,
+      "grad_norm": 0.27784864211749893,
+      "learning_rate": 7.543242468985057e-06,
+      "loss": 0.4604,
+      "step": 8246
+    },
+    {
+      "epoch": 1.3543817872025947,
+      "grad_norm": 0.47344567937835635,
+      "learning_rate": 7.542896984913885e-06,
+      "loss": 0.4714,
+      "step": 8247
+    },
+    {
+      "epoch": 1.3545460144109376,
+      "grad_norm": 0.3088897586133829,
+      "learning_rate": 7.542551467793858e-06,
+      "loss": 0.4707,
+      "step": 8248
+    },
+    {
+      "epoch": 1.3547102416192802,
+      "grad_norm": 0.3436137560670549,
+      "learning_rate": 7.542205917628729e-06,
+      "loss": 0.4804,
+      "step": 8249
+    },
+    {
+      "epoch": 1.354874468827623,
+      "grad_norm": 0.35880366032473915,
+      "learning_rate": 7.54186033442225e-06,
+      "loss": 0.4797,
+      "step": 8250
+    },
+    {
+      "epoch": 1.3550386960359657,
+      "grad_norm": 0.4984973501764961,
+      "learning_rate": 7.541514718178174e-06,
+      "loss": 0.4688,
+      "step": 8251
+    },
+    {
+      "epoch": 1.3552029232443086,
+      "grad_norm": 0.33572784628150065,
+      "learning_rate": 7.541169068900258e-06,
+      "loss": 0.4816,
+      "step": 8252
+    },
+    {
+      "epoch": 1.3553671504526512,
+      "grad_norm": 0.2685346799680185,
+      "learning_rate": 7.540823386592252e-06,
+      "loss": 0.4689,
+      "step": 8253
+    },
+    {
+      "epoch": 1.355531377660994,
+      "grad_norm": 0.32616794767271073,
+      "learning_rate": 7.540477671257913e-06,
+      "loss": 0.4785,
+      "step": 8254
+    },
+    {
+      "epoch": 1.3556956048693367,
+      "grad_norm": 0.297867131998836,
+      "learning_rate": 7.540131922900995e-06,
+      "loss": 0.4597,
+      "step": 8255
+    },
+    {
+      "epoch": 1.3558598320776793,
+      "grad_norm": 0.4019680636042429,
+      "learning_rate": 7.5397861415252526e-06,
+      "loss": 0.4575,
+      "step": 8256
+    },
+    {
+      "epoch": 1.3560240592860222,
+      "grad_norm": 0.3480970659977098,
+      "learning_rate": 7.539440327134442e-06,
+      "loss": 0.4863,
+      "step": 8257
+    },
+    {
+      "epoch": 1.356188286494365,
+      "grad_norm": 0.2997521494595294,
+      "learning_rate": 7.53909447973232e-06,
+      "loss": 0.4875,
+      "step": 8258
+    },
+    {
+      "epoch": 1.3563525137027077,
+      "grad_norm": 0.3412320676442106,
+      "learning_rate": 7.538748599322642e-06,
+      "loss": 0.4846,
+      "step": 8259
+    },
+    {
+      "epoch": 1.3565167409110503,
+      "grad_norm": 0.30124717220604347,
+      "learning_rate": 7.538402685909164e-06,
+      "loss": 0.4913,
+      "step": 8260
+    },
+    {
+      "epoch": 1.3566809681193932,
+      "grad_norm": 0.5869707815625042,
+      "learning_rate": 7.538056739495643e-06,
+      "loss": 0.4619,
+      "step": 8261
+    },
+    {
+      "epoch": 1.3568451953277358,
+      "grad_norm": 0.41886149665065575,
+      "learning_rate": 7.537710760085837e-06,
+      "loss": 0.4819,
+      "step": 8262
+    },
+    {
+      "epoch": 1.3570094225360787,
+      "grad_norm": 0.3076873203590548,
+      "learning_rate": 7.537364747683502e-06,
+      "loss": 0.461,
+      "step": 8263
+    },
+    {
+      "epoch": 1.3571736497444213,
+      "grad_norm": 0.339881383453682,
+      "learning_rate": 7.537018702292401e-06,
+      "loss": 0.4584,
+      "step": 8264
+    },
+    {
+      "epoch": 1.3573378769527642,
+      "grad_norm": 0.4871044368615955,
+      "learning_rate": 7.536672623916286e-06,
+      "loss": 0.4749,
+      "step": 8265
+    },
+    {
+      "epoch": 1.3575021041611068,
+      "grad_norm": 0.2778485736129083,
+      "learning_rate": 7.5363265125589195e-06,
+      "loss": 0.4515,
+      "step": 8266
+    },
+    {
+      "epoch": 1.3576663313694497,
+      "grad_norm": 0.29364065488596336,
+      "learning_rate": 7.535980368224061e-06,
+      "loss": 0.4816,
+      "step": 8267
+    },
+    {
+      "epoch": 1.3578305585777923,
+      "grad_norm": 0.2838595820478707,
+      "learning_rate": 7.5356341909154665e-06,
+      "loss": 0.4566,
+      "step": 8268
+    },
+    {
+      "epoch": 1.3579947857861352,
+      "grad_norm": 0.3392614946332992,
+      "learning_rate": 7.5352879806369e-06,
+      "loss": 0.4608,
+      "step": 8269
+    },
+    {
+      "epoch": 1.3581590129944778,
+      "grad_norm": 0.3391061577152166,
+      "learning_rate": 7.5349417373921175e-06,
+      "loss": 0.4797,
+      "step": 8270
+    },
+    {
+      "epoch": 1.3583232402028207,
+      "grad_norm": 0.3055348595500309,
+      "learning_rate": 7.534595461184884e-06,
+      "loss": 0.4872,
+      "step": 8271
+    },
+    {
+      "epoch": 1.3584874674111633,
+      "grad_norm": 0.3142799258083078,
+      "learning_rate": 7.534249152018957e-06,
+      "loss": 0.4582,
+      "step": 8272
+    },
+    {
+      "epoch": 1.358651694619506,
+      "grad_norm": 0.314329556767126,
+      "learning_rate": 7.533902809898098e-06,
+      "loss": 0.4672,
+      "step": 8273
+    },
+    {
+      "epoch": 1.3588159218278488,
+      "grad_norm": 0.2967056278298423,
+      "learning_rate": 7.533556434826072e-06,
+      "loss": 0.476,
+      "step": 8274
+    },
+    {
+      "epoch": 1.3589801490361917,
+      "grad_norm": 0.2837665629837183,
+      "learning_rate": 7.533210026806636e-06,
+      "loss": 0.4681,
+      "step": 8275
+    },
+    {
+      "epoch": 1.3591443762445343,
+      "grad_norm": 0.32916697819824303,
+      "learning_rate": 7.532863585843556e-06,
+      "loss": 0.4799,
+      "step": 8276
+    },
+    {
+      "epoch": 1.359308603452877,
+      "grad_norm": 0.3801296291612379,
+      "learning_rate": 7.532517111940593e-06,
+      "loss": 0.482,
+      "step": 8277
+    },
+    {
+      "epoch": 1.3594728306612198,
+      "grad_norm": 0.30518418701039557,
+      "learning_rate": 7.5321706051015115e-06,
+      "loss": 0.4902,
+      "step": 8278
+    },
+    {
+      "epoch": 1.3596370578695625,
+      "grad_norm": 0.31494213222449563,
+      "learning_rate": 7.531824065330073e-06,
+      "loss": 0.4801,
+      "step": 8279
+    },
+    {
+      "epoch": 1.3598012850779053,
+      "grad_norm": 0.27913060649487614,
+      "learning_rate": 7.5314774926300425e-06,
+      "loss": 0.4674,
+      "step": 8280
+    },
+    {
+      "epoch": 1.359965512286248,
+      "grad_norm": 0.4272270596215226,
+      "learning_rate": 7.531130887005185e-06,
+      "loss": 0.4564,
+      "step": 8281
+    },
+    {
+      "epoch": 1.3601297394945908,
+      "grad_norm": 0.33419238881683055,
+      "learning_rate": 7.5307842484592625e-06,
+      "loss": 0.4829,
+      "step": 8282
+    },
+    {
+      "epoch": 1.3602939667029335,
+      "grad_norm": 0.3286978349267236,
+      "learning_rate": 7.530437576996042e-06,
+      "loss": 0.4811,
+      "step": 8283
+    },
+    {
+      "epoch": 1.3604581939112763,
+      "grad_norm": 0.29069776505360395,
+      "learning_rate": 7.530090872619287e-06,
+      "loss": 0.4869,
+      "step": 8284
+    },
+    {
+      "epoch": 1.360622421119619,
+      "grad_norm": 0.3464972039264677,
+      "learning_rate": 7.529744135332765e-06,
+      "loss": 0.4772,
+      "step": 8285
+    },
+    {
+      "epoch": 1.3607866483279618,
+      "grad_norm": 0.2661431197402841,
+      "learning_rate": 7.52939736514024e-06,
+      "loss": 0.4928,
+      "step": 8286
+    },
+    {
+      "epoch": 1.3609508755363044,
+      "grad_norm": 0.3143496844872588,
+      "learning_rate": 7.5290505620454785e-06,
+      "loss": 0.4915,
+      "step": 8287
+    },
+    {
+      "epoch": 1.3611151027446473,
+      "grad_norm": 0.379624731415056,
+      "learning_rate": 7.528703726052248e-06,
+      "loss": 0.4606,
+      "step": 8288
+    },
+    {
+      "epoch": 1.36127932995299,
+      "grad_norm": 0.3309339659074846,
+      "learning_rate": 7.528356857164315e-06,
+      "loss": 0.4516,
+      "step": 8289
+    },
+    {
+      "epoch": 1.3614435571613326,
+      "grad_norm": 0.30095414860162006,
+      "learning_rate": 7.5280099553854495e-06,
+      "loss": 0.453,
+      "step": 8290
+    },
+    {
+      "epoch": 1.3616077843696754,
+      "grad_norm": 0.5214225092991446,
+      "learning_rate": 7.527663020719415e-06,
+      "loss": 0.4589,
+      "step": 8291
+    },
+    {
+      "epoch": 1.3617720115780183,
+      "grad_norm": 0.29060618148335327,
+      "learning_rate": 7.52731605316998e-06,
+      "loss": 0.4746,
+      "step": 8292
+    },
+    {
+      "epoch": 1.361936238786361,
+      "grad_norm": 0.3803658344262675,
+      "learning_rate": 7.526969052740916e-06,
+      "loss": 0.4697,
+      "step": 8293
+    },
+    {
+      "epoch": 1.3621004659947036,
+      "grad_norm": 0.34410484988053686,
+      "learning_rate": 7.52662201943599e-06,
+      "loss": 0.4707,
+      "step": 8294
+    },
+    {
+      "epoch": 1.3622646932030464,
+      "grad_norm": 0.32125725777151576,
+      "learning_rate": 7.52627495325897e-06,
+      "loss": 0.472,
+      "step": 8295
+    },
+    {
+      "epoch": 1.362428920411389,
+      "grad_norm": 0.3261931445987548,
+      "learning_rate": 7.525927854213627e-06,
+      "loss": 0.4928,
+      "step": 8296
+    },
+    {
+      "epoch": 1.362593147619732,
+      "grad_norm": 0.4204108817507413,
+      "learning_rate": 7.52558072230373e-06,
+      "loss": 0.4595,
+      "step": 8297
+    },
+    {
+      "epoch": 1.3627573748280746,
+      "grad_norm": 0.32134642277397796,
+      "learning_rate": 7.5252335575330514e-06,
+      "loss": 0.4782,
+      "step": 8298
+    },
+    {
+      "epoch": 1.3629216020364174,
+      "grad_norm": 0.3515706624302091,
+      "learning_rate": 7.524886359905357e-06,
+      "loss": 0.4944,
+      "step": 8299
+    },
+    {
+      "epoch": 1.36308582924476,
+      "grad_norm": 0.3573167400693006,
+      "learning_rate": 7.5245391294244225e-06,
+      "loss": 0.4852,
+      "step": 8300
+    },
+    {
+      "epoch": 1.363250056453103,
+      "grad_norm": 0.3120269607498423,
+      "learning_rate": 7.524191866094016e-06,
+      "loss": 0.4793,
+      "step": 8301
+    },
+    {
+      "epoch": 1.3634142836614456,
+      "grad_norm": 0.30100585842420313,
+      "learning_rate": 7.523844569917912e-06,
+      "loss": 0.4567,
+      "step": 8302
+    },
+    {
+      "epoch": 1.3635785108697884,
+      "grad_norm": 0.33931025334201326,
+      "learning_rate": 7.523497240899881e-06,
+      "loss": 0.495,
+      "step": 8303
+    },
+    {
+      "epoch": 1.363742738078131,
+      "grad_norm": 0.30310450315920817,
+      "learning_rate": 7.523149879043694e-06,
+      "loss": 0.4844,
+      "step": 8304
+    },
+    {
+      "epoch": 1.363906965286474,
+      "grad_norm": 0.36035296781002274,
+      "learning_rate": 7.522802484353125e-06,
+      "loss": 0.4844,
+      "step": 8305
+    },
+    {
+      "epoch": 1.3640711924948166,
+      "grad_norm": 0.32849349322708093,
+      "learning_rate": 7.522455056831948e-06,
+      "loss": 0.4785,
+      "step": 8306
+    },
+    {
+      "epoch": 1.3642354197031592,
+      "grad_norm": 0.3416519194854029,
+      "learning_rate": 7.522107596483934e-06,
+      "loss": 0.4575,
+      "step": 8307
+    },
+    {
+      "epoch": 1.364399646911502,
+      "grad_norm": 0.29685825846548575,
+      "learning_rate": 7.5217601033128604e-06,
+      "loss": 0.4852,
+      "step": 8308
+    },
+    {
+      "epoch": 1.364563874119845,
+      "grad_norm": 0.31485021328089774,
+      "learning_rate": 7.5214125773224975e-06,
+      "loss": 0.4671,
+      "step": 8309
+    },
+    {
+      "epoch": 1.3647281013281876,
+      "grad_norm": 0.28896520824254684,
+      "learning_rate": 7.5210650185166205e-06,
+      "loss": 0.4626,
+      "step": 8310
+    },
+    {
+      "epoch": 1.3648923285365302,
+      "grad_norm": 0.3173799606341833,
+      "learning_rate": 7.520717426899007e-06,
+      "loss": 0.4505,
+      "step": 8311
+    },
+    {
+      "epoch": 1.365056555744873,
+      "grad_norm": 0.45661343907823004,
+      "learning_rate": 7.520369802473429e-06,
+      "loss": 0.46,
+      "step": 8312
+    },
+    {
+      "epoch": 1.3652207829532157,
+      "grad_norm": 0.31662459661797954,
+      "learning_rate": 7.520022145243664e-06,
+      "loss": 0.4689,
+      "step": 8313
+    },
+    {
+      "epoch": 1.3653850101615586,
+      "grad_norm": 0.3206366426822329,
+      "learning_rate": 7.5196744552134866e-06,
+      "loss": 0.4767,
+      "step": 8314
+    },
+    {
+      "epoch": 1.3655492373699012,
+      "grad_norm": 0.31121042162784335,
+      "learning_rate": 7.519326732386674e-06,
+      "loss": 0.4463,
+      "step": 8315
+    },
+    {
+      "epoch": 1.365713464578244,
+      "grad_norm": 0.3287820896136919,
+      "learning_rate": 7.518978976767001e-06,
+      "loss": 0.4833,
+      "step": 8316
+    },
+    {
+      "epoch": 1.3658776917865867,
+      "grad_norm": 0.44335887820682335,
+      "learning_rate": 7.518631188358249e-06,
+      "loss": 0.4892,
+      "step": 8317
+    },
+    {
+      "epoch": 1.3660419189949295,
+      "grad_norm": 0.2763335722270975,
+      "learning_rate": 7.51828336716419e-06,
+      "loss": 0.4862,
+      "step": 8318
+    },
+    {
+      "epoch": 1.3662061462032722,
+      "grad_norm": 0.3209361127665086,
+      "learning_rate": 7.517935513188605e-06,
+      "loss": 0.4563,
+      "step": 8319
+    },
+    {
+      "epoch": 1.366370373411615,
+      "grad_norm": 0.35302259974899775,
+      "learning_rate": 7.517587626435271e-06,
+      "loss": 0.4781,
+      "step": 8320
+    },
+    {
+      "epoch": 1.3665346006199577,
+      "grad_norm": 0.3165049719005419,
+      "learning_rate": 7.5172397069079656e-06,
+      "loss": 0.4508,
+      "step": 8321
+    },
+    {
+      "epoch": 1.3666988278283005,
+      "grad_norm": 0.292617607086799,
+      "learning_rate": 7.516891754610469e-06,
+      "loss": 0.4726,
+      "step": 8322
+    },
+    {
+      "epoch": 1.3668630550366432,
+      "grad_norm": 0.3238974592916656,
+      "learning_rate": 7.5165437695465605e-06,
+      "loss": 0.4742,
+      "step": 8323
+    },
+    {
+      "epoch": 1.3670272822449858,
+      "grad_norm": 0.32845712957705675,
+      "learning_rate": 7.516195751720018e-06,
+      "loss": 0.4692,
+      "step": 8324
+    },
+    {
+      "epoch": 1.3671915094533287,
+      "grad_norm": 0.3973395741657114,
+      "learning_rate": 7.515847701134623e-06,
+      "loss": 0.4554,
+      "step": 8325
+    },
+    {
+      "epoch": 1.3673557366616715,
+      "grad_norm": 0.3248357261446677,
+      "learning_rate": 7.5154996177941544e-06,
+      "loss": 0.4755,
+      "step": 8326
+    },
+    {
+      "epoch": 1.3675199638700142,
+      "grad_norm": 0.29548108080992236,
+      "learning_rate": 7.515151501702392e-06,
+      "loss": 0.5038,
+      "step": 8327
+    },
+    {
+      "epoch": 1.3676841910783568,
+      "grad_norm": 0.3793355108500688,
+      "learning_rate": 7.514803352863119e-06,
+      "loss": 0.4573,
+      "step": 8328
+    },
+    {
+      "epoch": 1.3678484182866997,
+      "grad_norm": 0.28719236328501363,
+      "learning_rate": 7.5144551712801146e-06,
+      "loss": 0.4873,
+      "step": 8329
+    },
+    {
+      "epoch": 1.3680126454950423,
+      "grad_norm": 0.31485206904199836,
+      "learning_rate": 7.514106956957162e-06,
+      "loss": 0.4553,
+      "step": 8330
+    },
+    {
+      "epoch": 1.3681768727033852,
+      "grad_norm": 0.4207883658274008,
+      "learning_rate": 7.513758709898041e-06,
+      "loss": 0.4754,
+      "step": 8331
+    },
+    {
+      "epoch": 1.3683410999117278,
+      "grad_norm": 0.3500059465239506,
+      "learning_rate": 7.513410430106538e-06,
+      "loss": 0.4904,
+      "step": 8332
+    },
+    {
+      "epoch": 1.3685053271200707,
+      "grad_norm": 0.29361003110067885,
+      "learning_rate": 7.5130621175864295e-06,
+      "loss": 0.4701,
+      "step": 8333
+    },
+    {
+      "epoch": 1.3686695543284133,
+      "grad_norm": 0.36962502570741995,
+      "learning_rate": 7.512713772341504e-06,
+      "loss": 0.4786,
+      "step": 8334
+    },
+    {
+      "epoch": 1.3688337815367562,
+      "grad_norm": 0.3727744016142195,
+      "learning_rate": 7.512365394375543e-06,
+      "loss": 0.4797,
+      "step": 8335
+    },
+    {
+      "epoch": 1.3689980087450988,
+      "grad_norm": 0.2792130647553743,
+      "learning_rate": 7.512016983692329e-06,
+      "loss": 0.4687,
+      "step": 8336
+    },
+    {
+      "epoch": 1.3691622359534417,
+      "grad_norm": 0.29179763332648856,
+      "learning_rate": 7.511668540295648e-06,
+      "loss": 0.481,
+      "step": 8337
+    },
+    {
+      "epoch": 1.3693264631617843,
+      "grad_norm": 0.3281601698226462,
+      "learning_rate": 7.5113200641892826e-06,
+      "loss": 0.4811,
+      "step": 8338
+    },
+    {
+      "epoch": 1.3694906903701272,
+      "grad_norm": 0.32319827408143587,
+      "learning_rate": 7.510971555377019e-06,
+      "loss": 0.4744,
+      "step": 8339
+    },
+    {
+      "epoch": 1.3696549175784698,
+      "grad_norm": 0.48402519535000715,
+      "learning_rate": 7.510623013862643e-06,
+      "loss": 0.4692,
+      "step": 8340
+    },
+    {
+      "epoch": 1.3698191447868124,
+      "grad_norm": 0.36689549011407746,
+      "learning_rate": 7.510274439649938e-06,
+      "loss": 0.472,
+      "step": 8341
+    },
+    {
+      "epoch": 1.3699833719951553,
+      "grad_norm": 0.2931911030026804,
+      "learning_rate": 7.509925832742691e-06,
+      "loss": 0.4815,
+      "step": 8342
+    },
+    {
+      "epoch": 1.3701475992034982,
+      "grad_norm": 0.38794231413739944,
+      "learning_rate": 7.5095771931446874e-06,
+      "loss": 0.4806,
+      "step": 8343
+    },
+    {
+      "epoch": 1.3703118264118408,
+      "grad_norm": 0.32982445067061356,
+      "learning_rate": 7.509228520859716e-06,
+      "loss": 0.4655,
+      "step": 8344
+    },
+    {
+      "epoch": 1.3704760536201834,
+      "grad_norm": 0.29391263723942956,
+      "learning_rate": 7.508879815891561e-06,
+      "loss": 0.4733,
+      "step": 8345
+    },
+    {
+      "epoch": 1.3706402808285263,
+      "grad_norm": 0.35197760791067445,
+      "learning_rate": 7.50853107824401e-06,
+      "loss": 0.4762,
+      "step": 8346
+    },
+    {
+      "epoch": 1.370804508036869,
+      "grad_norm": 0.3392993683378087,
+      "learning_rate": 7.508182307920853e-06,
+      "loss": 0.4749,
+      "step": 8347
+    },
+    {
+      "epoch": 1.3709687352452118,
+      "grad_norm": 0.4550315133155526,
+      "learning_rate": 7.507833504925876e-06,
+      "loss": 0.461,
+      "step": 8348
+    },
+    {
+      "epoch": 1.3711329624535544,
+      "grad_norm": 0.31311924905014327,
+      "learning_rate": 7.507484669262869e-06,
+      "loss": 0.4714,
+      "step": 8349
+    },
+    {
+      "epoch": 1.3712971896618973,
+      "grad_norm": 0.39457589400869614,
+      "learning_rate": 7.507135800935618e-06,
+      "loss": 0.4842,
+      "step": 8350
+    },
+    {
+      "epoch": 1.37146141687024,
+      "grad_norm": 0.32116871340010233,
+      "learning_rate": 7.506786899947914e-06,
+      "loss": 0.4808,
+      "step": 8351
+    },
+    {
+      "epoch": 1.3716256440785828,
+      "grad_norm": 0.28265646058257143,
+      "learning_rate": 7.506437966303546e-06,
+      "loss": 0.4592,
+      "step": 8352
+    },
+    {
+      "epoch": 1.3717898712869254,
+      "grad_norm": 0.2988692411952469,
+      "learning_rate": 7.5060890000063035e-06,
+      "loss": 0.4847,
+      "step": 8353
+    },
+    {
+      "epoch": 1.3719540984952683,
+      "grad_norm": 0.3962595768036921,
+      "learning_rate": 7.505740001059977e-06,
+      "loss": 0.4694,
+      "step": 8354
+    },
+    {
+      "epoch": 1.372118325703611,
+      "grad_norm": 0.27664766973560395,
+      "learning_rate": 7.5053909694683575e-06,
+      "loss": 0.4652,
+      "step": 8355
+    },
+    {
+      "epoch": 1.3722825529119538,
+      "grad_norm": 0.34306614080647385,
+      "learning_rate": 7.505041905235234e-06,
+      "loss": 0.4609,
+      "step": 8356
+    },
+    {
+      "epoch": 1.3724467801202964,
+      "grad_norm": 0.3243406451551074,
+      "learning_rate": 7.5046928083644e-06,
+      "loss": 0.4756,
+      "step": 8357
+    },
+    {
+      "epoch": 1.372611007328639,
+      "grad_norm": 0.32581858539649494,
+      "learning_rate": 7.504343678859645e-06,
+      "loss": 0.4865,
+      "step": 8358
+    },
+    {
+      "epoch": 1.372775234536982,
+      "grad_norm": 0.3316606832659808,
+      "learning_rate": 7.5039945167247625e-06,
+      "loss": 0.4613,
+      "step": 8359
+    },
+    {
+      "epoch": 1.3729394617453248,
+      "grad_norm": 0.3043342240235991,
+      "learning_rate": 7.503645321963543e-06,
+      "loss": 0.4445,
+      "step": 8360
+    },
+    {
+      "epoch": 1.3731036889536674,
+      "grad_norm": 0.36325843544559216,
+      "learning_rate": 7.503296094579782e-06,
+      "loss": 0.4766,
+      "step": 8361
+    },
+    {
+      "epoch": 1.37326791616201,
+      "grad_norm": 0.3295513937739328,
+      "learning_rate": 7.502946834577269e-06,
+      "loss": 0.4764,
+      "step": 8362
+    },
+    {
+      "epoch": 1.373432143370353,
+      "grad_norm": 0.29301402397809956,
+      "learning_rate": 7.5025975419597995e-06,
+      "loss": 0.453,
+      "step": 8363
+    },
+    {
+      "epoch": 1.3735963705786955,
+      "grad_norm": 0.3265871905901499,
+      "learning_rate": 7.502248216731166e-06,
+      "loss": 0.485,
+      "step": 8364
+    },
+    {
+      "epoch": 1.3737605977870384,
+      "grad_norm": 0.3307848184863391,
+      "learning_rate": 7.501898858895163e-06,
+      "loss": 0.482,
+      "step": 8365
+    },
+    {
+      "epoch": 1.373924824995381,
+      "grad_norm": 0.318539651593937,
+      "learning_rate": 7.501549468455586e-06,
+      "loss": 0.4635,
+      "step": 8366
+    },
+    {
+      "epoch": 1.374089052203724,
+      "grad_norm": 0.4299176845852357,
+      "learning_rate": 7.501200045416228e-06,
+      "loss": 0.4746,
+      "step": 8367
+    },
+    {
+      "epoch": 1.3742532794120665,
+      "grad_norm": 0.33954283941162516,
+      "learning_rate": 7.500850589780885e-06,
+      "loss": 0.4581,
+      "step": 8368
+    },
+    {
+      "epoch": 1.3744175066204094,
+      "grad_norm": 0.4102003323616825,
+      "learning_rate": 7.5005011015533515e-06,
+      "loss": 0.4652,
+      "step": 8369
+    },
+    {
+      "epoch": 1.374581733828752,
+      "grad_norm": 0.37286895421142485,
+      "learning_rate": 7.500151580737423e-06,
+      "loss": 0.4749,
+      "step": 8370
+    },
+    {
+      "epoch": 1.374745961037095,
+      "grad_norm": 0.4328474063000734,
+      "learning_rate": 7.4998020273368985e-06,
+      "loss": 0.4734,
+      "step": 8371
+    },
+    {
+      "epoch": 1.3749101882454375,
+      "grad_norm": 0.3229313804670453,
+      "learning_rate": 7.499452441355571e-06,
+      "loss": 0.4854,
+      "step": 8372
+    },
+    {
+      "epoch": 1.3750744154537804,
+      "grad_norm": 0.338741035436834,
+      "learning_rate": 7.49910282279724e-06,
+      "loss": 0.4485,
+      "step": 8373
+    },
+    {
+      "epoch": 1.375238642662123,
+      "grad_norm": 0.3835064638344377,
+      "learning_rate": 7.498753171665702e-06,
+      "loss": 0.4763,
+      "step": 8374
+    },
+    {
+      "epoch": 1.3754028698704657,
+      "grad_norm": 0.3824516415335608,
+      "learning_rate": 7.498403487964754e-06,
+      "loss": 0.4885,
+      "step": 8375
+    },
+    {
+      "epoch": 1.3755670970788085,
+      "grad_norm": 0.3196719968516,
+      "learning_rate": 7.4980537716981935e-06,
+      "loss": 0.4901,
+      "step": 8376
+    },
+    {
+      "epoch": 1.3757313242871514,
+      "grad_norm": 0.3553565763368463,
+      "learning_rate": 7.49770402286982e-06,
+      "loss": 0.4437,
+      "step": 8377
+    },
+    {
+      "epoch": 1.375895551495494,
+      "grad_norm": 0.4544423987932916,
+      "learning_rate": 7.49735424148343e-06,
+      "loss": 0.4518,
+      "step": 8378
+    },
+    {
+      "epoch": 1.3760597787038367,
+      "grad_norm": 0.3659010510622629,
+      "learning_rate": 7.497004427542827e-06,
+      "loss": 0.5014,
+      "step": 8379
+    },
+    {
+      "epoch": 1.3762240059121795,
+      "grad_norm": 0.30706994581649516,
+      "learning_rate": 7.4966545810518046e-06,
+      "loss": 0.4702,
+      "step": 8380
+    },
+    {
+      "epoch": 1.3763882331205222,
+      "grad_norm": 0.3120020596631399,
+      "learning_rate": 7.496304702014165e-06,
+      "loss": 0.4603,
+      "step": 8381
+    },
+    {
+      "epoch": 1.376552460328865,
+      "grad_norm": 0.3428900780388786,
+      "learning_rate": 7.49595479043371e-06,
+      "loss": 0.4851,
+      "step": 8382
+    },
+    {
+      "epoch": 1.3767166875372077,
+      "grad_norm": 0.28232317726051354,
+      "learning_rate": 7.495604846314236e-06,
+      "loss": 0.473,
+      "step": 8383
+    },
+    {
+      "epoch": 1.3768809147455505,
+      "grad_norm": 0.40601124361359603,
+      "learning_rate": 7.495254869659548e-06,
+      "loss": 0.4825,
+      "step": 8384
+    },
+    {
+      "epoch": 1.3770451419538932,
+      "grad_norm": 0.30385377720199874,
+      "learning_rate": 7.494904860473446e-06,
+      "loss": 0.4827,
+      "step": 8385
+    },
+    {
+      "epoch": 1.377209369162236,
+      "grad_norm": 0.3585977445148269,
+      "learning_rate": 7.494554818759729e-06,
+      "loss": 0.4686,
+      "step": 8386
+    },
+    {
+      "epoch": 1.3773735963705787,
+      "grad_norm": 0.296032640267518,
+      "learning_rate": 7.4942047445222005e-06,
+      "loss": 0.4749,
+      "step": 8387
+    },
+    {
+      "epoch": 1.3775378235789215,
+      "grad_norm": 0.3929412698877934,
+      "learning_rate": 7.493854637764663e-06,
+      "loss": 0.471,
+      "step": 8388
+    },
+    {
+      "epoch": 1.3777020507872642,
+      "grad_norm": 0.3584205071049299,
+      "learning_rate": 7.493504498490919e-06,
+      "loss": 0.4703,
+      "step": 8389
+    },
+    {
+      "epoch": 1.377866277995607,
+      "grad_norm": 0.26009475523382186,
+      "learning_rate": 7.49315432670477e-06,
+      "loss": 0.4767,
+      "step": 8390
+    },
+    {
+      "epoch": 1.3780305052039497,
+      "grad_norm": 0.5168875923714398,
+      "learning_rate": 7.492804122410021e-06,
+      "loss": 0.467,
+      "step": 8391
+    },
+    {
+      "epoch": 1.3781947324122923,
+      "grad_norm": 0.2963401806601922,
+      "learning_rate": 7.492453885610474e-06,
+      "loss": 0.4768,
+      "step": 8392
+    },
+    {
+      "epoch": 1.3783589596206351,
+      "grad_norm": 0.31659003611238296,
+      "learning_rate": 7.492103616309933e-06,
+      "loss": 0.4711,
+      "step": 8393
+    },
+    {
+      "epoch": 1.378523186828978,
+      "grad_norm": 0.33403184376342293,
+      "learning_rate": 7.491753314512205e-06,
+      "loss": 0.4788,
+      "step": 8394
+    },
+    {
+      "epoch": 1.3786874140373206,
+      "grad_norm": 0.6322963767384263,
+      "learning_rate": 7.491402980221091e-06,
+      "loss": 0.484,
+      "step": 8395
+    },
+    {
+      "epoch": 1.3788516412456633,
+      "grad_norm": 0.3105545863957514,
+      "learning_rate": 7.491052613440398e-06,
+      "loss": 0.4804,
+      "step": 8396
+    },
+    {
+      "epoch": 1.3790158684540061,
+      "grad_norm": 0.37750790683795843,
+      "learning_rate": 7.4907022141739305e-06,
+      "loss": 0.4725,
+      "step": 8397
+    },
+    {
+      "epoch": 1.3791800956623488,
+      "grad_norm": 4.6596300526714955,
+      "learning_rate": 7.490351782425494e-06,
+      "loss": 0.4773,
+      "step": 8398
+    },
+    {
+      "epoch": 1.3793443228706916,
+      "grad_norm": 0.35738475045237333,
+      "learning_rate": 7.490001318198896e-06,
+      "loss": 0.4715,
+      "step": 8399
+    },
+    {
+      "epoch": 1.3795085500790343,
+      "grad_norm": 0.2786126847265779,
+      "learning_rate": 7.489650821497942e-06,
+      "loss": 0.4735,
+      "step": 8400
+    },
+    {
+      "epoch": 1.3796727772873771,
+      "grad_norm": 0.33428832912866097,
+      "learning_rate": 7.489300292326438e-06,
+      "loss": 0.4639,
+      "step": 8401
+    },
+    {
+      "epoch": 1.3798370044957198,
+      "grad_norm": 0.30964514582994457,
+      "learning_rate": 7.4889497306881924e-06,
+      "loss": 0.466,
+      "step": 8402
+    },
+    {
+      "epoch": 1.3800012317040626,
+      "grad_norm": 0.2889957503949191,
+      "learning_rate": 7.488599136587012e-06,
+      "loss": 0.4659,
+      "step": 8403
+    },
+    {
+      "epoch": 1.3801654589124053,
+      "grad_norm": 0.3925535751253334,
+      "learning_rate": 7.488248510026704e-06,
+      "loss": 0.4676,
+      "step": 8404
+    },
+    {
+      "epoch": 1.3803296861207481,
+      "grad_norm": 0.4829764742485803,
+      "learning_rate": 7.487897851011077e-06,
+      "loss": 0.4857,
+      "step": 8405
+    },
+    {
+      "epoch": 1.3804939133290908,
+      "grad_norm": 0.5775087390936786,
+      "learning_rate": 7.4875471595439395e-06,
+      "loss": 0.4627,
+      "step": 8406
+    },
+    {
+      "epoch": 1.3806581405374336,
+      "grad_norm": 0.41355222116583273,
+      "learning_rate": 7.4871964356291015e-06,
+      "loss": 0.5126,
+      "step": 8407
+    },
+    {
+      "epoch": 1.3808223677457763,
+      "grad_norm": 0.43117946888282094,
+      "learning_rate": 7.4868456792703715e-06,
+      "loss": 0.4567,
+      "step": 8408
+    },
+    {
+      "epoch": 1.380986594954119,
+      "grad_norm": 0.5106408016749285,
+      "learning_rate": 7.486494890471557e-06,
+      "loss": 0.4632,
+      "step": 8409
+    },
+    {
+      "epoch": 1.3811508221624618,
+      "grad_norm": 0.4436767147179429,
+      "learning_rate": 7.48614406923647e-06,
+      "loss": 0.4679,
+      "step": 8410
+    },
+    {
+      "epoch": 1.3813150493708046,
+      "grad_norm": 0.4034600956616641,
+      "learning_rate": 7.4857932155689216e-06,
+      "loss": 0.4875,
+      "step": 8411
+    },
+    {
+      "epoch": 1.3814792765791473,
+      "grad_norm": 0.33297679444132877,
+      "learning_rate": 7.485442329472721e-06,
+      "loss": 0.4634,
+      "step": 8412
+    },
+    {
+      "epoch": 1.38164350378749,
+      "grad_norm": 0.3543346865632646,
+      "learning_rate": 7.485091410951679e-06,
+      "loss": 0.4761,
+      "step": 8413
+    },
+    {
+      "epoch": 1.3818077309958328,
+      "grad_norm": 0.34842494937452245,
+      "learning_rate": 7.484740460009608e-06,
+      "loss": 0.4822,
+      "step": 8414
+    },
+    {
+      "epoch": 1.3819719582041754,
+      "grad_norm": 0.36443048186557747,
+      "learning_rate": 7.484389476650317e-06,
+      "loss": 0.4781,
+      "step": 8415
+    },
+    {
+      "epoch": 1.3821361854125183,
+      "grad_norm": 0.343766357021871,
+      "learning_rate": 7.484038460877623e-06,
+      "loss": 0.4706,
+      "step": 8416
+    },
+    {
+      "epoch": 1.382300412620861,
+      "grad_norm": 0.32675746001653816,
+      "learning_rate": 7.483687412695334e-06,
+      "loss": 0.4476,
+      "step": 8417
+    },
+    {
+      "epoch": 1.3824646398292038,
+      "grad_norm": 0.3221113117267582,
+      "learning_rate": 7.483336332107262e-06,
+      "loss": 0.4647,
+      "step": 8418
+    },
+    {
+      "epoch": 1.3826288670375464,
+      "grad_norm": 0.3215797449648108,
+      "learning_rate": 7.482985219117225e-06,
+      "loss": 0.4721,
+      "step": 8419
+    },
+    {
+      "epoch": 1.3827930942458893,
+      "grad_norm": 0.2928507581174783,
+      "learning_rate": 7.482634073729034e-06,
+      "loss": 0.4581,
+      "step": 8420
+    },
+    {
+      "epoch": 1.382957321454232,
+      "grad_norm": 0.35576242260440516,
+      "learning_rate": 7.482282895946501e-06,
+      "loss": 0.4617,
+      "step": 8421
+    },
+    {
+      "epoch": 1.3831215486625748,
+      "grad_norm": 0.3366737168411468,
+      "learning_rate": 7.481931685773442e-06,
+      "loss": 0.4536,
+      "step": 8422
+    },
+    {
+      "epoch": 1.3832857758709174,
+      "grad_norm": 0.46285205063668183,
+      "learning_rate": 7.481580443213671e-06,
+      "loss": 0.4628,
+      "step": 8423
+    },
+    {
+      "epoch": 1.3834500030792602,
+      "grad_norm": 0.2907414212608209,
+      "learning_rate": 7.481229168271003e-06,
+      "loss": 0.4778,
+      "step": 8424
+    },
+    {
+      "epoch": 1.3836142302876029,
+      "grad_norm": 0.3830255011938799,
+      "learning_rate": 7.480877860949253e-06,
+      "loss": 0.483,
+      "step": 8425
+    },
+    {
+      "epoch": 1.3837784574959455,
+      "grad_norm": 0.3698854430170395,
+      "learning_rate": 7.480526521252237e-06,
+      "loss": 0.4766,
+      "step": 8426
+    },
+    {
+      "epoch": 1.3839426847042884,
+      "grad_norm": 0.41376734309397795,
+      "learning_rate": 7.480175149183771e-06,
+      "loss": 0.491,
+      "step": 8427
+    },
+    {
+      "epoch": 1.3841069119126312,
+      "grad_norm": 0.39701920731416607,
+      "learning_rate": 7.479823744747669e-06,
+      "loss": 0.4913,
+      "step": 8428
+    },
+    {
+      "epoch": 1.3842711391209739,
+      "grad_norm": 0.2992212234035658,
+      "learning_rate": 7.479472307947752e-06,
+      "loss": 0.4996,
+      "step": 8429
+    },
+    {
+      "epoch": 1.3844353663293165,
+      "grad_norm": 0.3945568889015645,
+      "learning_rate": 7.479120838787832e-06,
+      "loss": 0.4843,
+      "step": 8430
+    },
+    {
+      "epoch": 1.3845995935376594,
+      "grad_norm": 0.32718184784481347,
+      "learning_rate": 7.478769337271729e-06,
+      "loss": 0.4688,
+      "step": 8431
+    },
+    {
+      "epoch": 1.384763820746002,
+      "grad_norm": 0.40322396401045213,
+      "learning_rate": 7.478417803403262e-06,
+      "loss": 0.4774,
+      "step": 8432
+    },
+    {
+      "epoch": 1.3849280479543449,
+      "grad_norm": 0.31278115808345475,
+      "learning_rate": 7.4780662371862454e-06,
+      "loss": 0.4887,
+      "step": 8433
+    },
+    {
+      "epoch": 1.3850922751626875,
+      "grad_norm": 0.2880177748586995,
+      "learning_rate": 7.4777146386245e-06,
+      "loss": 0.4622,
+      "step": 8434
+    },
+    {
+      "epoch": 1.3852565023710304,
+      "grad_norm": 0.2922145141022478,
+      "learning_rate": 7.477363007721842e-06,
+      "loss": 0.472,
+      "step": 8435
+    },
+    {
+      "epoch": 1.385420729579373,
+      "grad_norm": 0.2595927763167588,
+      "learning_rate": 7.477011344482097e-06,
+      "loss": 0.4598,
+      "step": 8436
+    },
+    {
+      "epoch": 1.3855849567877159,
+      "grad_norm": 0.3827827123321603,
+      "learning_rate": 7.4766596489090765e-06,
+      "loss": 0.4768,
+      "step": 8437
+    },
+    {
+      "epoch": 1.3857491839960585,
+      "grad_norm": 0.3046199138275548,
+      "learning_rate": 7.476307921006603e-06,
+      "loss": 0.4708,
+      "step": 8438
+    },
+    {
+      "epoch": 1.3859134112044014,
+      "grad_norm": 0.36047761509934007,
+      "learning_rate": 7.475956160778499e-06,
+      "loss": 0.4779,
+      "step": 8439
+    },
+    {
+      "epoch": 1.386077638412744,
+      "grad_norm": 0.2876281404368738,
+      "learning_rate": 7.475604368228583e-06,
+      "loss": 0.4676,
+      "step": 8440
+    },
+    {
+      "epoch": 1.3862418656210869,
+      "grad_norm": 0.3415868267007575,
+      "learning_rate": 7.475252543360676e-06,
+      "loss": 0.4821,
+      "step": 8441
+    },
+    {
+      "epoch": 1.3864060928294295,
+      "grad_norm": 0.3071306454877626,
+      "learning_rate": 7.474900686178598e-06,
+      "loss": 0.4661,
+      "step": 8442
+    },
+    {
+      "epoch": 1.3865703200377721,
+      "grad_norm": 0.3816098854453923,
+      "learning_rate": 7.474548796686172e-06,
+      "loss": 0.4777,
+      "step": 8443
+    },
+    {
+      "epoch": 1.386734547246115,
+      "grad_norm": 0.26175587149765556,
+      "learning_rate": 7.474196874887219e-06,
+      "loss": 0.4637,
+      "step": 8444
+    },
+    {
+      "epoch": 1.3868987744544579,
+      "grad_norm": 0.3957223210428203,
+      "learning_rate": 7.473844920785564e-06,
+      "loss": 0.4536,
+      "step": 8445
+    },
+    {
+      "epoch": 1.3870630016628005,
+      "grad_norm": 0.42887048967429414,
+      "learning_rate": 7.473492934385025e-06,
+      "loss": 0.4608,
+      "step": 8446
+    },
+    {
+      "epoch": 1.3872272288711431,
+      "grad_norm": 0.3039193509641038,
+      "learning_rate": 7.473140915689428e-06,
+      "loss": 0.487,
+      "step": 8447
+    },
+    {
+      "epoch": 1.387391456079486,
+      "grad_norm": 0.2924723132481886,
+      "learning_rate": 7.472788864702596e-06,
+      "loss": 0.4658,
+      "step": 8448
+    },
+    {
+      "epoch": 1.3875556832878286,
+      "grad_norm": 0.3104130699213828,
+      "learning_rate": 7.4724367814283515e-06,
+      "loss": 0.4661,
+      "step": 8449
+    },
+    {
+      "epoch": 1.3877199104961715,
+      "grad_norm": 0.2516987078850876,
+      "learning_rate": 7.472084665870519e-06,
+      "loss": 0.459,
+      "step": 8450
+    },
+    {
+      "epoch": 1.3878841377045141,
+      "grad_norm": 0.3246890939691412,
+      "learning_rate": 7.4717325180329246e-06,
+      "loss": 0.481,
+      "step": 8451
+    },
+    {
+      "epoch": 1.388048364912857,
+      "grad_norm": 0.3079833527004406,
+      "learning_rate": 7.4713803379193885e-06,
+      "loss": 0.4644,
+      "step": 8452
+    },
+    {
+      "epoch": 1.3882125921211996,
+      "grad_norm": 0.29180556133406804,
+      "learning_rate": 7.47102812553374e-06,
+      "loss": 0.4614,
+      "step": 8453
+    },
+    {
+      "epoch": 1.3883768193295425,
+      "grad_norm": 0.2886969888932696,
+      "learning_rate": 7.470675880879802e-06,
+      "loss": 0.4736,
+      "step": 8454
+    },
+    {
+      "epoch": 1.3885410465378851,
+      "grad_norm": 0.32198851042761334,
+      "learning_rate": 7.470323603961402e-06,
+      "loss": 0.4488,
+      "step": 8455
+    },
+    {
+      "epoch": 1.388705273746228,
+      "grad_norm": 0.27850170833038657,
+      "learning_rate": 7.469971294782366e-06,
+      "loss": 0.4719,
+      "step": 8456
+    },
+    {
+      "epoch": 1.3888695009545706,
+      "grad_norm": 0.3167504981040049,
+      "learning_rate": 7.469618953346519e-06,
+      "loss": 0.4733,
+      "step": 8457
+    },
+    {
+      "epoch": 1.3890337281629135,
+      "grad_norm": 0.2952002433702235,
+      "learning_rate": 7.469266579657688e-06,
+      "loss": 0.4689,
+      "step": 8458
+    },
+    {
+      "epoch": 1.3891979553712561,
+      "grad_norm": 0.3653309348288008,
+      "learning_rate": 7.468914173719701e-06,
+      "loss": 0.4991,
+      "step": 8459
+    },
+    {
+      "epoch": 1.3893621825795988,
+      "grad_norm": 0.33786108288862315,
+      "learning_rate": 7.468561735536384e-06,
+      "loss": 0.4691,
+      "step": 8460
+    },
+    {
+      "epoch": 1.3895264097879416,
+      "grad_norm": 0.38459079965390597,
+      "learning_rate": 7.468209265111568e-06,
+      "loss": 0.4428,
+      "step": 8461
+    },
+    {
+      "epoch": 1.3896906369962845,
+      "grad_norm": 0.26526099944477444,
+      "learning_rate": 7.467856762449077e-06,
+      "loss": 0.4388,
+      "step": 8462
+    },
+    {
+      "epoch": 1.3898548642046271,
+      "grad_norm": 0.29889257287695886,
+      "learning_rate": 7.467504227552743e-06,
+      "loss": 0.4768,
+      "step": 8463
+    },
+    {
+      "epoch": 1.3900190914129698,
+      "grad_norm": 0.4112551605074209,
+      "learning_rate": 7.467151660426393e-06,
+      "loss": 0.473,
+      "step": 8464
+    },
+    {
+      "epoch": 1.3901833186213126,
+      "grad_norm": 0.41186499581502484,
+      "learning_rate": 7.466799061073857e-06,
+      "loss": 0.475,
+      "step": 8465
+    },
+    {
+      "epoch": 1.3903475458296553,
+      "grad_norm": 0.5636465016362382,
+      "learning_rate": 7.466446429498963e-06,
+      "loss": 0.469,
+      "step": 8466
+    },
+    {
+      "epoch": 1.3905117730379981,
+      "grad_norm": 0.32784216972789626,
+      "learning_rate": 7.4660937657055426e-06,
+      "loss": 0.4753,
+      "step": 8467
+    },
+    {
+      "epoch": 1.3906760002463407,
+      "grad_norm": 0.357908475606493,
+      "learning_rate": 7.465741069697428e-06,
+      "loss": 0.488,
+      "step": 8468
+    },
+    {
+      "epoch": 1.3908402274546836,
+      "grad_norm": 0.33925619079683195,
+      "learning_rate": 7.4653883414784445e-06,
+      "loss": 0.472,
+      "step": 8469
+    },
+    {
+      "epoch": 1.3910044546630262,
+      "grad_norm": 0.29056325498193813,
+      "learning_rate": 7.465035581052428e-06,
+      "loss": 0.4641,
+      "step": 8470
+    },
+    {
+      "epoch": 1.391168681871369,
+      "grad_norm": 0.3130794677226531,
+      "learning_rate": 7.464682788423206e-06,
+      "loss": 0.4607,
+      "step": 8471
+    },
+    {
+      "epoch": 1.3913329090797117,
+      "grad_norm": 0.2999242321369178,
+      "learning_rate": 7.464329963594613e-06,
+      "loss": 0.4615,
+      "step": 8472
+    },
+    {
+      "epoch": 1.3914971362880546,
+      "grad_norm": 0.31033618826953613,
+      "learning_rate": 7.463977106570481e-06,
+      "loss": 0.4693,
+      "step": 8473
+    },
+    {
+      "epoch": 1.3916613634963972,
+      "grad_norm": 0.40690685776686436,
+      "learning_rate": 7.463624217354641e-06,
+      "loss": 0.4811,
+      "step": 8474
+    },
+    {
+      "epoch": 1.39182559070474,
+      "grad_norm": 0.5071406863141119,
+      "learning_rate": 7.463271295950926e-06,
+      "loss": 0.4661,
+      "step": 8475
+    },
+    {
+      "epoch": 1.3919898179130827,
+      "grad_norm": 0.33631701642143286,
+      "learning_rate": 7.462918342363169e-06,
+      "loss": 0.4771,
+      "step": 8476
+    },
+    {
+      "epoch": 1.3921540451214254,
+      "grad_norm": 0.30037934836463953,
+      "learning_rate": 7.462565356595202e-06,
+      "loss": 0.4802,
+      "step": 8477
+    },
+    {
+      "epoch": 1.3923182723297682,
+      "grad_norm": 0.3193277779391227,
+      "learning_rate": 7.4622123386508635e-06,
+      "loss": 0.4876,
+      "step": 8478
+    },
+    {
+      "epoch": 1.392482499538111,
+      "grad_norm": 0.270676534414637,
+      "learning_rate": 7.461859288533983e-06,
+      "loss": 0.4594,
+      "step": 8479
+    },
+    {
+      "epoch": 1.3926467267464537,
+      "grad_norm": 0.4127141520008906,
+      "learning_rate": 7.461506206248397e-06,
+      "loss": 0.4653,
+      "step": 8480
+    },
+    {
+      "epoch": 1.3928109539547964,
+      "grad_norm": 0.4991721815266884,
+      "learning_rate": 7.46115309179794e-06,
+      "loss": 0.4844,
+      "step": 8481
+    },
+    {
+      "epoch": 1.3929751811631392,
+      "grad_norm": 0.29698984397598005,
+      "learning_rate": 7.460799945186447e-06,
+      "loss": 0.4613,
+      "step": 8482
+    },
+    {
+      "epoch": 1.3931394083714819,
+      "grad_norm": 0.3248795253734255,
+      "learning_rate": 7.4604467664177535e-06,
+      "loss": 0.4638,
+      "step": 8483
+    },
+    {
+      "epoch": 1.3933036355798247,
+      "grad_norm": 0.37999833355929574,
+      "learning_rate": 7.460093555495695e-06,
+      "loss": 0.483,
+      "step": 8484
+    },
+    {
+      "epoch": 1.3934678627881674,
+      "grad_norm": 0.3038401995582087,
+      "learning_rate": 7.459740312424108e-06,
+      "loss": 0.4846,
+      "step": 8485
+    },
+    {
+      "epoch": 1.3936320899965102,
+      "grad_norm": 0.33297402648887964,
+      "learning_rate": 7.459387037206831e-06,
+      "loss": 0.4907,
+      "step": 8486
+    },
+    {
+      "epoch": 1.3937963172048529,
+      "grad_norm": 0.5686034853236254,
+      "learning_rate": 7.459033729847697e-06,
+      "loss": 0.4726,
+      "step": 8487
+    },
+    {
+      "epoch": 1.3939605444131957,
+      "grad_norm": 0.3948521265247895,
+      "learning_rate": 7.458680390350547e-06,
+      "loss": 0.4641,
+      "step": 8488
+    },
+    {
+      "epoch": 1.3941247716215384,
+      "grad_norm": 0.45439200880210123,
+      "learning_rate": 7.458327018719214e-06,
+      "loss": 0.4778,
+      "step": 8489
+    },
+    {
+      "epoch": 1.3942889988298812,
+      "grad_norm": 0.3042339933458967,
+      "learning_rate": 7.457973614957541e-06,
+      "loss": 0.4634,
+      "step": 8490
+    },
+    {
+      "epoch": 1.3944532260382239,
+      "grad_norm": 0.3305137784668754,
+      "learning_rate": 7.457620179069366e-06,
+      "loss": 0.4842,
+      "step": 8491
+    },
+    {
+      "epoch": 1.3946174532465667,
+      "grad_norm": 0.27760176510723344,
+      "learning_rate": 7.457266711058524e-06,
+      "loss": 0.4752,
+      "step": 8492
+    },
+    {
+      "epoch": 1.3947816804549094,
+      "grad_norm": 0.4102659742863658,
+      "learning_rate": 7.4569132109288555e-06,
+      "loss": 0.4685,
+      "step": 8493
+    },
+    {
+      "epoch": 1.394945907663252,
+      "grad_norm": 0.31841425627424486,
+      "learning_rate": 7.456559678684201e-06,
+      "loss": 0.469,
+      "step": 8494
+    },
+    {
+      "epoch": 1.3951101348715949,
+      "grad_norm": 0.2770736991011022,
+      "learning_rate": 7.456206114328402e-06,
+      "loss": 0.4522,
+      "step": 8495
+    },
+    {
+      "epoch": 1.3952743620799377,
+      "grad_norm": 0.36425396742082483,
+      "learning_rate": 7.455852517865292e-06,
+      "loss": 0.4661,
+      "step": 8496
+    },
+    {
+      "epoch": 1.3954385892882804,
+      "grad_norm": 0.29309115832848304,
+      "learning_rate": 7.4554988892987186e-06,
+      "loss": 0.4566,
+      "step": 8497
+    },
+    {
+      "epoch": 1.395602816496623,
+      "grad_norm": 0.3467605734878834,
+      "learning_rate": 7.455145228632518e-06,
+      "loss": 0.4851,
+      "step": 8498
+    },
+    {
+      "epoch": 1.3957670437049658,
+      "grad_norm": 0.2803211107231899,
+      "learning_rate": 7.454791535870533e-06,
+      "loss": 0.4698,
+      "step": 8499
+    },
+    {
+      "epoch": 1.3959312709133085,
+      "grad_norm": 0.30707032322686445,
+      "learning_rate": 7.454437811016605e-06,
+      "loss": 0.4637,
+      "step": 8500
+    },
+    {
+      "epoch": 1.3960954981216513,
+      "grad_norm": 0.292408272146633,
+      "learning_rate": 7.454084054074575e-06,
+      "loss": 0.4647,
+      "step": 8501
+    },
+    {
+      "epoch": 1.396259725329994,
+      "grad_norm": 0.3176060833917574,
+      "learning_rate": 7.453730265048285e-06,
+      "loss": 0.4789,
+      "step": 8502
+    },
+    {
+      "epoch": 1.3964239525383368,
+      "grad_norm": 0.2712677509040151,
+      "learning_rate": 7.45337644394158e-06,
+      "loss": 0.4637,
+      "step": 8503
+    },
+    {
+      "epoch": 1.3965881797466795,
+      "grad_norm": 0.2992868914528555,
+      "learning_rate": 7.453022590758301e-06,
+      "loss": 0.4662,
+      "step": 8504
+    },
+    {
+      "epoch": 1.3967524069550223,
+      "grad_norm": 0.2862272436304518,
+      "learning_rate": 7.45266870550229e-06,
+      "loss": 0.4684,
+      "step": 8505
+    },
+    {
+      "epoch": 1.396916634163365,
+      "grad_norm": 0.34703558397614,
+      "learning_rate": 7.452314788177391e-06,
+      "loss": 0.4745,
+      "step": 8506
+    },
+    {
+      "epoch": 1.3970808613717078,
+      "grad_norm": 0.3593107459557686,
+      "learning_rate": 7.451960838787452e-06,
+      "loss": 0.4701,
+      "step": 8507
+    },
+    {
+      "epoch": 1.3972450885800505,
+      "grad_norm": 0.27753636142750543,
+      "learning_rate": 7.451606857336312e-06,
+      "loss": 0.454,
+      "step": 8508
+    },
+    {
+      "epoch": 1.3974093157883933,
+      "grad_norm": 0.32519945938169814,
+      "learning_rate": 7.4512528438278174e-06,
+      "loss": 0.4653,
+      "step": 8509
+    },
+    {
+      "epoch": 1.397573542996736,
+      "grad_norm": 0.37636601508996753,
+      "learning_rate": 7.4508987982658135e-06,
+      "loss": 0.4723,
+      "step": 8510
+    },
+    {
+      "epoch": 1.3977377702050786,
+      "grad_norm": 0.3242174929140608,
+      "learning_rate": 7.450544720654145e-06,
+      "loss": 0.4673,
+      "step": 8511
+    },
+    {
+      "epoch": 1.3979019974134215,
+      "grad_norm": 0.3152017954856842,
+      "learning_rate": 7.4501906109966595e-06,
+      "loss": 0.4684,
+      "step": 8512
+    },
+    {
+      "epoch": 1.3980662246217643,
+      "grad_norm": 0.273731924885685,
+      "learning_rate": 7.4498364692971996e-06,
+      "loss": 0.4686,
+      "step": 8513
+    },
+    {
+      "epoch": 1.398230451830107,
+      "grad_norm": 0.4412982637682514,
+      "learning_rate": 7.449482295559614e-06,
+      "loss": 0.4689,
+      "step": 8514
+    },
+    {
+      "epoch": 1.3983946790384496,
+      "grad_norm": 0.6275413982705256,
+      "learning_rate": 7.4491280897877475e-06,
+      "loss": 0.4428,
+      "step": 8515
+    },
+    {
+      "epoch": 1.3985589062467925,
+      "grad_norm": 0.3754192144297303,
+      "learning_rate": 7.44877385198545e-06,
+      "loss": 0.4798,
+      "step": 8516
+    },
+    {
+      "epoch": 1.398723133455135,
+      "grad_norm": 1.2042144009390472,
+      "learning_rate": 7.448419582156568e-06,
+      "loss": 0.4711,
+      "step": 8517
+    },
+    {
+      "epoch": 1.398887360663478,
+      "grad_norm": 0.32201331477470474,
+      "learning_rate": 7.448065280304946e-06,
+      "loss": 0.4691,
+      "step": 8518
+    },
+    {
+      "epoch": 1.3990515878718206,
+      "grad_norm": 0.36748107501569693,
+      "learning_rate": 7.447710946434438e-06,
+      "loss": 0.4736,
+      "step": 8519
+    },
+    {
+      "epoch": 1.3992158150801635,
+      "grad_norm": 0.30196407746713133,
+      "learning_rate": 7.447356580548886e-06,
+      "loss": 0.461,
+      "step": 8520
+    },
+    {
+      "epoch": 1.399380042288506,
+      "grad_norm": 0.40934242162584306,
+      "learning_rate": 7.447002182652143e-06,
+      "loss": 0.4674,
+      "step": 8521
+    },
+    {
+      "epoch": 1.399544269496849,
+      "grad_norm": 0.2764712262536976,
+      "learning_rate": 7.446647752748056e-06,
+      "loss": 0.442,
+      "step": 8522
+    },
+    {
+      "epoch": 1.3997084967051916,
+      "grad_norm": 0.2964002281736302,
+      "learning_rate": 7.446293290840475e-06,
+      "loss": 0.466,
+      "step": 8523
+    },
+    {
+      "epoch": 1.3998727239135345,
+      "grad_norm": 0.31240139575336234,
+      "learning_rate": 7.4459387969332514e-06,
+      "loss": 0.4669,
+      "step": 8524
+    },
+    {
+      "epoch": 1.400036951121877,
+      "grad_norm": 0.5236529764809225,
+      "learning_rate": 7.4455842710302346e-06,
+      "loss": 0.4811,
+      "step": 8525
+    },
+    {
+      "epoch": 1.40020117833022,
+      "grad_norm": 0.2690862497073942,
+      "learning_rate": 7.445229713135273e-06,
+      "loss": 0.4989,
+      "step": 8526
+    },
+    {
+      "epoch": 1.4003654055385626,
+      "grad_norm": 0.31760863779848236,
+      "learning_rate": 7.444875123252219e-06,
+      "loss": 0.4686,
+      "step": 8527
+    },
+    {
+      "epoch": 1.4005296327469052,
+      "grad_norm": 0.3106443897577975,
+      "learning_rate": 7.444520501384925e-06,
+      "loss": 0.4796,
+      "step": 8528
+    },
+    {
+      "epoch": 1.400693859955248,
+      "grad_norm": 0.45939652043207874,
+      "learning_rate": 7.44416584753724e-06,
+      "loss": 0.465,
+      "step": 8529
+    },
+    {
+      "epoch": 1.400858087163591,
+      "grad_norm": 0.3001121880109803,
+      "learning_rate": 7.443811161713018e-06,
+      "loss": 0.4606,
+      "step": 8530
+    },
+    {
+      "epoch": 1.4010223143719336,
+      "grad_norm": 0.32939005836550994,
+      "learning_rate": 7.443456443916111e-06,
+      "loss": 0.4665,
+      "step": 8531
+    },
+    {
+      "epoch": 1.4011865415802762,
+      "grad_norm": 0.31886006819648094,
+      "learning_rate": 7.443101694150371e-06,
+      "loss": 0.4913,
+      "step": 8532
+    },
+    {
+      "epoch": 1.401350768788619,
+      "grad_norm": 0.2781872322411299,
+      "learning_rate": 7.442746912419649e-06,
+      "loss": 0.4668,
+      "step": 8533
+    },
+    {
+      "epoch": 1.4015149959969617,
+      "grad_norm": 0.9270944124758137,
+      "learning_rate": 7.442392098727801e-06,
+      "loss": 0.4763,
+      "step": 8534
+    },
+    {
+      "epoch": 1.4016792232053046,
+      "grad_norm": 0.35113450993008694,
+      "learning_rate": 7.442037253078681e-06,
+      "loss": 0.465,
+      "step": 8535
+    },
+    {
+      "epoch": 1.4018434504136472,
+      "grad_norm": 0.4559463003574399,
+      "learning_rate": 7.441682375476141e-06,
+      "loss": 0.4837,
+      "step": 8536
+    },
+    {
+      "epoch": 1.40200767762199,
+      "grad_norm": 0.4216442006644967,
+      "learning_rate": 7.441327465924038e-06,
+      "loss": 0.4726,
+      "step": 8537
+    },
+    {
+      "epoch": 1.4021719048303327,
+      "grad_norm": 0.2786121891687586,
+      "learning_rate": 7.440972524426222e-06,
+      "loss": 0.4744,
+      "step": 8538
+    },
+    {
+      "epoch": 1.4023361320386756,
+      "grad_norm": 0.38727303750505243,
+      "learning_rate": 7.440617550986552e-06,
+      "loss": 0.4792,
+      "step": 8539
+    },
+    {
+      "epoch": 1.4025003592470182,
+      "grad_norm": 0.30560195767242304,
+      "learning_rate": 7.4402625456088826e-06,
+      "loss": 0.4695,
+      "step": 8540
+    },
+    {
+      "epoch": 1.402664586455361,
+      "grad_norm": 0.28880668327933817,
+      "learning_rate": 7.43990750829707e-06,
+      "loss": 0.4721,
+      "step": 8541
+    },
+    {
+      "epoch": 1.4028288136637037,
+      "grad_norm": 0.4172695597566841,
+      "learning_rate": 7.439552439054967e-06,
+      "loss": 0.4622,
+      "step": 8542
+    },
+    {
+      "epoch": 1.4029930408720466,
+      "grad_norm": 0.30842138351354353,
+      "learning_rate": 7.439197337886435e-06,
+      "loss": 0.4837,
+      "step": 8543
+    },
+    {
+      "epoch": 1.4031572680803892,
+      "grad_norm": 0.29297139265934846,
+      "learning_rate": 7.438842204795327e-06,
+      "loss": 0.4716,
+      "step": 8544
+    },
+    {
+      "epoch": 1.4033214952887318,
+      "grad_norm": 0.59448376485058,
+      "learning_rate": 7.4384870397855e-06,
+      "loss": 0.4695,
+      "step": 8545
+    },
+    {
+      "epoch": 1.4034857224970747,
+      "grad_norm": 0.27269194109427797,
+      "learning_rate": 7.438131842860813e-06,
+      "loss": 0.4929,
+      "step": 8546
+    },
+    {
+      "epoch": 1.4036499497054176,
+      "grad_norm": 0.341391040294893,
+      "learning_rate": 7.437776614025125e-06,
+      "loss": 0.4908,
+      "step": 8547
+    },
+    {
+      "epoch": 1.4038141769137602,
+      "grad_norm": 0.32200140091490614,
+      "learning_rate": 7.4374213532822915e-06,
+      "loss": 0.4745,
+      "step": 8548
+    },
+    {
+      "epoch": 1.4039784041221028,
+      "grad_norm": 0.2661207928603546,
+      "learning_rate": 7.437066060636174e-06,
+      "loss": 0.4899,
+      "step": 8549
+    },
+    {
+      "epoch": 1.4041426313304457,
+      "grad_norm": 0.29696741496995765,
+      "learning_rate": 7.436710736090627e-06,
+      "loss": 0.4584,
+      "step": 8550
+    },
+    {
+      "epoch": 1.4043068585387883,
+      "grad_norm": 0.29306385562883797,
+      "learning_rate": 7.436355379649513e-06,
+      "loss": 0.4588,
+      "step": 8551
+    },
+    {
+      "epoch": 1.4044710857471312,
+      "grad_norm": 0.3970937075104934,
+      "learning_rate": 7.43599999131669e-06,
+      "loss": 0.4891,
+      "step": 8552
+    },
+    {
+      "epoch": 1.4046353129554738,
+      "grad_norm": 0.34058582880145505,
+      "learning_rate": 7.435644571096019e-06,
+      "loss": 0.4523,
+      "step": 8553
+    },
+    {
+      "epoch": 1.4047995401638167,
+      "grad_norm": 0.45701364527626853,
+      "learning_rate": 7.435289118991359e-06,
+      "loss": 0.4677,
+      "step": 8554
+    },
+    {
+      "epoch": 1.4049637673721593,
+      "grad_norm": 0.31318959131377144,
+      "learning_rate": 7.434933635006573e-06,
+      "loss": 0.4598,
+      "step": 8555
+    },
+    {
+      "epoch": 1.4051279945805022,
+      "grad_norm": 0.3489953705021745,
+      "learning_rate": 7.4345781191455184e-06,
+      "loss": 0.4679,
+      "step": 8556
+    },
+    {
+      "epoch": 1.4052922217888448,
+      "grad_norm": 0.5725145140493672,
+      "learning_rate": 7.434222571412059e-06,
+      "loss": 0.4708,
+      "step": 8557
+    },
+    {
+      "epoch": 1.4054564489971877,
+      "grad_norm": 0.2986176145115553,
+      "learning_rate": 7.433866991810055e-06,
+      "loss": 0.4637,
+      "step": 8558
+    },
+    {
+      "epoch": 1.4056206762055303,
+      "grad_norm": 0.3467896030354416,
+      "learning_rate": 7.433511380343369e-06,
+      "loss": 0.4869,
+      "step": 8559
+    },
+    {
+      "epoch": 1.4057849034138732,
+      "grad_norm": 0.5115851153170545,
+      "learning_rate": 7.433155737015863e-06,
+      "loss": 0.4801,
+      "step": 8560
+    },
+    {
+      "epoch": 1.4059491306222158,
+      "grad_norm": 0.3898500557859638,
+      "learning_rate": 7.432800061831401e-06,
+      "loss": 0.4636,
+      "step": 8561
+    },
+    {
+      "epoch": 1.4061133578305585,
+      "grad_norm": 0.3072252086568954,
+      "learning_rate": 7.432444354793844e-06,
+      "loss": 0.4574,
+      "step": 8562
+    },
+    {
+      "epoch": 1.4062775850389013,
+      "grad_norm": 0.38470436228122484,
+      "learning_rate": 7.432088615907057e-06,
+      "loss": 0.4824,
+      "step": 8563
+    },
+    {
+      "epoch": 1.4064418122472442,
+      "grad_norm": 0.35917326620407675,
+      "learning_rate": 7.431732845174901e-06,
+      "loss": 0.4609,
+      "step": 8564
+    },
+    {
+      "epoch": 1.4066060394555868,
+      "grad_norm": 0.3190169868757679,
+      "learning_rate": 7.4313770426012435e-06,
+      "loss": 0.4755,
+      "step": 8565
+    },
+    {
+      "epoch": 1.4067702666639295,
+      "grad_norm": 0.30937207655898136,
+      "learning_rate": 7.4310212081899475e-06,
+      "loss": 0.4747,
+      "step": 8566
+    },
+    {
+      "epoch": 1.4069344938722723,
+      "grad_norm": 0.30277533292576975,
+      "learning_rate": 7.430665341944877e-06,
+      "loss": 0.4801,
+      "step": 8567
+    },
+    {
+      "epoch": 1.407098721080615,
+      "grad_norm": 0.3104919922663232,
+      "learning_rate": 7.430309443869896e-06,
+      "loss": 0.4733,
+      "step": 8568
+    },
+    {
+      "epoch": 1.4072629482889578,
+      "grad_norm": 0.4631342370065941,
+      "learning_rate": 7.429953513968873e-06,
+      "loss": 0.4611,
+      "step": 8569
+    },
+    {
+      "epoch": 1.4074271754973005,
+      "grad_norm": 0.2910622862956142,
+      "learning_rate": 7.429597552245673e-06,
+      "loss": 0.4707,
+      "step": 8570
+    },
+    {
+      "epoch": 1.4075914027056433,
+      "grad_norm": 0.3070539773881218,
+      "learning_rate": 7.429241558704159e-06,
+      "loss": 0.4674,
+      "step": 8571
+    },
+    {
+      "epoch": 1.407755629913986,
+      "grad_norm": 0.3795506728011231,
+      "learning_rate": 7.428885533348201e-06,
+      "loss": 0.4866,
+      "step": 8572
+    },
+    {
+      "epoch": 1.4079198571223288,
+      "grad_norm": 0.3507655176583296,
+      "learning_rate": 7.428529476181664e-06,
+      "loss": 0.4611,
+      "step": 8573
+    },
+    {
+      "epoch": 1.4080840843306714,
+      "grad_norm": 0.34017407358390006,
+      "learning_rate": 7.428173387208416e-06,
+      "loss": 0.4712,
+      "step": 8574
+    },
+    {
+      "epoch": 1.4082483115390143,
+      "grad_norm": 0.30794027444638433,
+      "learning_rate": 7.427817266432324e-06,
+      "loss": 0.4707,
+      "step": 8575
+    },
+    {
+      "epoch": 1.408412538747357,
+      "grad_norm": 0.3143973834229577,
+      "learning_rate": 7.427461113857256e-06,
+      "loss": 0.451,
+      "step": 8576
+    },
+    {
+      "epoch": 1.4085767659556998,
+      "grad_norm": 0.5979945971913595,
+      "learning_rate": 7.42710492948708e-06,
+      "loss": 0.4639,
+      "step": 8577
+    },
+    {
+      "epoch": 1.4087409931640424,
+      "grad_norm": 0.2887234277853099,
+      "learning_rate": 7.426748713325664e-06,
+      "loss": 0.47,
+      "step": 8578
+    },
+    {
+      "epoch": 1.408905220372385,
+      "grad_norm": 0.33756145494788314,
+      "learning_rate": 7.426392465376879e-06,
+      "loss": 0.4775,
+      "step": 8579
+    },
+    {
+      "epoch": 1.409069447580728,
+      "grad_norm": 0.2800702567687854,
+      "learning_rate": 7.426036185644591e-06,
+      "loss": 0.4517,
+      "step": 8580
+    },
+    {
+      "epoch": 1.4092336747890708,
+      "grad_norm": 0.46678552703550924,
+      "learning_rate": 7.425679874132672e-06,
+      "loss": 0.4797,
+      "step": 8581
+    },
+    {
+      "epoch": 1.4093979019974134,
+      "grad_norm": 0.3356274085717647,
+      "learning_rate": 7.42532353084499e-06,
+      "loss": 0.4575,
+      "step": 8582
+    },
+    {
+      "epoch": 1.409562129205756,
+      "grad_norm": 0.3116451615310341,
+      "learning_rate": 7.424967155785418e-06,
+      "loss": 0.4616,
+      "step": 8583
+    },
+    {
+      "epoch": 1.409726356414099,
+      "grad_norm": 0.2696206283265771,
+      "learning_rate": 7.424610748957823e-06,
+      "loss": 0.4636,
+      "step": 8584
+    },
+    {
+      "epoch": 1.4098905836224416,
+      "grad_norm": 0.33772238707792296,
+      "learning_rate": 7.424254310366079e-06,
+      "loss": 0.4704,
+      "step": 8585
+    },
+    {
+      "epoch": 1.4100548108307844,
+      "grad_norm": 0.3055990131999076,
+      "learning_rate": 7.4238978400140565e-06,
+      "loss": 0.476,
+      "step": 8586
+    },
+    {
+      "epoch": 1.410219038039127,
+      "grad_norm": 0.3551794459285267,
+      "learning_rate": 7.423541337905626e-06,
+      "loss": 0.4573,
+      "step": 8587
+    },
+    {
+      "epoch": 1.41038326524747,
+      "grad_norm": 0.3186811436613112,
+      "learning_rate": 7.4231848040446605e-06,
+      "loss": 0.4709,
+      "step": 8588
+    },
+    {
+      "epoch": 1.4105474924558126,
+      "grad_norm": 0.49816413275093707,
+      "learning_rate": 7.4228282384350315e-06,
+      "loss": 0.4923,
+      "step": 8589
+    },
+    {
+      "epoch": 1.4107117196641554,
+      "grad_norm": 0.35030694382788935,
+      "learning_rate": 7.4224716410806126e-06,
+      "loss": 0.4781,
+      "step": 8590
+    },
+    {
+      "epoch": 1.410875946872498,
+      "grad_norm": 0.30048156452761454,
+      "learning_rate": 7.422115011985278e-06,
+      "loss": 0.4796,
+      "step": 8591
+    },
+    {
+      "epoch": 1.411040174080841,
+      "grad_norm": 0.3196508862449292,
+      "learning_rate": 7.421758351152898e-06,
+      "loss": 0.4611,
+      "step": 8592
+    },
+    {
+      "epoch": 1.4112044012891836,
+      "grad_norm": 0.4013596427282612,
+      "learning_rate": 7.421401658587347e-06,
+      "loss": 0.4694,
+      "step": 8593
+    },
+    {
+      "epoch": 1.4113686284975264,
+      "grad_norm": 0.2652885562504778,
+      "learning_rate": 7.4210449342924995e-06,
+      "loss": 0.4387,
+      "step": 8594
+    },
+    {
+      "epoch": 1.411532855705869,
+      "grad_norm": 0.30374163694020784,
+      "learning_rate": 7.4206881782722305e-06,
+      "loss": 0.4537,
+      "step": 8595
+    },
+    {
+      "epoch": 1.4116970829142117,
+      "grad_norm": 0.4612477117901834,
+      "learning_rate": 7.420331390530415e-06,
+      "loss": 0.4827,
+      "step": 8596
+    },
+    {
+      "epoch": 1.4118613101225546,
+      "grad_norm": 0.2844872627001613,
+      "learning_rate": 7.419974571070927e-06,
+      "loss": 0.4617,
+      "step": 8597
+    },
+    {
+      "epoch": 1.4120255373308974,
+      "grad_norm": 0.27905217514415914,
+      "learning_rate": 7.419617719897642e-06,
+      "loss": 0.4682,
+      "step": 8598
+    },
+    {
+      "epoch": 1.41218976453924,
+      "grad_norm": 0.27318352991817796,
+      "learning_rate": 7.4192608370144355e-06,
+      "loss": 0.4905,
+      "step": 8599
+    },
+    {
+      "epoch": 1.4123539917475827,
+      "grad_norm": 0.3544091023529306,
+      "learning_rate": 7.418903922425187e-06,
+      "loss": 0.4396,
+      "step": 8600
+    },
+    {
+      "epoch": 1.4125182189559256,
+      "grad_norm": 0.32549410255272687,
+      "learning_rate": 7.418546976133766e-06,
+      "loss": 0.457,
+      "step": 8601
+    },
+    {
+      "epoch": 1.4126824461642682,
+      "grad_norm": 0.33155795928650694,
+      "learning_rate": 7.418189998144056e-06,
+      "loss": 0.4621,
+      "step": 8602
+    },
+    {
+      "epoch": 1.412846673372611,
+      "grad_norm": 0.2963843090235932,
+      "learning_rate": 7.417832988459932e-06,
+      "loss": 0.4527,
+      "step": 8603
+    },
+    {
+      "epoch": 1.4130109005809537,
+      "grad_norm": 0.3009253973532018,
+      "learning_rate": 7.417475947085269e-06,
+      "loss": 0.4651,
+      "step": 8604
+    },
+    {
+      "epoch": 1.4131751277892965,
+      "grad_norm": 0.31122937905474224,
+      "learning_rate": 7.4171188740239475e-06,
+      "loss": 0.4778,
+      "step": 8605
+    },
+    {
+      "epoch": 1.4133393549976392,
+      "grad_norm": 0.35058399361945486,
+      "learning_rate": 7.416761769279846e-06,
+      "loss": 0.4452,
+      "step": 8606
+    },
+    {
+      "epoch": 1.413503582205982,
+      "grad_norm": 0.32076975303698535,
+      "learning_rate": 7.4164046328568404e-06,
+      "loss": 0.441,
+      "step": 8607
+    },
+    {
+      "epoch": 1.4136678094143247,
+      "grad_norm": 0.5070947218880237,
+      "learning_rate": 7.416047464758812e-06,
+      "loss": 0.4604,
+      "step": 8608
+    },
+    {
+      "epoch": 1.4138320366226675,
+      "grad_norm": 0.37100148779464004,
+      "learning_rate": 7.415690264989639e-06,
+      "loss": 0.4645,
+      "step": 8609
+    },
+    {
+      "epoch": 1.4139962638310102,
+      "grad_norm": 0.3308140782663967,
+      "learning_rate": 7.415333033553201e-06,
+      "loss": 0.47,
+      "step": 8610
+    },
+    {
+      "epoch": 1.414160491039353,
+      "grad_norm": 0.36507773865455445,
+      "learning_rate": 7.414975770453378e-06,
+      "loss": 0.4506,
+      "step": 8611
+    },
+    {
+      "epoch": 1.4143247182476957,
+      "grad_norm": 0.3049033070114644,
+      "learning_rate": 7.414618475694051e-06,
+      "loss": 0.4712,
+      "step": 8612
+    },
+    {
+      "epoch": 1.4144889454560383,
+      "grad_norm": 0.3199448512706157,
+      "learning_rate": 7.414261149279099e-06,
+      "loss": 0.4779,
+      "step": 8613
+    },
+    {
+      "epoch": 1.4146531726643812,
+      "grad_norm": 0.4204919239325658,
+      "learning_rate": 7.413903791212403e-06,
+      "loss": 0.4661,
+      "step": 8614
+    },
+    {
+      "epoch": 1.414817399872724,
+      "grad_norm": 0.31521651354486385,
+      "learning_rate": 7.413546401497846e-06,
+      "loss": 0.4816,
+      "step": 8615
+    },
+    {
+      "epoch": 1.4149816270810667,
+      "grad_norm": 0.35333273072663923,
+      "learning_rate": 7.413188980139309e-06,
+      "loss": 0.4628,
+      "step": 8616
+    },
+    {
+      "epoch": 1.4151458542894093,
+      "grad_norm": 0.289281221874642,
+      "learning_rate": 7.412831527140672e-06,
+      "loss": 0.4563,
+      "step": 8617
+    },
+    {
+      "epoch": 1.4153100814977522,
+      "grad_norm": 0.32144669061901465,
+      "learning_rate": 7.412474042505819e-06,
+      "loss": 0.4588,
+      "step": 8618
+    },
+    {
+      "epoch": 1.4154743087060948,
+      "grad_norm": 0.46956177346878347,
+      "learning_rate": 7.412116526238633e-06,
+      "loss": 0.4575,
+      "step": 8619
+    },
+    {
+      "epoch": 1.4156385359144377,
+      "grad_norm": 0.38669633047785396,
+      "learning_rate": 7.411758978342996e-06,
+      "loss": 0.4939,
+      "step": 8620
+    },
+    {
+      "epoch": 1.4158027631227803,
+      "grad_norm": 0.3337701977281101,
+      "learning_rate": 7.411401398822792e-06,
+      "loss": 0.455,
+      "step": 8621
+    },
+    {
+      "epoch": 1.4159669903311232,
+      "grad_norm": 0.27205267115254267,
+      "learning_rate": 7.411043787681904e-06,
+      "loss": 0.4595,
+      "step": 8622
+    },
+    {
+      "epoch": 1.4161312175394658,
+      "grad_norm": 0.31901257150921336,
+      "learning_rate": 7.410686144924216e-06,
+      "loss": 0.4744,
+      "step": 8623
+    },
+    {
+      "epoch": 1.4162954447478087,
+      "grad_norm": 0.29990230673339774,
+      "learning_rate": 7.410328470553614e-06,
+      "loss": 0.4677,
+      "step": 8624
+    },
+    {
+      "epoch": 1.4164596719561513,
+      "grad_norm": 0.432034413195062,
+      "learning_rate": 7.409970764573981e-06,
+      "loss": 0.4718,
+      "step": 8625
+    },
+    {
+      "epoch": 1.4166238991644942,
+      "grad_norm": 0.45117493263050745,
+      "learning_rate": 7.409613026989202e-06,
+      "loss": 0.4503,
+      "step": 8626
+    },
+    {
+      "epoch": 1.4167881263728368,
+      "grad_norm": 0.288399028366705,
+      "learning_rate": 7.409255257803164e-06,
+      "loss": 0.477,
+      "step": 8627
+    },
+    {
+      "epoch": 1.4169523535811797,
+      "grad_norm": 0.2742665130026796,
+      "learning_rate": 7.40889745701975e-06,
+      "loss": 0.4793,
+      "step": 8628
+    },
+    {
+      "epoch": 1.4171165807895223,
+      "grad_norm": 0.28388631767832506,
+      "learning_rate": 7.408539624642849e-06,
+      "loss": 0.4707,
+      "step": 8629
+    },
+    {
+      "epoch": 1.417280807997865,
+      "grad_norm": 0.3762696631629608,
+      "learning_rate": 7.408181760676345e-06,
+      "loss": 0.4742,
+      "step": 8630
+    },
+    {
+      "epoch": 1.4174450352062078,
+      "grad_norm": 0.30604710663862517,
+      "learning_rate": 7.407823865124126e-06,
+      "loss": 0.4893,
+      "step": 8631
+    },
+    {
+      "epoch": 1.4176092624145507,
+      "grad_norm": 0.30730097863583844,
+      "learning_rate": 7.407465937990079e-06,
+      "loss": 0.4583,
+      "step": 8632
+    },
+    {
+      "epoch": 1.4177734896228933,
+      "grad_norm": 0.28177094476792225,
+      "learning_rate": 7.407107979278093e-06,
+      "loss": 0.4893,
+      "step": 8633
+    },
+    {
+      "epoch": 1.417937716831236,
+      "grad_norm": 0.3339766813808621,
+      "learning_rate": 7.406749988992052e-06,
+      "loss": 0.4782,
+      "step": 8634
+    },
+    {
+      "epoch": 1.4181019440395788,
+      "grad_norm": 0.3550560311986116,
+      "learning_rate": 7.4063919671358456e-06,
+      "loss": 0.4521,
+      "step": 8635
+    },
+    {
+      "epoch": 1.4182661712479214,
+      "grad_norm": 0.31422102493092624,
+      "learning_rate": 7.406033913713365e-06,
+      "loss": 0.4574,
+      "step": 8636
+    },
+    {
+      "epoch": 1.4184303984562643,
+      "grad_norm": 0.36862943811813514,
+      "learning_rate": 7.405675828728497e-06,
+      "loss": 0.4612,
+      "step": 8637
+    },
+    {
+      "epoch": 1.418594625664607,
+      "grad_norm": 0.3823455427636954,
+      "learning_rate": 7.405317712185129e-06,
+      "loss": 0.4642,
+      "step": 8638
+    },
+    {
+      "epoch": 1.4187588528729498,
+      "grad_norm": 0.6697363617447964,
+      "learning_rate": 7.4049595640871534e-06,
+      "loss": 0.4611,
+      "step": 8639
+    },
+    {
+      "epoch": 1.4189230800812924,
+      "grad_norm": 0.3082877536536946,
+      "learning_rate": 7.404601384438458e-06,
+      "loss": 0.4553,
+      "step": 8640
+    },
+    {
+      "epoch": 1.4190873072896353,
+      "grad_norm": 0.3866697717486414,
+      "learning_rate": 7.404243173242936e-06,
+      "loss": 0.4717,
+      "step": 8641
+    },
+    {
+      "epoch": 1.419251534497978,
+      "grad_norm": 0.3095639973516765,
+      "learning_rate": 7.403884930504474e-06,
+      "loss": 0.4899,
+      "step": 8642
+    },
+    {
+      "epoch": 1.4194157617063208,
+      "grad_norm": 0.3281301085749949,
+      "learning_rate": 7.403526656226965e-06,
+      "loss": 0.4433,
+      "step": 8643
+    },
+    {
+      "epoch": 1.4195799889146634,
+      "grad_norm": 0.3093984334312015,
+      "learning_rate": 7.4031683504142985e-06,
+      "loss": 0.4617,
+      "step": 8644
+    },
+    {
+      "epoch": 1.4197442161230063,
+      "grad_norm": 0.3132767270598841,
+      "learning_rate": 7.402810013070369e-06,
+      "loss": 0.4648,
+      "step": 8645
+    },
+    {
+      "epoch": 1.419908443331349,
+      "grad_norm": 0.35555840733129107,
+      "learning_rate": 7.4024516441990665e-06,
+      "loss": 0.4825,
+      "step": 8646
+    },
+    {
+      "epoch": 1.4200726705396916,
+      "grad_norm": 0.3605843369817561,
+      "learning_rate": 7.402093243804283e-06,
+      "loss": 0.5036,
+      "step": 8647
+    },
+    {
+      "epoch": 1.4202368977480344,
+      "grad_norm": 0.31417513230023925,
+      "learning_rate": 7.401734811889911e-06,
+      "loss": 0.4756,
+      "step": 8648
+    },
+    {
+      "epoch": 1.4204011249563773,
+      "grad_norm": 0.3143056077749063,
+      "learning_rate": 7.401376348459846e-06,
+      "loss": 0.463,
+      "step": 8649
+    },
+    {
+      "epoch": 1.42056535216472,
+      "grad_norm": 0.33212448977759,
+      "learning_rate": 7.401017853517978e-06,
+      "loss": 0.4869,
+      "step": 8650
+    },
+    {
+      "epoch": 1.4207295793730625,
+      "grad_norm": 0.4137208235695577,
+      "learning_rate": 7.400659327068202e-06,
+      "loss": 0.4866,
+      "step": 8651
+    },
+    {
+      "epoch": 1.4208938065814054,
+      "grad_norm": 0.3671152000569904,
+      "learning_rate": 7.400300769114411e-06,
+      "loss": 0.4672,
+      "step": 8652
+    },
+    {
+      "epoch": 1.421058033789748,
+      "grad_norm": 0.36134698599484183,
+      "learning_rate": 7.399942179660502e-06,
+      "loss": 0.4803,
+      "step": 8653
+    },
+    {
+      "epoch": 1.421222260998091,
+      "grad_norm": 0.3226969808681514,
+      "learning_rate": 7.399583558710367e-06,
+      "loss": 0.4564,
+      "step": 8654
+    },
+    {
+      "epoch": 1.4213864882064335,
+      "grad_norm": 0.3360087789001294,
+      "learning_rate": 7.399224906267901e-06,
+      "loss": 0.4695,
+      "step": 8655
+    },
+    {
+      "epoch": 1.4215507154147764,
+      "grad_norm": 0.3999377209987789,
+      "learning_rate": 7.398866222337e-06,
+      "loss": 0.4524,
+      "step": 8656
+    },
+    {
+      "epoch": 1.421714942623119,
+      "grad_norm": 0.2619494941802104,
+      "learning_rate": 7.39850750692156e-06,
+      "loss": 0.4738,
+      "step": 8657
+    },
+    {
+      "epoch": 1.421879169831462,
+      "grad_norm": 0.3377812599960666,
+      "learning_rate": 7.398148760025479e-06,
+      "loss": 0.4783,
+      "step": 8658
+    },
+    {
+      "epoch": 1.4220433970398045,
+      "grad_norm": 0.5649331309834287,
+      "learning_rate": 7.397789981652648e-06,
+      "loss": 0.4686,
+      "step": 8659
+    },
+    {
+      "epoch": 1.4222076242481472,
+      "grad_norm": 0.266427925002802,
+      "learning_rate": 7.3974311718069685e-06,
+      "loss": 0.4613,
+      "step": 8660
+    },
+    {
+      "epoch": 1.42237185145649,
+      "grad_norm": 0.3444625103911929,
+      "learning_rate": 7.397072330492334e-06,
+      "loss": 0.4537,
+      "step": 8661
+    },
+    {
+      "epoch": 1.422536078664833,
+      "grad_norm": 0.34059735816769793,
+      "learning_rate": 7.396713457712646e-06,
+      "loss": 0.4899,
+      "step": 8662
+    },
+    {
+      "epoch": 1.4227003058731755,
+      "grad_norm": 0.3790439525703914,
+      "learning_rate": 7.396354553471799e-06,
+      "loss": 0.4625,
+      "step": 8663
+    },
+    {
+      "epoch": 1.4228645330815182,
+      "grad_norm": 0.39460946921143725,
+      "learning_rate": 7.3959956177736906e-06,
+      "loss": 0.4553,
+      "step": 8664
+    },
+    {
+      "epoch": 1.423028760289861,
+      "grad_norm": 0.2817981735232756,
+      "learning_rate": 7.3956366506222225e-06,
+      "loss": 0.4691,
+      "step": 8665
+    },
+    {
+      "epoch": 1.4231929874982039,
+      "grad_norm": 0.4378902762621768,
+      "learning_rate": 7.39527765202129e-06,
+      "loss": 0.4843,
+      "step": 8666
+    },
+    {
+      "epoch": 1.4233572147065465,
+      "grad_norm": 0.975584940866872,
+      "learning_rate": 7.394918621974795e-06,
+      "loss": 0.4811,
+      "step": 8667
+    },
+    {
+      "epoch": 1.4235214419148892,
+      "grad_norm": 0.7960565831153253,
+      "learning_rate": 7.394559560486634e-06,
+      "loss": 0.4538,
+      "step": 8668
+    },
+    {
+      "epoch": 1.423685669123232,
+      "grad_norm": 0.30691411758843296,
+      "learning_rate": 7.394200467560708e-06,
+      "loss": 0.4838,
+      "step": 8669
+    },
+    {
+      "epoch": 1.4238498963315747,
+      "grad_norm": 0.2963936182755594,
+      "learning_rate": 7.39384134320092e-06,
+      "loss": 0.4711,
+      "step": 8670
+    },
+    {
+      "epoch": 1.4240141235399175,
+      "grad_norm": 0.4976685910588599,
+      "learning_rate": 7.393482187411165e-06,
+      "loss": 0.4713,
+      "step": 8671
+    },
+    {
+      "epoch": 1.4241783507482602,
+      "grad_norm": 0.3144718758400609,
+      "learning_rate": 7.393123000195349e-06,
+      "loss": 0.4841,
+      "step": 8672
+    },
+    {
+      "epoch": 1.424342577956603,
+      "grad_norm": 0.30284813504731734,
+      "learning_rate": 7.392763781557369e-06,
+      "loss": 0.4538,
+      "step": 8673
+    },
+    {
+      "epoch": 1.4245068051649457,
+      "grad_norm": 0.34444613471749935,
+      "learning_rate": 7.3924045315011294e-06,
+      "loss": 0.4684,
+      "step": 8674
+    },
+    {
+      "epoch": 1.4246710323732885,
+      "grad_norm": 0.30198678685548713,
+      "learning_rate": 7.39204525003053e-06,
+      "loss": 0.4796,
+      "step": 8675
+    },
+    {
+      "epoch": 1.4248352595816312,
+      "grad_norm": 0.29695760309471625,
+      "learning_rate": 7.391685937149474e-06,
+      "loss": 0.4972,
+      "step": 8676
+    },
+    {
+      "epoch": 1.4249994867899738,
+      "grad_norm": 0.3142088847097196,
+      "learning_rate": 7.391326592861863e-06,
+      "loss": 0.4796,
+      "step": 8677
+    },
+    {
+      "epoch": 1.4251637139983167,
+      "grad_norm": 0.3184273247226719,
+      "learning_rate": 7.390967217171602e-06,
+      "loss": 0.4991,
+      "step": 8678
+    },
+    {
+      "epoch": 1.4253279412066595,
+      "grad_norm": 0.3155069256631096,
+      "learning_rate": 7.390607810082593e-06,
+      "loss": 0.4629,
+      "step": 8679
+    },
+    {
+      "epoch": 1.4254921684150021,
+      "grad_norm": 0.39120624749395266,
+      "learning_rate": 7.390248371598738e-06,
+      "loss": 0.4489,
+      "step": 8680
+    },
+    {
+      "epoch": 1.4256563956233448,
+      "grad_norm": 0.3245465147182549,
+      "learning_rate": 7.389888901723942e-06,
+      "loss": 0.4579,
+      "step": 8681
+    },
+    {
+      "epoch": 1.4258206228316876,
+      "grad_norm": 0.3394055494225815,
+      "learning_rate": 7.38952940046211e-06,
+      "loss": 0.4757,
+      "step": 8682
+    },
+    {
+      "epoch": 1.4259848500400305,
+      "grad_norm": 0.32787591201901345,
+      "learning_rate": 7.389169867817145e-06,
+      "loss": 0.4814,
+      "step": 8683
+    },
+    {
+      "epoch": 1.4261490772483731,
+      "grad_norm": 0.49850515920788757,
+      "learning_rate": 7.388810303792953e-06,
+      "loss": 0.4779,
+      "step": 8684
+    },
+    {
+      "epoch": 1.4263133044567158,
+      "grad_norm": 0.4000878439452594,
+      "learning_rate": 7.388450708393439e-06,
+      "loss": 0.4735,
+      "step": 8685
+    },
+    {
+      "epoch": 1.4264775316650586,
+      "grad_norm": 0.3884268909927122,
+      "learning_rate": 7.388091081622508e-06,
+      "loss": 0.4615,
+      "step": 8686
+    },
+    {
+      "epoch": 1.4266417588734013,
+      "grad_norm": 0.47546552185240776,
+      "learning_rate": 7.387731423484068e-06,
+      "loss": 0.4673,
+      "step": 8687
+    },
+    {
+      "epoch": 1.4268059860817441,
+      "grad_norm": 0.3451691022903009,
+      "learning_rate": 7.387371733982022e-06,
+      "loss": 0.4527,
+      "step": 8688
+    },
+    {
+      "epoch": 1.4269702132900868,
+      "grad_norm": 0.3351213061571729,
+      "learning_rate": 7.387012013120278e-06,
+      "loss": 0.4505,
+      "step": 8689
+    },
+    {
+      "epoch": 1.4271344404984296,
+      "grad_norm": 0.33697957511100923,
+      "learning_rate": 7.386652260902743e-06,
+      "loss": 0.4526,
+      "step": 8690
+    },
+    {
+      "epoch": 1.4272986677067723,
+      "grad_norm": 0.3288480616443145,
+      "learning_rate": 7.3862924773333266e-06,
+      "loss": 0.4605,
+      "step": 8691
+    },
+    {
+      "epoch": 1.4274628949151151,
+      "grad_norm": 0.2741337484391843,
+      "learning_rate": 7.385932662415932e-06,
+      "loss": 0.4733,
+      "step": 8692
+    },
+    {
+      "epoch": 1.4276271221234578,
+      "grad_norm": 0.35755160262639774,
+      "learning_rate": 7.38557281615447e-06,
+      "loss": 0.4764,
+      "step": 8693
+    },
+    {
+      "epoch": 1.4277913493318004,
+      "grad_norm": 0.42377404035966615,
+      "learning_rate": 7.3852129385528476e-06,
+      "loss": 0.4806,
+      "step": 8694
+    },
+    {
+      "epoch": 1.4279555765401433,
+      "grad_norm": 0.3601017434270535,
+      "learning_rate": 7.3848530296149756e-06,
+      "loss": 0.4878,
+      "step": 8695
+    },
+    {
+      "epoch": 1.4281198037484861,
+      "grad_norm": 0.2642369260132183,
+      "learning_rate": 7.38449308934476e-06,
+      "loss": 0.46,
+      "step": 8696
+    },
+    {
+      "epoch": 1.4282840309568288,
+      "grad_norm": 0.3231395826179842,
+      "learning_rate": 7.3841331177461114e-06,
+      "loss": 0.4513,
+      "step": 8697
+    },
+    {
+      "epoch": 1.4284482581651714,
+      "grad_norm": 0.5167362775726638,
+      "learning_rate": 7.38377311482294e-06,
+      "loss": 0.4731,
+      "step": 8698
+    },
+    {
+      "epoch": 1.4286124853735143,
+      "grad_norm": 0.31768177825252264,
+      "learning_rate": 7.383413080579156e-06,
+      "loss": 0.4808,
+      "step": 8699
+    },
+    {
+      "epoch": 1.4287767125818571,
+      "grad_norm": 0.2670760327678991,
+      "learning_rate": 7.383053015018668e-06,
+      "loss": 0.4463,
+      "step": 8700
+    },
+    {
+      "epoch": 1.4289409397901998,
+      "grad_norm": 0.7385714108339203,
+      "learning_rate": 7.382692918145388e-06,
+      "loss": 0.4714,
+      "step": 8701
+    },
+    {
+      "epoch": 1.4291051669985424,
+      "grad_norm": 0.3665760851986582,
+      "learning_rate": 7.382332789963226e-06,
+      "loss": 0.45,
+      "step": 8702
+    },
+    {
+      "epoch": 1.4292693942068853,
+      "grad_norm": 0.37412406306082946,
+      "learning_rate": 7.381972630476095e-06,
+      "loss": 0.4606,
+      "step": 8703
+    },
+    {
+      "epoch": 1.429433621415228,
+      "grad_norm": 0.282501205455874,
+      "learning_rate": 7.381612439687906e-06,
+      "loss": 0.4713,
+      "step": 8704
+    },
+    {
+      "epoch": 1.4295978486235708,
+      "grad_norm": 0.3175716080706612,
+      "learning_rate": 7.3812522176025705e-06,
+      "loss": 0.4505,
+      "step": 8705
+    },
+    {
+      "epoch": 1.4297620758319134,
+      "grad_norm": 0.30021046264890927,
+      "learning_rate": 7.380891964224001e-06,
+      "loss": 0.468,
+      "step": 8706
+    },
+    {
+      "epoch": 1.4299263030402563,
+      "grad_norm": 0.36843733896681896,
+      "learning_rate": 7.38053167955611e-06,
+      "loss": 0.4797,
+      "step": 8707
+    },
+    {
+      "epoch": 1.430090530248599,
+      "grad_norm": 0.4091002966536038,
+      "learning_rate": 7.380171363602812e-06,
+      "loss": 0.4429,
+      "step": 8708
+    },
+    {
+      "epoch": 1.4302547574569417,
+      "grad_norm": 0.34868960892958695,
+      "learning_rate": 7.379811016368018e-06,
+      "loss": 0.4822,
+      "step": 8709
+    },
+    {
+      "epoch": 1.4304189846652844,
+      "grad_norm": 0.30935125392275414,
+      "learning_rate": 7.379450637855644e-06,
+      "loss": 0.4717,
+      "step": 8710
+    },
+    {
+      "epoch": 1.430583211873627,
+      "grad_norm": 0.39039071388522256,
+      "learning_rate": 7.379090228069602e-06,
+      "loss": 0.4512,
+      "step": 8711
+    },
+    {
+      "epoch": 1.4307474390819699,
+      "grad_norm": 0.34733265965538085,
+      "learning_rate": 7.378729787013809e-06,
+      "loss": 0.4544,
+      "step": 8712
+    },
+    {
+      "epoch": 1.4309116662903127,
+      "grad_norm": 0.30168326488643704,
+      "learning_rate": 7.3783693146921765e-06,
+      "loss": 0.4871,
+      "step": 8713
+    },
+    {
+      "epoch": 1.4310758934986554,
+      "grad_norm": 0.30953253420001814,
+      "learning_rate": 7.378008811108622e-06,
+      "loss": 0.4725,
+      "step": 8714
+    },
+    {
+      "epoch": 1.431240120706998,
+      "grad_norm": 0.38832020065310613,
+      "learning_rate": 7.377648276267061e-06,
+      "loss": 0.4744,
+      "step": 8715
+    },
+    {
+      "epoch": 1.4314043479153409,
+      "grad_norm": 0.3305266235225147,
+      "learning_rate": 7.377287710171408e-06,
+      "loss": 0.4819,
+      "step": 8716
+    },
+    {
+      "epoch": 1.4315685751236837,
+      "grad_norm": 0.34863300149843024,
+      "learning_rate": 7.376927112825579e-06,
+      "loss": 0.4465,
+      "step": 8717
+    },
+    {
+      "epoch": 1.4317328023320264,
+      "grad_norm": 0.3006649152307014,
+      "learning_rate": 7.376566484233492e-06,
+      "loss": 0.4851,
+      "step": 8718
+    },
+    {
+      "epoch": 1.431897029540369,
+      "grad_norm": 0.4097936421003854,
+      "learning_rate": 7.3762058243990615e-06,
+      "loss": 0.4564,
+      "step": 8719
+    },
+    {
+      "epoch": 1.4320612567487119,
+      "grad_norm": 0.29230226483539684,
+      "learning_rate": 7.3758451333262075e-06,
+      "loss": 0.4433,
+      "step": 8720
+    },
+    {
+      "epoch": 1.4322254839570545,
+      "grad_norm": 0.48553587471592025,
+      "learning_rate": 7.375484411018845e-06,
+      "loss": 0.4547,
+      "step": 8721
+    },
+    {
+      "epoch": 1.4323897111653974,
+      "grad_norm": 0.322278646726219,
+      "learning_rate": 7.375123657480893e-06,
+      "loss": 0.4635,
+      "step": 8722
+    },
+    {
+      "epoch": 1.43255393837374,
+      "grad_norm": 0.3545662737503202,
+      "learning_rate": 7.374762872716269e-06,
+      "loss": 0.4687,
+      "step": 8723
+    },
+    {
+      "epoch": 1.4327181655820829,
+      "grad_norm": 0.3099671112478105,
+      "learning_rate": 7.374402056728893e-06,
+      "loss": 0.4638,
+      "step": 8724
+    },
+    {
+      "epoch": 1.4328823927904255,
+      "grad_norm": 0.3716005362691296,
+      "learning_rate": 7.374041209522682e-06,
+      "loss": 0.4784,
+      "step": 8725
+    },
+    {
+      "epoch": 1.4330466199987684,
+      "grad_norm": 0.27171238151285476,
+      "learning_rate": 7.373680331101554e-06,
+      "loss": 0.4728,
+      "step": 8726
+    },
+    {
+      "epoch": 1.433210847207111,
+      "grad_norm": 0.29662911135939474,
+      "learning_rate": 7.373319421469432e-06,
+      "loss": 0.4717,
+      "step": 8727
+    },
+    {
+      "epoch": 1.4333750744154536,
+      "grad_norm": 0.2888872805223644,
+      "learning_rate": 7.3729584806302346e-06,
+      "loss": 0.4556,
+      "step": 8728
+    },
+    {
+      "epoch": 1.4335393016237965,
+      "grad_norm": 0.8980499551263118,
+      "learning_rate": 7.372597508587881e-06,
+      "loss": 0.4434,
+      "step": 8729
+    },
+    {
+      "epoch": 1.4337035288321394,
+      "grad_norm": 0.31920864527645193,
+      "learning_rate": 7.372236505346292e-06,
+      "loss": 0.4707,
+      "step": 8730
+    },
+    {
+      "epoch": 1.433867756040482,
+      "grad_norm": 0.31494192651957037,
+      "learning_rate": 7.371875470909388e-06,
+      "loss": 0.4751,
+      "step": 8731
+    },
+    {
+      "epoch": 1.4340319832488246,
+      "grad_norm": 0.3646059146697081,
+      "learning_rate": 7.371514405281091e-06,
+      "loss": 0.4542,
+      "step": 8732
+    },
+    {
+      "epoch": 1.4341962104571675,
+      "grad_norm": 0.29678542187972806,
+      "learning_rate": 7.371153308465324e-06,
+      "loss": 0.4688,
+      "step": 8733
+    },
+    {
+      "epoch": 1.4343604376655104,
+      "grad_norm": 0.30307859935590387,
+      "learning_rate": 7.370792180466006e-06,
+      "loss": 0.4811,
+      "step": 8734
+    },
+    {
+      "epoch": 1.434524664873853,
+      "grad_norm": 0.35944445084086313,
+      "learning_rate": 7.370431021287059e-06,
+      "loss": 0.4892,
+      "step": 8735
+    },
+    {
+      "epoch": 1.4346888920821956,
+      "grad_norm": 0.3054816830042426,
+      "learning_rate": 7.37006983093241e-06,
+      "loss": 0.4489,
+      "step": 8736
+    },
+    {
+      "epoch": 1.4348531192905385,
+      "grad_norm": 0.3480155627029634,
+      "learning_rate": 7.369708609405977e-06,
+      "loss": 0.4521,
+      "step": 8737
+    },
+    {
+      "epoch": 1.4350173464988811,
+      "grad_norm": 0.33385036419258024,
+      "learning_rate": 7.369347356711686e-06,
+      "loss": 0.4694,
+      "step": 8738
+    },
+    {
+      "epoch": 1.435181573707224,
+      "grad_norm": 0.2680938800862615,
+      "learning_rate": 7.368986072853459e-06,
+      "loss": 0.4641,
+      "step": 8739
+    },
+    {
+      "epoch": 1.4353458009155666,
+      "grad_norm": 0.45577800534453167,
+      "learning_rate": 7.36862475783522e-06,
+      "loss": 0.4672,
+      "step": 8740
+    },
+    {
+      "epoch": 1.4355100281239095,
+      "grad_norm": 0.3124628033703303,
+      "learning_rate": 7.3682634116608955e-06,
+      "loss": 0.4613,
+      "step": 8741
+    },
+    {
+      "epoch": 1.4356742553322521,
+      "grad_norm": 0.29964203647181226,
+      "learning_rate": 7.367902034334407e-06,
+      "loss": 0.4702,
+      "step": 8742
+    },
+    {
+      "epoch": 1.435838482540595,
+      "grad_norm": 0.3592963394992377,
+      "learning_rate": 7.367540625859681e-06,
+      "loss": 0.4741,
+      "step": 8743
+    },
+    {
+      "epoch": 1.4360027097489376,
+      "grad_norm": 0.3646717741074932,
+      "learning_rate": 7.367179186240642e-06,
+      "loss": 0.459,
+      "step": 8744
+    },
+    {
+      "epoch": 1.4361669369572803,
+      "grad_norm": 0.3194017770756867,
+      "learning_rate": 7.366817715481216e-06,
+      "loss": 0.4661,
+      "step": 8745
+    },
+    {
+      "epoch": 1.4363311641656231,
+      "grad_norm": 0.8041761901096787,
+      "learning_rate": 7.36645621358533e-06,
+      "loss": 0.4723,
+      "step": 8746
+    },
+    {
+      "epoch": 1.436495391373966,
+      "grad_norm": 0.3466083547248436,
+      "learning_rate": 7.36609468055691e-06,
+      "loss": 0.4528,
+      "step": 8747
+    },
+    {
+      "epoch": 1.4366596185823086,
+      "grad_norm": 0.43236098119183647,
+      "learning_rate": 7.36573311639988e-06,
+      "loss": 0.4912,
+      "step": 8748
+    },
+    {
+      "epoch": 1.4368238457906513,
+      "grad_norm": 0.48201449975080524,
+      "learning_rate": 7.365371521118171e-06,
+      "loss": 0.4662,
+      "step": 8749
+    },
+    {
+      "epoch": 1.4369880729989941,
+      "grad_norm": 0.2897789566298985,
+      "learning_rate": 7.365009894715706e-06,
+      "loss": 0.4697,
+      "step": 8750
+    },
+    {
+      "epoch": 1.437152300207337,
+      "grad_norm": 0.4285770698212275,
+      "learning_rate": 7.364648237196416e-06,
+      "loss": 0.4645,
+      "step": 8751
+    },
+    {
+      "epoch": 1.4373165274156796,
+      "grad_norm": 0.30895575283505117,
+      "learning_rate": 7.364286548564226e-06,
+      "loss": 0.4635,
+      "step": 8752
+    },
+    {
+      "epoch": 1.4374807546240222,
+      "grad_norm": 0.3878587785336352,
+      "learning_rate": 7.3639248288230685e-06,
+      "loss": 0.4717,
+      "step": 8753
+    },
+    {
+      "epoch": 1.437644981832365,
+      "grad_norm": 0.31587243157295675,
+      "learning_rate": 7.3635630779768694e-06,
+      "loss": 0.4807,
+      "step": 8754
+    },
+    {
+      "epoch": 1.4378092090407077,
+      "grad_norm": 0.29345313599614753,
+      "learning_rate": 7.363201296029556e-06,
+      "loss": 0.4743,
+      "step": 8755
+    },
+    {
+      "epoch": 1.4379734362490506,
+      "grad_norm": 0.29824132861417796,
+      "learning_rate": 7.362839482985061e-06,
+      "loss": 0.4708,
+      "step": 8756
+    },
+    {
+      "epoch": 1.4381376634573932,
+      "grad_norm": 0.3150943702529251,
+      "learning_rate": 7.362477638847312e-06,
+      "loss": 0.4803,
+      "step": 8757
+    },
+    {
+      "epoch": 1.438301890665736,
+      "grad_norm": 0.3035150249486316,
+      "learning_rate": 7.362115763620241e-06,
+      "loss": 0.4642,
+      "step": 8758
+    },
+    {
+      "epoch": 1.4384661178740787,
+      "grad_norm": 0.49228834262052035,
+      "learning_rate": 7.361753857307775e-06,
+      "loss": 0.4644,
+      "step": 8759
+    },
+    {
+      "epoch": 1.4386303450824216,
+      "grad_norm": 0.425149859034616,
+      "learning_rate": 7.3613919199138464e-06,
+      "loss": 0.4573,
+      "step": 8760
+    },
+    {
+      "epoch": 1.4387945722907642,
+      "grad_norm": 0.38725281483202917,
+      "learning_rate": 7.361029951442388e-06,
+      "loss": 0.4717,
+      "step": 8761
+    },
+    {
+      "epoch": 1.4389587994991069,
+      "grad_norm": 0.304654791341761,
+      "learning_rate": 7.360667951897329e-06,
+      "loss": 0.475,
+      "step": 8762
+    },
+    {
+      "epoch": 1.4391230267074497,
+      "grad_norm": 0.29352327279914797,
+      "learning_rate": 7.3603059212826e-06,
+      "loss": 0.4814,
+      "step": 8763
+    },
+    {
+      "epoch": 1.4392872539157926,
+      "grad_norm": 0.39246611782479257,
+      "learning_rate": 7.359943859602135e-06,
+      "loss": 0.4847,
+      "step": 8764
+    },
+    {
+      "epoch": 1.4394514811241352,
+      "grad_norm": 0.3311896456985306,
+      "learning_rate": 7.359581766859867e-06,
+      "loss": 0.4603,
+      "step": 8765
+    },
+    {
+      "epoch": 1.4396157083324779,
+      "grad_norm": 0.359044957795356,
+      "learning_rate": 7.359219643059727e-06,
+      "loss": 0.4447,
+      "step": 8766
+    },
+    {
+      "epoch": 1.4397799355408207,
+      "grad_norm": 0.26938397873450154,
+      "learning_rate": 7.3588574882056485e-06,
+      "loss": 0.4674,
+      "step": 8767
+    },
+    {
+      "epoch": 1.4399441627491636,
+      "grad_norm": 0.4728328295191253,
+      "learning_rate": 7.358495302301566e-06,
+      "loss": 0.4584,
+      "step": 8768
+    },
+    {
+      "epoch": 1.4401083899575062,
+      "grad_norm": 0.28764939085722807,
+      "learning_rate": 7.35813308535141e-06,
+      "loss": 0.5044,
+      "step": 8769
+    },
+    {
+      "epoch": 1.4402726171658489,
+      "grad_norm": 0.64332414790041,
+      "learning_rate": 7.357770837359119e-06,
+      "loss": 0.4699,
+      "step": 8770
+    },
+    {
+      "epoch": 1.4404368443741917,
+      "grad_norm": 0.26700818401796017,
+      "learning_rate": 7.357408558328623e-06,
+      "loss": 0.4667,
+      "step": 8771
+    },
+    {
+      "epoch": 1.4406010715825344,
+      "grad_norm": 0.3857951942674286,
+      "learning_rate": 7.35704624826386e-06,
+      "loss": 0.4705,
+      "step": 8772
+    },
+    {
+      "epoch": 1.4407652987908772,
+      "grad_norm": 0.2941631079784591,
+      "learning_rate": 7.356683907168762e-06,
+      "loss": 0.4678,
+      "step": 8773
+    },
+    {
+      "epoch": 1.4409295259992199,
+      "grad_norm": 0.35313285672785943,
+      "learning_rate": 7.356321535047269e-06,
+      "loss": 0.4609,
+      "step": 8774
+    },
+    {
+      "epoch": 1.4410937532075627,
+      "grad_norm": 0.30558355477754984,
+      "learning_rate": 7.355959131903313e-06,
+      "loss": 0.491,
+      "step": 8775
+    },
+    {
+      "epoch": 1.4412579804159054,
+      "grad_norm": 0.30028824124065273,
+      "learning_rate": 7.355596697740828e-06,
+      "loss": 0.4782,
+      "step": 8776
+    },
+    {
+      "epoch": 1.4414222076242482,
+      "grad_norm": 0.3974043321955806,
+      "learning_rate": 7.355234232563758e-06,
+      "loss": 0.466,
+      "step": 8777
+    },
+    {
+      "epoch": 1.4415864348325909,
+      "grad_norm": 0.32176678659026303,
+      "learning_rate": 7.354871736376031e-06,
+      "loss": 0.4519,
+      "step": 8778
+    },
+    {
+      "epoch": 1.4417506620409335,
+      "grad_norm": 0.32792249317240885,
+      "learning_rate": 7.354509209181591e-06,
+      "loss": 0.4738,
+      "step": 8779
+    },
+    {
+      "epoch": 1.4419148892492764,
+      "grad_norm": 0.30658790737339253,
+      "learning_rate": 7.354146650984372e-06,
+      "loss": 0.462,
+      "step": 8780
+    },
+    {
+      "epoch": 1.4420791164576192,
+      "grad_norm": 0.4592510866299841,
+      "learning_rate": 7.35378406178831e-06,
+      "loss": 0.4529,
+      "step": 8781
+    },
+    {
+      "epoch": 1.4422433436659619,
+      "grad_norm": 0.2998740701505465,
+      "learning_rate": 7.353421441597348e-06,
+      "loss": 0.4682,
+      "step": 8782
+    },
+    {
+      "epoch": 1.4424075708743045,
+      "grad_norm": 0.355911208435906,
+      "learning_rate": 7.353058790415422e-06,
+      "loss": 0.4655,
+      "step": 8783
+    },
+    {
+      "epoch": 1.4425717980826473,
+      "grad_norm": 0.32644863293812215,
+      "learning_rate": 7.352696108246469e-06,
+      "loss": 0.4618,
+      "step": 8784
+    },
+    {
+      "epoch": 1.4427360252909902,
+      "grad_norm": 0.2730444038859821,
+      "learning_rate": 7.352333395094429e-06,
+      "loss": 0.4931,
+      "step": 8785
+    },
+    {
+      "epoch": 1.4429002524993328,
+      "grad_norm": 0.32896780601935,
+      "learning_rate": 7.351970650963243e-06,
+      "loss": 0.4527,
+      "step": 8786
+    },
+    {
+      "epoch": 1.4430644797076755,
+      "grad_norm": 0.3141642761360534,
+      "learning_rate": 7.35160787585685e-06,
+      "loss": 0.4731,
+      "step": 8787
+    },
+    {
+      "epoch": 1.4432287069160183,
+      "grad_norm": 0.3259242243345612,
+      "learning_rate": 7.35124506977919e-06,
+      "loss": 0.472,
+      "step": 8788
+    },
+    {
+      "epoch": 1.443392934124361,
+      "grad_norm": 0.28421921594410116,
+      "learning_rate": 7.350882232734202e-06,
+      "loss": 0.4423,
+      "step": 8789
+    },
+    {
+      "epoch": 1.4435571613327038,
+      "grad_norm": 0.3161737021968298,
+      "learning_rate": 7.350519364725829e-06,
+      "loss": 0.4716,
+      "step": 8790
+    },
+    {
+      "epoch": 1.4437213885410465,
+      "grad_norm": 0.2932958787210009,
+      "learning_rate": 7.350156465758012e-06,
+      "loss": 0.4774,
+      "step": 8791
+    },
+    {
+      "epoch": 1.4438856157493893,
+      "grad_norm": 0.6697209643610852,
+      "learning_rate": 7.3497935358346894e-06,
+      "loss": 0.4357,
+      "step": 8792
+    },
+    {
+      "epoch": 1.444049842957732,
+      "grad_norm": 0.3264213977418509,
+      "learning_rate": 7.349430574959807e-06,
+      "loss": 0.4725,
+      "step": 8793
+    },
+    {
+      "epoch": 1.4442140701660748,
+      "grad_norm": 0.32557575314880266,
+      "learning_rate": 7.349067583137305e-06,
+      "loss": 0.4615,
+      "step": 8794
+    },
+    {
+      "epoch": 1.4443782973744175,
+      "grad_norm": 0.31900811247252975,
+      "learning_rate": 7.348704560371126e-06,
+      "loss": 0.4705,
+      "step": 8795
+    },
+    {
+      "epoch": 1.4445425245827601,
+      "grad_norm": 0.35273304255974214,
+      "learning_rate": 7.348341506665211e-06,
+      "loss": 0.4836,
+      "step": 8796
+    },
+    {
+      "epoch": 1.444706751791103,
+      "grad_norm": 0.34808125964253456,
+      "learning_rate": 7.347978422023507e-06,
+      "loss": 0.4652,
+      "step": 8797
+    },
+    {
+      "epoch": 1.4448709789994458,
+      "grad_norm": 0.4687086087072129,
+      "learning_rate": 7.347615306449954e-06,
+      "loss": 0.473,
+      "step": 8798
+    },
+    {
+      "epoch": 1.4450352062077885,
+      "grad_norm": 0.3125715957895599,
+      "learning_rate": 7.347252159948498e-06,
+      "loss": 0.4606,
+      "step": 8799
+    },
+    {
+      "epoch": 1.445199433416131,
+      "grad_norm": 0.2823950209487057,
+      "learning_rate": 7.346888982523081e-06,
+      "loss": 0.4761,
+      "step": 8800
+    },
+    {
+      "epoch": 1.445363660624474,
+      "grad_norm": 0.27396154510210274,
+      "learning_rate": 7.3465257741776495e-06,
+      "loss": 0.4841,
+      "step": 8801
+    },
+    {
+      "epoch": 1.4455278878328168,
+      "grad_norm": 0.38488801653092736,
+      "learning_rate": 7.346162534916148e-06,
+      "loss": 0.4717,
+      "step": 8802
+    },
+    {
+      "epoch": 1.4456921150411595,
+      "grad_norm": 0.3296881924583521,
+      "learning_rate": 7.34579926474252e-06,
+      "loss": 0.4636,
+      "step": 8803
+    },
+    {
+      "epoch": 1.445856342249502,
+      "grad_norm": 0.4285832646198502,
+      "learning_rate": 7.345435963660713e-06,
+      "loss": 0.4575,
+      "step": 8804
+    },
+    {
+      "epoch": 1.446020569457845,
+      "grad_norm": 0.34978061084114653,
+      "learning_rate": 7.345072631674672e-06,
+      "loss": 0.4777,
+      "step": 8805
+    },
+    {
+      "epoch": 1.4461847966661876,
+      "grad_norm": 0.308759693755405,
+      "learning_rate": 7.344709268788342e-06,
+      "loss": 0.4674,
+      "step": 8806
+    },
+    {
+      "epoch": 1.4463490238745305,
+      "grad_norm": 0.37776335856045,
+      "learning_rate": 7.344345875005671e-06,
+      "loss": 0.4734,
+      "step": 8807
+    },
+    {
+      "epoch": 1.446513251082873,
+      "grad_norm": 0.30213707597068723,
+      "learning_rate": 7.343982450330605e-06,
+      "loss": 0.4677,
+      "step": 8808
+    },
+    {
+      "epoch": 1.446677478291216,
+      "grad_norm": 0.2863664625153726,
+      "learning_rate": 7.343618994767093e-06,
+      "loss": 0.4757,
+      "step": 8809
+    },
+    {
+      "epoch": 1.4468417054995586,
+      "grad_norm": 0.4567100127507095,
+      "learning_rate": 7.343255508319079e-06,
+      "loss": 0.4826,
+      "step": 8810
+    },
+    {
+      "epoch": 1.4470059327079015,
+      "grad_norm": 0.3235014519664871,
+      "learning_rate": 7.342891990990514e-06,
+      "loss": 0.4877,
+      "step": 8811
+    },
+    {
+      "epoch": 1.447170159916244,
+      "grad_norm": 0.48932063009730403,
+      "learning_rate": 7.342528442785346e-06,
+      "loss": 0.4597,
+      "step": 8812
+    },
+    {
+      "epoch": 1.4473343871245867,
+      "grad_norm": 0.29500404335881325,
+      "learning_rate": 7.34216486370752e-06,
+      "loss": 0.4617,
+      "step": 8813
+    },
+    {
+      "epoch": 1.4474986143329296,
+      "grad_norm": 0.32921988580938477,
+      "learning_rate": 7.341801253760988e-06,
+      "loss": 0.4409,
+      "step": 8814
+    },
+    {
+      "epoch": 1.4476628415412724,
+      "grad_norm": 0.32024711284584884,
+      "learning_rate": 7.341437612949699e-06,
+      "loss": 0.459,
+      "step": 8815
+    },
+    {
+      "epoch": 1.447827068749615,
+      "grad_norm": 0.2953490698527201,
+      "learning_rate": 7.341073941277602e-06,
+      "loss": 0.4596,
+      "step": 8816
+    },
+    {
+      "epoch": 1.4479912959579577,
+      "grad_norm": 0.3474854112500576,
+      "learning_rate": 7.340710238748646e-06,
+      "loss": 0.4713,
+      "step": 8817
+    },
+    {
+      "epoch": 1.4481555231663006,
+      "grad_norm": 0.6411897652901434,
+      "learning_rate": 7.340346505366782e-06,
+      "loss": 0.4792,
+      "step": 8818
+    },
+    {
+      "epoch": 1.4483197503746434,
+      "grad_norm": 0.3431135250449065,
+      "learning_rate": 7.3399827411359615e-06,
+      "loss": 0.4734,
+      "step": 8819
+    },
+    {
+      "epoch": 1.448483977582986,
+      "grad_norm": 0.3095387941404658,
+      "learning_rate": 7.3396189460601325e-06,
+      "loss": 0.4672,
+      "step": 8820
+    },
+    {
+      "epoch": 1.4486482047913287,
+      "grad_norm": 0.29983920611558473,
+      "learning_rate": 7.339255120143251e-06,
+      "loss": 0.4751,
+      "step": 8821
+    },
+    {
+      "epoch": 1.4488124319996716,
+      "grad_norm": 0.29211550519830737,
+      "learning_rate": 7.338891263389263e-06,
+      "loss": 0.4871,
+      "step": 8822
+    },
+    {
+      "epoch": 1.4489766592080142,
+      "grad_norm": 0.3023119216930271,
+      "learning_rate": 7.338527375802123e-06,
+      "loss": 0.4751,
+      "step": 8823
+    },
+    {
+      "epoch": 1.449140886416357,
+      "grad_norm": 0.2741409984560532,
+      "learning_rate": 7.338163457385783e-06,
+      "loss": 0.4576,
+      "step": 8824
+    },
+    {
+      "epoch": 1.4493051136246997,
+      "grad_norm": 0.7815452059668843,
+      "learning_rate": 7.337799508144196e-06,
+      "loss": 0.477,
+      "step": 8825
+    },
+    {
+      "epoch": 1.4494693408330426,
+      "grad_norm": 0.28544871860888393,
+      "learning_rate": 7.337435528081315e-06,
+      "loss": 0.4815,
+      "step": 8826
+    },
+    {
+      "epoch": 1.4496335680413852,
+      "grad_norm": 0.3701675073844943,
+      "learning_rate": 7.33707151720109e-06,
+      "loss": 0.478,
+      "step": 8827
+    },
+    {
+      "epoch": 1.449797795249728,
+      "grad_norm": 0.32291784357585807,
+      "learning_rate": 7.336707475507479e-06,
+      "loss": 0.4671,
+      "step": 8828
+    },
+    {
+      "epoch": 1.4499620224580707,
+      "grad_norm": 0.2915111220168,
+      "learning_rate": 7.336343403004434e-06,
+      "loss": 0.4804,
+      "step": 8829
+    },
+    {
+      "epoch": 1.4501262496664133,
+      "grad_norm": 0.3140881121761215,
+      "learning_rate": 7.335979299695907e-06,
+      "loss": 0.4604,
+      "step": 8830
+    },
+    {
+      "epoch": 1.4502904768747562,
+      "grad_norm": 0.2802099567639804,
+      "learning_rate": 7.335615165585857e-06,
+      "loss": 0.4703,
+      "step": 8831
+    },
+    {
+      "epoch": 1.450454704083099,
+      "grad_norm": 0.3050059252111744,
+      "learning_rate": 7.335251000678235e-06,
+      "loss": 0.4572,
+      "step": 8832
+    },
+    {
+      "epoch": 1.4506189312914417,
+      "grad_norm": 0.3919127908954277,
+      "learning_rate": 7.334886804976999e-06,
+      "loss": 0.4758,
+      "step": 8833
+    },
+    {
+      "epoch": 1.4507831584997843,
+      "grad_norm": 0.31520124295877416,
+      "learning_rate": 7.334522578486102e-06,
+      "loss": 0.4934,
+      "step": 8834
+    },
+    {
+      "epoch": 1.4509473857081272,
+      "grad_norm": 0.302542810820108,
+      "learning_rate": 7.334158321209502e-06,
+      "loss": 0.4637,
+      "step": 8835
+    },
+    {
+      "epoch": 1.45111161291647,
+      "grad_norm": 0.2983970505298541,
+      "learning_rate": 7.333794033151153e-06,
+      "loss": 0.456,
+      "step": 8836
+    },
+    {
+      "epoch": 1.4512758401248127,
+      "grad_norm": 0.2992793685686754,
+      "learning_rate": 7.333429714315014e-06,
+      "loss": 0.4765,
+      "step": 8837
+    },
+    {
+      "epoch": 1.4514400673331553,
+      "grad_norm": 0.30608183883212225,
+      "learning_rate": 7.333065364705039e-06,
+      "loss": 0.4544,
+      "step": 8838
+    },
+    {
+      "epoch": 1.4516042945414982,
+      "grad_norm": 0.33612070244669445,
+      "learning_rate": 7.332700984325188e-06,
+      "loss": 0.4473,
+      "step": 8839
+    },
+    {
+      "epoch": 1.4517685217498408,
+      "grad_norm": 0.36256371284524325,
+      "learning_rate": 7.332336573179417e-06,
+      "loss": 0.4773,
+      "step": 8840
+    },
+    {
+      "epoch": 1.4519327489581837,
+      "grad_norm": 0.5163192135177029,
+      "learning_rate": 7.331972131271683e-06,
+      "loss": 0.4803,
+      "step": 8841
+    },
+    {
+      "epoch": 1.4520969761665263,
+      "grad_norm": 0.291303436118226,
+      "learning_rate": 7.331607658605947e-06,
+      "loss": 0.4666,
+      "step": 8842
+    },
+    {
+      "epoch": 1.4522612033748692,
+      "grad_norm": 0.30054038001077393,
+      "learning_rate": 7.331243155186165e-06,
+      "loss": 0.4596,
+      "step": 8843
+    },
+    {
+      "epoch": 1.4524254305832118,
+      "grad_norm": 0.32552628814577383,
+      "learning_rate": 7.330878621016298e-06,
+      "loss": 0.4655,
+      "step": 8844
+    },
+    {
+      "epoch": 1.4525896577915547,
+      "grad_norm": 0.3255490210170379,
+      "learning_rate": 7.330514056100302e-06,
+      "loss": 0.4728,
+      "step": 8845
+    },
+    {
+      "epoch": 1.4527538849998973,
+      "grad_norm": 0.3732452570361322,
+      "learning_rate": 7.330149460442139e-06,
+      "loss": 0.4678,
+      "step": 8846
+    },
+    {
+      "epoch": 1.45291811220824,
+      "grad_norm": 0.3403884811461902,
+      "learning_rate": 7.329784834045769e-06,
+      "loss": 0.473,
+      "step": 8847
+    },
+    {
+      "epoch": 1.4530823394165828,
+      "grad_norm": 0.35037990887786574,
+      "learning_rate": 7.329420176915151e-06,
+      "loss": 0.4578,
+      "step": 8848
+    },
+    {
+      "epoch": 1.4532465666249257,
+      "grad_norm": 0.33906292449742215,
+      "learning_rate": 7.329055489054248e-06,
+      "loss": 0.473,
+      "step": 8849
+    },
+    {
+      "epoch": 1.4534107938332683,
+      "grad_norm": 0.3476995567742656,
+      "learning_rate": 7.3286907704670175e-06,
+      "loss": 0.489,
+      "step": 8850
+    },
+    {
+      "epoch": 1.453575021041611,
+      "grad_norm": 0.3307868667516689,
+      "learning_rate": 7.328326021157423e-06,
+      "loss": 0.4775,
+      "step": 8851
+    },
+    {
+      "epoch": 1.4537392482499538,
+      "grad_norm": 0.47335856697556894,
+      "learning_rate": 7.327961241129423e-06,
+      "loss": 0.471,
+      "step": 8852
+    },
+    {
+      "epoch": 1.4539034754582967,
+      "grad_norm": 0.27687802887216756,
+      "learning_rate": 7.327596430386984e-06,
+      "loss": 0.4493,
+      "step": 8853
+    },
+    {
+      "epoch": 1.4540677026666393,
+      "grad_norm": 0.3294214659206888,
+      "learning_rate": 7.327231588934065e-06,
+      "loss": 0.4644,
+      "step": 8854
+    },
+    {
+      "epoch": 1.454231929874982,
+      "grad_norm": 0.321441907890736,
+      "learning_rate": 7.326866716774629e-06,
+      "loss": 0.4733,
+      "step": 8855
+    },
+    {
+      "epoch": 1.4543961570833248,
+      "grad_norm": 0.3409296447226374,
+      "learning_rate": 7.326501813912639e-06,
+      "loss": 0.4459,
+      "step": 8856
+    },
+    {
+      "epoch": 1.4545603842916675,
+      "grad_norm": 0.40546821658975873,
+      "learning_rate": 7.326136880352058e-06,
+      "loss": 0.4736,
+      "step": 8857
+    },
+    {
+      "epoch": 1.4547246115000103,
+      "grad_norm": 0.3157113879724182,
+      "learning_rate": 7.325771916096853e-06,
+      "loss": 0.4581,
+      "step": 8858
+    },
+    {
+      "epoch": 1.454888838708353,
+      "grad_norm": 0.43633886623309787,
+      "learning_rate": 7.325406921150981e-06,
+      "loss": 0.4638,
+      "step": 8859
+    },
+    {
+      "epoch": 1.4550530659166958,
+      "grad_norm": 0.6933423819272897,
+      "learning_rate": 7.3250418955184115e-06,
+      "loss": 0.4936,
+      "step": 8860
+    },
+    {
+      "epoch": 1.4552172931250384,
+      "grad_norm": 0.30856150710693425,
+      "learning_rate": 7.324676839203108e-06,
+      "loss": 0.4739,
+      "step": 8861
+    },
+    {
+      "epoch": 1.4553815203333813,
+      "grad_norm": 0.37647065666549806,
+      "learning_rate": 7.324311752209035e-06,
+      "loss": 0.4774,
+      "step": 8862
+    },
+    {
+      "epoch": 1.455545747541724,
+      "grad_norm": 0.31474858543251255,
+      "learning_rate": 7.323946634540156e-06,
+      "loss": 0.4775,
+      "step": 8863
+    },
+    {
+      "epoch": 1.4557099747500666,
+      "grad_norm": 0.35550058465976125,
+      "learning_rate": 7.32358148620044e-06,
+      "loss": 0.475,
+      "step": 8864
+    },
+    {
+      "epoch": 1.4558742019584094,
+      "grad_norm": 0.3180664131895398,
+      "learning_rate": 7.32321630719385e-06,
+      "loss": 0.4534,
+      "step": 8865
+    },
+    {
+      "epoch": 1.4560384291667523,
+      "grad_norm": 0.4056724611982341,
+      "learning_rate": 7.322851097524354e-06,
+      "loss": 0.4915,
+      "step": 8866
+    },
+    {
+      "epoch": 1.456202656375095,
+      "grad_norm": 0.3267146528014822,
+      "learning_rate": 7.322485857195916e-06,
+      "loss": 0.4957,
+      "step": 8867
+    },
+    {
+      "epoch": 1.4563668835834376,
+      "grad_norm": 0.429087209244443,
+      "learning_rate": 7.322120586212507e-06,
+      "loss": 0.4619,
+      "step": 8868
+    },
+    {
+      "epoch": 1.4565311107917804,
+      "grad_norm": 0.2921345299614119,
+      "learning_rate": 7.32175528457809e-06,
+      "loss": 0.4468,
+      "step": 8869
+    },
+    {
+      "epoch": 1.4566953380001233,
+      "grad_norm": 0.326412519394861,
+      "learning_rate": 7.321389952296635e-06,
+      "loss": 0.4696,
+      "step": 8870
+    },
+    {
+      "epoch": 1.456859565208466,
+      "grad_norm": 0.30064282080456756,
+      "learning_rate": 7.32102458937211e-06,
+      "loss": 0.4659,
+      "step": 8871
+    },
+    {
+      "epoch": 1.4570237924168086,
+      "grad_norm": 0.3096831053057281,
+      "learning_rate": 7.320659195808482e-06,
+      "loss": 0.4621,
+      "step": 8872
+    },
+    {
+      "epoch": 1.4571880196251514,
+      "grad_norm": 0.29535434369055585,
+      "learning_rate": 7.32029377160972e-06,
+      "loss": 0.4546,
+      "step": 8873
+    },
+    {
+      "epoch": 1.457352246833494,
+      "grad_norm": 0.32787363679249254,
+      "learning_rate": 7.319928316779792e-06,
+      "loss": 0.4522,
+      "step": 8874
+    },
+    {
+      "epoch": 1.457516474041837,
+      "grad_norm": 0.35674569817713586,
+      "learning_rate": 7.3195628313226685e-06,
+      "loss": 0.4668,
+      "step": 8875
+    },
+    {
+      "epoch": 1.4576807012501796,
+      "grad_norm": 0.3534886154900476,
+      "learning_rate": 7.31919731524232e-06,
+      "loss": 0.463,
+      "step": 8876
+    },
+    {
+      "epoch": 1.4578449284585224,
+      "grad_norm": 0.37215114639067504,
+      "learning_rate": 7.318831768542713e-06,
+      "loss": 0.4606,
+      "step": 8877
+    },
+    {
+      "epoch": 1.458009155666865,
+      "grad_norm": 0.412766233405986,
+      "learning_rate": 7.31846619122782e-06,
+      "loss": 0.4659,
+      "step": 8878
+    },
+    {
+      "epoch": 1.458173382875208,
+      "grad_norm": 0.322259459292066,
+      "learning_rate": 7.318100583301612e-06,
+      "loss": 0.4712,
+      "step": 8879
+    },
+    {
+      "epoch": 1.4583376100835506,
+      "grad_norm": 0.38762871022432,
+      "learning_rate": 7.31773494476806e-06,
+      "loss": 0.4848,
+      "step": 8880
+    },
+    {
+      "epoch": 1.4585018372918932,
+      "grad_norm": 0.33110964123463305,
+      "learning_rate": 7.317369275631131e-06,
+      "loss": 0.4534,
+      "step": 8881
+    },
+    {
+      "epoch": 1.458666064500236,
+      "grad_norm": 0.29537116751045295,
+      "learning_rate": 7.317003575894802e-06,
+      "loss": 0.4618,
+      "step": 8882
+    },
+    {
+      "epoch": 1.458830291708579,
+      "grad_norm": 0.3093608228609226,
+      "learning_rate": 7.316637845563043e-06,
+      "loss": 0.4649,
+      "step": 8883
+    },
+    {
+      "epoch": 1.4589945189169216,
+      "grad_norm": 0.3557175296933889,
+      "learning_rate": 7.316272084639824e-06,
+      "loss": 0.4524,
+      "step": 8884
+    },
+    {
+      "epoch": 1.4591587461252642,
+      "grad_norm": 0.3408766965625624,
+      "learning_rate": 7.31590629312912e-06,
+      "loss": 0.4615,
+      "step": 8885
+    },
+    {
+      "epoch": 1.459322973333607,
+      "grad_norm": 0.31371414979257384,
+      "learning_rate": 7.315540471034903e-06,
+      "loss": 0.4605,
+      "step": 8886
+    },
+    {
+      "epoch": 1.45948720054195,
+      "grad_norm": 0.3204753006979815,
+      "learning_rate": 7.3151746183611485e-06,
+      "loss": 0.4685,
+      "step": 8887
+    },
+    {
+      "epoch": 1.4596514277502926,
+      "grad_norm": 0.2763367425927962,
+      "learning_rate": 7.314808735111825e-06,
+      "loss": 0.4723,
+      "step": 8888
+    },
+    {
+      "epoch": 1.4598156549586352,
+      "grad_norm": 0.30501998510386186,
+      "learning_rate": 7.314442821290911e-06,
+      "loss": 0.4755,
+      "step": 8889
+    },
+    {
+      "epoch": 1.459979882166978,
+      "grad_norm": 0.31957072046857943,
+      "learning_rate": 7.314076876902378e-06,
+      "loss": 0.4508,
+      "step": 8890
+    },
+    {
+      "epoch": 1.4601441093753207,
+      "grad_norm": 0.4569959723681631,
+      "learning_rate": 7.313710901950203e-06,
+      "loss": 0.4711,
+      "step": 8891
+    },
+    {
+      "epoch": 1.4603083365836635,
+      "grad_norm": 0.3275277934527127,
+      "learning_rate": 7.3133448964383575e-06,
+      "loss": 0.4531,
+      "step": 8892
+    },
+    {
+      "epoch": 1.4604725637920062,
+      "grad_norm": 0.31450641388610223,
+      "learning_rate": 7.312978860370818e-06,
+      "loss": 0.4699,
+      "step": 8893
+    },
+    {
+      "epoch": 1.460636791000349,
+      "grad_norm": 0.35690484276864787,
+      "learning_rate": 7.312612793751563e-06,
+      "loss": 0.4571,
+      "step": 8894
+    },
+    {
+      "epoch": 1.4608010182086917,
+      "grad_norm": 0.32055550173392267,
+      "learning_rate": 7.312246696584564e-06,
+      "loss": 0.4687,
+      "step": 8895
+    },
+    {
+      "epoch": 1.4609652454170345,
+      "grad_norm": 0.31626984369788685,
+      "learning_rate": 7.311880568873799e-06,
+      "loss": 0.4789,
+      "step": 8896
+    },
+    {
+      "epoch": 1.4611294726253772,
+      "grad_norm": 0.27444977756984074,
+      "learning_rate": 7.311514410623244e-06,
+      "loss": 0.4762,
+      "step": 8897
+    },
+    {
+      "epoch": 1.4612936998337198,
+      "grad_norm": 0.2779313804007929,
+      "learning_rate": 7.311148221836878e-06,
+      "loss": 0.4485,
+      "step": 8898
+    },
+    {
+      "epoch": 1.4614579270420627,
+      "grad_norm": 0.32831057129001473,
+      "learning_rate": 7.310782002518676e-06,
+      "loss": 0.4724,
+      "step": 8899
+    },
+    {
+      "epoch": 1.4616221542504055,
+      "grad_norm": 0.36700689821516946,
+      "learning_rate": 7.310415752672616e-06,
+      "loss": 0.4913,
+      "step": 8900
+    },
+    {
+      "epoch": 1.4617863814587482,
+      "grad_norm": 0.24873549957452207,
+      "learning_rate": 7.310049472302676e-06,
+      "loss": 0.4701,
+      "step": 8901
+    },
+    {
+      "epoch": 1.4619506086670908,
+      "grad_norm": 0.271400064217122,
+      "learning_rate": 7.3096831614128336e-06,
+      "loss": 0.4849,
+      "step": 8902
+    },
+    {
+      "epoch": 1.4621148358754337,
+      "grad_norm": 0.30503298928358774,
+      "learning_rate": 7.309316820007066e-06,
+      "loss": 0.4793,
+      "step": 8903
+    },
+    {
+      "epoch": 1.4622790630837765,
+      "grad_norm": 0.32951277679543467,
+      "learning_rate": 7.3089504480893565e-06,
+      "loss": 0.4634,
+      "step": 8904
+    },
+    {
+      "epoch": 1.4624432902921192,
+      "grad_norm": 0.3415409673036727,
+      "learning_rate": 7.30858404566368e-06,
+      "loss": 0.456,
+      "step": 8905
+    },
+    {
+      "epoch": 1.4626075175004618,
+      "grad_norm": 0.2969992645784687,
+      "learning_rate": 7.308217612734019e-06,
+      "loss": 0.4706,
+      "step": 8906
+    },
+    {
+      "epoch": 1.4627717447088047,
+      "grad_norm": 0.28013964963023535,
+      "learning_rate": 7.30785114930435e-06,
+      "loss": 0.4544,
+      "step": 8907
+    },
+    {
+      "epoch": 1.4629359719171473,
+      "grad_norm": 0.32904919821824175,
+      "learning_rate": 7.307484655378656e-06,
+      "loss": 0.4442,
+      "step": 8908
+    },
+    {
+      "epoch": 1.4631001991254902,
+      "grad_norm": 0.4378861663903591,
+      "learning_rate": 7.307118130960917e-06,
+      "loss": 0.4411,
+      "step": 8909
+    },
+    {
+      "epoch": 1.4632644263338328,
+      "grad_norm": 0.2887836195173391,
+      "learning_rate": 7.306751576055111e-06,
+      "loss": 0.4565,
+      "step": 8910
+    },
+    {
+      "epoch": 1.4634286535421757,
+      "grad_norm": 0.3921748038982994,
+      "learning_rate": 7.306384990665223e-06,
+      "loss": 0.4707,
+      "step": 8911
+    },
+    {
+      "epoch": 1.4635928807505183,
+      "grad_norm": 0.33692717526667476,
+      "learning_rate": 7.306018374795234e-06,
+      "loss": 0.4521,
+      "step": 8912
+    },
+    {
+      "epoch": 1.4637571079588612,
+      "grad_norm": 0.3331839944553744,
+      "learning_rate": 7.305651728449123e-06,
+      "loss": 0.4676,
+      "step": 8913
+    },
+    {
+      "epoch": 1.4639213351672038,
+      "grad_norm": 0.4512463545321785,
+      "learning_rate": 7.305285051630875e-06,
+      "loss": 0.4694,
+      "step": 8914
+    },
+    {
+      "epoch": 1.4640855623755464,
+      "grad_norm": 0.3429881607623171,
+      "learning_rate": 7.3049183443444694e-06,
+      "loss": 0.4444,
+      "step": 8915
+    },
+    {
+      "epoch": 1.4642497895838893,
+      "grad_norm": 0.2916652689698767,
+      "learning_rate": 7.304551606593892e-06,
+      "loss": 0.4534,
+      "step": 8916
+    },
+    {
+      "epoch": 1.4644140167922322,
+      "grad_norm": 0.3177553803478125,
+      "learning_rate": 7.3041848383831254e-06,
+      "loss": 0.4787,
+      "step": 8917
+    },
+    {
+      "epoch": 1.4645782440005748,
+      "grad_norm": 0.2936657930249479,
+      "learning_rate": 7.303818039716152e-06,
+      "loss": 0.4576,
+      "step": 8918
+    },
+    {
+      "epoch": 1.4647424712089174,
+      "grad_norm": 0.3535099457184482,
+      "learning_rate": 7.303451210596957e-06,
+      "loss": 0.4742,
+      "step": 8919
+    },
+    {
+      "epoch": 1.4649066984172603,
+      "grad_norm": 0.36301749310474385,
+      "learning_rate": 7.303084351029522e-06,
+      "loss": 0.4771,
+      "step": 8920
+    },
+    {
+      "epoch": 1.4650709256256031,
+      "grad_norm": 0.27236820017666563,
+      "learning_rate": 7.302717461017834e-06,
+      "loss": 0.476,
+      "step": 8921
+    },
+    {
+      "epoch": 1.4652351528339458,
+      "grad_norm": 0.31864032094301475,
+      "learning_rate": 7.302350540565877e-06,
+      "loss": 0.4626,
+      "step": 8922
+    },
+    {
+      "epoch": 1.4653993800422884,
+      "grad_norm": 0.44159276047496465,
+      "learning_rate": 7.301983589677637e-06,
+      "loss": 0.4661,
+      "step": 8923
+    },
+    {
+      "epoch": 1.4655636072506313,
+      "grad_norm": 0.6163261811363203,
+      "learning_rate": 7.301616608357096e-06,
+      "loss": 0.4687,
+      "step": 8924
+    },
+    {
+      "epoch": 1.465727834458974,
+      "grad_norm": 0.6132487217057926,
+      "learning_rate": 7.301249596608244e-06,
+      "loss": 0.4746,
+      "step": 8925
+    },
+    {
+      "epoch": 1.4658920616673168,
+      "grad_norm": 0.30355394160206417,
+      "learning_rate": 7.300882554435065e-06,
+      "loss": 0.4899,
+      "step": 8926
+    },
+    {
+      "epoch": 1.4660562888756594,
+      "grad_norm": 0.3028776881337748,
+      "learning_rate": 7.3005154818415446e-06,
+      "loss": 0.4624,
+      "step": 8927
+    },
+    {
+      "epoch": 1.4662205160840023,
+      "grad_norm": 0.3490294656355712,
+      "learning_rate": 7.300148378831672e-06,
+      "loss": 0.4762,
+      "step": 8928
+    },
+    {
+      "epoch": 1.466384743292345,
+      "grad_norm": 0.38672288617898776,
+      "learning_rate": 7.2997812454094325e-06,
+      "loss": 0.4563,
+      "step": 8929
+    },
+    {
+      "epoch": 1.4665489705006878,
+      "grad_norm": 0.3460975897825717,
+      "learning_rate": 7.2994140815788146e-06,
+      "loss": 0.4558,
+      "step": 8930
+    },
+    {
+      "epoch": 1.4667131977090304,
+      "grad_norm": 0.3817812381648602,
+      "learning_rate": 7.299046887343805e-06,
+      "loss": 0.4676,
+      "step": 8931
+    },
+    {
+      "epoch": 1.466877424917373,
+      "grad_norm": 0.29091502397352326,
+      "learning_rate": 7.298679662708392e-06,
+      "loss": 0.4535,
+      "step": 8932
+    },
+    {
+      "epoch": 1.467041652125716,
+      "grad_norm": 0.3720886033235574,
+      "learning_rate": 7.298312407676565e-06,
+      "loss": 0.4719,
+      "step": 8933
+    },
+    {
+      "epoch": 1.4672058793340588,
+      "grad_norm": 0.2982734402976886,
+      "learning_rate": 7.297945122252312e-06,
+      "loss": 0.4642,
+      "step": 8934
+    },
+    {
+      "epoch": 1.4673701065424014,
+      "grad_norm": 0.2767453925146361,
+      "learning_rate": 7.297577806439622e-06,
+      "loss": 0.4618,
+      "step": 8935
+    },
+    {
+      "epoch": 1.467534333750744,
+      "grad_norm": 0.36345402905506835,
+      "learning_rate": 7.297210460242484e-06,
+      "loss": 0.4714,
+      "step": 8936
+    },
+    {
+      "epoch": 1.467698560959087,
+      "grad_norm": 0.32274055597623746,
+      "learning_rate": 7.2968430836648885e-06,
+      "loss": 0.4605,
+      "step": 8937
+    },
+    {
+      "epoch": 1.4678627881674298,
+      "grad_norm": 0.2545913876862383,
+      "learning_rate": 7.2964756767108265e-06,
+      "loss": 0.4743,
+      "step": 8938
+    },
+    {
+      "epoch": 1.4680270153757724,
+      "grad_norm": 0.3158754146163357,
+      "learning_rate": 7.296108239384287e-06,
+      "loss": 0.4785,
+      "step": 8939
+    },
+    {
+      "epoch": 1.468191242584115,
+      "grad_norm": 0.2791997204542906,
+      "learning_rate": 7.2957407716892604e-06,
+      "loss": 0.4799,
+      "step": 8940
+    },
+    {
+      "epoch": 1.468355469792458,
+      "grad_norm": 0.26346926896439843,
+      "learning_rate": 7.295373273629739e-06,
+      "loss": 0.481,
+      "step": 8941
+    },
+    {
+      "epoch": 1.4685196970008005,
+      "grad_norm": 0.27357834356219624,
+      "learning_rate": 7.295005745209713e-06,
+      "loss": 0.446,
+      "step": 8942
+    },
+    {
+      "epoch": 1.4686839242091434,
+      "grad_norm": 0.2852558176896654,
+      "learning_rate": 7.294638186433175e-06,
+      "loss": 0.4663,
+      "step": 8943
+    },
+    {
+      "epoch": 1.468848151417486,
+      "grad_norm": 0.3705262899315953,
+      "learning_rate": 7.294270597304117e-06,
+      "loss": 0.4699,
+      "step": 8944
+    },
+    {
+      "epoch": 1.469012378625829,
+      "grad_norm": 0.31750281046154766,
+      "learning_rate": 7.2939029778265295e-06,
+      "loss": 0.4582,
+      "step": 8945
+    },
+    {
+      "epoch": 1.4691766058341715,
+      "grad_norm": 0.3038648764501812,
+      "learning_rate": 7.29353532800441e-06,
+      "loss": 0.4509,
+      "step": 8946
+    },
+    {
+      "epoch": 1.4693408330425144,
+      "grad_norm": 0.29723681969502824,
+      "learning_rate": 7.293167647841745e-06,
+      "loss": 0.472,
+      "step": 8947
+    },
+    {
+      "epoch": 1.469505060250857,
+      "grad_norm": 0.3876434334700834,
+      "learning_rate": 7.292799937342534e-06,
+      "loss": 0.4829,
+      "step": 8948
+    },
+    {
+      "epoch": 1.4696692874591997,
+      "grad_norm": 0.2695956978075607,
+      "learning_rate": 7.292432196510766e-06,
+      "loss": 0.449,
+      "step": 8949
+    },
+    {
+      "epoch": 1.4698335146675425,
+      "grad_norm": 0.2974485527980798,
+      "learning_rate": 7.292064425350438e-06,
+      "loss": 0.4747,
+      "step": 8950
+    },
+    {
+      "epoch": 1.4699977418758854,
+      "grad_norm": 0.2859895877854272,
+      "learning_rate": 7.2916966238655424e-06,
+      "loss": 0.448,
+      "step": 8951
+    },
+    {
+      "epoch": 1.470161969084228,
+      "grad_norm": 0.31309097836143424,
+      "learning_rate": 7.291328792060075e-06,
+      "loss": 0.4844,
+      "step": 8952
+    },
+    {
+      "epoch": 1.4703261962925707,
+      "grad_norm": 0.31950330577085617,
+      "learning_rate": 7.290960929938032e-06,
+      "loss": 0.4677,
+      "step": 8953
+    },
+    {
+      "epoch": 1.4704904235009135,
+      "grad_norm": 0.356031776293735,
+      "learning_rate": 7.290593037503405e-06,
+      "loss": 0.4631,
+      "step": 8954
+    },
+    {
+      "epoch": 1.4706546507092564,
+      "grad_norm": 0.3892179659188917,
+      "learning_rate": 7.290225114760195e-06,
+      "loss": 0.4788,
+      "step": 8955
+    },
+    {
+      "epoch": 1.470818877917599,
+      "grad_norm": 0.3088060110020469,
+      "learning_rate": 7.289857161712393e-06,
+      "loss": 0.4773,
+      "step": 8956
+    },
+    {
+      "epoch": 1.4709831051259417,
+      "grad_norm": 0.34766766878995486,
+      "learning_rate": 7.289489178363997e-06,
+      "loss": 0.4641,
+      "step": 8957
+    },
+    {
+      "epoch": 1.4711473323342845,
+      "grad_norm": 0.987689661044155,
+      "learning_rate": 7.289121164719006e-06,
+      "loss": 0.4616,
+      "step": 8958
+    },
+    {
+      "epoch": 1.4713115595426272,
+      "grad_norm": 0.37323779225069614,
+      "learning_rate": 7.288753120781414e-06,
+      "loss": 0.4725,
+      "step": 8959
+    },
+    {
+      "epoch": 1.47147578675097,
+      "grad_norm": 0.2999592106766716,
+      "learning_rate": 7.288385046555218e-06,
+      "loss": 0.4874,
+      "step": 8960
+    },
+    {
+      "epoch": 1.4716400139593127,
+      "grad_norm": 0.4071471939682784,
+      "learning_rate": 7.288016942044418e-06,
+      "loss": 0.476,
+      "step": 8961
+    },
+    {
+      "epoch": 1.4718042411676555,
+      "grad_norm": 0.4291504258791555,
+      "learning_rate": 7.287648807253012e-06,
+      "loss": 0.4571,
+      "step": 8962
+    },
+    {
+      "epoch": 1.4719684683759982,
+      "grad_norm": 0.3502198595250842,
+      "learning_rate": 7.287280642184996e-06,
+      "loss": 0.4828,
+      "step": 8963
+    },
+    {
+      "epoch": 1.472132695584341,
+      "grad_norm": 0.375414310412938,
+      "learning_rate": 7.286912446844369e-06,
+      "loss": 0.4879,
+      "step": 8964
+    },
+    {
+      "epoch": 1.4722969227926836,
+      "grad_norm": 0.26277910260284426,
+      "learning_rate": 7.286544221235134e-06,
+      "loss": 0.4556,
+      "step": 8965
+    },
+    {
+      "epoch": 1.4724611500010263,
+      "grad_norm": 0.31490149171126364,
+      "learning_rate": 7.286175965361285e-06,
+      "loss": 0.4638,
+      "step": 8966
+    },
+    {
+      "epoch": 1.4726253772093691,
+      "grad_norm": 0.2892208531501333,
+      "learning_rate": 7.285807679226825e-06,
+      "loss": 0.4601,
+      "step": 8967
+    },
+    {
+      "epoch": 1.472789604417712,
+      "grad_norm": 0.2918932647093718,
+      "learning_rate": 7.285439362835751e-06,
+      "loss": 0.4796,
+      "step": 8968
+    },
+    {
+      "epoch": 1.4729538316260546,
+      "grad_norm": 0.2945099295955126,
+      "learning_rate": 7.285071016192067e-06,
+      "loss": 0.4636,
+      "step": 8969
+    },
+    {
+      "epoch": 1.4731180588343973,
+      "grad_norm": 0.28597820917530037,
+      "learning_rate": 7.28470263929977e-06,
+      "loss": 0.4905,
+      "step": 8970
+    },
+    {
+      "epoch": 1.4732822860427401,
+      "grad_norm": 0.4109484557655828,
+      "learning_rate": 7.284334232162864e-06,
+      "loss": 0.4654,
+      "step": 8971
+    },
+    {
+      "epoch": 1.473446513251083,
+      "grad_norm": 0.2621363643613177,
+      "learning_rate": 7.283965794785346e-06,
+      "loss": 0.4665,
+      "step": 8972
+    },
+    {
+      "epoch": 1.4736107404594256,
+      "grad_norm": 0.3483351477212068,
+      "learning_rate": 7.283597327171223e-06,
+      "loss": 0.4772,
+      "step": 8973
+    },
+    {
+      "epoch": 1.4737749676677683,
+      "grad_norm": 0.3508989330249181,
+      "learning_rate": 7.2832288293244935e-06,
+      "loss": 0.4476,
+      "step": 8974
+    },
+    {
+      "epoch": 1.4739391948761111,
+      "grad_norm": 0.44165630426238167,
+      "learning_rate": 7.282860301249162e-06,
+      "loss": 0.4698,
+      "step": 8975
+    },
+    {
+      "epoch": 1.4741034220844538,
+      "grad_norm": 0.3086075564657962,
+      "learning_rate": 7.2824917429492275e-06,
+      "loss": 0.4756,
+      "step": 8976
+    },
+    {
+      "epoch": 1.4742676492927966,
+      "grad_norm": 0.5470619925050496,
+      "learning_rate": 7.282123154428696e-06,
+      "loss": 0.4881,
+      "step": 8977
+    },
+    {
+      "epoch": 1.4744318765011393,
+      "grad_norm": 0.28503437164757617,
+      "learning_rate": 7.28175453569157e-06,
+      "loss": 0.4909,
+      "step": 8978
+    },
+    {
+      "epoch": 1.4745961037094821,
+      "grad_norm": 0.3183139987359893,
+      "learning_rate": 7.281385886741852e-06,
+      "loss": 0.4508,
+      "step": 8979
+    },
+    {
+      "epoch": 1.4747603309178248,
+      "grad_norm": 0.30594851299796155,
+      "learning_rate": 7.281017207583548e-06,
+      "loss": 0.4369,
+      "step": 8980
+    },
+    {
+      "epoch": 1.4749245581261676,
+      "grad_norm": 0.32208508578505773,
+      "learning_rate": 7.28064849822066e-06,
+      "loss": 0.471,
+      "step": 8981
+    },
+    {
+      "epoch": 1.4750887853345103,
+      "grad_norm": 0.35597240221197657,
+      "learning_rate": 7.280279758657194e-06,
+      "loss": 0.4673,
+      "step": 8982
+    },
+    {
+      "epoch": 1.475253012542853,
+      "grad_norm": 0.30726013389705537,
+      "learning_rate": 7.2799109888971544e-06,
+      "loss": 0.4376,
+      "step": 8983
+    },
+    {
+      "epoch": 1.4754172397511958,
+      "grad_norm": 0.3061948703740364,
+      "learning_rate": 7.279542188944548e-06,
+      "loss": 0.4453,
+      "step": 8984
+    },
+    {
+      "epoch": 1.4755814669595386,
+      "grad_norm": 0.3801963399207698,
+      "learning_rate": 7.279173358803376e-06,
+      "loss": 0.4601,
+      "step": 8985
+    },
+    {
+      "epoch": 1.4757456941678813,
+      "grad_norm": 0.30531711719679955,
+      "learning_rate": 7.2788044984776475e-06,
+      "loss": 0.4602,
+      "step": 8986
+    },
+    {
+      "epoch": 1.475909921376224,
+      "grad_norm": 0.26931105089669216,
+      "learning_rate": 7.2784356079713695e-06,
+      "loss": 0.458,
+      "step": 8987
+    },
+    {
+      "epoch": 1.4760741485845668,
+      "grad_norm": 0.3842988883695888,
+      "learning_rate": 7.278066687288547e-06,
+      "loss": 0.4869,
+      "step": 8988
+    },
+    {
+      "epoch": 1.4762383757929096,
+      "grad_norm": 0.28218722853074585,
+      "learning_rate": 7.277697736433186e-06,
+      "loss": 0.4642,
+      "step": 8989
+    },
+    {
+      "epoch": 1.4764026030012523,
+      "grad_norm": 0.2897705573643494,
+      "learning_rate": 7.277328755409295e-06,
+      "loss": 0.4625,
+      "step": 8990
+    },
+    {
+      "epoch": 1.476566830209595,
+      "grad_norm": 0.27586866292250783,
+      "learning_rate": 7.276959744220881e-06,
+      "loss": 0.4664,
+      "step": 8991
+    },
+    {
+      "epoch": 1.4767310574179378,
+      "grad_norm": 0.2827774853066872,
+      "learning_rate": 7.276590702871954e-06,
+      "loss": 0.4744,
+      "step": 8992
+    },
+    {
+      "epoch": 1.4768952846262804,
+      "grad_norm": 0.32702311421279057,
+      "learning_rate": 7.276221631366516e-06,
+      "loss": 0.475,
+      "step": 8993
+    },
+    {
+      "epoch": 1.4770595118346233,
+      "grad_norm": 0.2617829134294582,
+      "learning_rate": 7.275852529708582e-06,
+      "loss": 0.4705,
+      "step": 8994
+    },
+    {
+      "epoch": 1.4772237390429659,
+      "grad_norm": 0.47054078391175314,
+      "learning_rate": 7.275483397902159e-06,
+      "loss": 0.4546,
+      "step": 8995
+    },
+    {
+      "epoch": 1.4773879662513087,
+      "grad_norm": 0.28021517153771475,
+      "learning_rate": 7.275114235951256e-06,
+      "loss": 0.4753,
+      "step": 8996
+    },
+    {
+      "epoch": 1.4775521934596514,
+      "grad_norm": 0.26960945938330666,
+      "learning_rate": 7.274745043859882e-06,
+      "loss": 0.4722,
+      "step": 8997
+    },
+    {
+      "epoch": 1.4777164206679942,
+      "grad_norm": 0.34790989377611464,
+      "learning_rate": 7.274375821632045e-06,
+      "loss": 0.4666,
+      "step": 8998
+    },
+    {
+      "epoch": 1.4778806478763369,
+      "grad_norm": 0.2692479263324211,
+      "learning_rate": 7.274006569271758e-06,
+      "loss": 0.4628,
+      "step": 8999
+    },
+    {
+      "epoch": 1.4780448750846795,
+      "grad_norm": 0.429200616981824,
+      "learning_rate": 7.273637286783031e-06,
+      "loss": 0.4727,
+      "step": 9000
+    },
+    {
+      "epoch": 1.4782091022930224,
+      "grad_norm": 0.27891585980617317,
+      "learning_rate": 7.273267974169874e-06,
+      "loss": 0.4539,
+      "step": 9001
+    },
+    {
+      "epoch": 1.4783733295013652,
+      "grad_norm": 0.3124890295725687,
+      "learning_rate": 7.272898631436298e-06,
+      "loss": 0.4667,
+      "step": 9002
+    },
+    {
+      "epoch": 1.4785375567097079,
+      "grad_norm": 0.3246369076421457,
+      "learning_rate": 7.272529258586314e-06,
+      "loss": 0.4608,
+      "step": 9003
+    },
+    {
+      "epoch": 1.4787017839180505,
+      "grad_norm": 0.4909357585608995,
+      "learning_rate": 7.272159855623936e-06,
+      "loss": 0.4267,
+      "step": 9004
+    },
+    {
+      "epoch": 1.4788660111263934,
+      "grad_norm": 0.2962438853516577,
+      "learning_rate": 7.271790422553172e-06,
+      "loss": 0.4808,
+      "step": 9005
+    },
+    {
+      "epoch": 1.4790302383347362,
+      "grad_norm": 0.3504770230621263,
+      "learning_rate": 7.27142095937804e-06,
+      "loss": 0.4711,
+      "step": 9006
+    },
+    {
+      "epoch": 1.4791944655430789,
+      "grad_norm": 0.2911014714288107,
+      "learning_rate": 7.271051466102547e-06,
+      "loss": 0.4596,
+      "step": 9007
+    },
+    {
+      "epoch": 1.4793586927514215,
+      "grad_norm": 0.31010532252304307,
+      "learning_rate": 7.27068194273071e-06,
+      "loss": 0.4781,
+      "step": 9008
+    },
+    {
+      "epoch": 1.4795229199597644,
+      "grad_norm": 0.286613279240997,
+      "learning_rate": 7.270312389266542e-06,
+      "loss": 0.4617,
+      "step": 9009
+    },
+    {
+      "epoch": 1.479687147168107,
+      "grad_norm": 0.3897309711832461,
+      "learning_rate": 7.2699428057140545e-06,
+      "loss": 0.4781,
+      "step": 9010
+    },
+    {
+      "epoch": 1.4798513743764499,
+      "grad_norm": 0.37326616468971,
+      "learning_rate": 7.269573192077263e-06,
+      "loss": 0.4661,
+      "step": 9011
+    },
+    {
+      "epoch": 1.4800156015847925,
+      "grad_norm": 0.2772497866687941,
+      "learning_rate": 7.269203548360182e-06,
+      "loss": 0.4527,
+      "step": 9012
+    },
+    {
+      "epoch": 1.4801798287931354,
+      "grad_norm": 0.2775398257236335,
+      "learning_rate": 7.2688338745668264e-06,
+      "loss": 0.4574,
+      "step": 9013
+    },
+    {
+      "epoch": 1.480344056001478,
+      "grad_norm": 0.365805191968904,
+      "learning_rate": 7.26846417070121e-06,
+      "loss": 0.4674,
+      "step": 9014
+    },
+    {
+      "epoch": 1.4805082832098209,
+      "grad_norm": 0.2995131322182461,
+      "learning_rate": 7.26809443676735e-06,
+      "loss": 0.4625,
+      "step": 9015
+    },
+    {
+      "epoch": 1.4806725104181635,
+      "grad_norm": 0.2920748910867749,
+      "learning_rate": 7.2677246727692605e-06,
+      "loss": 0.449,
+      "step": 9016
+    },
+    {
+      "epoch": 1.4808367376265061,
+      "grad_norm": 0.3134085998958177,
+      "learning_rate": 7.267354878710957e-06,
+      "loss": 0.4752,
+      "step": 9017
+    },
+    {
+      "epoch": 1.481000964834849,
+      "grad_norm": 0.2910762798061273,
+      "learning_rate": 7.266985054596457e-06,
+      "loss": 0.4668,
+      "step": 9018
+    },
+    {
+      "epoch": 1.4811651920431919,
+      "grad_norm": 0.3620826935634441,
+      "learning_rate": 7.266615200429778e-06,
+      "loss": 0.4694,
+      "step": 9019
+    },
+    {
+      "epoch": 1.4813294192515345,
+      "grad_norm": 0.3239612400513165,
+      "learning_rate": 7.266245316214935e-06,
+      "loss": 0.4685,
+      "step": 9020
+    },
+    {
+      "epoch": 1.4814936464598771,
+      "grad_norm": 0.512868777403178,
+      "learning_rate": 7.265875401955947e-06,
+      "loss": 0.4471,
+      "step": 9021
+    },
+    {
+      "epoch": 1.48165787366822,
+      "grad_norm": 0.35965132682584766,
+      "learning_rate": 7.265505457656831e-06,
+      "loss": 0.475,
+      "step": 9022
+    },
+    {
+      "epoch": 1.4818221008765629,
+      "grad_norm": 0.3268419705604959,
+      "learning_rate": 7.265135483321604e-06,
+      "loss": 0.4804,
+      "step": 9023
+    },
+    {
+      "epoch": 1.4819863280849055,
+      "grad_norm": 0.3992553047998647,
+      "learning_rate": 7.264765478954286e-06,
+      "loss": 0.449,
+      "step": 9024
+    },
+    {
+      "epoch": 1.4821505552932481,
+      "grad_norm": 0.2918082085432597,
+      "learning_rate": 7.264395444558895e-06,
+      "loss": 0.4584,
+      "step": 9025
+    },
+    {
+      "epoch": 1.482314782501591,
+      "grad_norm": 0.270882114076403,
+      "learning_rate": 7.264025380139448e-06,
+      "loss": 0.4664,
+      "step": 9026
+    },
+    {
+      "epoch": 1.4824790097099336,
+      "grad_norm": 0.3532222289683168,
+      "learning_rate": 7.263655285699966e-06,
+      "loss": 0.4719,
+      "step": 9027
+    },
+    {
+      "epoch": 1.4826432369182765,
+      "grad_norm": 0.2849963487764551,
+      "learning_rate": 7.263285161244469e-06,
+      "loss": 0.4576,
+      "step": 9028
+    },
+    {
+      "epoch": 1.4828074641266191,
+      "grad_norm": 0.3139206838920788,
+      "learning_rate": 7.262915006776978e-06,
+      "loss": 0.4721,
+      "step": 9029
+    },
+    {
+      "epoch": 1.482971691334962,
+      "grad_norm": 0.48344010886321853,
+      "learning_rate": 7.26254482230151e-06,
+      "loss": 0.4539,
+      "step": 9030
+    },
+    {
+      "epoch": 1.4831359185433046,
+      "grad_norm": 0.3287835530250805,
+      "learning_rate": 7.262174607822088e-06,
+      "loss": 0.4511,
+      "step": 9031
+    },
+    {
+      "epoch": 1.4833001457516475,
+      "grad_norm": 0.2750891894079266,
+      "learning_rate": 7.261804363342731e-06,
+      "loss": 0.4718,
+      "step": 9032
+    },
+    {
+      "epoch": 1.4834643729599901,
+      "grad_norm": 0.3425506478423419,
+      "learning_rate": 7.261434088867463e-06,
+      "loss": 0.4768,
+      "step": 9033
+    },
+    {
+      "epoch": 1.4836286001683328,
+      "grad_norm": 0.298876986879865,
+      "learning_rate": 7.261063784400304e-06,
+      "loss": 0.4703,
+      "step": 9034
+    },
+    {
+      "epoch": 1.4837928273766756,
+      "grad_norm": 0.34455142421650525,
+      "learning_rate": 7.260693449945274e-06,
+      "loss": 0.478,
+      "step": 9035
+    },
+    {
+      "epoch": 1.4839570545850185,
+      "grad_norm": 0.34054511742199567,
+      "learning_rate": 7.260323085506398e-06,
+      "loss": 0.4653,
+      "step": 9036
+    },
+    {
+      "epoch": 1.4841212817933611,
+      "grad_norm": 0.2849754511923055,
+      "learning_rate": 7.259952691087697e-06,
+      "loss": 0.4581,
+      "step": 9037
+    },
+    {
+      "epoch": 1.4842855090017038,
+      "grad_norm": 0.31908669245109694,
+      "learning_rate": 7.259582266693196e-06,
+      "loss": 0.4709,
+      "step": 9038
+    },
+    {
+      "epoch": 1.4844497362100466,
+      "grad_norm": 0.261760821048014,
+      "learning_rate": 7.259211812326916e-06,
+      "loss": 0.4595,
+      "step": 9039
+    },
+    {
+      "epoch": 1.4846139634183895,
+      "grad_norm": 0.5699588523422938,
+      "learning_rate": 7.258841327992879e-06,
+      "loss": 0.4435,
+      "step": 9040
+    },
+    {
+      "epoch": 1.484778190626732,
+      "grad_norm": 0.3436625253191333,
+      "learning_rate": 7.258470813695112e-06,
+      "loss": 0.49,
+      "step": 9041
+    },
+    {
+      "epoch": 1.4849424178350747,
+      "grad_norm": 0.29229791987285086,
+      "learning_rate": 7.258100269437637e-06,
+      "loss": 0.448,
+      "step": 9042
+    },
+    {
+      "epoch": 1.4851066450434176,
+      "grad_norm": 0.2696646005456477,
+      "learning_rate": 7.257729695224482e-06,
+      "loss": 0.4741,
+      "step": 9043
+    },
+    {
+      "epoch": 1.4852708722517602,
+      "grad_norm": 0.5564414897893172,
+      "learning_rate": 7.257359091059668e-06,
+      "loss": 0.4651,
+      "step": 9044
+    },
+    {
+      "epoch": 1.485435099460103,
+      "grad_norm": 0.37273520474470473,
+      "learning_rate": 7.256988456947221e-06,
+      "loss": 0.4658,
+      "step": 9045
+    },
+    {
+      "epoch": 1.4855993266684457,
+      "grad_norm": 0.26984808619858086,
+      "learning_rate": 7.256617792891168e-06,
+      "loss": 0.4763,
+      "step": 9046
+    },
+    {
+      "epoch": 1.4857635538767886,
+      "grad_norm": 0.30726132225802555,
+      "learning_rate": 7.256247098895533e-06,
+      "loss": 0.4564,
+      "step": 9047
+    },
+    {
+      "epoch": 1.4859277810851312,
+      "grad_norm": 0.2964204921364908,
+      "learning_rate": 7.255876374964341e-06,
+      "loss": 0.449,
+      "step": 9048
+    },
+    {
+      "epoch": 1.486092008293474,
+      "grad_norm": 0.34428474234477524,
+      "learning_rate": 7.255505621101623e-06,
+      "loss": 0.4738,
+      "step": 9049
+    },
+    {
+      "epoch": 1.4862562355018167,
+      "grad_norm": 0.3100524445547254,
+      "learning_rate": 7.255134837311402e-06,
+      "loss": 0.4513,
+      "step": 9050
+    },
+    {
+      "epoch": 1.4864204627101594,
+      "grad_norm": 0.2968065640191466,
+      "learning_rate": 7.254764023597705e-06,
+      "loss": 0.4488,
+      "step": 9051
+    },
+    {
+      "epoch": 1.4865846899185022,
+      "grad_norm": 0.42991865968143095,
+      "learning_rate": 7.254393179964561e-06,
+      "loss": 0.4656,
+      "step": 9052
+    },
+    {
+      "epoch": 1.486748917126845,
+      "grad_norm": 0.2840509670038126,
+      "learning_rate": 7.254022306415996e-06,
+      "loss": 0.4576,
+      "step": 9053
+    },
+    {
+      "epoch": 1.4869131443351877,
+      "grad_norm": 0.27885545284201946,
+      "learning_rate": 7.25365140295604e-06,
+      "loss": 0.4566,
+      "step": 9054
+    },
+    {
+      "epoch": 1.4870773715435304,
+      "grad_norm": 0.2880230279300889,
+      "learning_rate": 7.253280469588722e-06,
+      "loss": 0.467,
+      "step": 9055
+    },
+    {
+      "epoch": 1.4872415987518732,
+      "grad_norm": 0.2713198311101982,
+      "learning_rate": 7.252909506318067e-06,
+      "loss": 0.4737,
+      "step": 9056
+    },
+    {
+      "epoch": 1.487405825960216,
+      "grad_norm": 0.29087205139831046,
+      "learning_rate": 7.252538513148108e-06,
+      "loss": 0.457,
+      "step": 9057
+    },
+    {
+      "epoch": 1.4875700531685587,
+      "grad_norm": 0.3218582414140392,
+      "learning_rate": 7.2521674900828705e-06,
+      "loss": 0.4679,
+      "step": 9058
+    },
+    {
+      "epoch": 1.4877342803769014,
+      "grad_norm": 0.3022988004594291,
+      "learning_rate": 7.251796437126388e-06,
+      "loss": 0.4762,
+      "step": 9059
+    },
+    {
+      "epoch": 1.4878985075852442,
+      "grad_norm": 0.3237420410141972,
+      "learning_rate": 7.251425354282689e-06,
+      "loss": 0.4806,
+      "step": 9060
+    },
+    {
+      "epoch": 1.4880627347935869,
+      "grad_norm": 0.40148076135141486,
+      "learning_rate": 7.251054241555803e-06,
+      "loss": 0.4479,
+      "step": 9061
+    },
+    {
+      "epoch": 1.4882269620019297,
+      "grad_norm": 0.3021622550741741,
+      "learning_rate": 7.250683098949761e-06,
+      "loss": 0.4489,
+      "step": 9062
+    },
+    {
+      "epoch": 1.4883911892102724,
+      "grad_norm": 0.2737397256477465,
+      "learning_rate": 7.250311926468595e-06,
+      "loss": 0.4622,
+      "step": 9063
+    },
+    {
+      "epoch": 1.4885554164186152,
+      "grad_norm": 0.2838829341107318,
+      "learning_rate": 7.249940724116335e-06,
+      "loss": 0.4788,
+      "step": 9064
+    },
+    {
+      "epoch": 1.4887196436269579,
+      "grad_norm": 0.2900387237367032,
+      "learning_rate": 7.2495694918970125e-06,
+      "loss": 0.5016,
+      "step": 9065
+    },
+    {
+      "epoch": 1.4888838708353007,
+      "grad_norm": 0.32723225864852523,
+      "learning_rate": 7.249198229814661e-06,
+      "loss": 0.4635,
+      "step": 9066
+    },
+    {
+      "epoch": 1.4890480980436434,
+      "grad_norm": 0.2929153755364235,
+      "learning_rate": 7.248826937873313e-06,
+      "loss": 0.4627,
+      "step": 9067
+    },
+    {
+      "epoch": 1.489212325251986,
+      "grad_norm": 0.29032930548652564,
+      "learning_rate": 7.248455616076998e-06,
+      "loss": 0.4736,
+      "step": 9068
+    },
+    {
+      "epoch": 1.4893765524603289,
+      "grad_norm": 0.35280273211851365,
+      "learning_rate": 7.248084264429751e-06,
+      "loss": 0.473,
+      "step": 9069
+    },
+    {
+      "epoch": 1.4895407796686717,
+      "grad_norm": 0.46850827088761615,
+      "learning_rate": 7.2477128829356055e-06,
+      "loss": 0.4495,
+      "step": 9070
+    },
+    {
+      "epoch": 1.4897050068770143,
+      "grad_norm": 0.3028158132164672,
+      "learning_rate": 7.247341471598596e-06,
+      "loss": 0.4797,
+      "step": 9071
+    },
+    {
+      "epoch": 1.489869234085357,
+      "grad_norm": 0.3222872181362865,
+      "learning_rate": 7.2469700304227535e-06,
+      "loss": 0.4687,
+      "step": 9072
+    },
+    {
+      "epoch": 1.4900334612936998,
+      "grad_norm": 0.3353242450473158,
+      "learning_rate": 7.246598559412115e-06,
+      "loss": 0.4442,
+      "step": 9073
+    },
+    {
+      "epoch": 1.4901976885020427,
+      "grad_norm": 0.3147282673915163,
+      "learning_rate": 7.246227058570714e-06,
+      "loss": 0.4727,
+      "step": 9074
+    },
+    {
+      "epoch": 1.4903619157103853,
+      "grad_norm": 0.26213956343160233,
+      "learning_rate": 7.2458555279025836e-06,
+      "loss": 0.4803,
+      "step": 9075
+    },
+    {
+      "epoch": 1.490526142918728,
+      "grad_norm": 0.43566408018354824,
+      "learning_rate": 7.245483967411762e-06,
+      "loss": 0.4805,
+      "step": 9076
+    },
+    {
+      "epoch": 1.4906903701270708,
+      "grad_norm": 0.49105090744394064,
+      "learning_rate": 7.2451123771022816e-06,
+      "loss": 0.4697,
+      "step": 9077
+    },
+    {
+      "epoch": 1.4908545973354135,
+      "grad_norm": 0.5978965467093454,
+      "learning_rate": 7.244740756978181e-06,
+      "loss": 0.4664,
+      "step": 9078
+    },
+    {
+      "epoch": 1.4910188245437563,
+      "grad_norm": 0.34767763702904925,
+      "learning_rate": 7.2443691070434955e-06,
+      "loss": 0.4618,
+      "step": 9079
+    },
+    {
+      "epoch": 1.491183051752099,
+      "grad_norm": 0.2835559414850179,
+      "learning_rate": 7.2439974273022625e-06,
+      "loss": 0.4952,
+      "step": 9080
+    },
+    {
+      "epoch": 1.4913472789604418,
+      "grad_norm": 0.36899619802931427,
+      "learning_rate": 7.243625717758516e-06,
+      "loss": 0.4588,
+      "step": 9081
+    },
+    {
+      "epoch": 1.4915115061687845,
+      "grad_norm": 0.31211856962328216,
+      "learning_rate": 7.243253978416294e-06,
+      "loss": 0.4719,
+      "step": 9082
+    },
+    {
+      "epoch": 1.4916757333771273,
+      "grad_norm": 0.2795658714787976,
+      "learning_rate": 7.242882209279637e-06,
+      "loss": 0.4635,
+      "step": 9083
+    },
+    {
+      "epoch": 1.49183996058547,
+      "grad_norm": 0.25916023604519495,
+      "learning_rate": 7.242510410352581e-06,
+      "loss": 0.4682,
+      "step": 9084
+    },
+    {
+      "epoch": 1.4920041877938126,
+      "grad_norm": 0.3129340495290174,
+      "learning_rate": 7.242138581639162e-06,
+      "loss": 0.4615,
+      "step": 9085
+    },
+    {
+      "epoch": 1.4921684150021555,
+      "grad_norm": 0.36526210508797624,
+      "learning_rate": 7.24176672314342e-06,
+      "loss": 0.4804,
+      "step": 9086
+    },
+    {
+      "epoch": 1.4923326422104983,
+      "grad_norm": 0.37940069172741353,
+      "learning_rate": 7.241394834869395e-06,
+      "loss": 0.444,
+      "step": 9087
+    },
+    {
+      "epoch": 1.492496869418841,
+      "grad_norm": 0.36743810559285073,
+      "learning_rate": 7.241022916821124e-06,
+      "loss": 0.4617,
+      "step": 9088
+    },
+    {
+      "epoch": 1.4926610966271836,
+      "grad_norm": 0.3399224851236104,
+      "learning_rate": 7.240650969002647e-06,
+      "loss": 0.4767,
+      "step": 9089
+    },
+    {
+      "epoch": 1.4928253238355265,
+      "grad_norm": 0.3568026906661504,
+      "learning_rate": 7.2402789914180045e-06,
+      "loss": 0.4583,
+      "step": 9090
+    },
+    {
+      "epoch": 1.4929895510438693,
+      "grad_norm": 0.47928165555028224,
+      "learning_rate": 7.239906984071238e-06,
+      "loss": 0.4717,
+      "step": 9091
+    },
+    {
+      "epoch": 1.493153778252212,
+      "grad_norm": 0.3438429100691715,
+      "learning_rate": 7.239534946966384e-06,
+      "loss": 0.4975,
+      "step": 9092
+    },
+    {
+      "epoch": 1.4933180054605546,
+      "grad_norm": 0.37410444420126543,
+      "learning_rate": 7.239162880107485e-06,
+      "loss": 0.4614,
+      "step": 9093
+    },
+    {
+      "epoch": 1.4934822326688975,
+      "grad_norm": 0.3130268740405012,
+      "learning_rate": 7.238790783498583e-06,
+      "loss": 0.4601,
+      "step": 9094
+    },
+    {
+      "epoch": 1.49364645987724,
+      "grad_norm": 0.34707548690935847,
+      "learning_rate": 7.238418657143716e-06,
+      "loss": 0.4751,
+      "step": 9095
+    },
+    {
+      "epoch": 1.493810687085583,
+      "grad_norm": 0.2975015764973547,
+      "learning_rate": 7.2380465010469316e-06,
+      "loss": 0.4566,
+      "step": 9096
+    },
+    {
+      "epoch": 1.4939749142939256,
+      "grad_norm": 0.2923288441893135,
+      "learning_rate": 7.237674315212267e-06,
+      "loss": 0.4669,
+      "step": 9097
+    },
+    {
+      "epoch": 1.4941391415022685,
+      "grad_norm": 0.31449192041777363,
+      "learning_rate": 7.237302099643766e-06,
+      "loss": 0.4665,
+      "step": 9098
+    },
+    {
+      "epoch": 1.494303368710611,
+      "grad_norm": 0.4705678262595334,
+      "learning_rate": 7.23692985434547e-06,
+      "loss": 0.4765,
+      "step": 9099
+    },
+    {
+      "epoch": 1.494467595918954,
+      "grad_norm": 0.3617709294721401,
+      "learning_rate": 7.236557579321424e-06,
+      "loss": 0.4689,
+      "step": 9100
+    },
+    {
+      "epoch": 1.4946318231272966,
+      "grad_norm": 0.3301324253026518,
+      "learning_rate": 7.23618527457567e-06,
+      "loss": 0.442,
+      "step": 9101
+    },
+    {
+      "epoch": 1.4947960503356392,
+      "grad_norm": 0.658918156487996,
+      "learning_rate": 7.235812940112252e-06,
+      "loss": 0.4592,
+      "step": 9102
+    },
+    {
+      "epoch": 1.494960277543982,
+      "grad_norm": 1.0723152036270371,
+      "learning_rate": 7.235440575935215e-06,
+      "loss": 0.472,
+      "step": 9103
+    },
+    {
+      "epoch": 1.495124504752325,
+      "grad_norm": 0.32671311937006464,
+      "learning_rate": 7.235068182048599e-06,
+      "loss": 0.4698,
+      "step": 9104
+    },
+    {
+      "epoch": 1.4952887319606676,
+      "grad_norm": 0.28952442856594846,
+      "learning_rate": 7.234695758456454e-06,
+      "loss": 0.4607,
+      "step": 9105
+    },
+    {
+      "epoch": 1.4954529591690102,
+      "grad_norm": 0.5227314578459401,
+      "learning_rate": 7.234323305162822e-06,
+      "loss": 0.4777,
+      "step": 9106
+    },
+    {
+      "epoch": 1.495617186377353,
+      "grad_norm": 0.2734681071879501,
+      "learning_rate": 7.233950822171748e-06,
+      "loss": 0.4682,
+      "step": 9107
+    },
+    {
+      "epoch": 1.495781413585696,
+      "grad_norm": 0.3778217861495209,
+      "learning_rate": 7.233578309487279e-06,
+      "loss": 0.46,
+      "step": 9108
+    },
+    {
+      "epoch": 1.4959456407940386,
+      "grad_norm": 0.3616470314992457,
+      "learning_rate": 7.23320576711346e-06,
+      "loss": 0.4457,
+      "step": 9109
+    },
+    {
+      "epoch": 1.4961098680023812,
+      "grad_norm": 0.31428696869504247,
+      "learning_rate": 7.232833195054337e-06,
+      "loss": 0.4759,
+      "step": 9110
+    },
+    {
+      "epoch": 1.496274095210724,
+      "grad_norm": 0.3320566406078576,
+      "learning_rate": 7.232460593313957e-06,
+      "loss": 0.4645,
+      "step": 9111
+    },
+    {
+      "epoch": 1.4964383224190667,
+      "grad_norm": 0.5345790329651504,
+      "learning_rate": 7.232087961896366e-06,
+      "loss": 0.4563,
+      "step": 9112
+    },
+    {
+      "epoch": 1.4966025496274096,
+      "grad_norm": 0.41574619219644277,
+      "learning_rate": 7.231715300805613e-06,
+      "loss": 0.473,
+      "step": 9113
+    },
+    {
+      "epoch": 1.4967667768357522,
+      "grad_norm": 0.2652127144284407,
+      "learning_rate": 7.231342610045744e-06,
+      "loss": 0.4606,
+      "step": 9114
+    },
+    {
+      "epoch": 1.496931004044095,
+      "grad_norm": 0.2908442547923296,
+      "learning_rate": 7.230969889620806e-06,
+      "loss": 0.4607,
+      "step": 9115
+    },
+    {
+      "epoch": 1.4970952312524377,
+      "grad_norm": 0.3965280190532817,
+      "learning_rate": 7.230597139534848e-06,
+      "loss": 0.465,
+      "step": 9116
+    },
+    {
+      "epoch": 1.4972594584607806,
+      "grad_norm": 0.37558384455595123,
+      "learning_rate": 7.230224359791918e-06,
+      "loss": 0.4845,
+      "step": 9117
+    },
+    {
+      "epoch": 1.4974236856691232,
+      "grad_norm": 0.2828617605706177,
+      "learning_rate": 7.2298515503960665e-06,
+      "loss": 0.4719,
+      "step": 9118
+    },
+    {
+      "epoch": 1.4975879128774658,
+      "grad_norm": 0.28998875121414835,
+      "learning_rate": 7.229478711351341e-06,
+      "loss": 0.4722,
+      "step": 9119
+    },
+    {
+      "epoch": 1.4977521400858087,
+      "grad_norm": 0.3699351941513517,
+      "learning_rate": 7.229105842661792e-06,
+      "loss": 0.4716,
+      "step": 9120
+    },
+    {
+      "epoch": 1.4979163672941516,
+      "grad_norm": 0.4265125664411172,
+      "learning_rate": 7.228732944331468e-06,
+      "loss": 0.4693,
+      "step": 9121
+    },
+    {
+      "epoch": 1.4980805945024942,
+      "grad_norm": 0.4204329568659064,
+      "learning_rate": 7.228360016364418e-06,
+      "loss": 0.4519,
+      "step": 9122
+    },
+    {
+      "epoch": 1.4982448217108368,
+      "grad_norm": 0.5794156378337116,
+      "learning_rate": 7.227987058764696e-06,
+      "loss": 0.4771,
+      "step": 9123
+    },
+    {
+      "epoch": 1.4984090489191797,
+      "grad_norm": 0.2791439561851294,
+      "learning_rate": 7.22761407153635e-06,
+      "loss": 0.4829,
+      "step": 9124
+    },
+    {
+      "epoch": 1.4985732761275226,
+      "grad_norm": 0.6543863320758504,
+      "learning_rate": 7.227241054683431e-06,
+      "loss": 0.466,
+      "step": 9125
+    },
+    {
+      "epoch": 1.4987375033358652,
+      "grad_norm": 0.4130277000795869,
+      "learning_rate": 7.2268680082099915e-06,
+      "loss": 0.4615,
+      "step": 9126
+    },
+    {
+      "epoch": 1.4989017305442078,
+      "grad_norm": 0.29982715625006073,
+      "learning_rate": 7.226494932120081e-06,
+      "loss": 0.4564,
+      "step": 9127
+    },
+    {
+      "epoch": 1.4990659577525507,
+      "grad_norm": 0.35911683308681835,
+      "learning_rate": 7.226121826417755e-06,
+      "loss": 0.4566,
+      "step": 9128
+    },
+    {
+      "epoch": 1.4992301849608933,
+      "grad_norm": 0.3561038189163182,
+      "learning_rate": 7.225748691107063e-06,
+      "loss": 0.4506,
+      "step": 9129
+    },
+    {
+      "epoch": 1.4993944121692362,
+      "grad_norm": 0.31919930536086555,
+      "learning_rate": 7.225375526192059e-06,
+      "loss": 0.4602,
+      "step": 9130
+    },
+    {
+      "epoch": 1.4995586393775788,
+      "grad_norm": 0.37857621222149634,
+      "learning_rate": 7.225002331676795e-06,
+      "loss": 0.4588,
+      "step": 9131
+    },
+    {
+      "epoch": 1.4997228665859217,
+      "grad_norm": 0.29728581761591794,
+      "learning_rate": 7.224629107565324e-06,
+      "loss": 0.4722,
+      "step": 9132
+    },
+    {
+      "epoch": 1.4998870937942643,
+      "grad_norm": 0.4309870688263262,
+      "learning_rate": 7.224255853861701e-06,
+      "loss": 0.4506,
+      "step": 9133
+    },
+    {
+      "epoch": 1.5000513210026072,
+      "grad_norm": 0.4385984762113458,
+      "learning_rate": 7.22388257056998e-06,
+      "loss": 0.4754,
+      "step": 9134
+    },
+    {
+      "epoch": 1.5002155482109498,
+      "grad_norm": 0.37723360627303476,
+      "learning_rate": 7.2235092576942125e-06,
+      "loss": 0.4835,
+      "step": 9135
+    },
+    {
+      "epoch": 1.5003797754192925,
+      "grad_norm": 0.2969754460478771,
+      "learning_rate": 7.223135915238455e-06,
+      "loss": 0.4753,
+      "step": 9136
+    },
+    {
+      "epoch": 1.5005440026276353,
+      "grad_norm": 0.3254557370821571,
+      "learning_rate": 7.222762543206763e-06,
+      "loss": 0.4764,
+      "step": 9137
+    },
+    {
+      "epoch": 1.5007082298359782,
+      "grad_norm": 0.3534731655018746,
+      "learning_rate": 7.222389141603192e-06,
+      "loss": 0.4517,
+      "step": 9138
+    },
+    {
+      "epoch": 1.5008724570443208,
+      "grad_norm": 1.0771222470195647,
+      "learning_rate": 7.222015710431795e-06,
+      "loss": 0.4836,
+      "step": 9139
+    },
+    {
+      "epoch": 1.5010366842526635,
+      "grad_norm": 0.4629425359743406,
+      "learning_rate": 7.221642249696629e-06,
+      "loss": 0.4474,
+      "step": 9140
+    },
+    {
+      "epoch": 1.5012009114610063,
+      "grad_norm": 0.3712616941989587,
+      "learning_rate": 7.221268759401751e-06,
+      "loss": 0.4755,
+      "step": 9141
+    },
+    {
+      "epoch": 1.5013651386693492,
+      "grad_norm": 0.334863403671556,
+      "learning_rate": 7.220895239551218e-06,
+      "loss": 0.4748,
+      "step": 9142
+    },
+    {
+      "epoch": 1.5015293658776918,
+      "grad_norm": 0.3404372632720664,
+      "learning_rate": 7.220521690149084e-06,
+      "loss": 0.4634,
+      "step": 9143
+    },
+    {
+      "epoch": 1.5016935930860345,
+      "grad_norm": 0.3129244116574231,
+      "learning_rate": 7.220148111199409e-06,
+      "loss": 0.4562,
+      "step": 9144
+    },
+    {
+      "epoch": 1.5018578202943773,
+      "grad_norm": 0.3654382271943735,
+      "learning_rate": 7.219774502706248e-06,
+      "loss": 0.4505,
+      "step": 9145
+    },
+    {
+      "epoch": 1.5020220475027202,
+      "grad_norm": 0.35624390931397154,
+      "learning_rate": 7.21940086467366e-06,
+      "loss": 0.4513,
+      "step": 9146
+    },
+    {
+      "epoch": 1.5021862747110628,
+      "grad_norm": 0.3228160726866767,
+      "learning_rate": 7.219027197105705e-06,
+      "loss": 0.4384,
+      "step": 9147
+    },
+    {
+      "epoch": 1.5023505019194054,
+      "grad_norm": 0.3362682026085022,
+      "learning_rate": 7.218653500006438e-06,
+      "loss": 0.4613,
+      "step": 9148
+    },
+    {
+      "epoch": 1.502514729127748,
+      "grad_norm": 0.3517407041807582,
+      "learning_rate": 7.21827977337992e-06,
+      "loss": 0.4515,
+      "step": 9149
+    },
+    {
+      "epoch": 1.502678956336091,
+      "grad_norm": 0.3059612901310846,
+      "learning_rate": 7.217906017230208e-06,
+      "loss": 0.4642,
+      "step": 9150
+    },
+    {
+      "epoch": 1.5028431835444338,
+      "grad_norm": 0.3005958427302342,
+      "learning_rate": 7.217532231561363e-06,
+      "loss": 0.4691,
+      "step": 9151
+    },
+    {
+      "epoch": 1.5030074107527764,
+      "grad_norm": 0.5129411936571094,
+      "learning_rate": 7.217158416377445e-06,
+      "loss": 0.465,
+      "step": 9152
+    },
+    {
+      "epoch": 1.503171637961119,
+      "grad_norm": 0.3666288166046643,
+      "learning_rate": 7.216784571682513e-06,
+      "loss": 0.4509,
+      "step": 9153
+    },
+    {
+      "epoch": 1.503335865169462,
+      "grad_norm": 0.294422504507227,
+      "learning_rate": 7.216410697480627e-06,
+      "loss": 0.4554,
+      "step": 9154
+    },
+    {
+      "epoch": 1.5035000923778048,
+      "grad_norm": 0.3113601579543826,
+      "learning_rate": 7.21603679377585e-06,
+      "loss": 0.4828,
+      "step": 9155
+    },
+    {
+      "epoch": 1.5036643195861474,
+      "grad_norm": 0.30263614060555033,
+      "learning_rate": 7.215662860572238e-06,
+      "loss": 0.4694,
+      "step": 9156
+    },
+    {
+      "epoch": 1.50382854679449,
+      "grad_norm": 0.4734181151005649,
+      "learning_rate": 7.2152888978738585e-06,
+      "loss": 0.4762,
+      "step": 9157
+    },
+    {
+      "epoch": 1.503992774002833,
+      "grad_norm": 0.3287225146005883,
+      "learning_rate": 7.214914905684769e-06,
+      "loss": 0.4666,
+      "step": 9158
+    },
+    {
+      "epoch": 1.5041570012111758,
+      "grad_norm": 0.305886165071671,
+      "learning_rate": 7.214540884009032e-06,
+      "loss": 0.4496,
+      "step": 9159
+    },
+    {
+      "epoch": 1.5043212284195184,
+      "grad_norm": 0.3011882096853734,
+      "learning_rate": 7.214166832850711e-06,
+      "loss": 0.4582,
+      "step": 9160
+    },
+    {
+      "epoch": 1.504485455627861,
+      "grad_norm": 0.41488963999633277,
+      "learning_rate": 7.213792752213867e-06,
+      "loss": 0.4648,
+      "step": 9161
+    },
+    {
+      "epoch": 1.504649682836204,
+      "grad_norm": 0.31544674799165684,
+      "learning_rate": 7.213418642102564e-06,
+      "loss": 0.4844,
+      "step": 9162
+    },
+    {
+      "epoch": 1.5048139100445468,
+      "grad_norm": 0.28949199485026034,
+      "learning_rate": 7.213044502520866e-06,
+      "loss": 0.4671,
+      "step": 9163
+    },
+    {
+      "epoch": 1.5049781372528894,
+      "grad_norm": 0.2939590327333749,
+      "learning_rate": 7.212670333472835e-06,
+      "loss": 0.4771,
+      "step": 9164
+    },
+    {
+      "epoch": 1.505142364461232,
+      "grad_norm": 0.3670101412740805,
+      "learning_rate": 7.212296134962533e-06,
+      "loss": 0.4688,
+      "step": 9165
+    },
+    {
+      "epoch": 1.5053065916695747,
+      "grad_norm": 0.28723258461262186,
+      "learning_rate": 7.2119219069940296e-06,
+      "loss": 0.4862,
+      "step": 9166
+    },
+    {
+      "epoch": 1.5054708188779176,
+      "grad_norm": 0.32542175887326896,
+      "learning_rate": 7.2115476495713846e-06,
+      "loss": 0.4553,
+      "step": 9167
+    },
+    {
+      "epoch": 1.5056350460862604,
+      "grad_norm": 0.3007208342378376,
+      "learning_rate": 7.211173362698664e-06,
+      "loss": 0.4659,
+      "step": 9168
+    },
+    {
+      "epoch": 1.505799273294603,
+      "grad_norm": 0.38737030393235344,
+      "learning_rate": 7.210799046379935e-06,
+      "loss": 0.4686,
+      "step": 9169
+    },
+    {
+      "epoch": 1.5059635005029457,
+      "grad_norm": 0.26963051081694284,
+      "learning_rate": 7.210424700619259e-06,
+      "loss": 0.4454,
+      "step": 9170
+    },
+    {
+      "epoch": 1.5061277277112886,
+      "grad_norm": 0.34262459734970785,
+      "learning_rate": 7.210050325420705e-06,
+      "loss": 0.4565,
+      "step": 9171
+    },
+    {
+      "epoch": 1.5062919549196314,
+      "grad_norm": 0.5174600911956027,
+      "learning_rate": 7.209675920788338e-06,
+      "loss": 0.4668,
+      "step": 9172
+    },
+    {
+      "epoch": 1.506456182127974,
+      "grad_norm": 0.29129980320508153,
+      "learning_rate": 7.209301486726226e-06,
+      "loss": 0.4603,
+      "step": 9173
+    },
+    {
+      "epoch": 1.5066204093363167,
+      "grad_norm": 0.2999281854094833,
+      "learning_rate": 7.208927023238432e-06,
+      "loss": 0.4459,
+      "step": 9174
+    },
+    {
+      "epoch": 1.5067846365446596,
+      "grad_norm": 0.34509789834469307,
+      "learning_rate": 7.208552530329026e-06,
+      "loss": 0.4825,
+      "step": 9175
+    },
+    {
+      "epoch": 1.5069488637530024,
+      "grad_norm": 0.33035112975237557,
+      "learning_rate": 7.208178008002076e-06,
+      "loss": 0.4466,
+      "step": 9176
+    },
+    {
+      "epoch": 1.507113090961345,
+      "grad_norm": 0.33090831360597694,
+      "learning_rate": 7.2078034562616465e-06,
+      "loss": 0.4671,
+      "step": 9177
+    },
+    {
+      "epoch": 1.5072773181696877,
+      "grad_norm": 0.2959356681925396,
+      "learning_rate": 7.207428875111809e-06,
+      "loss": 0.4513,
+      "step": 9178
+    },
+    {
+      "epoch": 1.5074415453780305,
+      "grad_norm": 0.27555625602430894,
+      "learning_rate": 7.207054264556629e-06,
+      "loss": 0.4824,
+      "step": 9179
+    },
+    {
+      "epoch": 1.5076057725863734,
+      "grad_norm": 0.3831379474711666,
+      "learning_rate": 7.206679624600177e-06,
+      "loss": 0.4436,
+      "step": 9180
+    },
+    {
+      "epoch": 1.507769999794716,
+      "grad_norm": 0.3182477305511219,
+      "learning_rate": 7.206304955246521e-06,
+      "loss": 0.4693,
+      "step": 9181
+    },
+    {
+      "epoch": 1.5079342270030587,
+      "grad_norm": 0.3604807686672664,
+      "learning_rate": 7.2059302564997295e-06,
+      "loss": 0.4604,
+      "step": 9182
+    },
+    {
+      "epoch": 1.5080984542114013,
+      "grad_norm": 0.5300038312700972,
+      "learning_rate": 7.205555528363875e-06,
+      "loss": 0.4699,
+      "step": 9183
+    },
+    {
+      "epoch": 1.5082626814197442,
+      "grad_norm": 0.2701038850534607,
+      "learning_rate": 7.205180770843024e-06,
+      "loss": 0.4949,
+      "step": 9184
+    },
+    {
+      "epoch": 1.508426908628087,
+      "grad_norm": 0.7164451544357832,
+      "learning_rate": 7.204805983941249e-06,
+      "loss": 0.4636,
+      "step": 9185
+    },
+    {
+      "epoch": 1.5085911358364297,
+      "grad_norm": 0.45159203892050004,
+      "learning_rate": 7.20443116766262e-06,
+      "loss": 0.4948,
+      "step": 9186
+    },
+    {
+      "epoch": 1.5087553630447723,
+      "grad_norm": 0.28798609716907286,
+      "learning_rate": 7.204056322011208e-06,
+      "loss": 0.4756,
+      "step": 9187
+    },
+    {
+      "epoch": 1.5089195902531152,
+      "grad_norm": 0.2880361608967198,
+      "learning_rate": 7.203681446991084e-06,
+      "loss": 0.4714,
+      "step": 9188
+    },
+    {
+      "epoch": 1.509083817461458,
+      "grad_norm": 0.5247905124145498,
+      "learning_rate": 7.2033065426063176e-06,
+      "loss": 0.4727,
+      "step": 9189
+    },
+    {
+      "epoch": 1.5092480446698007,
+      "grad_norm": 0.6860853909871801,
+      "learning_rate": 7.202931608860984e-06,
+      "loss": 0.4934,
+      "step": 9190
+    },
+    {
+      "epoch": 1.5094122718781433,
+      "grad_norm": 0.28275800142480045,
+      "learning_rate": 7.202556645759153e-06,
+      "loss": 0.4777,
+      "step": 9191
+    },
+    {
+      "epoch": 1.5095764990864862,
+      "grad_norm": 0.31027777276533725,
+      "learning_rate": 7.2021816533049e-06,
+      "loss": 0.4692,
+      "step": 9192
+    },
+    {
+      "epoch": 1.509740726294829,
+      "grad_norm": 0.3172137811682593,
+      "learning_rate": 7.2018066315022925e-06,
+      "loss": 0.4735,
+      "step": 9193
+    },
+    {
+      "epoch": 1.5099049535031717,
+      "grad_norm": 0.3372622416767625,
+      "learning_rate": 7.201431580355408e-06,
+      "loss": 0.4537,
+      "step": 9194
+    },
+    {
+      "epoch": 1.5100691807115143,
+      "grad_norm": 0.31133923346230863,
+      "learning_rate": 7.201056499868319e-06,
+      "loss": 0.4743,
+      "step": 9195
+    },
+    {
+      "epoch": 1.5102334079198572,
+      "grad_norm": 0.3781581429663542,
+      "learning_rate": 7.2006813900451e-06,
+      "loss": 0.4559,
+      "step": 9196
+    },
+    {
+      "epoch": 1.5103976351282,
+      "grad_norm": 0.30980865474592634,
+      "learning_rate": 7.2003062508898224e-06,
+      "loss": 0.471,
+      "step": 9197
+    },
+    {
+      "epoch": 1.5105618623365427,
+      "grad_norm": 0.3164800874921807,
+      "learning_rate": 7.1999310824065624e-06,
+      "loss": 0.4684,
+      "step": 9198
+    },
+    {
+      "epoch": 1.5107260895448853,
+      "grad_norm": 0.33008071514670867,
+      "learning_rate": 7.199555884599395e-06,
+      "loss": 0.4589,
+      "step": 9199
+    },
+    {
+      "epoch": 1.510890316753228,
+      "grad_norm": 0.3651681256005668,
+      "learning_rate": 7.199180657472395e-06,
+      "loss": 0.4586,
+      "step": 9200
+    },
+    {
+      "epoch": 1.5110545439615708,
+      "grad_norm": 0.3513853825676322,
+      "learning_rate": 7.198805401029636e-06,
+      "loss": 0.4498,
+      "step": 9201
+    },
+    {
+      "epoch": 1.5112187711699137,
+      "grad_norm": 0.3528677624533436,
+      "learning_rate": 7.1984301152751956e-06,
+      "loss": 0.4606,
+      "step": 9202
+    },
+    {
+      "epoch": 1.5113829983782563,
+      "grad_norm": 0.4279584123728494,
+      "learning_rate": 7.198054800213151e-06,
+      "loss": 0.452,
+      "step": 9203
+    },
+    {
+      "epoch": 1.511547225586599,
+      "grad_norm": 0.2921779445976517,
+      "learning_rate": 7.1976794558475745e-06,
+      "loss": 0.4739,
+      "step": 9204
+    },
+    {
+      "epoch": 1.5117114527949418,
+      "grad_norm": 0.2706453446276023,
+      "learning_rate": 7.1973040821825465e-06,
+      "loss": 0.4623,
+      "step": 9205
+    },
+    {
+      "epoch": 1.5118756800032846,
+      "grad_norm": 0.2520182801619176,
+      "learning_rate": 7.196928679222141e-06,
+      "loss": 0.4442,
+      "step": 9206
+    },
+    {
+      "epoch": 1.5120399072116273,
+      "grad_norm": 0.29407078062875935,
+      "learning_rate": 7.196553246970438e-06,
+      "loss": 0.4711,
+      "step": 9207
+    },
+    {
+      "epoch": 1.51220413441997,
+      "grad_norm": 0.3291641852823142,
+      "learning_rate": 7.196177785431513e-06,
+      "loss": 0.4488,
+      "step": 9208
+    },
+    {
+      "epoch": 1.5123683616283128,
+      "grad_norm": 0.314166634713391,
+      "learning_rate": 7.195802294609444e-06,
+      "loss": 0.458,
+      "step": 9209
+    },
+    {
+      "epoch": 1.5125325888366556,
+      "grad_norm": 0.30014099078366124,
+      "learning_rate": 7.195426774508311e-06,
+      "loss": 0.4725,
+      "step": 9210
+    },
+    {
+      "epoch": 1.5126968160449983,
+      "grad_norm": 0.28804167027539607,
+      "learning_rate": 7.19505122513219e-06,
+      "loss": 0.4813,
+      "step": 9211
+    },
+    {
+      "epoch": 1.512861043253341,
+      "grad_norm": 0.27774905359300783,
+      "learning_rate": 7.194675646485161e-06,
+      "loss": 0.4516,
+      "step": 9212
+    },
+    {
+      "epoch": 1.5130252704616838,
+      "grad_norm": 0.3376729483173663,
+      "learning_rate": 7.194300038571305e-06,
+      "loss": 0.4664,
+      "step": 9213
+    },
+    {
+      "epoch": 1.5131894976700266,
+      "grad_norm": 0.2990700174753199,
+      "learning_rate": 7.193924401394699e-06,
+      "loss": 0.4613,
+      "step": 9214
+    },
+    {
+      "epoch": 1.5133537248783693,
+      "grad_norm": 0.2570384432016577,
+      "learning_rate": 7.193548734959423e-06,
+      "loss": 0.4556,
+      "step": 9215
+    },
+    {
+      "epoch": 1.513517952086712,
+      "grad_norm": 0.3070126716863017,
+      "learning_rate": 7.193173039269558e-06,
+      "loss": 0.464,
+      "step": 9216
+    },
+    {
+      "epoch": 1.5136821792950546,
+      "grad_norm": 0.29290536910157305,
+      "learning_rate": 7.1927973143291835e-06,
+      "loss": 0.4396,
+      "step": 9217
+    },
+    {
+      "epoch": 1.5138464065033974,
+      "grad_norm": 0.3581538018351269,
+      "learning_rate": 7.192421560142382e-06,
+      "loss": 0.4615,
+      "step": 9218
+    },
+    {
+      "epoch": 1.5140106337117403,
+      "grad_norm": 0.36202199207173685,
+      "learning_rate": 7.192045776713232e-06,
+      "loss": 0.4596,
+      "step": 9219
+    },
+    {
+      "epoch": 1.514174860920083,
+      "grad_norm": 0.29895826521978897,
+      "learning_rate": 7.191669964045818e-06,
+      "loss": 0.466,
+      "step": 9220
+    },
+    {
+      "epoch": 1.5143390881284255,
+      "grad_norm": 0.3183924651940098,
+      "learning_rate": 7.191294122144217e-06,
+      "loss": 0.48,
+      "step": 9221
+    },
+    {
+      "epoch": 1.5145033153367684,
+      "grad_norm": 0.28198083034908455,
+      "learning_rate": 7.190918251012517e-06,
+      "loss": 0.4663,
+      "step": 9222
+    },
+    {
+      "epoch": 1.5146675425451113,
+      "grad_norm": 0.3303474365929383,
+      "learning_rate": 7.190542350654796e-06,
+      "loss": 0.4923,
+      "step": 9223
+    },
+    {
+      "epoch": 1.514831769753454,
+      "grad_norm": 0.4024297610145994,
+      "learning_rate": 7.190166421075138e-06,
+      "loss": 0.4718,
+      "step": 9224
+    },
+    {
+      "epoch": 1.5149959969617965,
+      "grad_norm": 0.25837992364309836,
+      "learning_rate": 7.189790462277626e-06,
+      "loss": 0.4564,
+      "step": 9225
+    },
+    {
+      "epoch": 1.5151602241701394,
+      "grad_norm": 0.39812967678782274,
+      "learning_rate": 7.1894144742663435e-06,
+      "loss": 0.4519,
+      "step": 9226
+    },
+    {
+      "epoch": 1.5153244513784823,
+      "grad_norm": 0.3450635706066045,
+      "learning_rate": 7.189038457045372e-06,
+      "loss": 0.453,
+      "step": 9227
+    },
+    {
+      "epoch": 1.515488678586825,
+      "grad_norm": 0.2783889625191982,
+      "learning_rate": 7.1886624106188e-06,
+      "loss": 0.475,
+      "step": 9228
+    },
+    {
+      "epoch": 1.5156529057951675,
+      "grad_norm": 0.3376557375783973,
+      "learning_rate": 7.1882863349907076e-06,
+      "loss": 0.4526,
+      "step": 9229
+    },
+    {
+      "epoch": 1.5158171330035104,
+      "grad_norm": 0.2896971589896287,
+      "learning_rate": 7.187910230165181e-06,
+      "loss": 0.4553,
+      "step": 9230
+    },
+    {
+      "epoch": 1.5159813602118533,
+      "grad_norm": 0.3157032612136202,
+      "learning_rate": 7.187534096146304e-06,
+      "loss": 0.472,
+      "step": 9231
+    },
+    {
+      "epoch": 1.516145587420196,
+      "grad_norm": 0.30994446974518053,
+      "learning_rate": 7.1871579329381625e-06,
+      "loss": 0.4639,
+      "step": 9232
+    },
+    {
+      "epoch": 1.5163098146285385,
+      "grad_norm": 0.33537568493394626,
+      "learning_rate": 7.186781740544842e-06,
+      "loss": 0.4527,
+      "step": 9233
+    },
+    {
+      "epoch": 1.5164740418368812,
+      "grad_norm": 0.3319121755581918,
+      "learning_rate": 7.18640551897043e-06,
+      "loss": 0.4741,
+      "step": 9234
+    },
+    {
+      "epoch": 1.516638269045224,
+      "grad_norm": 0.29375157989929307,
+      "learning_rate": 7.18602926821901e-06,
+      "loss": 0.4805,
+      "step": 9235
+    },
+    {
+      "epoch": 1.516802496253567,
+      "grad_norm": 0.278719060087101,
+      "learning_rate": 7.185652988294668e-06,
+      "loss": 0.459,
+      "step": 9236
+    },
+    {
+      "epoch": 1.5169667234619095,
+      "grad_norm": 0.28645821773642316,
+      "learning_rate": 7.185276679201494e-06,
+      "loss": 0.4561,
+      "step": 9237
+    },
+    {
+      "epoch": 1.5171309506702522,
+      "grad_norm": 0.3963936354834165,
+      "learning_rate": 7.184900340943574e-06,
+      "loss": 0.4494,
+      "step": 9238
+    },
+    {
+      "epoch": 1.517295177878595,
+      "grad_norm": 0.26055817683213023,
+      "learning_rate": 7.184523973524993e-06,
+      "loss": 0.4488,
+      "step": 9239
+    },
+    {
+      "epoch": 1.5174594050869379,
+      "grad_norm": 0.31956318401914047,
+      "learning_rate": 7.184147576949841e-06,
+      "loss": 0.4846,
+      "step": 9240
+    },
+    {
+      "epoch": 1.5176236322952805,
+      "grad_norm": 0.6295143635584098,
+      "learning_rate": 7.183771151222205e-06,
+      "loss": 0.4712,
+      "step": 9241
+    },
+    {
+      "epoch": 1.5177878595036232,
+      "grad_norm": 0.35754560129622853,
+      "learning_rate": 7.183394696346175e-06,
+      "loss": 0.4528,
+      "step": 9242
+    },
+    {
+      "epoch": 1.517952086711966,
+      "grad_norm": 0.44717398192916935,
+      "learning_rate": 7.1830182123258376e-06,
+      "loss": 0.4685,
+      "step": 9243
+    },
+    {
+      "epoch": 1.5181163139203089,
+      "grad_norm": 0.44012743422950795,
+      "learning_rate": 7.1826416991652835e-06,
+      "loss": 0.4705,
+      "step": 9244
+    },
+    {
+      "epoch": 1.5182805411286515,
+      "grad_norm": 0.26859917683990014,
+      "learning_rate": 7.182265156868599e-06,
+      "loss": 0.4438,
+      "step": 9245
+    },
+    {
+      "epoch": 1.5184447683369942,
+      "grad_norm": 0.28892403464626515,
+      "learning_rate": 7.181888585439879e-06,
+      "loss": 0.4714,
+      "step": 9246
+    },
+    {
+      "epoch": 1.518608995545337,
+      "grad_norm": 0.38304813283172645,
+      "learning_rate": 7.181511984883208e-06,
+      "loss": 0.4867,
+      "step": 9247
+    },
+    {
+      "epoch": 1.5187732227536799,
+      "grad_norm": 0.30305384307379923,
+      "learning_rate": 7.181135355202679e-06,
+      "loss": 0.4791,
+      "step": 9248
+    },
+    {
+      "epoch": 1.5189374499620225,
+      "grad_norm": 0.3155490544939365,
+      "learning_rate": 7.180758696402382e-06,
+      "loss": 0.4766,
+      "step": 9249
+    },
+    {
+      "epoch": 1.5191016771703652,
+      "grad_norm": 0.2906595932861458,
+      "learning_rate": 7.180382008486409e-06,
+      "loss": 0.4532,
+      "step": 9250
+    },
+    {
+      "epoch": 1.5192659043787078,
+      "grad_norm": 0.4474126938917336,
+      "learning_rate": 7.18000529145885e-06,
+      "loss": 0.4571,
+      "step": 9251
+    },
+    {
+      "epoch": 1.5194301315870506,
+      "grad_norm": 0.31794116167934067,
+      "learning_rate": 7.179628545323797e-06,
+      "loss": 0.4786,
+      "step": 9252
+    },
+    {
+      "epoch": 1.5195943587953935,
+      "grad_norm": 0.3263806739083232,
+      "learning_rate": 7.1792517700853405e-06,
+      "loss": 0.4735,
+      "step": 9253
+    },
+    {
+      "epoch": 1.5197585860037361,
+      "grad_norm": 0.3827263780224837,
+      "learning_rate": 7.178874965747575e-06,
+      "loss": 0.4625,
+      "step": 9254
+    },
+    {
+      "epoch": 1.5199228132120788,
+      "grad_norm": 0.3192602339087632,
+      "learning_rate": 7.178498132314591e-06,
+      "loss": 0.4676,
+      "step": 9255
+    },
+    {
+      "epoch": 1.5200870404204216,
+      "grad_norm": 0.39440388028059575,
+      "learning_rate": 7.1781212697904815e-06,
+      "loss": 0.4611,
+      "step": 9256
+    },
+    {
+      "epoch": 1.5202512676287645,
+      "grad_norm": 0.32399627276062054,
+      "learning_rate": 7.177744378179342e-06,
+      "loss": 0.4473,
+      "step": 9257
+    },
+    {
+      "epoch": 1.5204154948371071,
+      "grad_norm": 0.2790344377044542,
+      "learning_rate": 7.177367457485262e-06,
+      "loss": 0.4805,
+      "step": 9258
+    },
+    {
+      "epoch": 1.5205797220454498,
+      "grad_norm": 0.3458950067863732,
+      "learning_rate": 7.176990507712338e-06,
+      "loss": 0.4459,
+      "step": 9259
+    },
+    {
+      "epoch": 1.5207439492537926,
+      "grad_norm": 0.3416751416590733,
+      "learning_rate": 7.176613528864664e-06,
+      "loss": 0.4549,
+      "step": 9260
+    },
+    {
+      "epoch": 1.5209081764621355,
+      "grad_norm": 0.3240434085579008,
+      "learning_rate": 7.176236520946333e-06,
+      "loss": 0.4541,
+      "step": 9261
+    },
+    {
+      "epoch": 1.5210724036704781,
+      "grad_norm": 0.28624661378771976,
+      "learning_rate": 7.175859483961441e-06,
+      "loss": 0.4687,
+      "step": 9262
+    },
+    {
+      "epoch": 1.5212366308788208,
+      "grad_norm": 0.31156116458197464,
+      "learning_rate": 7.175482417914081e-06,
+      "loss": 0.4775,
+      "step": 9263
+    },
+    {
+      "epoch": 1.5214008580871636,
+      "grad_norm": 0.3201213181079633,
+      "learning_rate": 7.1751053228083495e-06,
+      "loss": 0.4542,
+      "step": 9264
+    },
+    {
+      "epoch": 1.5215650852955065,
+      "grad_norm": 0.2896366480561736,
+      "learning_rate": 7.174728198648343e-06,
+      "loss": 0.4613,
+      "step": 9265
+    },
+    {
+      "epoch": 1.5217293125038491,
+      "grad_norm": 1.5454545712227965,
+      "learning_rate": 7.174351045438156e-06,
+      "loss": 0.4528,
+      "step": 9266
+    },
+    {
+      "epoch": 1.5218935397121918,
+      "grad_norm": 0.4130809329340727,
+      "learning_rate": 7.173973863181886e-06,
+      "loss": 0.4612,
+      "step": 9267
+    },
+    {
+      "epoch": 1.5220577669205344,
+      "grad_norm": 0.5392643599819068,
+      "learning_rate": 7.173596651883629e-06,
+      "loss": 0.4734,
+      "step": 9268
+    },
+    {
+      "epoch": 1.5222219941288773,
+      "grad_norm": 0.29554384020455526,
+      "learning_rate": 7.173219411547483e-06,
+      "loss": 0.4572,
+      "step": 9269
+    },
+    {
+      "epoch": 1.5223862213372201,
+      "grad_norm": 0.37929201827686776,
+      "learning_rate": 7.172842142177543e-06,
+      "loss": 0.4761,
+      "step": 9270
+    },
+    {
+      "epoch": 1.5225504485455628,
+      "grad_norm": 0.27942306578844844,
+      "learning_rate": 7.172464843777907e-06,
+      "loss": 0.4546,
+      "step": 9271
+    },
+    {
+      "epoch": 1.5227146757539054,
+      "grad_norm": 0.31283795812998494,
+      "learning_rate": 7.172087516352674e-06,
+      "loss": 0.4632,
+      "step": 9272
+    },
+    {
+      "epoch": 1.5228789029622483,
+      "grad_norm": 0.28612405394145773,
+      "learning_rate": 7.171710159905943e-06,
+      "loss": 0.4581,
+      "step": 9273
+    },
+    {
+      "epoch": 1.5230431301705911,
+      "grad_norm": 0.4814978265268251,
+      "learning_rate": 7.171332774441809e-06,
+      "loss": 0.4654,
+      "step": 9274
+    },
+    {
+      "epoch": 1.5232073573789338,
+      "grad_norm": 0.3974772117244747,
+      "learning_rate": 7.170955359964373e-06,
+      "loss": 0.4625,
+      "step": 9275
+    },
+    {
+      "epoch": 1.5233715845872764,
+      "grad_norm": 0.3978478495668505,
+      "learning_rate": 7.170577916477736e-06,
+      "loss": 0.4508,
+      "step": 9276
+    },
+    {
+      "epoch": 1.5235358117956193,
+      "grad_norm": 0.47194095448274265,
+      "learning_rate": 7.170200443985993e-06,
+      "loss": 0.4553,
+      "step": 9277
+    },
+    {
+      "epoch": 1.5237000390039621,
+      "grad_norm": 0.37774460441156743,
+      "learning_rate": 7.1698229424932476e-06,
+      "loss": 0.4612,
+      "step": 9278
+    },
+    {
+      "epoch": 1.5238642662123048,
+      "grad_norm": 0.36763814623347385,
+      "learning_rate": 7.169445412003598e-06,
+      "loss": 0.4604,
+      "step": 9279
+    },
+    {
+      "epoch": 1.5240284934206474,
+      "grad_norm": 0.5037516746821502,
+      "learning_rate": 7.169067852521144e-06,
+      "loss": 0.478,
+      "step": 9280
+    },
+    {
+      "epoch": 1.5241927206289902,
+      "grad_norm": 0.4212263940251124,
+      "learning_rate": 7.1686902640499876e-06,
+      "loss": 0.4634,
+      "step": 9281
+    },
+    {
+      "epoch": 1.524356947837333,
+      "grad_norm": 0.639577271055163,
+      "learning_rate": 7.168312646594228e-06,
+      "loss": 0.4659,
+      "step": 9282
+    },
+    {
+      "epoch": 1.5245211750456757,
+      "grad_norm": 0.31425965396324984,
+      "learning_rate": 7.16793500015797e-06,
+      "loss": 0.458,
+      "step": 9283
+    },
+    {
+      "epoch": 1.5246854022540184,
+      "grad_norm": 0.3549092300339072,
+      "learning_rate": 7.167557324745312e-06,
+      "loss": 0.4847,
+      "step": 9284
+    },
+    {
+      "epoch": 1.524849629462361,
+      "grad_norm": 0.464940002037064,
+      "learning_rate": 7.167179620360357e-06,
+      "loss": 0.4851,
+      "step": 9285
+    },
+    {
+      "epoch": 1.5250138566707039,
+      "grad_norm": 0.7404536442112363,
+      "learning_rate": 7.166801887007208e-06,
+      "loss": 0.4645,
+      "step": 9286
+    },
+    {
+      "epoch": 1.5251780838790467,
+      "grad_norm": 0.3797169402295166,
+      "learning_rate": 7.166424124689965e-06,
+      "loss": 0.4729,
+      "step": 9287
+    },
+    {
+      "epoch": 1.5253423110873894,
+      "grad_norm": 0.4800510956035743,
+      "learning_rate": 7.1660463334127345e-06,
+      "loss": 0.473,
+      "step": 9288
+    },
+    {
+      "epoch": 1.525506538295732,
+      "grad_norm": 0.29060595534750716,
+      "learning_rate": 7.165668513179617e-06,
+      "loss": 0.4534,
+      "step": 9289
+    },
+    {
+      "epoch": 1.5256707655040749,
+      "grad_norm": 0.30087575863625227,
+      "learning_rate": 7.165290663994715e-06,
+      "loss": 0.4956,
+      "step": 9290
+    },
+    {
+      "epoch": 1.5258349927124177,
+      "grad_norm": 0.339029574512295,
+      "learning_rate": 7.1649127858621354e-06,
+      "loss": 0.4774,
+      "step": 9291
+    },
+    {
+      "epoch": 1.5259992199207604,
+      "grad_norm": 0.2916719047991023,
+      "learning_rate": 7.164534878785982e-06,
+      "loss": 0.4867,
+      "step": 9292
+    },
+    {
+      "epoch": 1.526163447129103,
+      "grad_norm": 0.30316356818056006,
+      "learning_rate": 7.1641569427703585e-06,
+      "loss": 0.4491,
+      "step": 9293
+    },
+    {
+      "epoch": 1.5263276743374459,
+      "grad_norm": 0.34009506124918254,
+      "learning_rate": 7.163778977819368e-06,
+      "loss": 0.4506,
+      "step": 9294
+    },
+    {
+      "epoch": 1.5264919015457887,
+      "grad_norm": 0.3158249532017147,
+      "learning_rate": 7.163400983937117e-06,
+      "loss": 0.4728,
+      "step": 9295
+    },
+    {
+      "epoch": 1.5266561287541314,
+      "grad_norm": 0.31123665592368355,
+      "learning_rate": 7.163022961127711e-06,
+      "loss": 0.4595,
+      "step": 9296
+    },
+    {
+      "epoch": 1.526820355962474,
+      "grad_norm": 0.3281051407634187,
+      "learning_rate": 7.162644909395256e-06,
+      "loss": 0.4652,
+      "step": 9297
+    },
+    {
+      "epoch": 1.5269845831708169,
+      "grad_norm": 0.29028559973315327,
+      "learning_rate": 7.162266828743857e-06,
+      "loss": 0.469,
+      "step": 9298
+    },
+    {
+      "epoch": 1.5271488103791597,
+      "grad_norm": 0.26500384107962766,
+      "learning_rate": 7.161888719177622e-06,
+      "loss": 0.4717,
+      "step": 9299
+    },
+    {
+      "epoch": 1.5273130375875024,
+      "grad_norm": 0.3146412183956378,
+      "learning_rate": 7.161510580700656e-06,
+      "loss": 0.4669,
+      "step": 9300
+    },
+    {
+      "epoch": 1.527477264795845,
+      "grad_norm": 0.30019999592266705,
+      "learning_rate": 7.161132413317068e-06,
+      "loss": 0.4657,
+      "step": 9301
+    },
+    {
+      "epoch": 1.5276414920041876,
+      "grad_norm": 0.28169330642062956,
+      "learning_rate": 7.160754217030962e-06,
+      "loss": 0.4525,
+      "step": 9302
+    },
+    {
+      "epoch": 1.5278057192125305,
+      "grad_norm": 0.5323783758611011,
+      "learning_rate": 7.1603759918464476e-06,
+      "loss": 0.4877,
+      "step": 9303
+    },
+    {
+      "epoch": 1.5279699464208734,
+      "grad_norm": 0.2690038165753098,
+      "learning_rate": 7.159997737767632e-06,
+      "loss": 0.4695,
+      "step": 9304
+    },
+    {
+      "epoch": 1.528134173629216,
+      "grad_norm": 0.709221517163938,
+      "learning_rate": 7.159619454798625e-06,
+      "loss": 0.4505,
+      "step": 9305
+    },
+    {
+      "epoch": 1.5282984008375586,
+      "grad_norm": 0.26711462113418744,
+      "learning_rate": 7.159241142943533e-06,
+      "loss": 0.4658,
+      "step": 9306
+    },
+    {
+      "epoch": 1.5284626280459015,
+      "grad_norm": 0.3039516845191958,
+      "learning_rate": 7.158862802206466e-06,
+      "loss": 0.4671,
+      "step": 9307
+    },
+    {
+      "epoch": 1.5286268552542444,
+      "grad_norm": 0.2631380237679185,
+      "learning_rate": 7.158484432591534e-06,
+      "loss": 0.457,
+      "step": 9308
+    },
+    {
+      "epoch": 1.528791082462587,
+      "grad_norm": 0.34880055704945134,
+      "learning_rate": 7.158106034102846e-06,
+      "loss": 0.459,
+      "step": 9309
+    },
+    {
+      "epoch": 1.5289553096709296,
+      "grad_norm": 0.4332345447097457,
+      "learning_rate": 7.1577276067445094e-06,
+      "loss": 0.481,
+      "step": 9310
+    },
+    {
+      "epoch": 1.5291195368792725,
+      "grad_norm": 0.2701428254345847,
+      "learning_rate": 7.157349150520636e-06,
+      "loss": 0.464,
+      "step": 9311
+    },
+    {
+      "epoch": 1.5292837640876153,
+      "grad_norm": 0.3191308013759308,
+      "learning_rate": 7.156970665435338e-06,
+      "loss": 0.461,
+      "step": 9312
+    },
+    {
+      "epoch": 1.529447991295958,
+      "grad_norm": 0.3030079025076875,
+      "learning_rate": 7.156592151492722e-06,
+      "loss": 0.4735,
+      "step": 9313
+    },
+    {
+      "epoch": 1.5296122185043006,
+      "grad_norm": 0.33155877053915367,
+      "learning_rate": 7.156213608696904e-06,
+      "loss": 0.5012,
+      "step": 9314
+    },
+    {
+      "epoch": 1.5297764457126435,
+      "grad_norm": 0.3130320047585034,
+      "learning_rate": 7.155835037051993e-06,
+      "loss": 0.4917,
+      "step": 9315
+    },
+    {
+      "epoch": 1.5299406729209863,
+      "grad_norm": 0.36060204593966044,
+      "learning_rate": 7.155456436562098e-06,
+      "loss": 0.4803,
+      "step": 9316
+    },
+    {
+      "epoch": 1.530104900129329,
+      "grad_norm": 0.32398573562386035,
+      "learning_rate": 7.155077807231336e-06,
+      "loss": 0.4495,
+      "step": 9317
+    },
+    {
+      "epoch": 1.5302691273376716,
+      "grad_norm": 0.27751479721468064,
+      "learning_rate": 7.154699149063816e-06,
+      "loss": 0.4826,
+      "step": 9318
+    },
+    {
+      "epoch": 1.5304333545460143,
+      "grad_norm": 0.42071383808180096,
+      "learning_rate": 7.1543204620636505e-06,
+      "loss": 0.4529,
+      "step": 9319
+    },
+    {
+      "epoch": 1.5305975817543571,
+      "grad_norm": 0.44110763680049825,
+      "learning_rate": 7.153941746234953e-06,
+      "loss": 0.4987,
+      "step": 9320
+    },
+    {
+      "epoch": 1.5307618089627,
+      "grad_norm": 0.289458468189357,
+      "learning_rate": 7.153563001581838e-06,
+      "loss": 0.4733,
+      "step": 9321
+    },
+    {
+      "epoch": 1.5309260361710426,
+      "grad_norm": 0.2512213588331616,
+      "learning_rate": 7.153184228108419e-06,
+      "loss": 0.4549,
+      "step": 9322
+    },
+    {
+      "epoch": 1.5310902633793853,
+      "grad_norm": 0.3004273617854931,
+      "learning_rate": 7.152805425818807e-06,
+      "loss": 0.454,
+      "step": 9323
+    },
+    {
+      "epoch": 1.5312544905877281,
+      "grad_norm": 0.32267924554730515,
+      "learning_rate": 7.152426594717119e-06,
+      "loss": 0.4796,
+      "step": 9324
+    },
+    {
+      "epoch": 1.531418717796071,
+      "grad_norm": 0.2957664116913034,
+      "learning_rate": 7.152047734807469e-06,
+      "loss": 0.4688,
+      "step": 9325
+    },
+    {
+      "epoch": 1.5315829450044136,
+      "grad_norm": 0.29607329025830087,
+      "learning_rate": 7.151668846093971e-06,
+      "loss": 0.4735,
+      "step": 9326
+    },
+    {
+      "epoch": 1.5317471722127562,
+      "grad_norm": 0.34564464756084806,
+      "learning_rate": 7.15128992858074e-06,
+      "loss": 0.4746,
+      "step": 9327
+    },
+    {
+      "epoch": 1.531911399421099,
+      "grad_norm": 0.3632349137268046,
+      "learning_rate": 7.1509109822718915e-06,
+      "loss": 0.4638,
+      "step": 9328
+    },
+    {
+      "epoch": 1.532075626629442,
+      "grad_norm": 0.3528582463422724,
+      "learning_rate": 7.150532007171542e-06,
+      "loss": 0.4739,
+      "step": 9329
+    },
+    {
+      "epoch": 1.5322398538377846,
+      "grad_norm": 0.2851865961310054,
+      "learning_rate": 7.150153003283807e-06,
+      "loss": 0.4625,
+      "step": 9330
+    },
+    {
+      "epoch": 1.5324040810461272,
+      "grad_norm": 0.2644495220903839,
+      "learning_rate": 7.149773970612804e-06,
+      "loss": 0.4744,
+      "step": 9331
+    },
+    {
+      "epoch": 1.53256830825447,
+      "grad_norm": 0.26070085543493976,
+      "learning_rate": 7.149394909162648e-06,
+      "loss": 0.4608,
+      "step": 9332
+    },
+    {
+      "epoch": 1.532732535462813,
+      "grad_norm": 0.29613010697864284,
+      "learning_rate": 7.149015818937456e-06,
+      "loss": 0.4521,
+      "step": 9333
+    },
+    {
+      "epoch": 1.5328967626711556,
+      "grad_norm": 0.31117686961305685,
+      "learning_rate": 7.148636699941347e-06,
+      "loss": 0.4644,
+      "step": 9334
+    },
+    {
+      "epoch": 1.5330609898794982,
+      "grad_norm": 0.29211819423313984,
+      "learning_rate": 7.148257552178438e-06,
+      "loss": 0.4821,
+      "step": 9335
+    },
+    {
+      "epoch": 1.5332252170878409,
+      "grad_norm": 0.2703181498755695,
+      "learning_rate": 7.147878375652844e-06,
+      "loss": 0.4721,
+      "step": 9336
+    },
+    {
+      "epoch": 1.5333894442961837,
+      "grad_norm": 0.33278351582547366,
+      "learning_rate": 7.147499170368688e-06,
+      "loss": 0.4536,
+      "step": 9337
+    },
+    {
+      "epoch": 1.5335536715045266,
+      "grad_norm": 0.28749069490029466,
+      "learning_rate": 7.1471199363300845e-06,
+      "loss": 0.4548,
+      "step": 9338
+    },
+    {
+      "epoch": 1.5337178987128692,
+      "grad_norm": 0.29073060662806555,
+      "learning_rate": 7.146740673541155e-06,
+      "loss": 0.4615,
+      "step": 9339
+    },
+    {
+      "epoch": 1.5338821259212119,
+      "grad_norm": 0.4272828402384313,
+      "learning_rate": 7.146361382006019e-06,
+      "loss": 0.4344,
+      "step": 9340
+    },
+    {
+      "epoch": 1.5340463531295547,
+      "grad_norm": 0.28147195710433215,
+      "learning_rate": 7.145982061728792e-06,
+      "loss": 0.4752,
+      "step": 9341
+    },
+    {
+      "epoch": 1.5342105803378976,
+      "grad_norm": 0.3439436486971339,
+      "learning_rate": 7.145602712713598e-06,
+      "loss": 0.4666,
+      "step": 9342
+    },
+    {
+      "epoch": 1.5343748075462402,
+      "grad_norm": 0.3488722878311697,
+      "learning_rate": 7.145223334964556e-06,
+      "loss": 0.4615,
+      "step": 9343
+    },
+    {
+      "epoch": 1.5345390347545829,
+      "grad_norm": 0.30250466922681474,
+      "learning_rate": 7.144843928485786e-06,
+      "loss": 0.4663,
+      "step": 9344
+    },
+    {
+      "epoch": 1.5347032619629257,
+      "grad_norm": 0.2913076112847636,
+      "learning_rate": 7.144464493281407e-06,
+      "loss": 0.493,
+      "step": 9345
+    },
+    {
+      "epoch": 1.5348674891712686,
+      "grad_norm": 0.27574286461367503,
+      "learning_rate": 7.144085029355544e-06,
+      "loss": 0.4568,
+      "step": 9346
+    },
+    {
+      "epoch": 1.5350317163796112,
+      "grad_norm": 0.3085587250021239,
+      "learning_rate": 7.143705536712316e-06,
+      "loss": 0.459,
+      "step": 9347
+    },
+    {
+      "epoch": 1.5351959435879539,
+      "grad_norm": 0.42885403672499844,
+      "learning_rate": 7.143326015355844e-06,
+      "loss": 0.4581,
+      "step": 9348
+    },
+    {
+      "epoch": 1.5353601707962967,
+      "grad_norm": 0.2740916414682097,
+      "learning_rate": 7.14294646529025e-06,
+      "loss": 0.4518,
+      "step": 9349
+    },
+    {
+      "epoch": 1.5355243980046396,
+      "grad_norm": 0.3450236813009234,
+      "learning_rate": 7.1425668865196585e-06,
+      "loss": 0.4523,
+      "step": 9350
+    },
+    {
+      "epoch": 1.5356886252129822,
+      "grad_norm": 0.38502119714992583,
+      "learning_rate": 7.14218727904819e-06,
+      "loss": 0.4633,
+      "step": 9351
+    },
+    {
+      "epoch": 1.5358528524213249,
+      "grad_norm": 0.4006288598907935,
+      "learning_rate": 7.1418076428799685e-06,
+      "loss": 0.4618,
+      "step": 9352
+    },
+    {
+      "epoch": 1.5360170796296675,
+      "grad_norm": 0.5293557699101507,
+      "learning_rate": 7.141427978019116e-06,
+      "loss": 0.4661,
+      "step": 9353
+    },
+    {
+      "epoch": 1.5361813068380104,
+      "grad_norm": 0.3574612680794133,
+      "learning_rate": 7.141048284469758e-06,
+      "loss": 0.4801,
+      "step": 9354
+    },
+    {
+      "epoch": 1.5363455340463532,
+      "grad_norm": 0.3274060576534963,
+      "learning_rate": 7.1406685622360174e-06,
+      "loss": 0.4723,
+      "step": 9355
+    },
+    {
+      "epoch": 1.5365097612546958,
+      "grad_norm": 0.3126087621576872,
+      "learning_rate": 7.140288811322017e-06,
+      "loss": 0.4491,
+      "step": 9356
+    },
+    {
+      "epoch": 1.5366739884630385,
+      "grad_norm": 0.43435643215596625,
+      "learning_rate": 7.139909031731883e-06,
+      "loss": 0.476,
+      "step": 9357
+    },
+    {
+      "epoch": 1.5368382156713813,
+      "grad_norm": 0.2922725077472286,
+      "learning_rate": 7.139529223469738e-06,
+      "loss": 0.4591,
+      "step": 9358
+    },
+    {
+      "epoch": 1.5370024428797242,
+      "grad_norm": 0.33144720651032605,
+      "learning_rate": 7.139149386539711e-06,
+      "loss": 0.4807,
+      "step": 9359
+    },
+    {
+      "epoch": 1.5371666700880668,
+      "grad_norm": 0.2945145824570666,
+      "learning_rate": 7.138769520945925e-06,
+      "loss": 0.4484,
+      "step": 9360
+    },
+    {
+      "epoch": 1.5373308972964095,
+      "grad_norm": 0.2699345270357702,
+      "learning_rate": 7.138389626692504e-06,
+      "loss": 0.4663,
+      "step": 9361
+    },
+    {
+      "epoch": 1.5374951245047523,
+      "grad_norm": 0.37423556894532223,
+      "learning_rate": 7.138009703783577e-06,
+      "loss": 0.4629,
+      "step": 9362
+    },
+    {
+      "epoch": 1.5376593517130952,
+      "grad_norm": 0.28643072132489733,
+      "learning_rate": 7.137629752223268e-06,
+      "loss": 0.4626,
+      "step": 9363
+    },
+    {
+      "epoch": 1.5378235789214378,
+      "grad_norm": 0.4011805145508984,
+      "learning_rate": 7.137249772015707e-06,
+      "loss": 0.4598,
+      "step": 9364
+    },
+    {
+      "epoch": 1.5379878061297805,
+      "grad_norm": 0.28589490123057115,
+      "learning_rate": 7.136869763165017e-06,
+      "loss": 0.4401,
+      "step": 9365
+    },
+    {
+      "epoch": 1.5381520333381233,
+      "grad_norm": 0.33402197609435524,
+      "learning_rate": 7.136489725675328e-06,
+      "loss": 0.4718,
+      "step": 9366
+    },
+    {
+      "epoch": 1.5383162605464662,
+      "grad_norm": 0.3515709813075105,
+      "learning_rate": 7.136109659550765e-06,
+      "loss": 0.4692,
+      "step": 9367
+    },
+    {
+      "epoch": 1.5384804877548088,
+      "grad_norm": 0.727042993615268,
+      "learning_rate": 7.13572956479546e-06,
+      "loss": 0.4434,
+      "step": 9368
+    },
+    {
+      "epoch": 1.5386447149631515,
+      "grad_norm": 0.2803236782587894,
+      "learning_rate": 7.135349441413538e-06,
+      "loss": 0.4727,
+      "step": 9369
+    },
+    {
+      "epoch": 1.538808942171494,
+      "grad_norm": 0.3226815325060115,
+      "learning_rate": 7.134969289409126e-06,
+      "loss": 0.4747,
+      "step": 9370
+    },
+    {
+      "epoch": 1.538973169379837,
+      "grad_norm": 0.36235437219598443,
+      "learning_rate": 7.134589108786357e-06,
+      "loss": 0.4468,
+      "step": 9371
+    },
+    {
+      "epoch": 1.5391373965881798,
+      "grad_norm": 0.35625948487457887,
+      "learning_rate": 7.134208899549359e-06,
+      "loss": 0.4486,
+      "step": 9372
+    },
+    {
+      "epoch": 1.5393016237965225,
+      "grad_norm": 0.34735670003053626,
+      "learning_rate": 7.133828661702259e-06,
+      "loss": 0.4335,
+      "step": 9373
+    },
+    {
+      "epoch": 1.539465851004865,
+      "grad_norm": 0.34199356720688007,
+      "learning_rate": 7.133448395249189e-06,
+      "loss": 0.4774,
+      "step": 9374
+    },
+    {
+      "epoch": 1.539630078213208,
+      "grad_norm": 0.3075669052642566,
+      "learning_rate": 7.133068100194278e-06,
+      "loss": 0.466,
+      "step": 9375
+    },
+    {
+      "epoch": 1.5397943054215508,
+      "grad_norm": 0.30933420404936507,
+      "learning_rate": 7.132687776541658e-06,
+      "loss": 0.4629,
+      "step": 9376
+    },
+    {
+      "epoch": 1.5399585326298935,
+      "grad_norm": 0.3167400643057456,
+      "learning_rate": 7.132307424295457e-06,
+      "loss": 0.4694,
+      "step": 9377
+    },
+    {
+      "epoch": 1.540122759838236,
+      "grad_norm": 0.3330410083602585,
+      "learning_rate": 7.1319270434598095e-06,
+      "loss": 0.4541,
+      "step": 9378
+    },
+    {
+      "epoch": 1.540286987046579,
+      "grad_norm": 0.2608004017915481,
+      "learning_rate": 7.131546634038843e-06,
+      "loss": 0.4609,
+      "step": 9379
+    },
+    {
+      "epoch": 1.5404512142549218,
+      "grad_norm": 0.32756145155673744,
+      "learning_rate": 7.131166196036692e-06,
+      "loss": 0.4679,
+      "step": 9380
+    },
+    {
+      "epoch": 1.5406154414632645,
+      "grad_norm": 0.3215737601475467,
+      "learning_rate": 7.130785729457487e-06,
+      "loss": 0.4712,
+      "step": 9381
+    },
+    {
+      "epoch": 1.540779668671607,
+      "grad_norm": 0.34381508752868223,
+      "learning_rate": 7.13040523430536e-06,
+      "loss": 0.4399,
+      "step": 9382
+    },
+    {
+      "epoch": 1.54094389587995,
+      "grad_norm": 0.3004485729292794,
+      "learning_rate": 7.1300247105844455e-06,
+      "loss": 0.4567,
+      "step": 9383
+    },
+    {
+      "epoch": 1.5411081230882928,
+      "grad_norm": 0.37942435111803424,
+      "learning_rate": 7.1296441582988745e-06,
+      "loss": 0.4677,
+      "step": 9384
+    },
+    {
+      "epoch": 1.5412723502966355,
+      "grad_norm": 0.4357473073673098,
+      "learning_rate": 7.129263577452781e-06,
+      "loss": 0.4518,
+      "step": 9385
+    },
+    {
+      "epoch": 1.541436577504978,
+      "grad_norm": 0.3087949525062071,
+      "learning_rate": 7.128882968050298e-06,
+      "loss": 0.4724,
+      "step": 9386
+    },
+    {
+      "epoch": 1.5416008047133207,
+      "grad_norm": 0.35014647578381863,
+      "learning_rate": 7.128502330095558e-06,
+      "loss": 0.46,
+      "step": 9387
+    },
+    {
+      "epoch": 1.5417650319216636,
+      "grad_norm": 0.3146885120825772,
+      "learning_rate": 7.1281216635926985e-06,
+      "loss": 0.4583,
+      "step": 9388
+    },
+    {
+      "epoch": 1.5419292591300064,
+      "grad_norm": 0.3136789279242204,
+      "learning_rate": 7.127740968545852e-06,
+      "loss": 0.4843,
+      "step": 9389
+    },
+    {
+      "epoch": 1.542093486338349,
+      "grad_norm": 0.5298025260222496,
+      "learning_rate": 7.127360244959151e-06,
+      "loss": 0.4812,
+      "step": 9390
+    },
+    {
+      "epoch": 1.5422577135466917,
+      "grad_norm": 0.3844808012012636,
+      "learning_rate": 7.126979492836736e-06,
+      "loss": 0.4544,
+      "step": 9391
+    },
+    {
+      "epoch": 1.5424219407550346,
+      "grad_norm": 0.39924791256605835,
+      "learning_rate": 7.126598712182736e-06,
+      "loss": 0.479,
+      "step": 9392
+    },
+    {
+      "epoch": 1.5425861679633774,
+      "grad_norm": 0.3602246778936246,
+      "learning_rate": 7.1262179030012925e-06,
+      "loss": 0.4783,
+      "step": 9393
+    },
+    {
+      "epoch": 1.54275039517172,
+      "grad_norm": 0.2732202036898396,
+      "learning_rate": 7.1258370652965375e-06,
+      "loss": 0.4785,
+      "step": 9394
+    },
+    {
+      "epoch": 1.5429146223800627,
+      "grad_norm": 0.3263140799564156,
+      "learning_rate": 7.125456199072609e-06,
+      "loss": 0.4563,
+      "step": 9395
+    },
+    {
+      "epoch": 1.5430788495884056,
+      "grad_norm": 0.3548333323535569,
+      "learning_rate": 7.125075304333642e-06,
+      "loss": 0.4517,
+      "step": 9396
+    },
+    {
+      "epoch": 1.5432430767967484,
+      "grad_norm": 0.8067372128015823,
+      "learning_rate": 7.1246943810837745e-06,
+      "loss": 0.453,
+      "step": 9397
+    },
+    {
+      "epoch": 1.543407304005091,
+      "grad_norm": 0.26504405806007353,
+      "learning_rate": 7.1243134293271445e-06,
+      "loss": 0.4349,
+      "step": 9398
+    },
+    {
+      "epoch": 1.5435715312134337,
+      "grad_norm": 0.45675103704189807,
+      "learning_rate": 7.123932449067888e-06,
+      "loss": 0.4476,
+      "step": 9399
+    },
+    {
+      "epoch": 1.5437357584217766,
+      "grad_norm": 0.32637847065326686,
+      "learning_rate": 7.123551440310144e-06,
+      "loss": 0.4862,
+      "step": 9400
+    },
+    {
+      "epoch": 1.5438999856301194,
+      "grad_norm": 0.33195374509870645,
+      "learning_rate": 7.1231704030580516e-06,
+      "loss": 0.4596,
+      "step": 9401
+    },
+    {
+      "epoch": 1.544064212838462,
+      "grad_norm": 0.255934132190599,
+      "learning_rate": 7.122789337315745e-06,
+      "loss": 0.4678,
+      "step": 9402
+    },
+    {
+      "epoch": 1.5442284400468047,
+      "grad_norm": 0.33693985909373964,
+      "learning_rate": 7.122408243087367e-06,
+      "loss": 0.4686,
+      "step": 9403
+    },
+    {
+      "epoch": 1.5443926672551473,
+      "grad_norm": 0.2996963732838922,
+      "learning_rate": 7.122027120377055e-06,
+      "loss": 0.4839,
+      "step": 9404
+    },
+    {
+      "epoch": 1.5445568944634902,
+      "grad_norm": 0.2773729203223255,
+      "learning_rate": 7.12164596918895e-06,
+      "loss": 0.4862,
+      "step": 9405
+    },
+    {
+      "epoch": 1.544721121671833,
+      "grad_norm": 0.25544315255098893,
+      "learning_rate": 7.121264789527189e-06,
+      "loss": 0.4691,
+      "step": 9406
+    },
+    {
+      "epoch": 1.5448853488801757,
+      "grad_norm": 0.3592750822464238,
+      "learning_rate": 7.120883581395914e-06,
+      "loss": 0.4467,
+      "step": 9407
+    },
+    {
+      "epoch": 1.5450495760885183,
+      "grad_norm": 0.28607035669715075,
+      "learning_rate": 7.120502344799264e-06,
+      "loss": 0.4544,
+      "step": 9408
+    },
+    {
+      "epoch": 1.5452138032968612,
+      "grad_norm": 0.25531993325233804,
+      "learning_rate": 7.120121079741381e-06,
+      "loss": 0.4449,
+      "step": 9409
+    },
+    {
+      "epoch": 1.545378030505204,
+      "grad_norm": 0.5343782428139733,
+      "learning_rate": 7.119739786226406e-06,
+      "loss": 0.473,
+      "step": 9410
+    },
+    {
+      "epoch": 1.5455422577135467,
+      "grad_norm": 0.2864075940659775,
+      "learning_rate": 7.119358464258478e-06,
+      "loss": 0.4694,
+      "step": 9411
+    },
+    {
+      "epoch": 1.5457064849218893,
+      "grad_norm": 0.2715392085229134,
+      "learning_rate": 7.118977113841741e-06,
+      "loss": 0.4486,
+      "step": 9412
+    },
+    {
+      "epoch": 1.5458707121302322,
+      "grad_norm": 0.3593589381329291,
+      "learning_rate": 7.1185957349803355e-06,
+      "loss": 0.4578,
+      "step": 9413
+    },
+    {
+      "epoch": 1.546034939338575,
+      "grad_norm": 0.3399272636588334,
+      "learning_rate": 7.118214327678404e-06,
+      "loss": 0.4822,
+      "step": 9414
+    },
+    {
+      "epoch": 1.5461991665469177,
+      "grad_norm": 0.3378526292093369,
+      "learning_rate": 7.1178328919400895e-06,
+      "loss": 0.459,
+      "step": 9415
+    },
+    {
+      "epoch": 1.5463633937552603,
+      "grad_norm": 0.29955207988397575,
+      "learning_rate": 7.117451427769532e-06,
+      "loss": 0.4605,
+      "step": 9416
+    },
+    {
+      "epoch": 1.5465276209636032,
+      "grad_norm": 0.31826406899690846,
+      "learning_rate": 7.117069935170879e-06,
+      "loss": 0.4502,
+      "step": 9417
+    },
+    {
+      "epoch": 1.546691848171946,
+      "grad_norm": 0.3419910247342894,
+      "learning_rate": 7.11668841414827e-06,
+      "loss": 0.481,
+      "step": 9418
+    },
+    {
+      "epoch": 1.5468560753802887,
+      "grad_norm": 0.3465516444944875,
+      "learning_rate": 7.1163068647058515e-06,
+      "loss": 0.4589,
+      "step": 9419
+    },
+    {
+      "epoch": 1.5470203025886313,
+      "grad_norm": 0.3299902597621926,
+      "learning_rate": 7.115925286847767e-06,
+      "loss": 0.4424,
+      "step": 9420
+    },
+    {
+      "epoch": 1.547184529796974,
+      "grad_norm": 0.30635057158537654,
+      "learning_rate": 7.115543680578159e-06,
+      "loss": 0.4842,
+      "step": 9421
+    },
+    {
+      "epoch": 1.5473487570053168,
+      "grad_norm": 0.34957654661682797,
+      "learning_rate": 7.115162045901173e-06,
+      "loss": 0.4825,
+      "step": 9422
+    },
+    {
+      "epoch": 1.5475129842136597,
+      "grad_norm": 0.2953423418282763,
+      "learning_rate": 7.114780382820955e-06,
+      "loss": 0.4719,
+      "step": 9423
+    },
+    {
+      "epoch": 1.5476772114220023,
+      "grad_norm": 0.3780817633359112,
+      "learning_rate": 7.1143986913416495e-06,
+      "loss": 0.4778,
+      "step": 9424
+    },
+    {
+      "epoch": 1.547841438630345,
+      "grad_norm": 0.34843838102503016,
+      "learning_rate": 7.114016971467401e-06,
+      "loss": 0.4526,
+      "step": 9425
+    },
+    {
+      "epoch": 1.5480056658386878,
+      "grad_norm": 0.355449539427888,
+      "learning_rate": 7.113635223202358e-06,
+      "loss": 0.4825,
+      "step": 9426
+    },
+    {
+      "epoch": 1.5481698930470307,
+      "grad_norm": 0.3234611083942094,
+      "learning_rate": 7.113253446550665e-06,
+      "loss": 0.4869,
+      "step": 9427
+    },
+    {
+      "epoch": 1.5483341202553733,
+      "grad_norm": 0.307306483199121,
+      "learning_rate": 7.112871641516466e-06,
+      "loss": 0.4694,
+      "step": 9428
+    },
+    {
+      "epoch": 1.548498347463716,
+      "grad_norm": 0.30794715743718676,
+      "learning_rate": 7.112489808103912e-06,
+      "loss": 0.4873,
+      "step": 9429
+    },
+    {
+      "epoch": 1.5486625746720588,
+      "grad_norm": 0.47467914042047893,
+      "learning_rate": 7.1121079463171485e-06,
+      "loss": 0.475,
+      "step": 9430
+    },
+    {
+      "epoch": 1.5488268018804017,
+      "grad_norm": 0.29763731599021,
+      "learning_rate": 7.111726056160322e-06,
+      "loss": 0.4722,
+      "step": 9431
+    },
+    {
+      "epoch": 1.5489910290887443,
+      "grad_norm": 0.5091962443448468,
+      "learning_rate": 7.1113441376375815e-06,
+      "loss": 0.4707,
+      "step": 9432
+    },
+    {
+      "epoch": 1.549155256297087,
+      "grad_norm": 0.37582684669527816,
+      "learning_rate": 7.110962190753074e-06,
+      "loss": 0.4578,
+      "step": 9433
+    },
+    {
+      "epoch": 1.5493194835054298,
+      "grad_norm": 0.28575068074639237,
+      "learning_rate": 7.110580215510948e-06,
+      "loss": 0.4823,
+      "step": 9434
+    },
+    {
+      "epoch": 1.5494837107137727,
+      "grad_norm": 0.2629592017210659,
+      "learning_rate": 7.1101982119153535e-06,
+      "loss": 0.4744,
+      "step": 9435
+    },
+    {
+      "epoch": 1.5496479379221153,
+      "grad_norm": 0.30478574127256014,
+      "learning_rate": 7.109816179970438e-06,
+      "loss": 0.4651,
+      "step": 9436
+    },
+    {
+      "epoch": 1.549812165130458,
+      "grad_norm": 0.30396636516802095,
+      "learning_rate": 7.10943411968035e-06,
+      "loss": 0.4578,
+      "step": 9437
+    },
+    {
+      "epoch": 1.5499763923388006,
+      "grad_norm": 0.34806655997770647,
+      "learning_rate": 7.109052031049241e-06,
+      "loss": 0.451,
+      "step": 9438
+    },
+    {
+      "epoch": 1.5501406195471434,
+      "grad_norm": 0.3268077745593211,
+      "learning_rate": 7.108669914081259e-06,
+      "loss": 0.4668,
+      "step": 9439
+    },
+    {
+      "epoch": 1.5503048467554863,
+      "grad_norm": 0.34825895403338536,
+      "learning_rate": 7.108287768780558e-06,
+      "loss": 0.4656,
+      "step": 9440
+    },
+    {
+      "epoch": 1.550469073963829,
+      "grad_norm": 0.34054761958330737,
+      "learning_rate": 7.107905595151283e-06,
+      "loss": 0.4579,
+      "step": 9441
+    },
+    {
+      "epoch": 1.5506333011721716,
+      "grad_norm": 0.32724562169367966,
+      "learning_rate": 7.107523393197588e-06,
+      "loss": 0.451,
+      "step": 9442
+    },
+    {
+      "epoch": 1.5507975283805144,
+      "grad_norm": 0.30315162837296666,
+      "learning_rate": 7.107141162923624e-06,
+      "loss": 0.4703,
+      "step": 9443
+    },
+    {
+      "epoch": 1.5509617555888573,
+      "grad_norm": 0.3294099327301469,
+      "learning_rate": 7.1067589043335415e-06,
+      "loss": 0.4831,
+      "step": 9444
+    },
+    {
+      "epoch": 1.5511259827972,
+      "grad_norm": 0.3131121838015957,
+      "learning_rate": 7.106376617431493e-06,
+      "loss": 0.4784,
+      "step": 9445
+    },
+    {
+      "epoch": 1.5512902100055426,
+      "grad_norm": 0.4850003427166011,
+      "learning_rate": 7.105994302221629e-06,
+      "loss": 0.4584,
+      "step": 9446
+    },
+    {
+      "epoch": 1.5514544372138854,
+      "grad_norm": 0.2824536374442186,
+      "learning_rate": 7.105611958708103e-06,
+      "loss": 0.4684,
+      "step": 9447
+    },
+    {
+      "epoch": 1.5516186644222283,
+      "grad_norm": 0.26487668424818395,
+      "learning_rate": 7.105229586895069e-06,
+      "loss": 0.4706,
+      "step": 9448
+    },
+    {
+      "epoch": 1.551782891630571,
+      "grad_norm": 0.30186016930036735,
+      "learning_rate": 7.104847186786678e-06,
+      "loss": 0.4763,
+      "step": 9449
+    },
+    {
+      "epoch": 1.5519471188389136,
+      "grad_norm": 0.2944963864288857,
+      "learning_rate": 7.104464758387083e-06,
+      "loss": 0.4645,
+      "step": 9450
+    },
+    {
+      "epoch": 1.5521113460472564,
+      "grad_norm": 0.34167036610371465,
+      "learning_rate": 7.104082301700439e-06,
+      "loss": 0.4659,
+      "step": 9451
+    },
+    {
+      "epoch": 1.5522755732555993,
+      "grad_norm": 0.29168994835833856,
+      "learning_rate": 7.103699816730898e-06,
+      "loss": 0.4683,
+      "step": 9452
+    },
+    {
+      "epoch": 1.552439800463942,
+      "grad_norm": 0.31947429820977025,
+      "learning_rate": 7.1033173034826165e-06,
+      "loss": 0.4653,
+      "step": 9453
+    },
+    {
+      "epoch": 1.5526040276722846,
+      "grad_norm": 0.3599259614381269,
+      "learning_rate": 7.102934761959746e-06,
+      "loss": 0.4491,
+      "step": 9454
+    },
+    {
+      "epoch": 1.5527682548806272,
+      "grad_norm": 0.3837674128150548,
+      "learning_rate": 7.102552192166445e-06,
+      "loss": 0.4579,
+      "step": 9455
+    },
+    {
+      "epoch": 1.55293248208897,
+      "grad_norm": 0.5835016569566674,
+      "learning_rate": 7.102169594106867e-06,
+      "loss": 0.4753,
+      "step": 9456
+    },
+    {
+      "epoch": 1.553096709297313,
+      "grad_norm": 0.35557963697118555,
+      "learning_rate": 7.101786967785166e-06,
+      "loss": 0.4752,
+      "step": 9457
+    },
+    {
+      "epoch": 1.5532609365056556,
+      "grad_norm": 0.36659569428207733,
+      "learning_rate": 7.101404313205499e-06,
+      "loss": 0.4628,
+      "step": 9458
+    },
+    {
+      "epoch": 1.5534251637139982,
+      "grad_norm": 0.3163237320621313,
+      "learning_rate": 7.101021630372021e-06,
+      "loss": 0.4731,
+      "step": 9459
+    },
+    {
+      "epoch": 1.553589390922341,
+      "grad_norm": 0.2771310527959539,
+      "learning_rate": 7.1006389192888896e-06,
+      "loss": 0.4713,
+      "step": 9460
+    },
+    {
+      "epoch": 1.553753618130684,
+      "grad_norm": 0.305105596101068,
+      "learning_rate": 7.100256179960261e-06,
+      "loss": 0.4741,
+      "step": 9461
+    },
+    {
+      "epoch": 1.5539178453390265,
+      "grad_norm": 0.31147161003315604,
+      "learning_rate": 7.099873412390292e-06,
+      "loss": 0.4736,
+      "step": 9462
+    },
+    {
+      "epoch": 1.5540820725473692,
+      "grad_norm": 0.2704854712401249,
+      "learning_rate": 7.099490616583138e-06,
+      "loss": 0.4605,
+      "step": 9463
+    },
+    {
+      "epoch": 1.554246299755712,
+      "grad_norm": 0.468442175046904,
+      "learning_rate": 7.099107792542962e-06,
+      "loss": 0.4638,
+      "step": 9464
+    },
+    {
+      "epoch": 1.554410526964055,
+      "grad_norm": 0.2761597581553079,
+      "learning_rate": 7.098724940273914e-06,
+      "loss": 0.4456,
+      "step": 9465
+    },
+    {
+      "epoch": 1.5545747541723975,
+      "grad_norm": 0.3244344994528494,
+      "learning_rate": 7.098342059780159e-06,
+      "loss": 0.4405,
+      "step": 9466
+    },
+    {
+      "epoch": 1.5547389813807402,
+      "grad_norm": 0.4858686845931506,
+      "learning_rate": 7.097959151065853e-06,
+      "loss": 0.4572,
+      "step": 9467
+    },
+    {
+      "epoch": 1.554903208589083,
+      "grad_norm": 0.24869488822776611,
+      "learning_rate": 7.097576214135153e-06,
+      "loss": 0.457,
+      "step": 9468
+    },
+    {
+      "epoch": 1.555067435797426,
+      "grad_norm": 0.3540991970589224,
+      "learning_rate": 7.097193248992222e-06,
+      "loss": 0.4747,
+      "step": 9469
+    },
+    {
+      "epoch": 1.5552316630057685,
+      "grad_norm": 0.3304816779100368,
+      "learning_rate": 7.096810255641214e-06,
+      "loss": 0.449,
+      "step": 9470
+    },
+    {
+      "epoch": 1.5553958902141112,
+      "grad_norm": 0.38004786342989794,
+      "learning_rate": 7.096427234086294e-06,
+      "loss": 0.4592,
+      "step": 9471
+    },
+    {
+      "epoch": 1.5555601174224538,
+      "grad_norm": 0.27383675654332823,
+      "learning_rate": 7.096044184331619e-06,
+      "loss": 0.4678,
+      "step": 9472
+    },
+    {
+      "epoch": 1.5557243446307967,
+      "grad_norm": 0.3790453414043705,
+      "learning_rate": 7.09566110638135e-06,
+      "loss": 0.4699,
+      "step": 9473
+    },
+    {
+      "epoch": 1.5558885718391395,
+      "grad_norm": 0.38324765092279667,
+      "learning_rate": 7.095278000239646e-06,
+      "loss": 0.4563,
+      "step": 9474
+    },
+    {
+      "epoch": 1.5560527990474822,
+      "grad_norm": 0.33095846652110594,
+      "learning_rate": 7.094894865910672e-06,
+      "loss": 0.4408,
+      "step": 9475
+    },
+    {
+      "epoch": 1.5562170262558248,
+      "grad_norm": 0.3003482283168898,
+      "learning_rate": 7.094511703398586e-06,
+      "loss": 0.448,
+      "step": 9476
+    },
+    {
+      "epoch": 1.5563812534641677,
+      "grad_norm": 0.2829919076745514,
+      "learning_rate": 7.094128512707549e-06,
+      "loss": 0.4416,
+      "step": 9477
+    },
+    {
+      "epoch": 1.5565454806725105,
+      "grad_norm": 0.4384164662797856,
+      "learning_rate": 7.093745293841726e-06,
+      "loss": 0.4616,
+      "step": 9478
+    },
+    {
+      "epoch": 1.5567097078808532,
+      "grad_norm": 0.31266180818887007,
+      "learning_rate": 7.093362046805277e-06,
+      "loss": 0.46,
+      "step": 9479
+    },
+    {
+      "epoch": 1.5568739350891958,
+      "grad_norm": 0.27640585641465,
+      "learning_rate": 7.092978771602363e-06,
+      "loss": 0.4869,
+      "step": 9480
+    },
+    {
+      "epoch": 1.5570381622975387,
+      "grad_norm": 0.3111860240242668,
+      "learning_rate": 7.0925954682371495e-06,
+      "loss": 0.4859,
+      "step": 9481
+    },
+    {
+      "epoch": 1.5572023895058815,
+      "grad_norm": 0.26418192548126807,
+      "learning_rate": 7.092212136713798e-06,
+      "loss": 0.4423,
+      "step": 9482
+    },
+    {
+      "epoch": 1.5573666167142242,
+      "grad_norm": 0.32839296991893036,
+      "learning_rate": 7.0918287770364725e-06,
+      "loss": 0.4608,
+      "step": 9483
+    },
+    {
+      "epoch": 1.5575308439225668,
+      "grad_norm": 0.29188899721436623,
+      "learning_rate": 7.091445389209336e-06,
+      "loss": 0.4668,
+      "step": 9484
+    },
+    {
+      "epoch": 1.5576950711309097,
+      "grad_norm": 0.7384461938939119,
+      "learning_rate": 7.091061973236555e-06,
+      "loss": 0.4934,
+      "step": 9485
+    },
+    {
+      "epoch": 1.5578592983392525,
+      "grad_norm": 0.325973263250149,
+      "learning_rate": 7.0906785291222905e-06,
+      "loss": 0.4679,
+      "step": 9486
+    },
+    {
+      "epoch": 1.5580235255475952,
+      "grad_norm": 0.4237636097354834,
+      "learning_rate": 7.090295056870706e-06,
+      "loss": 0.447,
+      "step": 9487
+    },
+    {
+      "epoch": 1.5581877527559378,
+      "grad_norm": 0.27640337167974977,
+      "learning_rate": 7.089911556485971e-06,
+      "loss": 0.4606,
+      "step": 9488
+    },
+    {
+      "epoch": 1.5583519799642804,
+      "grad_norm": 0.2826604533459565,
+      "learning_rate": 7.08952802797225e-06,
+      "loss": 0.4604,
+      "step": 9489
+    },
+    {
+      "epoch": 1.5585162071726233,
+      "grad_norm": 0.33442396415131775,
+      "learning_rate": 7.089144471333703e-06,
+      "loss": 0.4526,
+      "step": 9490
+    },
+    {
+      "epoch": 1.5586804343809662,
+      "grad_norm": 0.3254632769404679,
+      "learning_rate": 7.088760886574502e-06,
+      "loss": 0.4542,
+      "step": 9491
+    },
+    {
+      "epoch": 1.5588446615893088,
+      "grad_norm": 0.3473371450325595,
+      "learning_rate": 7.08837727369881e-06,
+      "loss": 0.4719,
+      "step": 9492
+    },
+    {
+      "epoch": 1.5590088887976514,
+      "grad_norm": 0.3354920157988492,
+      "learning_rate": 7.087993632710796e-06,
+      "loss": 0.4702,
+      "step": 9493
+    },
+    {
+      "epoch": 1.5591731160059943,
+      "grad_norm": 0.30499776435563025,
+      "learning_rate": 7.087609963614622e-06,
+      "loss": 0.4514,
+      "step": 9494
+    },
+    {
+      "epoch": 1.5593373432143371,
+      "grad_norm": 0.3328412721849873,
+      "learning_rate": 7.08722626641446e-06,
+      "loss": 0.4576,
+      "step": 9495
+    },
+    {
+      "epoch": 1.5595015704226798,
+      "grad_norm": 0.31693393803170694,
+      "learning_rate": 7.086842541114474e-06,
+      "loss": 0.4571,
+      "step": 9496
+    },
+    {
+      "epoch": 1.5596657976310224,
+      "grad_norm": 0.35129703393382067,
+      "learning_rate": 7.086458787718834e-06,
+      "loss": 0.4543,
+      "step": 9497
+    },
+    {
+      "epoch": 1.5598300248393653,
+      "grad_norm": 0.294075165507447,
+      "learning_rate": 7.086075006231707e-06,
+      "loss": 0.4705,
+      "step": 9498
+    },
+    {
+      "epoch": 1.5599942520477081,
+      "grad_norm": 0.3011262860718532,
+      "learning_rate": 7.085691196657259e-06,
+      "loss": 0.464,
+      "step": 9499
+    },
+    {
+      "epoch": 1.5601584792560508,
+      "grad_norm": 0.2934159176282551,
+      "learning_rate": 7.085307358999662e-06,
+      "loss": 0.4623,
+      "step": 9500
+    },
+    {
+      "epoch": 1.5603227064643934,
+      "grad_norm": 0.31840048492371464,
+      "learning_rate": 7.084923493263085e-06,
+      "loss": 0.435,
+      "step": 9501
+    },
+    {
+      "epoch": 1.5604869336727363,
+      "grad_norm": 0.38778892493126904,
+      "learning_rate": 7.084539599451693e-06,
+      "loss": 0.4533,
+      "step": 9502
+    },
+    {
+      "epoch": 1.5606511608810791,
+      "grad_norm": 0.3121985866882979,
+      "learning_rate": 7.084155677569659e-06,
+      "loss": 0.476,
+      "step": 9503
+    },
+    {
+      "epoch": 1.5608153880894218,
+      "grad_norm": 0.4007484244591844,
+      "learning_rate": 7.0837717276211524e-06,
+      "loss": 0.4822,
+      "step": 9504
+    },
+    {
+      "epoch": 1.5609796152977644,
+      "grad_norm": 0.3096911825413162,
+      "learning_rate": 7.0833877496103425e-06,
+      "loss": 0.4582,
+      "step": 9505
+    },
+    {
+      "epoch": 1.561143842506107,
+      "grad_norm": 0.3039014707731322,
+      "learning_rate": 7.0830037435414e-06,
+      "loss": 0.4361,
+      "step": 9506
+    },
+    {
+      "epoch": 1.56130806971445,
+      "grad_norm": 0.3185728949963143,
+      "learning_rate": 7.082619709418496e-06,
+      "loss": 0.4464,
+      "step": 9507
+    },
+    {
+      "epoch": 1.5614722969227928,
+      "grad_norm": 0.29381394262617466,
+      "learning_rate": 7.0822356472458e-06,
+      "loss": 0.4765,
+      "step": 9508
+    },
+    {
+      "epoch": 1.5616365241311354,
+      "grad_norm": 0.30664391728474316,
+      "learning_rate": 7.081851557027485e-06,
+      "loss": 0.4621,
+      "step": 9509
+    },
+    {
+      "epoch": 1.561800751339478,
+      "grad_norm": 0.5913865040452077,
+      "learning_rate": 7.0814674387677216e-06,
+      "loss": 0.4729,
+      "step": 9510
+    },
+    {
+      "epoch": 1.561964978547821,
+      "grad_norm": 0.31667754417883187,
+      "learning_rate": 7.081083292470681e-06,
+      "loss": 0.4783,
+      "step": 9511
+    },
+    {
+      "epoch": 1.5621292057561638,
+      "grad_norm": 0.3476930914201371,
+      "learning_rate": 7.080699118140538e-06,
+      "loss": 0.4527,
+      "step": 9512
+    },
+    {
+      "epoch": 1.5622934329645064,
+      "grad_norm": 0.31491112655895354,
+      "learning_rate": 7.080314915781463e-06,
+      "loss": 0.462,
+      "step": 9513
+    },
+    {
+      "epoch": 1.562457660172849,
+      "grad_norm": 0.3392259237310825,
+      "learning_rate": 7.07993068539763e-06,
+      "loss": 0.477,
+      "step": 9514
+    },
+    {
+      "epoch": 1.562621887381192,
+      "grad_norm": 0.34770909353808194,
+      "learning_rate": 7.079546426993211e-06,
+      "loss": 0.4507,
+      "step": 9515
+    },
+    {
+      "epoch": 1.5627861145895348,
+      "grad_norm": 0.31253636621680053,
+      "learning_rate": 7.0791621405723785e-06,
+      "loss": 0.4686,
+      "step": 9516
+    },
+    {
+      "epoch": 1.5629503417978774,
+      "grad_norm": 0.319327369627498,
+      "learning_rate": 7.078777826139309e-06,
+      "loss": 0.4569,
+      "step": 9517
+    },
+    {
+      "epoch": 1.56311456900622,
+      "grad_norm": 0.30529715870836194,
+      "learning_rate": 7.078393483698175e-06,
+      "loss": 0.453,
+      "step": 9518
+    },
+    {
+      "epoch": 1.563278796214563,
+      "grad_norm": 0.265733050448355,
+      "learning_rate": 7.078009113253151e-06,
+      "loss": 0.4722,
+      "step": 9519
+    },
+    {
+      "epoch": 1.5634430234229058,
+      "grad_norm": 0.31408912922552107,
+      "learning_rate": 7.077624714808411e-06,
+      "loss": 0.4641,
+      "step": 9520
+    },
+    {
+      "epoch": 1.5636072506312484,
+      "grad_norm": 0.3132960285785013,
+      "learning_rate": 7.077240288368131e-06,
+      "loss": 0.454,
+      "step": 9521
+    },
+    {
+      "epoch": 1.563771477839591,
+      "grad_norm": 0.3063188137151771,
+      "learning_rate": 7.076855833936486e-06,
+      "loss": 0.4662,
+      "step": 9522
+    },
+    {
+      "epoch": 1.5639357050479337,
+      "grad_norm": 0.2720545580356203,
+      "learning_rate": 7.0764713515176505e-06,
+      "loss": 0.4629,
+      "step": 9523
+    },
+    {
+      "epoch": 1.5640999322562765,
+      "grad_norm": 0.2956064257938657,
+      "learning_rate": 7.076086841115802e-06,
+      "loss": 0.4465,
+      "step": 9524
+    },
+    {
+      "epoch": 1.5642641594646194,
+      "grad_norm": 0.2800826541100162,
+      "learning_rate": 7.075702302735116e-06,
+      "loss": 0.4575,
+      "step": 9525
+    },
+    {
+      "epoch": 1.564428386672962,
+      "grad_norm": 0.41341271140566976,
+      "learning_rate": 7.075317736379768e-06,
+      "loss": 0.4594,
+      "step": 9526
+    },
+    {
+      "epoch": 1.5645926138813047,
+      "grad_norm": 0.34150965202806,
+      "learning_rate": 7.0749331420539365e-06,
+      "loss": 0.4561,
+      "step": 9527
+    },
+    {
+      "epoch": 1.5647568410896475,
+      "grad_norm": 0.24492469743953443,
+      "learning_rate": 7.074548519761797e-06,
+      "loss": 0.4475,
+      "step": 9528
+    },
+    {
+      "epoch": 1.5649210682979904,
+      "grad_norm": 0.34562330583944856,
+      "learning_rate": 7.074163869507528e-06,
+      "loss": 0.472,
+      "step": 9529
+    },
+    {
+      "epoch": 1.565085295506333,
+      "grad_norm": 0.292325354044597,
+      "learning_rate": 7.073779191295306e-06,
+      "loss": 0.4573,
+      "step": 9530
+    },
+    {
+      "epoch": 1.5652495227146757,
+      "grad_norm": 0.4229715578432917,
+      "learning_rate": 7.0733944851293116e-06,
+      "loss": 0.4577,
+      "step": 9531
+    },
+    {
+      "epoch": 1.5654137499230185,
+      "grad_norm": 0.32096121679499595,
+      "learning_rate": 7.0730097510137204e-06,
+      "loss": 0.4503,
+      "step": 9532
+    },
+    {
+      "epoch": 1.5655779771313614,
+      "grad_norm": 0.27339951728229317,
+      "learning_rate": 7.072624988952711e-06,
+      "loss": 0.4557,
+      "step": 9533
+    },
+    {
+      "epoch": 1.565742204339704,
+      "grad_norm": 0.30943985914503824,
+      "learning_rate": 7.0722401989504634e-06,
+      "loss": 0.4557,
+      "step": 9534
+    },
+    {
+      "epoch": 1.5659064315480467,
+      "grad_norm": 0.33863811933806,
+      "learning_rate": 7.071855381011157e-06,
+      "loss": 0.4532,
+      "step": 9535
+    },
+    {
+      "epoch": 1.5660706587563895,
+      "grad_norm": 0.32722389646224836,
+      "learning_rate": 7.071470535138969e-06,
+      "loss": 0.4534,
+      "step": 9536
+    },
+    {
+      "epoch": 1.5662348859647324,
+      "grad_norm": 0.3616589600678771,
+      "learning_rate": 7.071085661338083e-06,
+      "loss": 0.4596,
+      "step": 9537
+    },
+    {
+      "epoch": 1.566399113173075,
+      "grad_norm": 0.290065563681788,
+      "learning_rate": 7.070700759612676e-06,
+      "loss": 0.4419,
+      "step": 9538
+    },
+    {
+      "epoch": 1.5665633403814176,
+      "grad_norm": 0.3681681484784924,
+      "learning_rate": 7.070315829966932e-06,
+      "loss": 0.4674,
+      "step": 9539
+    },
+    {
+      "epoch": 1.5667275675897603,
+      "grad_norm": 0.29433431026892365,
+      "learning_rate": 7.069930872405026e-06,
+      "loss": 0.4681,
+      "step": 9540
+    },
+    {
+      "epoch": 1.5668917947981031,
+      "grad_norm": 0.28165471409438464,
+      "learning_rate": 7.069545886931144e-06,
+      "loss": 0.4381,
+      "step": 9541
+    },
+    {
+      "epoch": 1.567056022006446,
+      "grad_norm": 0.2646752230437082,
+      "learning_rate": 7.069160873549464e-06,
+      "loss": 0.4711,
+      "step": 9542
+    },
+    {
+      "epoch": 1.5672202492147886,
+      "grad_norm": 0.29628517431852486,
+      "learning_rate": 7.068775832264172e-06,
+      "loss": 0.4607,
+      "step": 9543
+    },
+    {
+      "epoch": 1.5673844764231313,
+      "grad_norm": 0.44146222206987495,
+      "learning_rate": 7.068390763079445e-06,
+      "loss": 0.4721,
+      "step": 9544
+    },
+    {
+      "epoch": 1.5675487036314741,
+      "grad_norm": 0.3452702849381547,
+      "learning_rate": 7.068005665999467e-06,
+      "loss": 0.4667,
+      "step": 9545
+    },
+    {
+      "epoch": 1.567712930839817,
+      "grad_norm": 0.32130074805891345,
+      "learning_rate": 7.067620541028422e-06,
+      "loss": 0.4643,
+      "step": 9546
+    },
+    {
+      "epoch": 1.5678771580481596,
+      "grad_norm": 0.2856839609238298,
+      "learning_rate": 7.0672353881704915e-06,
+      "loss": 0.45,
+      "step": 9547
+    },
+    {
+      "epoch": 1.5680413852565023,
+      "grad_norm": 0.41361729342399817,
+      "learning_rate": 7.066850207429859e-06,
+      "loss": 0.457,
+      "step": 9548
+    },
+    {
+      "epoch": 1.5682056124648451,
+      "grad_norm": 0.4783024490657136,
+      "learning_rate": 7.066464998810707e-06,
+      "loss": 0.4673,
+      "step": 9549
+    },
+    {
+      "epoch": 1.568369839673188,
+      "grad_norm": 0.26854722532828895,
+      "learning_rate": 7.066079762317221e-06,
+      "loss": 0.4533,
+      "step": 9550
+    },
+    {
+      "epoch": 1.5685340668815306,
+      "grad_norm": 0.3243123197398592,
+      "learning_rate": 7.0656944979535836e-06,
+      "loss": 0.4703,
+      "step": 9551
+    },
+    {
+      "epoch": 1.5686982940898733,
+      "grad_norm": 0.31882320116029533,
+      "learning_rate": 7.06530920572398e-06,
+      "loss": 0.4334,
+      "step": 9552
+    },
+    {
+      "epoch": 1.5688625212982161,
+      "grad_norm": 0.3153813981363007,
+      "learning_rate": 7.064923885632595e-06,
+      "loss": 0.4629,
+      "step": 9553
+    },
+    {
+      "epoch": 1.569026748506559,
+      "grad_norm": 0.32022797273174436,
+      "learning_rate": 7.064538537683612e-06,
+      "loss": 0.4655,
+      "step": 9554
+    },
+    {
+      "epoch": 1.5691909757149016,
+      "grad_norm": 0.3000384583865678,
+      "learning_rate": 7.0641531618812165e-06,
+      "loss": 0.4616,
+      "step": 9555
+    },
+    {
+      "epoch": 1.5693552029232443,
+      "grad_norm": 0.3007245952874813,
+      "learning_rate": 7.063767758229597e-06,
+      "loss": 0.458,
+      "step": 9556
+    },
+    {
+      "epoch": 1.569519430131587,
+      "grad_norm": 0.3370373862995186,
+      "learning_rate": 7.063382326732936e-06,
+      "loss": 0.4564,
+      "step": 9557
+    },
+    {
+      "epoch": 1.5696836573399298,
+      "grad_norm": 0.2668854475743722,
+      "learning_rate": 7.062996867395421e-06,
+      "loss": 0.4637,
+      "step": 9558
+    },
+    {
+      "epoch": 1.5698478845482726,
+      "grad_norm": 0.32082269614047243,
+      "learning_rate": 7.062611380221239e-06,
+      "loss": 0.4557,
+      "step": 9559
+    },
+    {
+      "epoch": 1.5700121117566153,
+      "grad_norm": 0.2642372498332476,
+      "learning_rate": 7.0622258652145755e-06,
+      "loss": 0.459,
+      "step": 9560
+    },
+    {
+      "epoch": 1.570176338964958,
+      "grad_norm": 0.32119739559651794,
+      "learning_rate": 7.061840322379618e-06,
+      "loss": 0.4826,
+      "step": 9561
+    },
+    {
+      "epoch": 1.5703405661733008,
+      "grad_norm": 0.33573520572974747,
+      "learning_rate": 7.061454751720556e-06,
+      "loss": 0.4773,
+      "step": 9562
+    },
+    {
+      "epoch": 1.5705047933816436,
+      "grad_norm": 0.3727906802239041,
+      "learning_rate": 7.061069153241572e-06,
+      "loss": 0.467,
+      "step": 9563
+    },
+    {
+      "epoch": 1.5706690205899863,
+      "grad_norm": 0.25635954624939483,
+      "learning_rate": 7.06068352694686e-06,
+      "loss": 0.4725,
+      "step": 9564
+    },
+    {
+      "epoch": 1.570833247798329,
+      "grad_norm": 0.3569861822095909,
+      "learning_rate": 7.060297872840604e-06,
+      "loss": 0.4674,
+      "step": 9565
+    },
+    {
+      "epoch": 1.5709974750066718,
+      "grad_norm": 0.4167333878551979,
+      "learning_rate": 7.059912190926995e-06,
+      "loss": 0.464,
+      "step": 9566
+    },
+    {
+      "epoch": 1.5711617022150146,
+      "grad_norm": 0.2988223076469179,
+      "learning_rate": 7.05952648121022e-06,
+      "loss": 0.4883,
+      "step": 9567
+    },
+    {
+      "epoch": 1.5713259294233572,
+      "grad_norm": 0.2830256329297102,
+      "learning_rate": 7.059140743694471e-06,
+      "loss": 0.4476,
+      "step": 9568
+    },
+    {
+      "epoch": 1.5714901566316999,
+      "grad_norm": 0.372258375978891,
+      "learning_rate": 7.058754978383934e-06,
+      "loss": 0.4893,
+      "step": 9569
+    },
+    {
+      "epoch": 1.5716543838400427,
+      "grad_norm": 0.2654171969672735,
+      "learning_rate": 7.0583691852828e-06,
+      "loss": 0.4525,
+      "step": 9570
+    },
+    {
+      "epoch": 1.5718186110483856,
+      "grad_norm": 0.280300815255278,
+      "learning_rate": 7.0579833643952605e-06,
+      "loss": 0.455,
+      "step": 9571
+    },
+    {
+      "epoch": 1.5719828382567282,
+      "grad_norm": 0.5008140822399365,
+      "learning_rate": 7.0575975157255044e-06,
+      "loss": 0.4672,
+      "step": 9572
+    },
+    {
+      "epoch": 1.5721470654650709,
+      "grad_norm": 0.31955287284252826,
+      "learning_rate": 7.057211639277725e-06,
+      "loss": 0.477,
+      "step": 9573
+    },
+    {
+      "epoch": 1.5723112926734135,
+      "grad_norm": 0.3489108249379957,
+      "learning_rate": 7.05682573505611e-06,
+      "loss": 0.4773,
+      "step": 9574
+    },
+    {
+      "epoch": 1.5724755198817564,
+      "grad_norm": 0.35593903885934475,
+      "learning_rate": 7.056439803064851e-06,
+      "loss": 0.4744,
+      "step": 9575
+    },
+    {
+      "epoch": 1.5726397470900992,
+      "grad_norm": 0.28457851992077404,
+      "learning_rate": 7.056053843308141e-06,
+      "loss": 0.4511,
+      "step": 9576
+    },
+    {
+      "epoch": 1.5728039742984419,
+      "grad_norm": 0.27240800898902373,
+      "learning_rate": 7.0556678557901725e-06,
+      "loss": 0.4603,
+      "step": 9577
+    },
+    {
+      "epoch": 1.5729682015067845,
+      "grad_norm": 0.2727708545101182,
+      "learning_rate": 7.055281840515136e-06,
+      "loss": 0.4814,
+      "step": 9578
+    },
+    {
+      "epoch": 1.5731324287151274,
+      "grad_norm": 0.31236355231569163,
+      "learning_rate": 7.054895797487223e-06,
+      "loss": 0.4415,
+      "step": 9579
+    },
+    {
+      "epoch": 1.5732966559234702,
+      "grad_norm": 0.27149555085526944,
+      "learning_rate": 7.054509726710629e-06,
+      "loss": 0.4602,
+      "step": 9580
+    },
+    {
+      "epoch": 1.5734608831318129,
+      "grad_norm": 0.26192212609758236,
+      "learning_rate": 7.054123628189548e-06,
+      "loss": 0.473,
+      "step": 9581
+    },
+    {
+      "epoch": 1.5736251103401555,
+      "grad_norm": 0.2604095204435766,
+      "learning_rate": 7.0537375019281705e-06,
+      "loss": 0.4786,
+      "step": 9582
+    },
+    {
+      "epoch": 1.5737893375484984,
+      "grad_norm": 0.3173000671655432,
+      "learning_rate": 7.05335134793069e-06,
+      "loss": 0.4431,
+      "step": 9583
+    },
+    {
+      "epoch": 1.5739535647568412,
+      "grad_norm": 0.4249155458956311,
+      "learning_rate": 7.052965166201301e-06,
+      "loss": 0.4422,
+      "step": 9584
+    },
+    {
+      "epoch": 1.5741177919651839,
+      "grad_norm": 0.326714175509031,
+      "learning_rate": 7.052578956744201e-06,
+      "loss": 0.4533,
+      "step": 9585
+    },
+    {
+      "epoch": 1.5742820191735265,
+      "grad_norm": 0.30076342807026163,
+      "learning_rate": 7.052192719563582e-06,
+      "loss": 0.4646,
+      "step": 9586
+    },
+    {
+      "epoch": 1.5744462463818694,
+      "grad_norm": 0.5574470434140718,
+      "learning_rate": 7.0518064546636375e-06,
+      "loss": 0.4785,
+      "step": 9587
+    },
+    {
+      "epoch": 1.5746104735902122,
+      "grad_norm": 0.3899021496848957,
+      "learning_rate": 7.051420162048565e-06,
+      "loss": 0.4645,
+      "step": 9588
+    },
+    {
+      "epoch": 1.5747747007985549,
+      "grad_norm": 0.5009365869861512,
+      "learning_rate": 7.051033841722559e-06,
+      "loss": 0.4616,
+      "step": 9589
+    },
+    {
+      "epoch": 1.5749389280068975,
+      "grad_norm": 0.3026411124363173,
+      "learning_rate": 7.050647493689816e-06,
+      "loss": 0.4617,
+      "step": 9590
+    },
+    {
+      "epoch": 1.5751031552152401,
+      "grad_norm": 0.29061469419822633,
+      "learning_rate": 7.050261117954531e-06,
+      "loss": 0.4668,
+      "step": 9591
+    },
+    {
+      "epoch": 1.575267382423583,
+      "grad_norm": 0.29963184439358226,
+      "learning_rate": 7.049874714520902e-06,
+      "loss": 0.4807,
+      "step": 9592
+    },
+    {
+      "epoch": 1.5754316096319259,
+      "grad_norm": 0.32685838124755867,
+      "learning_rate": 7.049488283393124e-06,
+      "loss": 0.4554,
+      "step": 9593
+    },
+    {
+      "epoch": 1.5755958368402685,
+      "grad_norm": 0.3098084135045518,
+      "learning_rate": 7.049101824575395e-06,
+      "loss": 0.4631,
+      "step": 9594
+    },
+    {
+      "epoch": 1.5757600640486111,
+      "grad_norm": 0.5182353120987377,
+      "learning_rate": 7.048715338071913e-06,
+      "loss": 0.4556,
+      "step": 9595
+    },
+    {
+      "epoch": 1.575924291256954,
+      "grad_norm": 0.33384098807841417,
+      "learning_rate": 7.048328823886873e-06,
+      "loss": 0.4752,
+      "step": 9596
+    },
+    {
+      "epoch": 1.5760885184652969,
+      "grad_norm": 0.5106129343711809,
+      "learning_rate": 7.047942282024477e-06,
+      "loss": 0.4458,
+      "step": 9597
+    },
+    {
+      "epoch": 1.5762527456736395,
+      "grad_norm": 0.2779558813719126,
+      "learning_rate": 7.047555712488921e-06,
+      "loss": 0.4627,
+      "step": 9598
+    },
+    {
+      "epoch": 1.5764169728819821,
+      "grad_norm": 0.3036165546910705,
+      "learning_rate": 7.047169115284401e-06,
+      "loss": 0.4542,
+      "step": 9599
+    },
+    {
+      "epoch": 1.576581200090325,
+      "grad_norm": 0.2813306295945631,
+      "learning_rate": 7.0467824904151205e-06,
+      "loss": 0.4574,
+      "step": 9600
+    },
+    {
+      "epoch": 1.5767454272986678,
+      "grad_norm": 0.3756631418000488,
+      "learning_rate": 7.046395837885276e-06,
+      "loss": 0.4665,
+      "step": 9601
+    },
+    {
+      "epoch": 1.5769096545070105,
+      "grad_norm": 0.3014834902402672,
+      "learning_rate": 7.046009157699068e-06,
+      "loss": 0.4686,
+      "step": 9602
+    },
+    {
+      "epoch": 1.5770738817153531,
+      "grad_norm": 0.37031689284883956,
+      "learning_rate": 7.045622449860695e-06,
+      "loss": 0.438,
+      "step": 9603
+    },
+    {
+      "epoch": 1.577238108923696,
+      "grad_norm": 0.5091529417201783,
+      "learning_rate": 7.045235714374358e-06,
+      "loss": 0.4596,
+      "step": 9604
+    },
+    {
+      "epoch": 1.5774023361320388,
+      "grad_norm": 0.316980806826266,
+      "learning_rate": 7.044848951244255e-06,
+      "loss": 0.4717,
+      "step": 9605
+    },
+    {
+      "epoch": 1.5775665633403815,
+      "grad_norm": 0.27563114539777,
+      "learning_rate": 7.04446216047459e-06,
+      "loss": 0.4652,
+      "step": 9606
+    },
+    {
+      "epoch": 1.5777307905487241,
+      "grad_norm": 0.3254061465093312,
+      "learning_rate": 7.044075342069563e-06,
+      "loss": 0.4721,
+      "step": 9607
+    },
+    {
+      "epoch": 1.5778950177570668,
+      "grad_norm": 0.3070070243971381,
+      "learning_rate": 7.0436884960333746e-06,
+      "loss": 0.4451,
+      "step": 9608
+    },
+    {
+      "epoch": 1.5780592449654096,
+      "grad_norm": 0.3046083263392374,
+      "learning_rate": 7.0433016223702255e-06,
+      "loss": 0.4646,
+      "step": 9609
+    },
+    {
+      "epoch": 1.5782234721737525,
+      "grad_norm": 0.3356082246504402,
+      "learning_rate": 7.04291472108432e-06,
+      "loss": 0.4773,
+      "step": 9610
+    },
+    {
+      "epoch": 1.578387699382095,
+      "grad_norm": 0.3154556445315861,
+      "learning_rate": 7.042527792179857e-06,
+      "loss": 0.4668,
+      "step": 9611
+    },
+    {
+      "epoch": 1.5785519265904377,
+      "grad_norm": 0.34093873199245545,
+      "learning_rate": 7.042140835661042e-06,
+      "loss": 0.4478,
+      "step": 9612
+    },
+    {
+      "epoch": 1.5787161537987806,
+      "grad_norm": 0.4638435804483628,
+      "learning_rate": 7.041753851532076e-06,
+      "loss": 0.4545,
+      "step": 9613
+    },
+    {
+      "epoch": 1.5788803810071235,
+      "grad_norm": 0.3080869937789505,
+      "learning_rate": 7.041366839797163e-06,
+      "loss": 0.4676,
+      "step": 9614
+    },
+    {
+      "epoch": 1.579044608215466,
+      "grad_norm": 0.3208085604236869,
+      "learning_rate": 7.040979800460505e-06,
+      "loss": 0.4618,
+      "step": 9615
+    },
+    {
+      "epoch": 1.5792088354238087,
+      "grad_norm": 0.37304515332678506,
+      "learning_rate": 7.040592733526307e-06,
+      "loss": 0.47,
+      "step": 9616
+    },
+    {
+      "epoch": 1.5793730626321516,
+      "grad_norm": 0.3489077569804723,
+      "learning_rate": 7.0402056389987716e-06,
+      "loss": 0.4719,
+      "step": 9617
+    },
+    {
+      "epoch": 1.5795372898404945,
+      "grad_norm": 0.3204051769035944,
+      "learning_rate": 7.039818516882105e-06,
+      "loss": 0.4643,
+      "step": 9618
+    },
+    {
+      "epoch": 1.579701517048837,
+      "grad_norm": 0.33360715457799445,
+      "learning_rate": 7.039431367180509e-06,
+      "loss": 0.4711,
+      "step": 9619
+    },
+    {
+      "epoch": 1.5798657442571797,
+      "grad_norm": 0.3188657125688107,
+      "learning_rate": 7.03904418989819e-06,
+      "loss": 0.478,
+      "step": 9620
+    },
+    {
+      "epoch": 1.5800299714655226,
+      "grad_norm": 0.3406891273173139,
+      "learning_rate": 7.038656985039353e-06,
+      "loss": 0.464,
+      "step": 9621
+    },
+    {
+      "epoch": 1.5801941986738655,
+      "grad_norm": 0.3344841323401284,
+      "learning_rate": 7.038269752608205e-06,
+      "loss": 0.4628,
+      "step": 9622
+    },
+    {
+      "epoch": 1.580358425882208,
+      "grad_norm": 0.2962372762879127,
+      "learning_rate": 7.0378824926089485e-06,
+      "loss": 0.4557,
+      "step": 9623
+    },
+    {
+      "epoch": 1.5805226530905507,
+      "grad_norm": 0.30877525105222614,
+      "learning_rate": 7.0374952050457925e-06,
+      "loss": 0.4545,
+      "step": 9624
+    },
+    {
+      "epoch": 1.5806868802988934,
+      "grad_norm": 0.5755131888295433,
+      "learning_rate": 7.037107889922941e-06,
+      "loss": 0.4472,
+      "step": 9625
+    },
+    {
+      "epoch": 1.5808511075072362,
+      "grad_norm": 0.25748170687714833,
+      "learning_rate": 7.036720547244602e-06,
+      "loss": 0.4641,
+      "step": 9626
+    },
+    {
+      "epoch": 1.581015334715579,
+      "grad_norm": 0.36057283649245553,
+      "learning_rate": 7.0363331770149826e-06,
+      "loss": 0.468,
+      "step": 9627
+    },
+    {
+      "epoch": 1.5811795619239217,
+      "grad_norm": 0.327370587050576,
+      "learning_rate": 7.035945779238288e-06,
+      "loss": 0.4446,
+      "step": 9628
+    },
+    {
+      "epoch": 1.5813437891322644,
+      "grad_norm": 0.32141525897584117,
+      "learning_rate": 7.035558353918728e-06,
+      "loss": 0.4408,
+      "step": 9629
+    },
+    {
+      "epoch": 1.5815080163406072,
+      "grad_norm": 0.2675636946309375,
+      "learning_rate": 7.035170901060509e-06,
+      "loss": 0.461,
+      "step": 9630
+    },
+    {
+      "epoch": 1.58167224354895,
+      "grad_norm": 0.3485598277383404,
+      "learning_rate": 7.034783420667841e-06,
+      "loss": 0.4599,
+      "step": 9631
+    },
+    {
+      "epoch": 1.5818364707572927,
+      "grad_norm": 0.40961186549980955,
+      "learning_rate": 7.03439591274493e-06,
+      "loss": 0.4485,
+      "step": 9632
+    },
+    {
+      "epoch": 1.5820006979656354,
+      "grad_norm": 0.3150450672652272,
+      "learning_rate": 7.034008377295987e-06,
+      "loss": 0.4458,
+      "step": 9633
+    },
+    {
+      "epoch": 1.5821649251739782,
+      "grad_norm": 0.2925616187190679,
+      "learning_rate": 7.033620814325219e-06,
+      "loss": 0.4864,
+      "step": 9634
+    },
+    {
+      "epoch": 1.582329152382321,
+      "grad_norm": 0.33606955689744944,
+      "learning_rate": 7.033233223836837e-06,
+      "loss": 0.466,
+      "step": 9635
+    },
+    {
+      "epoch": 1.5824933795906637,
+      "grad_norm": 0.4824169072771863,
+      "learning_rate": 7.03284560583505e-06,
+      "loss": 0.4528,
+      "step": 9636
+    },
+    {
+      "epoch": 1.5826576067990064,
+      "grad_norm": 0.2906977714611012,
+      "learning_rate": 7.032457960324066e-06,
+      "loss": 0.4537,
+      "step": 9637
+    },
+    {
+      "epoch": 1.5828218340073492,
+      "grad_norm": 0.279391411557559,
+      "learning_rate": 7.032070287308099e-06,
+      "loss": 0.4664,
+      "step": 9638
+    },
+    {
+      "epoch": 1.582986061215692,
+      "grad_norm": 0.321052045706845,
+      "learning_rate": 7.031682586791356e-06,
+      "loss": 0.4575,
+      "step": 9639
+    },
+    {
+      "epoch": 1.5831502884240347,
+      "grad_norm": 0.3399100775023264,
+      "learning_rate": 7.03129485877805e-06,
+      "loss": 0.4729,
+      "step": 9640
+    },
+    {
+      "epoch": 1.5833145156323774,
+      "grad_norm": 0.32977031429239967,
+      "learning_rate": 7.030907103272391e-06,
+      "loss": 0.4716,
+      "step": 9641
+    },
+    {
+      "epoch": 1.58347874284072,
+      "grad_norm": 0.31242153963448743,
+      "learning_rate": 7.0305193202785905e-06,
+      "loss": 0.473,
+      "step": 9642
+    },
+    {
+      "epoch": 1.5836429700490628,
+      "grad_norm": 0.3019615032920058,
+      "learning_rate": 7.030131509800861e-06,
+      "loss": 0.4414,
+      "step": 9643
+    },
+    {
+      "epoch": 1.5838071972574057,
+      "grad_norm": 0.3483822784558959,
+      "learning_rate": 7.029743671843415e-06,
+      "loss": 0.4646,
+      "step": 9644
+    },
+    {
+      "epoch": 1.5839714244657483,
+      "grad_norm": 0.2991796623781081,
+      "learning_rate": 7.029355806410462e-06,
+      "loss": 0.4656,
+      "step": 9645
+    },
+    {
+      "epoch": 1.584135651674091,
+      "grad_norm": 0.36838639847658394,
+      "learning_rate": 7.028967913506217e-06,
+      "loss": 0.4784,
+      "step": 9646
+    },
+    {
+      "epoch": 1.5842998788824338,
+      "grad_norm": 0.30208778163975347,
+      "learning_rate": 7.028579993134892e-06,
+      "loss": 0.4827,
+      "step": 9647
+    },
+    {
+      "epoch": 1.5844641060907767,
+      "grad_norm": 0.27407402707941875,
+      "learning_rate": 7.028192045300701e-06,
+      "loss": 0.4668,
+      "step": 9648
+    },
+    {
+      "epoch": 1.5846283332991193,
+      "grad_norm": 0.2618196125378437,
+      "learning_rate": 7.027804070007858e-06,
+      "loss": 0.4807,
+      "step": 9649
+    },
+    {
+      "epoch": 1.584792560507462,
+      "grad_norm": 0.3152951828088361,
+      "learning_rate": 7.027416067260574e-06,
+      "loss": 0.4632,
+      "step": 9650
+    },
+    {
+      "epoch": 1.5849567877158048,
+      "grad_norm": 0.3463447699192336,
+      "learning_rate": 7.027028037063066e-06,
+      "loss": 0.45,
+      "step": 9651
+    },
+    {
+      "epoch": 1.5851210149241477,
+      "grad_norm": 0.2943307574825071,
+      "learning_rate": 7.0266399794195465e-06,
+      "loss": 0.4598,
+      "step": 9652
+    },
+    {
+      "epoch": 1.5852852421324903,
+      "grad_norm": 0.3006986761693188,
+      "learning_rate": 7.026251894334231e-06,
+      "loss": 0.4703,
+      "step": 9653
+    },
+    {
+      "epoch": 1.585449469340833,
+      "grad_norm": 0.3068826226953967,
+      "learning_rate": 7.025863781811335e-06,
+      "loss": 0.4569,
+      "step": 9654
+    },
+    {
+      "epoch": 1.5856136965491758,
+      "grad_norm": 0.2975266953823899,
+      "learning_rate": 7.025475641855074e-06,
+      "loss": 0.4541,
+      "step": 9655
+    },
+    {
+      "epoch": 1.5857779237575187,
+      "grad_norm": 0.3719118805987487,
+      "learning_rate": 7.025087474469661e-06,
+      "loss": 0.4693,
+      "step": 9656
+    },
+    {
+      "epoch": 1.5859421509658613,
+      "grad_norm": 0.316004408532983,
+      "learning_rate": 7.024699279659314e-06,
+      "loss": 0.4846,
+      "step": 9657
+    },
+    {
+      "epoch": 1.586106378174204,
+      "grad_norm": 0.2750964430209116,
+      "learning_rate": 7.024311057428249e-06,
+      "loss": 0.45,
+      "step": 9658
+    },
+    {
+      "epoch": 1.5862706053825466,
+      "grad_norm": 0.25877806352237737,
+      "learning_rate": 7.023922807780682e-06,
+      "loss": 0.4712,
+      "step": 9659
+    },
+    {
+      "epoch": 1.5864348325908895,
+      "grad_norm": 0.30726318388169743,
+      "learning_rate": 7.023534530720832e-06,
+      "loss": 0.4505,
+      "step": 9660
+    },
+    {
+      "epoch": 1.5865990597992323,
+      "grad_norm": 0.277175420733913,
+      "learning_rate": 7.0231462262529125e-06,
+      "loss": 0.4462,
+      "step": 9661
+    },
+    {
+      "epoch": 1.586763287007575,
+      "grad_norm": 0.2480687810862842,
+      "learning_rate": 7.022757894381143e-06,
+      "loss": 0.4578,
+      "step": 9662
+    },
+    {
+      "epoch": 1.5869275142159176,
+      "grad_norm": 0.310062585639788,
+      "learning_rate": 7.02236953510974e-06,
+      "loss": 0.4529,
+      "step": 9663
+    },
+    {
+      "epoch": 1.5870917414242605,
+      "grad_norm": 0.29278447216754283,
+      "learning_rate": 7.021981148442923e-06,
+      "loss": 0.4479,
+      "step": 9664
+    },
+    {
+      "epoch": 1.5872559686326033,
+      "grad_norm": 0.34205414183900085,
+      "learning_rate": 7.0215927343849095e-06,
+      "loss": 0.4567,
+      "step": 9665
+    },
+    {
+      "epoch": 1.587420195840946,
+      "grad_norm": 0.30235224417403145,
+      "learning_rate": 7.021204292939917e-06,
+      "loss": 0.4504,
+      "step": 9666
+    },
+    {
+      "epoch": 1.5875844230492886,
+      "grad_norm": 0.3204273273921009,
+      "learning_rate": 7.020815824112166e-06,
+      "loss": 0.4555,
+      "step": 9667
+    },
+    {
+      "epoch": 1.5877486502576315,
+      "grad_norm": 0.29059833973456584,
+      "learning_rate": 7.020427327905875e-06,
+      "loss": 0.4524,
+      "step": 9668
+    },
+    {
+      "epoch": 1.5879128774659743,
+      "grad_norm": 0.312532284678278,
+      "learning_rate": 7.0200388043252635e-06,
+      "loss": 0.4531,
+      "step": 9669
+    },
+    {
+      "epoch": 1.588077104674317,
+      "grad_norm": 0.4001305657269944,
+      "learning_rate": 7.019650253374552e-06,
+      "loss": 0.4649,
+      "step": 9670
+    },
+    {
+      "epoch": 1.5882413318826596,
+      "grad_norm": 0.29994775751366703,
+      "learning_rate": 7.019261675057958e-06,
+      "loss": 0.454,
+      "step": 9671
+    },
+    {
+      "epoch": 1.5884055590910025,
+      "grad_norm": 0.3719043680773999,
+      "learning_rate": 7.018873069379705e-06,
+      "loss": 0.4631,
+      "step": 9672
+    },
+    {
+      "epoch": 1.5885697862993453,
+      "grad_norm": 0.2949768760769632,
+      "learning_rate": 7.018484436344012e-06,
+      "loss": 0.4696,
+      "step": 9673
+    },
+    {
+      "epoch": 1.588734013507688,
+      "grad_norm": 0.28422629123361914,
+      "learning_rate": 7.018095775955099e-06,
+      "loss": 0.4618,
+      "step": 9674
+    },
+    {
+      "epoch": 1.5888982407160306,
+      "grad_norm": 0.27145460449253184,
+      "learning_rate": 7.0177070882171895e-06,
+      "loss": 0.4839,
+      "step": 9675
+    },
+    {
+      "epoch": 1.5890624679243732,
+      "grad_norm": 0.3286664513768836,
+      "learning_rate": 7.017318373134504e-06,
+      "loss": 0.4594,
+      "step": 9676
+    },
+    {
+      "epoch": 1.589226695132716,
+      "grad_norm": 0.3619168102990762,
+      "learning_rate": 7.0169296307112635e-06,
+      "loss": 0.4444,
+      "step": 9677
+    },
+    {
+      "epoch": 1.589390922341059,
+      "grad_norm": 0.28446907903214214,
+      "learning_rate": 7.0165408609516916e-06,
+      "loss": 0.4572,
+      "step": 9678
+    },
+    {
+      "epoch": 1.5895551495494016,
+      "grad_norm": 0.352917143530713,
+      "learning_rate": 7.0161520638600105e-06,
+      "loss": 0.4514,
+      "step": 9679
+    },
+    {
+      "epoch": 1.5897193767577442,
+      "grad_norm": 0.35002679253958296,
+      "learning_rate": 7.01576323944044e-06,
+      "loss": 0.4588,
+      "step": 9680
+    },
+    {
+      "epoch": 1.589883603966087,
+      "grad_norm": 0.32197653074915533,
+      "learning_rate": 7.015374387697208e-06,
+      "loss": 0.4674,
+      "step": 9681
+    },
+    {
+      "epoch": 1.59004783117443,
+      "grad_norm": 0.29434146515200116,
+      "learning_rate": 7.014985508634535e-06,
+      "loss": 0.4843,
+      "step": 9682
+    },
+    {
+      "epoch": 1.5902120583827726,
+      "grad_norm": 0.32987470643692046,
+      "learning_rate": 7.014596602256644e-06,
+      "loss": 0.4927,
+      "step": 9683
+    },
+    {
+      "epoch": 1.5903762855911152,
+      "grad_norm": 0.43707017640769624,
+      "learning_rate": 7.014207668567761e-06,
+      "loss": 0.4806,
+      "step": 9684
+    },
+    {
+      "epoch": 1.590540512799458,
+      "grad_norm": 0.37174042084246073,
+      "learning_rate": 7.013818707572109e-06,
+      "loss": 0.483,
+      "step": 9685
+    },
+    {
+      "epoch": 1.590704740007801,
+      "grad_norm": 0.29865649575691555,
+      "learning_rate": 7.013429719273912e-06,
+      "loss": 0.4724,
+      "step": 9686
+    },
+    {
+      "epoch": 1.5908689672161436,
+      "grad_norm": 0.28389824575492095,
+      "learning_rate": 7.0130407036773955e-06,
+      "loss": 0.4617,
+      "step": 9687
+    },
+    {
+      "epoch": 1.5910331944244862,
+      "grad_norm": 0.4856752735113594,
+      "learning_rate": 7.0126516607867845e-06,
+      "loss": 0.4588,
+      "step": 9688
+    },
+    {
+      "epoch": 1.591197421632829,
+      "grad_norm": 0.3280719264639045,
+      "learning_rate": 7.012262590606304e-06,
+      "loss": 0.4744,
+      "step": 9689
+    },
+    {
+      "epoch": 1.591361648841172,
+      "grad_norm": 0.3017016566131072,
+      "learning_rate": 7.011873493140181e-06,
+      "loss": 0.4561,
+      "step": 9690
+    },
+    {
+      "epoch": 1.5915258760495146,
+      "grad_norm": 0.27235999254119797,
+      "learning_rate": 7.011484368392639e-06,
+      "loss": 0.4589,
+      "step": 9691
+    },
+    {
+      "epoch": 1.5916901032578572,
+      "grad_norm": 0.293826082990073,
+      "learning_rate": 7.011095216367906e-06,
+      "loss": 0.4538,
+      "step": 9692
+    },
+    {
+      "epoch": 1.5918543304661998,
+      "grad_norm": 0.30234293241414956,
+      "learning_rate": 7.010706037070209e-06,
+      "loss": 0.461,
+      "step": 9693
+    },
+    {
+      "epoch": 1.5920185576745427,
+      "grad_norm": 0.2807863448579779,
+      "learning_rate": 7.010316830503775e-06,
+      "loss": 0.4579,
+      "step": 9694
+    },
+    {
+      "epoch": 1.5921827848828856,
+      "grad_norm": 0.25374035826197255,
+      "learning_rate": 7.009927596672829e-06,
+      "loss": 0.4783,
+      "step": 9695
+    },
+    {
+      "epoch": 1.5923470120912282,
+      "grad_norm": 0.28295051020587353,
+      "learning_rate": 7.009538335581601e-06,
+      "loss": 0.4487,
+      "step": 9696
+    },
+    {
+      "epoch": 1.5925112392995708,
+      "grad_norm": 0.9402708462138283,
+      "learning_rate": 7.0091490472343165e-06,
+      "loss": 0.4652,
+      "step": 9697
+    },
+    {
+      "epoch": 1.5926754665079137,
+      "grad_norm": 0.3089325539254181,
+      "learning_rate": 7.008759731635206e-06,
+      "loss": 0.4736,
+      "step": 9698
+    },
+    {
+      "epoch": 1.5928396937162566,
+      "grad_norm": 0.28210806545171774,
+      "learning_rate": 7.008370388788496e-06,
+      "loss": 0.4424,
+      "step": 9699
+    },
+    {
+      "epoch": 1.5930039209245992,
+      "grad_norm": 0.33788709073616296,
+      "learning_rate": 7.007981018698415e-06,
+      "loss": 0.4705,
+      "step": 9700
+    },
+    {
+      "epoch": 1.5931681481329418,
+      "grad_norm": 0.2810052632243497,
+      "learning_rate": 7.007591621369193e-06,
+      "loss": 0.4618,
+      "step": 9701
+    },
+    {
+      "epoch": 1.5933323753412847,
+      "grad_norm": 0.3861673845526481,
+      "learning_rate": 7.00720219680506e-06,
+      "loss": 0.453,
+      "step": 9702
+    },
+    {
+      "epoch": 1.5934966025496275,
+      "grad_norm": 0.29160915803694276,
+      "learning_rate": 7.006812745010243e-06,
+      "loss": 0.4733,
+      "step": 9703
+    },
+    {
+      "epoch": 1.5936608297579702,
+      "grad_norm": 0.33914569659292165,
+      "learning_rate": 7.006423265988972e-06,
+      "loss": 0.4515,
+      "step": 9704
+    },
+    {
+      "epoch": 1.5938250569663128,
+      "grad_norm": 0.2550988195503919,
+      "learning_rate": 7.006033759745481e-06,
+      "loss": 0.4619,
+      "step": 9705
+    },
+    {
+      "epoch": 1.5939892841746557,
+      "grad_norm": 0.29945375984337574,
+      "learning_rate": 7.005644226283997e-06,
+      "loss": 0.4713,
+      "step": 9706
+    },
+    {
+      "epoch": 1.5941535113829983,
+      "grad_norm": 0.25626086854928126,
+      "learning_rate": 7.005254665608751e-06,
+      "loss": 0.4694,
+      "step": 9707
+    },
+    {
+      "epoch": 1.5943177385913412,
+      "grad_norm": 0.2478955917014123,
+      "learning_rate": 7.004865077723974e-06,
+      "loss": 0.4469,
+      "step": 9708
+    },
+    {
+      "epoch": 1.5944819657996838,
+      "grad_norm": 0.261073698927775,
+      "learning_rate": 7.0044754626339e-06,
+      "loss": 0.4424,
+      "step": 9709
+    },
+    {
+      "epoch": 1.5946461930080265,
+      "grad_norm": 0.29944897321085306,
+      "learning_rate": 7.0040858203427555e-06,
+      "loss": 0.4598,
+      "step": 9710
+    },
+    {
+      "epoch": 1.5948104202163693,
+      "grad_norm": 0.3029852969342873,
+      "learning_rate": 7.003696150854777e-06,
+      "loss": 0.4568,
+      "step": 9711
+    },
+    {
+      "epoch": 1.5949746474247122,
+      "grad_norm": 0.3545481979767949,
+      "learning_rate": 7.003306454174195e-06,
+      "loss": 0.4739,
+      "step": 9712
+    },
+    {
+      "epoch": 1.5951388746330548,
+      "grad_norm": 0.2915600645900402,
+      "learning_rate": 7.002916730305242e-06,
+      "loss": 0.4731,
+      "step": 9713
+    },
+    {
+      "epoch": 1.5953031018413975,
+      "grad_norm": 0.30375434938571066,
+      "learning_rate": 7.00252697925215e-06,
+      "loss": 0.4594,
+      "step": 9714
+    },
+    {
+      "epoch": 1.5954673290497403,
+      "grad_norm": 0.3858113833425676,
+      "learning_rate": 7.002137201019153e-06,
+      "loss": 0.4711,
+      "step": 9715
+    },
+    {
+      "epoch": 1.5956315562580832,
+      "grad_norm": 0.306509544782491,
+      "learning_rate": 7.001747395610485e-06,
+      "loss": 0.4478,
+      "step": 9716
+    },
+    {
+      "epoch": 1.5957957834664258,
+      "grad_norm": 0.3747588996266856,
+      "learning_rate": 7.001357563030378e-06,
+      "loss": 0.4668,
+      "step": 9717
+    },
+    {
+      "epoch": 1.5959600106747684,
+      "grad_norm": 0.24590682956768997,
+      "learning_rate": 7.000967703283067e-06,
+      "loss": 0.4767,
+      "step": 9718
+    },
+    {
+      "epoch": 1.5961242378831113,
+      "grad_norm": 0.389786665880499,
+      "learning_rate": 7.000577816372787e-06,
+      "loss": 0.4499,
+      "step": 9719
+    },
+    {
+      "epoch": 1.5962884650914542,
+      "grad_norm": 0.3980182584891584,
+      "learning_rate": 7.0001879023037704e-06,
+      "loss": 0.4691,
+      "step": 9720
+    },
+    {
+      "epoch": 1.5964526922997968,
+      "grad_norm": 1.5256900334767118,
+      "learning_rate": 6.999797961080255e-06,
+      "loss": 0.4411,
+      "step": 9721
+    },
+    {
+      "epoch": 1.5966169195081394,
+      "grad_norm": 0.341679700793873,
+      "learning_rate": 6.999407992706472e-06,
+      "loss": 0.4636,
+      "step": 9722
+    },
+    {
+      "epoch": 1.5967811467164823,
+      "grad_norm": 0.3189785275812659,
+      "learning_rate": 6.999017997186662e-06,
+      "loss": 0.4502,
+      "step": 9723
+    },
+    {
+      "epoch": 1.596945373924825,
+      "grad_norm": 0.3096324528487003,
+      "learning_rate": 6.998627974525056e-06,
+      "loss": 0.4658,
+      "step": 9724
+    },
+    {
+      "epoch": 1.5971096011331678,
+      "grad_norm": 0.32705409411953407,
+      "learning_rate": 6.998237924725891e-06,
+      "loss": 0.473,
+      "step": 9725
+    },
+    {
+      "epoch": 1.5972738283415104,
+      "grad_norm": 0.2877085709462597,
+      "learning_rate": 6.997847847793406e-06,
+      "loss": 0.4643,
+      "step": 9726
+    },
+    {
+      "epoch": 1.597438055549853,
+      "grad_norm": 0.3923418706541653,
+      "learning_rate": 6.997457743731836e-06,
+      "loss": 0.4555,
+      "step": 9727
+    },
+    {
+      "epoch": 1.597602282758196,
+      "grad_norm": 0.3107603017489276,
+      "learning_rate": 6.997067612545416e-06,
+      "loss": 0.4443,
+      "step": 9728
+    },
+    {
+      "epoch": 1.5977665099665388,
+      "grad_norm": 0.32297011284910904,
+      "learning_rate": 6.996677454238386e-06,
+      "loss": 0.4674,
+      "step": 9729
+    },
+    {
+      "epoch": 1.5979307371748814,
+      "grad_norm": 0.31297648005174855,
+      "learning_rate": 6.996287268814981e-06,
+      "loss": 0.4502,
+      "step": 9730
+    },
+    {
+      "epoch": 1.598094964383224,
+      "grad_norm": 0.33511246862645244,
+      "learning_rate": 6.9958970562794435e-06,
+      "loss": 0.4729,
+      "step": 9731
+    },
+    {
+      "epoch": 1.598259191591567,
+      "grad_norm": 0.34563244047173136,
+      "learning_rate": 6.995506816636005e-06,
+      "loss": 0.4593,
+      "step": 9732
+    },
+    {
+      "epoch": 1.5984234187999098,
+      "grad_norm": 0.2745624451603022,
+      "learning_rate": 6.9951165498889085e-06,
+      "loss": 0.4536,
+      "step": 9733
+    },
+    {
+      "epoch": 1.5985876460082524,
+      "grad_norm": 0.3344724710839866,
+      "learning_rate": 6.994726256042392e-06,
+      "loss": 0.4574,
+      "step": 9734
+    },
+    {
+      "epoch": 1.598751873216595,
+      "grad_norm": 0.9641068009989658,
+      "learning_rate": 6.994335935100693e-06,
+      "loss": 0.4694,
+      "step": 9735
+    },
+    {
+      "epoch": 1.598916100424938,
+      "grad_norm": 0.32579840476566196,
+      "learning_rate": 6.993945587068053e-06,
+      "loss": 0.4849,
+      "step": 9736
+    },
+    {
+      "epoch": 1.5990803276332808,
+      "grad_norm": 0.33205921777574166,
+      "learning_rate": 6.993555211948709e-06,
+      "loss": 0.4692,
+      "step": 9737
+    },
+    {
+      "epoch": 1.5992445548416234,
+      "grad_norm": 0.2770176558361628,
+      "learning_rate": 6.993164809746901e-06,
+      "loss": 0.4531,
+      "step": 9738
+    },
+    {
+      "epoch": 1.599408782049966,
+      "grad_norm": 0.2972289466677687,
+      "learning_rate": 6.992774380466872e-06,
+      "loss": 0.4647,
+      "step": 9739
+    },
+    {
+      "epoch": 1.599573009258309,
+      "grad_norm": 0.5161134035043563,
+      "learning_rate": 6.99238392411286e-06,
+      "loss": 0.4444,
+      "step": 9740
+    },
+    {
+      "epoch": 1.5997372364666516,
+      "grad_norm": 0.3632176478786719,
+      "learning_rate": 6.991993440689107e-06,
+      "loss": 0.4536,
+      "step": 9741
+    },
+    {
+      "epoch": 1.5999014636749944,
+      "grad_norm": 0.3972599771589535,
+      "learning_rate": 6.991602930199853e-06,
+      "loss": 0.4565,
+      "step": 9742
+    },
+    {
+      "epoch": 1.600065690883337,
+      "grad_norm": 0.34852573709915075,
+      "learning_rate": 6.991212392649341e-06,
+      "loss": 0.4395,
+      "step": 9743
+    },
+    {
+      "epoch": 1.6002299180916797,
+      "grad_norm": 0.378242100179647,
+      "learning_rate": 6.990821828041809e-06,
+      "loss": 0.4815,
+      "step": 9744
+    },
+    {
+      "epoch": 1.6003941453000226,
+      "grad_norm": 0.2997246411725691,
+      "learning_rate": 6.990431236381503e-06,
+      "loss": 0.4667,
+      "step": 9745
+    },
+    {
+      "epoch": 1.6005583725083654,
+      "grad_norm": 0.2928757734583082,
+      "learning_rate": 6.990040617672663e-06,
+      "loss": 0.464,
+      "step": 9746
+    },
+    {
+      "epoch": 1.600722599716708,
+      "grad_norm": 0.2671624667674582,
+      "learning_rate": 6.989649971919531e-06,
+      "loss": 0.4643,
+      "step": 9747
+    },
+    {
+      "epoch": 1.6008868269250507,
+      "grad_norm": 0.2615783823407928,
+      "learning_rate": 6.989259299126353e-06,
+      "loss": 0.45,
+      "step": 9748
+    },
+    {
+      "epoch": 1.6010510541333935,
+      "grad_norm": 0.31506464269600354,
+      "learning_rate": 6.988868599297368e-06,
+      "loss": 0.456,
+      "step": 9749
+    },
+    {
+      "epoch": 1.6012152813417364,
+      "grad_norm": 0.2857207363724411,
+      "learning_rate": 6.988477872436822e-06,
+      "loss": 0.4596,
+      "step": 9750
+    },
+    {
+      "epoch": 1.601379508550079,
+      "grad_norm": 0.3117589736533758,
+      "learning_rate": 6.988087118548959e-06,
+      "loss": 0.4498,
+      "step": 9751
+    },
+    {
+      "epoch": 1.6015437357584217,
+      "grad_norm": 0.26792546533856765,
+      "learning_rate": 6.9876963376380206e-06,
+      "loss": 0.4418,
+      "step": 9752
+    },
+    {
+      "epoch": 1.6017079629667645,
+      "grad_norm": 0.3510529580662092,
+      "learning_rate": 6.987305529708253e-06,
+      "loss": 0.4754,
+      "step": 9753
+    },
+    {
+      "epoch": 1.6018721901751074,
+      "grad_norm": 0.34665871073162313,
+      "learning_rate": 6.986914694763899e-06,
+      "loss": 0.4495,
+      "step": 9754
+    },
+    {
+      "epoch": 1.60203641738345,
+      "grad_norm": 0.2704560117494492,
+      "learning_rate": 6.986523832809207e-06,
+      "loss": 0.4563,
+      "step": 9755
+    },
+    {
+      "epoch": 1.6022006445917927,
+      "grad_norm": 0.39103153382891753,
+      "learning_rate": 6.986132943848418e-06,
+      "loss": 0.4657,
+      "step": 9756
+    },
+    {
+      "epoch": 1.6023648718001355,
+      "grad_norm": 0.27479927593805215,
+      "learning_rate": 6.985742027885779e-06,
+      "loss": 0.4703,
+      "step": 9757
+    },
+    {
+      "epoch": 1.6025290990084782,
+      "grad_norm": 0.34785620793652167,
+      "learning_rate": 6.985351084925537e-06,
+      "loss": 0.4459,
+      "step": 9758
+    },
+    {
+      "epoch": 1.602693326216821,
+      "grad_norm": 0.3060687588327717,
+      "learning_rate": 6.984960114971936e-06,
+      "loss": 0.4774,
+      "step": 9759
+    },
+    {
+      "epoch": 1.6028575534251637,
+      "grad_norm": 0.37563535546956034,
+      "learning_rate": 6.9845691180292235e-06,
+      "loss": 0.4805,
+      "step": 9760
+    },
+    {
+      "epoch": 1.6030217806335063,
+      "grad_norm": 0.24377779296879026,
+      "learning_rate": 6.984178094101647e-06,
+      "loss": 0.4809,
+      "step": 9761
+    },
+    {
+      "epoch": 1.6031860078418492,
+      "grad_norm": 0.2571146440045776,
+      "learning_rate": 6.983787043193452e-06,
+      "loss": 0.4634,
+      "step": 9762
+    },
+    {
+      "epoch": 1.603350235050192,
+      "grad_norm": 0.3778507318993349,
+      "learning_rate": 6.983395965308885e-06,
+      "loss": 0.4464,
+      "step": 9763
+    },
+    {
+      "epoch": 1.6035144622585347,
+      "grad_norm": 0.30104596005073675,
+      "learning_rate": 6.983004860452195e-06,
+      "loss": 0.463,
+      "step": 9764
+    },
+    {
+      "epoch": 1.6036786894668773,
+      "grad_norm": 0.29668588814681535,
+      "learning_rate": 6.982613728627629e-06,
+      "loss": 0.4557,
+      "step": 9765
+    },
+    {
+      "epoch": 1.6038429166752202,
+      "grad_norm": 0.4248514538815655,
+      "learning_rate": 6.982222569839436e-06,
+      "loss": 0.4238,
+      "step": 9766
+    },
+    {
+      "epoch": 1.604007143883563,
+      "grad_norm": 0.36176526973408535,
+      "learning_rate": 6.981831384091863e-06,
+      "loss": 0.4639,
+      "step": 9767
+    },
+    {
+      "epoch": 1.6041713710919057,
+      "grad_norm": 0.3367393011027692,
+      "learning_rate": 6.981440171389158e-06,
+      "loss": 0.4729,
+      "step": 9768
+    },
+    {
+      "epoch": 1.6043355983002483,
+      "grad_norm": 0.29669673074478037,
+      "learning_rate": 6.981048931735574e-06,
+      "loss": 0.4804,
+      "step": 9769
+    },
+    {
+      "epoch": 1.6044998255085912,
+      "grad_norm": 0.2980190900348066,
+      "learning_rate": 6.980657665135357e-06,
+      "loss": 0.4661,
+      "step": 9770
+    },
+    {
+      "epoch": 1.604664052716934,
+      "grad_norm": 0.332201811865359,
+      "learning_rate": 6.980266371592756e-06,
+      "loss": 0.4418,
+      "step": 9771
+    },
+    {
+      "epoch": 1.6048282799252767,
+      "grad_norm": 0.40769861897148146,
+      "learning_rate": 6.979875051112023e-06,
+      "loss": 0.4509,
+      "step": 9772
+    },
+    {
+      "epoch": 1.6049925071336193,
+      "grad_norm": 0.28501944575324484,
+      "learning_rate": 6.979483703697408e-06,
+      "loss": 0.4693,
+      "step": 9773
+    },
+    {
+      "epoch": 1.6051567343419622,
+      "grad_norm": 0.2760017459121368,
+      "learning_rate": 6.979092329353159e-06,
+      "loss": 0.4627,
+      "step": 9774
+    },
+    {
+      "epoch": 1.6053209615503048,
+      "grad_norm": 0.3752488518830528,
+      "learning_rate": 6.978700928083527e-06,
+      "loss": 0.4763,
+      "step": 9775
+    },
+    {
+      "epoch": 1.6054851887586477,
+      "grad_norm": 0.3251274051588068,
+      "learning_rate": 6.9783094998927655e-06,
+      "loss": 0.4671,
+      "step": 9776
+    },
+    {
+      "epoch": 1.6056494159669903,
+      "grad_norm": 0.36151954149272725,
+      "learning_rate": 6.977918044785125e-06,
+      "loss": 0.453,
+      "step": 9777
+    },
+    {
+      "epoch": 1.605813643175333,
+      "grad_norm": 0.37377513986589384,
+      "learning_rate": 6.9775265627648565e-06,
+      "loss": 0.4644,
+      "step": 9778
+    },
+    {
+      "epoch": 1.6059778703836758,
+      "grad_norm": 0.28806517291802397,
+      "learning_rate": 6.977135053836211e-06,
+      "loss": 0.4641,
+      "step": 9779
+    },
+    {
+      "epoch": 1.6061420975920186,
+      "grad_norm": 0.25543690612044667,
+      "learning_rate": 6.976743518003443e-06,
+      "loss": 0.4595,
+      "step": 9780
+    },
+    {
+      "epoch": 1.6063063248003613,
+      "grad_norm": 0.2927473928733643,
+      "learning_rate": 6.976351955270803e-06,
+      "loss": 0.4852,
+      "step": 9781
+    },
+    {
+      "epoch": 1.606470552008704,
+      "grad_norm": 0.32663703524442445,
+      "learning_rate": 6.975960365642544e-06,
+      "loss": 0.4828,
+      "step": 9782
+    },
+    {
+      "epoch": 1.6066347792170468,
+      "grad_norm": 0.33126405619204863,
+      "learning_rate": 6.9755687491229195e-06,
+      "loss": 0.4574,
+      "step": 9783
+    },
+    {
+      "epoch": 1.6067990064253896,
+      "grad_norm": 0.27115302052887735,
+      "learning_rate": 6.975177105716184e-06,
+      "loss": 0.4554,
+      "step": 9784
+    },
+    {
+      "epoch": 1.6069632336337323,
+      "grad_norm": 0.5938177654376247,
+      "learning_rate": 6.974785435426588e-06,
+      "loss": 0.4748,
+      "step": 9785
+    },
+    {
+      "epoch": 1.607127460842075,
+      "grad_norm": 0.3709570909137645,
+      "learning_rate": 6.974393738258388e-06,
+      "loss": 0.4737,
+      "step": 9786
+    },
+    {
+      "epoch": 1.6072916880504178,
+      "grad_norm": 0.29217409026482316,
+      "learning_rate": 6.974002014215839e-06,
+      "loss": 0.4594,
+      "step": 9787
+    },
+    {
+      "epoch": 1.6074559152587606,
+      "grad_norm": 0.5522762635217223,
+      "learning_rate": 6.973610263303191e-06,
+      "loss": 0.4405,
+      "step": 9788
+    },
+    {
+      "epoch": 1.6076201424671033,
+      "grad_norm": 0.2700062333742084,
+      "learning_rate": 6.973218485524704e-06,
+      "loss": 0.4527,
+      "step": 9789
+    },
+    {
+      "epoch": 1.607784369675446,
+      "grad_norm": 0.2752977343062548,
+      "learning_rate": 6.972826680884631e-06,
+      "loss": 0.4579,
+      "step": 9790
+    },
+    {
+      "epoch": 1.6079485968837888,
+      "grad_norm": 0.351632135170519,
+      "learning_rate": 6.972434849387226e-06,
+      "loss": 0.4488,
+      "step": 9791
+    },
+    {
+      "epoch": 1.6081128240921314,
+      "grad_norm": 0.30396405938296145,
+      "learning_rate": 6.972042991036747e-06,
+      "loss": 0.4552,
+      "step": 9792
+    },
+    {
+      "epoch": 1.6082770513004743,
+      "grad_norm": 0.2923401297274902,
+      "learning_rate": 6.971651105837449e-06,
+      "loss": 0.4671,
+      "step": 9793
+    },
+    {
+      "epoch": 1.608441278508817,
+      "grad_norm": 0.36307972078813605,
+      "learning_rate": 6.971259193793588e-06,
+      "loss": 0.4382,
+      "step": 9794
+    },
+    {
+      "epoch": 1.6086055057171595,
+      "grad_norm": 0.31702386181625597,
+      "learning_rate": 6.970867254909422e-06,
+      "loss": 0.4675,
+      "step": 9795
+    },
+    {
+      "epoch": 1.6087697329255024,
+      "grad_norm": 0.3142731818838145,
+      "learning_rate": 6.970475289189204e-06,
+      "loss": 0.459,
+      "step": 9796
+    },
+    {
+      "epoch": 1.6089339601338453,
+      "grad_norm": 0.36241745736101644,
+      "learning_rate": 6.970083296637196e-06,
+      "loss": 0.4748,
+      "step": 9797
+    },
+    {
+      "epoch": 1.609098187342188,
+      "grad_norm": 0.3470417056667347,
+      "learning_rate": 6.969691277257652e-06,
+      "loss": 0.469,
+      "step": 9798
+    },
+    {
+      "epoch": 1.6092624145505305,
+      "grad_norm": 0.41937451840307666,
+      "learning_rate": 6.969299231054831e-06,
+      "loss": 0.4564,
+      "step": 9799
+    },
+    {
+      "epoch": 1.6094266417588734,
+      "grad_norm": 0.3262573052858504,
+      "learning_rate": 6.9689071580329905e-06,
+      "loss": 0.4587,
+      "step": 9800
+    },
+    {
+      "epoch": 1.6095908689672163,
+      "grad_norm": 0.2947421917980309,
+      "learning_rate": 6.96851505819639e-06,
+      "loss": 0.4533,
+      "step": 9801
+    },
+    {
+      "epoch": 1.609755096175559,
+      "grad_norm": 0.2521016894403766,
+      "learning_rate": 6.968122931549288e-06,
+      "loss": 0.4593,
+      "step": 9802
+    },
+    {
+      "epoch": 1.6099193233839015,
+      "grad_norm": 0.2955673416323675,
+      "learning_rate": 6.9677307780959416e-06,
+      "loss": 0.4666,
+      "step": 9803
+    },
+    {
+      "epoch": 1.6100835505922444,
+      "grad_norm": 0.27659905950201386,
+      "learning_rate": 6.96733859784061e-06,
+      "loss": 0.4489,
+      "step": 9804
+    },
+    {
+      "epoch": 1.6102477778005873,
+      "grad_norm": 0.3404542615909959,
+      "learning_rate": 6.966946390787554e-06,
+      "loss": 0.4602,
+      "step": 9805
+    },
+    {
+      "epoch": 1.61041200500893,
+      "grad_norm": 0.2845051029226587,
+      "learning_rate": 6.9665541569410355e-06,
+      "loss": 0.4692,
+      "step": 9806
+    },
+    {
+      "epoch": 1.6105762322172725,
+      "grad_norm": 0.38260420051370825,
+      "learning_rate": 6.96616189630531e-06,
+      "loss": 0.4759,
+      "step": 9807
+    },
+    {
+      "epoch": 1.6107404594256154,
+      "grad_norm": 0.26683465973181375,
+      "learning_rate": 6.96576960888464e-06,
+      "loss": 0.4739,
+      "step": 9808
+    },
+    {
+      "epoch": 1.610904686633958,
+      "grad_norm": 0.384929446621508,
+      "learning_rate": 6.965377294683286e-06,
+      "loss": 0.438,
+      "step": 9809
+    },
+    {
+      "epoch": 1.6110689138423009,
+      "grad_norm": 0.38374467343084134,
+      "learning_rate": 6.964984953705509e-06,
+      "loss": 0.4729,
+      "step": 9810
+    },
+    {
+      "epoch": 1.6112331410506435,
+      "grad_norm": 0.2694526018583905,
+      "learning_rate": 6.964592585955571e-06,
+      "loss": 0.4557,
+      "step": 9811
+    },
+    {
+      "epoch": 1.6113973682589862,
+      "grad_norm": 0.3115303986468197,
+      "learning_rate": 6.964200191437732e-06,
+      "loss": 0.4506,
+      "step": 9812
+    },
+    {
+      "epoch": 1.611561595467329,
+      "grad_norm": 0.3035293723273839,
+      "learning_rate": 6.963807770156254e-06,
+      "loss": 0.4641,
+      "step": 9813
+    },
+    {
+      "epoch": 1.6117258226756719,
+      "grad_norm": 0.2992017924804873,
+      "learning_rate": 6.963415322115402e-06,
+      "loss": 0.4536,
+      "step": 9814
+    },
+    {
+      "epoch": 1.6118900498840145,
+      "grad_norm": 0.31667569525527806,
+      "learning_rate": 6.963022847319434e-06,
+      "loss": 0.4551,
+      "step": 9815
+    },
+    {
+      "epoch": 1.6120542770923572,
+      "grad_norm": 0.2540364674293299,
+      "learning_rate": 6.962630345772615e-06,
+      "loss": 0.4708,
+      "step": 9816
+    },
+    {
+      "epoch": 1.6122185043007,
+      "grad_norm": 0.4442076282948899,
+      "learning_rate": 6.962237817479207e-06,
+      "loss": 0.4743,
+      "step": 9817
+    },
+    {
+      "epoch": 1.6123827315090429,
+      "grad_norm": 0.2903014811736007,
+      "learning_rate": 6.961845262443474e-06,
+      "loss": 0.4645,
+      "step": 9818
+    },
+    {
+      "epoch": 1.6125469587173855,
+      "grad_norm": 0.36913889456903465,
+      "learning_rate": 6.96145268066968e-06,
+      "loss": 0.4619,
+      "step": 9819
+    },
+    {
+      "epoch": 1.6127111859257282,
+      "grad_norm": 0.4006840087480322,
+      "learning_rate": 6.961060072162087e-06,
+      "loss": 0.4707,
+      "step": 9820
+    },
+    {
+      "epoch": 1.612875413134071,
+      "grad_norm": 0.3353180038710587,
+      "learning_rate": 6.960667436924961e-06,
+      "loss": 0.4773,
+      "step": 9821
+    },
+    {
+      "epoch": 1.6130396403424139,
+      "grad_norm": 0.28696136203207506,
+      "learning_rate": 6.960274774962565e-06,
+      "loss": 0.4412,
+      "step": 9822
+    },
+    {
+      "epoch": 1.6132038675507565,
+      "grad_norm": 0.2864030214755703,
+      "learning_rate": 6.959882086279166e-06,
+      "loss": 0.4622,
+      "step": 9823
+    },
+    {
+      "epoch": 1.6133680947590991,
+      "grad_norm": 0.2985379366016787,
+      "learning_rate": 6.959489370879026e-06,
+      "loss": 0.4611,
+      "step": 9824
+    },
+    {
+      "epoch": 1.613532321967442,
+      "grad_norm": 0.28314934992789903,
+      "learning_rate": 6.959096628766411e-06,
+      "loss": 0.4729,
+      "step": 9825
+    },
+    {
+      "epoch": 1.6136965491757846,
+      "grad_norm": 0.38930289972293297,
+      "learning_rate": 6.9587038599455874e-06,
+      "loss": 0.4515,
+      "step": 9826
+    },
+    {
+      "epoch": 1.6138607763841275,
+      "grad_norm": 0.787582127701228,
+      "learning_rate": 6.958311064420822e-06,
+      "loss": 0.4569,
+      "step": 9827
+    },
+    {
+      "epoch": 1.6140250035924701,
+      "grad_norm": 0.27450397102919527,
+      "learning_rate": 6.957918242196379e-06,
+      "loss": 0.4619,
+      "step": 9828
+    },
+    {
+      "epoch": 1.6141892308008128,
+      "grad_norm": 0.4634130614046554,
+      "learning_rate": 6.957525393276526e-06,
+      "loss": 0.4574,
+      "step": 9829
+    },
+    {
+      "epoch": 1.6143534580091556,
+      "grad_norm": 0.32687976750827596,
+      "learning_rate": 6.957132517665529e-06,
+      "loss": 0.4478,
+      "step": 9830
+    },
+    {
+      "epoch": 1.6145176852174985,
+      "grad_norm": 0.2749625418731786,
+      "learning_rate": 6.9567396153676556e-06,
+      "loss": 0.4787,
+      "step": 9831
+    },
+    {
+      "epoch": 1.6146819124258411,
+      "grad_norm": 0.2686148625927595,
+      "learning_rate": 6.956346686387174e-06,
+      "loss": 0.4641,
+      "step": 9832
+    },
+    {
+      "epoch": 1.6148461396341838,
+      "grad_norm": 0.27749373643361147,
+      "learning_rate": 6.955953730728349e-06,
+      "loss": 0.4478,
+      "step": 9833
+    },
+    {
+      "epoch": 1.6150103668425266,
+      "grad_norm": 0.33035669372922283,
+      "learning_rate": 6.95556074839545e-06,
+      "loss": 0.4574,
+      "step": 9834
+    },
+    {
+      "epoch": 1.6151745940508695,
+      "grad_norm": 0.319743000157342,
+      "learning_rate": 6.955167739392747e-06,
+      "loss": 0.4471,
+      "step": 9835
+    },
+    {
+      "epoch": 1.6153388212592121,
+      "grad_norm": 0.26117931070035566,
+      "learning_rate": 6.954774703724506e-06,
+      "loss": 0.4639,
+      "step": 9836
+    },
+    {
+      "epoch": 1.6155030484675548,
+      "grad_norm": 0.27760241909976285,
+      "learning_rate": 6.9543816413949965e-06,
+      "loss": 0.4572,
+      "step": 9837
+    },
+    {
+      "epoch": 1.6156672756758976,
+      "grad_norm": 0.2998866914995418,
+      "learning_rate": 6.953988552408487e-06,
+      "loss": 0.4551,
+      "step": 9838
+    },
+    {
+      "epoch": 1.6158315028842405,
+      "grad_norm": 0.30021599310248603,
+      "learning_rate": 6.953595436769248e-06,
+      "loss": 0.4615,
+      "step": 9839
+    },
+    {
+      "epoch": 1.6159957300925831,
+      "grad_norm": 0.31256899332836824,
+      "learning_rate": 6.953202294481548e-06,
+      "loss": 0.4541,
+      "step": 9840
+    },
+    {
+      "epoch": 1.6161599573009258,
+      "grad_norm": 0.3899196834494022,
+      "learning_rate": 6.952809125549658e-06,
+      "loss": 0.4771,
+      "step": 9841
+    },
+    {
+      "epoch": 1.6163241845092686,
+      "grad_norm": 0.3010082901322875,
+      "learning_rate": 6.952415929977848e-06,
+      "loss": 0.4719,
+      "step": 9842
+    },
+    {
+      "epoch": 1.6164884117176113,
+      "grad_norm": 0.3576168449635639,
+      "learning_rate": 6.952022707770387e-06,
+      "loss": 0.4404,
+      "step": 9843
+    },
+    {
+      "epoch": 1.6166526389259541,
+      "grad_norm": 0.2760702244129268,
+      "learning_rate": 6.951629458931548e-06,
+      "loss": 0.4578,
+      "step": 9844
+    },
+    {
+      "epoch": 1.6168168661342968,
+      "grad_norm": 0.2749007165277947,
+      "learning_rate": 6.951236183465601e-06,
+      "loss": 0.45,
+      "step": 9845
+    },
+    {
+      "epoch": 1.6169810933426394,
+      "grad_norm": 0.31390505673084157,
+      "learning_rate": 6.950842881376816e-06,
+      "loss": 0.4495,
+      "step": 9846
+    },
+    {
+      "epoch": 1.6171453205509823,
+      "grad_norm": 0.2868629433154881,
+      "learning_rate": 6.9504495526694675e-06,
+      "loss": 0.4534,
+      "step": 9847
+    },
+    {
+      "epoch": 1.6173095477593251,
+      "grad_norm": 0.312588550829176,
+      "learning_rate": 6.9500561973478264e-06,
+      "loss": 0.4478,
+      "step": 9848
+    },
+    {
+      "epoch": 1.6174737749676678,
+      "grad_norm": 0.2958477322250829,
+      "learning_rate": 6.949662815416163e-06,
+      "loss": 0.464,
+      "step": 9849
+    },
+    {
+      "epoch": 1.6176380021760104,
+      "grad_norm": 0.29871300762114433,
+      "learning_rate": 6.949269406878752e-06,
+      "loss": 0.4636,
+      "step": 9850
+    },
+    {
+      "epoch": 1.6178022293843533,
+      "grad_norm": 0.40407588926968097,
+      "learning_rate": 6.9488759717398645e-06,
+      "loss": 0.4673,
+      "step": 9851
+    },
+    {
+      "epoch": 1.6179664565926961,
+      "grad_norm": 0.27940144990316745,
+      "learning_rate": 6.948482510003776e-06,
+      "loss": 0.4671,
+      "step": 9852
+    },
+    {
+      "epoch": 1.6181306838010387,
+      "grad_norm": 0.30202586554252114,
+      "learning_rate": 6.948089021674758e-06,
+      "loss": 0.4433,
+      "step": 9853
+    },
+    {
+      "epoch": 1.6182949110093814,
+      "grad_norm": 0.285583505265655,
+      "learning_rate": 6.947695506757084e-06,
+      "loss": 0.4786,
+      "step": 9854
+    },
+    {
+      "epoch": 1.6184591382177242,
+      "grad_norm": 0.39905382824632013,
+      "learning_rate": 6.947301965255029e-06,
+      "loss": 0.4724,
+      "step": 9855
+    },
+    {
+      "epoch": 1.618623365426067,
+      "grad_norm": 0.307863136962308,
+      "learning_rate": 6.946908397172866e-06,
+      "loss": 0.4232,
+      "step": 9856
+    },
+    {
+      "epoch": 1.6187875926344097,
+      "grad_norm": 0.43057042653410355,
+      "learning_rate": 6.946514802514872e-06,
+      "loss": 0.451,
+      "step": 9857
+    },
+    {
+      "epoch": 1.6189518198427524,
+      "grad_norm": 0.3171082589247781,
+      "learning_rate": 6.94612118128532e-06,
+      "loss": 0.4396,
+      "step": 9858
+    },
+    {
+      "epoch": 1.6191160470510952,
+      "grad_norm": 0.29330682282221376,
+      "learning_rate": 6.945727533488483e-06,
+      "loss": 0.4546,
+      "step": 9859
+    },
+    {
+      "epoch": 1.6192802742594379,
+      "grad_norm": 0.2870797937588179,
+      "learning_rate": 6.945333859128642e-06,
+      "loss": 0.4634,
+      "step": 9860
+    },
+    {
+      "epoch": 1.6194445014677807,
+      "grad_norm": 0.3051040740644912,
+      "learning_rate": 6.944940158210067e-06,
+      "loss": 0.4536,
+      "step": 9861
+    },
+    {
+      "epoch": 1.6196087286761234,
+      "grad_norm": 0.41361444535474173,
+      "learning_rate": 6.9445464307370375e-06,
+      "loss": 0.458,
+      "step": 9862
+    },
+    {
+      "epoch": 1.619772955884466,
+      "grad_norm": 0.25243414567729117,
+      "learning_rate": 6.944152676713828e-06,
+      "loss": 0.4663,
+      "step": 9863
+    },
+    {
+      "epoch": 1.6199371830928089,
+      "grad_norm": 0.3815051393474824,
+      "learning_rate": 6.943758896144715e-06,
+      "loss": 0.444,
+      "step": 9864
+    },
+    {
+      "epoch": 1.6201014103011517,
+      "grad_norm": 0.38218490759074814,
+      "learning_rate": 6.943365089033979e-06,
+      "loss": 0.4714,
+      "step": 9865
+    },
+    {
+      "epoch": 1.6202656375094944,
+      "grad_norm": 0.28814253453168126,
+      "learning_rate": 6.942971255385893e-06,
+      "loss": 0.4753,
+      "step": 9866
+    },
+    {
+      "epoch": 1.620429864717837,
+      "grad_norm": 0.337629238841208,
+      "learning_rate": 6.9425773952047335e-06,
+      "loss": 0.4506,
+      "step": 9867
+    },
+    {
+      "epoch": 1.6205940919261799,
+      "grad_norm": 0.25032929018512107,
+      "learning_rate": 6.942183508494782e-06,
+      "loss": 0.4624,
+      "step": 9868
+    },
+    {
+      "epoch": 1.6207583191345227,
+      "grad_norm": 0.28775404020259854,
+      "learning_rate": 6.941789595260315e-06,
+      "loss": 0.4854,
+      "step": 9869
+    },
+    {
+      "epoch": 1.6209225463428654,
+      "grad_norm": 0.28379877031843614,
+      "learning_rate": 6.941395655505611e-06,
+      "loss": 0.4747,
+      "step": 9870
+    },
+    {
+      "epoch": 1.621086773551208,
+      "grad_norm": 0.8202653712403721,
+      "learning_rate": 6.941001689234946e-06,
+      "loss": 0.4387,
+      "step": 9871
+    },
+    {
+      "epoch": 1.6212510007595509,
+      "grad_norm": 0.42955934255178224,
+      "learning_rate": 6.940607696452603e-06,
+      "loss": 0.4522,
+      "step": 9872
+    },
+    {
+      "epoch": 1.6214152279678937,
+      "grad_norm": 0.3259530932672982,
+      "learning_rate": 6.940213677162859e-06,
+      "loss": 0.4552,
+      "step": 9873
+    },
+    {
+      "epoch": 1.6215794551762364,
+      "grad_norm": 0.38494136138535234,
+      "learning_rate": 6.939819631369992e-06,
+      "loss": 0.4636,
+      "step": 9874
+    },
+    {
+      "epoch": 1.621743682384579,
+      "grad_norm": 0.2880001166580629,
+      "learning_rate": 6.939425559078285e-06,
+      "loss": 0.4682,
+      "step": 9875
+    },
+    {
+      "epoch": 1.6219079095929219,
+      "grad_norm": 0.2530128315191015,
+      "learning_rate": 6.939031460292016e-06,
+      "loss": 0.4465,
+      "step": 9876
+    },
+    {
+      "epoch": 1.6220721368012645,
+      "grad_norm": 0.29707957324021506,
+      "learning_rate": 6.938637335015466e-06,
+      "loss": 0.4585,
+      "step": 9877
+    },
+    {
+      "epoch": 1.6222363640096074,
+      "grad_norm": 0.2835923151500123,
+      "learning_rate": 6.938243183252914e-06,
+      "loss": 0.4799,
+      "step": 9878
+    },
+    {
+      "epoch": 1.62240059121795,
+      "grad_norm": 0.3185739450507327,
+      "learning_rate": 6.937849005008643e-06,
+      "loss": 0.4754,
+      "step": 9879
+    },
+    {
+      "epoch": 1.6225648184262926,
+      "grad_norm": 0.2816869463340398,
+      "learning_rate": 6.937454800286933e-06,
+      "loss": 0.4432,
+      "step": 9880
+    },
+    {
+      "epoch": 1.6227290456346355,
+      "grad_norm": 0.2856348714662334,
+      "learning_rate": 6.937060569092066e-06,
+      "loss": 0.4584,
+      "step": 9881
+    },
+    {
+      "epoch": 1.6228932728429784,
+      "grad_norm": 0.39852623208599847,
+      "learning_rate": 6.936666311428324e-06,
+      "loss": 0.4808,
+      "step": 9882
+    },
+    {
+      "epoch": 1.623057500051321,
+      "grad_norm": 0.2881186208127396,
+      "learning_rate": 6.936272027299987e-06,
+      "loss": 0.4663,
+      "step": 9883
+    },
+    {
+      "epoch": 1.6232217272596636,
+      "grad_norm": 0.28622926544540417,
+      "learning_rate": 6.935877716711341e-06,
+      "loss": 0.4604,
+      "step": 9884
+    },
+    {
+      "epoch": 1.6233859544680065,
+      "grad_norm": 0.25634085668901097,
+      "learning_rate": 6.935483379666665e-06,
+      "loss": 0.4489,
+      "step": 9885
+    },
+    {
+      "epoch": 1.6235501816763493,
+      "grad_norm": 0.35586491846358337,
+      "learning_rate": 6.935089016170243e-06,
+      "loss": 0.4747,
+      "step": 9886
+    },
+    {
+      "epoch": 1.623714408884692,
+      "grad_norm": 0.45630619008435724,
+      "learning_rate": 6.934694626226359e-06,
+      "loss": 0.4803,
+      "step": 9887
+    },
+    {
+      "epoch": 1.6238786360930346,
+      "grad_norm": 0.414835910725295,
+      "learning_rate": 6.934300209839296e-06,
+      "loss": 0.4619,
+      "step": 9888
+    },
+    {
+      "epoch": 1.6240428633013775,
+      "grad_norm": 0.39488106594808875,
+      "learning_rate": 6.933905767013337e-06,
+      "loss": 0.4804,
+      "step": 9889
+    },
+    {
+      "epoch": 1.6242070905097203,
+      "grad_norm": 0.25434236302766866,
+      "learning_rate": 6.933511297752769e-06,
+      "loss": 0.4511,
+      "step": 9890
+    },
+    {
+      "epoch": 1.624371317718063,
+      "grad_norm": 0.2852061649295925,
+      "learning_rate": 6.93311680206187e-06,
+      "loss": 0.4452,
+      "step": 9891
+    },
+    {
+      "epoch": 1.6245355449264056,
+      "grad_norm": 0.26183435681063855,
+      "learning_rate": 6.932722279944933e-06,
+      "loss": 0.443,
+      "step": 9892
+    },
+    {
+      "epoch": 1.6246997721347485,
+      "grad_norm": 0.3371400824807553,
+      "learning_rate": 6.932327731406235e-06,
+      "loss": 0.4562,
+      "step": 9893
+    },
+    {
+      "epoch": 1.6248639993430911,
+      "grad_norm": 0.259702383047695,
+      "learning_rate": 6.931933156450068e-06,
+      "loss": 0.453,
+      "step": 9894
+    },
+    {
+      "epoch": 1.625028226551434,
+      "grad_norm": 0.2845006703877606,
+      "learning_rate": 6.931538555080712e-06,
+      "loss": 0.4577,
+      "step": 9895
+    },
+    {
+      "epoch": 1.6251924537597766,
+      "grad_norm": 0.2900661570644532,
+      "learning_rate": 6.931143927302455e-06,
+      "loss": 0.4455,
+      "step": 9896
+    },
+    {
+      "epoch": 1.6253566809681192,
+      "grad_norm": 0.2464969379598356,
+      "learning_rate": 6.930749273119583e-06,
+      "loss": 0.4546,
+      "step": 9897
+    },
+    {
+      "epoch": 1.625520908176462,
+      "grad_norm": 0.2571761908465857,
+      "learning_rate": 6.930354592536384e-06,
+      "loss": 0.4719,
+      "step": 9898
+    },
+    {
+      "epoch": 1.625685135384805,
+      "grad_norm": 0.3131139124840701,
+      "learning_rate": 6.929959885557142e-06,
+      "loss": 0.4357,
+      "step": 9899
+    },
+    {
+      "epoch": 1.6258493625931476,
+      "grad_norm": 0.2794163509230121,
+      "learning_rate": 6.929565152186145e-06,
+      "loss": 0.4645,
+      "step": 9900
+    },
+    {
+      "epoch": 1.6260135898014902,
+      "grad_norm": 0.3279839951388422,
+      "learning_rate": 6.9291703924276795e-06,
+      "loss": 0.4668,
+      "step": 9901
+    },
+    {
+      "epoch": 1.626177817009833,
+      "grad_norm": 0.28360673429548217,
+      "learning_rate": 6.928775606286034e-06,
+      "loss": 0.4366,
+      "step": 9902
+    },
+    {
+      "epoch": 1.626342044218176,
+      "grad_norm": 0.26088450799046986,
+      "learning_rate": 6.928380793765495e-06,
+      "loss": 0.4621,
+      "step": 9903
+    },
+    {
+      "epoch": 1.6265062714265186,
+      "grad_norm": 0.31179713297315687,
+      "learning_rate": 6.927985954870352e-06,
+      "loss": 0.4639,
+      "step": 9904
+    },
+    {
+      "epoch": 1.6266704986348612,
+      "grad_norm": 0.31748956718889826,
+      "learning_rate": 6.927591089604894e-06,
+      "loss": 0.4539,
+      "step": 9905
+    },
+    {
+      "epoch": 1.626834725843204,
+      "grad_norm": 0.2803655727982917,
+      "learning_rate": 6.927196197973406e-06,
+      "loss": 0.4618,
+      "step": 9906
+    },
+    {
+      "epoch": 1.626998953051547,
+      "grad_norm": 0.27920456966557644,
+      "learning_rate": 6.926801279980181e-06,
+      "loss": 0.4754,
+      "step": 9907
+    },
+    {
+      "epoch": 1.6271631802598896,
+      "grad_norm": 0.35933877215535787,
+      "learning_rate": 6.926406335629506e-06,
+      "loss": 0.4694,
+      "step": 9908
+    },
+    {
+      "epoch": 1.6273274074682322,
+      "grad_norm": 0.32910657102073315,
+      "learning_rate": 6.926011364925671e-06,
+      "loss": 0.4563,
+      "step": 9909
+    },
+    {
+      "epoch": 1.627491634676575,
+      "grad_norm": 0.28716458893408375,
+      "learning_rate": 6.925616367872966e-06,
+      "loss": 0.4486,
+      "step": 9910
+    },
+    {
+      "epoch": 1.6276558618849177,
+      "grad_norm": 0.3223186286773691,
+      "learning_rate": 6.92522134447568e-06,
+      "loss": 0.4487,
+      "step": 9911
+    },
+    {
+      "epoch": 1.6278200890932606,
+      "grad_norm": 0.28144093144592025,
+      "learning_rate": 6.924826294738104e-06,
+      "loss": 0.4602,
+      "step": 9912
+    },
+    {
+      "epoch": 1.6279843163016032,
+      "grad_norm": 0.30356735564352616,
+      "learning_rate": 6.924431218664529e-06,
+      "loss": 0.4542,
+      "step": 9913
+    },
+    {
+      "epoch": 1.6281485435099459,
+      "grad_norm": 0.27955424056774203,
+      "learning_rate": 6.9240361162592466e-06,
+      "loss": 0.4532,
+      "step": 9914
+    },
+    {
+      "epoch": 1.6283127707182887,
+      "grad_norm": 0.27703469310058093,
+      "learning_rate": 6.923640987526547e-06,
+      "loss": 0.4641,
+      "step": 9915
+    },
+    {
+      "epoch": 1.6284769979266316,
+      "grad_norm": 0.2741045415309227,
+      "learning_rate": 6.92324583247072e-06,
+      "loss": 0.4624,
+      "step": 9916
+    },
+    {
+      "epoch": 1.6286412251349742,
+      "grad_norm": 0.566318151594307,
+      "learning_rate": 6.9228506510960595e-06,
+      "loss": 0.4463,
+      "step": 9917
+    },
+    {
+      "epoch": 1.6288054523433169,
+      "grad_norm": 0.4132857679172401,
+      "learning_rate": 6.922455443406858e-06,
+      "loss": 0.4639,
+      "step": 9918
+    },
+    {
+      "epoch": 1.6289696795516597,
+      "grad_norm": 0.3322732173475824,
+      "learning_rate": 6.922060209407407e-06,
+      "loss": 0.4523,
+      "step": 9919
+    },
+    {
+      "epoch": 1.6291339067600026,
+      "grad_norm": 0.26147560000120657,
+      "learning_rate": 6.921664949102e-06,
+      "loss": 0.432,
+      "step": 9920
+    },
+    {
+      "epoch": 1.6292981339683452,
+      "grad_norm": 0.26057303179752594,
+      "learning_rate": 6.921269662494927e-06,
+      "loss": 0.4612,
+      "step": 9921
+    },
+    {
+      "epoch": 1.6294623611766879,
+      "grad_norm": 0.40536918485757667,
+      "learning_rate": 6.9208743495904846e-06,
+      "loss": 0.4676,
+      "step": 9922
+    },
+    {
+      "epoch": 1.6296265883850307,
+      "grad_norm": 0.2811103795445873,
+      "learning_rate": 6.920479010392964e-06,
+      "loss": 0.4848,
+      "step": 9923
+    },
+    {
+      "epoch": 1.6297908155933736,
+      "grad_norm": 0.28369942619293775,
+      "learning_rate": 6.920083644906659e-06,
+      "loss": 0.4575,
+      "step": 9924
+    },
+    {
+      "epoch": 1.6299550428017162,
+      "grad_norm": 0.9300496984621315,
+      "learning_rate": 6.919688253135867e-06,
+      "loss": 0.4596,
+      "step": 9925
+    },
+    {
+      "epoch": 1.6301192700100589,
+      "grad_norm": 0.629522730790281,
+      "learning_rate": 6.919292835084879e-06,
+      "loss": 0.4461,
+      "step": 9926
+    },
+    {
+      "epoch": 1.6302834972184017,
+      "grad_norm": 0.2516198118246472,
+      "learning_rate": 6.91889739075799e-06,
+      "loss": 0.4709,
+      "step": 9927
+    },
+    {
+      "epoch": 1.6304477244267443,
+      "grad_norm": 0.3026698120551612,
+      "learning_rate": 6.918501920159496e-06,
+      "loss": 0.4531,
+      "step": 9928
+    },
+    {
+      "epoch": 1.6306119516350872,
+      "grad_norm": 0.26043558936113975,
+      "learning_rate": 6.9181064232936926e-06,
+      "loss": 0.454,
+      "step": 9929
+    },
+    {
+      "epoch": 1.6307761788434298,
+      "grad_norm": 0.2549099028558008,
+      "learning_rate": 6.917710900164873e-06,
+      "loss": 0.4673,
+      "step": 9930
+    },
+    {
+      "epoch": 1.6309404060517725,
+      "grad_norm": 0.3021568552153403,
+      "learning_rate": 6.917315350777335e-06,
+      "loss": 0.4532,
+      "step": 9931
+    },
+    {
+      "epoch": 1.6311046332601153,
+      "grad_norm": 0.3000004873374244,
+      "learning_rate": 6.916919775135374e-06,
+      "loss": 0.4772,
+      "step": 9932
+    },
+    {
+      "epoch": 1.6312688604684582,
+      "grad_norm": 0.32110553303472866,
+      "learning_rate": 6.916524173243285e-06,
+      "loss": 0.4524,
+      "step": 9933
+    },
+    {
+      "epoch": 1.6314330876768008,
+      "grad_norm": 0.31698803998851016,
+      "learning_rate": 6.916128545105368e-06,
+      "loss": 0.4602,
+      "step": 9934
+    },
+    {
+      "epoch": 1.6315973148851435,
+      "grad_norm": 0.28149338690697373,
+      "learning_rate": 6.915732890725915e-06,
+      "loss": 0.4535,
+      "step": 9935
+    },
+    {
+      "epoch": 1.6317615420934863,
+      "grad_norm": 0.3373716032595957,
+      "learning_rate": 6.91533721010923e-06,
+      "loss": 0.4603,
+      "step": 9936
+    },
+    {
+      "epoch": 1.6319257693018292,
+      "grad_norm": 0.2547340712685884,
+      "learning_rate": 6.914941503259604e-06,
+      "loss": 0.4513,
+      "step": 9937
+    },
+    {
+      "epoch": 1.6320899965101718,
+      "grad_norm": 0.2737540154463338,
+      "learning_rate": 6.9145457701813365e-06,
+      "loss": 0.4644,
+      "step": 9938
+    },
+    {
+      "epoch": 1.6322542237185145,
+      "grad_norm": 0.3191457901931209,
+      "learning_rate": 6.9141500108787284e-06,
+      "loss": 0.4609,
+      "step": 9939
+    },
+    {
+      "epoch": 1.6324184509268573,
+      "grad_norm": 0.30132448751628016,
+      "learning_rate": 6.913754225356075e-06,
+      "loss": 0.4529,
+      "step": 9940
+    },
+    {
+      "epoch": 1.6325826781352002,
+      "grad_norm": 0.3397790738337766,
+      "learning_rate": 6.9133584136176755e-06,
+      "loss": 0.4414,
+      "step": 9941
+    },
+    {
+      "epoch": 1.6327469053435428,
+      "grad_norm": 0.294003400100843,
+      "learning_rate": 6.91296257566783e-06,
+      "loss": 0.4663,
+      "step": 9942
+    },
+    {
+      "epoch": 1.6329111325518855,
+      "grad_norm": 0.3010856698778934,
+      "learning_rate": 6.912566711510836e-06,
+      "loss": 0.4773,
+      "step": 9943
+    },
+    {
+      "epoch": 1.6330753597602283,
+      "grad_norm": 0.4470639674793449,
+      "learning_rate": 6.912170821150994e-06,
+      "loss": 0.4485,
+      "step": 9944
+    },
+    {
+      "epoch": 1.633239586968571,
+      "grad_norm": 0.3158933653529828,
+      "learning_rate": 6.911774904592605e-06,
+      "loss": 0.4678,
+      "step": 9945
+    },
+    {
+      "epoch": 1.6334038141769138,
+      "grad_norm": 0.35320107780642485,
+      "learning_rate": 6.911378961839966e-06,
+      "loss": 0.4766,
+      "step": 9946
+    },
+    {
+      "epoch": 1.6335680413852565,
+      "grad_norm": 0.32146417670514194,
+      "learning_rate": 6.9109829928973794e-06,
+      "loss": 0.4507,
+      "step": 9947
+    },
+    {
+      "epoch": 1.633732268593599,
+      "grad_norm": 0.39318528095594146,
+      "learning_rate": 6.910586997769147e-06,
+      "loss": 0.4354,
+      "step": 9948
+    },
+    {
+      "epoch": 1.633896495801942,
+      "grad_norm": 0.31839866812565076,
+      "learning_rate": 6.910190976459568e-06,
+      "loss": 0.4575,
+      "step": 9949
+    },
+    {
+      "epoch": 1.6340607230102848,
+      "grad_norm": 0.4040545593174407,
+      "learning_rate": 6.9097949289729416e-06,
+      "loss": 0.4536,
+      "step": 9950
+    },
+    {
+      "epoch": 1.6342249502186275,
+      "grad_norm": 0.3204970703934227,
+      "learning_rate": 6.909398855313572e-06,
+      "loss": 0.4437,
+      "step": 9951
+    },
+    {
+      "epoch": 1.63438917742697,
+      "grad_norm": 0.35954942581069016,
+      "learning_rate": 6.90900275548576e-06,
+      "loss": 0.4657,
+      "step": 9952
+    },
+    {
+      "epoch": 1.634553404635313,
+      "grad_norm": 0.3009896626505442,
+      "learning_rate": 6.908606629493809e-06,
+      "loss": 0.4508,
+      "step": 9953
+    },
+    {
+      "epoch": 1.6347176318436558,
+      "grad_norm": 0.32257264785454315,
+      "learning_rate": 6.908210477342019e-06,
+      "loss": 0.4328,
+      "step": 9954
+    },
+    {
+      "epoch": 1.6348818590519985,
+      "grad_norm": 0.4503379019831764,
+      "learning_rate": 6.907814299034695e-06,
+      "loss": 0.4706,
+      "step": 9955
+    },
+    {
+      "epoch": 1.635046086260341,
+      "grad_norm": 0.3434774458899971,
+      "learning_rate": 6.907418094576138e-06,
+      "loss": 0.4926,
+      "step": 9956
+    },
+    {
+      "epoch": 1.635210313468684,
+      "grad_norm": 0.4039715002404461,
+      "learning_rate": 6.9070218639706535e-06,
+      "loss": 0.4568,
+      "step": 9957
+    },
+    {
+      "epoch": 1.6353745406770268,
+      "grad_norm": 0.32002629237152624,
+      "learning_rate": 6.906625607222541e-06,
+      "loss": 0.4694,
+      "step": 9958
+    },
+    {
+      "epoch": 1.6355387678853694,
+      "grad_norm": 0.28880005235928163,
+      "learning_rate": 6.906229324336109e-06,
+      "loss": 0.4752,
+      "step": 9959
+    },
+    {
+      "epoch": 1.635702995093712,
+      "grad_norm": 0.300776456819465,
+      "learning_rate": 6.905833015315657e-06,
+      "loss": 0.469,
+      "step": 9960
+    },
+    {
+      "epoch": 1.635867222302055,
+      "grad_norm": 0.3689131755069731,
+      "learning_rate": 6.905436680165493e-06,
+      "loss": 0.4483,
+      "step": 9961
+    },
+    {
+      "epoch": 1.6360314495103976,
+      "grad_norm": 0.34492706183414207,
+      "learning_rate": 6.905040318889919e-06,
+      "loss": 0.4574,
+      "step": 9962
+    },
+    {
+      "epoch": 1.6361956767187404,
+      "grad_norm": 0.3119084180868803,
+      "learning_rate": 6.904643931493241e-06,
+      "loss": 0.4612,
+      "step": 9963
+    },
+    {
+      "epoch": 1.636359903927083,
+      "grad_norm": 0.25489213577479997,
+      "learning_rate": 6.904247517979764e-06,
+      "loss": 0.4543,
+      "step": 9964
+    },
+    {
+      "epoch": 1.6365241311354257,
+      "grad_norm": 0.29549691545907814,
+      "learning_rate": 6.903851078353795e-06,
+      "loss": 0.4733,
+      "step": 9965
+    },
+    {
+      "epoch": 1.6366883583437686,
+      "grad_norm": 0.28335212002696497,
+      "learning_rate": 6.903454612619636e-06,
+      "loss": 0.4608,
+      "step": 9966
+    },
+    {
+      "epoch": 1.6368525855521114,
+      "grad_norm": 0.2993104367804374,
+      "learning_rate": 6.903058120781597e-06,
+      "loss": 0.4672,
+      "step": 9967
+    },
+    {
+      "epoch": 1.637016812760454,
+      "grad_norm": 0.28332822889276027,
+      "learning_rate": 6.902661602843981e-06,
+      "loss": 0.4487,
+      "step": 9968
+    },
+    {
+      "epoch": 1.6371810399687967,
+      "grad_norm": 0.25729560828720943,
+      "learning_rate": 6.9022650588110985e-06,
+      "loss": 0.4184,
+      "step": 9969
+    },
+    {
+      "epoch": 1.6373452671771396,
+      "grad_norm": 0.30264712506892016,
+      "learning_rate": 6.901868488687251e-06,
+      "loss": 0.4596,
+      "step": 9970
+    },
+    {
+      "epoch": 1.6375094943854824,
+      "grad_norm": 0.2672502724121429,
+      "learning_rate": 6.901471892476751e-06,
+      "loss": 0.4603,
+      "step": 9971
+    },
+    {
+      "epoch": 1.637673721593825,
+      "grad_norm": 0.28382573152561147,
+      "learning_rate": 6.901075270183901e-06,
+      "loss": 0.4805,
+      "step": 9972
+    },
+    {
+      "epoch": 1.6378379488021677,
+      "grad_norm": 0.29678729782940755,
+      "learning_rate": 6.900678621813014e-06,
+      "loss": 0.4696,
+      "step": 9973
+    },
+    {
+      "epoch": 1.6380021760105106,
+      "grad_norm": 0.2438202408021163,
+      "learning_rate": 6.900281947368394e-06,
+      "loss": 0.4582,
+      "step": 9974
+    },
+    {
+      "epoch": 1.6381664032188534,
+      "grad_norm": 0.3363567609882665,
+      "learning_rate": 6.8998852468543495e-06,
+      "loss": 0.4349,
+      "step": 9975
+    },
+    {
+      "epoch": 1.638330630427196,
+      "grad_norm": 0.2897556757604139,
+      "learning_rate": 6.8994885202751905e-06,
+      "loss": 0.4581,
+      "step": 9976
+    },
+    {
+      "epoch": 1.6384948576355387,
+      "grad_norm": 0.3406024616786714,
+      "learning_rate": 6.899091767635226e-06,
+      "loss": 0.4669,
+      "step": 9977
+    },
+    {
+      "epoch": 1.6386590848438816,
+      "grad_norm": 0.3571028862398934,
+      "learning_rate": 6.8986949889387655e-06,
+      "loss": 0.4622,
+      "step": 9978
+    },
+    {
+      "epoch": 1.6388233120522242,
+      "grad_norm": 0.28486072032707144,
+      "learning_rate": 6.898298184190116e-06,
+      "loss": 0.4584,
+      "step": 9979
+    },
+    {
+      "epoch": 1.638987539260567,
+      "grad_norm": 0.6487043597141194,
+      "learning_rate": 6.897901353393588e-06,
+      "loss": 0.474,
+      "step": 9980
+    },
+    {
+      "epoch": 1.6391517664689097,
+      "grad_norm": 0.2955606209398538,
+      "learning_rate": 6.897504496553493e-06,
+      "loss": 0.4613,
+      "step": 9981
+    },
+    {
+      "epoch": 1.6393159936772523,
+      "grad_norm": 0.3960711855428685,
+      "learning_rate": 6.89710761367414e-06,
+      "loss": 0.4384,
+      "step": 9982
+    },
+    {
+      "epoch": 1.6394802208855952,
+      "grad_norm": 0.32856200591871165,
+      "learning_rate": 6.8967107047598405e-06,
+      "loss": 0.4834,
+      "step": 9983
+    },
+    {
+      "epoch": 1.639644448093938,
+      "grad_norm": 0.2726115586771324,
+      "learning_rate": 6.896313769814905e-06,
+      "loss": 0.4681,
+      "step": 9984
+    },
+    {
+      "epoch": 1.6398086753022807,
+      "grad_norm": 0.2689421461804761,
+      "learning_rate": 6.895916808843643e-06,
+      "loss": 0.4533,
+      "step": 9985
+    },
+    {
+      "epoch": 1.6399729025106233,
+      "grad_norm": 0.2627988789746831,
+      "learning_rate": 6.895519821850368e-06,
+      "loss": 0.4709,
+      "step": 9986
+    },
+    {
+      "epoch": 1.6401371297189662,
+      "grad_norm": 0.32357716022869204,
+      "learning_rate": 6.895122808839391e-06,
+      "loss": 0.4752,
+      "step": 9987
+    },
+    {
+      "epoch": 1.640301356927309,
+      "grad_norm": 0.35351722543048564,
+      "learning_rate": 6.894725769815023e-06,
+      "loss": 0.4596,
+      "step": 9988
+    },
+    {
+      "epoch": 1.6404655841356517,
+      "grad_norm": 0.30283147058524224,
+      "learning_rate": 6.894328704781578e-06,
+      "loss": 0.4602,
+      "step": 9989
+    },
+    {
+      "epoch": 1.6406298113439943,
+      "grad_norm": 0.32871065213422807,
+      "learning_rate": 6.893931613743367e-06,
+      "loss": 0.4398,
+      "step": 9990
+    },
+    {
+      "epoch": 1.6407940385523372,
+      "grad_norm": 0.2909440710280135,
+      "learning_rate": 6.893534496704704e-06,
+      "loss": 0.4562,
+      "step": 9991
+    },
+    {
+      "epoch": 1.64095826576068,
+      "grad_norm": 0.26730344436772585,
+      "learning_rate": 6.893137353669899e-06,
+      "loss": 0.4363,
+      "step": 9992
+    },
+    {
+      "epoch": 1.6411224929690227,
+      "grad_norm": 0.3766542819585991,
+      "learning_rate": 6.892740184643271e-06,
+      "loss": 0.4577,
+      "step": 9993
+    },
+    {
+      "epoch": 1.6412867201773653,
+      "grad_norm": 0.2845863213745585,
+      "learning_rate": 6.8923429896291295e-06,
+      "loss": 0.4547,
+      "step": 9994
+    },
+    {
+      "epoch": 1.6414509473857082,
+      "grad_norm": 0.28734066949876436,
+      "learning_rate": 6.891945768631789e-06,
+      "loss": 0.47,
+      "step": 9995
+    },
+    {
+      "epoch": 1.6416151745940508,
+      "grad_norm": 0.4035762230136988,
+      "learning_rate": 6.891548521655563e-06,
+      "loss": 0.4615,
+      "step": 9996
+    },
+    {
+      "epoch": 1.6417794018023937,
+      "grad_norm": 0.27005177748526993,
+      "learning_rate": 6.891151248704769e-06,
+      "loss": 0.4527,
+      "step": 9997
+    },
+    {
+      "epoch": 1.6419436290107363,
+      "grad_norm": 1.0453460937427654,
+      "learning_rate": 6.890753949783719e-06,
+      "loss": 0.4578,
+      "step": 9998
+    },
+    {
+      "epoch": 1.642107856219079,
+      "grad_norm": 0.38637735669787354,
+      "learning_rate": 6.8903566248967296e-06,
+      "loss": 0.4528,
+      "step": 9999
+    },
+    {
+      "epoch": 1.6422720834274218,
+      "grad_norm": 0.26584240115120206,
+      "learning_rate": 6.889959274048115e-06,
+      "loss": 0.4463,
+      "step": 10000
+    },
+    {
+      "epoch": 1.6424363106357647,
+      "grad_norm": 0.390070342324435,
+      "learning_rate": 6.889561897242191e-06,
+      "loss": 0.4599,
+      "step": 10001
+    },
+    {
+      "epoch": 1.6426005378441073,
+      "grad_norm": 0.24050547042083945,
+      "learning_rate": 6.889164494483274e-06,
+      "loss": 0.4414,
+      "step": 10002
+    },
+    {
+      "epoch": 1.64276476505245,
+      "grad_norm": 0.2653471004908911,
+      "learning_rate": 6.888767065775681e-06,
+      "loss": 0.4624,
+      "step": 10003
+    },
+    {
+      "epoch": 1.6429289922607928,
+      "grad_norm": 0.2749182908648262,
+      "learning_rate": 6.888369611123726e-06,
+      "loss": 0.4613,
+      "step": 10004
+    },
+    {
+      "epoch": 1.6430932194691357,
+      "grad_norm": 0.3815791887017438,
+      "learning_rate": 6.887972130531728e-06,
+      "loss": 0.445,
+      "step": 10005
+    },
+    {
+      "epoch": 1.6432574466774783,
+      "grad_norm": 0.3873361429992468,
+      "learning_rate": 6.887574624004002e-06,
+      "loss": 0.4693,
+      "step": 10006
+    },
+    {
+      "epoch": 1.643421673885821,
+      "grad_norm": 0.40267621321284247,
+      "learning_rate": 6.8871770915448666e-06,
+      "loss": 0.4621,
+      "step": 10007
+    },
+    {
+      "epoch": 1.6435859010941638,
+      "grad_norm": 0.2863259481455292,
+      "learning_rate": 6.886779533158642e-06,
+      "loss": 0.4698,
+      "step": 10008
+    },
+    {
+      "epoch": 1.6437501283025067,
+      "grad_norm": 0.3518634637082607,
+      "learning_rate": 6.88638194884964e-06,
+      "loss": 0.4403,
+      "step": 10009
+    },
+    {
+      "epoch": 1.6439143555108493,
+      "grad_norm": 0.2604027741918767,
+      "learning_rate": 6.885984338622183e-06,
+      "loss": 0.4512,
+      "step": 10010
+    },
+    {
+      "epoch": 1.644078582719192,
+      "grad_norm": 0.28233015945831985,
+      "learning_rate": 6.8855867024805904e-06,
+      "loss": 0.4521,
+      "step": 10011
+    },
+    {
+      "epoch": 1.6442428099275348,
+      "grad_norm": 0.3275936381929935,
+      "learning_rate": 6.885189040429179e-06,
+      "loss": 0.4734,
+      "step": 10012
+    },
+    {
+      "epoch": 1.6444070371358774,
+      "grad_norm": 0.2791456316545117,
+      "learning_rate": 6.884791352472266e-06,
+      "loss": 0.4579,
+      "step": 10013
+    },
+    {
+      "epoch": 1.6445712643442203,
+      "grad_norm": 0.32228759885779773,
+      "learning_rate": 6.884393638614173e-06,
+      "loss": 0.4712,
+      "step": 10014
+    },
+    {
+      "epoch": 1.644735491552563,
+      "grad_norm": 0.43417519786081893,
+      "learning_rate": 6.88399589885922e-06,
+      "loss": 0.4557,
+      "step": 10015
+    },
+    {
+      "epoch": 1.6448997187609056,
+      "grad_norm": 0.36224933578030566,
+      "learning_rate": 6.8835981332117256e-06,
+      "loss": 0.4663,
+      "step": 10016
+    },
+    {
+      "epoch": 1.6450639459692484,
+      "grad_norm": 0.4044672096369628,
+      "learning_rate": 6.883200341676011e-06,
+      "loss": 0.4646,
+      "step": 10017
+    },
+    {
+      "epoch": 1.6452281731775913,
+      "grad_norm": 0.5066739693604935,
+      "learning_rate": 6.882802524256395e-06,
+      "loss": 0.4484,
+      "step": 10018
+    },
+    {
+      "epoch": 1.645392400385934,
+      "grad_norm": 0.31496716290507054,
+      "learning_rate": 6.8824046809571985e-06,
+      "loss": 0.4647,
+      "step": 10019
+    },
+    {
+      "epoch": 1.6455566275942766,
+      "grad_norm": 0.28390866682913946,
+      "learning_rate": 6.882006811782745e-06,
+      "loss": 0.4529,
+      "step": 10020
+    },
+    {
+      "epoch": 1.6457208548026194,
+      "grad_norm": 0.3395815099915771,
+      "learning_rate": 6.881608916737352e-06,
+      "loss": 0.4805,
+      "step": 10021
+    },
+    {
+      "epoch": 1.6458850820109623,
+      "grad_norm": 0.4099494282068759,
+      "learning_rate": 6.881210995825344e-06,
+      "loss": 0.4478,
+      "step": 10022
+    },
+    {
+      "epoch": 1.646049309219305,
+      "grad_norm": 0.30184948710719806,
+      "learning_rate": 6.880813049051043e-06,
+      "loss": 0.45,
+      "step": 10023
+    },
+    {
+      "epoch": 1.6462135364276476,
+      "grad_norm": 0.38529826858875593,
+      "learning_rate": 6.880415076418768e-06,
+      "loss": 0.4433,
+      "step": 10024
+    },
+    {
+      "epoch": 1.6463777636359904,
+      "grad_norm": 0.3318275138640135,
+      "learning_rate": 6.880017077932844e-06,
+      "loss": 0.4388,
+      "step": 10025
+    },
+    {
+      "epoch": 1.6465419908443333,
+      "grad_norm": 0.5649606434374844,
+      "learning_rate": 6.879619053597593e-06,
+      "loss": 0.4452,
+      "step": 10026
+    },
+    {
+      "epoch": 1.646706218052676,
+      "grad_norm": 0.30558762969639924,
+      "learning_rate": 6.879221003417338e-06,
+      "loss": 0.4533,
+      "step": 10027
+    },
+    {
+      "epoch": 1.6468704452610186,
+      "grad_norm": 0.3017474883745442,
+      "learning_rate": 6.878822927396402e-06,
+      "loss": 0.4363,
+      "step": 10028
+    },
+    {
+      "epoch": 1.6470346724693614,
+      "grad_norm": 0.3065943760906928,
+      "learning_rate": 6.878424825539108e-06,
+      "loss": 0.4582,
+      "step": 10029
+    },
+    {
+      "epoch": 1.647198899677704,
+      "grad_norm": 0.33933242129544633,
+      "learning_rate": 6.8780266978497805e-06,
+      "loss": 0.4664,
+      "step": 10030
+    },
+    {
+      "epoch": 1.647363126886047,
+      "grad_norm": 0.6614223853975514,
+      "learning_rate": 6.877628544332744e-06,
+      "loss": 0.4433,
+      "step": 10031
+    },
+    {
+      "epoch": 1.6475273540943896,
+      "grad_norm": 0.3048403727802867,
+      "learning_rate": 6.877230364992322e-06,
+      "loss": 0.4547,
+      "step": 10032
+    },
+    {
+      "epoch": 1.6476915813027322,
+      "grad_norm": 0.29375069237212753,
+      "learning_rate": 6.87683215983284e-06,
+      "loss": 0.4735,
+      "step": 10033
+    },
+    {
+      "epoch": 1.647855808511075,
+      "grad_norm": 0.2877086753406038,
+      "learning_rate": 6.876433928858621e-06,
+      "loss": 0.4635,
+      "step": 10034
+    },
+    {
+      "epoch": 1.648020035719418,
+      "grad_norm": 0.3292038999675611,
+      "learning_rate": 6.8760356720739906e-06,
+      "loss": 0.4748,
+      "step": 10035
+    },
+    {
+      "epoch": 1.6481842629277605,
+      "grad_norm": 0.28454457736434846,
+      "learning_rate": 6.875637389483278e-06,
+      "loss": 0.4575,
+      "step": 10036
+    },
+    {
+      "epoch": 1.6483484901361032,
+      "grad_norm": 0.46952622537098326,
+      "learning_rate": 6.875239081090805e-06,
+      "loss": 0.4685,
+      "step": 10037
+    },
+    {
+      "epoch": 1.648512717344446,
+      "grad_norm": 0.48947938405660096,
+      "learning_rate": 6.8748407469008975e-06,
+      "loss": 0.4523,
+      "step": 10038
+    },
+    {
+      "epoch": 1.648676944552789,
+      "grad_norm": 0.296064157512238,
+      "learning_rate": 6.874442386917883e-06,
+      "loss": 0.4504,
+      "step": 10039
+    },
+    {
+      "epoch": 1.6488411717611315,
+      "grad_norm": 0.29222940294924704,
+      "learning_rate": 6.87404400114609e-06,
+      "loss": 0.463,
+      "step": 10040
+    },
+    {
+      "epoch": 1.6490053989694742,
+      "grad_norm": 0.3095877404451654,
+      "learning_rate": 6.873645589589842e-06,
+      "loss": 0.4656,
+      "step": 10041
+    },
+    {
+      "epoch": 1.649169626177817,
+      "grad_norm": 0.6592237090493654,
+      "learning_rate": 6.8732471522534675e-06,
+      "loss": 0.4323,
+      "step": 10042
+    },
+    {
+      "epoch": 1.64933385338616,
+      "grad_norm": 0.36761926537069284,
+      "learning_rate": 6.872848689141294e-06,
+      "loss": 0.4479,
+      "step": 10043
+    },
+    {
+      "epoch": 1.6494980805945025,
+      "grad_norm": 0.4189296399366135,
+      "learning_rate": 6.872450200257648e-06,
+      "loss": 0.4572,
+      "step": 10044
+    },
+    {
+      "epoch": 1.6496623078028452,
+      "grad_norm": 0.3915907380874519,
+      "learning_rate": 6.872051685606861e-06,
+      "loss": 0.4438,
+      "step": 10045
+    },
+    {
+      "epoch": 1.649826535011188,
+      "grad_norm": 0.7110489512948105,
+      "learning_rate": 6.871653145193258e-06,
+      "loss": 0.4649,
+      "step": 10046
+    },
+    {
+      "epoch": 1.6499907622195307,
+      "grad_norm": 0.34437855980356424,
+      "learning_rate": 6.871254579021168e-06,
+      "loss": 0.4823,
+      "step": 10047
+    },
+    {
+      "epoch": 1.6501549894278735,
+      "grad_norm": 0.30378428528681245,
+      "learning_rate": 6.8708559870949205e-06,
+      "loss": 0.4661,
+      "step": 10048
+    },
+    {
+      "epoch": 1.6503192166362162,
+      "grad_norm": 0.3439466847782822,
+      "learning_rate": 6.8704573694188455e-06,
+      "loss": 0.4542,
+      "step": 10049
+    },
+    {
+      "epoch": 1.6504834438445588,
+      "grad_norm": 0.27429610500978663,
+      "learning_rate": 6.870058725997269e-06,
+      "loss": 0.4437,
+      "step": 10050
+    },
+    {
+      "epoch": 1.6506476710529017,
+      "grad_norm": 0.3041048998211635,
+      "learning_rate": 6.8696600568345235e-06,
+      "loss": 0.4444,
+      "step": 10051
+    },
+    {
+      "epoch": 1.6508118982612445,
+      "grad_norm": 0.3819352975581534,
+      "learning_rate": 6.869261361934939e-06,
+      "loss": 0.4556,
+      "step": 10052
+    },
+    {
+      "epoch": 1.6509761254695872,
+      "grad_norm": 0.3133404616783929,
+      "learning_rate": 6.8688626413028455e-06,
+      "loss": 0.4369,
+      "step": 10053
+    },
+    {
+      "epoch": 1.6511403526779298,
+      "grad_norm": 0.30859949689052946,
+      "learning_rate": 6.868463894942572e-06,
+      "loss": 0.4505,
+      "step": 10054
+    },
+    {
+      "epoch": 1.6513045798862727,
+      "grad_norm": 0.2569151609312324,
+      "learning_rate": 6.868065122858452e-06,
+      "loss": 0.4494,
+      "step": 10055
+    },
+    {
+      "epoch": 1.6514688070946155,
+      "grad_norm": 0.29726416618644635,
+      "learning_rate": 6.867666325054813e-06,
+      "loss": 0.4643,
+      "step": 10056
+    },
+    {
+      "epoch": 1.6516330343029582,
+      "grad_norm": 0.30981990269827364,
+      "learning_rate": 6.86726750153599e-06,
+      "loss": 0.4526,
+      "step": 10057
+    },
+    {
+      "epoch": 1.6517972615113008,
+      "grad_norm": 0.3135006032323148,
+      "learning_rate": 6.866868652306312e-06,
+      "loss": 0.4547,
+      "step": 10058
+    },
+    {
+      "epoch": 1.6519614887196437,
+      "grad_norm": 0.31218503934827935,
+      "learning_rate": 6.866469777370111e-06,
+      "loss": 0.4615,
+      "step": 10059
+    },
+    {
+      "epoch": 1.6521257159279865,
+      "grad_norm": 0.28950250972669206,
+      "learning_rate": 6.8660708767317204e-06,
+      "loss": 0.4606,
+      "step": 10060
+    },
+    {
+      "epoch": 1.6522899431363292,
+      "grad_norm": 0.32905922986746405,
+      "learning_rate": 6.865671950395474e-06,
+      "loss": 0.4595,
+      "step": 10061
+    },
+    {
+      "epoch": 1.6524541703446718,
+      "grad_norm": 0.28969266444342057,
+      "learning_rate": 6.8652729983656995e-06,
+      "loss": 0.4603,
+      "step": 10062
+    },
+    {
+      "epoch": 1.6526183975530147,
+      "grad_norm": 0.39683467383680576,
+      "learning_rate": 6.8648740206467345e-06,
+      "loss": 0.4676,
+      "step": 10063
+    },
+    {
+      "epoch": 1.6527826247613573,
+      "grad_norm": 0.5331896140060073,
+      "learning_rate": 6.864475017242911e-06,
+      "loss": 0.4477,
+      "step": 10064
+    },
+    {
+      "epoch": 1.6529468519697001,
+      "grad_norm": 0.2953487815232206,
+      "learning_rate": 6.8640759881585635e-06,
+      "loss": 0.4531,
+      "step": 10065
+    },
+    {
+      "epoch": 1.6531110791780428,
+      "grad_norm": 0.3145480336655805,
+      "learning_rate": 6.863676933398024e-06,
+      "loss": 0.4483,
+      "step": 10066
+    },
+    {
+      "epoch": 1.6532753063863854,
+      "grad_norm": 0.27112246157431824,
+      "learning_rate": 6.863277852965627e-06,
+      "loss": 0.4821,
+      "step": 10067
+    },
+    {
+      "epoch": 1.6534395335947283,
+      "grad_norm": 0.33070118687121697,
+      "learning_rate": 6.862878746865708e-06,
+      "loss": 0.47,
+      "step": 10068
+    },
+    {
+      "epoch": 1.6536037608030711,
+      "grad_norm": 0.3125265983720962,
+      "learning_rate": 6.8624796151026e-06,
+      "loss": 0.4612,
+      "step": 10069
+    },
+    {
+      "epoch": 1.6537679880114138,
+      "grad_norm": 0.3305711511682732,
+      "learning_rate": 6.862080457680641e-06,
+      "loss": 0.4693,
+      "step": 10070
+    },
+    {
+      "epoch": 1.6539322152197564,
+      "grad_norm": 0.29359255539804346,
+      "learning_rate": 6.861681274604163e-06,
+      "loss": 0.4603,
+      "step": 10071
+    },
+    {
+      "epoch": 1.6540964424280993,
+      "grad_norm": 0.3449893242439603,
+      "learning_rate": 6.861282065877503e-06,
+      "loss": 0.4705,
+      "step": 10072
+    },
+    {
+      "epoch": 1.6542606696364421,
+      "grad_norm": 0.44332745359425263,
+      "learning_rate": 6.860882831504996e-06,
+      "loss": 0.4621,
+      "step": 10073
+    },
+    {
+      "epoch": 1.6544248968447848,
+      "grad_norm": 0.3169582268764672,
+      "learning_rate": 6.86048357149098e-06,
+      "loss": 0.4606,
+      "step": 10074
+    },
+    {
+      "epoch": 1.6545891240531274,
+      "grad_norm": 0.2812684006764611,
+      "learning_rate": 6.860084285839787e-06,
+      "loss": 0.4624,
+      "step": 10075
+    },
+    {
+      "epoch": 1.6547533512614703,
+      "grad_norm": 0.3030134510583051,
+      "learning_rate": 6.85968497455576e-06,
+      "loss": 0.4628,
+      "step": 10076
+    },
+    {
+      "epoch": 1.6549175784698131,
+      "grad_norm": 0.5621866248997615,
+      "learning_rate": 6.859285637643231e-06,
+      "loss": 0.4475,
+      "step": 10077
+    },
+    {
+      "epoch": 1.6550818056781558,
+      "grad_norm": 0.2846808009420695,
+      "learning_rate": 6.85888627510654e-06,
+      "loss": 0.4467,
+      "step": 10078
+    },
+    {
+      "epoch": 1.6552460328864984,
+      "grad_norm": 0.5413683398262233,
+      "learning_rate": 6.858486886950022e-06,
+      "loss": 0.4525,
+      "step": 10079
+    },
+    {
+      "epoch": 1.6554102600948413,
+      "grad_norm": 0.26694224344114137,
+      "learning_rate": 6.858087473178015e-06,
+      "loss": 0.4613,
+      "step": 10080
+    },
+    {
+      "epoch": 1.655574487303184,
+      "grad_norm": 0.31430274465533675,
+      "learning_rate": 6.85768803379486e-06,
+      "loss": 0.4487,
+      "step": 10081
+    },
+    {
+      "epoch": 1.6557387145115268,
+      "grad_norm": 0.3048388379627296,
+      "learning_rate": 6.8572885688048935e-06,
+      "loss": 0.4686,
+      "step": 10082
+    },
+    {
+      "epoch": 1.6559029417198694,
+      "grad_norm": 0.4029978714207957,
+      "learning_rate": 6.856889078212452e-06,
+      "loss": 0.4556,
+      "step": 10083
+    },
+    {
+      "epoch": 1.656067168928212,
+      "grad_norm": 0.2713347612783681,
+      "learning_rate": 6.856489562021877e-06,
+      "loss": 0.4522,
+      "step": 10084
+    },
+    {
+      "epoch": 1.656231396136555,
+      "grad_norm": 0.3227466412547536,
+      "learning_rate": 6.856090020237507e-06,
+      "loss": 0.4526,
+      "step": 10085
+    },
+    {
+      "epoch": 1.6563956233448978,
+      "grad_norm": 0.30791311562056967,
+      "learning_rate": 6.855690452863681e-06,
+      "loss": 0.4426,
+      "step": 10086
+    },
+    {
+      "epoch": 1.6565598505532404,
+      "grad_norm": 0.28536385334132947,
+      "learning_rate": 6.85529085990474e-06,
+      "loss": 0.4503,
+      "step": 10087
+    },
+    {
+      "epoch": 1.656724077761583,
+      "grad_norm": 0.3003422483466558,
+      "learning_rate": 6.854891241365023e-06,
+      "loss": 0.484,
+      "step": 10088
+    },
+    {
+      "epoch": 1.656888304969926,
+      "grad_norm": 0.3026878855756908,
+      "learning_rate": 6.8544915972488685e-06,
+      "loss": 0.4474,
+      "step": 10089
+    },
+    {
+      "epoch": 1.6570525321782688,
+      "grad_norm": 0.2629547019528613,
+      "learning_rate": 6.85409192756062e-06,
+      "loss": 0.4556,
+      "step": 10090
+    },
+    {
+      "epoch": 1.6572167593866114,
+      "grad_norm": 0.5892721961165369,
+      "learning_rate": 6.8536922323046175e-06,
+      "loss": 0.4635,
+      "step": 10091
+    },
+    {
+      "epoch": 1.657380986594954,
+      "grad_norm": 0.31233894549114827,
+      "learning_rate": 6.8532925114852005e-06,
+      "loss": 0.4504,
+      "step": 10092
+    },
+    {
+      "epoch": 1.657545213803297,
+      "grad_norm": 0.9927882559085216,
+      "learning_rate": 6.852892765106712e-06,
+      "loss": 0.4446,
+      "step": 10093
+    },
+    {
+      "epoch": 1.6577094410116398,
+      "grad_norm": 0.27389626760484753,
+      "learning_rate": 6.852492993173493e-06,
+      "loss": 0.4438,
+      "step": 10094
+    },
+    {
+      "epoch": 1.6578736682199824,
+      "grad_norm": 0.28999814533998897,
+      "learning_rate": 6.852093195689886e-06,
+      "loss": 0.4749,
+      "step": 10095
+    },
+    {
+      "epoch": 1.658037895428325,
+      "grad_norm": 0.28344610002066106,
+      "learning_rate": 6.851693372660232e-06,
+      "loss": 0.4621,
+      "step": 10096
+    },
+    {
+      "epoch": 1.6582021226366677,
+      "grad_norm": 0.29129088168434936,
+      "learning_rate": 6.851293524088875e-06,
+      "loss": 0.4675,
+      "step": 10097
+    },
+    {
+      "epoch": 1.6583663498450105,
+      "grad_norm": 0.26928559045198974,
+      "learning_rate": 6.850893649980156e-06,
+      "loss": 0.4559,
+      "step": 10098
+    },
+    {
+      "epoch": 1.6585305770533534,
+      "grad_norm": 0.3608111549504681,
+      "learning_rate": 6.85049375033842e-06,
+      "loss": 0.4795,
+      "step": 10099
+    },
+    {
+      "epoch": 1.658694804261696,
+      "grad_norm": 0.31003894818762734,
+      "learning_rate": 6.850093825168009e-06,
+      "loss": 0.4503,
+      "step": 10100
+    },
+    {
+      "epoch": 1.6588590314700387,
+      "grad_norm": 0.2982802514934513,
+      "learning_rate": 6.849693874473266e-06,
+      "loss": 0.4603,
+      "step": 10101
+    },
+    {
+      "epoch": 1.6590232586783815,
+      "grad_norm": 0.36819864548273795,
+      "learning_rate": 6.849293898258537e-06,
+      "loss": 0.4573,
+      "step": 10102
+    },
+    {
+      "epoch": 1.6591874858867244,
+      "grad_norm": 0.3051585703764991,
+      "learning_rate": 6.8488938965281645e-06,
+      "loss": 0.4663,
+      "step": 10103
+    },
+    {
+      "epoch": 1.659351713095067,
+      "grad_norm": 0.33465803831927327,
+      "learning_rate": 6.848493869286493e-06,
+      "loss": 0.4502,
+      "step": 10104
+    },
+    {
+      "epoch": 1.6595159403034097,
+      "grad_norm": 0.38292256300891603,
+      "learning_rate": 6.848093816537868e-06,
+      "loss": 0.455,
+      "step": 10105
+    },
+    {
+      "epoch": 1.6596801675117525,
+      "grad_norm": 0.4288122393071528,
+      "learning_rate": 6.847693738286633e-06,
+      "loss": 0.4597,
+      "step": 10106
+    },
+    {
+      "epoch": 1.6598443947200954,
+      "grad_norm": 0.24811801866347535,
+      "learning_rate": 6.847293634537135e-06,
+      "loss": 0.4505,
+      "step": 10107
+    },
+    {
+      "epoch": 1.660008621928438,
+      "grad_norm": 0.3088836539894647,
+      "learning_rate": 6.846893505293719e-06,
+      "loss": 0.4509,
+      "step": 10108
+    },
+    {
+      "epoch": 1.6601728491367806,
+      "grad_norm": 0.3003313846361625,
+      "learning_rate": 6.846493350560729e-06,
+      "loss": 0.4523,
+      "step": 10109
+    },
+    {
+      "epoch": 1.6603370763451235,
+      "grad_norm": 0.5284141833491072,
+      "learning_rate": 6.846093170342515e-06,
+      "loss": 0.4548,
+      "step": 10110
+    },
+    {
+      "epoch": 1.6605013035534664,
+      "grad_norm": 0.3918097946018369,
+      "learning_rate": 6.84569296464342e-06,
+      "loss": 0.4502,
+      "step": 10111
+    },
+    {
+      "epoch": 1.660665530761809,
+      "grad_norm": 0.3162058034183177,
+      "learning_rate": 6.84529273346779e-06,
+      "loss": 0.4537,
+      "step": 10112
+    },
+    {
+      "epoch": 1.6608297579701516,
+      "grad_norm": 0.3712791403292991,
+      "learning_rate": 6.844892476819973e-06,
+      "loss": 0.4753,
+      "step": 10113
+    },
+    {
+      "epoch": 1.6609939851784943,
+      "grad_norm": 0.2919157998372654,
+      "learning_rate": 6.84449219470432e-06,
+      "loss": 0.4555,
+      "step": 10114
+    },
+    {
+      "epoch": 1.6611582123868371,
+      "grad_norm": 0.3444696305060915,
+      "learning_rate": 6.844091887125172e-06,
+      "loss": 0.4596,
+      "step": 10115
+    },
+    {
+      "epoch": 1.66132243959518,
+      "grad_norm": 0.3506674472551571,
+      "learning_rate": 6.843691554086882e-06,
+      "loss": 0.4696,
+      "step": 10116
+    },
+    {
+      "epoch": 1.6614866668035226,
+      "grad_norm": 0.3444951783000774,
+      "learning_rate": 6.8432911955937935e-06,
+      "loss": 0.4573,
+      "step": 10117
+    },
+    {
+      "epoch": 1.6616508940118653,
+      "grad_norm": 0.4158734419460173,
+      "learning_rate": 6.8428908116502595e-06,
+      "loss": 0.4593,
+      "step": 10118
+    },
+    {
+      "epoch": 1.6618151212202081,
+      "grad_norm": 0.30382622051819197,
+      "learning_rate": 6.842490402260625e-06,
+      "loss": 0.4731,
+      "step": 10119
+    },
+    {
+      "epoch": 1.661979348428551,
+      "grad_norm": 0.2730468701565896,
+      "learning_rate": 6.8420899674292405e-06,
+      "loss": 0.4655,
+      "step": 10120
+    },
+    {
+      "epoch": 1.6621435756368936,
+      "grad_norm": 0.3113059653283631,
+      "learning_rate": 6.8416895071604545e-06,
+      "loss": 0.4715,
+      "step": 10121
+    },
+    {
+      "epoch": 1.6623078028452363,
+      "grad_norm": 0.4526111377968724,
+      "learning_rate": 6.841289021458617e-06,
+      "loss": 0.4748,
+      "step": 10122
+    },
+    {
+      "epoch": 1.6624720300535791,
+      "grad_norm": 0.3076876962554403,
+      "learning_rate": 6.840888510328075e-06,
+      "loss": 0.4424,
+      "step": 10123
+    },
+    {
+      "epoch": 1.662636257261922,
+      "grad_norm": 0.4323855586209853,
+      "learning_rate": 6.840487973773183e-06,
+      "loss": 0.4513,
+      "step": 10124
+    },
+    {
+      "epoch": 1.6628004844702646,
+      "grad_norm": 0.2948138042150791,
+      "learning_rate": 6.840087411798289e-06,
+      "loss": 0.4572,
+      "step": 10125
+    },
+    {
+      "epoch": 1.6629647116786073,
+      "grad_norm": 0.3235230615740369,
+      "learning_rate": 6.839686824407742e-06,
+      "loss": 0.4566,
+      "step": 10126
+    },
+    {
+      "epoch": 1.6631289388869501,
+      "grad_norm": 0.35818271322707407,
+      "learning_rate": 6.8392862116058945e-06,
+      "loss": 0.4557,
+      "step": 10127
+    },
+    {
+      "epoch": 1.663293166095293,
+      "grad_norm": 0.28832099909946796,
+      "learning_rate": 6.8388855733970975e-06,
+      "loss": 0.464,
+      "step": 10128
+    },
+    {
+      "epoch": 1.6634573933036356,
+      "grad_norm": 0.32200541250106973,
+      "learning_rate": 6.838484909785702e-06,
+      "loss": 0.4468,
+      "step": 10129
+    },
+    {
+      "epoch": 1.6636216205119783,
+      "grad_norm": 0.2759151795686372,
+      "learning_rate": 6.838084220776061e-06,
+      "loss": 0.4549,
+      "step": 10130
+    },
+    {
+      "epoch": 1.663785847720321,
+      "grad_norm": 0.38764451430397096,
+      "learning_rate": 6.8376835063725216e-06,
+      "loss": 0.4591,
+      "step": 10131
+    },
+    {
+      "epoch": 1.6639500749286638,
+      "grad_norm": 0.3689015443052184,
+      "learning_rate": 6.837282766579442e-06,
+      "loss": 0.4512,
+      "step": 10132
+    },
+    {
+      "epoch": 1.6641143021370066,
+      "grad_norm": 0.3136526314101458,
+      "learning_rate": 6.836882001401171e-06,
+      "loss": 0.4673,
+      "step": 10133
+    },
+    {
+      "epoch": 1.6642785293453493,
+      "grad_norm": 0.29260575496034674,
+      "learning_rate": 6.836481210842064e-06,
+      "loss": 0.4675,
+      "step": 10134
+    },
+    {
+      "epoch": 1.664442756553692,
+      "grad_norm": 0.3097745397723282,
+      "learning_rate": 6.83608039490647e-06,
+      "loss": 0.4534,
+      "step": 10135
+    },
+    {
+      "epoch": 1.6646069837620348,
+      "grad_norm": 0.2974482169941179,
+      "learning_rate": 6.835679553598746e-06,
+      "loss": 0.4403,
+      "step": 10136
+    },
+    {
+      "epoch": 1.6647712109703776,
+      "grad_norm": 0.5267663399483836,
+      "learning_rate": 6.835278686923242e-06,
+      "loss": 0.4649,
+      "step": 10137
+    },
+    {
+      "epoch": 1.6649354381787203,
+      "grad_norm": 0.28156728840426715,
+      "learning_rate": 6.834877794884314e-06,
+      "loss": 0.4638,
+      "step": 10138
+    },
+    {
+      "epoch": 1.6650996653870629,
+      "grad_norm": 0.44869677574882944,
+      "learning_rate": 6.834476877486318e-06,
+      "loss": 0.453,
+      "step": 10139
+    },
+    {
+      "epoch": 1.6652638925954057,
+      "grad_norm": 0.30123810086379144,
+      "learning_rate": 6.8340759347336056e-06,
+      "loss": 0.4601,
+      "step": 10140
+    },
+    {
+      "epoch": 1.6654281198037486,
+      "grad_norm": 0.5795294618527556,
+      "learning_rate": 6.833674966630533e-06,
+      "loss": 0.4615,
+      "step": 10141
+    },
+    {
+      "epoch": 1.6655923470120912,
+      "grad_norm": 0.40046742999378,
+      "learning_rate": 6.833273973181453e-06,
+      "loss": 0.47,
+      "step": 10142
+    },
+    {
+      "epoch": 1.6657565742204339,
+      "grad_norm": 0.2708138308319611,
+      "learning_rate": 6.832872954390723e-06,
+      "loss": 0.4649,
+      "step": 10143
+    },
+    {
+      "epoch": 1.6659208014287767,
+      "grad_norm": 0.4570428523808035,
+      "learning_rate": 6.832471910262696e-06,
+      "loss": 0.4627,
+      "step": 10144
+    },
+    {
+      "epoch": 1.6660850286371196,
+      "grad_norm": 0.4293640775402717,
+      "learning_rate": 6.8320708408017305e-06,
+      "loss": 0.4579,
+      "step": 10145
+    },
+    {
+      "epoch": 1.6662492558454622,
+      "grad_norm": 0.2685026647291335,
+      "learning_rate": 6.83166974601218e-06,
+      "loss": 0.4509,
+      "step": 10146
+    },
+    {
+      "epoch": 1.6664134830538049,
+      "grad_norm": 0.31228397957090265,
+      "learning_rate": 6.831268625898402e-06,
+      "loss": 0.479,
+      "step": 10147
+    },
+    {
+      "epoch": 1.6665777102621475,
+      "grad_norm": 0.3137721705986118,
+      "learning_rate": 6.830867480464754e-06,
+      "loss": 0.4437,
+      "step": 10148
+    },
+    {
+      "epoch": 1.6667419374704904,
+      "grad_norm": 0.5708636843935152,
+      "learning_rate": 6.830466309715593e-06,
+      "loss": 0.4512,
+      "step": 10149
+    },
+    {
+      "epoch": 1.6669061646788332,
+      "grad_norm": 0.2653212103007819,
+      "learning_rate": 6.830065113655272e-06,
+      "loss": 0.4496,
+      "step": 10150
+    },
+    {
+      "epoch": 1.6670703918871759,
+      "grad_norm": 0.7572840674435325,
+      "learning_rate": 6.829663892288155e-06,
+      "loss": 0.4401,
+      "step": 10151
+    },
+    {
+      "epoch": 1.6672346190955185,
+      "grad_norm": 0.425478297583331,
+      "learning_rate": 6.829262645618592e-06,
+      "loss": 0.4681,
+      "step": 10152
+    },
+    {
+      "epoch": 1.6673988463038614,
+      "grad_norm": 0.2559404822460795,
+      "learning_rate": 6.8288613736509485e-06,
+      "loss": 0.4491,
+      "step": 10153
+    },
+    {
+      "epoch": 1.6675630735122042,
+      "grad_norm": 0.2704039864595664,
+      "learning_rate": 6.828460076389577e-06,
+      "loss": 0.4651,
+      "step": 10154
+    },
+    {
+      "epoch": 1.6677273007205469,
+      "grad_norm": 0.29909003380134896,
+      "learning_rate": 6.82805875383884e-06,
+      "loss": 0.4391,
+      "step": 10155
+    },
+    {
+      "epoch": 1.6678915279288895,
+      "grad_norm": 0.3071232266044963,
+      "learning_rate": 6.827657406003092e-06,
+      "loss": 0.4543,
+      "step": 10156
+    },
+    {
+      "epoch": 1.6680557551372324,
+      "grad_norm": 0.33138094181705335,
+      "learning_rate": 6.8272560328866965e-06,
+      "loss": 0.4758,
+      "step": 10157
+    },
+    {
+      "epoch": 1.6682199823455752,
+      "grad_norm": 0.7843415418694715,
+      "learning_rate": 6.826854634494011e-06,
+      "loss": 0.4437,
+      "step": 10158
+    },
+    {
+      "epoch": 1.6683842095539179,
+      "grad_norm": 0.2925059010158654,
+      "learning_rate": 6.8264532108293936e-06,
+      "loss": 0.4652,
+      "step": 10159
+    },
+    {
+      "epoch": 1.6685484367622605,
+      "grad_norm": 0.31203921925175077,
+      "learning_rate": 6.826051761897205e-06,
+      "loss": 0.4407,
+      "step": 10160
+    },
+    {
+      "epoch": 1.6687126639706034,
+      "grad_norm": 0.3378418281840202,
+      "learning_rate": 6.825650287701807e-06,
+      "loss": 0.4396,
+      "step": 10161
+    },
+    {
+      "epoch": 1.6688768911789462,
+      "grad_norm": 0.6728722464644683,
+      "learning_rate": 6.825248788247557e-06,
+      "loss": 0.4732,
+      "step": 10162
+    },
+    {
+      "epoch": 1.6690411183872889,
+      "grad_norm": 0.4629013914353093,
+      "learning_rate": 6.82484726353882e-06,
+      "loss": 0.4691,
+      "step": 10163
+    },
+    {
+      "epoch": 1.6692053455956315,
+      "grad_norm": 0.31808519111750916,
+      "learning_rate": 6.824445713579954e-06,
+      "loss": 0.4594,
+      "step": 10164
+    },
+    {
+      "epoch": 1.6693695728039741,
+      "grad_norm": 0.2995549769221818,
+      "learning_rate": 6.824044138375318e-06,
+      "loss": 0.4584,
+      "step": 10165
+    },
+    {
+      "epoch": 1.669533800012317,
+      "grad_norm": 1.8078725563334275,
+      "learning_rate": 6.823642537929278e-06,
+      "loss": 0.4445,
+      "step": 10166
+    },
+    {
+      "epoch": 1.6696980272206599,
+      "grad_norm": 0.2907891043954342,
+      "learning_rate": 6.823240912246193e-06,
+      "loss": 0.4537,
+      "step": 10167
+    },
+    {
+      "epoch": 1.6698622544290025,
+      "grad_norm": 0.5084326002415936,
+      "learning_rate": 6.8228392613304285e-06,
+      "loss": 0.4627,
+      "step": 10168
+    },
+    {
+      "epoch": 1.6700264816373451,
+      "grad_norm": 0.29511217662791017,
+      "learning_rate": 6.822437585186341e-06,
+      "loss": 0.4779,
+      "step": 10169
+    },
+    {
+      "epoch": 1.670190708845688,
+      "grad_norm": 0.2693223126141609,
+      "learning_rate": 6.822035883818299e-06,
+      "loss": 0.4466,
+      "step": 10170
+    },
+    {
+      "epoch": 1.6703549360540308,
+      "grad_norm": 0.3933448320738575,
+      "learning_rate": 6.821634157230661e-06,
+      "loss": 0.4772,
+      "step": 10171
+    },
+    {
+      "epoch": 1.6705191632623735,
+      "grad_norm": 0.2942491456962361,
+      "learning_rate": 6.821232405427791e-06,
+      "loss": 0.452,
+      "step": 10172
+    },
+    {
+      "epoch": 1.6706833904707161,
+      "grad_norm": 0.29925065753639407,
+      "learning_rate": 6.820830628414056e-06,
+      "loss": 0.4626,
+      "step": 10173
+    },
+    {
+      "epoch": 1.670847617679059,
+      "grad_norm": 0.3433489098652025,
+      "learning_rate": 6.820428826193816e-06,
+      "loss": 0.4565,
+      "step": 10174
+    },
+    {
+      "epoch": 1.6710118448874018,
+      "grad_norm": 0.297289729417751,
+      "learning_rate": 6.820026998771435e-06,
+      "loss": 0.468,
+      "step": 10175
+    },
+    {
+      "epoch": 1.6711760720957445,
+      "grad_norm": 0.2536699644456774,
+      "learning_rate": 6.819625146151278e-06,
+      "loss": 0.4534,
+      "step": 10176
+    },
+    {
+      "epoch": 1.6713402993040871,
+      "grad_norm": 0.25733154085588766,
+      "learning_rate": 6.81922326833771e-06,
+      "loss": 0.4698,
+      "step": 10177
+    },
+    {
+      "epoch": 1.67150452651243,
+      "grad_norm": 0.7738151356879273,
+      "learning_rate": 6.818821365335097e-06,
+      "loss": 0.4686,
+      "step": 10178
+    },
+    {
+      "epoch": 1.6716687537207728,
+      "grad_norm": 0.2918725037566939,
+      "learning_rate": 6.818419437147802e-06,
+      "loss": 0.4822,
+      "step": 10179
+    },
+    {
+      "epoch": 1.6718329809291155,
+      "grad_norm": 0.3340710811833185,
+      "learning_rate": 6.81801748378019e-06,
+      "loss": 0.4595,
+      "step": 10180
+    },
+    {
+      "epoch": 1.6719972081374581,
+      "grad_norm": 0.30776634460261865,
+      "learning_rate": 6.817615505236627e-06,
+      "loss": 0.4676,
+      "step": 10181
+    },
+    {
+      "epoch": 1.6721614353458008,
+      "grad_norm": 0.2942163724177741,
+      "learning_rate": 6.81721350152148e-06,
+      "loss": 0.4607,
+      "step": 10182
+    },
+    {
+      "epoch": 1.6723256625541436,
+      "grad_norm": 0.6801375014109976,
+      "learning_rate": 6.816811472639116e-06,
+      "loss": 0.4689,
+      "step": 10183
+    },
+    {
+      "epoch": 1.6724898897624865,
+      "grad_norm": 0.31395840415757875,
+      "learning_rate": 6.8164094185938986e-06,
+      "loss": 0.4637,
+      "step": 10184
+    },
+    {
+      "epoch": 1.672654116970829,
+      "grad_norm": 0.3058483374593646,
+      "learning_rate": 6.816007339390195e-06,
+      "loss": 0.4504,
+      "step": 10185
+    },
+    {
+      "epoch": 1.6728183441791717,
+      "grad_norm": 0.31385256374761467,
+      "learning_rate": 6.815605235032374e-06,
+      "loss": 0.4518,
+      "step": 10186
+    },
+    {
+      "epoch": 1.6729825713875146,
+      "grad_norm": 0.29066102734867516,
+      "learning_rate": 6.815203105524803e-06,
+      "loss": 0.4303,
+      "step": 10187
+    },
+    {
+      "epoch": 1.6731467985958575,
+      "grad_norm": 0.3231898610109705,
+      "learning_rate": 6.814800950871848e-06,
+      "loss": 0.4578,
+      "step": 10188
+    },
+    {
+      "epoch": 1.6733110258042,
+      "grad_norm": 0.305140836240839,
+      "learning_rate": 6.8143987710778764e-06,
+      "loss": 0.4656,
+      "step": 10189
+    },
+    {
+      "epoch": 1.6734752530125427,
+      "grad_norm": 0.301483387719651,
+      "learning_rate": 6.813996566147257e-06,
+      "loss": 0.4879,
+      "step": 10190
+    },
+    {
+      "epoch": 1.6736394802208856,
+      "grad_norm": 1.2463700741489872,
+      "learning_rate": 6.813594336084359e-06,
+      "loss": 0.4543,
+      "step": 10191
+    },
+    {
+      "epoch": 1.6738037074292285,
+      "grad_norm": 0.39223622977769473,
+      "learning_rate": 6.81319208089355e-06,
+      "loss": 0.4681,
+      "step": 10192
+    },
+    {
+      "epoch": 1.673967934637571,
+      "grad_norm": 0.3176465043128236,
+      "learning_rate": 6.812789800579198e-06,
+      "loss": 0.4456,
+      "step": 10193
+    },
+    {
+      "epoch": 1.6741321618459137,
+      "grad_norm": 0.2903606994722687,
+      "learning_rate": 6.812387495145675e-06,
+      "loss": 0.4687,
+      "step": 10194
+    },
+    {
+      "epoch": 1.6742963890542566,
+      "grad_norm": 0.3094943548301401,
+      "learning_rate": 6.811985164597348e-06,
+      "loss": 0.4565,
+      "step": 10195
+    },
+    {
+      "epoch": 1.6744606162625995,
+      "grad_norm": 0.4317000637297288,
+      "learning_rate": 6.811582808938587e-06,
+      "loss": 0.449,
+      "step": 10196
+    },
+    {
+      "epoch": 1.674624843470942,
+      "grad_norm": 0.45232778899457615,
+      "learning_rate": 6.8111804281737636e-06,
+      "loss": 0.4724,
+      "step": 10197
+    },
+    {
+      "epoch": 1.6747890706792847,
+      "grad_norm": 0.33289301304418345,
+      "learning_rate": 6.810778022307245e-06,
+      "loss": 0.4624,
+      "step": 10198
+    },
+    {
+      "epoch": 1.6749532978876274,
+      "grad_norm": 0.8288740572151884,
+      "learning_rate": 6.810375591343405e-06,
+      "loss": 0.4414,
+      "step": 10199
+    },
+    {
+      "epoch": 1.6751175250959702,
+      "grad_norm": 0.5574232589287477,
+      "learning_rate": 6.809973135286613e-06,
+      "loss": 0.4157,
+      "step": 10200
+    },
+    {
+      "epoch": 1.675281752304313,
+      "grad_norm": 0.40580094718204485,
+      "learning_rate": 6.809570654141239e-06,
+      "loss": 0.4702,
+      "step": 10201
+    },
+    {
+      "epoch": 1.6754459795126557,
+      "grad_norm": 0.337536087662044,
+      "learning_rate": 6.809168147911656e-06,
+      "loss": 0.4573,
+      "step": 10202
+    },
+    {
+      "epoch": 1.6756102067209984,
+      "grad_norm": 0.3024412512663363,
+      "learning_rate": 6.808765616602236e-06,
+      "loss": 0.4379,
+      "step": 10203
+    },
+    {
+      "epoch": 1.6757744339293412,
+      "grad_norm": 0.36910762147354104,
+      "learning_rate": 6.808363060217348e-06,
+      "loss": 0.4482,
+      "step": 10204
+    },
+    {
+      "epoch": 1.675938661137684,
+      "grad_norm": 0.4531128443431689,
+      "learning_rate": 6.8079604787613664e-06,
+      "loss": 0.4615,
+      "step": 10205
+    },
+    {
+      "epoch": 1.6761028883460267,
+      "grad_norm": 0.25406746010569026,
+      "learning_rate": 6.8075578722386646e-06,
+      "loss": 0.4625,
+      "step": 10206
+    },
+    {
+      "epoch": 1.6762671155543694,
+      "grad_norm": 0.4489461763784889,
+      "learning_rate": 6.807155240653614e-06,
+      "loss": 0.4557,
+      "step": 10207
+    },
+    {
+      "epoch": 1.6764313427627122,
+      "grad_norm": 0.32368261589902353,
+      "learning_rate": 6.806752584010586e-06,
+      "loss": 0.4512,
+      "step": 10208
+    },
+    {
+      "epoch": 1.676595569971055,
+      "grad_norm": 0.30499813536039555,
+      "learning_rate": 6.8063499023139565e-06,
+      "loss": 0.4653,
+      "step": 10209
+    },
+    {
+      "epoch": 1.6767597971793977,
+      "grad_norm": 0.28669103974464233,
+      "learning_rate": 6.805947195568096e-06,
+      "loss": 0.4656,
+      "step": 10210
+    },
+    {
+      "epoch": 1.6769240243877404,
+      "grad_norm": 0.342212151906084,
+      "learning_rate": 6.805544463777383e-06,
+      "loss": 0.4629,
+      "step": 10211
+    },
+    {
+      "epoch": 1.6770882515960832,
+      "grad_norm": 0.2878312678429544,
+      "learning_rate": 6.805141706946188e-06,
+      "loss": 0.4674,
+      "step": 10212
+    },
+    {
+      "epoch": 1.677252478804426,
+      "grad_norm": 0.2932394604869272,
+      "learning_rate": 6.804738925078885e-06,
+      "loss": 0.4713,
+      "step": 10213
+    },
+    {
+      "epoch": 1.6774167060127687,
+      "grad_norm": 0.29840250165017845,
+      "learning_rate": 6.8043361181798515e-06,
+      "loss": 0.4747,
+      "step": 10214
+    },
+    {
+      "epoch": 1.6775809332211113,
+      "grad_norm": 0.3088400439236326,
+      "learning_rate": 6.803933286253458e-06,
+      "loss": 0.4733,
+      "step": 10215
+    },
+    {
+      "epoch": 1.677745160429454,
+      "grad_norm": 0.3900409979120372,
+      "learning_rate": 6.803530429304084e-06,
+      "loss": 0.4675,
+      "step": 10216
+    },
+    {
+      "epoch": 1.6779093876377968,
+      "grad_norm": 0.30596119386700804,
+      "learning_rate": 6.803127547336104e-06,
+      "loss": 0.4477,
+      "step": 10217
+    },
+    {
+      "epoch": 1.6780736148461397,
+      "grad_norm": 0.3396698020804204,
+      "learning_rate": 6.802724640353891e-06,
+      "loss": 0.4533,
+      "step": 10218
+    },
+    {
+      "epoch": 1.6782378420544823,
+      "grad_norm": 0.36186525495151045,
+      "learning_rate": 6.802321708361823e-06,
+      "loss": 0.4737,
+      "step": 10219
+    },
+    {
+      "epoch": 1.678402069262825,
+      "grad_norm": 0.32655803334282024,
+      "learning_rate": 6.8019187513642775e-06,
+      "loss": 0.4307,
+      "step": 10220
+    },
+    {
+      "epoch": 1.6785662964711678,
+      "grad_norm": 0.2625813564960309,
+      "learning_rate": 6.801515769365629e-06,
+      "loss": 0.4823,
+      "step": 10221
+    },
+    {
+      "epoch": 1.6787305236795107,
+      "grad_norm": 0.29774752894997863,
+      "learning_rate": 6.801112762370254e-06,
+      "loss": 0.4524,
+      "step": 10222
+    },
+    {
+      "epoch": 1.6788947508878533,
+      "grad_norm": 0.3127505478958254,
+      "learning_rate": 6.800709730382531e-06,
+      "loss": 0.4403,
+      "step": 10223
+    },
+    {
+      "epoch": 1.679058978096196,
+      "grad_norm": 0.30376527080229004,
+      "learning_rate": 6.8003066734068374e-06,
+      "loss": 0.4531,
+      "step": 10224
+    },
+    {
+      "epoch": 1.6792232053045388,
+      "grad_norm": 0.3136688315705886,
+      "learning_rate": 6.799903591447548e-06,
+      "loss": 0.4698,
+      "step": 10225
+    },
+    {
+      "epoch": 1.6793874325128817,
+      "grad_norm": 0.28186818901257743,
+      "learning_rate": 6.799500484509046e-06,
+      "loss": 0.4312,
+      "step": 10226
+    },
+    {
+      "epoch": 1.6795516597212243,
+      "grad_norm": 0.30242428453226355,
+      "learning_rate": 6.799097352595704e-06,
+      "loss": 0.4545,
+      "step": 10227
+    },
+    {
+      "epoch": 1.679715886929567,
+      "grad_norm": 0.33302698471723835,
+      "learning_rate": 6.798694195711903e-06,
+      "loss": 0.4716,
+      "step": 10228
+    },
+    {
+      "epoch": 1.6798801141379098,
+      "grad_norm": 0.34080784140890574,
+      "learning_rate": 6.798291013862023e-06,
+      "loss": 0.4677,
+      "step": 10229
+    },
+    {
+      "epoch": 1.6800443413462527,
+      "grad_norm": 0.3374671449414891,
+      "learning_rate": 6.797887807050439e-06,
+      "loss": 0.4534,
+      "step": 10230
+    },
+    {
+      "epoch": 1.6802085685545953,
+      "grad_norm": 0.3153387804540685,
+      "learning_rate": 6.797484575281535e-06,
+      "loss": 0.463,
+      "step": 10231
+    },
+    {
+      "epoch": 1.680372795762938,
+      "grad_norm": 0.402795624677278,
+      "learning_rate": 6.797081318559686e-06,
+      "loss": 0.4644,
+      "step": 10232
+    },
+    {
+      "epoch": 1.6805370229712806,
+      "grad_norm": 0.30953687154793585,
+      "learning_rate": 6.796678036889275e-06,
+      "loss": 0.4674,
+      "step": 10233
+    },
+    {
+      "epoch": 1.6807012501796235,
+      "grad_norm": 0.3317048558219616,
+      "learning_rate": 6.79627473027468e-06,
+      "loss": 0.4857,
+      "step": 10234
+    },
+    {
+      "epoch": 1.6808654773879663,
+      "grad_norm": 0.3802312298770638,
+      "learning_rate": 6.795871398720282e-06,
+      "loss": 0.4726,
+      "step": 10235
+    },
+    {
+      "epoch": 1.681029704596309,
+      "grad_norm": 0.2997172652219126,
+      "learning_rate": 6.795468042230464e-06,
+      "loss": 0.4712,
+      "step": 10236
+    },
+    {
+      "epoch": 1.6811939318046516,
+      "grad_norm": 0.3335225931142483,
+      "learning_rate": 6.795064660809604e-06,
+      "loss": 0.4387,
+      "step": 10237
+    },
+    {
+      "epoch": 1.6813581590129945,
+      "grad_norm": 0.2824447347067514,
+      "learning_rate": 6.7946612544620825e-06,
+      "loss": 0.4538,
+      "step": 10238
+    },
+    {
+      "epoch": 1.6815223862213373,
+      "grad_norm": 0.29554559306343914,
+      "learning_rate": 6.794257823192282e-06,
+      "loss": 0.4443,
+      "step": 10239
+    },
+    {
+      "epoch": 1.68168661342968,
+      "grad_norm": 0.2963918421520866,
+      "learning_rate": 6.793854367004585e-06,
+      "loss": 0.4639,
+      "step": 10240
+    },
+    {
+      "epoch": 1.6818508406380226,
+      "grad_norm": 0.36615082692498374,
+      "learning_rate": 6.793450885903374e-06,
+      "loss": 0.4534,
+      "step": 10241
+    },
+    {
+      "epoch": 1.6820150678463655,
+      "grad_norm": 0.32269538559501254,
+      "learning_rate": 6.793047379893027e-06,
+      "loss": 0.4743,
+      "step": 10242
+    },
+    {
+      "epoch": 1.6821792950547083,
+      "grad_norm": 0.35870366829155603,
+      "learning_rate": 6.7926438489779315e-06,
+      "loss": 0.4643,
+      "step": 10243
+    },
+    {
+      "epoch": 1.682343522263051,
+      "grad_norm": 0.3287341790332416,
+      "learning_rate": 6.792240293162467e-06,
+      "loss": 0.4611,
+      "step": 10244
+    },
+    {
+      "epoch": 1.6825077494713936,
+      "grad_norm": 0.36393328763510463,
+      "learning_rate": 6.791836712451018e-06,
+      "loss": 0.4727,
+      "step": 10245
+    },
+    {
+      "epoch": 1.6826719766797364,
+      "grad_norm": 0.32174905680140364,
+      "learning_rate": 6.791433106847968e-06,
+      "loss": 0.4446,
+      "step": 10246
+    },
+    {
+      "epoch": 1.6828362038880793,
+      "grad_norm": 0.341965634713859,
+      "learning_rate": 6.7910294763577e-06,
+      "loss": 0.4682,
+      "step": 10247
+    },
+    {
+      "epoch": 1.683000431096422,
+      "grad_norm": 0.3729904454177416,
+      "learning_rate": 6.790625820984597e-06,
+      "loss": 0.4374,
+      "step": 10248
+    },
+    {
+      "epoch": 1.6831646583047646,
+      "grad_norm": 0.3070748646257641,
+      "learning_rate": 6.790222140733044e-06,
+      "loss": 0.4516,
+      "step": 10249
+    },
+    {
+      "epoch": 1.6833288855131072,
+      "grad_norm": 0.33665130707619334,
+      "learning_rate": 6.789818435607426e-06,
+      "loss": 0.4436,
+      "step": 10250
+    },
+    {
+      "epoch": 1.68349311272145,
+      "grad_norm": 0.2835961697412146,
+      "learning_rate": 6.789414705612128e-06,
+      "loss": 0.4449,
+      "step": 10251
+    },
+    {
+      "epoch": 1.683657339929793,
+      "grad_norm": 0.33480743026989657,
+      "learning_rate": 6.7890109507515315e-06,
+      "loss": 0.4691,
+      "step": 10252
+    },
+    {
+      "epoch": 1.6838215671381356,
+      "grad_norm": 0.3456039394184217,
+      "learning_rate": 6.788607171030025e-06,
+      "loss": 0.4514,
+      "step": 10253
+    },
+    {
+      "epoch": 1.6839857943464782,
+      "grad_norm": 0.27234082924457936,
+      "learning_rate": 6.788203366451993e-06,
+      "loss": 0.4447,
+      "step": 10254
+    },
+    {
+      "epoch": 1.684150021554821,
+      "grad_norm": 0.34880013350903516,
+      "learning_rate": 6.78779953702182e-06,
+      "loss": 0.4709,
+      "step": 10255
+    },
+    {
+      "epoch": 1.684314248763164,
+      "grad_norm": 0.33395070260387977,
+      "learning_rate": 6.787395682743895e-06,
+      "loss": 0.4554,
+      "step": 10256
+    },
+    {
+      "epoch": 1.6844784759715066,
+      "grad_norm": 0.28524661530688,
+      "learning_rate": 6.786991803622602e-06,
+      "loss": 0.485,
+      "step": 10257
+    },
+    {
+      "epoch": 1.6846427031798492,
+      "grad_norm": 0.3068358586063563,
+      "learning_rate": 6.786587899662327e-06,
+      "loss": 0.474,
+      "step": 10258
+    },
+    {
+      "epoch": 1.684806930388192,
+      "grad_norm": 0.2848755321460969,
+      "learning_rate": 6.786183970867458e-06,
+      "loss": 0.4548,
+      "step": 10259
+    },
+    {
+      "epoch": 1.684971157596535,
+      "grad_norm": 0.3643720044509586,
+      "learning_rate": 6.785780017242382e-06,
+      "loss": 0.4391,
+      "step": 10260
+    },
+    {
+      "epoch": 1.6851353848048776,
+      "grad_norm": 0.2542353434490754,
+      "learning_rate": 6.785376038791486e-06,
+      "loss": 0.4491,
+      "step": 10261
+    },
+    {
+      "epoch": 1.6852996120132202,
+      "grad_norm": 0.6157108993224596,
+      "learning_rate": 6.784972035519159e-06,
+      "loss": 0.4631,
+      "step": 10262
+    },
+    {
+      "epoch": 1.685463839221563,
+      "grad_norm": 0.2588069234838315,
+      "learning_rate": 6.784568007429786e-06,
+      "loss": 0.4453,
+      "step": 10263
+    },
+    {
+      "epoch": 1.685628066429906,
+      "grad_norm": 0.3417690274270287,
+      "learning_rate": 6.784163954527755e-06,
+      "loss": 0.4588,
+      "step": 10264
+    },
+    {
+      "epoch": 1.6857922936382486,
+      "grad_norm": 0.41849951712013345,
+      "learning_rate": 6.7837598768174595e-06,
+      "loss": 0.4612,
+      "step": 10265
+    },
+    {
+      "epoch": 1.6859565208465912,
+      "grad_norm": 0.7638550414307494,
+      "learning_rate": 6.783355774303284e-06,
+      "loss": 0.4342,
+      "step": 10266
+    },
+    {
+      "epoch": 1.6861207480549338,
+      "grad_norm": 0.34278375489175383,
+      "learning_rate": 6.782951646989617e-06,
+      "loss": 0.4427,
+      "step": 10267
+    },
+    {
+      "epoch": 1.6862849752632767,
+      "grad_norm": 0.36146825412263345,
+      "learning_rate": 6.7825474948808495e-06,
+      "loss": 0.4443,
+      "step": 10268
+    },
+    {
+      "epoch": 1.6864492024716196,
+      "grad_norm": 0.29089142141868596,
+      "learning_rate": 6.782143317981371e-06,
+      "loss": 0.4551,
+      "step": 10269
+    },
+    {
+      "epoch": 1.6866134296799622,
+      "grad_norm": 0.30031960919690487,
+      "learning_rate": 6.78173911629557e-06,
+      "loss": 0.4703,
+      "step": 10270
+    },
+    {
+      "epoch": 1.6867776568883048,
+      "grad_norm": 0.3250331596103951,
+      "learning_rate": 6.78133488982784e-06,
+      "loss": 0.4714,
+      "step": 10271
+    },
+    {
+      "epoch": 1.6869418840966477,
+      "grad_norm": 0.3971300159279936,
+      "learning_rate": 6.780930638582566e-06,
+      "loss": 0.4546,
+      "step": 10272
+    },
+    {
+      "epoch": 1.6871061113049906,
+      "grad_norm": 0.29162352290664173,
+      "learning_rate": 6.7805263625641394e-06,
+      "loss": 0.4755,
+      "step": 10273
+    },
+    {
+      "epoch": 1.6872703385133332,
+      "grad_norm": 0.26532718169105196,
+      "learning_rate": 6.780122061776957e-06,
+      "loss": 0.443,
+      "step": 10274
+    },
+    {
+      "epoch": 1.6874345657216758,
+      "grad_norm": 0.2722156753431404,
+      "learning_rate": 6.7797177362254035e-06,
+      "loss": 0.4593,
+      "step": 10275
+    },
+    {
+      "epoch": 1.6875987929300187,
+      "grad_norm": 0.26193853080166557,
+      "learning_rate": 6.779313385913872e-06,
+      "loss": 0.4471,
+      "step": 10276
+    },
+    {
+      "epoch": 1.6877630201383615,
+      "grad_norm": 0.37316229355423847,
+      "learning_rate": 6.778909010846754e-06,
+      "loss": 0.4663,
+      "step": 10277
+    },
+    {
+      "epoch": 1.6879272473467042,
+      "grad_norm": 0.3047724813156029,
+      "learning_rate": 6.778504611028443e-06,
+      "loss": 0.4729,
+      "step": 10278
+    },
+    {
+      "epoch": 1.6880914745550468,
+      "grad_norm": 0.6649079988395612,
+      "learning_rate": 6.778100186463331e-06,
+      "loss": 0.4553,
+      "step": 10279
+    },
+    {
+      "epoch": 1.6882557017633897,
+      "grad_norm": 0.29209916143149306,
+      "learning_rate": 6.777695737155809e-06,
+      "loss": 0.4535,
+      "step": 10280
+    },
+    {
+      "epoch": 1.6884199289717325,
+      "grad_norm": 0.31165383824990045,
+      "learning_rate": 6.77729126311027e-06,
+      "loss": 0.4529,
+      "step": 10281
+    },
+    {
+      "epoch": 1.6885841561800752,
+      "grad_norm": 0.30989188157862446,
+      "learning_rate": 6.776886764331108e-06,
+      "loss": 0.4538,
+      "step": 10282
+    },
+    {
+      "epoch": 1.6887483833884178,
+      "grad_norm": 0.3244201798525414,
+      "learning_rate": 6.776482240822715e-06,
+      "loss": 0.4484,
+      "step": 10283
+    },
+    {
+      "epoch": 1.6889126105967605,
+      "grad_norm": 0.29740012853549874,
+      "learning_rate": 6.776077692589484e-06,
+      "loss": 0.4392,
+      "step": 10284
+    },
+    {
+      "epoch": 1.6890768378051033,
+      "grad_norm": 0.31728604602189964,
+      "learning_rate": 6.775673119635812e-06,
+      "loss": 0.4747,
+      "step": 10285
+    },
+    {
+      "epoch": 1.6892410650134462,
+      "grad_norm": 0.25451478660174215,
+      "learning_rate": 6.775268521966091e-06,
+      "loss": 0.4455,
+      "step": 10286
+    },
+    {
+      "epoch": 1.6894052922217888,
+      "grad_norm": 0.28516624095149246,
+      "learning_rate": 6.774863899584714e-06,
+      "loss": 0.4463,
+      "step": 10287
+    },
+    {
+      "epoch": 1.6895695194301315,
+      "grad_norm": 0.38342411007047644,
+      "learning_rate": 6.774459252496077e-06,
+      "loss": 0.4654,
+      "step": 10288
+    },
+    {
+      "epoch": 1.6897337466384743,
+      "grad_norm": 0.26453823657971653,
+      "learning_rate": 6.774054580704576e-06,
+      "loss": 0.4571,
+      "step": 10289
+    },
+    {
+      "epoch": 1.6898979738468172,
+      "grad_norm": 0.2918895135048132,
+      "learning_rate": 6.773649884214603e-06,
+      "loss": 0.443,
+      "step": 10290
+    },
+    {
+      "epoch": 1.6900622010551598,
+      "grad_norm": 0.3251405163584019,
+      "learning_rate": 6.773245163030557e-06,
+      "loss": 0.4578,
+      "step": 10291
+    },
+    {
+      "epoch": 1.6902264282635024,
+      "grad_norm": 0.2912874050807492,
+      "learning_rate": 6.772840417156831e-06,
+      "loss": 0.4555,
+      "step": 10292
+    },
+    {
+      "epoch": 1.6903906554718453,
+      "grad_norm": 0.30295294439343956,
+      "learning_rate": 6.77243564659782e-06,
+      "loss": 0.4802,
+      "step": 10293
+    },
+    {
+      "epoch": 1.6905548826801882,
+      "grad_norm": 0.2914439106468315,
+      "learning_rate": 6.7720308513579255e-06,
+      "loss": 0.4621,
+      "step": 10294
+    },
+    {
+      "epoch": 1.6907191098885308,
+      "grad_norm": 0.30084357916317045,
+      "learning_rate": 6.77162603144154e-06,
+      "loss": 0.4574,
+      "step": 10295
+    },
+    {
+      "epoch": 1.6908833370968734,
+      "grad_norm": 0.31309506377140034,
+      "learning_rate": 6.771221186853059e-06,
+      "loss": 0.4383,
+      "step": 10296
+    },
+    {
+      "epoch": 1.6910475643052163,
+      "grad_norm": 0.32088881385022494,
+      "learning_rate": 6.770816317596882e-06,
+      "loss": 0.4492,
+      "step": 10297
+    },
+    {
+      "epoch": 1.6912117915135592,
+      "grad_norm": 0.2979143781946795,
+      "learning_rate": 6.770411423677406e-06,
+      "loss": 0.4547,
+      "step": 10298
+    },
+    {
+      "epoch": 1.6913760187219018,
+      "grad_norm": 0.24331679942835663,
+      "learning_rate": 6.770006505099029e-06,
+      "loss": 0.4569,
+      "step": 10299
+    },
+    {
+      "epoch": 1.6915402459302444,
+      "grad_norm": 0.29674736960822723,
+      "learning_rate": 6.769601561866147e-06,
+      "loss": 0.452,
+      "step": 10300
+    },
+    {
+      "epoch": 1.691704473138587,
+      "grad_norm": 0.27199210828461945,
+      "learning_rate": 6.76919659398316e-06,
+      "loss": 0.4464,
+      "step": 10301
+    },
+    {
+      "epoch": 1.69186870034693,
+      "grad_norm": 0.26438612536458783,
+      "learning_rate": 6.7687916014544635e-06,
+      "loss": 0.4382,
+      "step": 10302
+    },
+    {
+      "epoch": 1.6920329275552728,
+      "grad_norm": 0.32917048248376596,
+      "learning_rate": 6.768386584284458e-06,
+      "loss": 0.4457,
+      "step": 10303
+    },
+    {
+      "epoch": 1.6921971547636154,
+      "grad_norm": 0.33675980630175334,
+      "learning_rate": 6.767981542477545e-06,
+      "loss": 0.4518,
+      "step": 10304
+    },
+    {
+      "epoch": 1.692361381971958,
+      "grad_norm": 0.32689547936397234,
+      "learning_rate": 6.767576476038119e-06,
+      "loss": 0.457,
+      "step": 10305
+    },
+    {
+      "epoch": 1.692525609180301,
+      "grad_norm": 0.33346484391816406,
+      "learning_rate": 6.767171384970583e-06,
+      "loss": 0.4844,
+      "step": 10306
+    },
+    {
+      "epoch": 1.6926898363886438,
+      "grad_norm": 0.308072789313566,
+      "learning_rate": 6.766766269279333e-06,
+      "loss": 0.4584,
+      "step": 10307
+    },
+    {
+      "epoch": 1.6928540635969864,
+      "grad_norm": 0.287864748449835,
+      "learning_rate": 6.7663611289687725e-06,
+      "loss": 0.4439,
+      "step": 10308
+    },
+    {
+      "epoch": 1.693018290805329,
+      "grad_norm": 0.3082917399631894,
+      "learning_rate": 6.765955964043302e-06,
+      "loss": 0.4588,
+      "step": 10309
+    },
+    {
+      "epoch": 1.693182518013672,
+      "grad_norm": 0.29546090273406916,
+      "learning_rate": 6.765550774507317e-06,
+      "loss": 0.4353,
+      "step": 10310
+    },
+    {
+      "epoch": 1.6933467452220148,
+      "grad_norm": 0.2858280741500684,
+      "learning_rate": 6.765145560365224e-06,
+      "loss": 0.4591,
+      "step": 10311
+    },
+    {
+      "epoch": 1.6935109724303574,
+      "grad_norm": 0.32008558469929194,
+      "learning_rate": 6.7647403216214205e-06,
+      "loss": 0.4529,
+      "step": 10312
+    },
+    {
+      "epoch": 1.6936751996387,
+      "grad_norm": 0.3146694953864376,
+      "learning_rate": 6.764335058280309e-06,
+      "loss": 0.4696,
+      "step": 10313
+    },
+    {
+      "epoch": 1.693839426847043,
+      "grad_norm": 0.3219358380569886,
+      "learning_rate": 6.7639297703462916e-06,
+      "loss": 0.4394,
+      "step": 10314
+    },
+    {
+      "epoch": 1.6940036540553858,
+      "grad_norm": 0.3004978132549195,
+      "learning_rate": 6.763524457823768e-06,
+      "loss": 0.4607,
+      "step": 10315
+    },
+    {
+      "epoch": 1.6941678812637284,
+      "grad_norm": 0.2733719772784754,
+      "learning_rate": 6.763119120717143e-06,
+      "loss": 0.467,
+      "step": 10316
+    },
+    {
+      "epoch": 1.694332108472071,
+      "grad_norm": 0.2706089474104434,
+      "learning_rate": 6.762713759030817e-06,
+      "loss": 0.449,
+      "step": 10317
+    },
+    {
+      "epoch": 1.6944963356804137,
+      "grad_norm": 0.42130473705280985,
+      "learning_rate": 6.762308372769194e-06,
+      "loss": 0.4529,
+      "step": 10318
+    },
+    {
+      "epoch": 1.6946605628887565,
+      "grad_norm": 0.36806728788529874,
+      "learning_rate": 6.761902961936676e-06,
+      "loss": 0.4709,
+      "step": 10319
+    },
+    {
+      "epoch": 1.6948247900970994,
+      "grad_norm": 0.30950909755417966,
+      "learning_rate": 6.761497526537668e-06,
+      "loss": 0.4662,
+      "step": 10320
+    },
+    {
+      "epoch": 1.694989017305442,
+      "grad_norm": 0.25509018884086976,
+      "learning_rate": 6.761092066576569e-06,
+      "loss": 0.4302,
+      "step": 10321
+    },
+    {
+      "epoch": 1.6951532445137847,
+      "grad_norm": 0.3542012605063986,
+      "learning_rate": 6.760686582057787e-06,
+      "loss": 0.4785,
+      "step": 10322
+    },
+    {
+      "epoch": 1.6953174717221275,
+      "grad_norm": 0.27189101345521893,
+      "learning_rate": 6.760281072985725e-06,
+      "loss": 0.471,
+      "step": 10323
+    },
+    {
+      "epoch": 1.6954816989304704,
+      "grad_norm": 0.2690817881529284,
+      "learning_rate": 6.7598755393647855e-06,
+      "loss": 0.4621,
+      "step": 10324
+    },
+    {
+      "epoch": 1.695645926138813,
+      "grad_norm": 0.2796687617198455,
+      "learning_rate": 6.759469981199375e-06,
+      "loss": 0.4409,
+      "step": 10325
+    },
+    {
+      "epoch": 1.6958101533471557,
+      "grad_norm": 0.34430847635917744,
+      "learning_rate": 6.7590643984938965e-06,
+      "loss": 0.4662,
+      "step": 10326
+    },
+    {
+      "epoch": 1.6959743805554985,
+      "grad_norm": 0.3545725970327559,
+      "learning_rate": 6.7586587912527575e-06,
+      "loss": 0.4882,
+      "step": 10327
+    },
+    {
+      "epoch": 1.6961386077638414,
+      "grad_norm": 0.45759440356995434,
+      "learning_rate": 6.758253159480362e-06,
+      "loss": 0.4548,
+      "step": 10328
+    },
+    {
+      "epoch": 1.696302834972184,
+      "grad_norm": 0.3371943712343659,
+      "learning_rate": 6.757847503181114e-06,
+      "loss": 0.4493,
+      "step": 10329
+    },
+    {
+      "epoch": 1.6964670621805267,
+      "grad_norm": 0.38712254849700195,
+      "learning_rate": 6.757441822359422e-06,
+      "loss": 0.4553,
+      "step": 10330
+    },
+    {
+      "epoch": 1.6966312893888695,
+      "grad_norm": 0.367661804842924,
+      "learning_rate": 6.757036117019689e-06,
+      "loss": 0.4605,
+      "step": 10331
+    },
+    {
+      "epoch": 1.6967955165972124,
+      "grad_norm": 0.28797269409836235,
+      "learning_rate": 6.756630387166324e-06,
+      "loss": 0.4527,
+      "step": 10332
+    },
+    {
+      "epoch": 1.696959743805555,
+      "grad_norm": 0.31382333583278105,
+      "learning_rate": 6.756224632803734e-06,
+      "loss": 0.4749,
+      "step": 10333
+    },
+    {
+      "epoch": 1.6971239710138977,
+      "grad_norm": 0.3185347219714828,
+      "learning_rate": 6.755818853936323e-06,
+      "loss": 0.4722,
+      "step": 10334
+    },
+    {
+      "epoch": 1.6972881982222403,
+      "grad_norm": 0.31584226638721585,
+      "learning_rate": 6.755413050568501e-06,
+      "loss": 0.4729,
+      "step": 10335
+    },
+    {
+      "epoch": 1.6974524254305832,
+      "grad_norm": 0.3821555035764605,
+      "learning_rate": 6.755007222704674e-06,
+      "loss": 0.468,
+      "step": 10336
+    },
+    {
+      "epoch": 1.697616652638926,
+      "grad_norm": 0.3381119959701247,
+      "learning_rate": 6.754601370349249e-06,
+      "loss": 0.4728,
+      "step": 10337
+    },
+    {
+      "epoch": 1.6977808798472687,
+      "grad_norm": 0.26045264306014676,
+      "learning_rate": 6.754195493506635e-06,
+      "loss": 0.4314,
+      "step": 10338
+    },
+    {
+      "epoch": 1.6979451070556113,
+      "grad_norm": 0.29197349217378316,
+      "learning_rate": 6.753789592181241e-06,
+      "loss": 0.4717,
+      "step": 10339
+    },
+    {
+      "epoch": 1.6981093342639542,
+      "grad_norm": 0.341468042844591,
+      "learning_rate": 6.753383666377474e-06,
+      "loss": 0.4558,
+      "step": 10340
+    },
+    {
+      "epoch": 1.698273561472297,
+      "grad_norm": 0.2817026922401525,
+      "learning_rate": 6.752977716099744e-06,
+      "loss": 0.4563,
+      "step": 10341
+    },
+    {
+      "epoch": 1.6984377886806397,
+      "grad_norm": 0.29656847976847733,
+      "learning_rate": 6.752571741352459e-06,
+      "loss": 0.4578,
+      "step": 10342
+    },
+    {
+      "epoch": 1.6986020158889823,
+      "grad_norm": 0.28247072661654443,
+      "learning_rate": 6.752165742140029e-06,
+      "loss": 0.439,
+      "step": 10343
+    },
+    {
+      "epoch": 1.6987662430973252,
+      "grad_norm": 0.28792432296827525,
+      "learning_rate": 6.751759718466862e-06,
+      "loss": 0.454,
+      "step": 10344
+    },
+    {
+      "epoch": 1.698930470305668,
+      "grad_norm": 0.3713145261232129,
+      "learning_rate": 6.751353670337371e-06,
+      "loss": 0.4646,
+      "step": 10345
+    },
+    {
+      "epoch": 1.6990946975140107,
+      "grad_norm": 0.3210942286713483,
+      "learning_rate": 6.7509475977559614e-06,
+      "loss": 0.4533,
+      "step": 10346
+    },
+    {
+      "epoch": 1.6992589247223533,
+      "grad_norm": 0.25200029858425016,
+      "learning_rate": 6.750541500727048e-06,
+      "loss": 0.4567,
+      "step": 10347
+    },
+    {
+      "epoch": 1.6994231519306962,
+      "grad_norm": 0.2854102830910053,
+      "learning_rate": 6.7501353792550404e-06,
+      "loss": 0.4514,
+      "step": 10348
+    },
+    {
+      "epoch": 1.699587379139039,
+      "grad_norm": 0.4609283687810497,
+      "learning_rate": 6.749729233344347e-06,
+      "loss": 0.462,
+      "step": 10349
+    },
+    {
+      "epoch": 1.6997516063473816,
+      "grad_norm": 0.2920906286091975,
+      "learning_rate": 6.749323062999382e-06,
+      "loss": 0.4496,
+      "step": 10350
+    },
+    {
+      "epoch": 1.6999158335557243,
+      "grad_norm": 0.34357740596563463,
+      "learning_rate": 6.748916868224554e-06,
+      "loss": 0.467,
+      "step": 10351
+    },
+    {
+      "epoch": 1.700080060764067,
+      "grad_norm": 0.35815701259371335,
+      "learning_rate": 6.748510649024277e-06,
+      "loss": 0.4445,
+      "step": 10352
+    },
+    {
+      "epoch": 1.7002442879724098,
+      "grad_norm": 0.2697156568134558,
+      "learning_rate": 6.748104405402963e-06,
+      "loss": 0.4506,
+      "step": 10353
+    },
+    {
+      "epoch": 1.7004085151807526,
+      "grad_norm": 0.3074531144110258,
+      "learning_rate": 6.747698137365023e-06,
+      "loss": 0.4523,
+      "step": 10354
+    },
+    {
+      "epoch": 1.7005727423890953,
+      "grad_norm": 0.33685069490300174,
+      "learning_rate": 6.7472918449148695e-06,
+      "loss": 0.4725,
+      "step": 10355
+    },
+    {
+      "epoch": 1.700736969597438,
+      "grad_norm": 0.3240178685891051,
+      "learning_rate": 6.746885528056915e-06,
+      "loss": 0.4745,
+      "step": 10356
+    },
+    {
+      "epoch": 1.7009011968057808,
+      "grad_norm": 0.30519208673786224,
+      "learning_rate": 6.746479186795573e-06,
+      "loss": 0.4722,
+      "step": 10357
+    },
+    {
+      "epoch": 1.7010654240141236,
+      "grad_norm": 0.284689567099896,
+      "learning_rate": 6.746072821135258e-06,
+      "loss": 0.469,
+      "step": 10358
+    },
+    {
+      "epoch": 1.7012296512224663,
+      "grad_norm": 0.3561436791068111,
+      "learning_rate": 6.745666431080382e-06,
+      "loss": 0.4613,
+      "step": 10359
+    },
+    {
+      "epoch": 1.701393878430809,
+      "grad_norm": 0.2637723404608694,
+      "learning_rate": 6.745260016635358e-06,
+      "loss": 0.4644,
+      "step": 10360
+    },
+    {
+      "epoch": 1.7015581056391518,
+      "grad_norm": 0.26636461148580687,
+      "learning_rate": 6.744853577804601e-06,
+      "loss": 0.4464,
+      "step": 10361
+    },
+    {
+      "epoch": 1.7017223328474946,
+      "grad_norm": 1.0327093958429159,
+      "learning_rate": 6.744447114592526e-06,
+      "loss": 0.4665,
+      "step": 10362
+    },
+    {
+      "epoch": 1.7018865600558373,
+      "grad_norm": 0.2762304703339504,
+      "learning_rate": 6.744040627003549e-06,
+      "loss": 0.473,
+      "step": 10363
+    },
+    {
+      "epoch": 1.70205078726418,
+      "grad_norm": 0.41730054692606017,
+      "learning_rate": 6.743634115042082e-06,
+      "loss": 0.4709,
+      "step": 10364
+    },
+    {
+      "epoch": 1.7022150144725228,
+      "grad_norm": 0.25904651355072494,
+      "learning_rate": 6.743227578712539e-06,
+      "loss": 0.444,
+      "step": 10365
+    },
+    {
+      "epoch": 1.7023792416808656,
+      "grad_norm": 0.980301374420022,
+      "learning_rate": 6.74282101801934e-06,
+      "loss": 0.4497,
+      "step": 10366
+    },
+    {
+      "epoch": 1.7025434688892083,
+      "grad_norm": 0.299626853323015,
+      "learning_rate": 6.742414432966896e-06,
+      "loss": 0.4636,
+      "step": 10367
+    },
+    {
+      "epoch": 1.702707696097551,
+      "grad_norm": 0.3115433025902387,
+      "learning_rate": 6.742007823559627e-06,
+      "loss": 0.4485,
+      "step": 10368
+    },
+    {
+      "epoch": 1.7028719233058935,
+      "grad_norm": 0.3507573006468152,
+      "learning_rate": 6.741601189801946e-06,
+      "loss": 0.456,
+      "step": 10369
+    },
+    {
+      "epoch": 1.7030361505142364,
+      "grad_norm": 0.36607655888975515,
+      "learning_rate": 6.741194531698271e-06,
+      "loss": 0.4585,
+      "step": 10370
+    },
+    {
+      "epoch": 1.7032003777225793,
+      "grad_norm": 0.283069528317907,
+      "learning_rate": 6.740787849253018e-06,
+      "loss": 0.4582,
+      "step": 10371
+    },
+    {
+      "epoch": 1.703364604930922,
+      "grad_norm": 0.3830473238004654,
+      "learning_rate": 6.740381142470605e-06,
+      "loss": 0.4409,
+      "step": 10372
+    },
+    {
+      "epoch": 1.7035288321392645,
+      "grad_norm": 0.2768537602884907,
+      "learning_rate": 6.739974411355448e-06,
+      "loss": 0.4537,
+      "step": 10373
+    },
+    {
+      "epoch": 1.7036930593476074,
+      "grad_norm": 0.3490536084340397,
+      "learning_rate": 6.739567655911965e-06,
+      "loss": 0.453,
+      "step": 10374
+    },
+    {
+      "epoch": 1.7038572865559503,
+      "grad_norm": 0.36771767087617424,
+      "learning_rate": 6.739160876144575e-06,
+      "loss": 0.4766,
+      "step": 10375
+    },
+    {
+      "epoch": 1.704021513764293,
+      "grad_norm": 0.2878841697769496,
+      "learning_rate": 6.738754072057693e-06,
+      "loss": 0.4759,
+      "step": 10376
+    },
+    {
+      "epoch": 1.7041857409726355,
+      "grad_norm": 0.3020927112421301,
+      "learning_rate": 6.738347243655741e-06,
+      "loss": 0.4874,
+      "step": 10377
+    },
+    {
+      "epoch": 1.7043499681809784,
+      "grad_norm": 0.29368059878224284,
+      "learning_rate": 6.737940390943134e-06,
+      "loss": 0.4657,
+      "step": 10378
+    },
+    {
+      "epoch": 1.7045141953893213,
+      "grad_norm": 0.339697488464221,
+      "learning_rate": 6.7375335139242936e-06,
+      "loss": 0.4602,
+      "step": 10379
+    },
+    {
+      "epoch": 1.704678422597664,
+      "grad_norm": 0.4986765984982801,
+      "learning_rate": 6.737126612603637e-06,
+      "loss": 0.4584,
+      "step": 10380
+    },
+    {
+      "epoch": 1.7048426498060065,
+      "grad_norm": 0.28694916511887175,
+      "learning_rate": 6.7367196869855845e-06,
+      "loss": 0.447,
+      "step": 10381
+    },
+    {
+      "epoch": 1.7050068770143494,
+      "grad_norm": 0.37107709671599887,
+      "learning_rate": 6.736312737074557e-06,
+      "loss": 0.4554,
+      "step": 10382
+    },
+    {
+      "epoch": 1.7051711042226922,
+      "grad_norm": 0.32138012549222017,
+      "learning_rate": 6.735905762874972e-06,
+      "loss": 0.4549,
+      "step": 10383
+    },
+    {
+      "epoch": 1.7053353314310349,
+      "grad_norm": 0.29340690471890696,
+      "learning_rate": 6.73549876439125e-06,
+      "loss": 0.4647,
+      "step": 10384
+    },
+    {
+      "epoch": 1.7054995586393775,
+      "grad_norm": 0.35984360098887486,
+      "learning_rate": 6.735091741627811e-06,
+      "loss": 0.4483,
+      "step": 10385
+    },
+    {
+      "epoch": 1.7056637858477202,
+      "grad_norm": 0.2811151158720807,
+      "learning_rate": 6.734684694589078e-06,
+      "loss": 0.4787,
+      "step": 10386
+    },
+    {
+      "epoch": 1.705828013056063,
+      "grad_norm": 0.3031459072365189,
+      "learning_rate": 6.7342776232794725e-06,
+      "loss": 0.4511,
+      "step": 10387
+    },
+    {
+      "epoch": 1.7059922402644059,
+      "grad_norm": 0.2571523484193839,
+      "learning_rate": 6.733870527703411e-06,
+      "loss": 0.4524,
+      "step": 10388
+    },
+    {
+      "epoch": 1.7061564674727485,
+      "grad_norm": 0.3361334913411509,
+      "learning_rate": 6.733463407865319e-06,
+      "loss": 0.4628,
+      "step": 10389
+    },
+    {
+      "epoch": 1.7063206946810912,
+      "grad_norm": 0.27863555359391223,
+      "learning_rate": 6.733056263769616e-06,
+      "loss": 0.4745,
+      "step": 10390
+    },
+    {
+      "epoch": 1.706484921889434,
+      "grad_norm": 0.31072955149297077,
+      "learning_rate": 6.732649095420726e-06,
+      "loss": 0.457,
+      "step": 10391
+    },
+    {
+      "epoch": 1.7066491490977769,
+      "grad_norm": 0.388313133889824,
+      "learning_rate": 6.73224190282307e-06,
+      "loss": 0.4623,
+      "step": 10392
+    },
+    {
+      "epoch": 1.7068133763061195,
+      "grad_norm": 1.0394211962031206,
+      "learning_rate": 6.73183468598107e-06,
+      "loss": 0.4495,
+      "step": 10393
+    },
+    {
+      "epoch": 1.7069776035144621,
+      "grad_norm": 0.40572430108763996,
+      "learning_rate": 6.73142744489915e-06,
+      "loss": 0.4471,
+      "step": 10394
+    },
+    {
+      "epoch": 1.707141830722805,
+      "grad_norm": 0.242324759201867,
+      "learning_rate": 6.731020179581732e-06,
+      "loss": 0.4317,
+      "step": 10395
+    },
+    {
+      "epoch": 1.7073060579311479,
+      "grad_norm": 0.31949901700226574,
+      "learning_rate": 6.7306128900332405e-06,
+      "loss": 0.4409,
+      "step": 10396
+    },
+    {
+      "epoch": 1.7074702851394905,
+      "grad_norm": 0.2662188504303161,
+      "learning_rate": 6.730205576258099e-06,
+      "loss": 0.4576,
+      "step": 10397
+    },
+    {
+      "epoch": 1.7076345123478331,
+      "grad_norm": 0.34371474136704083,
+      "learning_rate": 6.7297982382607295e-06,
+      "loss": 0.4603,
+      "step": 10398
+    },
+    {
+      "epoch": 1.707798739556176,
+      "grad_norm": 0.5706657024675674,
+      "learning_rate": 6.7293908760455575e-06,
+      "loss": 0.4607,
+      "step": 10399
+    },
+    {
+      "epoch": 1.7079629667645189,
+      "grad_norm": 0.4611294102411975,
+      "learning_rate": 6.728983489617008e-06,
+      "loss": 0.4773,
+      "step": 10400
+    },
+    {
+      "epoch": 1.7081271939728615,
+      "grad_norm": 0.2924681797853017,
+      "learning_rate": 6.728576078979503e-06,
+      "loss": 0.4431,
+      "step": 10401
+    },
+    {
+      "epoch": 1.7082914211812041,
+      "grad_norm": 0.34761075327557933,
+      "learning_rate": 6.7281686441374705e-06,
+      "loss": 0.4424,
+      "step": 10402
+    },
+    {
+      "epoch": 1.7084556483895468,
+      "grad_norm": 0.26384299959420315,
+      "learning_rate": 6.727761185095334e-06,
+      "loss": 0.4524,
+      "step": 10403
+    },
+    {
+      "epoch": 1.7086198755978896,
+      "grad_norm": 0.3084678291137836,
+      "learning_rate": 6.727353701857519e-06,
+      "loss": 0.4659,
+      "step": 10404
+    },
+    {
+      "epoch": 1.7087841028062325,
+      "grad_norm": 0.2901345922105834,
+      "learning_rate": 6.7269461944284525e-06,
+      "loss": 0.4788,
+      "step": 10405
+    },
+    {
+      "epoch": 1.7089483300145751,
+      "grad_norm": 0.27440293975090463,
+      "learning_rate": 6.7265386628125584e-06,
+      "loss": 0.4579,
+      "step": 10406
+    },
+    {
+      "epoch": 1.7091125572229178,
+      "grad_norm": 0.3972081442525347,
+      "learning_rate": 6.726131107014264e-06,
+      "loss": 0.4322,
+      "step": 10407
+    },
+    {
+      "epoch": 1.7092767844312606,
+      "grad_norm": 0.334535225703064,
+      "learning_rate": 6.7257235270379955e-06,
+      "loss": 0.4618,
+      "step": 10408
+    },
+    {
+      "epoch": 1.7094410116396035,
+      "grad_norm": 0.2765674491027994,
+      "learning_rate": 6.725315922888179e-06,
+      "loss": 0.4544,
+      "step": 10409
+    },
+    {
+      "epoch": 1.7096052388479461,
+      "grad_norm": 0.34975187365135374,
+      "learning_rate": 6.724908294569242e-06,
+      "loss": 0.4603,
+      "step": 10410
+    },
+    {
+      "epoch": 1.7097694660562888,
+      "grad_norm": 0.2856071808433099,
+      "learning_rate": 6.724500642085614e-06,
+      "loss": 0.4412,
+      "step": 10411
+    },
+    {
+      "epoch": 1.7099336932646316,
+      "grad_norm": 0.8417017909558503,
+      "learning_rate": 6.7240929654417196e-06,
+      "loss": 0.4508,
+      "step": 10412
+    },
+    {
+      "epoch": 1.7100979204729745,
+      "grad_norm": 0.2939192815435623,
+      "learning_rate": 6.723685264641986e-06,
+      "loss": 0.4666,
+      "step": 10413
+    },
+    {
+      "epoch": 1.7102621476813171,
+      "grad_norm": 0.26199379189429295,
+      "learning_rate": 6.723277539690842e-06,
+      "loss": 0.4474,
+      "step": 10414
+    },
+    {
+      "epoch": 1.7104263748896598,
+      "grad_norm": 0.29879744281049325,
+      "learning_rate": 6.7228697905927184e-06,
+      "loss": 0.4793,
+      "step": 10415
+    },
+    {
+      "epoch": 1.7105906020980026,
+      "grad_norm": 0.2918493779231738,
+      "learning_rate": 6.722462017352042e-06,
+      "loss": 0.4618,
+      "step": 10416
+    },
+    {
+      "epoch": 1.7107548293063455,
+      "grad_norm": 0.3098008325301708,
+      "learning_rate": 6.722054219973242e-06,
+      "loss": 0.4552,
+      "step": 10417
+    },
+    {
+      "epoch": 1.7109190565146881,
+      "grad_norm": 0.3165158551796311,
+      "learning_rate": 6.721646398460745e-06,
+      "loss": 0.4707,
+      "step": 10418
+    },
+    {
+      "epoch": 1.7110832837230308,
+      "grad_norm": 0.30672549273761096,
+      "learning_rate": 6.721238552818983e-06,
+      "loss": 0.4417,
+      "step": 10419
+    },
+    {
+      "epoch": 1.7112475109313734,
+      "grad_norm": 0.2584646055648473,
+      "learning_rate": 6.7208306830523844e-06,
+      "loss": 0.4383,
+      "step": 10420
+    },
+    {
+      "epoch": 1.7114117381397163,
+      "grad_norm": 0.2717102838108191,
+      "learning_rate": 6.720422789165382e-06,
+      "loss": 0.4503,
+      "step": 10421
+    },
+    {
+      "epoch": 1.7115759653480591,
+      "grad_norm": 0.37441039548738764,
+      "learning_rate": 6.720014871162402e-06,
+      "loss": 0.448,
+      "step": 10422
+    },
+    {
+      "epoch": 1.7117401925564018,
+      "grad_norm": 0.3053067938918457,
+      "learning_rate": 6.719606929047875e-06,
+      "loss": 0.4555,
+      "step": 10423
+    },
+    {
+      "epoch": 1.7119044197647444,
+      "grad_norm": 0.33951489319508066,
+      "learning_rate": 6.719198962826234e-06,
+      "loss": 0.4508,
+      "step": 10424
+    },
+    {
+      "epoch": 1.7120686469730872,
+      "grad_norm": 0.40669330323048297,
+      "learning_rate": 6.718790972501909e-06,
+      "loss": 0.4462,
+      "step": 10425
+    },
+    {
+      "epoch": 1.71223287418143,
+      "grad_norm": 0.31788264969946245,
+      "learning_rate": 6.718382958079332e-06,
+      "loss": 0.4448,
+      "step": 10426
+    },
+    {
+      "epoch": 1.7123971013897727,
+      "grad_norm": 0.2972005514734704,
+      "learning_rate": 6.717974919562932e-06,
+      "loss": 0.4546,
+      "step": 10427
+    },
+    {
+      "epoch": 1.7125613285981154,
+      "grad_norm": 0.3405902179433068,
+      "learning_rate": 6.717566856957143e-06,
+      "loss": 0.4451,
+      "step": 10428
+    },
+    {
+      "epoch": 1.7127255558064582,
+      "grad_norm": 0.27053272243654325,
+      "learning_rate": 6.717158770266396e-06,
+      "loss": 0.4484,
+      "step": 10429
+    },
+    {
+      "epoch": 1.712889783014801,
+      "grad_norm": 0.2926589561710632,
+      "learning_rate": 6.716750659495123e-06,
+      "loss": 0.4565,
+      "step": 10430
+    },
+    {
+      "epoch": 1.7130540102231437,
+      "grad_norm": 0.3146258821459904,
+      "learning_rate": 6.716342524647757e-06,
+      "loss": 0.4634,
+      "step": 10431
+    },
+    {
+      "epoch": 1.7132182374314864,
+      "grad_norm": 0.45155855350560037,
+      "learning_rate": 6.71593436572873e-06,
+      "loss": 0.456,
+      "step": 10432
+    },
+    {
+      "epoch": 1.7133824646398292,
+      "grad_norm": 0.31378119401303933,
+      "learning_rate": 6.715526182742477e-06,
+      "loss": 0.4758,
+      "step": 10433
+    },
+    {
+      "epoch": 1.713546691848172,
+      "grad_norm": 0.26711786147597105,
+      "learning_rate": 6.715117975693429e-06,
+      "loss": 0.4353,
+      "step": 10434
+    },
+    {
+      "epoch": 1.7137109190565147,
+      "grad_norm": 0.28845664173005386,
+      "learning_rate": 6.71470974458602e-06,
+      "loss": 0.4627,
+      "step": 10435
+    },
+    {
+      "epoch": 1.7138751462648574,
+      "grad_norm": 0.4823107200435918,
+      "learning_rate": 6.7143014894246846e-06,
+      "loss": 0.4466,
+      "step": 10436
+    },
+    {
+      "epoch": 1.7140393734732,
+      "grad_norm": 0.2786151787980394,
+      "learning_rate": 6.713893210213857e-06,
+      "loss": 0.45,
+      "step": 10437
+    },
+    {
+      "epoch": 1.7142036006815429,
+      "grad_norm": 0.2935876494132793,
+      "learning_rate": 6.713484906957971e-06,
+      "loss": 0.492,
+      "step": 10438
+    },
+    {
+      "epoch": 1.7143678278898857,
+      "grad_norm": 0.28236963636794876,
+      "learning_rate": 6.71307657966146e-06,
+      "loss": 0.4516,
+      "step": 10439
+    },
+    {
+      "epoch": 1.7145320550982284,
+      "grad_norm": 0.6496068241742552,
+      "learning_rate": 6.712668228328761e-06,
+      "loss": 0.4313,
+      "step": 10440
+    },
+    {
+      "epoch": 1.714696282306571,
+      "grad_norm": 0.6204538490273221,
+      "learning_rate": 6.712259852964308e-06,
+      "loss": 0.4574,
+      "step": 10441
+    },
+    {
+      "epoch": 1.7148605095149139,
+      "grad_norm": 0.2793118226529691,
+      "learning_rate": 6.711851453572535e-06,
+      "loss": 0.4689,
+      "step": 10442
+    },
+    {
+      "epoch": 1.7150247367232567,
+      "grad_norm": 0.2744098757190267,
+      "learning_rate": 6.71144303015788e-06,
+      "loss": 0.4442,
+      "step": 10443
+    },
+    {
+      "epoch": 1.7151889639315994,
+      "grad_norm": 0.34174424481501375,
+      "learning_rate": 6.711034582724778e-06,
+      "loss": 0.4657,
+      "step": 10444
+    },
+    {
+      "epoch": 1.715353191139942,
+      "grad_norm": 0.2955658019877106,
+      "learning_rate": 6.710626111277666e-06,
+      "loss": 0.4418,
+      "step": 10445
+    },
+    {
+      "epoch": 1.7155174183482849,
+      "grad_norm": 0.2634439956111268,
+      "learning_rate": 6.710217615820979e-06,
+      "loss": 0.4597,
+      "step": 10446
+    },
+    {
+      "epoch": 1.7156816455566277,
+      "grad_norm": 0.3298989104289385,
+      "learning_rate": 6.709809096359153e-06,
+      "loss": 0.4435,
+      "step": 10447
+    },
+    {
+      "epoch": 1.7158458727649704,
+      "grad_norm": 0.314188393745869,
+      "learning_rate": 6.709400552896627e-06,
+      "loss": 0.4441,
+      "step": 10448
+    },
+    {
+      "epoch": 1.716010099973313,
+      "grad_norm": 0.2666842409720632,
+      "learning_rate": 6.708991985437836e-06,
+      "loss": 0.4552,
+      "step": 10449
+    },
+    {
+      "epoch": 1.7161743271816559,
+      "grad_norm": 0.3231120636371941,
+      "learning_rate": 6.70858339398722e-06,
+      "loss": 0.4706,
+      "step": 10450
+    },
+    {
+      "epoch": 1.7163385543899987,
+      "grad_norm": 0.2752932839799465,
+      "learning_rate": 6.708174778549216e-06,
+      "loss": 0.4474,
+      "step": 10451
+    },
+    {
+      "epoch": 1.7165027815983414,
+      "grad_norm": 0.30247348745969427,
+      "learning_rate": 6.70776613912826e-06,
+      "loss": 0.4403,
+      "step": 10452
+    },
+    {
+      "epoch": 1.716667008806684,
+      "grad_norm": 0.37340998671693054,
+      "learning_rate": 6.707357475728792e-06,
+      "loss": 0.4617,
+      "step": 10453
+    },
+    {
+      "epoch": 1.7168312360150266,
+      "grad_norm": 0.2745424306375286,
+      "learning_rate": 6.70694878835525e-06,
+      "loss": 0.4498,
+      "step": 10454
+    },
+    {
+      "epoch": 1.7169954632233695,
+      "grad_norm": 0.43213806007948663,
+      "learning_rate": 6.706540077012074e-06,
+      "loss": 0.4333,
+      "step": 10455
+    },
+    {
+      "epoch": 1.7171596904317123,
+      "grad_norm": 0.31565939323718656,
+      "learning_rate": 6.7061313417037e-06,
+      "loss": 0.4752,
+      "step": 10456
+    },
+    {
+      "epoch": 1.717323917640055,
+      "grad_norm": 0.5251729427395571,
+      "learning_rate": 6.705722582434569e-06,
+      "loss": 0.4505,
+      "step": 10457
+    },
+    {
+      "epoch": 1.7174881448483976,
+      "grad_norm": 0.3968038722142534,
+      "learning_rate": 6.705313799209123e-06,
+      "loss": 0.46,
+      "step": 10458
+    },
+    {
+      "epoch": 1.7176523720567405,
+      "grad_norm": 0.36237150280496244,
+      "learning_rate": 6.704904992031796e-06,
+      "loss": 0.4517,
+      "step": 10459
+    },
+    {
+      "epoch": 1.7178165992650833,
+      "grad_norm": 0.3266331407754335,
+      "learning_rate": 6.704496160907035e-06,
+      "loss": 0.4553,
+      "step": 10460
+    },
+    {
+      "epoch": 1.717980826473426,
+      "grad_norm": 0.35061228607963685,
+      "learning_rate": 6.704087305839275e-06,
+      "loss": 0.449,
+      "step": 10461
+    },
+    {
+      "epoch": 1.7181450536817686,
+      "grad_norm": 0.3011777257227322,
+      "learning_rate": 6.703678426832958e-06,
+      "loss": 0.4715,
+      "step": 10462
+    },
+    {
+      "epoch": 1.7183092808901115,
+      "grad_norm": 0.3016498684711106,
+      "learning_rate": 6.703269523892526e-06,
+      "loss": 0.4533,
+      "step": 10463
+    },
+    {
+      "epoch": 1.7184735080984543,
+      "grad_norm": 0.3188339916151969,
+      "learning_rate": 6.7028605970224175e-06,
+      "loss": 0.4638,
+      "step": 10464
+    },
+    {
+      "epoch": 1.718637735306797,
+      "grad_norm": 0.2868947673367226,
+      "learning_rate": 6.702451646227077e-06,
+      "loss": 0.4535,
+      "step": 10465
+    },
+    {
+      "epoch": 1.7188019625151396,
+      "grad_norm": 0.3513689796636359,
+      "learning_rate": 6.702042671510945e-06,
+      "loss": 0.4534,
+      "step": 10466
+    },
+    {
+      "epoch": 1.7189661897234825,
+      "grad_norm": 0.2835644141881405,
+      "learning_rate": 6.701633672878463e-06,
+      "loss": 0.4514,
+      "step": 10467
+    },
+    {
+      "epoch": 1.7191304169318253,
+      "grad_norm": 0.4246394600668017,
+      "learning_rate": 6.701224650334072e-06,
+      "loss": 0.4525,
+      "step": 10468
+    },
+    {
+      "epoch": 1.719294644140168,
+      "grad_norm": 0.4832348262491999,
+      "learning_rate": 6.700815603882218e-06,
+      "loss": 0.467,
+      "step": 10469
+    },
+    {
+      "epoch": 1.7194588713485106,
+      "grad_norm": 0.2393193885818513,
+      "learning_rate": 6.700406533527338e-06,
+      "loss": 0.4606,
+      "step": 10470
+    },
+    {
+      "epoch": 1.7196230985568532,
+      "grad_norm": 0.28340704067344497,
+      "learning_rate": 6.699997439273881e-06,
+      "loss": 0.473,
+      "step": 10471
+    },
+    {
+      "epoch": 1.719787325765196,
+      "grad_norm": 0.29619829364482575,
+      "learning_rate": 6.6995883211262855e-06,
+      "loss": 0.467,
+      "step": 10472
+    },
+    {
+      "epoch": 1.719951552973539,
+      "grad_norm": 0.2680896560474138,
+      "learning_rate": 6.699179179088998e-06,
+      "loss": 0.4677,
+      "step": 10473
+    },
+    {
+      "epoch": 1.7201157801818816,
+      "grad_norm": 0.29926212745864955,
+      "learning_rate": 6.69877001316646e-06,
+      "loss": 0.4517,
+      "step": 10474
+    },
+    {
+      "epoch": 1.7202800073902242,
+      "grad_norm": 0.30751652357242387,
+      "learning_rate": 6.698360823363117e-06,
+      "loss": 0.4574,
+      "step": 10475
+    },
+    {
+      "epoch": 1.720444234598567,
+      "grad_norm": 0.3328073603049628,
+      "learning_rate": 6.697951609683412e-06,
+      "loss": 0.4514,
+      "step": 10476
+    },
+    {
+      "epoch": 1.72060846180691,
+      "grad_norm": 0.24904973625178245,
+      "learning_rate": 6.697542372131789e-06,
+      "loss": 0.4691,
+      "step": 10477
+    },
+    {
+      "epoch": 1.7207726890152526,
+      "grad_norm": 0.3055567768298029,
+      "learning_rate": 6.697133110712695e-06,
+      "loss": 0.4415,
+      "step": 10478
+    },
+    {
+      "epoch": 1.7209369162235952,
+      "grad_norm": 0.25915721154981863,
+      "learning_rate": 6.696723825430574e-06,
+      "loss": 0.4787,
+      "step": 10479
+    },
+    {
+      "epoch": 1.721101143431938,
+      "grad_norm": 0.2802197692823335,
+      "learning_rate": 6.696314516289872e-06,
+      "loss": 0.4601,
+      "step": 10480
+    },
+    {
+      "epoch": 1.721265370640281,
+      "grad_norm": 0.3525676666157653,
+      "learning_rate": 6.695905183295031e-06,
+      "loss": 0.4588,
+      "step": 10481
+    },
+    {
+      "epoch": 1.7214295978486236,
+      "grad_norm": 0.2727601150327771,
+      "learning_rate": 6.695495826450501e-06,
+      "loss": 0.4584,
+      "step": 10482
+    },
+    {
+      "epoch": 1.7215938250569662,
+      "grad_norm": 0.32149351547771154,
+      "learning_rate": 6.695086445760725e-06,
+      "loss": 0.456,
+      "step": 10483
+    },
+    {
+      "epoch": 1.721758052265309,
+      "grad_norm": 0.39117096357713665,
+      "learning_rate": 6.694677041230152e-06,
+      "loss": 0.4586,
+      "step": 10484
+    },
+    {
+      "epoch": 1.721922279473652,
+      "grad_norm": 0.3322750991270649,
+      "learning_rate": 6.694267612863227e-06,
+      "loss": 0.4635,
+      "step": 10485
+    },
+    {
+      "epoch": 1.7220865066819946,
+      "grad_norm": 0.32418802055402957,
+      "learning_rate": 6.693858160664394e-06,
+      "loss": 0.447,
+      "step": 10486
+    },
+    {
+      "epoch": 1.7222507338903372,
+      "grad_norm": 0.2675172834941799,
+      "learning_rate": 6.693448684638106e-06,
+      "loss": 0.435,
+      "step": 10487
+    },
+    {
+      "epoch": 1.7224149610986799,
+      "grad_norm": 0.2947083214764386,
+      "learning_rate": 6.693039184788806e-06,
+      "loss": 0.4471,
+      "step": 10488
+    },
+    {
+      "epoch": 1.7225791883070227,
+      "grad_norm": 0.29546155758230347,
+      "learning_rate": 6.692629661120944e-06,
+      "loss": 0.4609,
+      "step": 10489
+    },
+    {
+      "epoch": 1.7227434155153656,
+      "grad_norm": 0.3072763602251934,
+      "learning_rate": 6.692220113638965e-06,
+      "loss": 0.4698,
+      "step": 10490
+    },
+    {
+      "epoch": 1.7229076427237082,
+      "grad_norm": 0.2612485647618418,
+      "learning_rate": 6.691810542347319e-06,
+      "loss": 0.4593,
+      "step": 10491
+    },
+    {
+      "epoch": 1.7230718699320509,
+      "grad_norm": 0.27610091859399394,
+      "learning_rate": 6.691400947250454e-06,
+      "loss": 0.4466,
+      "step": 10492
+    },
+    {
+      "epoch": 1.7232360971403937,
+      "grad_norm": 0.2994420717725685,
+      "learning_rate": 6.690991328352819e-06,
+      "loss": 0.4596,
+      "step": 10493
+    },
+    {
+      "epoch": 1.7234003243487366,
+      "grad_norm": 0.3337391439613997,
+      "learning_rate": 6.690581685658863e-06,
+      "loss": 0.4519,
+      "step": 10494
+    },
+    {
+      "epoch": 1.7235645515570792,
+      "grad_norm": 0.283817248370631,
+      "learning_rate": 6.690172019173035e-06,
+      "loss": 0.4538,
+      "step": 10495
+    },
+    {
+      "epoch": 1.7237287787654219,
+      "grad_norm": 0.27336394201895664,
+      "learning_rate": 6.689762328899783e-06,
+      "loss": 0.4741,
+      "step": 10496
+    },
+    {
+      "epoch": 1.7238930059737647,
+      "grad_norm": 0.4125973348785593,
+      "learning_rate": 6.689352614843557e-06,
+      "loss": 0.476,
+      "step": 10497
+    },
+    {
+      "epoch": 1.7240572331821076,
+      "grad_norm": 0.35607773234902174,
+      "learning_rate": 6.688942877008809e-06,
+      "loss": 0.4361,
+      "step": 10498
+    },
+    {
+      "epoch": 1.7242214603904502,
+      "grad_norm": 0.3158612111856992,
+      "learning_rate": 6.688533115399987e-06,
+      "loss": 0.4657,
+      "step": 10499
+    },
+    {
+      "epoch": 1.7243856875987928,
+      "grad_norm": 0.30310414980860667,
+      "learning_rate": 6.688123330021543e-06,
+      "loss": 0.4622,
+      "step": 10500
+    },
+    {
+      "epoch": 1.7245499148071357,
+      "grad_norm": 0.29849694832345836,
+      "learning_rate": 6.687713520877926e-06,
+      "loss": 0.4584,
+      "step": 10501
+    },
+    {
+      "epoch": 1.7247141420154786,
+      "grad_norm": 0.2651875391881616,
+      "learning_rate": 6.687303687973588e-06,
+      "loss": 0.4753,
+      "step": 10502
+    },
+    {
+      "epoch": 1.7248783692238212,
+      "grad_norm": 0.31880123609608296,
+      "learning_rate": 6.68689383131298e-06,
+      "loss": 0.4422,
+      "step": 10503
+    },
+    {
+      "epoch": 1.7250425964321638,
+      "grad_norm": 0.30651866368968245,
+      "learning_rate": 6.6864839509005534e-06,
+      "loss": 0.4674,
+      "step": 10504
+    },
+    {
+      "epoch": 1.7252068236405065,
+      "grad_norm": 0.273843576978905,
+      "learning_rate": 6.6860740467407594e-06,
+      "loss": 0.4542,
+      "step": 10505
+    },
+    {
+      "epoch": 1.7253710508488493,
+      "grad_norm": 0.33736322841446215,
+      "learning_rate": 6.685664118838051e-06,
+      "loss": 0.4733,
+      "step": 10506
+    },
+    {
+      "epoch": 1.7255352780571922,
+      "grad_norm": 0.28310879953417917,
+      "learning_rate": 6.685254167196879e-06,
+      "loss": 0.4325,
+      "step": 10507
+    },
+    {
+      "epoch": 1.7256995052655348,
+      "grad_norm": 0.30625720125880435,
+      "learning_rate": 6.684844191821698e-06,
+      "loss": 0.4654,
+      "step": 10508
+    },
+    {
+      "epoch": 1.7258637324738775,
+      "grad_norm": 0.4041210802480253,
+      "learning_rate": 6.684434192716959e-06,
+      "loss": 0.4629,
+      "step": 10509
+    },
+    {
+      "epoch": 1.7260279596822203,
+      "grad_norm": 0.3285621834318125,
+      "learning_rate": 6.684024169887115e-06,
+      "loss": 0.4517,
+      "step": 10510
+    },
+    {
+      "epoch": 1.7261921868905632,
+      "grad_norm": 0.5371314885182225,
+      "learning_rate": 6.68361412333662e-06,
+      "loss": 0.4544,
+      "step": 10511
+    },
+    {
+      "epoch": 1.7263564140989058,
+      "grad_norm": 0.3346484736752237,
+      "learning_rate": 6.683204053069928e-06,
+      "loss": 0.4483,
+      "step": 10512
+    },
+    {
+      "epoch": 1.7265206413072485,
+      "grad_norm": 0.265535426288644,
+      "learning_rate": 6.68279395909149e-06,
+      "loss": 0.455,
+      "step": 10513
+    },
+    {
+      "epoch": 1.7266848685155913,
+      "grad_norm": 0.272151834494228,
+      "learning_rate": 6.682383841405764e-06,
+      "loss": 0.4627,
+      "step": 10514
+    },
+    {
+      "epoch": 1.7268490957239342,
+      "grad_norm": 0.2768667797354831,
+      "learning_rate": 6.681973700017202e-06,
+      "loss": 0.4622,
+      "step": 10515
+    },
+    {
+      "epoch": 1.7270133229322768,
+      "grad_norm": 0.2611288871413042,
+      "learning_rate": 6.681563534930258e-06,
+      "loss": 0.4692,
+      "step": 10516
+    },
+    {
+      "epoch": 1.7271775501406195,
+      "grad_norm": 0.30032015789813,
+      "learning_rate": 6.681153346149388e-06,
+      "loss": 0.4425,
+      "step": 10517
+    },
+    {
+      "epoch": 1.7273417773489623,
+      "grad_norm": 0.2765399919998032,
+      "learning_rate": 6.680743133679048e-06,
+      "loss": 0.4418,
+      "step": 10518
+    },
+    {
+      "epoch": 1.7275060045573052,
+      "grad_norm": 0.2637540347081802,
+      "learning_rate": 6.6803328975236904e-06,
+      "loss": 0.4579,
+      "step": 10519
+    },
+    {
+      "epoch": 1.7276702317656478,
+      "grad_norm": 0.405385672011455,
+      "learning_rate": 6.679922637687772e-06,
+      "loss": 0.4557,
+      "step": 10520
+    },
+    {
+      "epoch": 1.7278344589739905,
+      "grad_norm": 0.2911681189325491,
+      "learning_rate": 6.679512354175751e-06,
+      "loss": 0.4586,
+      "step": 10521
+    },
+    {
+      "epoch": 1.727998686182333,
+      "grad_norm": 0.2675727811944536,
+      "learning_rate": 6.679102046992079e-06,
+      "loss": 0.4376,
+      "step": 10522
+    },
+    {
+      "epoch": 1.728162913390676,
+      "grad_norm": 0.2817047223528335,
+      "learning_rate": 6.678691716141217e-06,
+      "loss": 0.4556,
+      "step": 10523
+    },
+    {
+      "epoch": 1.7283271405990188,
+      "grad_norm": 0.30171073273169,
+      "learning_rate": 6.678281361627619e-06,
+      "loss": 0.433,
+      "step": 10524
+    },
+    {
+      "epoch": 1.7284913678073615,
+      "grad_norm": 0.33379983997579427,
+      "learning_rate": 6.677870983455741e-06,
+      "loss": 0.4654,
+      "step": 10525
+    },
+    {
+      "epoch": 1.728655595015704,
+      "grad_norm": 0.2722756149058063,
+      "learning_rate": 6.677460581630043e-06,
+      "loss": 0.4506,
+      "step": 10526
+    },
+    {
+      "epoch": 1.728819822224047,
+      "grad_norm": 0.8752736675862947,
+      "learning_rate": 6.677050156154979e-06,
+      "loss": 0.4607,
+      "step": 10527
+    },
+    {
+      "epoch": 1.7289840494323898,
+      "grad_norm": 0.8938119846179062,
+      "learning_rate": 6.67663970703501e-06,
+      "loss": 0.4628,
+      "step": 10528
+    },
+    {
+      "epoch": 1.7291482766407325,
+      "grad_norm": 0.29687175309764746,
+      "learning_rate": 6.676229234274592e-06,
+      "loss": 0.4473,
+      "step": 10529
+    },
+    {
+      "epoch": 1.729312503849075,
+      "grad_norm": 0.2983652847761805,
+      "learning_rate": 6.675818737878183e-06,
+      "loss": 0.464,
+      "step": 10530
+    },
+    {
+      "epoch": 1.729476731057418,
+      "grad_norm": 0.2882164982567439,
+      "learning_rate": 6.67540821785024e-06,
+      "loss": 0.4656,
+      "step": 10531
+    },
+    {
+      "epoch": 1.7296409582657608,
+      "grad_norm": 0.3597729555329601,
+      "learning_rate": 6.674997674195225e-06,
+      "loss": 0.4854,
+      "step": 10532
+    },
+    {
+      "epoch": 1.7298051854741034,
+      "grad_norm": 0.23773034919142239,
+      "learning_rate": 6.674587106917597e-06,
+      "loss": 0.419,
+      "step": 10533
+    },
+    {
+      "epoch": 1.729969412682446,
+      "grad_norm": 0.3158190970863376,
+      "learning_rate": 6.674176516021812e-06,
+      "loss": 0.457,
+      "step": 10534
+    },
+    {
+      "epoch": 1.730133639890789,
+      "grad_norm": 0.32428977605563186,
+      "learning_rate": 6.6737659015123315e-06,
+      "loss": 0.4549,
+      "step": 10535
+    },
+    {
+      "epoch": 1.7302978670991318,
+      "grad_norm": 0.2866692455873746,
+      "learning_rate": 6.673355263393612e-06,
+      "loss": 0.4351,
+      "step": 10536
+    },
+    {
+      "epoch": 1.7304620943074744,
+      "grad_norm": 0.38551922492958735,
+      "learning_rate": 6.6729446016701195e-06,
+      "loss": 0.475,
+      "step": 10537
+    },
+    {
+      "epoch": 1.730626321515817,
+      "grad_norm": 0.5243113966027705,
+      "learning_rate": 6.672533916346309e-06,
+      "loss": 0.475,
+      "step": 10538
+    },
+    {
+      "epoch": 1.7307905487241597,
+      "grad_norm": 0.2515301096578204,
+      "learning_rate": 6.672123207426644e-06,
+      "loss": 0.4438,
+      "step": 10539
+    },
+    {
+      "epoch": 1.7309547759325026,
+      "grad_norm": 0.3074191829940396,
+      "learning_rate": 6.671712474915583e-06,
+      "loss": 0.4616,
+      "step": 10540
+    },
+    {
+      "epoch": 1.7311190031408454,
+      "grad_norm": 0.2880000746192107,
+      "learning_rate": 6.671301718817586e-06,
+      "loss": 0.4564,
+      "step": 10541
+    },
+    {
+      "epoch": 1.731283230349188,
+      "grad_norm": 0.2962969244487233,
+      "learning_rate": 6.67089093913712e-06,
+      "loss": 0.471,
+      "step": 10542
+    },
+    {
+      "epoch": 1.7314474575575307,
+      "grad_norm": 0.27218947530068305,
+      "learning_rate": 6.67048013587864e-06,
+      "loss": 0.4569,
+      "step": 10543
+    },
+    {
+      "epoch": 1.7316116847658736,
+      "grad_norm": 0.3374040851327258,
+      "learning_rate": 6.670069309046611e-06,
+      "loss": 0.4434,
+      "step": 10544
+    },
+    {
+      "epoch": 1.7317759119742164,
+      "grad_norm": 0.3706121145690397,
+      "learning_rate": 6.669658458645493e-06,
+      "loss": 0.4644,
+      "step": 10545
+    },
+    {
+      "epoch": 1.731940139182559,
+      "grad_norm": 0.3597240408303545,
+      "learning_rate": 6.669247584679751e-06,
+      "loss": 0.4718,
+      "step": 10546
+    },
+    {
+      "epoch": 1.7321043663909017,
+      "grad_norm": 0.34526245876029543,
+      "learning_rate": 6.668836687153844e-06,
+      "loss": 0.4502,
+      "step": 10547
+    },
+    {
+      "epoch": 1.7322685935992446,
+      "grad_norm": 0.25465094933578114,
+      "learning_rate": 6.668425766072239e-06,
+      "loss": 0.4451,
+      "step": 10548
+    },
+    {
+      "epoch": 1.7324328208075874,
+      "grad_norm": 0.2753508045274311,
+      "learning_rate": 6.6680148214393965e-06,
+      "loss": 0.4478,
+      "step": 10549
+    },
+    {
+      "epoch": 1.73259704801593,
+      "grad_norm": 0.2400898508376296,
+      "learning_rate": 6.667603853259779e-06,
+      "loss": 0.4785,
+      "step": 10550
+    },
+    {
+      "epoch": 1.7327612752242727,
+      "grad_norm": 0.2739499399340202,
+      "learning_rate": 6.667192861537851e-06,
+      "loss": 0.4763,
+      "step": 10551
+    },
+    {
+      "epoch": 1.7329255024326156,
+      "grad_norm": 0.2838378785010479,
+      "learning_rate": 6.666781846278077e-06,
+      "loss": 0.4431,
+      "step": 10552
+    },
+    {
+      "epoch": 1.7330897296409584,
+      "grad_norm": 0.3275742837311978,
+      "learning_rate": 6.6663708074849195e-06,
+      "loss": 0.4555,
+      "step": 10553
+    },
+    {
+      "epoch": 1.733253956849301,
+      "grad_norm": 0.35097480269581555,
+      "learning_rate": 6.665959745162845e-06,
+      "loss": 0.4681,
+      "step": 10554
+    },
+    {
+      "epoch": 1.7334181840576437,
+      "grad_norm": 0.9181025874191894,
+      "learning_rate": 6.6655486593163155e-06,
+      "loss": 0.4616,
+      "step": 10555
+    },
+    {
+      "epoch": 1.7335824112659863,
+      "grad_norm": 0.7894254312466302,
+      "learning_rate": 6.665137549949797e-06,
+      "loss": 0.4539,
+      "step": 10556
+    },
+    {
+      "epoch": 1.7337466384743292,
+      "grad_norm": 0.43038099202683344,
+      "learning_rate": 6.664726417067755e-06,
+      "loss": 0.4593,
+      "step": 10557
+    },
+    {
+      "epoch": 1.733910865682672,
+      "grad_norm": 0.4093328771869033,
+      "learning_rate": 6.664315260674654e-06,
+      "loss": 0.4361,
+      "step": 10558
+    },
+    {
+      "epoch": 1.7340750928910147,
+      "grad_norm": 0.3082230765897813,
+      "learning_rate": 6.66390408077496e-06,
+      "loss": 0.4586,
+      "step": 10559
+    },
+    {
+      "epoch": 1.7342393200993573,
+      "grad_norm": 0.32275708574988343,
+      "learning_rate": 6.663492877373138e-06,
+      "loss": 0.4461,
+      "step": 10560
+    },
+    {
+      "epoch": 1.7344035473077002,
+      "grad_norm": 0.31035788426403466,
+      "learning_rate": 6.663081650473655e-06,
+      "loss": 0.4617,
+      "step": 10561
+    },
+    {
+      "epoch": 1.734567774516043,
+      "grad_norm": 0.31601675827456194,
+      "learning_rate": 6.662670400080978e-06,
+      "loss": 0.4706,
+      "step": 10562
+    },
+    {
+      "epoch": 1.7347320017243857,
+      "grad_norm": 0.33103811875238215,
+      "learning_rate": 6.662259126199573e-06,
+      "loss": 0.4919,
+      "step": 10563
+    },
+    {
+      "epoch": 1.7348962289327283,
+      "grad_norm": 0.41689720596390145,
+      "learning_rate": 6.661847828833905e-06,
+      "loss": 0.46,
+      "step": 10564
+    },
+    {
+      "epoch": 1.7350604561410712,
+      "grad_norm": 0.35258440749097886,
+      "learning_rate": 6.661436507988442e-06,
+      "loss": 0.4874,
+      "step": 10565
+    },
+    {
+      "epoch": 1.735224683349414,
+      "grad_norm": 0.38367876380098803,
+      "learning_rate": 6.6610251636676536e-06,
+      "loss": 0.4711,
+      "step": 10566
+    },
+    {
+      "epoch": 1.7353889105577567,
+      "grad_norm": 0.35911279259469897,
+      "learning_rate": 6.660613795876007e-06,
+      "loss": 0.4483,
+      "step": 10567
+    },
+    {
+      "epoch": 1.7355531377660993,
+      "grad_norm": 0.2888127223228145,
+      "learning_rate": 6.6602024046179665e-06,
+      "loss": 0.4533,
+      "step": 10568
+    },
+    {
+      "epoch": 1.7357173649744422,
+      "grad_norm": 0.2798102991264217,
+      "learning_rate": 6.659790989898002e-06,
+      "loss": 0.4543,
+      "step": 10569
+    },
+    {
+      "epoch": 1.735881592182785,
+      "grad_norm": 0.32600684688107623,
+      "learning_rate": 6.659379551720584e-06,
+      "loss": 0.4565,
+      "step": 10570
+    },
+    {
+      "epoch": 1.7360458193911277,
+      "grad_norm": 0.3049264362217875,
+      "learning_rate": 6.658968090090179e-06,
+      "loss": 0.4613,
+      "step": 10571
+    },
+    {
+      "epoch": 1.7362100465994703,
+      "grad_norm": 0.27547939138967337,
+      "learning_rate": 6.658556605011257e-06,
+      "loss": 0.4423,
+      "step": 10572
+    },
+    {
+      "epoch": 1.736374273807813,
+      "grad_norm": 0.31268202539302475,
+      "learning_rate": 6.6581450964882865e-06,
+      "loss": 0.4649,
+      "step": 10573
+    },
+    {
+      "epoch": 1.7365385010161558,
+      "grad_norm": 0.2947480651912641,
+      "learning_rate": 6.6577335645257356e-06,
+      "loss": 0.4667,
+      "step": 10574
+    },
+    {
+      "epoch": 1.7367027282244987,
+      "grad_norm": 0.3829571158229656,
+      "learning_rate": 6.657322009128077e-06,
+      "loss": 0.4471,
+      "step": 10575
+    },
+    {
+      "epoch": 1.7368669554328413,
+      "grad_norm": 0.9143886610861583,
+      "learning_rate": 6.656910430299777e-06,
+      "loss": 0.4641,
+      "step": 10576
+    },
+    {
+      "epoch": 1.737031182641184,
+      "grad_norm": 0.27884086419625775,
+      "learning_rate": 6.65649882804531e-06,
+      "loss": 0.4774,
+      "step": 10577
+    },
+    {
+      "epoch": 1.7371954098495268,
+      "grad_norm": 0.33456889981474314,
+      "learning_rate": 6.656087202369142e-06,
+      "loss": 0.4481,
+      "step": 10578
+    },
+    {
+      "epoch": 1.7373596370578697,
+      "grad_norm": 0.30338483694912066,
+      "learning_rate": 6.655675553275747e-06,
+      "loss": 0.4408,
+      "step": 10579
+    },
+    {
+      "epoch": 1.7375238642662123,
+      "grad_norm": 0.29204954685142154,
+      "learning_rate": 6.655263880769593e-06,
+      "loss": 0.4561,
+      "step": 10580
+    },
+    {
+      "epoch": 1.737688091474555,
+      "grad_norm": 0.28477453765180205,
+      "learning_rate": 6.654852184855153e-06,
+      "loss": 0.473,
+      "step": 10581
+    },
+    {
+      "epoch": 1.7378523186828978,
+      "grad_norm": 0.3611993605343738,
+      "learning_rate": 6.654440465536899e-06,
+      "loss": 0.4816,
+      "step": 10582
+    },
+    {
+      "epoch": 1.7380165458912407,
+      "grad_norm": 0.2788356026362653,
+      "learning_rate": 6.654028722819303e-06,
+      "loss": 0.4878,
+      "step": 10583
+    },
+    {
+      "epoch": 1.7381807730995833,
+      "grad_norm": 0.33148173481642546,
+      "learning_rate": 6.653616956706834e-06,
+      "loss": 0.4545,
+      "step": 10584
+    },
+    {
+      "epoch": 1.738345000307926,
+      "grad_norm": 0.3035096767740506,
+      "learning_rate": 6.653205167203966e-06,
+      "loss": 0.4748,
+      "step": 10585
+    },
+    {
+      "epoch": 1.7385092275162688,
+      "grad_norm": 0.3493164699571448,
+      "learning_rate": 6.652793354315173e-06,
+      "loss": 0.4618,
+      "step": 10586
+    },
+    {
+      "epoch": 1.7386734547246117,
+      "grad_norm": 0.3404445961452418,
+      "learning_rate": 6.652381518044924e-06,
+      "loss": 0.471,
+      "step": 10587
+    },
+    {
+      "epoch": 1.7388376819329543,
+      "grad_norm": 0.33019863805209754,
+      "learning_rate": 6.651969658397696e-06,
+      "loss": 0.4659,
+      "step": 10588
+    },
+    {
+      "epoch": 1.739001909141297,
+      "grad_norm": 0.26683706560142134,
+      "learning_rate": 6.651557775377958e-06,
+      "loss": 0.4395,
+      "step": 10589
+    },
+    {
+      "epoch": 1.7391661363496396,
+      "grad_norm": 0.3633868009011593,
+      "learning_rate": 6.651145868990188e-06,
+      "loss": 0.4582,
+      "step": 10590
+    },
+    {
+      "epoch": 1.7393303635579824,
+      "grad_norm": 0.35217765494100506,
+      "learning_rate": 6.650733939238857e-06,
+      "loss": 0.4377,
+      "step": 10591
+    },
+    {
+      "epoch": 1.7394945907663253,
+      "grad_norm": 0.36982069809605017,
+      "learning_rate": 6.650321986128439e-06,
+      "loss": 0.4503,
+      "step": 10592
+    },
+    {
+      "epoch": 1.739658817974668,
+      "grad_norm": 0.307761272677984,
+      "learning_rate": 6.649910009663408e-06,
+      "loss": 0.4752,
+      "step": 10593
+    },
+    {
+      "epoch": 1.7398230451830106,
+      "grad_norm": 0.28942145592305923,
+      "learning_rate": 6.649498009848239e-06,
+      "loss": 0.4665,
+      "step": 10594
+    },
+    {
+      "epoch": 1.7399872723913534,
+      "grad_norm": 0.27124401320804664,
+      "learning_rate": 6.649085986687406e-06,
+      "loss": 0.4398,
+      "step": 10595
+    },
+    {
+      "epoch": 1.7401514995996963,
+      "grad_norm": 0.3261971310442718,
+      "learning_rate": 6.648673940185388e-06,
+      "loss": 0.4601,
+      "step": 10596
+    },
+    {
+      "epoch": 1.740315726808039,
+      "grad_norm": 0.2843498854236714,
+      "learning_rate": 6.6482618703466545e-06,
+      "loss": 0.4694,
+      "step": 10597
+    },
+    {
+      "epoch": 1.7404799540163816,
+      "grad_norm": 0.37333399523345706,
+      "learning_rate": 6.6478497771756845e-06,
+      "loss": 0.4482,
+      "step": 10598
+    },
+    {
+      "epoch": 1.7406441812247244,
+      "grad_norm": 0.2788317231506996,
+      "learning_rate": 6.647437660676951e-06,
+      "loss": 0.4513,
+      "step": 10599
+    },
+    {
+      "epoch": 1.7408084084330673,
+      "grad_norm": 0.31375562829679565,
+      "learning_rate": 6.647025520854934e-06,
+      "loss": 0.4624,
+      "step": 10600
+    },
+    {
+      "epoch": 1.74097263564141,
+      "grad_norm": 0.3615619772581296,
+      "learning_rate": 6.646613357714107e-06,
+      "loss": 0.4436,
+      "step": 10601
+    },
+    {
+      "epoch": 1.7411368628497526,
+      "grad_norm": 0.3315910833247448,
+      "learning_rate": 6.646201171258946e-06,
+      "loss": 0.4512,
+      "step": 10602
+    },
+    {
+      "epoch": 1.7413010900580954,
+      "grad_norm": 0.2708299030595272,
+      "learning_rate": 6.645788961493929e-06,
+      "loss": 0.4426,
+      "step": 10603
+    },
+    {
+      "epoch": 1.7414653172664383,
+      "grad_norm": 0.27840725603169686,
+      "learning_rate": 6.645376728423533e-06,
+      "loss": 0.4385,
+      "step": 10604
+    },
+    {
+      "epoch": 1.741629544474781,
+      "grad_norm": 0.2569884855267863,
+      "learning_rate": 6.644964472052234e-06,
+      "loss": 0.4575,
+      "step": 10605
+    },
+    {
+      "epoch": 1.7417937716831235,
+      "grad_norm": 0.28479141295519006,
+      "learning_rate": 6.644552192384512e-06,
+      "loss": 0.4424,
+      "step": 10606
+    },
+    {
+      "epoch": 1.7419579988914662,
+      "grad_norm": 0.2951760227251854,
+      "learning_rate": 6.644139889424842e-06,
+      "loss": 0.4657,
+      "step": 10607
+    },
+    {
+      "epoch": 1.742122226099809,
+      "grad_norm": 0.30186113653350727,
+      "learning_rate": 6.643727563177704e-06,
+      "loss": 0.4754,
+      "step": 10608
+    },
+    {
+      "epoch": 1.742286453308152,
+      "grad_norm": 0.2612584685088557,
+      "learning_rate": 6.643315213647575e-06,
+      "loss": 0.4453,
+      "step": 10609
+    },
+    {
+      "epoch": 1.7424506805164945,
+      "grad_norm": 0.6947221272165853,
+      "learning_rate": 6.642902840838934e-06,
+      "loss": 0.4571,
+      "step": 10610
+    },
+    {
+      "epoch": 1.7426149077248372,
+      "grad_norm": 0.3120014744182741,
+      "learning_rate": 6.6424904447562615e-06,
+      "loss": 0.4474,
+      "step": 10611
+    },
+    {
+      "epoch": 1.74277913493318,
+      "grad_norm": 0.3068805224277791,
+      "learning_rate": 6.642078025404033e-06,
+      "loss": 0.4649,
+      "step": 10612
+    },
+    {
+      "epoch": 1.742943362141523,
+      "grad_norm": 0.37785825035727366,
+      "learning_rate": 6.641665582786731e-06,
+      "loss": 0.4277,
+      "step": 10613
+    },
+    {
+      "epoch": 1.7431075893498655,
+      "grad_norm": 0.32257286709905886,
+      "learning_rate": 6.6412531169088325e-06,
+      "loss": 0.4476,
+      "step": 10614
+    },
+    {
+      "epoch": 1.7432718165582082,
+      "grad_norm": 0.27288879363747537,
+      "learning_rate": 6.6408406277748176e-06,
+      "loss": 0.4439,
+      "step": 10615
+    },
+    {
+      "epoch": 1.743436043766551,
+      "grad_norm": 0.29328108390781554,
+      "learning_rate": 6.640428115389168e-06,
+      "loss": 0.4833,
+      "step": 10616
+    },
+    {
+      "epoch": 1.743600270974894,
+      "grad_norm": 0.4506991914279768,
+      "learning_rate": 6.640015579756364e-06,
+      "loss": 0.4652,
+      "step": 10617
+    },
+    {
+      "epoch": 1.7437644981832365,
+      "grad_norm": 0.3087380589024369,
+      "learning_rate": 6.639603020880885e-06,
+      "loss": 0.4793,
+      "step": 10618
+    },
+    {
+      "epoch": 1.7439287253915792,
+      "grad_norm": 0.34152603811976384,
+      "learning_rate": 6.639190438767211e-06,
+      "loss": 0.4486,
+      "step": 10619
+    },
+    {
+      "epoch": 1.744092952599922,
+      "grad_norm": 0.3100709807557343,
+      "learning_rate": 6.638777833419825e-06,
+      "loss": 0.4602,
+      "step": 10620
+    },
+    {
+      "epoch": 1.744257179808265,
+      "grad_norm": 0.3054017542582119,
+      "learning_rate": 6.638365204843209e-06,
+      "loss": 0.4642,
+      "step": 10621
+    },
+    {
+      "epoch": 1.7444214070166075,
+      "grad_norm": 0.34891463707536224,
+      "learning_rate": 6.637952553041842e-06,
+      "loss": 0.4696,
+      "step": 10622
+    },
+    {
+      "epoch": 1.7445856342249502,
+      "grad_norm": 0.3379641255537083,
+      "learning_rate": 6.637539878020205e-06,
+      "loss": 0.46,
+      "step": 10623
+    },
+    {
+      "epoch": 1.7447498614332928,
+      "grad_norm": 0.4668367355877062,
+      "learning_rate": 6.637127179782782e-06,
+      "loss": 0.4542,
+      "step": 10624
+    },
+    {
+      "epoch": 1.7449140886416357,
+      "grad_norm": 0.8209300021742912,
+      "learning_rate": 6.636714458334057e-06,
+      "loss": 0.4623,
+      "step": 10625
+    },
+    {
+      "epoch": 1.7450783158499785,
+      "grad_norm": 0.28705239054834886,
+      "learning_rate": 6.636301713678511e-06,
+      "loss": 0.4507,
+      "step": 10626
+    },
+    {
+      "epoch": 1.7452425430583212,
+      "grad_norm": 0.3355537786733906,
+      "learning_rate": 6.635888945820625e-06,
+      "loss": 0.4302,
+      "step": 10627
+    },
+    {
+      "epoch": 1.7454067702666638,
+      "grad_norm": 0.37192872092910106,
+      "learning_rate": 6.635476154764884e-06,
+      "loss": 0.452,
+      "step": 10628
+    },
+    {
+      "epoch": 1.7455709974750067,
+      "grad_norm": 0.26525282249639376,
+      "learning_rate": 6.635063340515772e-06,
+      "loss": 0.4537,
+      "step": 10629
+    },
+    {
+      "epoch": 1.7457352246833495,
+      "grad_norm": 0.30958646418527175,
+      "learning_rate": 6.63465050307777e-06,
+      "loss": 0.4485,
+      "step": 10630
+    },
+    {
+      "epoch": 1.7458994518916922,
+      "grad_norm": 0.33668426160861986,
+      "learning_rate": 6.634237642455365e-06,
+      "loss": 0.4624,
+      "step": 10631
+    },
+    {
+      "epoch": 1.7460636791000348,
+      "grad_norm": 0.3366417579129416,
+      "learning_rate": 6.633824758653038e-06,
+      "loss": 0.4421,
+      "step": 10632
+    },
+    {
+      "epoch": 1.7462279063083777,
+      "grad_norm": 0.3456242095455785,
+      "learning_rate": 6.633411851675275e-06,
+      "loss": 0.4655,
+      "step": 10633
+    },
+    {
+      "epoch": 1.7463921335167205,
+      "grad_norm": 0.32458901913886634,
+      "learning_rate": 6.63299892152656e-06,
+      "loss": 0.4458,
+      "step": 10634
+    },
+    {
+      "epoch": 1.7465563607250632,
+      "grad_norm": 0.29732800671455417,
+      "learning_rate": 6.632585968211379e-06,
+      "loss": 0.4663,
+      "step": 10635
+    },
+    {
+      "epoch": 1.7467205879334058,
+      "grad_norm": 0.35012774391383994,
+      "learning_rate": 6.632172991734216e-06,
+      "loss": 0.4716,
+      "step": 10636
+    },
+    {
+      "epoch": 1.7468848151417486,
+      "grad_norm": 0.37338988277876983,
+      "learning_rate": 6.6317599920995555e-06,
+      "loss": 0.4591,
+      "step": 10637
+    },
+    {
+      "epoch": 1.7470490423500915,
+      "grad_norm": 0.3799089211952642,
+      "learning_rate": 6.631346969311886e-06,
+      "loss": 0.4604,
+      "step": 10638
+    },
+    {
+      "epoch": 1.7472132695584341,
+      "grad_norm": 0.2894958279907993,
+      "learning_rate": 6.6309339233756894e-06,
+      "loss": 0.4603,
+      "step": 10639
+    },
+    {
+      "epoch": 1.7473774967667768,
+      "grad_norm": 0.3575252568448904,
+      "learning_rate": 6.630520854295455e-06,
+      "loss": 0.4781,
+      "step": 10640
+    },
+    {
+      "epoch": 1.7475417239751194,
+      "grad_norm": 0.27491260487392993,
+      "learning_rate": 6.630107762075668e-06,
+      "loss": 0.4522,
+      "step": 10641
+    },
+    {
+      "epoch": 1.7477059511834623,
+      "grad_norm": 0.28858432050151406,
+      "learning_rate": 6.629694646720815e-06,
+      "loss": 0.4569,
+      "step": 10642
+    },
+    {
+      "epoch": 1.7478701783918051,
+      "grad_norm": 0.25855245592740905,
+      "learning_rate": 6.6292815082353825e-06,
+      "loss": 0.4733,
+      "step": 10643
+    },
+    {
+      "epoch": 1.7480344056001478,
+      "grad_norm": 0.28350226598092165,
+      "learning_rate": 6.628868346623858e-06,
+      "loss": 0.4555,
+      "step": 10644
+    },
+    {
+      "epoch": 1.7481986328084904,
+      "grad_norm": 0.3637068588878069,
+      "learning_rate": 6.6284551618907284e-06,
+      "loss": 0.4548,
+      "step": 10645
+    },
+    {
+      "epoch": 1.7483628600168333,
+      "grad_norm": 0.3527726874714712,
+      "learning_rate": 6.628041954040482e-06,
+      "loss": 0.4544,
+      "step": 10646
+    },
+    {
+      "epoch": 1.7485270872251761,
+      "grad_norm": 0.26024031603615716,
+      "learning_rate": 6.627628723077606e-06,
+      "loss": 0.46,
+      "step": 10647
+    },
+    {
+      "epoch": 1.7486913144335188,
+      "grad_norm": 0.3288597058710364,
+      "learning_rate": 6.627215469006589e-06,
+      "loss": 0.447,
+      "step": 10648
+    },
+    {
+      "epoch": 1.7488555416418614,
+      "grad_norm": 0.25892417052957034,
+      "learning_rate": 6.626802191831919e-06,
+      "loss": 0.4713,
+      "step": 10649
+    },
+    {
+      "epoch": 1.7490197688502043,
+      "grad_norm": 0.4091246145076616,
+      "learning_rate": 6.626388891558086e-06,
+      "loss": 0.4735,
+      "step": 10650
+    },
+    {
+      "epoch": 1.7491839960585471,
+      "grad_norm": 0.25989391360676,
+      "learning_rate": 6.625975568189575e-06,
+      "loss": 0.4459,
+      "step": 10651
+    },
+    {
+      "epoch": 1.7493482232668898,
+      "grad_norm": 0.34067425289731734,
+      "learning_rate": 6.62556222173088e-06,
+      "loss": 0.4395,
+      "step": 10652
+    },
+    {
+      "epoch": 1.7495124504752324,
+      "grad_norm": 0.36332303622096074,
+      "learning_rate": 6.625148852186485e-06,
+      "loss": 0.466,
+      "step": 10653
+    },
+    {
+      "epoch": 1.7496766776835753,
+      "grad_norm": 0.40636172529249653,
+      "learning_rate": 6.624735459560886e-06,
+      "loss": 0.4526,
+      "step": 10654
+    },
+    {
+      "epoch": 1.7498409048919181,
+      "grad_norm": 0.35763187050169826,
+      "learning_rate": 6.6243220438585685e-06,
+      "loss": 0.4465,
+      "step": 10655
+    },
+    {
+      "epoch": 1.7500051321002608,
+      "grad_norm": 0.2928042263761072,
+      "learning_rate": 6.623908605084023e-06,
+      "loss": 0.4452,
+      "step": 10656
+    },
+    {
+      "epoch": 1.7501693593086034,
+      "grad_norm": 0.32058362491336445,
+      "learning_rate": 6.623495143241739e-06,
+      "loss": 0.4424,
+      "step": 10657
+    },
+    {
+      "epoch": 1.750333586516946,
+      "grad_norm": 0.3051735078726969,
+      "learning_rate": 6.623081658336211e-06,
+      "loss": 0.4478,
+      "step": 10658
+    },
+    {
+      "epoch": 1.750497813725289,
+      "grad_norm": 0.2846557283714331,
+      "learning_rate": 6.622668150371925e-06,
+      "loss": 0.4746,
+      "step": 10659
+    },
+    {
+      "epoch": 1.7506620409336318,
+      "grad_norm": 0.2631461096010346,
+      "learning_rate": 6.622254619353377e-06,
+      "loss": 0.4806,
+      "step": 10660
+    },
+    {
+      "epoch": 1.7508262681419744,
+      "grad_norm": 0.27165081583272227,
+      "learning_rate": 6.621841065285054e-06,
+      "loss": 0.4629,
+      "step": 10661
+    },
+    {
+      "epoch": 1.750990495350317,
+      "grad_norm": 0.34897299039333585,
+      "learning_rate": 6.62142748817145e-06,
+      "loss": 0.4733,
+      "step": 10662
+    },
+    {
+      "epoch": 1.75115472255866,
+      "grad_norm": 0.2967986687922061,
+      "learning_rate": 6.621013888017057e-06,
+      "loss": 0.4721,
+      "step": 10663
+    },
+    {
+      "epoch": 1.7513189497670028,
+      "grad_norm": 0.2696164577840737,
+      "learning_rate": 6.620600264826365e-06,
+      "loss": 0.4533,
+      "step": 10664
+    },
+    {
+      "epoch": 1.7514831769753454,
+      "grad_norm": 0.30955662042741533,
+      "learning_rate": 6.620186618603869e-06,
+      "loss": 0.4681,
+      "step": 10665
+    },
+    {
+      "epoch": 1.751647404183688,
+      "grad_norm": 0.2663600728753576,
+      "learning_rate": 6.6197729493540595e-06,
+      "loss": 0.4527,
+      "step": 10666
+    },
+    {
+      "epoch": 1.7518116313920309,
+      "grad_norm": 0.39339227377818786,
+      "learning_rate": 6.619359257081431e-06,
+      "loss": 0.4404,
+      "step": 10667
+    },
+    {
+      "epoch": 1.7519758586003737,
+      "grad_norm": 0.2943000935613823,
+      "learning_rate": 6.618945541790474e-06,
+      "loss": 0.4437,
+      "step": 10668
+    },
+    {
+      "epoch": 1.7521400858087164,
+      "grad_norm": 0.29804979688660815,
+      "learning_rate": 6.618531803485686e-06,
+      "loss": 0.4505,
+      "step": 10669
+    },
+    {
+      "epoch": 1.752304313017059,
+      "grad_norm": 0.29331311411628713,
+      "learning_rate": 6.6181180421715574e-06,
+      "loss": 0.4579,
+      "step": 10670
+    },
+    {
+      "epoch": 1.7524685402254019,
+      "grad_norm": 0.2996471216969548,
+      "learning_rate": 6.617704257852583e-06,
+      "loss": 0.4488,
+      "step": 10671
+    },
+    {
+      "epoch": 1.7526327674337447,
+      "grad_norm": 0.45540878251191624,
+      "learning_rate": 6.6172904505332555e-06,
+      "loss": 0.4634,
+      "step": 10672
+    },
+    {
+      "epoch": 1.7527969946420874,
+      "grad_norm": 0.3484322554734612,
+      "learning_rate": 6.616876620218071e-06,
+      "loss": 0.4566,
+      "step": 10673
+    },
+    {
+      "epoch": 1.75296122185043,
+      "grad_norm": 0.293755535000031,
+      "learning_rate": 6.616462766911525e-06,
+      "loss": 0.4753,
+      "step": 10674
+    },
+    {
+      "epoch": 1.7531254490587727,
+      "grad_norm": 0.33996324247139664,
+      "learning_rate": 6.616048890618111e-06,
+      "loss": 0.4636,
+      "step": 10675
+    },
+    {
+      "epoch": 1.7532896762671155,
+      "grad_norm": 0.41606836561116783,
+      "learning_rate": 6.615634991342323e-06,
+      "loss": 0.4418,
+      "step": 10676
+    },
+    {
+      "epoch": 1.7534539034754584,
+      "grad_norm": 0.3132030327583419,
+      "learning_rate": 6.615221069088658e-06,
+      "loss": 0.4685,
+      "step": 10677
+    },
+    {
+      "epoch": 1.753618130683801,
+      "grad_norm": 0.2993572772718032,
+      "learning_rate": 6.614807123861611e-06,
+      "loss": 0.4639,
+      "step": 10678
+    },
+    {
+      "epoch": 1.7537823578921437,
+      "grad_norm": 0.4478176444935102,
+      "learning_rate": 6.614393155665678e-06,
+      "loss": 0.4475,
+      "step": 10679
+    },
+    {
+      "epoch": 1.7539465851004865,
+      "grad_norm": 0.30048533535202293,
+      "learning_rate": 6.613979164505355e-06,
+      "loss": 0.4403,
+      "step": 10680
+    },
+    {
+      "epoch": 1.7541108123088294,
+      "grad_norm": 0.2908512457695989,
+      "learning_rate": 6.613565150385138e-06,
+      "loss": 0.4559,
+      "step": 10681
+    },
+    {
+      "epoch": 1.754275039517172,
+      "grad_norm": 0.30416948079747663,
+      "learning_rate": 6.613151113309524e-06,
+      "loss": 0.4624,
+      "step": 10682
+    },
+    {
+      "epoch": 1.7544392667255146,
+      "grad_norm": 0.3456723939038373,
+      "learning_rate": 6.6127370532830105e-06,
+      "loss": 0.4458,
+      "step": 10683
+    },
+    {
+      "epoch": 1.7546034939338575,
+      "grad_norm": 0.2688557993639089,
+      "learning_rate": 6.612322970310094e-06,
+      "loss": 0.4395,
+      "step": 10684
+    },
+    {
+      "epoch": 1.7547677211422004,
+      "grad_norm": 0.2857108442653925,
+      "learning_rate": 6.6119088643952715e-06,
+      "loss": 0.4616,
+      "step": 10685
+    },
+    {
+      "epoch": 1.754931948350543,
+      "grad_norm": 0.2946948925070282,
+      "learning_rate": 6.611494735543041e-06,
+      "loss": 0.4428,
+      "step": 10686
+    },
+    {
+      "epoch": 1.7550961755588856,
+      "grad_norm": 0.2816529774982687,
+      "learning_rate": 6.611080583757899e-06,
+      "loss": 0.4602,
+      "step": 10687
+    },
+    {
+      "epoch": 1.7552604027672285,
+      "grad_norm": 0.290157879388694,
+      "learning_rate": 6.610666409044347e-06,
+      "loss": 0.4696,
+      "step": 10688
+    },
+    {
+      "epoch": 1.7554246299755714,
+      "grad_norm": 0.36955605038324496,
+      "learning_rate": 6.61025221140688e-06,
+      "loss": 0.4493,
+      "step": 10689
+    },
+    {
+      "epoch": 1.755588857183914,
+      "grad_norm": 0.3095960553714322,
+      "learning_rate": 6.609837990849999e-06,
+      "loss": 0.4315,
+      "step": 10690
+    },
+    {
+      "epoch": 1.7557530843922566,
+      "grad_norm": 0.47185918895177503,
+      "learning_rate": 6.609423747378199e-06,
+      "loss": 0.4817,
+      "step": 10691
+    },
+    {
+      "epoch": 1.7559173116005993,
+      "grad_norm": 0.28115335143313835,
+      "learning_rate": 6.609009480995984e-06,
+      "loss": 0.4312,
+      "step": 10692
+    },
+    {
+      "epoch": 1.7560815388089421,
+      "grad_norm": 0.30650451169190884,
+      "learning_rate": 6.608595191707851e-06,
+      "loss": 0.4793,
+      "step": 10693
+    },
+    {
+      "epoch": 1.756245766017285,
+      "grad_norm": 0.34971127443580935,
+      "learning_rate": 6.608180879518299e-06,
+      "loss": 0.4604,
+      "step": 10694
+    },
+    {
+      "epoch": 1.7564099932256276,
+      "grad_norm": 0.3094900938094681,
+      "learning_rate": 6.607766544431828e-06,
+      "loss": 0.4473,
+      "step": 10695
+    },
+    {
+      "epoch": 1.7565742204339703,
+      "grad_norm": 0.3187215929719953,
+      "learning_rate": 6.6073521864529395e-06,
+      "loss": 0.4662,
+      "step": 10696
+    },
+    {
+      "epoch": 1.7567384476423131,
+      "grad_norm": 0.3338168033020335,
+      "learning_rate": 6.606937805586132e-06,
+      "loss": 0.4582,
+      "step": 10697
+    },
+    {
+      "epoch": 1.756902674850656,
+      "grad_norm": 0.28507866277484556,
+      "learning_rate": 6.606523401835907e-06,
+      "loss": 0.4632,
+      "step": 10698
+    },
+    {
+      "epoch": 1.7570669020589986,
+      "grad_norm": 1.3822681586763514,
+      "learning_rate": 6.606108975206767e-06,
+      "loss": 0.4564,
+      "step": 10699
+    },
+    {
+      "epoch": 1.7572311292673413,
+      "grad_norm": 0.2818682573662767,
+      "learning_rate": 6.605694525703209e-06,
+      "loss": 0.4744,
+      "step": 10700
+    },
+    {
+      "epoch": 1.7573953564756841,
+      "grad_norm": 0.3082291536949848,
+      "learning_rate": 6.605280053329738e-06,
+      "loss": 0.4434,
+      "step": 10701
+    },
+    {
+      "epoch": 1.757559583684027,
+      "grad_norm": 0.30796809315309,
+      "learning_rate": 6.604865558090854e-06,
+      "loss": 0.4537,
+      "step": 10702
+    },
+    {
+      "epoch": 1.7577238108923696,
+      "grad_norm": 0.2785442087505172,
+      "learning_rate": 6.604451039991059e-06,
+      "loss": 0.4688,
+      "step": 10703
+    },
+    {
+      "epoch": 1.7578880381007123,
+      "grad_norm": 0.26598099036361045,
+      "learning_rate": 6.6040364990348556e-06,
+      "loss": 0.4381,
+      "step": 10704
+    },
+    {
+      "epoch": 1.7580522653090551,
+      "grad_norm": 0.3066077419739165,
+      "learning_rate": 6.603621935226746e-06,
+      "loss": 0.4783,
+      "step": 10705
+    },
+    {
+      "epoch": 1.758216492517398,
+      "grad_norm": 0.28792843857434686,
+      "learning_rate": 6.603207348571231e-06,
+      "loss": 0.4533,
+      "step": 10706
+    },
+    {
+      "epoch": 1.7583807197257406,
+      "grad_norm": 0.2912059270129017,
+      "learning_rate": 6.602792739072817e-06,
+      "loss": 0.4634,
+      "step": 10707
+    },
+    {
+      "epoch": 1.7585449469340833,
+      "grad_norm": 0.2938487886132544,
+      "learning_rate": 6.6023781067360035e-06,
+      "loss": 0.4729,
+      "step": 10708
+    },
+    {
+      "epoch": 1.758709174142426,
+      "grad_norm": 0.32239400228435916,
+      "learning_rate": 6.601963451565297e-06,
+      "loss": 0.461,
+      "step": 10709
+    },
+    {
+      "epoch": 1.7588734013507688,
+      "grad_norm": 0.37857542453820436,
+      "learning_rate": 6.601548773565197e-06,
+      "loss": 0.4429,
+      "step": 10710
+    },
+    {
+      "epoch": 1.7590376285591116,
+      "grad_norm": 0.3075074168414816,
+      "learning_rate": 6.601134072740211e-06,
+      "loss": 0.4403,
+      "step": 10711
+    },
+    {
+      "epoch": 1.7592018557674542,
+      "grad_norm": 0.7430190716889086,
+      "learning_rate": 6.600719349094841e-06,
+      "loss": 0.4448,
+      "step": 10712
+    },
+    {
+      "epoch": 1.7593660829757969,
+      "grad_norm": 0.44458514655301934,
+      "learning_rate": 6.600304602633594e-06,
+      "loss": 0.4563,
+      "step": 10713
+    },
+    {
+      "epoch": 1.7595303101841397,
+      "grad_norm": 0.5312318676070004,
+      "learning_rate": 6.5998898333609715e-06,
+      "loss": 0.4531,
+      "step": 10714
+    },
+    {
+      "epoch": 1.7596945373924826,
+      "grad_norm": 0.29153224613219314,
+      "learning_rate": 6.599475041281479e-06,
+      "loss": 0.4433,
+      "step": 10715
+    },
+    {
+      "epoch": 1.7598587646008252,
+      "grad_norm": 0.8133344159717972,
+      "learning_rate": 6.59906022639962e-06,
+      "loss": 0.4555,
+      "step": 10716
+    },
+    {
+      "epoch": 1.7600229918091679,
+      "grad_norm": 0.29733696086637973,
+      "learning_rate": 6.598645388719905e-06,
+      "loss": 0.448,
+      "step": 10717
+    },
+    {
+      "epoch": 1.7601872190175107,
+      "grad_norm": 0.3456375401995861,
+      "learning_rate": 6.598230528246835e-06,
+      "loss": 0.4571,
+      "step": 10718
+    },
+    {
+      "epoch": 1.7603514462258536,
+      "grad_norm": 0.3311963527266412,
+      "learning_rate": 6.597815644984918e-06,
+      "loss": 0.4506,
+      "step": 10719
+    },
+    {
+      "epoch": 1.7605156734341962,
+      "grad_norm": 0.38540894382757607,
+      "learning_rate": 6.597400738938658e-06,
+      "loss": 0.4657,
+      "step": 10720
+    },
+    {
+      "epoch": 1.7606799006425389,
+      "grad_norm": 0.3139327570135372,
+      "learning_rate": 6.596985810112563e-06,
+      "loss": 0.4371,
+      "step": 10721
+    },
+    {
+      "epoch": 1.7608441278508817,
+      "grad_norm": 0.2927446497689957,
+      "learning_rate": 6.596570858511138e-06,
+      "loss": 0.4377,
+      "step": 10722
+    },
+    {
+      "epoch": 1.7610083550592246,
+      "grad_norm": 0.3170970048028251,
+      "learning_rate": 6.5961558841388915e-06,
+      "loss": 0.4459,
+      "step": 10723
+    },
+    {
+      "epoch": 1.7611725822675672,
+      "grad_norm": 0.2895232317380216,
+      "learning_rate": 6.59574088700033e-06,
+      "loss": 0.4463,
+      "step": 10724
+    },
+    {
+      "epoch": 1.7613368094759099,
+      "grad_norm": 0.30132577539583216,
+      "learning_rate": 6.59532586709996e-06,
+      "loss": 0.4479,
+      "step": 10725
+    },
+    {
+      "epoch": 1.7615010366842525,
+      "grad_norm": 0.297712209809135,
+      "learning_rate": 6.59491082444229e-06,
+      "loss": 0.4812,
+      "step": 10726
+    },
+    {
+      "epoch": 1.7616652638925954,
+      "grad_norm": 0.30872107113301706,
+      "learning_rate": 6.594495759031826e-06,
+      "loss": 0.4532,
+      "step": 10727
+    },
+    {
+      "epoch": 1.7618294911009382,
+      "grad_norm": 0.28669341623275213,
+      "learning_rate": 6.594080670873079e-06,
+      "loss": 0.4717,
+      "step": 10728
+    },
+    {
+      "epoch": 1.7619937183092809,
+      "grad_norm": 0.3179139571288722,
+      "learning_rate": 6.593665559970555e-06,
+      "loss": 0.4736,
+      "step": 10729
+    },
+    {
+      "epoch": 1.7621579455176235,
+      "grad_norm": 0.32432497751797623,
+      "learning_rate": 6.5932504263287636e-06,
+      "loss": 0.4646,
+      "step": 10730
+    },
+    {
+      "epoch": 1.7623221727259664,
+      "grad_norm": 0.29703929983463007,
+      "learning_rate": 6.592835269952212e-06,
+      "loss": 0.4648,
+      "step": 10731
+    },
+    {
+      "epoch": 1.7624863999343092,
+      "grad_norm": 0.34219272476991125,
+      "learning_rate": 6.592420090845412e-06,
+      "loss": 0.4648,
+      "step": 10732
+    },
+    {
+      "epoch": 1.7626506271426519,
+      "grad_norm": 0.3240563250775775,
+      "learning_rate": 6.59200488901287e-06,
+      "loss": 0.4554,
+      "step": 10733
+    },
+    {
+      "epoch": 1.7628148543509945,
+      "grad_norm": 0.3625809225011669,
+      "learning_rate": 6.591589664459096e-06,
+      "loss": 0.4584,
+      "step": 10734
+    },
+    {
+      "epoch": 1.7629790815593374,
+      "grad_norm": 0.6478536789807493,
+      "learning_rate": 6.5911744171886016e-06,
+      "loss": 0.4771,
+      "step": 10735
+    },
+    {
+      "epoch": 1.7631433087676802,
+      "grad_norm": 0.3228958516773163,
+      "learning_rate": 6.590759147205895e-06,
+      "loss": 0.4507,
+      "step": 10736
+    },
+    {
+      "epoch": 1.7633075359760229,
+      "grad_norm": 0.26105724500638866,
+      "learning_rate": 6.590343854515487e-06,
+      "loss": 0.4801,
+      "step": 10737
+    },
+    {
+      "epoch": 1.7634717631843655,
+      "grad_norm": 0.3364783517266538,
+      "learning_rate": 6.589928539121889e-06,
+      "loss": 0.4448,
+      "step": 10738
+    },
+    {
+      "epoch": 1.7636359903927084,
+      "grad_norm": 0.2677527481328561,
+      "learning_rate": 6.589513201029609e-06,
+      "loss": 0.4705,
+      "step": 10739
+    },
+    {
+      "epoch": 1.7638002176010512,
+      "grad_norm": 0.28157915471787903,
+      "learning_rate": 6.589097840243162e-06,
+      "loss": 0.4264,
+      "step": 10740
+    },
+    {
+      "epoch": 1.7639644448093939,
+      "grad_norm": 0.2877591902737641,
+      "learning_rate": 6.588682456767055e-06,
+      "loss": 0.4307,
+      "step": 10741
+    },
+    {
+      "epoch": 1.7641286720177365,
+      "grad_norm": 0.3628181269741827,
+      "learning_rate": 6.588267050605803e-06,
+      "loss": 0.4588,
+      "step": 10742
+    },
+    {
+      "epoch": 1.7642928992260791,
+      "grad_norm": 0.3245284684820889,
+      "learning_rate": 6.587851621763916e-06,
+      "loss": 0.4455,
+      "step": 10743
+    },
+    {
+      "epoch": 1.764457126434422,
+      "grad_norm": 0.4464116642828582,
+      "learning_rate": 6.587436170245907e-06,
+      "loss": 0.446,
+      "step": 10744
+    },
+    {
+      "epoch": 1.7646213536427648,
+      "grad_norm": 0.3618431017109906,
+      "learning_rate": 6.587020696056285e-06,
+      "loss": 0.4663,
+      "step": 10745
+    },
+    {
+      "epoch": 1.7647855808511075,
+      "grad_norm": 0.7835075103239901,
+      "learning_rate": 6.586605199199567e-06,
+      "loss": 0.4709,
+      "step": 10746
+    },
+    {
+      "epoch": 1.7649498080594501,
+      "grad_norm": 0.30836303357783795,
+      "learning_rate": 6.586189679680263e-06,
+      "loss": 0.4497,
+      "step": 10747
+    },
+    {
+      "epoch": 1.765114035267793,
+      "grad_norm": 0.5197716220523791,
+      "learning_rate": 6.585774137502887e-06,
+      "loss": 0.4672,
+      "step": 10748
+    },
+    {
+      "epoch": 1.7652782624761358,
+      "grad_norm": 0.37370315931183823,
+      "learning_rate": 6.585358572671951e-06,
+      "loss": 0.4688,
+      "step": 10749
+    },
+    {
+      "epoch": 1.7654424896844785,
+      "grad_norm": 0.30915793884894854,
+      "learning_rate": 6.584942985191969e-06,
+      "loss": 0.4522,
+      "step": 10750
+    },
+    {
+      "epoch": 1.7656067168928211,
+      "grad_norm": 0.2850764149367957,
+      "learning_rate": 6.584527375067456e-06,
+      "loss": 0.4426,
+      "step": 10751
+    },
+    {
+      "epoch": 1.765770944101164,
+      "grad_norm": 0.3175691534431449,
+      "learning_rate": 6.584111742302924e-06,
+      "loss": 0.4661,
+      "step": 10752
+    },
+    {
+      "epoch": 1.7659351713095068,
+      "grad_norm": 0.2726331377292973,
+      "learning_rate": 6.583696086902888e-06,
+      "loss": 0.4386,
+      "step": 10753
+    },
+    {
+      "epoch": 1.7660993985178495,
+      "grad_norm": 0.3159980495549621,
+      "learning_rate": 6.583280408871862e-06,
+      "loss": 0.4281,
+      "step": 10754
+    },
+    {
+      "epoch": 1.766263625726192,
+      "grad_norm": 0.30476048593373595,
+      "learning_rate": 6.5828647082143624e-06,
+      "loss": 0.4386,
+      "step": 10755
+    },
+    {
+      "epoch": 1.766427852934535,
+      "grad_norm": 0.5184037524101708,
+      "learning_rate": 6.582448984934901e-06,
+      "loss": 0.4596,
+      "step": 10756
+    },
+    {
+      "epoch": 1.7665920801428778,
+      "grad_norm": 0.3719752223361219,
+      "learning_rate": 6.582033239037997e-06,
+      "loss": 0.4586,
+      "step": 10757
+    },
+    {
+      "epoch": 1.7667563073512205,
+      "grad_norm": 0.42817232060131705,
+      "learning_rate": 6.581617470528162e-06,
+      "loss": 0.4301,
+      "step": 10758
+    },
+    {
+      "epoch": 1.766920534559563,
+      "grad_norm": 0.274656189593246,
+      "learning_rate": 6.5812016794099144e-06,
+      "loss": 0.4472,
+      "step": 10759
+    },
+    {
+      "epoch": 1.7670847617679057,
+      "grad_norm": 0.3350111125117778,
+      "learning_rate": 6.5807858656877675e-06,
+      "loss": 0.4384,
+      "step": 10760
+    },
+    {
+      "epoch": 1.7672489889762486,
+      "grad_norm": 0.3961802579412799,
+      "learning_rate": 6.580370029366239e-06,
+      "loss": 0.4514,
+      "step": 10761
+    },
+    {
+      "epoch": 1.7674132161845915,
+      "grad_norm": 0.2810769194265865,
+      "learning_rate": 6.579954170449847e-06,
+      "loss": 0.4389,
+      "step": 10762
+    },
+    {
+      "epoch": 1.767577443392934,
+      "grad_norm": 0.2939767282116434,
+      "learning_rate": 6.5795382889431045e-06,
+      "loss": 0.4733,
+      "step": 10763
+    },
+    {
+      "epoch": 1.7677416706012767,
+      "grad_norm": 0.35139246676594316,
+      "learning_rate": 6.5791223848505305e-06,
+      "loss": 0.46,
+      "step": 10764
+    },
+    {
+      "epoch": 1.7679058978096196,
+      "grad_norm": 0.2923901618315435,
+      "learning_rate": 6.578706458176642e-06,
+      "loss": 0.4569,
+      "step": 10765
+    },
+    {
+      "epoch": 1.7680701250179625,
+      "grad_norm": 0.3661090744129942,
+      "learning_rate": 6.578290508925957e-06,
+      "loss": 0.4572,
+      "step": 10766
+    },
+    {
+      "epoch": 1.768234352226305,
+      "grad_norm": 0.3365126819034772,
+      "learning_rate": 6.577874537102991e-06,
+      "loss": 0.4611,
+      "step": 10767
+    },
+    {
+      "epoch": 1.7683985794346477,
+      "grad_norm": 0.3998899420071638,
+      "learning_rate": 6.577458542712263e-06,
+      "loss": 0.4459,
+      "step": 10768
+    },
+    {
+      "epoch": 1.7685628066429906,
+      "grad_norm": 0.32391092810763855,
+      "learning_rate": 6.5770425257582926e-06,
+      "loss": 0.4644,
+      "step": 10769
+    },
+    {
+      "epoch": 1.7687270338513335,
+      "grad_norm": 0.3529853865117939,
+      "learning_rate": 6.576626486245596e-06,
+      "loss": 0.4698,
+      "step": 10770
+    },
+    {
+      "epoch": 1.768891261059676,
+      "grad_norm": 0.2915016282853621,
+      "learning_rate": 6.576210424178693e-06,
+      "loss": 0.4494,
+      "step": 10771
+    },
+    {
+      "epoch": 1.7690554882680187,
+      "grad_norm": 0.29888633281109755,
+      "learning_rate": 6.575794339562103e-06,
+      "loss": 0.4623,
+      "step": 10772
+    },
+    {
+      "epoch": 1.7692197154763616,
+      "grad_norm": 0.33782021114077626,
+      "learning_rate": 6.575378232400343e-06,
+      "loss": 0.4722,
+      "step": 10773
+    },
+    {
+      "epoch": 1.7693839426847044,
+      "grad_norm": 0.33197396780011823,
+      "learning_rate": 6.574962102697932e-06,
+      "loss": 0.446,
+      "step": 10774
+    },
+    {
+      "epoch": 1.769548169893047,
+      "grad_norm": 0.3498934678387365,
+      "learning_rate": 6.574545950459393e-06,
+      "loss": 0.471,
+      "step": 10775
+    },
+    {
+      "epoch": 1.7697123971013897,
+      "grad_norm": 0.2838093993191466,
+      "learning_rate": 6.574129775689244e-06,
+      "loss": 0.4541,
+      "step": 10776
+    },
+    {
+      "epoch": 1.7698766243097324,
+      "grad_norm": 0.31636922709474874,
+      "learning_rate": 6.573713578392005e-06,
+      "loss": 0.4339,
+      "step": 10777
+    },
+    {
+      "epoch": 1.7700408515180752,
+      "grad_norm": 0.49559821090645556,
+      "learning_rate": 6.5732973585721955e-06,
+      "loss": 0.4695,
+      "step": 10778
+    },
+    {
+      "epoch": 1.770205078726418,
+      "grad_norm": 0.28710379603213987,
+      "learning_rate": 6.572881116234337e-06,
+      "loss": 0.4597,
+      "step": 10779
+    },
+    {
+      "epoch": 1.7703693059347607,
+      "grad_norm": 0.2641290921404477,
+      "learning_rate": 6.57246485138295e-06,
+      "loss": 0.4399,
+      "step": 10780
+    },
+    {
+      "epoch": 1.7705335331431034,
+      "grad_norm": 0.31354698332702563,
+      "learning_rate": 6.572048564022557e-06,
+      "loss": 0.4761,
+      "step": 10781
+    },
+    {
+      "epoch": 1.7706977603514462,
+      "grad_norm": 0.27293143060893843,
+      "learning_rate": 6.571632254157676e-06,
+      "loss": 0.4702,
+      "step": 10782
+    },
+    {
+      "epoch": 1.770861987559789,
+      "grad_norm": 0.2789244175301241,
+      "learning_rate": 6.571215921792832e-06,
+      "loss": 0.4639,
+      "step": 10783
+    },
+    {
+      "epoch": 1.7710262147681317,
+      "grad_norm": 0.3119449859394376,
+      "learning_rate": 6.570799566932545e-06,
+      "loss": 0.4298,
+      "step": 10784
+    },
+    {
+      "epoch": 1.7711904419764744,
+      "grad_norm": 0.4196702056062995,
+      "learning_rate": 6.570383189581336e-06,
+      "loss": 0.4453,
+      "step": 10785
+    },
+    {
+      "epoch": 1.7713546691848172,
+      "grad_norm": 0.3100662001480286,
+      "learning_rate": 6.569966789743731e-06,
+      "loss": 0.4666,
+      "step": 10786
+    },
+    {
+      "epoch": 1.77151889639316,
+      "grad_norm": 0.3417203555284162,
+      "learning_rate": 6.569550367424248e-06,
+      "loss": 0.4489,
+      "step": 10787
+    },
+    {
+      "epoch": 1.7716831236015027,
+      "grad_norm": 0.2997760203724878,
+      "learning_rate": 6.569133922627413e-06,
+      "loss": 0.4672,
+      "step": 10788
+    },
+    {
+      "epoch": 1.7718473508098453,
+      "grad_norm": 0.3650237571988705,
+      "learning_rate": 6.5687174553577475e-06,
+      "loss": 0.4398,
+      "step": 10789
+    },
+    {
+      "epoch": 1.7720115780181882,
+      "grad_norm": 0.40735904021853114,
+      "learning_rate": 6.568300965619775e-06,
+      "loss": 0.4582,
+      "step": 10790
+    },
+    {
+      "epoch": 1.772175805226531,
+      "grad_norm": 0.41041486738406796,
+      "learning_rate": 6.56788445341802e-06,
+      "loss": 0.4596,
+      "step": 10791
+    },
+    {
+      "epoch": 1.7723400324348737,
+      "grad_norm": 0.34668412608676713,
+      "learning_rate": 6.567467918757004e-06,
+      "loss": 0.4445,
+      "step": 10792
+    },
+    {
+      "epoch": 1.7725042596432163,
+      "grad_norm": 0.31068727442963523,
+      "learning_rate": 6.5670513616412525e-06,
+      "loss": 0.4421,
+      "step": 10793
+    },
+    {
+      "epoch": 1.772668486851559,
+      "grad_norm": 0.3138059771083177,
+      "learning_rate": 6.5666347820752895e-06,
+      "loss": 0.4358,
+      "step": 10794
+    },
+    {
+      "epoch": 1.7728327140599018,
+      "grad_norm": 0.2586739324491047,
+      "learning_rate": 6.5662181800636395e-06,
+      "loss": 0.4436,
+      "step": 10795
+    },
+    {
+      "epoch": 1.7729969412682447,
+      "grad_norm": 0.30324835791213506,
+      "learning_rate": 6.565801555610827e-06,
+      "loss": 0.4382,
+      "step": 10796
+    },
+    {
+      "epoch": 1.7731611684765873,
+      "grad_norm": 0.33744280019832007,
+      "learning_rate": 6.565384908721379e-06,
+      "loss": 0.4297,
+      "step": 10797
+    },
+    {
+      "epoch": 1.77332539568493,
+      "grad_norm": 0.2838435130478732,
+      "learning_rate": 6.564968239399816e-06,
+      "loss": 0.4328,
+      "step": 10798
+    },
+    {
+      "epoch": 1.7734896228932728,
+      "grad_norm": 0.29658021102022636,
+      "learning_rate": 6.564551547650668e-06,
+      "loss": 0.4576,
+      "step": 10799
+    },
+    {
+      "epoch": 1.7736538501016157,
+      "grad_norm": 0.3007907744005072,
+      "learning_rate": 6.564134833478459e-06,
+      "loss": 0.4695,
+      "step": 10800
+    },
+    {
+      "epoch": 1.7738386057110012,
+      "grad_norm": 0.285764971226959,
+      "learning_rate": 6.563718096887715e-06,
+      "loss": 0.4597,
+      "step": 10801
+    },
+    {
+      "epoch": 1.774002832919344,
+      "grad_norm": 0.28967808908806614,
+      "learning_rate": 6.56330133788296e-06,
+      "loss": 0.4587,
+      "step": 10802
+    },
+    {
+      "epoch": 1.7741670601276867,
+      "grad_norm": 0.2703467117021694,
+      "learning_rate": 6.562884556468725e-06,
+      "loss": 0.445,
+      "step": 10803
+    },
+    {
+      "epoch": 1.7743312873360293,
+      "grad_norm": 0.2627977347482234,
+      "learning_rate": 6.562467752649532e-06,
+      "loss": 0.4707,
+      "step": 10804
+    },
+    {
+      "epoch": 1.7744955145443722,
+      "grad_norm": 0.2966307541960856,
+      "learning_rate": 6.562050926429912e-06,
+      "loss": 0.4636,
+      "step": 10805
+    },
+    {
+      "epoch": 1.774659741752715,
+      "grad_norm": 0.39316975585417724,
+      "learning_rate": 6.561634077814389e-06,
+      "loss": 0.4592,
+      "step": 10806
+    },
+    {
+      "epoch": 1.7748239689610577,
+      "grad_norm": 0.28644623247691015,
+      "learning_rate": 6.561217206807491e-06,
+      "loss": 0.4627,
+      "step": 10807
+    },
+    {
+      "epoch": 1.7749881961694003,
+      "grad_norm": 0.38443342350967297,
+      "learning_rate": 6.5608003134137465e-06,
+      "loss": 0.4639,
+      "step": 10808
+    },
+    {
+      "epoch": 1.7751524233777431,
+      "grad_norm": 0.388144163612288,
+      "learning_rate": 6.560383397637684e-06,
+      "loss": 0.4733,
+      "step": 10809
+    },
+    {
+      "epoch": 1.775316650586086,
+      "grad_norm": 0.32983894347945997,
+      "learning_rate": 6.55996645948383e-06,
+      "loss": 0.4477,
+      "step": 10810
+    },
+    {
+      "epoch": 1.7754808777944286,
+      "grad_norm": 0.4331970351898739,
+      "learning_rate": 6.559549498956715e-06,
+      "loss": 0.4576,
+      "step": 10811
+    },
+    {
+      "epoch": 1.7756451050027713,
+      "grad_norm": 0.298687059323067,
+      "learning_rate": 6.559132516060865e-06,
+      "loss": 0.4534,
+      "step": 10812
+    },
+    {
+      "epoch": 1.775809332211114,
+      "grad_norm": 0.2718148542045194,
+      "learning_rate": 6.55871551080081e-06,
+      "loss": 0.4431,
+      "step": 10813
+    },
+    {
+      "epoch": 1.7759735594194568,
+      "grad_norm": 0.3192924286136304,
+      "learning_rate": 6.558298483181078e-06,
+      "loss": 0.4637,
+      "step": 10814
+    },
+    {
+      "epoch": 1.7761377866277996,
+      "grad_norm": 0.3229275703672246,
+      "learning_rate": 6.5578814332062e-06,
+      "loss": 0.4517,
+      "step": 10815
+    },
+    {
+      "epoch": 1.7763020138361423,
+      "grad_norm": 0.3074386241153964,
+      "learning_rate": 6.557464360880704e-06,
+      "loss": 0.4425,
+      "step": 10816
+    },
+    {
+      "epoch": 1.776466241044485,
+      "grad_norm": 0.2904906686506346,
+      "learning_rate": 6.557047266209123e-06,
+      "loss": 0.4571,
+      "step": 10817
+    },
+    {
+      "epoch": 1.7766304682528278,
+      "grad_norm": 0.27502918069961163,
+      "learning_rate": 6.556630149195984e-06,
+      "loss": 0.4583,
+      "step": 10818
+    },
+    {
+      "epoch": 1.7767946954611706,
+      "grad_norm": 0.34980850402681873,
+      "learning_rate": 6.5562130098458175e-06,
+      "loss": 0.4665,
+      "step": 10819
+    },
+    {
+      "epoch": 1.7769589226695133,
+      "grad_norm": 0.32905826580936176,
+      "learning_rate": 6.555795848163155e-06,
+      "loss": 0.4326,
+      "step": 10820
+    },
+    {
+      "epoch": 1.777123149877856,
+      "grad_norm": 0.31399784018566634,
+      "learning_rate": 6.5553786641525266e-06,
+      "loss": 0.4728,
+      "step": 10821
+    },
+    {
+      "epoch": 1.7772873770861988,
+      "grad_norm": 0.28220727942929263,
+      "learning_rate": 6.554961457818464e-06,
+      "loss": 0.4689,
+      "step": 10822
+    },
+    {
+      "epoch": 1.7774516042945416,
+      "grad_norm": 0.350203078583294,
+      "learning_rate": 6.554544229165498e-06,
+      "loss": 0.4778,
+      "step": 10823
+    },
+    {
+      "epoch": 1.7776158315028843,
+      "grad_norm": 0.43535540364275027,
+      "learning_rate": 6.55412697819816e-06,
+      "loss": 0.4483,
+      "step": 10824
+    },
+    {
+      "epoch": 1.777780058711227,
+      "grad_norm": 0.30355242827628565,
+      "learning_rate": 6.553709704920984e-06,
+      "loss": 0.4592,
+      "step": 10825
+    },
+    {
+      "epoch": 1.7779442859195698,
+      "grad_norm": 0.3460292247959962,
+      "learning_rate": 6.553292409338499e-06,
+      "loss": 0.4647,
+      "step": 10826
+    },
+    {
+      "epoch": 1.7781085131279126,
+      "grad_norm": 0.3144679395520833,
+      "learning_rate": 6.552875091455237e-06,
+      "loss": 0.4783,
+      "step": 10827
+    },
+    {
+      "epoch": 1.7782727403362553,
+      "grad_norm": 0.2956641392050019,
+      "learning_rate": 6.552457751275732e-06,
+      "loss": 0.463,
+      "step": 10828
+    },
+    {
+      "epoch": 1.778436967544598,
+      "grad_norm": 1.2627729567562462,
+      "learning_rate": 6.5520403888045175e-06,
+      "loss": 0.4429,
+      "step": 10829
+    },
+    {
+      "epoch": 1.7786011947529405,
+      "grad_norm": 0.2850528949825912,
+      "learning_rate": 6.551623004046125e-06,
+      "loss": 0.4428,
+      "step": 10830
+    },
+    {
+      "epoch": 1.7787654219612834,
+      "grad_norm": 0.3196432030904305,
+      "learning_rate": 6.551205597005088e-06,
+      "loss": 0.4663,
+      "step": 10831
+    },
+    {
+      "epoch": 1.7789296491696263,
+      "grad_norm": 0.2575516148408011,
+      "learning_rate": 6.550788167685941e-06,
+      "loss": 0.4293,
+      "step": 10832
+    },
+    {
+      "epoch": 1.779093876377969,
+      "grad_norm": 0.25567605917187286,
+      "learning_rate": 6.550370716093215e-06,
+      "loss": 0.4511,
+      "step": 10833
+    },
+    {
+      "epoch": 1.7792581035863115,
+      "grad_norm": 0.3174637713246571,
+      "learning_rate": 6.549953242231447e-06,
+      "loss": 0.4611,
+      "step": 10834
+    },
+    {
+      "epoch": 1.7794223307946544,
+      "grad_norm": 0.5439191573102424,
+      "learning_rate": 6.549535746105171e-06,
+      "loss": 0.4454,
+      "step": 10835
+    },
+    {
+      "epoch": 1.7795865580029973,
+      "grad_norm": 0.3170447961287215,
+      "learning_rate": 6.549118227718918e-06,
+      "loss": 0.4606,
+      "step": 10836
+    },
+    {
+      "epoch": 1.77975078521134,
+      "grad_norm": 0.36447664378677996,
+      "learning_rate": 6.548700687077226e-06,
+      "loss": 0.4711,
+      "step": 10837
+    },
+    {
+      "epoch": 1.7799150124196825,
+      "grad_norm": 0.2904792276977439,
+      "learning_rate": 6.5482831241846284e-06,
+      "loss": 0.4585,
+      "step": 10838
+    },
+    {
+      "epoch": 1.7800792396280254,
+      "grad_norm": 0.2719058857151288,
+      "learning_rate": 6.547865539045661e-06,
+      "loss": 0.467,
+      "step": 10839
+    },
+    {
+      "epoch": 1.7802434668363682,
+      "grad_norm": 0.3222765664208403,
+      "learning_rate": 6.54744793166486e-06,
+      "loss": 0.4411,
+      "step": 10840
+    },
+    {
+      "epoch": 1.7804076940447109,
+      "grad_norm": 0.3264407930977664,
+      "learning_rate": 6.547030302046759e-06,
+      "loss": 0.4493,
+      "step": 10841
+    },
+    {
+      "epoch": 1.7805719212530535,
+      "grad_norm": 0.3601154985976864,
+      "learning_rate": 6.546612650195897e-06,
+      "loss": 0.4593,
+      "step": 10842
+    },
+    {
+      "epoch": 1.7807361484613964,
+      "grad_norm": 0.29387278390753235,
+      "learning_rate": 6.546194976116805e-06,
+      "loss": 0.4364,
+      "step": 10843
+    },
+    {
+      "epoch": 1.7809003756697392,
+      "grad_norm": 0.32486471922003535,
+      "learning_rate": 6.545777279814024e-06,
+      "loss": 0.4613,
+      "step": 10844
+    },
+    {
+      "epoch": 1.7810646028780819,
+      "grad_norm": 0.3041127864331923,
+      "learning_rate": 6.5453595612920885e-06,
+      "loss": 0.4567,
+      "step": 10845
+    },
+    {
+      "epoch": 1.7812288300864245,
+      "grad_norm": 0.3548441166557751,
+      "learning_rate": 6.544941820555536e-06,
+      "loss": 0.4446,
+      "step": 10846
+    },
+    {
+      "epoch": 1.7813930572947672,
+      "grad_norm": 0.33470252658410204,
+      "learning_rate": 6.544524057608904e-06,
+      "loss": 0.4537,
+      "step": 10847
+    },
+    {
+      "epoch": 1.78155728450311,
+      "grad_norm": 0.2989380276370097,
+      "learning_rate": 6.544106272456727e-06,
+      "loss": 0.4557,
+      "step": 10848
+    },
+    {
+      "epoch": 1.7817215117114529,
+      "grad_norm": 0.2997943871744751,
+      "learning_rate": 6.543688465103548e-06,
+      "loss": 0.4559,
+      "step": 10849
+    },
+    {
+      "epoch": 1.7818857389197955,
+      "grad_norm": 0.2936293590695143,
+      "learning_rate": 6.5432706355538985e-06,
+      "loss": 0.4467,
+      "step": 10850
+    },
+    {
+      "epoch": 1.7820499661281382,
+      "grad_norm": 0.25972399431795706,
+      "learning_rate": 6.5428527838123215e-06,
+      "loss": 0.4515,
+      "step": 10851
+    },
+    {
+      "epoch": 1.782214193336481,
+      "grad_norm": 0.34117830092565526,
+      "learning_rate": 6.5424349098833534e-06,
+      "loss": 0.4497,
+      "step": 10852
+    },
+    {
+      "epoch": 1.7823784205448239,
+      "grad_norm": 0.40292122778145445,
+      "learning_rate": 6.542017013771531e-06,
+      "loss": 0.4445,
+      "step": 10853
+    },
+    {
+      "epoch": 1.7825426477531665,
+      "grad_norm": 0.29530119778119707,
+      "learning_rate": 6.541599095481396e-06,
+      "loss": 0.465,
+      "step": 10854
+    },
+    {
+      "epoch": 1.7827068749615091,
+      "grad_norm": 0.31448321131899243,
+      "learning_rate": 6.541181155017487e-06,
+      "loss": 0.4569,
+      "step": 10855
+    },
+    {
+      "epoch": 1.782871102169852,
+      "grad_norm": 0.28643134276254606,
+      "learning_rate": 6.540763192384341e-06,
+      "loss": 0.455,
+      "step": 10856
+    },
+    {
+      "epoch": 1.7830353293781949,
+      "grad_norm": 0.30659806432999265,
+      "learning_rate": 6.540345207586498e-06,
+      "loss": 0.4507,
+      "step": 10857
+    },
+    {
+      "epoch": 1.7831995565865375,
+      "grad_norm": 0.26947636257084084,
+      "learning_rate": 6.5399272006285e-06,
+      "loss": 0.4635,
+      "step": 10858
+    },
+    {
+      "epoch": 1.7833637837948801,
+      "grad_norm": 0.30243946446801434,
+      "learning_rate": 6.539509171514888e-06,
+      "loss": 0.4348,
+      "step": 10859
+    },
+    {
+      "epoch": 1.783528011003223,
+      "grad_norm": 0.29964948420515214,
+      "learning_rate": 6.539091120250196e-06,
+      "loss": 0.4556,
+      "step": 10860
+    },
+    {
+      "epoch": 1.7836922382115659,
+      "grad_norm": 0.4948757323995327,
+      "learning_rate": 6.53867304683897e-06,
+      "loss": 0.4437,
+      "step": 10861
+    },
+    {
+      "epoch": 1.7838564654199085,
+      "grad_norm": 0.27079970013608673,
+      "learning_rate": 6.538254951285747e-06,
+      "loss": 0.4639,
+      "step": 10862
+    },
+    {
+      "epoch": 1.7840206926282511,
+      "grad_norm": 0.4382147974067117,
+      "learning_rate": 6.5378368335950716e-06,
+      "loss": 0.4524,
+      "step": 10863
+    },
+    {
+      "epoch": 1.7841849198365938,
+      "grad_norm": 0.3276204275339086,
+      "learning_rate": 6.537418693771484e-06,
+      "loss": 0.4729,
+      "step": 10864
+    },
+    {
+      "epoch": 1.7843491470449366,
+      "grad_norm": 0.44798273637267755,
+      "learning_rate": 6.537000531819523e-06,
+      "loss": 0.4515,
+      "step": 10865
+    },
+    {
+      "epoch": 1.7845133742532795,
+      "grad_norm": 0.2958121807443287,
+      "learning_rate": 6.536582347743732e-06,
+      "loss": 0.4713,
+      "step": 10866
+    },
+    {
+      "epoch": 1.7846776014616221,
+      "grad_norm": 0.3204812397349806,
+      "learning_rate": 6.536164141548654e-06,
+      "loss": 0.4734,
+      "step": 10867
+    },
+    {
+      "epoch": 1.7848418286699648,
+      "grad_norm": 0.27522351558354125,
+      "learning_rate": 6.535745913238831e-06,
+      "loss": 0.4507,
+      "step": 10868
+    },
+    {
+      "epoch": 1.7850060558783076,
+      "grad_norm": 0.31930333118326126,
+      "learning_rate": 6.535327662818804e-06,
+      "loss": 0.4608,
+      "step": 10869
+    },
+    {
+      "epoch": 1.7851702830866505,
+      "grad_norm": 0.3058979140390943,
+      "learning_rate": 6.534909390293115e-06,
+      "loss": 0.4617,
+      "step": 10870
+    },
+    {
+      "epoch": 1.7853345102949931,
+      "grad_norm": 0.29387217939437715,
+      "learning_rate": 6.534491095666308e-06,
+      "loss": 0.4646,
+      "step": 10871
+    },
+    {
+      "epoch": 1.7854987375033358,
+      "grad_norm": 0.2930336057878312,
+      "learning_rate": 6.534072778942927e-06,
+      "loss": 0.4641,
+      "step": 10872
+    },
+    {
+      "epoch": 1.7856629647116786,
+      "grad_norm": 0.2594508867244782,
+      "learning_rate": 6.533654440127514e-06,
+      "loss": 0.4437,
+      "step": 10873
+    },
+    {
+      "epoch": 1.7858271919200215,
+      "grad_norm": 0.2716593847223371,
+      "learning_rate": 6.5332360792246125e-06,
+      "loss": 0.4558,
+      "step": 10874
+    },
+    {
+      "epoch": 1.7859914191283641,
+      "grad_norm": 0.28558707454108323,
+      "learning_rate": 6.532817696238766e-06,
+      "loss": 0.4584,
+      "step": 10875
+    },
+    {
+      "epoch": 1.7861556463367068,
+      "grad_norm": 0.35979502883415454,
+      "learning_rate": 6.532399291174521e-06,
+      "loss": 0.4617,
+      "step": 10876
+    },
+    {
+      "epoch": 1.7863198735450496,
+      "grad_norm": 0.28513575240988615,
+      "learning_rate": 6.531980864036419e-06,
+      "loss": 0.4546,
+      "step": 10877
+    },
+    {
+      "epoch": 1.7864841007533925,
+      "grad_norm": 0.3096452283539566,
+      "learning_rate": 6.531562414829007e-06,
+      "loss": 0.4776,
+      "step": 10878
+    },
+    {
+      "epoch": 1.7866483279617351,
+      "grad_norm": 0.3335856723002851,
+      "learning_rate": 6.5311439435568275e-06,
+      "loss": 0.4468,
+      "step": 10879
+    },
+    {
+      "epoch": 1.7868125551700778,
+      "grad_norm": 0.2934169575309989,
+      "learning_rate": 6.530725450224426e-06,
+      "loss": 0.4379,
+      "step": 10880
+    },
+    {
+      "epoch": 1.7869767823784204,
+      "grad_norm": 0.27505220382016676,
+      "learning_rate": 6.530306934836349e-06,
+      "loss": 0.4481,
+      "step": 10881
+    },
+    {
+      "epoch": 1.7871410095867633,
+      "grad_norm": 0.30139640369351783,
+      "learning_rate": 6.52988839739714e-06,
+      "loss": 0.4537,
+      "step": 10882
+    },
+    {
+      "epoch": 1.7873052367951061,
+      "grad_norm": 0.2989131874209402,
+      "learning_rate": 6.529469837911347e-06,
+      "loss": 0.427,
+      "step": 10883
+    },
+    {
+      "epoch": 1.7874694640034487,
+      "grad_norm": 0.5210680923032412,
+      "learning_rate": 6.529051256383515e-06,
+      "loss": 0.4633,
+      "step": 10884
+    },
+    {
+      "epoch": 1.7876336912117914,
+      "grad_norm": 0.29927676663364566,
+      "learning_rate": 6.528632652818189e-06,
+      "loss": 0.482,
+      "step": 10885
+    },
+    {
+      "epoch": 1.7877979184201342,
+      "grad_norm": 0.30191373230381446,
+      "learning_rate": 6.528214027219916e-06,
+      "loss": 0.4607,
+      "step": 10886
+    },
+    {
+      "epoch": 1.787962145628477,
+      "grad_norm": 0.293916130122298,
+      "learning_rate": 6.527795379593244e-06,
+      "loss": 0.4591,
+      "step": 10887
+    },
+    {
+      "epoch": 1.7881263728368197,
+      "grad_norm": 0.4352563222467592,
+      "learning_rate": 6.52737670994272e-06,
+      "loss": 0.4788,
+      "step": 10888
+    },
+    {
+      "epoch": 1.7882906000451624,
+      "grad_norm": 0.9674724434130556,
+      "learning_rate": 6.52695801827289e-06,
+      "loss": 0.4533,
+      "step": 10889
+    },
+    {
+      "epoch": 1.7884548272535052,
+      "grad_norm": 0.2905113824899683,
+      "learning_rate": 6.5265393045882995e-06,
+      "loss": 0.442,
+      "step": 10890
+    },
+    {
+      "epoch": 1.788619054461848,
+      "grad_norm": 0.2708018683006927,
+      "learning_rate": 6.5261205688935e-06,
+      "loss": 0.4559,
+      "step": 10891
+    },
+    {
+      "epoch": 1.7887832816701907,
+      "grad_norm": 0.32087345113081384,
+      "learning_rate": 6.525701811193037e-06,
+      "loss": 0.4751,
+      "step": 10892
+    },
+    {
+      "epoch": 1.7889475088785334,
+      "grad_norm": 0.37785896852038653,
+      "learning_rate": 6.52528303149146e-06,
+      "loss": 0.4552,
+      "step": 10893
+    },
+    {
+      "epoch": 1.7891117360868762,
+      "grad_norm": 0.430886228611665,
+      "learning_rate": 6.524864229793317e-06,
+      "loss": 0.4806,
+      "step": 10894
+    },
+    {
+      "epoch": 1.789275963295219,
+      "grad_norm": 0.290804423309023,
+      "learning_rate": 6.524445406103155e-06,
+      "loss": 0.4715,
+      "step": 10895
+    },
+    {
+      "epoch": 1.7894401905035617,
+      "grad_norm": 0.503444044730968,
+      "learning_rate": 6.524026560425525e-06,
+      "loss": 0.4692,
+      "step": 10896
+    },
+    {
+      "epoch": 1.7896044177119044,
+      "grad_norm": 0.31224124708725315,
+      "learning_rate": 6.523607692764976e-06,
+      "loss": 0.4474,
+      "step": 10897
+    },
+    {
+      "epoch": 1.789768644920247,
+      "grad_norm": 0.2946202941953378,
+      "learning_rate": 6.523188803126056e-06,
+      "loss": 0.4633,
+      "step": 10898
+    },
+    {
+      "epoch": 1.7899328721285899,
+      "grad_norm": 0.2758037376117042,
+      "learning_rate": 6.522769891513314e-06,
+      "loss": 0.4693,
+      "step": 10899
+    },
+    {
+      "epoch": 1.7900970993369327,
+      "grad_norm": 0.32328162302086266,
+      "learning_rate": 6.522350957931301e-06,
+      "loss": 0.4612,
+      "step": 10900
+    },
+    {
+      "epoch": 1.7902613265452754,
+      "grad_norm": 0.3282936126324164,
+      "learning_rate": 6.521932002384568e-06,
+      "loss": 0.4419,
+      "step": 10901
+    },
+    {
+      "epoch": 1.790425553753618,
+      "grad_norm": 0.389470936266045,
+      "learning_rate": 6.5215130248776625e-06,
+      "loss": 0.4575,
+      "step": 10902
+    },
+    {
+      "epoch": 1.7905897809619609,
+      "grad_norm": 0.35674404043827046,
+      "learning_rate": 6.521094025415138e-06,
+      "loss": 0.4636,
+      "step": 10903
+    },
+    {
+      "epoch": 1.7907540081703037,
+      "grad_norm": 0.45021462450787103,
+      "learning_rate": 6.520675004001544e-06,
+      "loss": 0.4805,
+      "step": 10904
+    },
+    {
+      "epoch": 1.7909182353786464,
+      "grad_norm": 0.3029242489250928,
+      "learning_rate": 6.520255960641431e-06,
+      "loss": 0.4538,
+      "step": 10905
+    },
+    {
+      "epoch": 1.791082462586989,
+      "grad_norm": 0.35349088557231206,
+      "learning_rate": 6.5198368953393505e-06,
+      "loss": 0.47,
+      "step": 10906
+    },
+    {
+      "epoch": 1.7912466897953319,
+      "grad_norm": 0.2986364223054352,
+      "learning_rate": 6.519417808099853e-06,
+      "loss": 0.4348,
+      "step": 10907
+    },
+    {
+      "epoch": 1.7914109170036747,
+      "grad_norm": 0.2913259353520882,
+      "learning_rate": 6.5189986989274925e-06,
+      "loss": 0.4525,
+      "step": 10908
+    },
+    {
+      "epoch": 1.7915751442120174,
+      "grad_norm": 0.3301936968659294,
+      "learning_rate": 6.518579567826821e-06,
+      "loss": 0.4729,
+      "step": 10909
+    },
+    {
+      "epoch": 1.79173937142036,
+      "grad_norm": 0.38417340767056996,
+      "learning_rate": 6.518160414802386e-06,
+      "loss": 0.4485,
+      "step": 10910
+    },
+    {
+      "epoch": 1.7919035986287029,
+      "grad_norm": 0.344684065145805,
+      "learning_rate": 6.517741239858746e-06,
+      "loss": 0.4345,
+      "step": 10911
+    },
+    {
+      "epoch": 1.7920678258370457,
+      "grad_norm": 0.2605300283061335,
+      "learning_rate": 6.5173220430004505e-06,
+      "loss": 0.4607,
+      "step": 10912
+    },
+    {
+      "epoch": 1.7922320530453884,
+      "grad_norm": 0.2659233305573,
+      "learning_rate": 6.5169028242320535e-06,
+      "loss": 0.4444,
+      "step": 10913
+    },
+    {
+      "epoch": 1.792396280253731,
+      "grad_norm": 0.4184644221606819,
+      "learning_rate": 6.516483583558105e-06,
+      "loss": 0.4402,
+      "step": 10914
+    },
+    {
+      "epoch": 1.7925605074620736,
+      "grad_norm": 0.3084055212966688,
+      "learning_rate": 6.5160643209831625e-06,
+      "loss": 0.4519,
+      "step": 10915
+    },
+    {
+      "epoch": 1.7927247346704165,
+      "grad_norm": 0.31404721713067635,
+      "learning_rate": 6.5156450365117775e-06,
+      "loss": 0.4551,
+      "step": 10916
+    },
+    {
+      "epoch": 1.7928889618787593,
+      "grad_norm": 0.3102731776923771,
+      "learning_rate": 6.515225730148504e-06,
+      "loss": 0.4377,
+      "step": 10917
+    },
+    {
+      "epoch": 1.793053189087102,
+      "grad_norm": 0.2839548794420706,
+      "learning_rate": 6.514806401897898e-06,
+      "loss": 0.457,
+      "step": 10918
+    },
+    {
+      "epoch": 1.7932174162954446,
+      "grad_norm": 0.3763346060238748,
+      "learning_rate": 6.51438705176451e-06,
+      "loss": 0.4337,
+      "step": 10919
+    },
+    {
+      "epoch": 1.7933816435037875,
+      "grad_norm": 0.3949426628078843,
+      "learning_rate": 6.513967679752898e-06,
+      "loss": 0.4731,
+      "step": 10920
+    },
+    {
+      "epoch": 1.7935458707121303,
+      "grad_norm": 0.3935792482843286,
+      "learning_rate": 6.513548285867615e-06,
+      "loss": 0.4436,
+      "step": 10921
+    },
+    {
+      "epoch": 1.793710097920473,
+      "grad_norm": 0.315815220534042,
+      "learning_rate": 6.513128870113217e-06,
+      "loss": 0.4442,
+      "step": 10922
+    },
+    {
+      "epoch": 1.7938743251288156,
+      "grad_norm": 0.34375092404008345,
+      "learning_rate": 6.51270943249426e-06,
+      "loss": 0.4646,
+      "step": 10923
+    },
+    {
+      "epoch": 1.7940385523371585,
+      "grad_norm": 0.31256531427721074,
+      "learning_rate": 6.512289973015296e-06,
+      "loss": 0.4668,
+      "step": 10924
+    },
+    {
+      "epoch": 1.7942027795455013,
+      "grad_norm": 0.37261118367690316,
+      "learning_rate": 6.511870491680884e-06,
+      "loss": 0.449,
+      "step": 10925
+    },
+    {
+      "epoch": 1.794367006753844,
+      "grad_norm": 0.3237806057691401,
+      "learning_rate": 6.511450988495579e-06,
+      "loss": 0.4706,
+      "step": 10926
+    },
+    {
+      "epoch": 1.7945312339621866,
+      "grad_norm": 0.3194253833424651,
+      "learning_rate": 6.511031463463938e-06,
+      "loss": 0.4614,
+      "step": 10927
+    },
+    {
+      "epoch": 1.7946954611705295,
+      "grad_norm": 0.32585747743809507,
+      "learning_rate": 6.510611916590516e-06,
+      "loss": 0.4661,
+      "step": 10928
+    },
+    {
+      "epoch": 1.7948596883788723,
+      "grad_norm": 0.3586057403191899,
+      "learning_rate": 6.51019234787987e-06,
+      "loss": 0.4613,
+      "step": 10929
+    },
+    {
+      "epoch": 1.795023915587215,
+      "grad_norm": 0.346056732069319,
+      "learning_rate": 6.5097727573365585e-06,
+      "loss": 0.4798,
+      "step": 10930
+    },
+    {
+      "epoch": 1.7951881427955576,
+      "grad_norm": 0.3126585681021275,
+      "learning_rate": 6.509353144965137e-06,
+      "loss": 0.4404,
+      "step": 10931
+    },
+    {
+      "epoch": 1.7953523700039002,
+      "grad_norm": 0.3535289870550478,
+      "learning_rate": 6.508933510770163e-06,
+      "loss": 0.4505,
+      "step": 10932
+    },
+    {
+      "epoch": 1.795516597212243,
+      "grad_norm": 0.3555341152205804,
+      "learning_rate": 6.508513854756194e-06,
+      "loss": 0.4561,
+      "step": 10933
+    },
+    {
+      "epoch": 1.795680824420586,
+      "grad_norm": 0.29507904491048875,
+      "learning_rate": 6.5080941769277895e-06,
+      "loss": 0.4435,
+      "step": 10934
+    },
+    {
+      "epoch": 1.7958450516289286,
+      "grad_norm": 0.5882288729640641,
+      "learning_rate": 6.5076744772895066e-06,
+      "loss": 0.456,
+      "step": 10935
+    },
+    {
+      "epoch": 1.7960092788372712,
+      "grad_norm": 0.45412600871907816,
+      "learning_rate": 6.507254755845903e-06,
+      "loss": 0.4564,
+      "step": 10936
+    },
+    {
+      "epoch": 1.796173506045614,
+      "grad_norm": 0.39575488914845575,
+      "learning_rate": 6.506835012601538e-06,
+      "loss": 0.4669,
+      "step": 10937
+    },
+    {
+      "epoch": 1.796337733253957,
+      "grad_norm": 0.48392008032248435,
+      "learning_rate": 6.50641524756097e-06,
+      "loss": 0.4518,
+      "step": 10938
+    },
+    {
+      "epoch": 1.7965019604622996,
+      "grad_norm": 0.5955498416084375,
+      "learning_rate": 6.505995460728759e-06,
+      "loss": 0.4419,
+      "step": 10939
+    },
+    {
+      "epoch": 1.7966661876706422,
+      "grad_norm": 0.32689911038398983,
+      "learning_rate": 6.505575652109464e-06,
+      "loss": 0.4531,
+      "step": 10940
+    },
+    {
+      "epoch": 1.796830414878985,
+      "grad_norm": 0.34195713441440434,
+      "learning_rate": 6.505155821707642e-06,
+      "loss": 0.4402,
+      "step": 10941
+    },
+    {
+      "epoch": 1.796994642087328,
+      "grad_norm": 0.2566054657671745,
+      "learning_rate": 6.504735969527858e-06,
+      "loss": 0.4404,
+      "step": 10942
+    },
+    {
+      "epoch": 1.7971588692956706,
+      "grad_norm": 0.32830307058291525,
+      "learning_rate": 6.504316095574668e-06,
+      "loss": 0.4713,
+      "step": 10943
+    },
+    {
+      "epoch": 1.7973230965040132,
+      "grad_norm": 0.2556867194390802,
+      "learning_rate": 6.503896199852632e-06,
+      "loss": 0.447,
+      "step": 10944
+    },
+    {
+      "epoch": 1.797487323712356,
+      "grad_norm": 0.33439323742377086,
+      "learning_rate": 6.503476282366313e-06,
+      "loss": 0.4523,
+      "step": 10945
+    },
+    {
+      "epoch": 1.797651550920699,
+      "grad_norm": 0.369790100012626,
+      "learning_rate": 6.50305634312027e-06,
+      "loss": 0.4555,
+      "step": 10946
+    },
+    {
+      "epoch": 1.7978157781290416,
+      "grad_norm": 0.31309996544936336,
+      "learning_rate": 6.502636382119064e-06,
+      "loss": 0.4618,
+      "step": 10947
+    },
+    {
+      "epoch": 1.7979800053373842,
+      "grad_norm": 0.28018315473575545,
+      "learning_rate": 6.5022163993672575e-06,
+      "loss": 0.47,
+      "step": 10948
+    },
+    {
+      "epoch": 1.7981442325457269,
+      "grad_norm": 0.33690209073758887,
+      "learning_rate": 6.5017963948694094e-06,
+      "loss": 0.4501,
+      "step": 10949
+    },
+    {
+      "epoch": 1.7983084597540697,
+      "grad_norm": 0.4252119518199385,
+      "learning_rate": 6.501376368630083e-06,
+      "loss": 0.4494,
+      "step": 10950
+    },
+    {
+      "epoch": 1.7984726869624126,
+      "grad_norm": 0.3752736500074051,
+      "learning_rate": 6.5009563206538426e-06,
+      "loss": 0.4607,
+      "step": 10951
+    },
+    {
+      "epoch": 1.7986369141707552,
+      "grad_norm": 0.35097152026242395,
+      "learning_rate": 6.500536250945247e-06,
+      "loss": 0.4517,
+      "step": 10952
+    },
+    {
+      "epoch": 1.7988011413790979,
+      "grad_norm": 0.3815338566089623,
+      "learning_rate": 6.500116159508858e-06,
+      "loss": 0.4392,
+      "step": 10953
+    },
+    {
+      "epoch": 1.7989653685874407,
+      "grad_norm": 0.47223690095875387,
+      "learning_rate": 6.49969604634924e-06,
+      "loss": 0.4736,
+      "step": 10954
+    },
+    {
+      "epoch": 1.7991295957957836,
+      "grad_norm": 0.35101255145850957,
+      "learning_rate": 6.499275911470957e-06,
+      "loss": 0.4562,
+      "step": 10955
+    },
+    {
+      "epoch": 1.7992938230041262,
+      "grad_norm": 0.6402324900212448,
+      "learning_rate": 6.498855754878569e-06,
+      "loss": 0.4366,
+      "step": 10956
+    },
+    {
+      "epoch": 1.7994580502124689,
+      "grad_norm": 0.3048648173075826,
+      "learning_rate": 6.498435576576641e-06,
+      "loss": 0.4421,
+      "step": 10957
+    },
+    {
+      "epoch": 1.7996222774208117,
+      "grad_norm": 0.34111976311589354,
+      "learning_rate": 6.498015376569737e-06,
+      "loss": 0.4375,
+      "step": 10958
+    },
+    {
+      "epoch": 1.7997865046291546,
+      "grad_norm": 0.30784589349110625,
+      "learning_rate": 6.49759515486242e-06,
+      "loss": 0.4567,
+      "step": 10959
+    },
+    {
+      "epoch": 1.7999507318374972,
+      "grad_norm": 0.31297748416124943,
+      "learning_rate": 6.497174911459255e-06,
+      "loss": 0.4631,
+      "step": 10960
+    },
+    {
+      "epoch": 1.8001149590458398,
+      "grad_norm": 0.3201823680748511,
+      "learning_rate": 6.496754646364805e-06,
+      "loss": 0.4813,
+      "step": 10961
+    },
+    {
+      "epoch": 1.8002791862541827,
+      "grad_norm": 0.2898450135675172,
+      "learning_rate": 6.496334359583635e-06,
+      "loss": 0.4512,
+      "step": 10962
+    },
+    {
+      "epoch": 1.8004434134625256,
+      "grad_norm": 0.3354166298240835,
+      "learning_rate": 6.4959140511203085e-06,
+      "loss": 0.4553,
+      "step": 10963
+    },
+    {
+      "epoch": 1.8006076406708682,
+      "grad_norm": 0.3826613416674712,
+      "learning_rate": 6.495493720979394e-06,
+      "loss": 0.4511,
+      "step": 10964
+    },
+    {
+      "epoch": 1.8007718678792108,
+      "grad_norm": 0.31587237057764445,
+      "learning_rate": 6.495073369165452e-06,
+      "loss": 0.4801,
+      "step": 10965
+    },
+    {
+      "epoch": 1.8009360950875535,
+      "grad_norm": 0.3074217908036942,
+      "learning_rate": 6.494652995683053e-06,
+      "loss": 0.4202,
+      "step": 10966
+    },
+    {
+      "epoch": 1.8011003222958963,
+      "grad_norm": 0.3223590332544111,
+      "learning_rate": 6.494232600536757e-06,
+      "loss": 0.4732,
+      "step": 10967
+    },
+    {
+      "epoch": 1.8012645495042392,
+      "grad_norm": 0.34342528444048104,
+      "learning_rate": 6.493812183731135e-06,
+      "loss": 0.4386,
+      "step": 10968
+    },
+    {
+      "epoch": 1.8014287767125818,
+      "grad_norm": 0.28160135914564627,
+      "learning_rate": 6.49339174527075e-06,
+      "loss": 0.428,
+      "step": 10969
+    },
+    {
+      "epoch": 1.8015930039209245,
+      "grad_norm": 0.32430009941633803,
+      "learning_rate": 6.492971285160169e-06,
+      "loss": 0.4391,
+      "step": 10970
+    },
+    {
+      "epoch": 1.8017572311292673,
+      "grad_norm": 0.364125940598261,
+      "learning_rate": 6.492550803403962e-06,
+      "loss": 0.4454,
+      "step": 10971
+    },
+    {
+      "epoch": 1.8019214583376102,
+      "grad_norm": 0.27861356365010376,
+      "learning_rate": 6.492130300006691e-06,
+      "loss": 0.4506,
+      "step": 10972
+    },
+    {
+      "epoch": 1.8020856855459528,
+      "grad_norm": 0.29793446395521767,
+      "learning_rate": 6.491709774972923e-06,
+      "loss": 0.4528,
+      "step": 10973
+    },
+    {
+      "epoch": 1.8022499127542955,
+      "grad_norm": 0.30617337367580927,
+      "learning_rate": 6.491289228307229e-06,
+      "loss": 0.4604,
+      "step": 10974
+    },
+    {
+      "epoch": 1.8024141399626383,
+      "grad_norm": 0.35553986727454073,
+      "learning_rate": 6.490868660014175e-06,
+      "loss": 0.4618,
+      "step": 10975
+    },
+    {
+      "epoch": 1.8025783671709812,
+      "grad_norm": 0.43792944372602643,
+      "learning_rate": 6.49044807009833e-06,
+      "loss": 0.444,
+      "step": 10976
+    },
+    {
+      "epoch": 1.8027425943793238,
+      "grad_norm": 0.309443097946476,
+      "learning_rate": 6.490027458564258e-06,
+      "loss": 0.4633,
+      "step": 10977
+    },
+    {
+      "epoch": 1.8029068215876665,
+      "grad_norm": 0.29035938841687275,
+      "learning_rate": 6.489606825416531e-06,
+      "loss": 0.4657,
+      "step": 10978
+    },
+    {
+      "epoch": 1.8030710487960093,
+      "grad_norm": 0.38280437482755186,
+      "learning_rate": 6.489186170659715e-06,
+      "loss": 0.4412,
+      "step": 10979
+    },
+    {
+      "epoch": 1.8032352760043522,
+      "grad_norm": 0.3268712043393675,
+      "learning_rate": 6.488765494298382e-06,
+      "loss": 0.461,
+      "step": 10980
+    },
+    {
+      "epoch": 1.8033995032126948,
+      "grad_norm": 0.5975993158922271,
+      "learning_rate": 6.488344796337099e-06,
+      "loss": 0.4575,
+      "step": 10981
+    },
+    {
+      "epoch": 1.8035637304210375,
+      "grad_norm": 0.28730055670229004,
+      "learning_rate": 6.487924076780434e-06,
+      "loss": 0.4605,
+      "step": 10982
+    },
+    {
+      "epoch": 1.80372795762938,
+      "grad_norm": 0.3067576354459948,
+      "learning_rate": 6.487503335632958e-06,
+      "loss": 0.4579,
+      "step": 10983
+    },
+    {
+      "epoch": 1.803892184837723,
+      "grad_norm": 0.3178578324928383,
+      "learning_rate": 6.48708257289924e-06,
+      "loss": 0.4553,
+      "step": 10984
+    },
+    {
+      "epoch": 1.8040564120460658,
+      "grad_norm": 0.31296466092563197,
+      "learning_rate": 6.486661788583851e-06,
+      "loss": 0.4701,
+      "step": 10985
+    },
+    {
+      "epoch": 1.8042206392544085,
+      "grad_norm": 0.27868046588972933,
+      "learning_rate": 6.4862409826913615e-06,
+      "loss": 0.4234,
+      "step": 10986
+    },
+    {
+      "epoch": 1.804384866462751,
+      "grad_norm": 0.2894081748376271,
+      "learning_rate": 6.485820155226339e-06,
+      "loss": 0.4575,
+      "step": 10987
+    },
+    {
+      "epoch": 1.804549093671094,
+      "grad_norm": 0.3840278851261531,
+      "learning_rate": 6.485399306193356e-06,
+      "loss": 0.4688,
+      "step": 10988
+    },
+    {
+      "epoch": 1.8047133208794368,
+      "grad_norm": 0.35329064565706103,
+      "learning_rate": 6.484978435596983e-06,
+      "loss": 0.4492,
+      "step": 10989
+    },
+    {
+      "epoch": 1.8048775480877794,
+      "grad_norm": 0.2666745016209278,
+      "learning_rate": 6.484557543441792e-06,
+      "loss": 0.451,
+      "step": 10990
+    },
+    {
+      "epoch": 1.805041775296122,
+      "grad_norm": 0.2727701044292131,
+      "learning_rate": 6.484136629732354e-06,
+      "loss": 0.4699,
+      "step": 10991
+    },
+    {
+      "epoch": 1.805206002504465,
+      "grad_norm": 0.2876975201051365,
+      "learning_rate": 6.483715694473239e-06,
+      "loss": 0.452,
+      "step": 10992
+    },
+    {
+      "epoch": 1.8053702297128078,
+      "grad_norm": 0.30486164251675607,
+      "learning_rate": 6.483294737669021e-06,
+      "loss": 0.4638,
+      "step": 10993
+    },
+    {
+      "epoch": 1.8055344569211504,
+      "grad_norm": 0.3242566132554888,
+      "learning_rate": 6.482873759324268e-06,
+      "loss": 0.4656,
+      "step": 10994
+    },
+    {
+      "epoch": 1.805698684129493,
+      "grad_norm": 0.27753825968593376,
+      "learning_rate": 6.4824527594435586e-06,
+      "loss": 0.4425,
+      "step": 10995
+    },
+    {
+      "epoch": 1.805862911337836,
+      "grad_norm": 0.32392349033908413,
+      "learning_rate": 6.48203173803146e-06,
+      "loss": 0.4592,
+      "step": 10996
+    },
+    {
+      "epoch": 1.8060271385461788,
+      "grad_norm": 0.2700411488707498,
+      "learning_rate": 6.481610695092547e-06,
+      "loss": 0.4622,
+      "step": 10997
+    },
+    {
+      "epoch": 1.8061913657545214,
+      "grad_norm": 0.3150824644921173,
+      "learning_rate": 6.481189630631392e-06,
+      "loss": 0.4466,
+      "step": 10998
+    },
+    {
+      "epoch": 1.806355592962864,
+      "grad_norm": 0.4169412074763378,
+      "learning_rate": 6.480768544652569e-06,
+      "loss": 0.4512,
+      "step": 10999
+    },
+    {
+      "epoch": 1.8065198201712067,
+      "grad_norm": 0.33008365407804763,
+      "learning_rate": 6.48034743716065e-06,
+      "loss": 0.4525,
+      "step": 11000
+    },
+    {
+      "epoch": 1.8066840473795496,
+      "grad_norm": 0.3184161468360843,
+      "learning_rate": 6.479926308160211e-06,
+      "loss": 0.4638,
+      "step": 11001
+    },
+    {
+      "epoch": 1.8068482745878924,
+      "grad_norm": 0.2894848075749476,
+      "learning_rate": 6.479505157655822e-06,
+      "loss": 0.443,
+      "step": 11002
+    },
+    {
+      "epoch": 1.807012501796235,
+      "grad_norm": 0.319981318393346,
+      "learning_rate": 6.4790839856520605e-06,
+      "loss": 0.4473,
+      "step": 11003
+    },
+    {
+      "epoch": 1.8071767290045777,
+      "grad_norm": 0.3540499062083888,
+      "learning_rate": 6.4786627921534985e-06,
+      "loss": 0.4304,
+      "step": 11004
+    },
+    {
+      "epoch": 1.8073409562129206,
+      "grad_norm": 0.294134878699911,
+      "learning_rate": 6.4782415771647145e-06,
+      "loss": 0.4544,
+      "step": 11005
+    },
+    {
+      "epoch": 1.8075051834212634,
+      "grad_norm": 0.4010107953685391,
+      "learning_rate": 6.47782034069028e-06,
+      "loss": 0.4226,
+      "step": 11006
+    },
+    {
+      "epoch": 1.807669410629606,
+      "grad_norm": 0.3958453380626719,
+      "learning_rate": 6.477399082734769e-06,
+      "loss": 0.4563,
+      "step": 11007
+    },
+    {
+      "epoch": 1.8078336378379487,
+      "grad_norm": 0.49181283823773797,
+      "learning_rate": 6.476977803302758e-06,
+      "loss": 0.4549,
+      "step": 11008
+    },
+    {
+      "epoch": 1.8079978650462916,
+      "grad_norm": 0.32707082518126557,
+      "learning_rate": 6.476556502398825e-06,
+      "loss": 0.437,
+      "step": 11009
+    },
+    {
+      "epoch": 1.8081620922546344,
+      "grad_norm": 0.32247170989748514,
+      "learning_rate": 6.476135180027544e-06,
+      "loss": 0.4545,
+      "step": 11010
+    },
+    {
+      "epoch": 1.808326319462977,
+      "grad_norm": 0.2571814591641584,
+      "learning_rate": 6.47571383619349e-06,
+      "loss": 0.4704,
+      "step": 11011
+    },
+    {
+      "epoch": 1.8084905466713197,
+      "grad_norm": 0.27433283480124443,
+      "learning_rate": 6.4752924709012385e-06,
+      "loss": 0.4397,
+      "step": 11012
+    },
+    {
+      "epoch": 1.8086547738796626,
+      "grad_norm": 0.30296470989492685,
+      "learning_rate": 6.474871084155368e-06,
+      "loss": 0.4827,
+      "step": 11013
+    },
+    {
+      "epoch": 1.8088190010880054,
+      "grad_norm": 0.329768978817267,
+      "learning_rate": 6.474449675960455e-06,
+      "loss": 0.4383,
+      "step": 11014
+    },
+    {
+      "epoch": 1.808983228296348,
+      "grad_norm": 0.3012850081596831,
+      "learning_rate": 6.474028246321077e-06,
+      "loss": 0.4354,
+      "step": 11015
+    },
+    {
+      "epoch": 1.8091474555046907,
+      "grad_norm": 0.3940716281617523,
+      "learning_rate": 6.47360679524181e-06,
+      "loss": 0.4473,
+      "step": 11016
+    },
+    {
+      "epoch": 1.8093116827130333,
+      "grad_norm": 0.29057796041348144,
+      "learning_rate": 6.473185322727228e-06,
+      "loss": 0.456,
+      "step": 11017
+    },
+    {
+      "epoch": 1.8094759099213762,
+      "grad_norm": 0.3647116420745295,
+      "learning_rate": 6.472763828781916e-06,
+      "loss": 0.4429,
+      "step": 11018
+    },
+    {
+      "epoch": 1.809640137129719,
+      "grad_norm": 0.4227525053223582,
+      "learning_rate": 6.472342313410446e-06,
+      "loss": 0.4582,
+      "step": 11019
+    },
+    {
+      "epoch": 1.8098043643380617,
+      "grad_norm": 0.3717185423227278,
+      "learning_rate": 6.471920776617399e-06,
+      "loss": 0.4441,
+      "step": 11020
+    },
+    {
+      "epoch": 1.8099685915464043,
+      "grad_norm": 0.27670914914335104,
+      "learning_rate": 6.471499218407351e-06,
+      "loss": 0.4658,
+      "step": 11021
+    },
+    {
+      "epoch": 1.8101328187547472,
+      "grad_norm": 0.34867878518606726,
+      "learning_rate": 6.471077638784882e-06,
+      "loss": 0.4565,
+      "step": 11022
+    },
+    {
+      "epoch": 1.81029704596309,
+      "grad_norm": 0.27587572155017304,
+      "learning_rate": 6.470656037754571e-06,
+      "loss": 0.4582,
+      "step": 11023
+    },
+    {
+      "epoch": 1.8104612731714327,
+      "grad_norm": 0.7304768832794305,
+      "learning_rate": 6.470234415320997e-06,
+      "loss": 0.4771,
+      "step": 11024
+    },
+    {
+      "epoch": 1.8106255003797753,
+      "grad_norm": 0.3823763993220649,
+      "learning_rate": 6.469812771488737e-06,
+      "loss": 0.4675,
+      "step": 11025
+    },
+    {
+      "epoch": 1.8107897275881182,
+      "grad_norm": 0.333588611062913,
+      "learning_rate": 6.469391106262375e-06,
+      "loss": 0.4481,
+      "step": 11026
+    },
+    {
+      "epoch": 1.810953954796461,
+      "grad_norm": 0.3375432724883843,
+      "learning_rate": 6.468969419646486e-06,
+      "loss": 0.4603,
+      "step": 11027
+    },
+    {
+      "epoch": 1.8111181820048037,
+      "grad_norm": 0.3195629064261066,
+      "learning_rate": 6.468547711645652e-06,
+      "loss": 0.4542,
+      "step": 11028
+    },
+    {
+      "epoch": 1.8112824092131463,
+      "grad_norm": 0.28394558610822923,
+      "learning_rate": 6.468125982264454e-06,
+      "loss": 0.4513,
+      "step": 11029
+    },
+    {
+      "epoch": 1.8114466364214892,
+      "grad_norm": 0.33674311931385514,
+      "learning_rate": 6.4677042315074715e-06,
+      "loss": 0.4662,
+      "step": 11030
+    },
+    {
+      "epoch": 1.811610863629832,
+      "grad_norm": 0.38153871812642226,
+      "learning_rate": 6.4672824593792835e-06,
+      "loss": 0.4618,
+      "step": 11031
+    },
+    {
+      "epoch": 1.8117750908381747,
+      "grad_norm": 0.30354053866531616,
+      "learning_rate": 6.466860665884473e-06,
+      "loss": 0.4538,
+      "step": 11032
+    },
+    {
+      "epoch": 1.8119393180465173,
+      "grad_norm": 0.3405080385317878,
+      "learning_rate": 6.466438851027622e-06,
+      "loss": 0.4433,
+      "step": 11033
+    },
+    {
+      "epoch": 1.81210354525486,
+      "grad_norm": 0.29623350679986754,
+      "learning_rate": 6.46601701481331e-06,
+      "loss": 0.4314,
+      "step": 11034
+    },
+    {
+      "epoch": 1.8122677724632028,
+      "grad_norm": 0.2935779674611224,
+      "learning_rate": 6.465595157246118e-06,
+      "loss": 0.4527,
+      "step": 11035
+    },
+    {
+      "epoch": 1.8124319996715457,
+      "grad_norm": 0.3068991690315416,
+      "learning_rate": 6.4651732783306285e-06,
+      "loss": 0.4497,
+      "step": 11036
+    },
+    {
+      "epoch": 1.8125962268798883,
+      "grad_norm": 0.40453061082958175,
+      "learning_rate": 6.464751378071424e-06,
+      "loss": 0.4602,
+      "step": 11037
+    },
+    {
+      "epoch": 1.812760454088231,
+      "grad_norm": 0.31945911053068904,
+      "learning_rate": 6.464329456473086e-06,
+      "loss": 0.464,
+      "step": 11038
+    },
+    {
+      "epoch": 1.8129246812965738,
+      "grad_norm": 0.311919813694301,
+      "learning_rate": 6.4639075135402e-06,
+      "loss": 0.4706,
+      "step": 11039
+    },
+    {
+      "epoch": 1.8130889085049167,
+      "grad_norm": 0.37696971232973997,
+      "learning_rate": 6.463485549277343e-06,
+      "loss": 0.4397,
+      "step": 11040
+    },
+    {
+      "epoch": 1.8132531357132593,
+      "grad_norm": 0.3312339585677867,
+      "learning_rate": 6.463063563689103e-06,
+      "loss": 0.4468,
+      "step": 11041
+    },
+    {
+      "epoch": 1.813417362921602,
+      "grad_norm": 1.0677719619640302,
+      "learning_rate": 6.46264155678006e-06,
+      "loss": 0.4389,
+      "step": 11042
+    },
+    {
+      "epoch": 1.8135815901299448,
+      "grad_norm": 0.3297430212323639,
+      "learning_rate": 6.4622195285548e-06,
+      "loss": 0.4531,
+      "step": 11043
+    },
+    {
+      "epoch": 1.8137458173382877,
+      "grad_norm": 0.2958232682654792,
+      "learning_rate": 6.461797479017906e-06,
+      "loss": 0.4471,
+      "step": 11044
+    },
+    {
+      "epoch": 1.8139100445466303,
+      "grad_norm": 0.327250642966852,
+      "learning_rate": 6.46137540817396e-06,
+      "loss": 0.4401,
+      "step": 11045
+    },
+    {
+      "epoch": 1.814074271754973,
+      "grad_norm": 0.4175480266474759,
+      "learning_rate": 6.4609533160275465e-06,
+      "loss": 0.4378,
+      "step": 11046
+    },
+    {
+      "epoch": 1.8142384989633158,
+      "grad_norm": 0.38417408251346336,
+      "learning_rate": 6.460531202583252e-06,
+      "loss": 0.4641,
+      "step": 11047
+    },
+    {
+      "epoch": 1.8144027261716587,
+      "grad_norm": 1.088990431170157,
+      "learning_rate": 6.460109067845658e-06,
+      "loss": 0.4307,
+      "step": 11048
+    },
+    {
+      "epoch": 1.8145669533800013,
+      "grad_norm": 0.4952726764230832,
+      "learning_rate": 6.459686911819353e-06,
+      "loss": 0.4457,
+      "step": 11049
+    },
+    {
+      "epoch": 1.814731180588344,
+      "grad_norm": 0.31313614766898384,
+      "learning_rate": 6.45926473450892e-06,
+      "loss": 0.4671,
+      "step": 11050
+    },
+    {
+      "epoch": 1.8148954077966866,
+      "grad_norm": 0.3303499375852876,
+      "learning_rate": 6.458842535918944e-06,
+      "loss": 0.4383,
+      "step": 11051
+    },
+    {
+      "epoch": 1.8150596350050294,
+      "grad_norm": 0.29281553354836926,
+      "learning_rate": 6.4584203160540105e-06,
+      "loss": 0.4327,
+      "step": 11052
+    },
+    {
+      "epoch": 1.8152238622133723,
+      "grad_norm": 0.38270445224557276,
+      "learning_rate": 6.457998074918705e-06,
+      "loss": 0.4503,
+      "step": 11053
+    },
+    {
+      "epoch": 1.815388089421715,
+      "grad_norm": 0.4228231426560049,
+      "learning_rate": 6.457575812517615e-06,
+      "loss": 0.4482,
+      "step": 11054
+    },
+    {
+      "epoch": 1.8155523166300576,
+      "grad_norm": 0.34167907967027655,
+      "learning_rate": 6.457153528855325e-06,
+      "loss": 0.4512,
+      "step": 11055
+    },
+    {
+      "epoch": 1.8157165438384004,
+      "grad_norm": 0.41697598086361165,
+      "learning_rate": 6.456731223936423e-06,
+      "loss": 0.4805,
+      "step": 11056
+    },
+    {
+      "epoch": 1.8158807710467433,
+      "grad_norm": 0.3569020023794603,
+      "learning_rate": 6.456308897765494e-06,
+      "loss": 0.4532,
+      "step": 11057
+    },
+    {
+      "epoch": 1.816044998255086,
+      "grad_norm": 0.3618170825730565,
+      "learning_rate": 6.455886550347127e-06,
+      "loss": 0.4494,
+      "step": 11058
+    },
+    {
+      "epoch": 1.8162092254634286,
+      "grad_norm": 0.3233994158463805,
+      "learning_rate": 6.455464181685904e-06,
+      "loss": 0.4636,
+      "step": 11059
+    },
+    {
+      "epoch": 1.8163734526717714,
+      "grad_norm": 0.2949245076583048,
+      "learning_rate": 6.45504179178642e-06,
+      "loss": 0.4542,
+      "step": 11060
+    },
+    {
+      "epoch": 1.8165376798801143,
+      "grad_norm": 0.3139588667284323,
+      "learning_rate": 6.454619380653257e-06,
+      "loss": 0.4656,
+      "step": 11061
+    },
+    {
+      "epoch": 1.816701907088457,
+      "grad_norm": 0.29795405508359124,
+      "learning_rate": 6.4541969482910044e-06,
+      "loss": 0.4467,
+      "step": 11062
+    },
+    {
+      "epoch": 1.8168661342967996,
+      "grad_norm": 0.3168979315435727,
+      "learning_rate": 6.453774494704251e-06,
+      "loss": 0.4384,
+      "step": 11063
+    },
+    {
+      "epoch": 1.8170303615051424,
+      "grad_norm": 0.32749872421298704,
+      "learning_rate": 6.453352019897584e-06,
+      "loss": 0.4554,
+      "step": 11064
+    },
+    {
+      "epoch": 1.8171945887134853,
+      "grad_norm": 0.42556222068988175,
+      "learning_rate": 6.452929523875592e-06,
+      "loss": 0.4552,
+      "step": 11065
+    },
+    {
+      "epoch": 1.817358815921828,
+      "grad_norm": 0.44368552613487633,
+      "learning_rate": 6.452507006642863e-06,
+      "loss": 0.4534,
+      "step": 11066
+    },
+    {
+      "epoch": 1.8175230431301705,
+      "grad_norm": 0.3223174577268706,
+      "learning_rate": 6.452084468203988e-06,
+      "loss": 0.4479,
+      "step": 11067
+    },
+    {
+      "epoch": 1.8176872703385132,
+      "grad_norm": 0.33800164745732303,
+      "learning_rate": 6.4516619085635555e-06,
+      "loss": 0.4734,
+      "step": 11068
+    },
+    {
+      "epoch": 1.817851497546856,
+      "grad_norm": 0.3452294395819774,
+      "learning_rate": 6.451239327726155e-06,
+      "loss": 0.4517,
+      "step": 11069
+    },
+    {
+      "epoch": 1.818015724755199,
+      "grad_norm": 0.4589685077156546,
+      "learning_rate": 6.4508167256963735e-06,
+      "loss": 0.4278,
+      "step": 11070
+    },
+    {
+      "epoch": 1.8181799519635415,
+      "grad_norm": 0.4040673809335704,
+      "learning_rate": 6.450394102478804e-06,
+      "loss": 0.4348,
+      "step": 11071
+    },
+    {
+      "epoch": 1.8183441791718842,
+      "grad_norm": 0.3518608635195822,
+      "learning_rate": 6.449971458078036e-06,
+      "loss": 0.4628,
+      "step": 11072
+    },
+    {
+      "epoch": 1.818508406380227,
+      "grad_norm": 0.40018200265748566,
+      "learning_rate": 6.44954879249866e-06,
+      "loss": 0.432,
+      "step": 11073
+    },
+    {
+      "epoch": 1.81867263358857,
+      "grad_norm": 0.3737532515019579,
+      "learning_rate": 6.4491261057452644e-06,
+      "loss": 0.4505,
+      "step": 11074
+    },
+    {
+      "epoch": 1.8188368607969125,
+      "grad_norm": 0.31055544483008213,
+      "learning_rate": 6.448703397822442e-06,
+      "loss": 0.4481,
+      "step": 11075
+    },
+    {
+      "epoch": 1.8190010880052552,
+      "grad_norm": 0.3225928339655794,
+      "learning_rate": 6.448280668734785e-06,
+      "loss": 0.463,
+      "step": 11076
+    },
+    {
+      "epoch": 1.819165315213598,
+      "grad_norm": 0.3259872807174308,
+      "learning_rate": 6.447857918486881e-06,
+      "loss": 0.4619,
+      "step": 11077
+    },
+    {
+      "epoch": 1.819329542421941,
+      "grad_norm": 0.3702360876886552,
+      "learning_rate": 6.447435147083326e-06,
+      "loss": 0.4555,
+      "step": 11078
+    },
+    {
+      "epoch": 1.8194937696302835,
+      "grad_norm": 0.3691408949553295,
+      "learning_rate": 6.447012354528708e-06,
+      "loss": 0.4432,
+      "step": 11079
+    },
+    {
+      "epoch": 1.8196579968386262,
+      "grad_norm": 0.3041321344097473,
+      "learning_rate": 6.446589540827619e-06,
+      "loss": 0.4201,
+      "step": 11080
+    },
+    {
+      "epoch": 1.819822224046969,
+      "grad_norm": 0.30894935051624056,
+      "learning_rate": 6.446166705984654e-06,
+      "loss": 0.4367,
+      "step": 11081
+    },
+    {
+      "epoch": 1.8199864512553119,
+      "grad_norm": 0.3979234912372484,
+      "learning_rate": 6.4457438500044025e-06,
+      "loss": 0.4442,
+      "step": 11082
+    },
+    {
+      "epoch": 1.8201506784636545,
+      "grad_norm": 0.35321748362975847,
+      "learning_rate": 6.44532097289146e-06,
+      "loss": 0.4585,
+      "step": 11083
+    },
+    {
+      "epoch": 1.8203149056719972,
+      "grad_norm": 0.32382149930434995,
+      "learning_rate": 6.444898074650416e-06,
+      "loss": 0.4504,
+      "step": 11084
+    },
+    {
+      "epoch": 1.8204791328803398,
+      "grad_norm": 0.3905287570407914,
+      "learning_rate": 6.444475155285867e-06,
+      "loss": 0.4671,
+      "step": 11085
+    },
+    {
+      "epoch": 1.8206433600886827,
+      "grad_norm": 0.3364174209242473,
+      "learning_rate": 6.444052214802404e-06,
+      "loss": 0.4514,
+      "step": 11086
+    },
+    {
+      "epoch": 1.8208075872970255,
+      "grad_norm": 0.3132902251492512,
+      "learning_rate": 6.443629253204621e-06,
+      "loss": 0.4593,
+      "step": 11087
+    },
+    {
+      "epoch": 1.8209718145053682,
+      "grad_norm": 0.31611007055962465,
+      "learning_rate": 6.443206270497113e-06,
+      "loss": 0.4289,
+      "step": 11088
+    },
+    {
+      "epoch": 1.8211360417137108,
+      "grad_norm": 0.3183713282494217,
+      "learning_rate": 6.4427832666844725e-06,
+      "loss": 0.4557,
+      "step": 11089
+    },
+    {
+      "epoch": 1.8213002689220537,
+      "grad_norm": 0.3550373150593315,
+      "learning_rate": 6.442360241771294e-06,
+      "loss": 0.4428,
+      "step": 11090
+    },
+    {
+      "epoch": 1.8214644961303965,
+      "grad_norm": 0.32252481795074844,
+      "learning_rate": 6.4419371957621726e-06,
+      "loss": 0.4485,
+      "step": 11091
+    },
+    {
+      "epoch": 1.8216287233387392,
+      "grad_norm": 0.31700873209102415,
+      "learning_rate": 6.441514128661702e-06,
+      "loss": 0.4577,
+      "step": 11092
+    },
+    {
+      "epoch": 1.8217929505470818,
+      "grad_norm": 0.3785064320067957,
+      "learning_rate": 6.44109104047448e-06,
+      "loss": 0.4385,
+      "step": 11093
+    },
+    {
+      "epoch": 1.8219571777554247,
+      "grad_norm": 0.29377408501971125,
+      "learning_rate": 6.440667931205097e-06,
+      "loss": 0.4443,
+      "step": 11094
+    },
+    {
+      "epoch": 1.8221214049637675,
+      "grad_norm": 0.30032625362802456,
+      "learning_rate": 6.440244800858152e-06,
+      "loss": 0.4331,
+      "step": 11095
+    },
+    {
+      "epoch": 1.8222856321721101,
+      "grad_norm": 0.3718303109112087,
+      "learning_rate": 6.4398216494382386e-06,
+      "loss": 0.4572,
+      "step": 11096
+    },
+    {
+      "epoch": 1.8224498593804528,
+      "grad_norm": 0.2960251316220687,
+      "learning_rate": 6.439398476949954e-06,
+      "loss": 0.4621,
+      "step": 11097
+    },
+    {
+      "epoch": 1.8226140865887956,
+      "grad_norm": 0.33123020320172913,
+      "learning_rate": 6.438975283397895e-06,
+      "loss": 0.4432,
+      "step": 11098
+    },
+    {
+      "epoch": 1.8227783137971385,
+      "grad_norm": 0.2742503490676697,
+      "learning_rate": 6.4385520687866554e-06,
+      "loss": 0.4535,
+      "step": 11099
+    },
+    {
+      "epoch": 1.8229425410054811,
+      "grad_norm": 0.2836794155117891,
+      "learning_rate": 6.438128833120833e-06,
+      "loss": 0.4517,
+      "step": 11100
+    },
+    {
+      "epoch": 1.8231067682138238,
+      "grad_norm": 0.30579928885691016,
+      "learning_rate": 6.437705576405025e-06,
+      "loss": 0.4649,
+      "step": 11101
+    },
+    {
+      "epoch": 1.8232709954221664,
+      "grad_norm": 0.3615411845518621,
+      "learning_rate": 6.437282298643828e-06,
+      "loss": 0.4617,
+      "step": 11102
+    },
+    {
+      "epoch": 1.8234352226305093,
+      "grad_norm": 0.30059915648933905,
+      "learning_rate": 6.43685899984184e-06,
+      "loss": 0.4397,
+      "step": 11103
+    },
+    {
+      "epoch": 1.8235994498388521,
+      "grad_norm": 0.9129523653870982,
+      "learning_rate": 6.4364356800036555e-06,
+      "loss": 0.4795,
+      "step": 11104
+    },
+    {
+      "epoch": 1.8237636770471948,
+      "grad_norm": 0.4191096623708467,
+      "learning_rate": 6.436012339133876e-06,
+      "loss": 0.4281,
+      "step": 11105
+    },
+    {
+      "epoch": 1.8239279042555374,
+      "grad_norm": 0.3097106124164847,
+      "learning_rate": 6.435588977237098e-06,
+      "loss": 0.4599,
+      "step": 11106
+    },
+    {
+      "epoch": 1.8240921314638803,
+      "grad_norm": 0.25366945090881754,
+      "learning_rate": 6.435165594317919e-06,
+      "loss": 0.452,
+      "step": 11107
+    },
+    {
+      "epoch": 1.8242563586722231,
+      "grad_norm": 0.42724123488796834,
+      "learning_rate": 6.434742190380938e-06,
+      "loss": 0.4489,
+      "step": 11108
+    },
+    {
+      "epoch": 1.8244205858805658,
+      "grad_norm": 0.4470158326079735,
+      "learning_rate": 6.4343187654307516e-06,
+      "loss": 0.4586,
+      "step": 11109
+    },
+    {
+      "epoch": 1.8245848130889084,
+      "grad_norm": 0.2843995188282995,
+      "learning_rate": 6.4338953194719625e-06,
+      "loss": 0.4499,
+      "step": 11110
+    },
+    {
+      "epoch": 1.8247490402972513,
+      "grad_norm": 0.3535728345730193,
+      "learning_rate": 6.433471852509166e-06,
+      "loss": 0.4527,
+      "step": 11111
+    },
+    {
+      "epoch": 1.8249132675055941,
+      "grad_norm": 0.390488775163116,
+      "learning_rate": 6.433048364546963e-06,
+      "loss": 0.4499,
+      "step": 11112
+    },
+    {
+      "epoch": 1.8250774947139368,
+      "grad_norm": 0.2856644175201384,
+      "learning_rate": 6.4326248555899535e-06,
+      "loss": 0.4562,
+      "step": 11113
+    },
+    {
+      "epoch": 1.8252417219222794,
+      "grad_norm": 0.25416212783117076,
+      "learning_rate": 6.432201325642737e-06,
+      "loss": 0.4399,
+      "step": 11114
+    },
+    {
+      "epoch": 1.8254059491306223,
+      "grad_norm": 0.41194388605844706,
+      "learning_rate": 6.431777774709912e-06,
+      "loss": 0.4593,
+      "step": 11115
+    },
+    {
+      "epoch": 1.8255701763389651,
+      "grad_norm": 0.3144676904089028,
+      "learning_rate": 6.43135420279608e-06,
+      "loss": 0.443,
+      "step": 11116
+    },
+    {
+      "epoch": 1.8257344035473078,
+      "grad_norm": 0.393080464105788,
+      "learning_rate": 6.430930609905842e-06,
+      "loss": 0.4569,
+      "step": 11117
+    },
+    {
+      "epoch": 1.8258986307556504,
+      "grad_norm": 0.33847400830923585,
+      "learning_rate": 6.430506996043798e-06,
+      "loss": 0.4589,
+      "step": 11118
+    },
+    {
+      "epoch": 1.826062857963993,
+      "grad_norm": 0.36415414734893137,
+      "learning_rate": 6.430083361214547e-06,
+      "loss": 0.468,
+      "step": 11119
+    },
+    {
+      "epoch": 1.826227085172336,
+      "grad_norm": 0.37557718876286633,
+      "learning_rate": 6.429659705422693e-06,
+      "loss": 0.4628,
+      "step": 11120
+    },
+    {
+      "epoch": 1.8263913123806788,
+      "grad_norm": 0.32069328835487537,
+      "learning_rate": 6.429236028672834e-06,
+      "loss": 0.4346,
+      "step": 11121
+    },
+    {
+      "epoch": 1.8265555395890214,
+      "grad_norm": 0.31544843970795344,
+      "learning_rate": 6.428812330969576e-06,
+      "loss": 0.4485,
+      "step": 11122
+    },
+    {
+      "epoch": 1.826719766797364,
+      "grad_norm": 0.29920140008668483,
+      "learning_rate": 6.428388612317519e-06,
+      "loss": 0.4633,
+      "step": 11123
+    },
+    {
+      "epoch": 1.826883994005707,
+      "grad_norm": 0.26314160528688985,
+      "learning_rate": 6.427964872721262e-06,
+      "loss": 0.4446,
+      "step": 11124
+    },
+    {
+      "epoch": 1.8270482212140498,
+      "grad_norm": 0.3417200243928961,
+      "learning_rate": 6.4275411121854095e-06,
+      "loss": 0.447,
+      "step": 11125
+    },
+    {
+      "epoch": 1.8272124484223924,
+      "grad_norm": 0.4288052637609021,
+      "learning_rate": 6.427117330714566e-06,
+      "loss": 0.4726,
+      "step": 11126
+    },
+    {
+      "epoch": 1.827376675630735,
+      "grad_norm": 0.3291954670980573,
+      "learning_rate": 6.426693528313333e-06,
+      "loss": 0.4301,
+      "step": 11127
+    },
+    {
+      "epoch": 1.8275409028390779,
+      "grad_norm": 0.3432703402034783,
+      "learning_rate": 6.4262697049863106e-06,
+      "loss": 0.4617,
+      "step": 11128
+    },
+    {
+      "epoch": 1.8277051300474207,
+      "grad_norm": 0.31460182638244183,
+      "learning_rate": 6.425845860738104e-06,
+      "loss": 0.4361,
+      "step": 11129
+    },
+    {
+      "epoch": 1.8278693572557634,
+      "grad_norm": 0.3356042150563821,
+      "learning_rate": 6.4254219955733166e-06,
+      "loss": 0.4502,
+      "step": 11130
+    },
+    {
+      "epoch": 1.828033584464106,
+      "grad_norm": 0.3417134062143049,
+      "learning_rate": 6.424998109496554e-06,
+      "loss": 0.4591,
+      "step": 11131
+    },
+    {
+      "epoch": 1.8281978116724489,
+      "grad_norm": 0.29443259361596996,
+      "learning_rate": 6.4245742025124165e-06,
+      "loss": 0.4615,
+      "step": 11132
+    },
+    {
+      "epoch": 1.8283620388807917,
+      "grad_norm": 0.3027837712996636,
+      "learning_rate": 6.424150274625509e-06,
+      "loss": 0.4432,
+      "step": 11133
+    },
+    {
+      "epoch": 1.8285262660891344,
+      "grad_norm": 0.3859255584228414,
+      "learning_rate": 6.423726325840437e-06,
+      "loss": 0.4525,
+      "step": 11134
+    },
+    {
+      "epoch": 1.828690493297477,
+      "grad_norm": 0.8084532470537148,
+      "learning_rate": 6.423302356161805e-06,
+      "loss": 0.4585,
+      "step": 11135
+    },
+    {
+      "epoch": 1.8288547205058197,
+      "grad_norm": 0.3222083942212579,
+      "learning_rate": 6.422878365594217e-06,
+      "loss": 0.4348,
+      "step": 11136
+    },
+    {
+      "epoch": 1.8290189477141625,
+      "grad_norm": 0.5974350173854517,
+      "learning_rate": 6.422454354142277e-06,
+      "loss": 0.445,
+      "step": 11137
+    },
+    {
+      "epoch": 1.8291831749225054,
+      "grad_norm": 0.48381322250594383,
+      "learning_rate": 6.422030321810592e-06,
+      "loss": 0.4752,
+      "step": 11138
+    },
+    {
+      "epoch": 1.829347402130848,
+      "grad_norm": 0.43774133544125177,
+      "learning_rate": 6.421606268603767e-06,
+      "loss": 0.4676,
+      "step": 11139
+    },
+    {
+      "epoch": 1.8295116293391906,
+      "grad_norm": 0.5109065743607105,
+      "learning_rate": 6.421182194526407e-06,
+      "loss": 0.4339,
+      "step": 11140
+    },
+    {
+      "epoch": 1.8296758565475335,
+      "grad_norm": 0.2663502067197082,
+      "learning_rate": 6.420758099583119e-06,
+      "loss": 0.4548,
+      "step": 11141
+    },
+    {
+      "epoch": 1.8298400837558764,
+      "grad_norm": 0.35162665093102086,
+      "learning_rate": 6.420333983778507e-06,
+      "loss": 0.4549,
+      "step": 11142
+    },
+    {
+      "epoch": 1.830004310964219,
+      "grad_norm": 0.3582562471813268,
+      "learning_rate": 6.419909847117179e-06,
+      "loss": 0.4602,
+      "step": 11143
+    },
+    {
+      "epoch": 1.8301685381725616,
+      "grad_norm": 0.3064404860661055,
+      "learning_rate": 6.4194856896037416e-06,
+      "loss": 0.4517,
+      "step": 11144
+    },
+    {
+      "epoch": 1.8303327653809045,
+      "grad_norm": 0.3968100953955666,
+      "learning_rate": 6.419061511242799e-06,
+      "loss": 0.4455,
+      "step": 11145
+    },
+    {
+      "epoch": 1.8304969925892474,
+      "grad_norm": 0.3523584283490642,
+      "learning_rate": 6.418637312038963e-06,
+      "loss": 0.4545,
+      "step": 11146
+    },
+    {
+      "epoch": 1.83066121979759,
+      "grad_norm": 0.335856440378906,
+      "learning_rate": 6.4182130919968375e-06,
+      "loss": 0.4475,
+      "step": 11147
+    },
+    {
+      "epoch": 1.8308254470059326,
+      "grad_norm": 0.3146367836178673,
+      "learning_rate": 6.41778885112103e-06,
+      "loss": 0.4426,
+      "step": 11148
+    },
+    {
+      "epoch": 1.8309896742142755,
+      "grad_norm": 0.3653371520959425,
+      "learning_rate": 6.417364589416148e-06,
+      "loss": 0.45,
+      "step": 11149
+    },
+    {
+      "epoch": 1.8311539014226184,
+      "grad_norm": 0.3708603837292037,
+      "learning_rate": 6.4169403068868e-06,
+      "loss": 0.4656,
+      "step": 11150
+    },
+    {
+      "epoch": 1.831318128630961,
+      "grad_norm": 0.29649438936926636,
+      "learning_rate": 6.416516003537597e-06,
+      "loss": 0.4495,
+      "step": 11151
+    },
+    {
+      "epoch": 1.8314823558393036,
+      "grad_norm": 0.3521235978642944,
+      "learning_rate": 6.416091679373144e-06,
+      "loss": 0.4578,
+      "step": 11152
+    },
+    {
+      "epoch": 1.8316465830476463,
+      "grad_norm": 0.39789361916375526,
+      "learning_rate": 6.415667334398047e-06,
+      "loss": 0.4686,
+      "step": 11153
+    },
+    {
+      "epoch": 1.8318108102559891,
+      "grad_norm": 3.7291295388929075,
+      "learning_rate": 6.4152429686169195e-06,
+      "loss": 0.4464,
+      "step": 11154
+    },
+    {
+      "epoch": 1.831975037464332,
+      "grad_norm": 0.3062939593911645,
+      "learning_rate": 6.414818582034371e-06,
+      "loss": 0.4477,
+      "step": 11155
+    },
+    {
+      "epoch": 1.8321392646726746,
+      "grad_norm": 0.42487906457102476,
+      "learning_rate": 6.414394174655007e-06,
+      "loss": 0.4378,
+      "step": 11156
+    },
+    {
+      "epoch": 1.8323034918810173,
+      "grad_norm": 0.4173646928078844,
+      "learning_rate": 6.413969746483439e-06,
+      "loss": 0.4466,
+      "step": 11157
+    },
+    {
+      "epoch": 1.8324677190893601,
+      "grad_norm": 0.29528763334595215,
+      "learning_rate": 6.413545297524276e-06,
+      "loss": 0.4338,
+      "step": 11158
+    },
+    {
+      "epoch": 1.832631946297703,
+      "grad_norm": 0.3550556187269253,
+      "learning_rate": 6.413120827782128e-06,
+      "loss": 0.4294,
+      "step": 11159
+    },
+    {
+      "epoch": 1.8327961735060456,
+      "grad_norm": 0.35767683689815905,
+      "learning_rate": 6.412696337261608e-06,
+      "loss": 0.458,
+      "step": 11160
+    },
+    {
+      "epoch": 1.8329604007143883,
+      "grad_norm": 0.28119233715005026,
+      "learning_rate": 6.412271825967322e-06,
+      "loss": 0.4302,
+      "step": 11161
+    },
+    {
+      "epoch": 1.8331246279227311,
+      "grad_norm": 0.4170109914341217,
+      "learning_rate": 6.411847293903883e-06,
+      "loss": 0.4465,
+      "step": 11162
+    },
+    {
+      "epoch": 1.833288855131074,
+      "grad_norm": 0.34109494612960684,
+      "learning_rate": 6.4114227410759004e-06,
+      "loss": 0.4252,
+      "step": 11163
+    },
+    {
+      "epoch": 1.8334530823394166,
+      "grad_norm": 0.3907746165563666,
+      "learning_rate": 6.410998167487988e-06,
+      "loss": 0.4729,
+      "step": 11164
+    },
+    {
+      "epoch": 1.8336173095477593,
+      "grad_norm": 0.2953268045869606,
+      "learning_rate": 6.410573573144754e-06,
+      "loss": 0.4539,
+      "step": 11165
+    },
+    {
+      "epoch": 1.8337815367561021,
+      "grad_norm": 0.3687647764909497,
+      "learning_rate": 6.410148958050813e-06,
+      "loss": 0.4563,
+      "step": 11166
+    },
+    {
+      "epoch": 1.833945763964445,
+      "grad_norm": 0.49949614089435895,
+      "learning_rate": 6.409724322210772e-06,
+      "loss": 0.4785,
+      "step": 11167
+    },
+    {
+      "epoch": 1.8341099911727876,
+      "grad_norm": 0.35319171075997546,
+      "learning_rate": 6.4092996656292495e-06,
+      "loss": 0.4475,
+      "step": 11168
+    },
+    {
+      "epoch": 1.8342742183811303,
+      "grad_norm": 0.3578914249392386,
+      "learning_rate": 6.408874988310852e-06,
+      "loss": 0.4555,
+      "step": 11169
+    },
+    {
+      "epoch": 1.8344384455894729,
+      "grad_norm": 0.5233504521881581,
+      "learning_rate": 6.4084502902601946e-06,
+      "loss": 0.4535,
+      "step": 11170
+    },
+    {
+      "epoch": 1.8346026727978157,
+      "grad_norm": 0.3695067674128277,
+      "learning_rate": 6.408025571481889e-06,
+      "loss": 0.4489,
+      "step": 11171
+    },
+    {
+      "epoch": 1.8347669000061586,
+      "grad_norm": 0.3542537195814928,
+      "learning_rate": 6.407600831980548e-06,
+      "loss": 0.444,
+      "step": 11172
+    },
+    {
+      "epoch": 1.8349311272145012,
+      "grad_norm": 0.433248517239817,
+      "learning_rate": 6.407176071760787e-06,
+      "loss": 0.4562,
+      "step": 11173
+    },
+    {
+      "epoch": 1.8350953544228439,
+      "grad_norm": 0.43227636924692336,
+      "learning_rate": 6.406751290827214e-06,
+      "loss": 0.4399,
+      "step": 11174
+    },
+    {
+      "epoch": 1.8352595816311867,
+      "grad_norm": 0.43145471772441574,
+      "learning_rate": 6.40632648918445e-06,
+      "loss": 0.4457,
+      "step": 11175
+    },
+    {
+      "epoch": 1.8354238088395296,
+      "grad_norm": 0.3710357940741841,
+      "learning_rate": 6.405901666837102e-06,
+      "loss": 0.4319,
+      "step": 11176
+    },
+    {
+      "epoch": 1.8355880360478722,
+      "grad_norm": 0.3588178442046596,
+      "learning_rate": 6.405476823789788e-06,
+      "loss": 0.4713,
+      "step": 11177
+    },
+    {
+      "epoch": 1.8357522632562149,
+      "grad_norm": 0.3383423727372397,
+      "learning_rate": 6.4050519600471205e-06,
+      "loss": 0.45,
+      "step": 11178
+    },
+    {
+      "epoch": 1.8359164904645577,
+      "grad_norm": 0.3492514766969178,
+      "learning_rate": 6.404627075613715e-06,
+      "loss": 0.471,
+      "step": 11179
+    },
+    {
+      "epoch": 1.8360807176729006,
+      "grad_norm": 0.41424669052303414,
+      "learning_rate": 6.404202170494184e-06,
+      "loss": 0.4609,
+      "step": 11180
+    },
+    {
+      "epoch": 1.8362449448812432,
+      "grad_norm": 0.36182965770449854,
+      "learning_rate": 6.403777244693146e-06,
+      "loss": 0.4654,
+      "step": 11181
+    },
+    {
+      "epoch": 1.8364091720895859,
+      "grad_norm": 0.3327741789191045,
+      "learning_rate": 6.403352298215212e-06,
+      "loss": 0.4618,
+      "step": 11182
+    },
+    {
+      "epoch": 1.8365733992979287,
+      "grad_norm": 0.41298626199924565,
+      "learning_rate": 6.402927331065001e-06,
+      "loss": 0.4606,
+      "step": 11183
+    },
+    {
+      "epoch": 1.8367376265062716,
+      "grad_norm": 0.2869633366868695,
+      "learning_rate": 6.402502343247126e-06,
+      "loss": 0.4676,
+      "step": 11184
+    },
+    {
+      "epoch": 1.8369018537146142,
+      "grad_norm": 0.37976823478720945,
+      "learning_rate": 6.402077334766204e-06,
+      "loss": 0.4525,
+      "step": 11185
+    },
+    {
+      "epoch": 1.8370660809229569,
+      "grad_norm": 0.41990835423104483,
+      "learning_rate": 6.401652305626852e-06,
+      "loss": 0.4548,
+      "step": 11186
+    },
+    {
+      "epoch": 1.8372303081312995,
+      "grad_norm": 0.3411060115422886,
+      "learning_rate": 6.401227255833683e-06,
+      "loss": 0.4496,
+      "step": 11187
+    },
+    {
+      "epoch": 1.8373945353396424,
+      "grad_norm": 1.1713248336301512,
+      "learning_rate": 6.400802185391317e-06,
+      "loss": 0.4754,
+      "step": 11188
+    },
+    {
+      "epoch": 1.8375587625479852,
+      "grad_norm": 1.1536385664772004,
+      "learning_rate": 6.4003770943043685e-06,
+      "loss": 0.4581,
+      "step": 11189
+    },
+    {
+      "epoch": 1.8377229897563279,
+      "grad_norm": 0.32862778514012486,
+      "learning_rate": 6.399951982577456e-06,
+      "loss": 0.4752,
+      "step": 11190
+    },
+    {
+      "epoch": 1.8378872169646705,
+      "grad_norm": 0.47493385870865834,
+      "learning_rate": 6.399526850215195e-06,
+      "loss": 0.4555,
+      "step": 11191
+    },
+    {
+      "epoch": 1.8380514441730134,
+      "grad_norm": 0.31223427158800315,
+      "learning_rate": 6.399101697222202e-06,
+      "loss": 0.4529,
+      "step": 11192
+    },
+    {
+      "epoch": 1.8382156713813562,
+      "grad_norm": 0.31446797620165784,
+      "learning_rate": 6.3986765236030975e-06,
+      "loss": 0.4471,
+      "step": 11193
+    },
+    {
+      "epoch": 1.8383798985896989,
+      "grad_norm": 0.44653437318643374,
+      "learning_rate": 6.398251329362498e-06,
+      "loss": 0.4879,
+      "step": 11194
+    },
+    {
+      "epoch": 1.8385441257980415,
+      "grad_norm": 0.3569741883200403,
+      "learning_rate": 6.397826114505022e-06,
+      "loss": 0.4637,
+      "step": 11195
+    },
+    {
+      "epoch": 1.8387083530063844,
+      "grad_norm": 0.2930468326068091,
+      "learning_rate": 6.397400879035285e-06,
+      "loss": 0.4349,
+      "step": 11196
+    },
+    {
+      "epoch": 1.8388725802147272,
+      "grad_norm": 0.3441165187924981,
+      "learning_rate": 6.3969756229579085e-06,
+      "loss": 0.4466,
+      "step": 11197
+    },
+    {
+      "epoch": 1.8390368074230699,
+      "grad_norm": 0.3680111685540521,
+      "learning_rate": 6.396550346277512e-06,
+      "loss": 0.4433,
+      "step": 11198
+    },
+    {
+      "epoch": 1.8392010346314125,
+      "grad_norm": 0.4387326278721702,
+      "learning_rate": 6.396125048998711e-06,
+      "loss": 0.4499,
+      "step": 11199
+    },
+    {
+      "epoch": 1.8393652618397553,
+      "grad_norm": 0.29303807948267785,
+      "learning_rate": 6.395699731126128e-06,
+      "loss": 0.4655,
+      "step": 11200
+    },
+    {
+      "epoch": 1.8395294890480982,
+      "grad_norm": 0.535697689652598,
+      "learning_rate": 6.3952743926643795e-06,
+      "loss": 0.4374,
+      "step": 11201
+    },
+    {
+      "epoch": 1.8396937162564408,
+      "grad_norm": 0.3118572177188081,
+      "learning_rate": 6.394849033618087e-06,
+      "loss": 0.4725,
+      "step": 11202
+    },
+    {
+      "epoch": 1.8398579434647835,
+      "grad_norm": 0.3140766818191945,
+      "learning_rate": 6.394423653991869e-06,
+      "loss": 0.4267,
+      "step": 11203
+    },
+    {
+      "epoch": 1.8400221706731261,
+      "grad_norm": 0.37368035891343115,
+      "learning_rate": 6.393998253790347e-06,
+      "loss": 0.4639,
+      "step": 11204
+    },
+    {
+      "epoch": 1.840186397881469,
+      "grad_norm": 0.2971672289149953,
+      "learning_rate": 6.39357283301814e-06,
+      "loss": 0.4433,
+      "step": 11205
+    },
+    {
+      "epoch": 1.8403506250898118,
+      "grad_norm": 0.3390585276018315,
+      "learning_rate": 6.3931473916798705e-06,
+      "loss": 0.4781,
+      "step": 11206
+    },
+    {
+      "epoch": 1.8405148522981545,
+      "grad_norm": 0.38748554649335143,
+      "learning_rate": 6.3927219297801555e-06,
+      "loss": 0.4494,
+      "step": 11207
+    },
+    {
+      "epoch": 1.8406790795064971,
+      "grad_norm": 0.49663527474275104,
+      "learning_rate": 6.39229644732362e-06,
+      "loss": 0.4611,
+      "step": 11208
+    },
+    {
+      "epoch": 1.84084330671484,
+      "grad_norm": 0.29911888732308106,
+      "learning_rate": 6.391870944314882e-06,
+      "loss": 0.4591,
+      "step": 11209
+    },
+    {
+      "epoch": 1.8410075339231828,
+      "grad_norm": 0.40218304374728053,
+      "learning_rate": 6.391445420758565e-06,
+      "loss": 0.4715,
+      "step": 11210
+    },
+    {
+      "epoch": 1.8411717611315255,
+      "grad_norm": 0.3619310185273555,
+      "learning_rate": 6.39101987665929e-06,
+      "loss": 0.4516,
+      "step": 11211
+    },
+    {
+      "epoch": 1.8413359883398681,
+      "grad_norm": 0.4285679890613997,
+      "learning_rate": 6.390594312021677e-06,
+      "loss": 0.4587,
+      "step": 11212
+    },
+    {
+      "epoch": 1.841500215548211,
+      "grad_norm": 0.2965380574492834,
+      "learning_rate": 6.390168726850351e-06,
+      "loss": 0.4338,
+      "step": 11213
+    },
+    {
+      "epoch": 1.8416644427565538,
+      "grad_norm": 0.3859482067158745,
+      "learning_rate": 6.3897431211499325e-06,
+      "loss": 0.4529,
+      "step": 11214
+    },
+    {
+      "epoch": 1.8418286699648965,
+      "grad_norm": 0.41682659190036797,
+      "learning_rate": 6.389317494925046e-06,
+      "loss": 0.4431,
+      "step": 11215
+    },
+    {
+      "epoch": 1.841992897173239,
+      "grad_norm": 0.3187290191100235,
+      "learning_rate": 6.388891848180311e-06,
+      "loss": 0.4507,
+      "step": 11216
+    },
+    {
+      "epoch": 1.842157124381582,
+      "grad_norm": 0.3103138081670325,
+      "learning_rate": 6.388466180920351e-06,
+      "loss": 0.4699,
+      "step": 11217
+    },
+    {
+      "epoch": 1.8423213515899248,
+      "grad_norm": 0.4371948820010515,
+      "learning_rate": 6.388040493149793e-06,
+      "loss": 0.4592,
+      "step": 11218
+    },
+    {
+      "epoch": 1.8424855787982675,
+      "grad_norm": 0.3792863854932938,
+      "learning_rate": 6.387614784873257e-06,
+      "loss": 0.4655,
+      "step": 11219
+    },
+    {
+      "epoch": 1.84264980600661,
+      "grad_norm": 0.454009914635368,
+      "learning_rate": 6.387189056095367e-06,
+      "loss": 0.4589,
+      "step": 11220
+    },
+    {
+      "epoch": 1.8428140332149527,
+      "grad_norm": 0.36236305614927683,
+      "learning_rate": 6.386763306820746e-06,
+      "loss": 0.4805,
+      "step": 11221
+    },
+    {
+      "epoch": 1.8429782604232956,
+      "grad_norm": 0.32699570043919124,
+      "learning_rate": 6.38633753705402e-06,
+      "loss": 0.4348,
+      "step": 11222
+    },
+    {
+      "epoch": 1.8431424876316385,
+      "grad_norm": 0.27557523057430056,
+      "learning_rate": 6.385911746799812e-06,
+      "loss": 0.4529,
+      "step": 11223
+    },
+    {
+      "epoch": 1.843306714839981,
+      "grad_norm": 0.3092554373958321,
+      "learning_rate": 6.385485936062749e-06,
+      "loss": 0.4562,
+      "step": 11224
+    },
+    {
+      "epoch": 1.8434709420483237,
+      "grad_norm": 0.36892198796982645,
+      "learning_rate": 6.3850601048474516e-06,
+      "loss": 0.4672,
+      "step": 11225
+    },
+    {
+      "epoch": 1.8436351692566666,
+      "grad_norm": 0.2739936180230769,
+      "learning_rate": 6.384634253158546e-06,
+      "loss": 0.4637,
+      "step": 11226
+    },
+    {
+      "epoch": 1.8437993964650095,
+      "grad_norm": 0.7730659708755419,
+      "learning_rate": 6.38420838100066e-06,
+      "loss": 0.4554,
+      "step": 11227
+    },
+    {
+      "epoch": 1.843963623673352,
+      "grad_norm": 0.691066161259853,
+      "learning_rate": 6.383782488378416e-06,
+      "loss": 0.456,
+      "step": 11228
+    },
+    {
+      "epoch": 1.8441278508816947,
+      "grad_norm": 0.3885199433170536,
+      "learning_rate": 6.3833565752964415e-06,
+      "loss": 0.4622,
+      "step": 11229
+    },
+    {
+      "epoch": 1.8442920780900376,
+      "grad_norm": 0.3272844786118214,
+      "learning_rate": 6.382930641759361e-06,
+      "loss": 0.4454,
+      "step": 11230
+    },
+    {
+      "epoch": 1.8444563052983804,
+      "grad_norm": 0.2924812894819204,
+      "learning_rate": 6.382504687771804e-06,
+      "loss": 0.4606,
+      "step": 11231
+    },
+    {
+      "epoch": 1.844620532506723,
+      "grad_norm": 0.30384658501138234,
+      "learning_rate": 6.382078713338391e-06,
+      "loss": 0.4614,
+      "step": 11232
+    },
+    {
+      "epoch": 1.8447847597150657,
+      "grad_norm": 0.2959473309078626,
+      "learning_rate": 6.3816527184637514e-06,
+      "loss": 0.4711,
+      "step": 11233
+    },
+    {
+      "epoch": 1.8449489869234086,
+      "grad_norm": 0.3025842567454638,
+      "learning_rate": 6.3812267031525125e-06,
+      "loss": 0.4502,
+      "step": 11234
+    },
+    {
+      "epoch": 1.8451132141317512,
+      "grad_norm": 0.2605784588380211,
+      "learning_rate": 6.3808006674093015e-06,
+      "loss": 0.4534,
+      "step": 11235
+    },
+    {
+      "epoch": 1.845277441340094,
+      "grad_norm": 0.48953924584015085,
+      "learning_rate": 6.380374611238743e-06,
+      "loss": 0.4455,
+      "step": 11236
+    },
+    {
+      "epoch": 1.8454416685484367,
+      "grad_norm": 0.629228985946868,
+      "learning_rate": 6.3799485346454685e-06,
+      "loss": 0.4351,
+      "step": 11237
+    },
+    {
+      "epoch": 1.8456058957567794,
+      "grad_norm": 0.3061764968428075,
+      "learning_rate": 6.379522437634102e-06,
+      "loss": 0.4563,
+      "step": 11238
+    },
+    {
+      "epoch": 1.8457701229651222,
+      "grad_norm": 1.4709802223407416,
+      "learning_rate": 6.379096320209273e-06,
+      "loss": 0.4539,
+      "step": 11239
+    },
+    {
+      "epoch": 1.845934350173465,
+      "grad_norm": 0.323637509736383,
+      "learning_rate": 6.378670182375609e-06,
+      "loss": 0.451,
+      "step": 11240
+    },
+    {
+      "epoch": 1.8460985773818077,
+      "grad_norm": 0.3057557907583729,
+      "learning_rate": 6.3782440241377375e-06,
+      "loss": 0.4581,
+      "step": 11241
+    },
+    {
+      "epoch": 1.8462628045901504,
+      "grad_norm": 0.304889386260149,
+      "learning_rate": 6.377817845500289e-06,
+      "loss": 0.4325,
+      "step": 11242
+    },
+    {
+      "epoch": 1.8464270317984932,
+      "grad_norm": 0.33035100993678235,
+      "learning_rate": 6.377391646467891e-06,
+      "loss": 0.4502,
+      "step": 11243
+    },
+    {
+      "epoch": 1.846591259006836,
+      "grad_norm": 0.29746641796158035,
+      "learning_rate": 6.376965427045173e-06,
+      "loss": 0.4503,
+      "step": 11244
+    },
+    {
+      "epoch": 1.8467554862151787,
+      "grad_norm": 0.32982674950063423,
+      "learning_rate": 6.376539187236764e-06,
+      "loss": 0.4692,
+      "step": 11245
+    },
+    {
+      "epoch": 1.8469197134235213,
+      "grad_norm": 0.3033746048395144,
+      "learning_rate": 6.376112927047292e-06,
+      "loss": 0.456,
+      "step": 11246
+    },
+    {
+      "epoch": 1.8470839406318642,
+      "grad_norm": 0.34344852873284026,
+      "learning_rate": 6.375686646481388e-06,
+      "loss": 0.4693,
+      "step": 11247
+    },
+    {
+      "epoch": 1.847248167840207,
+      "grad_norm": 0.3152472718518401,
+      "learning_rate": 6.375260345543683e-06,
+      "loss": 0.4738,
+      "step": 11248
+    },
+    {
+      "epoch": 1.8474123950485497,
+      "grad_norm": 0.34465837766764895,
+      "learning_rate": 6.374834024238805e-06,
+      "loss": 0.4589,
+      "step": 11249
+    },
+    {
+      "epoch": 1.8475766222568923,
+      "grad_norm": 0.33303273260166083,
+      "learning_rate": 6.374407682571384e-06,
+      "loss": 0.481,
+      "step": 11250
+    },
+    {
+      "epoch": 1.8477408494652352,
+      "grad_norm": 0.2974928118083777,
+      "learning_rate": 6.373981320546051e-06,
+      "loss": 0.4489,
+      "step": 11251
+    },
+    {
+      "epoch": 1.8479050766735778,
+      "grad_norm": 0.3249645450729173,
+      "learning_rate": 6.373554938167439e-06,
+      "loss": 0.4445,
+      "step": 11252
+    },
+    {
+      "epoch": 1.8480693038819207,
+      "grad_norm": 0.3307822690469634,
+      "learning_rate": 6.373128535440177e-06,
+      "loss": 0.471,
+      "step": 11253
+    },
+    {
+      "epoch": 1.8482335310902633,
+      "grad_norm": 0.3783593787704891,
+      "learning_rate": 6.372702112368894e-06,
+      "loss": 0.4261,
+      "step": 11254
+    },
+    {
+      "epoch": 1.848397758298606,
+      "grad_norm": 0.3394670412812821,
+      "learning_rate": 6.372275668958225e-06,
+      "loss": 0.472,
+      "step": 11255
+    },
+    {
+      "epoch": 1.8485619855069488,
+      "grad_norm": 0.33822575181629483,
+      "learning_rate": 6.371849205212801e-06,
+      "loss": 0.4439,
+      "step": 11256
+    },
+    {
+      "epoch": 1.8487262127152917,
+      "grad_norm": 0.3142075629376298,
+      "learning_rate": 6.371422721137252e-06,
+      "loss": 0.4464,
+      "step": 11257
+    },
+    {
+      "epoch": 1.8488904399236343,
+      "grad_norm": 0.3342951923344951,
+      "learning_rate": 6.370996216736211e-06,
+      "loss": 0.4779,
+      "step": 11258
+    },
+    {
+      "epoch": 1.849054667131977,
+      "grad_norm": 0.349350432116345,
+      "learning_rate": 6.370569692014309e-06,
+      "loss": 0.4233,
+      "step": 11259
+    },
+    {
+      "epoch": 1.8492188943403198,
+      "grad_norm": 0.29426312829915247,
+      "learning_rate": 6.370143146976182e-06,
+      "loss": 0.4605,
+      "step": 11260
+    },
+    {
+      "epoch": 1.8493831215486627,
+      "grad_norm": 0.6891572867343618,
+      "learning_rate": 6.369716581626459e-06,
+      "loss": 0.4539,
+      "step": 11261
+    },
+    {
+      "epoch": 1.8495473487570053,
+      "grad_norm": 0.3566966922620718,
+      "learning_rate": 6.3692899959697735e-06,
+      "loss": 0.4473,
+      "step": 11262
+    },
+    {
+      "epoch": 1.849711575965348,
+      "grad_norm": 0.3671286591601414,
+      "learning_rate": 6.368863390010762e-06,
+      "loss": 0.4626,
+      "step": 11263
+    },
+    {
+      "epoch": 1.8498758031736908,
+      "grad_norm": 0.4516765235822067,
+      "learning_rate": 6.368436763754055e-06,
+      "loss": 0.4572,
+      "step": 11264
+    },
+    {
+      "epoch": 1.8500400303820337,
+      "grad_norm": 0.2954919181729513,
+      "learning_rate": 6.368010117204286e-06,
+      "loss": 0.4393,
+      "step": 11265
+    },
+    {
+      "epoch": 1.8502042575903763,
+      "grad_norm": 0.3493134392224947,
+      "learning_rate": 6.367583450366087e-06,
+      "loss": 0.4567,
+      "step": 11266
+    },
+    {
+      "epoch": 1.850368484798719,
+      "grad_norm": 0.35872074525718806,
+      "learning_rate": 6.367156763244097e-06,
+      "loss": 0.4561,
+      "step": 11267
+    },
+    {
+      "epoch": 1.8505327120070618,
+      "grad_norm": 0.42804427248337146,
+      "learning_rate": 6.3667300558429475e-06,
+      "loss": 0.4678,
+      "step": 11268
+    },
+    {
+      "epoch": 1.8506969392154045,
+      "grad_norm": 0.39557124706471175,
+      "learning_rate": 6.366303328167273e-06,
+      "loss": 0.4527,
+      "step": 11269
+    },
+    {
+      "epoch": 1.8508611664237473,
+      "grad_norm": 0.3645519450986868,
+      "learning_rate": 6.365876580221706e-06,
+      "loss": 0.4439,
+      "step": 11270
+    },
+    {
+      "epoch": 1.85102539363209,
+      "grad_norm": 0.2611586402914713,
+      "learning_rate": 6.365449812010884e-06,
+      "loss": 0.4624,
+      "step": 11271
+    },
+    {
+      "epoch": 1.8511896208404326,
+      "grad_norm": 0.4224856050672986,
+      "learning_rate": 6.365023023539444e-06,
+      "loss": 0.4262,
+      "step": 11272
+    },
+    {
+      "epoch": 1.8513538480487755,
+      "grad_norm": 0.34426711144349553,
+      "learning_rate": 6.364596214812018e-06,
+      "loss": 0.4635,
+      "step": 11273
+    },
+    {
+      "epoch": 1.8515180752571183,
+      "grad_norm": 0.33044232606364654,
+      "learning_rate": 6.364169385833242e-06,
+      "loss": 0.4413,
+      "step": 11274
+    },
+    {
+      "epoch": 1.851682302465461,
+      "grad_norm": 0.3566030564790324,
+      "learning_rate": 6.363742536607753e-06,
+      "loss": 0.4657,
+      "step": 11275
+    },
+    {
+      "epoch": 1.8518465296738036,
+      "grad_norm": 0.32664316500962876,
+      "learning_rate": 6.363315667140185e-06,
+      "loss": 0.4231,
+      "step": 11276
+    },
+    {
+      "epoch": 1.8520107568821464,
+      "grad_norm": 0.3192712622806749,
+      "learning_rate": 6.362888777435177e-06,
+      "loss": 0.4429,
+      "step": 11277
+    },
+    {
+      "epoch": 1.8521749840904893,
+      "grad_norm": 0.3483004909516952,
+      "learning_rate": 6.362461867497364e-06,
+      "loss": 0.4459,
+      "step": 11278
+    },
+    {
+      "epoch": 1.852339211298832,
+      "grad_norm": 0.3238437276102331,
+      "learning_rate": 6.362034937331382e-06,
+      "loss": 0.4658,
+      "step": 11279
+    },
+    {
+      "epoch": 1.8525034385071746,
+      "grad_norm": 0.4155190721399048,
+      "learning_rate": 6.361607986941869e-06,
+      "loss": 0.4594,
+      "step": 11280
+    },
+    {
+      "epoch": 1.8526676657155174,
+      "grad_norm": 0.4259720186833352,
+      "learning_rate": 6.361181016333462e-06,
+      "loss": 0.4427,
+      "step": 11281
+    },
+    {
+      "epoch": 1.8528318929238603,
+      "grad_norm": 0.39788743048437325,
+      "learning_rate": 6.360754025510797e-06,
+      "loss": 0.4389,
+      "step": 11282
+    },
+    {
+      "epoch": 1.852996120132203,
+      "grad_norm": 0.4959653416913325,
+      "learning_rate": 6.360327014478513e-06,
+      "loss": 0.4683,
+      "step": 11283
+    },
+    {
+      "epoch": 1.8531603473405456,
+      "grad_norm": 0.9145142089162646,
+      "learning_rate": 6.359899983241248e-06,
+      "loss": 0.4678,
+      "step": 11284
+    },
+    {
+      "epoch": 1.8533245745488884,
+      "grad_norm": 0.34958711678494037,
+      "learning_rate": 6.3594729318036395e-06,
+      "loss": 0.4392,
+      "step": 11285
+    },
+    {
+      "epoch": 1.853488801757231,
+      "grad_norm": 0.2915456832169805,
+      "learning_rate": 6.3590458601703234e-06,
+      "loss": 0.4396,
+      "step": 11286
+    },
+    {
+      "epoch": 1.853653028965574,
+      "grad_norm": 0.3523839879611003,
+      "learning_rate": 6.358618768345943e-06,
+      "loss": 0.4452,
+      "step": 11287
+    },
+    {
+      "epoch": 1.8538172561739166,
+      "grad_norm": 0.39252336971886814,
+      "learning_rate": 6.358191656335133e-06,
+      "loss": 0.4503,
+      "step": 11288
+    },
+    {
+      "epoch": 1.8539814833822592,
+      "grad_norm": 0.29758768939411184,
+      "learning_rate": 6.357764524142533e-06,
+      "loss": 0.4371,
+      "step": 11289
+    },
+    {
+      "epoch": 1.854145710590602,
+      "grad_norm": 0.2923988521358031,
+      "learning_rate": 6.3573373717727814e-06,
+      "loss": 0.4435,
+      "step": 11290
+    },
+    {
+      "epoch": 1.854309937798945,
+      "grad_norm": 0.3797175874288087,
+      "learning_rate": 6.35691019923052e-06,
+      "loss": 0.4735,
+      "step": 11291
+    },
+    {
+      "epoch": 1.8544741650072876,
+      "grad_norm": 0.3943571084660884,
+      "learning_rate": 6.356483006520387e-06,
+      "loss": 0.4385,
+      "step": 11292
+    },
+    {
+      "epoch": 1.8546383922156302,
+      "grad_norm": 0.422371885454595,
+      "learning_rate": 6.356055793647021e-06,
+      "loss": 0.4549,
+      "step": 11293
+    },
+    {
+      "epoch": 1.854802619423973,
+      "grad_norm": 0.2881492886554232,
+      "learning_rate": 6.3556285606150645e-06,
+      "loss": 0.4354,
+      "step": 11294
+    },
+    {
+      "epoch": 1.854966846632316,
+      "grad_norm": 0.2895915499671857,
+      "learning_rate": 6.355201307429155e-06,
+      "loss": 0.4554,
+      "step": 11295
+    },
+    {
+      "epoch": 1.8551310738406586,
+      "grad_norm": 0.27380290883797564,
+      "learning_rate": 6.354774034093934e-06,
+      "loss": 0.4457,
+      "step": 11296
+    },
+    {
+      "epoch": 1.8552953010490012,
+      "grad_norm": 0.3123756640040171,
+      "learning_rate": 6.354346740614043e-06,
+      "loss": 0.4317,
+      "step": 11297
+    },
+    {
+      "epoch": 1.855459528257344,
+      "grad_norm": 0.3097783710970955,
+      "learning_rate": 6.353919426994121e-06,
+      "loss": 0.4458,
+      "step": 11298
+    },
+    {
+      "epoch": 1.855623755465687,
+      "grad_norm": 0.31806916318986367,
+      "learning_rate": 6.353492093238811e-06,
+      "loss": 0.4459,
+      "step": 11299
+    },
+    {
+      "epoch": 1.8557879826740296,
+      "grad_norm": 0.42610233945463255,
+      "learning_rate": 6.353064739352752e-06,
+      "loss": 0.452,
+      "step": 11300
+    },
+    {
+      "epoch": 1.8559522098823722,
+      "grad_norm": 0.3118649517740116,
+      "learning_rate": 6.352637365340588e-06,
+      "loss": 0.4414,
+      "step": 11301
+    },
+    {
+      "epoch": 1.856116437090715,
+      "grad_norm": 0.2883834157653145,
+      "learning_rate": 6.352209971206959e-06,
+      "loss": 0.4447,
+      "step": 11302
+    },
+    {
+      "epoch": 1.8562806642990577,
+      "grad_norm": 0.30425811238812217,
+      "learning_rate": 6.3517825569565074e-06,
+      "loss": 0.4677,
+      "step": 11303
+    },
+    {
+      "epoch": 1.8564448915074006,
+      "grad_norm": 0.3157632228032478,
+      "learning_rate": 6.351355122593875e-06,
+      "loss": 0.451,
+      "step": 11304
+    },
+    {
+      "epoch": 1.8566091187157432,
+      "grad_norm": 0.31606115308966193,
+      "learning_rate": 6.350927668123704e-06,
+      "loss": 0.4615,
+      "step": 11305
+    },
+    {
+      "epoch": 1.8567733459240858,
+      "grad_norm": 0.2724015541206075,
+      "learning_rate": 6.350500193550638e-06,
+      "loss": 0.4353,
+      "step": 11306
+    },
+    {
+      "epoch": 1.8569375731324287,
+      "grad_norm": 0.2767441122373085,
+      "learning_rate": 6.35007269887932e-06,
+      "loss": 0.4567,
+      "step": 11307
+    },
+    {
+      "epoch": 1.8571018003407715,
+      "grad_norm": 0.29273901495114507,
+      "learning_rate": 6.349645184114392e-06,
+      "loss": 0.4567,
+      "step": 11308
+    },
+    {
+      "epoch": 1.8572660275491142,
+      "grad_norm": 0.2969139208502846,
+      "learning_rate": 6.349217649260497e-06,
+      "loss": 0.4342,
+      "step": 11309
+    },
+    {
+      "epoch": 1.8574302547574568,
+      "grad_norm": 0.35579388252131244,
+      "learning_rate": 6.34879009432228e-06,
+      "loss": 0.4611,
+      "step": 11310
+    },
+    {
+      "epoch": 1.8575944819657997,
+      "grad_norm": 0.34158770325629234,
+      "learning_rate": 6.348362519304382e-06,
+      "loss": 0.4616,
+      "step": 11311
+    },
+    {
+      "epoch": 1.8577587091741425,
+      "grad_norm": 0.28067291327273947,
+      "learning_rate": 6.34793492421145e-06,
+      "loss": 0.4455,
+      "step": 11312
+    },
+    {
+      "epoch": 1.8579229363824852,
+      "grad_norm": 0.30837793967609967,
+      "learning_rate": 6.347507309048125e-06,
+      "loss": 0.4578,
+      "step": 11313
+    },
+    {
+      "epoch": 1.8580871635908278,
+      "grad_norm": 0.3125025367261328,
+      "learning_rate": 6.347079673819053e-06,
+      "loss": 0.4472,
+      "step": 11314
+    },
+    {
+      "epoch": 1.8582513907991707,
+      "grad_norm": 0.5052978416465356,
+      "learning_rate": 6.34665201852888e-06,
+      "loss": 0.4471,
+      "step": 11315
+    },
+    {
+      "epoch": 1.8584156180075135,
+      "grad_norm": 0.31536032442555645,
+      "learning_rate": 6.346224343182248e-06,
+      "loss": 0.483,
+      "step": 11316
+    },
+    {
+      "epoch": 1.8585798452158562,
+      "grad_norm": 0.4470949513656592,
+      "learning_rate": 6.345796647783804e-06,
+      "loss": 0.4606,
+      "step": 11317
+    },
+    {
+      "epoch": 1.8587440724241988,
+      "grad_norm": 0.3170649277211733,
+      "learning_rate": 6.345368932338192e-06,
+      "loss": 0.4342,
+      "step": 11318
+    },
+    {
+      "epoch": 1.8589082996325417,
+      "grad_norm": 0.3424326385751127,
+      "learning_rate": 6.344941196850058e-06,
+      "loss": 0.463,
+      "step": 11319
+    },
+    {
+      "epoch": 1.8590725268408843,
+      "grad_norm": 0.3128336163128946,
+      "learning_rate": 6.344513441324048e-06,
+      "loss": 0.4494,
+      "step": 11320
+    },
+    {
+      "epoch": 1.8592367540492272,
+      "grad_norm": 0.4693741408208788,
+      "learning_rate": 6.344085665764806e-06,
+      "loss": 0.4228,
+      "step": 11321
+    },
+    {
+      "epoch": 1.8594009812575698,
+      "grad_norm": 0.3938432577957824,
+      "learning_rate": 6.343657870176979e-06,
+      "loss": 0.4579,
+      "step": 11322
+    },
+    {
+      "epoch": 1.8595652084659124,
+      "grad_norm": 0.2665157963971842,
+      "learning_rate": 6.343230054565215e-06,
+      "loss": 0.4559,
+      "step": 11323
+    },
+    {
+      "epoch": 1.8597294356742553,
+      "grad_norm": 0.5765119530939804,
+      "learning_rate": 6.342802218934159e-06,
+      "loss": 0.4366,
+      "step": 11324
+    },
+    {
+      "epoch": 1.8598936628825982,
+      "grad_norm": 0.6677510627653236,
+      "learning_rate": 6.342374363288456e-06,
+      "loss": 0.4634,
+      "step": 11325
+    },
+    {
+      "epoch": 1.8600578900909408,
+      "grad_norm": 0.3405761435040425,
+      "learning_rate": 6.341946487632758e-06,
+      "loss": 0.4518,
+      "step": 11326
+    },
+    {
+      "epoch": 1.8602221172992834,
+      "grad_norm": 0.28552553151580606,
+      "learning_rate": 6.341518591971707e-06,
+      "loss": 0.4468,
+      "step": 11327
+    },
+    {
+      "epoch": 1.8603863445076263,
+      "grad_norm": 0.3218955469025756,
+      "learning_rate": 6.341090676309951e-06,
+      "loss": 0.4628,
+      "step": 11328
+    },
+    {
+      "epoch": 1.8605505717159692,
+      "grad_norm": 0.3214289953278923,
+      "learning_rate": 6.340662740652141e-06,
+      "loss": 0.4289,
+      "step": 11329
+    },
+    {
+      "epoch": 1.8607147989243118,
+      "grad_norm": 0.2527221904206092,
+      "learning_rate": 6.340234785002922e-06,
+      "loss": 0.4347,
+      "step": 11330
+    },
+    {
+      "epoch": 1.8608790261326544,
+      "grad_norm": 0.39361376413646876,
+      "learning_rate": 6.339806809366942e-06,
+      "loss": 0.4583,
+      "step": 11331
+    },
+    {
+      "epoch": 1.8610432533409973,
+      "grad_norm": 0.36043320353855923,
+      "learning_rate": 6.339378813748852e-06,
+      "loss": 0.4427,
+      "step": 11332
+    },
+    {
+      "epoch": 1.8612074805493402,
+      "grad_norm": 0.8114253982192287,
+      "learning_rate": 6.338950798153295e-06,
+      "loss": 0.4365,
+      "step": 11333
+    },
+    {
+      "epoch": 1.8613717077576828,
+      "grad_norm": 0.2778367874294855,
+      "learning_rate": 6.338522762584925e-06,
+      "loss": 0.447,
+      "step": 11334
+    },
+    {
+      "epoch": 1.8615359349660254,
+      "grad_norm": 0.3560822146400267,
+      "learning_rate": 6.338094707048389e-06,
+      "loss": 0.4356,
+      "step": 11335
+    },
+    {
+      "epoch": 1.8617001621743683,
+      "grad_norm": 0.27192416670571484,
+      "learning_rate": 6.337666631548337e-06,
+      "loss": 0.4498,
+      "step": 11336
+    },
+    {
+      "epoch": 1.861864389382711,
+      "grad_norm": 0.4327915472121516,
+      "learning_rate": 6.337238536089416e-06,
+      "loss": 0.4366,
+      "step": 11337
+    },
+    {
+      "epoch": 1.8620286165910538,
+      "grad_norm": 0.4133339322356034,
+      "learning_rate": 6.336810420676277e-06,
+      "loss": 0.4509,
+      "step": 11338
+    },
+    {
+      "epoch": 1.8621928437993964,
+      "grad_norm": 0.2788288955428656,
+      "learning_rate": 6.336382285313569e-06,
+      "loss": 0.4571,
+      "step": 11339
+    },
+    {
+      "epoch": 1.862357071007739,
+      "grad_norm": 0.3527929954753714,
+      "learning_rate": 6.335954130005945e-06,
+      "loss": 0.4516,
+      "step": 11340
+    },
+    {
+      "epoch": 1.862521298216082,
+      "grad_norm": 0.4160277813715837,
+      "learning_rate": 6.335525954758051e-06,
+      "loss": 0.4483,
+      "step": 11341
+    },
+    {
+      "epoch": 1.8626855254244248,
+      "grad_norm": 0.41947127135248397,
+      "learning_rate": 6.335097759574539e-06,
+      "loss": 0.4715,
+      "step": 11342
+    },
+    {
+      "epoch": 1.8628497526327674,
+      "grad_norm": 0.30891972256795924,
+      "learning_rate": 6.33466954446006e-06,
+      "loss": 0.4757,
+      "step": 11343
+    },
+    {
+      "epoch": 1.86301397984111,
+      "grad_norm": 0.32281669512362693,
+      "learning_rate": 6.334241309419265e-06,
+      "loss": 0.4487,
+      "step": 11344
+    },
+    {
+      "epoch": 1.863178207049453,
+      "grad_norm": 0.3710096514926429,
+      "learning_rate": 6.333813054456805e-06,
+      "loss": 0.4512,
+      "step": 11345
+    },
+    {
+      "epoch": 1.8633424342577958,
+      "grad_norm": 0.297366741235457,
+      "learning_rate": 6.33338477957733e-06,
+      "loss": 0.4876,
+      "step": 11346
+    },
+    {
+      "epoch": 1.8635066614661384,
+      "grad_norm": 0.287759306786484,
+      "learning_rate": 6.332956484785495e-06,
+      "loss": 0.4574,
+      "step": 11347
+    },
+    {
+      "epoch": 1.863670888674481,
+      "grad_norm": 0.5127446230381244,
+      "learning_rate": 6.332528170085947e-06,
+      "loss": 0.4564,
+      "step": 11348
+    },
+    {
+      "epoch": 1.863835115882824,
+      "grad_norm": 0.3181484291823922,
+      "learning_rate": 6.33209983548334e-06,
+      "loss": 0.4511,
+      "step": 11349
+    },
+    {
+      "epoch": 1.8639993430911668,
+      "grad_norm": 0.36182292779200886,
+      "learning_rate": 6.331671480982328e-06,
+      "loss": 0.4509,
+      "step": 11350
+    },
+    {
+      "epoch": 1.8641635702995094,
+      "grad_norm": 0.3005556130697668,
+      "learning_rate": 6.3312431065875596e-06,
+      "loss": 0.4439,
+      "step": 11351
+    },
+    {
+      "epoch": 1.864327797507852,
+      "grad_norm": 0.31933998986102374,
+      "learning_rate": 6.33081471230369e-06,
+      "loss": 0.4796,
+      "step": 11352
+    },
+    {
+      "epoch": 1.864492024716195,
+      "grad_norm": 0.43163246367148256,
+      "learning_rate": 6.330386298135372e-06,
+      "loss": 0.4495,
+      "step": 11353
+    },
+    {
+      "epoch": 1.8646562519245375,
+      "grad_norm": 0.37365493661552046,
+      "learning_rate": 6.329957864087256e-06,
+      "loss": 0.4458,
+      "step": 11354
+    },
+    {
+      "epoch": 1.8648204791328804,
+      "grad_norm": 0.28485223530004894,
+      "learning_rate": 6.329529410163999e-06,
+      "loss": 0.4384,
+      "step": 11355
+    },
+    {
+      "epoch": 1.864984706341223,
+      "grad_norm": 0.3407623561451519,
+      "learning_rate": 6.329100936370253e-06,
+      "loss": 0.4628,
+      "step": 11356
+    },
+    {
+      "epoch": 1.8651489335495657,
+      "grad_norm": 0.3862070798121785,
+      "learning_rate": 6.328672442710671e-06,
+      "loss": 0.4508,
+      "step": 11357
+    },
+    {
+      "epoch": 1.8653131607579085,
+      "grad_norm": 0.3748507949476453,
+      "learning_rate": 6.328243929189905e-06,
+      "loss": 0.4458,
+      "step": 11358
+    },
+    {
+      "epoch": 1.8654773879662514,
+      "grad_norm": 0.38990948562174965,
+      "learning_rate": 6.327815395812613e-06,
+      "loss": 0.4455,
+      "step": 11359
+    },
+    {
+      "epoch": 1.865641615174594,
+      "grad_norm": 0.37073709138352673,
+      "learning_rate": 6.327386842583447e-06,
+      "loss": 0.4339,
+      "step": 11360
+    },
+    {
+      "epoch": 1.8658058423829367,
+      "grad_norm": 0.3471538957326833,
+      "learning_rate": 6.326958269507063e-06,
+      "loss": 0.4536,
+      "step": 11361
+    },
+    {
+      "epoch": 1.8659700695912795,
+      "grad_norm": 0.3642542342861628,
+      "learning_rate": 6.326529676588114e-06,
+      "loss": 0.4466,
+      "step": 11362
+    },
+    {
+      "epoch": 1.8661342967996224,
+      "grad_norm": 0.3782161620612055,
+      "learning_rate": 6.326101063831254e-06,
+      "loss": 0.4516,
+      "step": 11363
+    },
+    {
+      "epoch": 1.866298524007965,
+      "grad_norm": 0.3331236609817597,
+      "learning_rate": 6.325672431241142e-06,
+      "loss": 0.4546,
+      "step": 11364
+    },
+    {
+      "epoch": 1.8664627512163077,
+      "grad_norm": 1.0242190816400079,
+      "learning_rate": 6.325243778822431e-06,
+      "loss": 0.4514,
+      "step": 11365
+    },
+    {
+      "epoch": 1.8666269784246505,
+      "grad_norm": 0.27978995786744143,
+      "learning_rate": 6.324815106579777e-06,
+      "loss": 0.459,
+      "step": 11366
+    },
+    {
+      "epoch": 1.8667912056329934,
+      "grad_norm": 0.3215904181313786,
+      "learning_rate": 6.324386414517834e-06,
+      "loss": 0.452,
+      "step": 11367
+    },
+    {
+      "epoch": 1.866955432841336,
+      "grad_norm": 0.36826949228960465,
+      "learning_rate": 6.32395770264126e-06,
+      "loss": 0.4561,
+      "step": 11368
+    },
+    {
+      "epoch": 1.8671196600496787,
+      "grad_norm": 0.3002113463960166,
+      "learning_rate": 6.323528970954711e-06,
+      "loss": 0.4545,
+      "step": 11369
+    },
+    {
+      "epoch": 1.8672838872580215,
+      "grad_norm": 0.327024602938653,
+      "learning_rate": 6.323100219462844e-06,
+      "loss": 0.4568,
+      "step": 11370
+    },
+    {
+      "epoch": 1.8674481144663642,
+      "grad_norm": 0.44713301545973794,
+      "learning_rate": 6.322671448170314e-06,
+      "loss": 0.4622,
+      "step": 11371
+    },
+    {
+      "epoch": 1.867612341674707,
+      "grad_norm": 0.37617685259242234,
+      "learning_rate": 6.322242657081779e-06,
+      "loss": 0.451,
+      "step": 11372
+    },
+    {
+      "epoch": 1.8677765688830497,
+      "grad_norm": 0.4999178729558093,
+      "learning_rate": 6.321813846201897e-06,
+      "loss": 0.4676,
+      "step": 11373
+    },
+    {
+      "epoch": 1.8679407960913923,
+      "grad_norm": 0.3485035319342385,
+      "learning_rate": 6.321385015535323e-06,
+      "loss": 0.4607,
+      "step": 11374
+    },
+    {
+      "epoch": 1.8681050232997352,
+      "grad_norm": 0.3385822781097892,
+      "learning_rate": 6.320956165086716e-06,
+      "loss": 0.4748,
+      "step": 11375
+    },
+    {
+      "epoch": 1.868269250508078,
+      "grad_norm": 0.3730575244192388,
+      "learning_rate": 6.320527294860734e-06,
+      "loss": 0.4713,
+      "step": 11376
+    },
+    {
+      "epoch": 1.8684334777164207,
+      "grad_norm": 0.3223886623461952,
+      "learning_rate": 6.3200984048620335e-06,
+      "loss": 0.4469,
+      "step": 11377
+    },
+    {
+      "epoch": 1.8685977049247633,
+      "grad_norm": 0.31862136491199516,
+      "learning_rate": 6.319669495095275e-06,
+      "loss": 0.4537,
+      "step": 11378
+    },
+    {
+      "epoch": 1.8687619321331062,
+      "grad_norm": 0.29791203455607024,
+      "learning_rate": 6.3192405655651125e-06,
+      "loss": 0.4598,
+      "step": 11379
+    },
+    {
+      "epoch": 1.868926159341449,
+      "grad_norm": 0.32176269966805215,
+      "learning_rate": 6.318811616276211e-06,
+      "loss": 0.4483,
+      "step": 11380
+    },
+    {
+      "epoch": 1.8690903865497916,
+      "grad_norm": 0.43102468936947047,
+      "learning_rate": 6.318382647233225e-06,
+      "loss": 0.4427,
+      "step": 11381
+    },
+    {
+      "epoch": 1.8692546137581343,
+      "grad_norm": 0.33077619486864357,
+      "learning_rate": 6.3179536584408135e-06,
+      "loss": 0.4667,
+      "step": 11382
+    },
+    {
+      "epoch": 1.8694188409664771,
+      "grad_norm": 0.34418501391086825,
+      "learning_rate": 6.317524649903637e-06,
+      "loss": 0.4298,
+      "step": 11383
+    },
+    {
+      "epoch": 1.86958306817482,
+      "grad_norm": 0.29682961680933806,
+      "learning_rate": 6.317095621626354e-06,
+      "loss": 0.4531,
+      "step": 11384
+    },
+    {
+      "epoch": 1.8697472953831626,
+      "grad_norm": 0.3043037241928832,
+      "learning_rate": 6.316666573613625e-06,
+      "loss": 0.4474,
+      "step": 11385
+    },
+    {
+      "epoch": 1.8699115225915053,
+      "grad_norm": 0.39672356587770125,
+      "learning_rate": 6.316237505870111e-06,
+      "loss": 0.4473,
+      "step": 11386
+    },
+    {
+      "epoch": 1.8700757497998481,
+      "grad_norm": 0.32709641612353896,
+      "learning_rate": 6.315808418400469e-06,
+      "loss": 0.4533,
+      "step": 11387
+    },
+    {
+      "epoch": 1.8702399770081908,
+      "grad_norm": 0.571241424685496,
+      "learning_rate": 6.315379311209362e-06,
+      "loss": 0.4535,
+      "step": 11388
+    },
+    {
+      "epoch": 1.8704042042165336,
+      "grad_norm": 0.37875251809804106,
+      "learning_rate": 6.31495018430145e-06,
+      "loss": 0.452,
+      "step": 11389
+    },
+    {
+      "epoch": 1.8705684314248763,
+      "grad_norm": 0.35808189075596303,
+      "learning_rate": 6.3145210376813925e-06,
+      "loss": 0.4142,
+      "step": 11390
+    },
+    {
+      "epoch": 1.870732658633219,
+      "grad_norm": 0.33499970260438333,
+      "learning_rate": 6.314091871353852e-06,
+      "loss": 0.4386,
+      "step": 11391
+    },
+    {
+      "epoch": 1.8708968858415618,
+      "grad_norm": 0.3271825235316504,
+      "learning_rate": 6.313662685323488e-06,
+      "loss": 0.4348,
+      "step": 11392
+    },
+    {
+      "epoch": 1.8710611130499046,
+      "grad_norm": 0.3161373090038315,
+      "learning_rate": 6.313233479594963e-06,
+      "loss": 0.4542,
+      "step": 11393
+    },
+    {
+      "epoch": 1.8712253402582473,
+      "grad_norm": 0.3103016233161706,
+      "learning_rate": 6.312804254172938e-06,
+      "loss": 0.4525,
+      "step": 11394
+    },
+    {
+      "epoch": 1.87138956746659,
+      "grad_norm": 0.4571370996671186,
+      "learning_rate": 6.312375009062078e-06,
+      "loss": 0.4334,
+      "step": 11395
+    },
+    {
+      "epoch": 1.8715537946749328,
+      "grad_norm": 0.31005290994916324,
+      "learning_rate": 6.311945744267039e-06,
+      "loss": 0.4298,
+      "step": 11396
+    },
+    {
+      "epoch": 1.8717180218832756,
+      "grad_norm": 0.32531698774417556,
+      "learning_rate": 6.311516459792488e-06,
+      "loss": 0.4134,
+      "step": 11397
+    },
+    {
+      "epoch": 1.8718822490916183,
+      "grad_norm": 0.2637088668005535,
+      "learning_rate": 6.311087155643087e-06,
+      "loss": 0.4277,
+      "step": 11398
+    },
+    {
+      "epoch": 1.872046476299961,
+      "grad_norm": 0.352195618024151,
+      "learning_rate": 6.310657831823495e-06,
+      "loss": 0.4655,
+      "step": 11399
+    },
+    {
+      "epoch": 1.8722107035083038,
+      "grad_norm": 0.36961737788626986,
+      "learning_rate": 6.310228488338379e-06,
+      "loss": 0.4602,
+      "step": 11400
+    },
+    {
+      "epoch": 1.8723749307166466,
+      "grad_norm": 0.29607974177504476,
+      "learning_rate": 6.3097991251923995e-06,
+      "loss": 0.4531,
+      "step": 11401
+    },
+    {
+      "epoch": 1.8725391579249893,
+      "grad_norm": 0.3992638430794492,
+      "learning_rate": 6.309369742390224e-06,
+      "loss": 0.4544,
+      "step": 11402
+    },
+    {
+      "epoch": 1.872703385133332,
+      "grad_norm": 0.2969144434983537,
+      "learning_rate": 6.308940339936509e-06,
+      "loss": 0.4434,
+      "step": 11403
+    },
+    {
+      "epoch": 1.8728676123416748,
+      "grad_norm": 0.2615427443565544,
+      "learning_rate": 6.3085109178359245e-06,
+      "loss": 0.4419,
+      "step": 11404
+    },
+    {
+      "epoch": 1.8730318395500174,
+      "grad_norm": 0.32425811970912993,
+      "learning_rate": 6.308081476093131e-06,
+      "loss": 0.4489,
+      "step": 11405
+    },
+    {
+      "epoch": 1.8731960667583603,
+      "grad_norm": 0.47585657206487836,
+      "learning_rate": 6.3076520147127956e-06,
+      "loss": 0.4424,
+      "step": 11406
+    },
+    {
+      "epoch": 1.873360293966703,
+      "grad_norm": 0.30516338170610596,
+      "learning_rate": 6.30722253369958e-06,
+      "loss": 0.4413,
+      "step": 11407
+    },
+    {
+      "epoch": 1.8735245211750455,
+      "grad_norm": 0.39396978310804875,
+      "learning_rate": 6.306793033058147e-06,
+      "loss": 0.4667,
+      "step": 11408
+    },
+    {
+      "epoch": 1.8736887483833884,
+      "grad_norm": 0.3038677691494423,
+      "learning_rate": 6.306363512793167e-06,
+      "loss": 0.4502,
+      "step": 11409
+    },
+    {
+      "epoch": 1.8738529755917313,
+      "grad_norm": 0.31531753590896405,
+      "learning_rate": 6.305933972909301e-06,
+      "loss": 0.4226,
+      "step": 11410
+    },
+    {
+      "epoch": 1.874017202800074,
+      "grad_norm": 0.3287934612429399,
+      "learning_rate": 6.3055044134112165e-06,
+      "loss": 0.4417,
+      "step": 11411
+    },
+    {
+      "epoch": 1.8741814300084165,
+      "grad_norm": 0.27368395270120466,
+      "learning_rate": 6.305074834303576e-06,
+      "loss": 0.44,
+      "step": 11412
+    },
+    {
+      "epoch": 1.8743456572167594,
+      "grad_norm": 0.33455145144424686,
+      "learning_rate": 6.3046452355910465e-06,
+      "loss": 0.4573,
+      "step": 11413
+    },
+    {
+      "epoch": 1.8745098844251022,
+      "grad_norm": 0.3506306212009912,
+      "learning_rate": 6.304215617278296e-06,
+      "loss": 0.4549,
+      "step": 11414
+    },
+    {
+      "epoch": 1.8746741116334449,
+      "grad_norm": 0.3501676831247077,
+      "learning_rate": 6.303785979369988e-06,
+      "loss": 0.4774,
+      "step": 11415
+    },
+    {
+      "epoch": 1.8748383388417875,
+      "grad_norm": 0.2696692929128657,
+      "learning_rate": 6.30335632187079e-06,
+      "loss": 0.4582,
+      "step": 11416
+    },
+    {
+      "epoch": 1.8750025660501304,
+      "grad_norm": 0.2685861402212932,
+      "learning_rate": 6.302926644785367e-06,
+      "loss": 0.4552,
+      "step": 11417
+    },
+    {
+      "epoch": 1.8751667932584732,
+      "grad_norm": 0.5705129421412283,
+      "learning_rate": 6.302496948118388e-06,
+      "loss": 0.4575,
+      "step": 11418
+    },
+    {
+      "epoch": 1.8753310204668159,
+      "grad_norm": 0.27189216908107755,
+      "learning_rate": 6.302067231874519e-06,
+      "loss": 0.4584,
+      "step": 11419
+    },
+    {
+      "epoch": 1.8754952476751585,
+      "grad_norm": 0.30435635762114543,
+      "learning_rate": 6.3016374960584245e-06,
+      "loss": 0.4211,
+      "step": 11420
+    },
+    {
+      "epoch": 1.8756594748835014,
+      "grad_norm": 0.35066223874902763,
+      "learning_rate": 6.301207740674776e-06,
+      "loss": 0.4382,
+      "step": 11421
+    },
+    {
+      "epoch": 1.875823702091844,
+      "grad_norm": 0.2863261329375365,
+      "learning_rate": 6.300777965728238e-06,
+      "loss": 0.4595,
+      "step": 11422
+    },
+    {
+      "epoch": 1.8759879293001869,
+      "grad_norm": 0.3336400798811265,
+      "learning_rate": 6.300348171223482e-06,
+      "loss": 0.4607,
+      "step": 11423
+    },
+    {
+      "epoch": 1.8761521565085295,
+      "grad_norm": 0.265241509790696,
+      "learning_rate": 6.299918357165172e-06,
+      "loss": 0.4402,
+      "step": 11424
+    },
+    {
+      "epoch": 1.8763163837168721,
+      "grad_norm": 0.3908903746403792,
+      "learning_rate": 6.299488523557977e-06,
+      "loss": 0.4574,
+      "step": 11425
+    },
+    {
+      "epoch": 1.876480610925215,
+      "grad_norm": 0.3081261018458057,
+      "learning_rate": 6.299058670406567e-06,
+      "loss": 0.4411,
+      "step": 11426
+    },
+    {
+      "epoch": 1.8766448381335579,
+      "grad_norm": 0.36235121919155894,
+      "learning_rate": 6.298628797715611e-06,
+      "loss": 0.4471,
+      "step": 11427
+    },
+    {
+      "epoch": 1.8768090653419005,
+      "grad_norm": 0.3715659252716504,
+      "learning_rate": 6.298198905489775e-06,
+      "loss": 0.4622,
+      "step": 11428
+    },
+    {
+      "epoch": 1.8769732925502431,
+      "grad_norm": 0.2812323777900102,
+      "learning_rate": 6.297768993733731e-06,
+      "loss": 0.455,
+      "step": 11429
+    },
+    {
+      "epoch": 1.877137519758586,
+      "grad_norm": 0.5437456561571391,
+      "learning_rate": 6.297339062452145e-06,
+      "loss": 0.4685,
+      "step": 11430
+    },
+    {
+      "epoch": 1.8773017469669289,
+      "grad_norm": 0.3013365443414987,
+      "learning_rate": 6.296909111649689e-06,
+      "loss": 0.4403,
+      "step": 11431
+    },
+    {
+      "epoch": 1.8774659741752715,
+      "grad_norm": 0.3600951978257436,
+      "learning_rate": 6.296479141331033e-06,
+      "loss": 0.4688,
+      "step": 11432
+    },
+    {
+      "epoch": 1.8776302013836141,
+      "grad_norm": 0.32973648411913603,
+      "learning_rate": 6.296049151500847e-06,
+      "loss": 0.4671,
+      "step": 11433
+    },
+    {
+      "epoch": 1.877794428591957,
+      "grad_norm": 0.77516515033505,
+      "learning_rate": 6.295619142163799e-06,
+      "loss": 0.448,
+      "step": 11434
+    },
+    {
+      "epoch": 1.8779586558002999,
+      "grad_norm": 0.2813318600336481,
+      "learning_rate": 6.295189113324559e-06,
+      "loss": 0.4392,
+      "step": 11435
+    },
+    {
+      "epoch": 1.8781228830086425,
+      "grad_norm": 0.38431140428798827,
+      "learning_rate": 6.294759064987801e-06,
+      "loss": 0.4476,
+      "step": 11436
+    },
+    {
+      "epoch": 1.8782871102169851,
+      "grad_norm": 0.35869495335306845,
+      "learning_rate": 6.294328997158193e-06,
+      "loss": 0.428,
+      "step": 11437
+    },
+    {
+      "epoch": 1.878451337425328,
+      "grad_norm": 0.34393025001890787,
+      "learning_rate": 6.293898909840407e-06,
+      "loss": 0.443,
+      "step": 11438
+    },
+    {
+      "epoch": 1.8786155646336706,
+      "grad_norm": 0.30989139809827226,
+      "learning_rate": 6.293468803039114e-06,
+      "loss": 0.4453,
+      "step": 11439
+    },
+    {
+      "epoch": 1.8787797918420135,
+      "grad_norm": 0.3166916712968177,
+      "learning_rate": 6.293038676758985e-06,
+      "loss": 0.451,
+      "step": 11440
+    },
+    {
+      "epoch": 1.8789440190503561,
+      "grad_norm": 0.27680712337488805,
+      "learning_rate": 6.292608531004692e-06,
+      "loss": 0.4412,
+      "step": 11441
+    },
+    {
+      "epoch": 1.8791082462586988,
+      "grad_norm": 0.2942760765393107,
+      "learning_rate": 6.292178365780906e-06,
+      "loss": 0.4561,
+      "step": 11442
+    },
+    {
+      "epoch": 1.8792724734670416,
+      "grad_norm": 0.39917989420901784,
+      "learning_rate": 6.2917481810923e-06,
+      "loss": 0.4643,
+      "step": 11443
+    },
+    {
+      "epoch": 1.8794367006753845,
+      "grad_norm": 0.6947523304053774,
+      "learning_rate": 6.291317976943547e-06,
+      "loss": 0.4524,
+      "step": 11444
+    },
+    {
+      "epoch": 1.8796009278837271,
+      "grad_norm": 0.41434515474515987,
+      "learning_rate": 6.2908877533393164e-06,
+      "loss": 0.4302,
+      "step": 11445
+    },
+    {
+      "epoch": 1.8797651550920698,
+      "grad_norm": 0.45832914292702964,
+      "learning_rate": 6.290457510284283e-06,
+      "loss": 0.4517,
+      "step": 11446
+    },
+    {
+      "epoch": 1.8799293823004126,
+      "grad_norm": 0.32247693907639524,
+      "learning_rate": 6.290027247783121e-06,
+      "loss": 0.4395,
+      "step": 11447
+    },
+    {
+      "epoch": 1.8800936095087555,
+      "grad_norm": 0.2892060133011612,
+      "learning_rate": 6.289596965840503e-06,
+      "loss": 0.4655,
+      "step": 11448
+    },
+    {
+      "epoch": 1.8802578367170981,
+      "grad_norm": 0.3216050782026567,
+      "learning_rate": 6.2891666644610985e-06,
+      "loss": 0.4504,
+      "step": 11449
+    },
+    {
+      "epoch": 1.8804220639254408,
+      "grad_norm": 0.38605541235934754,
+      "learning_rate": 6.288736343649584e-06,
+      "loss": 0.4442,
+      "step": 11450
+    },
+    {
+      "epoch": 1.8805862911337836,
+      "grad_norm": 0.6160848873275179,
+      "learning_rate": 6.288306003410633e-06,
+      "loss": 0.468,
+      "step": 11451
+    },
+    {
+      "epoch": 1.8807505183421265,
+      "grad_norm": 0.3142613591082338,
+      "learning_rate": 6.287875643748921e-06,
+      "loss": 0.4312,
+      "step": 11452
+    },
+    {
+      "epoch": 1.8809147455504691,
+      "grad_norm": 0.3910821269458435,
+      "learning_rate": 6.2874452646691205e-06,
+      "loss": 0.4533,
+      "step": 11453
+    },
+    {
+      "epoch": 1.8810789727588118,
+      "grad_norm": 0.3190445080531141,
+      "learning_rate": 6.287014866175905e-06,
+      "loss": 0.4482,
+      "step": 11454
+    },
+    {
+      "epoch": 1.8812431999671546,
+      "grad_norm": 0.3013278803822531,
+      "learning_rate": 6.286584448273949e-06,
+      "loss": 0.4392,
+      "step": 11455
+    },
+    {
+      "epoch": 1.8814074271754972,
+      "grad_norm": 0.33607970980024,
+      "learning_rate": 6.286154010967928e-06,
+      "loss": 0.4596,
+      "step": 11456
+    },
+    {
+      "epoch": 1.88157165438384,
+      "grad_norm": 0.28505845262642243,
+      "learning_rate": 6.285723554262519e-06,
+      "loss": 0.4511,
+      "step": 11457
+    },
+    {
+      "epoch": 1.8817358815921827,
+      "grad_norm": 0.34519722662923064,
+      "learning_rate": 6.285293078162394e-06,
+      "loss": 0.4544,
+      "step": 11458
+    },
+    {
+      "epoch": 1.8819001088005254,
+      "grad_norm": 0.26225405676276736,
+      "learning_rate": 6.28486258267223e-06,
+      "loss": 0.4177,
+      "step": 11459
+    },
+    {
+      "epoch": 1.8820643360088682,
+      "grad_norm": 0.40104514826593735,
+      "learning_rate": 6.284432067796701e-06,
+      "loss": 0.4517,
+      "step": 11460
+    },
+    {
+      "epoch": 1.882228563217211,
+      "grad_norm": 0.29963155183195445,
+      "learning_rate": 6.284001533540486e-06,
+      "loss": 0.45,
+      "step": 11461
+    },
+    {
+      "epoch": 1.8823927904255537,
+      "grad_norm": 0.30673249122938356,
+      "learning_rate": 6.283570979908258e-06,
+      "loss": 0.4439,
+      "step": 11462
+    },
+    {
+      "epoch": 1.8825570176338964,
+      "grad_norm": 0.3151831837335785,
+      "learning_rate": 6.283140406904695e-06,
+      "loss": 0.4615,
+      "step": 11463
+    },
+    {
+      "epoch": 1.8827212448422392,
+      "grad_norm": 0.2774161243114612,
+      "learning_rate": 6.282709814534472e-06,
+      "loss": 0.448,
+      "step": 11464
+    },
+    {
+      "epoch": 1.882885472050582,
+      "grad_norm": 0.32876418823333714,
+      "learning_rate": 6.282279202802268e-06,
+      "loss": 0.4354,
+      "step": 11465
+    },
+    {
+      "epoch": 1.8830496992589247,
+      "grad_norm": 0.36150603135257353,
+      "learning_rate": 6.281848571712756e-06,
+      "loss": 0.4361,
+      "step": 11466
+    },
+    {
+      "epoch": 1.8832139264672674,
+      "grad_norm": 0.36968781944725243,
+      "learning_rate": 6.281417921270618e-06,
+      "loss": 0.4453,
+      "step": 11467
+    },
+    {
+      "epoch": 1.8833781536756102,
+      "grad_norm": 0.37179395407086213,
+      "learning_rate": 6.280987251480527e-06,
+      "loss": 0.4458,
+      "step": 11468
+    },
+    {
+      "epoch": 1.883542380883953,
+      "grad_norm": 0.4567677928565284,
+      "learning_rate": 6.280556562347163e-06,
+      "loss": 0.4602,
+      "step": 11469
+    },
+    {
+      "epoch": 1.8837066080922957,
+      "grad_norm": 0.3907571110244243,
+      "learning_rate": 6.280125853875202e-06,
+      "loss": 0.4568,
+      "step": 11470
+    },
+    {
+      "epoch": 1.8838708353006384,
+      "grad_norm": 0.3045018771953139,
+      "learning_rate": 6.279695126069323e-06,
+      "loss": 0.4285,
+      "step": 11471
+    },
+    {
+      "epoch": 1.8840350625089812,
+      "grad_norm": 0.38405524916279843,
+      "learning_rate": 6.279264378934205e-06,
+      "loss": 0.4835,
+      "step": 11472
+    },
+    {
+      "epoch": 1.8841992897173239,
+      "grad_norm": 0.2940885242084993,
+      "learning_rate": 6.278833612474525e-06,
+      "loss": 0.4656,
+      "step": 11473
+    },
+    {
+      "epoch": 1.8843635169256667,
+      "grad_norm": 0.34809134178677464,
+      "learning_rate": 6.2784028266949615e-06,
+      "loss": 0.4426,
+      "step": 11474
+    },
+    {
+      "epoch": 1.8845277441340094,
+      "grad_norm": 0.27515836675655314,
+      "learning_rate": 6.277972021600192e-06,
+      "loss": 0.4433,
+      "step": 11475
+    },
+    {
+      "epoch": 1.884691971342352,
+      "grad_norm": 0.33595930419622255,
+      "learning_rate": 6.277541197194899e-06,
+      "loss": 0.4445,
+      "step": 11476
+    },
+    {
+      "epoch": 1.8848561985506949,
+      "grad_norm": 0.32328908953131097,
+      "learning_rate": 6.27711035348376e-06,
+      "loss": 0.4423,
+      "step": 11477
+    },
+    {
+      "epoch": 1.8850204257590377,
+      "grad_norm": 0.2819189057919243,
+      "learning_rate": 6.276679490471454e-06,
+      "loss": 0.4349,
+      "step": 11478
+    },
+    {
+      "epoch": 1.8851846529673804,
+      "grad_norm": 0.3256501015452309,
+      "learning_rate": 6.276248608162659e-06,
+      "loss": 0.4511,
+      "step": 11479
+    },
+    {
+      "epoch": 1.885348880175723,
+      "grad_norm": 0.3552499408145123,
+      "learning_rate": 6.275817706562058e-06,
+      "loss": 0.4528,
+      "step": 11480
+    },
+    {
+      "epoch": 1.8855131073840659,
+      "grad_norm": 0.2841417854307412,
+      "learning_rate": 6.275386785674329e-06,
+      "loss": 0.4485,
+      "step": 11481
+    },
+    {
+      "epoch": 1.8856773345924087,
+      "grad_norm": 0.31758879872950563,
+      "learning_rate": 6.274955845504154e-06,
+      "loss": 0.4472,
+      "step": 11482
+    },
+    {
+      "epoch": 1.8858415618007514,
+      "grad_norm": 0.43039374723598556,
+      "learning_rate": 6.2745248860562105e-06,
+      "loss": 0.4618,
+      "step": 11483
+    },
+    {
+      "epoch": 1.886005789009094,
+      "grad_norm": 0.3325004206289839,
+      "learning_rate": 6.274093907335181e-06,
+      "loss": 0.4484,
+      "step": 11484
+    },
+    {
+      "epoch": 1.8861700162174369,
+      "grad_norm": 0.3247843329476943,
+      "learning_rate": 6.273662909345747e-06,
+      "loss": 0.4326,
+      "step": 11485
+    },
+    {
+      "epoch": 1.8863342434257797,
+      "grad_norm": 0.3280630386694321,
+      "learning_rate": 6.273231892092589e-06,
+      "loss": 0.4396,
+      "step": 11486
+    },
+    {
+      "epoch": 1.8864984706341223,
+      "grad_norm": 0.3923324865040793,
+      "learning_rate": 6.272800855580388e-06,
+      "loss": 0.4365,
+      "step": 11487
+    },
+    {
+      "epoch": 1.886662697842465,
+      "grad_norm": 0.33057770575928486,
+      "learning_rate": 6.272369799813824e-06,
+      "loss": 0.443,
+      "step": 11488
+    },
+    {
+      "epoch": 1.8868269250508078,
+      "grad_norm": 0.7173285870423338,
+      "learning_rate": 6.271938724797581e-06,
+      "loss": 0.4398,
+      "step": 11489
+    },
+    {
+      "epoch": 1.8869911522591505,
+      "grad_norm": 0.4121293372702511,
+      "learning_rate": 6.27150763053634e-06,
+      "loss": 0.4651,
+      "step": 11490
+    },
+    {
+      "epoch": 1.8871553794674933,
+      "grad_norm": 0.29936010036350014,
+      "learning_rate": 6.271076517034784e-06,
+      "loss": 0.4499,
+      "step": 11491
+    },
+    {
+      "epoch": 1.887319606675836,
+      "grad_norm": 0.3528959255900349,
+      "learning_rate": 6.270645384297594e-06,
+      "loss": 0.4273,
+      "step": 11492
+    },
+    {
+      "epoch": 1.8874838338841786,
+      "grad_norm": 0.39731600061297867,
+      "learning_rate": 6.270214232329453e-06,
+      "loss": 0.4497,
+      "step": 11493
+    },
+    {
+      "epoch": 1.8876480610925215,
+      "grad_norm": 0.277693424780527,
+      "learning_rate": 6.269783061135044e-06,
+      "loss": 0.444,
+      "step": 11494
+    },
+    {
+      "epoch": 1.8878122883008643,
+      "grad_norm": 0.40844379179595136,
+      "learning_rate": 6.269351870719049e-06,
+      "loss": 0.4735,
+      "step": 11495
+    },
+    {
+      "epoch": 1.887976515509207,
+      "grad_norm": 0.31135073552625936,
+      "learning_rate": 6.268920661086153e-06,
+      "loss": 0.437,
+      "step": 11496
+    },
+    {
+      "epoch": 1.8881407427175496,
+      "grad_norm": 0.32411155738375347,
+      "learning_rate": 6.268489432241038e-06,
+      "loss": 0.4472,
+      "step": 11497
+    },
+    {
+      "epoch": 1.8883049699258925,
+      "grad_norm": 0.28558926169895243,
+      "learning_rate": 6.268058184188387e-06,
+      "loss": 0.4326,
+      "step": 11498
+    },
+    {
+      "epoch": 1.8884691971342353,
+      "grad_norm": 0.3449194860672656,
+      "learning_rate": 6.267626916932886e-06,
+      "loss": 0.4565,
+      "step": 11499
+    },
+    {
+      "epoch": 1.888633424342578,
+      "grad_norm": 0.34609846937748573,
+      "learning_rate": 6.267195630479215e-06,
+      "loss": 0.45,
+      "step": 11500
+    },
+    {
+      "epoch": 1.8887976515509206,
+      "grad_norm": 0.30123922815432147,
+      "learning_rate": 6.266764324832063e-06,
+      "loss": 0.4421,
+      "step": 11501
+    },
+    {
+      "epoch": 1.8889618787592635,
+      "grad_norm": 0.33185310581070543,
+      "learning_rate": 6.266332999996111e-06,
+      "loss": 0.4628,
+      "step": 11502
+    },
+    {
+      "epoch": 1.8891261059676063,
+      "grad_norm": 0.45745920406436913,
+      "learning_rate": 6.265901655976046e-06,
+      "loss": 0.4583,
+      "step": 11503
+    },
+    {
+      "epoch": 1.889290333175949,
+      "grad_norm": 0.2995880781951445,
+      "learning_rate": 6.265470292776551e-06,
+      "loss": 0.4449,
+      "step": 11504
+    },
+    {
+      "epoch": 1.8894545603842916,
+      "grad_norm": 0.45810579993366735,
+      "learning_rate": 6.265038910402311e-06,
+      "loss": 0.4526,
+      "step": 11505
+    },
+    {
+      "epoch": 1.8896187875926345,
+      "grad_norm": 0.28333685775743284,
+      "learning_rate": 6.264607508858013e-06,
+      "loss": 0.4512,
+      "step": 11506
+    },
+    {
+      "epoch": 1.889783014800977,
+      "grad_norm": 0.35274460640803634,
+      "learning_rate": 6.2641760881483415e-06,
+      "loss": 0.4415,
+      "step": 11507
+    },
+    {
+      "epoch": 1.88994724200932,
+      "grad_norm": 0.32639796842673624,
+      "learning_rate": 6.26374464827798e-06,
+      "loss": 0.4678,
+      "step": 11508
+    },
+    {
+      "epoch": 1.8901114692176626,
+      "grad_norm": 0.35721090968404956,
+      "learning_rate": 6.263313189251618e-06,
+      "loss": 0.4635,
+      "step": 11509
+    },
+    {
+      "epoch": 1.8902756964260052,
+      "grad_norm": 0.3324618961782957,
+      "learning_rate": 6.262881711073939e-06,
+      "loss": 0.4408,
+      "step": 11510
+    },
+    {
+      "epoch": 1.890439923634348,
+      "grad_norm": 0.3248140728701892,
+      "learning_rate": 6.262450213749631e-06,
+      "loss": 0.4484,
+      "step": 11511
+    },
+    {
+      "epoch": 1.890604150842691,
+      "grad_norm": 0.30240343332479325,
+      "learning_rate": 6.26201869728338e-06,
+      "loss": 0.4579,
+      "step": 11512
+    },
+    {
+      "epoch": 1.8907683780510336,
+      "grad_norm": 0.3734330252729453,
+      "learning_rate": 6.261587161679871e-06,
+      "loss": 0.4604,
+      "step": 11513
+    },
+    {
+      "epoch": 1.8909326052593762,
+      "grad_norm": 0.31340663492426546,
+      "learning_rate": 6.261155606943793e-06,
+      "loss": 0.4493,
+      "step": 11514
+    },
+    {
+      "epoch": 1.891096832467719,
+      "grad_norm": 0.3291571447089467,
+      "learning_rate": 6.260724033079832e-06,
+      "loss": 0.4614,
+      "step": 11515
+    },
+    {
+      "epoch": 1.891261059676062,
+      "grad_norm": 0.29024380784731046,
+      "learning_rate": 6.260292440092677e-06,
+      "loss": 0.4427,
+      "step": 11516
+    },
+    {
+      "epoch": 1.8914252868844046,
+      "grad_norm": 0.9818095952183078,
+      "learning_rate": 6.259860827987014e-06,
+      "loss": 0.4554,
+      "step": 11517
+    },
+    {
+      "epoch": 1.8915895140927472,
+      "grad_norm": 0.59070892702716,
+      "learning_rate": 6.259429196767529e-06,
+      "loss": 0.4225,
+      "step": 11518
+    },
+    {
+      "epoch": 1.89175374130109,
+      "grad_norm": 0.35026140088051805,
+      "learning_rate": 6.258997546438914e-06,
+      "loss": 0.4421,
+      "step": 11519
+    },
+    {
+      "epoch": 1.891917968509433,
+      "grad_norm": 0.455132303164521,
+      "learning_rate": 6.258565877005853e-06,
+      "loss": 0.4801,
+      "step": 11520
+    },
+    {
+      "epoch": 1.8920821957177756,
+      "grad_norm": 0.3062190096578685,
+      "learning_rate": 6.258134188473038e-06,
+      "loss": 0.4525,
+      "step": 11521
+    },
+    {
+      "epoch": 1.8922464229261182,
+      "grad_norm": 0.33415070443741085,
+      "learning_rate": 6.257702480845155e-06,
+      "loss": 0.4534,
+      "step": 11522
+    },
+    {
+      "epoch": 1.892410650134461,
+      "grad_norm": 0.30454079751989627,
+      "learning_rate": 6.257270754126895e-06,
+      "loss": 0.4444,
+      "step": 11523
+    },
+    {
+      "epoch": 1.8925748773428037,
+      "grad_norm": 0.28195699746422137,
+      "learning_rate": 6.256839008322944e-06,
+      "loss": 0.4761,
+      "step": 11524
+    },
+    {
+      "epoch": 1.8927391045511466,
+      "grad_norm": 0.32220012186937447,
+      "learning_rate": 6.256407243437993e-06,
+      "loss": 0.4656,
+      "step": 11525
+    },
+    {
+      "epoch": 1.8929033317594892,
+      "grad_norm": 0.29079011686059164,
+      "learning_rate": 6.255975459476733e-06,
+      "loss": 0.4481,
+      "step": 11526
+    },
+    {
+      "epoch": 1.8930675589678319,
+      "grad_norm": 0.27533179125606577,
+      "learning_rate": 6.25554365644385e-06,
+      "loss": 0.4347,
+      "step": 11527
+    },
+    {
+      "epoch": 1.8932317861761747,
+      "grad_norm": 0.2722721620887308,
+      "learning_rate": 6.255111834344037e-06,
+      "loss": 0.447,
+      "step": 11528
+    },
+    {
+      "epoch": 1.8933960133845176,
+      "grad_norm": 0.40678158182691243,
+      "learning_rate": 6.254679993181982e-06,
+      "loss": 0.4652,
+      "step": 11529
+    },
+    {
+      "epoch": 1.8935602405928602,
+      "grad_norm": 0.32142588418213813,
+      "learning_rate": 6.254248132962377e-06,
+      "loss": 0.4635,
+      "step": 11530
+    },
+    {
+      "epoch": 1.8937244678012028,
+      "grad_norm": 0.29323029764232855,
+      "learning_rate": 6.253816253689909e-06,
+      "loss": 0.4447,
+      "step": 11531
+    },
+    {
+      "epoch": 1.8938886950095457,
+      "grad_norm": 0.29601891468051994,
+      "learning_rate": 6.2533843553692736e-06,
+      "loss": 0.4559,
+      "step": 11532
+    },
+    {
+      "epoch": 1.8940529222178886,
+      "grad_norm": 0.286254401704754,
+      "learning_rate": 6.252952438005157e-06,
+      "loss": 0.4337,
+      "step": 11533
+    },
+    {
+      "epoch": 1.8942171494262312,
+      "grad_norm": 0.3827539442976956,
+      "learning_rate": 6.252520501602252e-06,
+      "loss": 0.4571,
+      "step": 11534
+    },
+    {
+      "epoch": 1.8943813766345738,
+      "grad_norm": 0.2871685184088086,
+      "learning_rate": 6.2520885461652515e-06,
+      "loss": 0.4257,
+      "step": 11535
+    },
+    {
+      "epoch": 1.8945456038429167,
+      "grad_norm": 0.8404977702623102,
+      "learning_rate": 6.251656571698846e-06,
+      "loss": 0.4478,
+      "step": 11536
+    },
+    {
+      "epoch": 1.8947098310512596,
+      "grad_norm": 0.32025923391449485,
+      "learning_rate": 6.251224578207725e-06,
+      "loss": 0.4484,
+      "step": 11537
+    },
+    {
+      "epoch": 1.8948740582596022,
+      "grad_norm": 0.286052981589957,
+      "learning_rate": 6.2507925656965825e-06,
+      "loss": 0.4484,
+      "step": 11538
+    },
+    {
+      "epoch": 1.8950382854679448,
+      "grad_norm": 0.2856432354318945,
+      "learning_rate": 6.25036053417011e-06,
+      "loss": 0.441,
+      "step": 11539
+    },
+    {
+      "epoch": 1.8952025126762877,
+      "grad_norm": 0.3021535240566939,
+      "learning_rate": 6.2499284836330014e-06,
+      "loss": 0.4523,
+      "step": 11540
+    },
+    {
+      "epoch": 1.8953667398846303,
+      "grad_norm": 0.4005247080570124,
+      "learning_rate": 6.249496414089948e-06,
+      "loss": 0.463,
+      "step": 11541
+    },
+    {
+      "epoch": 1.8955309670929732,
+      "grad_norm": 0.2994150789421531,
+      "learning_rate": 6.24906432554564e-06,
+      "loss": 0.4448,
+      "step": 11542
+    },
+    {
+      "epoch": 1.8956951943013158,
+      "grad_norm": 0.3034711094777697,
+      "learning_rate": 6.248632218004773e-06,
+      "loss": 0.4422,
+      "step": 11543
+    },
+    {
+      "epoch": 1.8958594215096585,
+      "grad_norm": 0.3241176095587336,
+      "learning_rate": 6.248200091472042e-06,
+      "loss": 0.4704,
+      "step": 11544
+    },
+    {
+      "epoch": 1.8960236487180013,
+      "grad_norm": 0.31980159348254694,
+      "learning_rate": 6.247767945952138e-06,
+      "loss": 0.4389,
+      "step": 11545
+    },
+    {
+      "epoch": 1.8961878759263442,
+      "grad_norm": 0.37791081930961246,
+      "learning_rate": 6.247335781449751e-06,
+      "loss": 0.4415,
+      "step": 11546
+    },
+    {
+      "epoch": 1.8963521031346868,
+      "grad_norm": 0.271887232267887,
+      "learning_rate": 6.2469035979695805e-06,
+      "loss": 0.449,
+      "step": 11547
+    },
+    {
+      "epoch": 1.8965163303430295,
+      "grad_norm": 0.33458075064495246,
+      "learning_rate": 6.246471395516319e-06,
+      "loss": 0.4447,
+      "step": 11548
+    },
+    {
+      "epoch": 1.8966805575513723,
+      "grad_norm": 0.8032382841821906,
+      "learning_rate": 6.2460391740946585e-06,
+      "loss": 0.4596,
+      "step": 11549
+    },
+    {
+      "epoch": 1.8968447847597152,
+      "grad_norm": 0.39747951496956996,
+      "learning_rate": 6.245606933709296e-06,
+      "loss": 0.4669,
+      "step": 11550
+    },
+    {
+      "epoch": 1.8970090119680578,
+      "grad_norm": 0.29049470000836963,
+      "learning_rate": 6.245174674364923e-06,
+      "loss": 0.4586,
+      "step": 11551
+    },
+    {
+      "epoch": 1.8971732391764005,
+      "grad_norm": 0.3086255776162428,
+      "learning_rate": 6.244742396066237e-06,
+      "loss": 0.4448,
+      "step": 11552
+    },
+    {
+      "epoch": 1.8973374663847433,
+      "grad_norm": 0.3656060035316904,
+      "learning_rate": 6.244310098817933e-06,
+      "loss": 0.4494,
+      "step": 11553
+    },
+    {
+      "epoch": 1.8975016935930862,
+      "grad_norm": 0.2870249267218026,
+      "learning_rate": 6.243877782624703e-06,
+      "loss": 0.4474,
+      "step": 11554
+    },
+    {
+      "epoch": 1.8976659208014288,
+      "grad_norm": 0.39457473479831795,
+      "learning_rate": 6.243445447491246e-06,
+      "loss": 0.4487,
+      "step": 11555
+    },
+    {
+      "epoch": 1.8978301480097715,
+      "grad_norm": 0.35379457137991177,
+      "learning_rate": 6.2430130934222545e-06,
+      "loss": 0.4588,
+      "step": 11556
+    },
+    {
+      "epoch": 1.8979943752181143,
+      "grad_norm": 0.39416032087160907,
+      "learning_rate": 6.242580720422428e-06,
+      "loss": 0.4667,
+      "step": 11557
+    },
+    {
+      "epoch": 1.898158602426457,
+      "grad_norm": 0.3312029003928857,
+      "learning_rate": 6.242148328496459e-06,
+      "loss": 0.4657,
+      "step": 11558
+    },
+    {
+      "epoch": 1.8983228296347998,
+      "grad_norm": 0.3977817496035618,
+      "learning_rate": 6.241715917649046e-06,
+      "loss": 0.4449,
+      "step": 11559
+    },
+    {
+      "epoch": 1.8984870568431425,
+      "grad_norm": 0.2861507297462149,
+      "learning_rate": 6.241283487884884e-06,
+      "loss": 0.4449,
+      "step": 11560
+    },
+    {
+      "epoch": 1.898651284051485,
+      "grad_norm": 0.2867583166903834,
+      "learning_rate": 6.2408510392086714e-06,
+      "loss": 0.4653,
+      "step": 11561
+    },
+    {
+      "epoch": 1.898815511259828,
+      "grad_norm": 0.3138206192945438,
+      "learning_rate": 6.240418571625102e-06,
+      "loss": 0.4344,
+      "step": 11562
+    },
+    {
+      "epoch": 1.8989797384681708,
+      "grad_norm": 0.33426289559226574,
+      "learning_rate": 6.239986085138875e-06,
+      "loss": 0.4586,
+      "step": 11563
+    },
+    {
+      "epoch": 1.8991439656765134,
+      "grad_norm": 0.3225829509226065,
+      "learning_rate": 6.2395535797546875e-06,
+      "loss": 0.4533,
+      "step": 11564
+    },
+    {
+      "epoch": 1.899308192884856,
+      "grad_norm": 0.32700753558233936,
+      "learning_rate": 6.239121055477237e-06,
+      "loss": 0.4444,
+      "step": 11565
+    },
+    {
+      "epoch": 1.899472420093199,
+      "grad_norm": 0.7395817042078737,
+      "learning_rate": 6.238688512311219e-06,
+      "loss": 0.4559,
+      "step": 11566
+    },
+    {
+      "epoch": 1.8996366473015418,
+      "grad_norm": 0.3176473441311553,
+      "learning_rate": 6.238255950261335e-06,
+      "loss": 0.4468,
+      "step": 11567
+    },
+    {
+      "epoch": 1.8998008745098844,
+      "grad_norm": 0.3308084452657348,
+      "learning_rate": 6.23782336933228e-06,
+      "loss": 0.4732,
+      "step": 11568
+    },
+    {
+      "epoch": 1.899965101718227,
+      "grad_norm": 0.3424847343425013,
+      "learning_rate": 6.237390769528754e-06,
+      "loss": 0.459,
+      "step": 11569
+    },
+    {
+      "epoch": 1.90012932892657,
+      "grad_norm": 0.3329594799384342,
+      "learning_rate": 6.236958150855456e-06,
+      "loss": 0.4605,
+      "step": 11570
+    },
+    {
+      "epoch": 1.9002935561349128,
+      "grad_norm": 0.5164135512378328,
+      "learning_rate": 6.236525513317083e-06,
+      "loss": 0.4517,
+      "step": 11571
+    },
+    {
+      "epoch": 1.9004577833432554,
+      "grad_norm": 0.2962729195900784,
+      "learning_rate": 6.236092856918333e-06,
+      "loss": 0.454,
+      "step": 11572
+    },
+    {
+      "epoch": 1.900622010551598,
+      "grad_norm": 0.4457999560545923,
+      "learning_rate": 6.235660181663906e-06,
+      "loss": 0.4499,
+      "step": 11573
+    },
+    {
+      "epoch": 1.900786237759941,
+      "grad_norm": 0.29085184015101817,
+      "learning_rate": 6.235227487558504e-06,
+      "loss": 0.4487,
+      "step": 11574
+    },
+    {
+      "epoch": 1.9009504649682836,
+      "grad_norm": 0.2955717712662332,
+      "learning_rate": 6.2347947746068245e-06,
+      "loss": 0.4578,
+      "step": 11575
+    },
+    {
+      "epoch": 1.9011146921766264,
+      "grad_norm": 0.3152314118744751,
+      "learning_rate": 6.234362042813565e-06,
+      "loss": 0.4325,
+      "step": 11576
+    },
+    {
+      "epoch": 1.901278919384969,
+      "grad_norm": 0.26407825448367234,
+      "learning_rate": 6.233929292183427e-06,
+      "loss": 0.4503,
+      "step": 11577
+    },
+    {
+      "epoch": 1.9014431465933117,
+      "grad_norm": 0.3915379290000229,
+      "learning_rate": 6.233496522721113e-06,
+      "loss": 0.4504,
+      "step": 11578
+    },
+    {
+      "epoch": 1.9016073738016546,
+      "grad_norm": 0.3412610510019888,
+      "learning_rate": 6.233063734431321e-06,
+      "loss": 0.4463,
+      "step": 11579
+    },
+    {
+      "epoch": 1.9017716010099974,
+      "grad_norm": 0.32050195639366946,
+      "learning_rate": 6.23263092731875e-06,
+      "loss": 0.4497,
+      "step": 11580
+    },
+    {
+      "epoch": 1.90193582821834,
+      "grad_norm": 0.38273878233161573,
+      "learning_rate": 6.232198101388104e-06,
+      "loss": 0.4403,
+      "step": 11581
+    },
+    {
+      "epoch": 1.9021000554266827,
+      "grad_norm": 0.27806982974287925,
+      "learning_rate": 6.2317652566440825e-06,
+      "loss": 0.4451,
+      "step": 11582
+    },
+    {
+      "epoch": 1.9022642826350256,
+      "grad_norm": 0.32856501937660954,
+      "learning_rate": 6.231332393091385e-06,
+      "loss": 0.4551,
+      "step": 11583
+    },
+    {
+      "epoch": 1.9024285098433684,
+      "grad_norm": 0.2682827626213739,
+      "learning_rate": 6.230899510734716e-06,
+      "loss": 0.44,
+      "step": 11584
+    },
+    {
+      "epoch": 1.902592737051711,
+      "grad_norm": 0.49272255538438214,
+      "learning_rate": 6.230466609578773e-06,
+      "loss": 0.4414,
+      "step": 11585
+    },
+    {
+      "epoch": 1.9027569642600537,
+      "grad_norm": 0.32260483225705006,
+      "learning_rate": 6.230033689628262e-06,
+      "loss": 0.4271,
+      "step": 11586
+    },
+    {
+      "epoch": 1.9029211914683966,
+      "grad_norm": 0.29093147267619285,
+      "learning_rate": 6.229600750887883e-06,
+      "loss": 0.4579,
+      "step": 11587
+    },
+    {
+      "epoch": 1.9030854186767394,
+      "grad_norm": 0.35265332639510805,
+      "learning_rate": 6.229167793362337e-06,
+      "loss": 0.4413,
+      "step": 11588
+    },
+    {
+      "epoch": 1.903249645885082,
+      "grad_norm": 0.3142574353716658,
+      "learning_rate": 6.228734817056328e-06,
+      "loss": 0.4415,
+      "step": 11589
+    },
+    {
+      "epoch": 1.9034138730934247,
+      "grad_norm": 0.5622430879448479,
+      "learning_rate": 6.228301821974559e-06,
+      "loss": 0.4357,
+      "step": 11590
+    },
+    {
+      "epoch": 1.9035781003017676,
+      "grad_norm": 0.3871498522685965,
+      "learning_rate": 6.227868808121731e-06,
+      "loss": 0.4503,
+      "step": 11591
+    },
+    {
+      "epoch": 1.9037423275101102,
+      "grad_norm": 0.3141566277760273,
+      "learning_rate": 6.227435775502547e-06,
+      "loss": 0.4639,
+      "step": 11592
+    },
+    {
+      "epoch": 1.903906554718453,
+      "grad_norm": 0.32246353834594804,
+      "learning_rate": 6.227002724121711e-06,
+      "loss": 0.4344,
+      "step": 11593
+    },
+    {
+      "epoch": 1.9040707819267957,
+      "grad_norm": 0.29801343541053976,
+      "learning_rate": 6.226569653983929e-06,
+      "loss": 0.4561,
+      "step": 11594
+    },
+    {
+      "epoch": 1.9042350091351383,
+      "grad_norm": 0.33855218720060254,
+      "learning_rate": 6.2261365650939e-06,
+      "loss": 0.445,
+      "step": 11595
+    },
+    {
+      "epoch": 1.9043992363434812,
+      "grad_norm": 0.3087953002873901,
+      "learning_rate": 6.2257034574563285e-06,
+      "loss": 0.4356,
+      "step": 11596
+    },
+    {
+      "epoch": 1.904563463551824,
+      "grad_norm": 0.448227658275254,
+      "learning_rate": 6.225270331075921e-06,
+      "loss": 0.4669,
+      "step": 11597
+    },
+    {
+      "epoch": 1.9047276907601667,
+      "grad_norm": 0.325155177303091,
+      "learning_rate": 6.22483718595738e-06,
+      "loss": 0.4615,
+      "step": 11598
+    },
+    {
+      "epoch": 1.9048919179685093,
+      "grad_norm": 0.4675952573516096,
+      "learning_rate": 6.2244040221054095e-06,
+      "loss": 0.4613,
+      "step": 11599
+    },
+    {
+      "epoch": 1.9050561451768522,
+      "grad_norm": 0.32235353184835497,
+      "learning_rate": 6.223970839524715e-06,
+      "loss": 0.4601,
+      "step": 11600
+    },
+    {
+      "epoch": 1.905220372385195,
+      "grad_norm": 0.33233067198397287,
+      "learning_rate": 6.223537638220001e-06,
+      "loss": 0.4715,
+      "step": 11601
+    },
+    {
+      "epoch": 1.9053845995935377,
+      "grad_norm": 0.2982512809301434,
+      "learning_rate": 6.223104418195972e-06,
+      "loss": 0.4555,
+      "step": 11602
+    },
+    {
+      "epoch": 1.9055488268018803,
+      "grad_norm": 0.3801442216186049,
+      "learning_rate": 6.2226711794573354e-06,
+      "loss": 0.4318,
+      "step": 11603
+    },
+    {
+      "epoch": 1.9057130540102232,
+      "grad_norm": 0.2957424808790963,
+      "learning_rate": 6.222237922008795e-06,
+      "loss": 0.4574,
+      "step": 11604
+    },
+    {
+      "epoch": 1.905877281218566,
+      "grad_norm": 0.3087230953401648,
+      "learning_rate": 6.221804645855054e-06,
+      "loss": 0.4411,
+      "step": 11605
+    },
+    {
+      "epoch": 1.9060415084269087,
+      "grad_norm": 0.3542054566112007,
+      "learning_rate": 6.221371351000822e-06,
+      "loss": 0.4476,
+      "step": 11606
+    },
+    {
+      "epoch": 1.9062057356352513,
+      "grad_norm": 0.3553897290376785,
+      "learning_rate": 6.2209380374508035e-06,
+      "loss": 0.4707,
+      "step": 11607
+    },
+    {
+      "epoch": 1.906369962843594,
+      "grad_norm": 0.3564058255833194,
+      "learning_rate": 6.220504705209705e-06,
+      "loss": 0.4495,
+      "step": 11608
+    },
+    {
+      "epoch": 1.9065341900519368,
+      "grad_norm": 0.3832925455459929,
+      "learning_rate": 6.220071354282232e-06,
+      "loss": 0.4625,
+      "step": 11609
+    },
+    {
+      "epoch": 1.9066984172602797,
+      "grad_norm": 0.2817875411638949,
+      "learning_rate": 6.219637984673092e-06,
+      "loss": 0.4453,
+      "step": 11610
+    },
+    {
+      "epoch": 1.9068626444686223,
+      "grad_norm": 0.4776547443948289,
+      "learning_rate": 6.219204596386991e-06,
+      "loss": 0.4558,
+      "step": 11611
+    },
+    {
+      "epoch": 1.907026871676965,
+      "grad_norm": 0.34962942789150825,
+      "learning_rate": 6.218771189428637e-06,
+      "loss": 0.455,
+      "step": 11612
+    },
+    {
+      "epoch": 1.9071910988853078,
+      "grad_norm": 0.30962434569385655,
+      "learning_rate": 6.218337763802738e-06,
+      "loss": 0.4462,
+      "step": 11613
+    },
+    {
+      "epoch": 1.9073553260936507,
+      "grad_norm": 0.3391862780256781,
+      "learning_rate": 6.2179043195139985e-06,
+      "loss": 0.4351,
+      "step": 11614
+    },
+    {
+      "epoch": 1.9075195533019933,
+      "grad_norm": 0.30719116774597893,
+      "learning_rate": 6.2174708565671296e-06,
+      "loss": 0.4492,
+      "step": 11615
+    },
+    {
+      "epoch": 1.907683780510336,
+      "grad_norm": 0.30054149364958066,
+      "learning_rate": 6.217037374966836e-06,
+      "loss": 0.4628,
+      "step": 11616
+    },
+    {
+      "epoch": 1.9078480077186788,
+      "grad_norm": 0.386653014036295,
+      "learning_rate": 6.216603874717828e-06,
+      "loss": 0.4479,
+      "step": 11617
+    },
+    {
+      "epoch": 1.9080122349270217,
+      "grad_norm": 0.31127433847704317,
+      "learning_rate": 6.216170355824812e-06,
+      "loss": 0.4377,
+      "step": 11618
+    },
+    {
+      "epoch": 1.9081764621353643,
+      "grad_norm": 0.3605960243523264,
+      "learning_rate": 6.215736818292499e-06,
+      "loss": 0.459,
+      "step": 11619
+    },
+    {
+      "epoch": 1.908340689343707,
+      "grad_norm": 0.33757270101982667,
+      "learning_rate": 6.215303262125595e-06,
+      "loss": 0.4627,
+      "step": 11620
+    },
+    {
+      "epoch": 1.9085049165520498,
+      "grad_norm": 0.38173723225560446,
+      "learning_rate": 6.21486968732881e-06,
+      "loss": 0.4569,
+      "step": 11621
+    },
+    {
+      "epoch": 1.9086691437603927,
+      "grad_norm": 0.32491942608170243,
+      "learning_rate": 6.214436093906852e-06,
+      "loss": 0.4513,
+      "step": 11622
+    },
+    {
+      "epoch": 1.9088333709687353,
+      "grad_norm": 0.38253325439429253,
+      "learning_rate": 6.214002481864434e-06,
+      "loss": 0.429,
+      "step": 11623
+    },
+    {
+      "epoch": 1.908997598177078,
+      "grad_norm": 0.5169229832581916,
+      "learning_rate": 6.213568851206261e-06,
+      "loss": 0.4398,
+      "step": 11624
+    },
+    {
+      "epoch": 1.9091618253854206,
+      "grad_norm": 0.36052604572634506,
+      "learning_rate": 6.2131352019370446e-06,
+      "loss": 0.4403,
+      "step": 11625
+    },
+    {
+      "epoch": 1.9093260525937634,
+      "grad_norm": 0.2924912097276931,
+      "learning_rate": 6.212701534061493e-06,
+      "loss": 0.4611,
+      "step": 11626
+    },
+    {
+      "epoch": 1.9094902798021063,
+      "grad_norm": 0.3404899101670224,
+      "learning_rate": 6.212267847584319e-06,
+      "loss": 0.4447,
+      "step": 11627
+    },
+    {
+      "epoch": 1.909654507010449,
+      "grad_norm": 0.36381309075662677,
+      "learning_rate": 6.211834142510232e-06,
+      "loss": 0.4577,
+      "step": 11628
+    },
+    {
+      "epoch": 1.9098187342187916,
+      "grad_norm": 0.38424637201933903,
+      "learning_rate": 6.211400418843942e-06,
+      "loss": 0.4495,
+      "step": 11629
+    },
+    {
+      "epoch": 1.9099829614271344,
+      "grad_norm": 0.9281977678260263,
+      "learning_rate": 6.2109666765901585e-06,
+      "loss": 0.4659,
+      "step": 11630
+    },
+    {
+      "epoch": 1.9101471886354773,
+      "grad_norm": 0.35161232363455713,
+      "learning_rate": 6.2105329157535935e-06,
+      "loss": 0.455,
+      "step": 11631
+    },
+    {
+      "epoch": 1.91031141584382,
+      "grad_norm": 1.1585271182741959,
+      "learning_rate": 6.21009913633896e-06,
+      "loss": 0.4312,
+      "step": 11632
+    },
+    {
+      "epoch": 1.9104756430521626,
+      "grad_norm": 0.3674122396524195,
+      "learning_rate": 6.209665338350967e-06,
+      "loss": 0.4796,
+      "step": 11633
+    },
+    {
+      "epoch": 1.9106398702605054,
+      "grad_norm": 0.36718309874128846,
+      "learning_rate": 6.209231521794324e-06,
+      "loss": 0.4304,
+      "step": 11634
+    },
+    {
+      "epoch": 1.9108040974688483,
+      "grad_norm": 0.6703568626303426,
+      "learning_rate": 6.208797686673746e-06,
+      "loss": 0.4637,
+      "step": 11635
+    },
+    {
+      "epoch": 1.910968324677191,
+      "grad_norm": 0.40632340164699454,
+      "learning_rate": 6.2083638329939455e-06,
+      "loss": 0.46,
+      "step": 11636
+    },
+    {
+      "epoch": 1.9111325518855335,
+      "grad_norm": 0.35780222483159585,
+      "learning_rate": 6.207929960759631e-06,
+      "loss": 0.4555,
+      "step": 11637
+    },
+    {
+      "epoch": 1.9112967790938764,
+      "grad_norm": 0.34875469521533164,
+      "learning_rate": 6.207496069975519e-06,
+      "loss": 0.4445,
+      "step": 11638
+    },
+    {
+      "epoch": 1.9114610063022193,
+      "grad_norm": 0.37116131198879804,
+      "learning_rate": 6.207062160646318e-06,
+      "loss": 0.4627,
+      "step": 11639
+    },
+    {
+      "epoch": 1.911625233510562,
+      "grad_norm": 0.3261872497806788,
+      "learning_rate": 6.206628232776743e-06,
+      "loss": 0.436,
+      "step": 11640
+    },
+    {
+      "epoch": 1.9117894607189045,
+      "grad_norm": 0.2917346097333719,
+      "learning_rate": 6.206194286371505e-06,
+      "loss": 0.438,
+      "step": 11641
+    },
+    {
+      "epoch": 1.9119536879272472,
+      "grad_norm": 0.318715515401316,
+      "learning_rate": 6.205760321435319e-06,
+      "loss": 0.4716,
+      "step": 11642
+    },
+    {
+      "epoch": 1.91211791513559,
+      "grad_norm": 0.3822885736630288,
+      "learning_rate": 6.205326337972899e-06,
+      "loss": 0.4612,
+      "step": 11643
+    },
+    {
+      "epoch": 1.912282142343933,
+      "grad_norm": 0.39512948614808435,
+      "learning_rate": 6.204892335988956e-06,
+      "loss": 0.4546,
+      "step": 11644
+    },
+    {
+      "epoch": 1.9124463695522755,
+      "grad_norm": 0.2906356956083419,
+      "learning_rate": 6.204458315488205e-06,
+      "loss": 0.4323,
+      "step": 11645
+    },
+    {
+      "epoch": 1.9126105967606182,
+      "grad_norm": 0.3151679721101704,
+      "learning_rate": 6.204024276475361e-06,
+      "loss": 0.4274,
+      "step": 11646
+    },
+    {
+      "epoch": 1.912774823968961,
+      "grad_norm": 0.280920093275785,
+      "learning_rate": 6.203590218955136e-06,
+      "loss": 0.4345,
+      "step": 11647
+    },
+    {
+      "epoch": 1.912939051177304,
+      "grad_norm": 0.3177305722530813,
+      "learning_rate": 6.203156142932243e-06,
+      "loss": 0.4561,
+      "step": 11648
+    },
+    {
+      "epoch": 1.9131032783856465,
+      "grad_norm": 0.35949250252352477,
+      "learning_rate": 6.202722048411402e-06,
+      "loss": 0.4366,
+      "step": 11649
+    },
+    {
+      "epoch": 1.9132675055939892,
+      "grad_norm": 0.37584685311355087,
+      "learning_rate": 6.202287935397321e-06,
+      "loss": 0.4478,
+      "step": 11650
+    },
+    {
+      "epoch": 1.913431732802332,
+      "grad_norm": 0.33655101627969175,
+      "learning_rate": 6.20185380389472e-06,
+      "loss": 0.4548,
+      "step": 11651
+    },
+    {
+      "epoch": 1.913595960010675,
+      "grad_norm": 0.30964429163933455,
+      "learning_rate": 6.201419653908313e-06,
+      "loss": 0.4538,
+      "step": 11652
+    },
+    {
+      "epoch": 1.9137601872190175,
+      "grad_norm": 0.29372286656988195,
+      "learning_rate": 6.200985485442815e-06,
+      "loss": 0.4318,
+      "step": 11653
+    },
+    {
+      "epoch": 1.9139244144273602,
+      "grad_norm": 1.6537874331386198,
+      "learning_rate": 6.200551298502939e-06,
+      "loss": 0.4327,
+      "step": 11654
+    },
+    {
+      "epoch": 1.914088641635703,
+      "grad_norm": 0.414880176031993,
+      "learning_rate": 6.2001170930934025e-06,
+      "loss": 0.4639,
+      "step": 11655
+    },
+    {
+      "epoch": 1.9142528688440459,
+      "grad_norm": 0.381572536194753,
+      "learning_rate": 6.199682869218922e-06,
+      "loss": 0.454,
+      "step": 11656
+    },
+    {
+      "epoch": 1.9144170960523885,
+      "grad_norm": 0.3618841751303404,
+      "learning_rate": 6.199248626884215e-06,
+      "loss": 0.4415,
+      "step": 11657
+    },
+    {
+      "epoch": 1.9145813232607312,
+      "grad_norm": 0.3091815517830806,
+      "learning_rate": 6.198814366093996e-06,
+      "loss": 0.4635,
+      "step": 11658
+    },
+    {
+      "epoch": 1.9147455504690738,
+      "grad_norm": 0.3461588940465848,
+      "learning_rate": 6.198380086852981e-06,
+      "loss": 0.4545,
+      "step": 11659
+    },
+    {
+      "epoch": 1.9149097776774167,
+      "grad_norm": 0.32914313950790325,
+      "learning_rate": 6.197945789165885e-06,
+      "loss": 0.4288,
+      "step": 11660
+    },
+    {
+      "epoch": 1.9150740048857595,
+      "grad_norm": 0.30501120382983676,
+      "learning_rate": 6.197511473037431e-06,
+      "loss": 0.4331,
+      "step": 11661
+    },
+    {
+      "epoch": 1.9152382320941022,
+      "grad_norm": 0.31219879028545594,
+      "learning_rate": 6.19707713847233e-06,
+      "loss": 0.4464,
+      "step": 11662
+    },
+    {
+      "epoch": 1.9154024593024448,
+      "grad_norm": 0.31460550974100093,
+      "learning_rate": 6.196642785475302e-06,
+      "loss": 0.4562,
+      "step": 11663
+    },
+    {
+      "epoch": 1.9155666865107877,
+      "grad_norm": 0.36270468986808635,
+      "learning_rate": 6.196208414051064e-06,
+      "loss": 0.4501,
+      "step": 11664
+    },
+    {
+      "epoch": 1.9157309137191305,
+      "grad_norm": 0.4112853774633795,
+      "learning_rate": 6.195774024204334e-06,
+      "loss": 0.4611,
+      "step": 11665
+    },
+    {
+      "epoch": 1.9158951409274732,
+      "grad_norm": 0.33448305601115735,
+      "learning_rate": 6.19533961593983e-06,
+      "loss": 0.4738,
+      "step": 11666
+    },
+    {
+      "epoch": 1.9160593681358158,
+      "grad_norm": 0.5320652524792794,
+      "learning_rate": 6.194905189262269e-06,
+      "loss": 0.4803,
+      "step": 11667
+    },
+    {
+      "epoch": 1.9162235953441586,
+      "grad_norm": 0.3020273231113173,
+      "learning_rate": 6.19447074417637e-06,
+      "loss": 0.4522,
+      "step": 11668
+    },
+    {
+      "epoch": 1.9163878225525015,
+      "grad_norm": 0.3270778515804615,
+      "learning_rate": 6.194036280686851e-06,
+      "loss": 0.4652,
+      "step": 11669
+    },
+    {
+      "epoch": 1.9165520497608441,
+      "grad_norm": 0.30776326745206567,
+      "learning_rate": 6.193601798798435e-06,
+      "loss": 0.4489,
+      "step": 11670
+    },
+    {
+      "epoch": 1.9167162769691868,
+      "grad_norm": 0.30728786187995366,
+      "learning_rate": 6.193167298515833e-06,
+      "loss": 0.4427,
+      "step": 11671
+    },
+    {
+      "epoch": 1.9168805041775296,
+      "grad_norm": 0.3927785447223489,
+      "learning_rate": 6.192732779843771e-06,
+      "loss": 0.4553,
+      "step": 11672
+    },
+    {
+      "epoch": 1.9170447313858725,
+      "grad_norm": 0.29282886384148993,
+      "learning_rate": 6.192298242786963e-06,
+      "loss": 0.4499,
+      "step": 11673
+    },
+    {
+      "epoch": 1.9172089585942151,
+      "grad_norm": 0.31577968887587804,
+      "learning_rate": 6.191863687350133e-06,
+      "loss": 0.4792,
+      "step": 11674
+    },
+    {
+      "epoch": 1.9173731858025578,
+      "grad_norm": 0.3761942499553469,
+      "learning_rate": 6.191429113537998e-06,
+      "loss": 0.4498,
+      "step": 11675
+    },
+    {
+      "epoch": 1.9175374130109004,
+      "grad_norm": 0.317601854958605,
+      "learning_rate": 6.190994521355279e-06,
+      "loss": 0.4429,
+      "step": 11676
+    },
+    {
+      "epoch": 1.9177016402192433,
+      "grad_norm": 0.3517208242879503,
+      "learning_rate": 6.190559910806696e-06,
+      "loss": 0.4644,
+      "step": 11677
+    },
+    {
+      "epoch": 1.9178658674275861,
+      "grad_norm": 0.3241901907564984,
+      "learning_rate": 6.190125281896969e-06,
+      "loss": 0.456,
+      "step": 11678
+    },
+    {
+      "epoch": 1.9180300946359288,
+      "grad_norm": 0.3025139517641792,
+      "learning_rate": 6.189690634630818e-06,
+      "loss": 0.4564,
+      "step": 11679
+    },
+    {
+      "epoch": 1.9181943218442714,
+      "grad_norm": 0.35524817549149657,
+      "learning_rate": 6.189255969012965e-06,
+      "loss": 0.4399,
+      "step": 11680
+    },
+    {
+      "epoch": 1.9183585490526143,
+      "grad_norm": 0.4244953064809638,
+      "learning_rate": 6.18882128504813e-06,
+      "loss": 0.4292,
+      "step": 11681
+    },
+    {
+      "epoch": 1.9185227762609571,
+      "grad_norm": 0.30648349992963597,
+      "learning_rate": 6.188386582741034e-06,
+      "loss": 0.4547,
+      "step": 11682
+    },
+    {
+      "epoch": 1.9186870034692998,
+      "grad_norm": 0.30163722557002615,
+      "learning_rate": 6.187951862096398e-06,
+      "loss": 0.4584,
+      "step": 11683
+    },
+    {
+      "epoch": 1.9188512306776424,
+      "grad_norm": 0.33493401319101296,
+      "learning_rate": 6.187517123118945e-06,
+      "loss": 0.4509,
+      "step": 11684
+    },
+    {
+      "epoch": 1.9190154578859853,
+      "grad_norm": 0.313911034591551,
+      "learning_rate": 6.187082365813395e-06,
+      "loss": 0.4539,
+      "step": 11685
+    },
+    {
+      "epoch": 1.9191796850943281,
+      "grad_norm": 0.5202964654233119,
+      "learning_rate": 6.186647590184471e-06,
+      "loss": 0.4664,
+      "step": 11686
+    },
+    {
+      "epoch": 1.9193439123026708,
+      "grad_norm": 0.2911861487320873,
+      "learning_rate": 6.186212796236896e-06,
+      "loss": 0.4493,
+      "step": 11687
+    },
+    {
+      "epoch": 1.9195081395110134,
+      "grad_norm": 0.3894019111871137,
+      "learning_rate": 6.185777983975389e-06,
+      "loss": 0.4719,
+      "step": 11688
+    },
+    {
+      "epoch": 1.9196723667193563,
+      "grad_norm": 0.3263915765382676,
+      "learning_rate": 6.185343153404675e-06,
+      "loss": 0.4603,
+      "step": 11689
+    },
+    {
+      "epoch": 1.9198365939276991,
+      "grad_norm": 0.2812099109496593,
+      "learning_rate": 6.184908304529477e-06,
+      "loss": 0.4397,
+      "step": 11690
+    },
+    {
+      "epoch": 1.9200008211360418,
+      "grad_norm": 0.3316284604588048,
+      "learning_rate": 6.184473437354517e-06,
+      "loss": 0.4634,
+      "step": 11691
+    },
+    {
+      "epoch": 1.9201650483443844,
+      "grad_norm": 0.3214264446614346,
+      "learning_rate": 6.184038551884518e-06,
+      "loss": 0.4567,
+      "step": 11692
+    },
+    {
+      "epoch": 1.920329275552727,
+      "grad_norm": 0.29348354172249397,
+      "learning_rate": 6.183603648124203e-06,
+      "loss": 0.4354,
+      "step": 11693
+    },
+    {
+      "epoch": 1.92049350276107,
+      "grad_norm": 0.42769052741256725,
+      "learning_rate": 6.183168726078295e-06,
+      "loss": 0.4522,
+      "step": 11694
+    },
+    {
+      "epoch": 1.9206577299694128,
+      "grad_norm": 0.3172711368471013,
+      "learning_rate": 6.182733785751521e-06,
+      "loss": 0.4598,
+      "step": 11695
+    },
+    {
+      "epoch": 1.9208219571777554,
+      "grad_norm": 0.3141973931650255,
+      "learning_rate": 6.182298827148602e-06,
+      "loss": 0.4464,
+      "step": 11696
+    },
+    {
+      "epoch": 1.920986184386098,
+      "grad_norm": 0.3561540020725479,
+      "learning_rate": 6.181863850274262e-06,
+      "loss": 0.4455,
+      "step": 11697
+    },
+    {
+      "epoch": 1.9211504115944409,
+      "grad_norm": 0.32393565694758736,
+      "learning_rate": 6.181428855133225e-06,
+      "loss": 0.4564,
+      "step": 11698
+    },
+    {
+      "epoch": 1.9213146388027837,
+      "grad_norm": 0.298550373846303,
+      "learning_rate": 6.1809938417302176e-06,
+      "loss": 0.4506,
+      "step": 11699
+    },
+    {
+      "epoch": 1.9214788660111264,
+      "grad_norm": 0.3943979027923778,
+      "learning_rate": 6.180558810069962e-06,
+      "loss": 0.4397,
+      "step": 11700
+    },
+    {
+      "epoch": 1.921643093219469,
+      "grad_norm": 0.33128113794288905,
+      "learning_rate": 6.180123760157187e-06,
+      "loss": 0.4649,
+      "step": 11701
+    },
+    {
+      "epoch": 1.9218073204278119,
+      "grad_norm": 0.33788593807856176,
+      "learning_rate": 6.179688691996611e-06,
+      "loss": 0.4505,
+      "step": 11702
+    },
+    {
+      "epoch": 1.9219715476361547,
+      "grad_norm": 0.29197053226973274,
+      "learning_rate": 6.179253605592966e-06,
+      "loss": 0.4484,
+      "step": 11703
+    },
+    {
+      "epoch": 1.9221357748444974,
+      "grad_norm": 0.4002773092580003,
+      "learning_rate": 6.178818500950975e-06,
+      "loss": 0.4368,
+      "step": 11704
+    },
+    {
+      "epoch": 1.92230000205284,
+      "grad_norm": 0.3225017152426909,
+      "learning_rate": 6.178383378075361e-06,
+      "loss": 0.4492,
+      "step": 11705
+    },
+    {
+      "epoch": 1.9224642292611829,
+      "grad_norm": 0.3292097446594349,
+      "learning_rate": 6.177948236970854e-06,
+      "loss": 0.4524,
+      "step": 11706
+    },
+    {
+      "epoch": 1.9226284564695257,
+      "grad_norm": 0.351995173082209,
+      "learning_rate": 6.177513077642178e-06,
+      "loss": 0.4453,
+      "step": 11707
+    },
+    {
+      "epoch": 1.9227926836778684,
+      "grad_norm": 0.31252517151783626,
+      "learning_rate": 6.177077900094058e-06,
+      "loss": 0.453,
+      "step": 11708
+    },
+    {
+      "epoch": 1.922956910886211,
+      "grad_norm": 0.2861624043707913,
+      "learning_rate": 6.176642704331224e-06,
+      "loss": 0.4326,
+      "step": 11709
+    },
+    {
+      "epoch": 1.9231211380945537,
+      "grad_norm": 0.3842211082952547,
+      "learning_rate": 6.176207490358399e-06,
+      "loss": 0.4414,
+      "step": 11710
+    },
+    {
+      "epoch": 1.9232853653028965,
+      "grad_norm": 0.7917893981486481,
+      "learning_rate": 6.175772258180314e-06,
+      "loss": 0.445,
+      "step": 11711
+    },
+    {
+      "epoch": 1.9234495925112394,
+      "grad_norm": 0.4790207371880203,
+      "learning_rate": 6.175337007801691e-06,
+      "loss": 0.4307,
+      "step": 11712
+    },
+    {
+      "epoch": 1.923613819719582,
+      "grad_norm": 0.40561312865159216,
+      "learning_rate": 6.174901739227259e-06,
+      "loss": 0.4498,
+      "step": 11713
+    },
+    {
+      "epoch": 1.9237780469279246,
+      "grad_norm": 0.47192462934888785,
+      "learning_rate": 6.174466452461749e-06,
+      "loss": 0.4539,
+      "step": 11714
+    },
+    {
+      "epoch": 1.9239422741362675,
+      "grad_norm": 0.3453248892639385,
+      "learning_rate": 6.174031147509885e-06,
+      "loss": 0.452,
+      "step": 11715
+    },
+    {
+      "epoch": 1.9241065013446104,
+      "grad_norm": 0.46993171084266266,
+      "learning_rate": 6.173595824376396e-06,
+      "loss": 0.441,
+      "step": 11716
+    },
+    {
+      "epoch": 1.924270728552953,
+      "grad_norm": 0.3132400675291704,
+      "learning_rate": 6.17316048306601e-06,
+      "loss": 0.4474,
+      "step": 11717
+    },
+    {
+      "epoch": 1.9244349557612956,
+      "grad_norm": 0.5141692658873547,
+      "learning_rate": 6.172725123583452e-06,
+      "loss": 0.4445,
+      "step": 11718
+    },
+    {
+      "epoch": 1.9245991829696385,
+      "grad_norm": 0.3123892509641973,
+      "learning_rate": 6.1722897459334554e-06,
+      "loss": 0.4195,
+      "step": 11719
+    },
+    {
+      "epoch": 1.9247634101779814,
+      "grad_norm": 0.32517861042004026,
+      "learning_rate": 6.171854350120748e-06,
+      "loss": 0.4694,
+      "step": 11720
+    },
+    {
+      "epoch": 1.924927637386324,
+      "grad_norm": 0.43041962733396594,
+      "learning_rate": 6.171418936150057e-06,
+      "loss": 0.4448,
+      "step": 11721
+    },
+    {
+      "epoch": 1.9250918645946666,
+      "grad_norm": 0.32581878191201175,
+      "learning_rate": 6.170983504026111e-06,
+      "loss": 0.4353,
+      "step": 11722
+    },
+    {
+      "epoch": 1.9252560918030095,
+      "grad_norm": 0.3221459896375791,
+      "learning_rate": 6.17054805375364e-06,
+      "loss": 0.4501,
+      "step": 11723
+    },
+    {
+      "epoch": 1.9254203190113524,
+      "grad_norm": 0.4751198372362801,
+      "learning_rate": 6.170112585337375e-06,
+      "loss": 0.466,
+      "step": 11724
+    },
+    {
+      "epoch": 1.925584546219695,
+      "grad_norm": 0.4111261993043956,
+      "learning_rate": 6.169677098782044e-06,
+      "loss": 0.4415,
+      "step": 11725
+    },
+    {
+      "epoch": 1.9257487734280376,
+      "grad_norm": 0.4096292396716481,
+      "learning_rate": 6.169241594092376e-06,
+      "loss": 0.4322,
+      "step": 11726
+    },
+    {
+      "epoch": 1.9259130006363803,
+      "grad_norm": 0.3758588386747688,
+      "learning_rate": 6.168806071273102e-06,
+      "loss": 0.4531,
+      "step": 11727
+    },
+    {
+      "epoch": 1.9260772278447231,
+      "grad_norm": 0.32238900503219686,
+      "learning_rate": 6.168370530328952e-06,
+      "loss": 0.4496,
+      "step": 11728
+    },
+    {
+      "epoch": 1.926241455053066,
+      "grad_norm": 0.3738260566944197,
+      "learning_rate": 6.167934971264657e-06,
+      "loss": 0.4533,
+      "step": 11729
+    },
+    {
+      "epoch": 1.9264056822614086,
+      "grad_norm": 0.3247033594038588,
+      "learning_rate": 6.167499394084947e-06,
+      "loss": 0.4448,
+      "step": 11730
+    },
+    {
+      "epoch": 1.9265699094697513,
+      "grad_norm": 0.3613742112719459,
+      "learning_rate": 6.167063798794553e-06,
+      "loss": 0.452,
+      "step": 11731
+    },
+    {
+      "epoch": 1.9267341366780941,
+      "grad_norm": 0.30402605193304383,
+      "learning_rate": 6.166628185398207e-06,
+      "loss": 0.4418,
+      "step": 11732
+    },
+    {
+      "epoch": 1.926898363886437,
+      "grad_norm": 0.36077690801634543,
+      "learning_rate": 6.166192553900637e-06,
+      "loss": 0.4634,
+      "step": 11733
+    },
+    {
+      "epoch": 1.9270625910947796,
+      "grad_norm": 0.3157741710943167,
+      "learning_rate": 6.165756904306578e-06,
+      "loss": 0.437,
+      "step": 11734
+    },
+    {
+      "epoch": 1.9272268183031223,
+      "grad_norm": 0.36396243689784985,
+      "learning_rate": 6.16532123662076e-06,
+      "loss": 0.4715,
+      "step": 11735
+    },
+    {
+      "epoch": 1.9273910455114651,
+      "grad_norm": 0.32922576893702915,
+      "learning_rate": 6.164885550847916e-06,
+      "loss": 0.4401,
+      "step": 11736
+    },
+    {
+      "epoch": 1.927555272719808,
+      "grad_norm": 0.40557599637326114,
+      "learning_rate": 6.1644498469927755e-06,
+      "loss": 0.4527,
+      "step": 11737
+    },
+    {
+      "epoch": 1.9277194999281506,
+      "grad_norm": 0.32617286436421994,
+      "learning_rate": 6.164014125060072e-06,
+      "loss": 0.4699,
+      "step": 11738
+    },
+    {
+      "epoch": 1.9278837271364933,
+      "grad_norm": 0.3296499446537119,
+      "learning_rate": 6.163578385054538e-06,
+      "loss": 0.4516,
+      "step": 11739
+    },
+    {
+      "epoch": 1.9280479543448361,
+      "grad_norm": 0.5284605915318895,
+      "learning_rate": 6.163142626980906e-06,
+      "loss": 0.4438,
+      "step": 11740
+    },
+    {
+      "epoch": 1.928212181553179,
+      "grad_norm": 0.304222483380602,
+      "learning_rate": 6.1627068508439095e-06,
+      "loss": 0.4493,
+      "step": 11741
+    },
+    {
+      "epoch": 1.9283764087615216,
+      "grad_norm": 0.37578326123625305,
+      "learning_rate": 6.1622710566482795e-06,
+      "loss": 0.4453,
+      "step": 11742
+    },
+    {
+      "epoch": 1.9285406359698642,
+      "grad_norm": 0.3316740779606767,
+      "learning_rate": 6.161835244398751e-06,
+      "loss": 0.4475,
+      "step": 11743
+    },
+    {
+      "epoch": 1.9287048631782069,
+      "grad_norm": 0.4724230151433168,
+      "learning_rate": 6.161399414100057e-06,
+      "loss": 0.464,
+      "step": 11744
+    },
+    {
+      "epoch": 1.9288690903865497,
+      "grad_norm": 0.33102520210764236,
+      "learning_rate": 6.160963565756932e-06,
+      "loss": 0.4482,
+      "step": 11745
+    },
+    {
+      "epoch": 1.9290333175948926,
+      "grad_norm": 0.3039412054856362,
+      "learning_rate": 6.160527699374107e-06,
+      "loss": 0.4579,
+      "step": 11746
+    },
+    {
+      "epoch": 1.9291975448032352,
+      "grad_norm": 0.4759493387711221,
+      "learning_rate": 6.160091814956317e-06,
+      "loss": 0.4543,
+      "step": 11747
+    },
+    {
+      "epoch": 1.9293617720115779,
+      "grad_norm": 0.30295570419838147,
+      "learning_rate": 6.159655912508297e-06,
+      "loss": 0.4502,
+      "step": 11748
+    },
+    {
+      "epoch": 1.9295259992199207,
+      "grad_norm": 0.3095851008676612,
+      "learning_rate": 6.159219992034782e-06,
+      "loss": 0.4476,
+      "step": 11749
+    },
+    {
+      "epoch": 1.9296902264282636,
+      "grad_norm": 0.30467728546486345,
+      "learning_rate": 6.158784053540504e-06,
+      "loss": 0.4536,
+      "step": 11750
+    },
+    {
+      "epoch": 1.9298544536366062,
+      "grad_norm": 0.3967570517620851,
+      "learning_rate": 6.1583480970301995e-06,
+      "loss": 0.436,
+      "step": 11751
+    },
+    {
+      "epoch": 1.9300186808449489,
+      "grad_norm": 0.3002495362111364,
+      "learning_rate": 6.157912122508603e-06,
+      "loss": 0.4264,
+      "step": 11752
+    },
+    {
+      "epoch": 1.9301829080532917,
+      "grad_norm": 0.3828269006370637,
+      "learning_rate": 6.157476129980451e-06,
+      "loss": 0.4421,
+      "step": 11753
+    },
+    {
+      "epoch": 1.9303471352616346,
+      "grad_norm": 0.30484222608414635,
+      "learning_rate": 6.157040119450475e-06,
+      "loss": 0.4594,
+      "step": 11754
+    },
+    {
+      "epoch": 1.9305113624699772,
+      "grad_norm": 0.4559966835142827,
+      "learning_rate": 6.156604090923415e-06,
+      "loss": 0.4618,
+      "step": 11755
+    },
+    {
+      "epoch": 1.9306755896783199,
+      "grad_norm": 0.2895503351697011,
+      "learning_rate": 6.1561680444040035e-06,
+      "loss": 0.4558,
+      "step": 11756
+    },
+    {
+      "epoch": 1.9308398168866627,
+      "grad_norm": 0.30856588658860873,
+      "learning_rate": 6.1557319798969785e-06,
+      "loss": 0.4452,
+      "step": 11757
+    },
+    {
+      "epoch": 1.9310040440950056,
+      "grad_norm": 0.49077377646148856,
+      "learning_rate": 6.155295897407075e-06,
+      "loss": 0.4692,
+      "step": 11758
+    },
+    {
+      "epoch": 1.9311682713033482,
+      "grad_norm": 0.2968529348737747,
+      "learning_rate": 6.154859796939029e-06,
+      "loss": 0.4615,
+      "step": 11759
+    },
+    {
+      "epoch": 1.9313324985116909,
+      "grad_norm": 0.3808550571539235,
+      "learning_rate": 6.154423678497578e-06,
+      "loss": 0.4438,
+      "step": 11760
+    },
+    {
+      "epoch": 1.9314967257200335,
+      "grad_norm": 0.33257583457981604,
+      "learning_rate": 6.153987542087457e-06,
+      "loss": 0.4715,
+      "step": 11761
+    },
+    {
+      "epoch": 1.9316609529283764,
+      "grad_norm": 0.3034797701744059,
+      "learning_rate": 6.153551387713406e-06,
+      "loss": 0.4335,
+      "step": 11762
+    },
+    {
+      "epoch": 1.9318251801367192,
+      "grad_norm": 0.3366439268492937,
+      "learning_rate": 6.153115215380159e-06,
+      "loss": 0.4536,
+      "step": 11763
+    },
+    {
+      "epoch": 1.9319894073450619,
+      "grad_norm": 0.333839910373322,
+      "learning_rate": 6.1526790250924545e-06,
+      "loss": 0.4588,
+      "step": 11764
+    },
+    {
+      "epoch": 1.9321536345534045,
+      "grad_norm": 0.29555048190918864,
+      "learning_rate": 6.1522428168550286e-06,
+      "loss": 0.441,
+      "step": 11765
+    },
+    {
+      "epoch": 1.9323178617617474,
+      "grad_norm": 0.3160434566532558,
+      "learning_rate": 6.151806590672622e-06,
+      "loss": 0.4259,
+      "step": 11766
+    },
+    {
+      "epoch": 1.9324820889700902,
+      "grad_norm": 0.3977836618478647,
+      "learning_rate": 6.151370346549969e-06,
+      "loss": 0.4411,
+      "step": 11767
+    },
+    {
+      "epoch": 1.9326463161784329,
+      "grad_norm": 0.2852990076026332,
+      "learning_rate": 6.150934084491809e-06,
+      "loss": 0.4394,
+      "step": 11768
+    },
+    {
+      "epoch": 1.9328105433867755,
+      "grad_norm": 0.34967502718124766,
+      "learning_rate": 6.1504978045028825e-06,
+      "loss": 0.4371,
+      "step": 11769
+    },
+    {
+      "epoch": 1.9329747705951184,
+      "grad_norm": 0.2949264653113903,
+      "learning_rate": 6.150061506587925e-06,
+      "loss": 0.4506,
+      "step": 11770
+    },
+    {
+      "epoch": 1.9331389978034612,
+      "grad_norm": 0.32292941521963847,
+      "learning_rate": 6.149625190751676e-06,
+      "loss": 0.4514,
+      "step": 11771
+    },
+    {
+      "epoch": 1.9333032250118038,
+      "grad_norm": 0.3901095676284324,
+      "learning_rate": 6.149188856998874e-06,
+      "loss": 0.4461,
+      "step": 11772
+    },
+    {
+      "epoch": 1.9334674522201465,
+      "grad_norm": 0.3148958430118748,
+      "learning_rate": 6.148752505334259e-06,
+      "loss": 0.439,
+      "step": 11773
+    },
+    {
+      "epoch": 1.9336316794284893,
+      "grad_norm": 0.4645426254687232,
+      "learning_rate": 6.148316135762571e-06,
+      "loss": 0.4662,
+      "step": 11774
+    },
+    {
+      "epoch": 1.9337959066368322,
+      "grad_norm": 0.3381456693550708,
+      "learning_rate": 6.147879748288546e-06,
+      "loss": 0.4397,
+      "step": 11775
+    },
+    {
+      "epoch": 1.9339601338451748,
+      "grad_norm": 0.2965471428211383,
+      "learning_rate": 6.1474433429169255e-06,
+      "loss": 0.4574,
+      "step": 11776
+    },
+    {
+      "epoch": 1.9341243610535175,
+      "grad_norm": 0.3603201346209414,
+      "learning_rate": 6.14700691965245e-06,
+      "loss": 0.4469,
+      "step": 11777
+    },
+    {
+      "epoch": 1.9342885882618601,
+      "grad_norm": 0.32054268239319894,
+      "learning_rate": 6.146570478499859e-06,
+      "loss": 0.437,
+      "step": 11778
+    },
+    {
+      "epoch": 1.934452815470203,
+      "grad_norm": 0.48292569009156594,
+      "learning_rate": 6.146134019463895e-06,
+      "loss": 0.4561,
+      "step": 11779
+    },
+    {
+      "epoch": 1.9346170426785458,
+      "grad_norm": 0.3414445379177717,
+      "learning_rate": 6.1456975425492925e-06,
+      "loss": 0.4601,
+      "step": 11780
+    },
+    {
+      "epoch": 1.9347812698868885,
+      "grad_norm": 0.32659418586761957,
+      "learning_rate": 6.145261047760797e-06,
+      "loss": 0.4548,
+      "step": 11781
+    },
+    {
+      "epoch": 1.9349454970952311,
+      "grad_norm": 0.35652097068907895,
+      "learning_rate": 6.144824535103147e-06,
+      "loss": 0.4782,
+      "step": 11782
+    },
+    {
+      "epoch": 1.935109724303574,
+      "grad_norm": 0.2759711761110947,
+      "learning_rate": 6.144388004581084e-06,
+      "loss": 0.4528,
+      "step": 11783
+    },
+    {
+      "epoch": 1.9352739515119168,
+      "grad_norm": 0.34655152524113225,
+      "learning_rate": 6.143951456199352e-06,
+      "loss": 0.4426,
+      "step": 11784
+    },
+    {
+      "epoch": 1.9354381787202595,
+      "grad_norm": 0.30678758926269073,
+      "learning_rate": 6.143514889962687e-06,
+      "loss": 0.4533,
+      "step": 11785
+    },
+    {
+      "epoch": 1.935602405928602,
+      "grad_norm": 0.3465024740320573,
+      "learning_rate": 6.143078305875834e-06,
+      "loss": 0.4457,
+      "step": 11786
+    },
+    {
+      "epoch": 1.935766633136945,
+      "grad_norm": 0.2944811158711843,
+      "learning_rate": 6.142641703943534e-06,
+      "loss": 0.4469,
+      "step": 11787
+    },
+    {
+      "epoch": 1.9359308603452878,
+      "grad_norm": 0.4588778909977551,
+      "learning_rate": 6.142205084170529e-06,
+      "loss": 0.4569,
+      "step": 11788
+    },
+    {
+      "epoch": 1.9360950875536305,
+      "grad_norm": 0.38001254582850585,
+      "learning_rate": 6.141768446561563e-06,
+      "loss": 0.4391,
+      "step": 11789
+    },
+    {
+      "epoch": 1.936259314761973,
+      "grad_norm": 0.36647792906297894,
+      "learning_rate": 6.141331791121374e-06,
+      "loss": 0.4675,
+      "step": 11790
+    },
+    {
+      "epoch": 1.936423541970316,
+      "grad_norm": 0.3147751631851471,
+      "learning_rate": 6.140895117854708e-06,
+      "loss": 0.4451,
+      "step": 11791
+    },
+    {
+      "epoch": 1.9365877691786588,
+      "grad_norm": 0.37708723383068626,
+      "learning_rate": 6.140458426766305e-06,
+      "loss": 0.4363,
+      "step": 11792
+    },
+    {
+      "epoch": 1.9367519963870015,
+      "grad_norm": 0.314419687793131,
+      "learning_rate": 6.140021717860911e-06,
+      "loss": 0.4344,
+      "step": 11793
+    },
+    {
+      "epoch": 1.936916223595344,
+      "grad_norm": 0.2978678788734563,
+      "learning_rate": 6.139584991143268e-06,
+      "loss": 0.4354,
+      "step": 11794
+    },
+    {
+      "epoch": 1.9370804508036867,
+      "grad_norm": 0.6290390429035364,
+      "learning_rate": 6.139148246618118e-06,
+      "loss": 0.4435,
+      "step": 11795
+    },
+    {
+      "epoch": 1.9372446780120296,
+      "grad_norm": 0.3930535583884199,
+      "learning_rate": 6.138711484290205e-06,
+      "loss": 0.4452,
+      "step": 11796
+    },
+    {
+      "epoch": 1.9374089052203725,
+      "grad_norm": 0.4058397974826341,
+      "learning_rate": 6.1382747041642735e-06,
+      "loss": 0.4424,
+      "step": 11797
+    },
+    {
+      "epoch": 1.937573132428715,
+      "grad_norm": 0.28754785524275933,
+      "learning_rate": 6.137837906245067e-06,
+      "loss": 0.4626,
+      "step": 11798
+    },
+    {
+      "epoch": 1.9377373596370577,
+      "grad_norm": 0.37145994996852144,
+      "learning_rate": 6.13740109053733e-06,
+      "loss": 0.4296,
+      "step": 11799
+    },
+    {
+      "epoch": 1.9379015868454006,
+      "grad_norm": 0.31987207942764523,
+      "learning_rate": 6.136964257045804e-06,
+      "loss": 0.4407,
+      "step": 11800
+    },
+    {
+      "epoch": 1.9380658140537435,
+      "grad_norm": 0.3120597556496171,
+      "learning_rate": 6.136527405775238e-06,
+      "loss": 0.4154,
+      "step": 11801
+    },
+    {
+      "epoch": 1.938230041262086,
+      "grad_norm": 0.32184214275987144,
+      "learning_rate": 6.136090536730372e-06,
+      "loss": 0.4416,
+      "step": 11802
+    },
+    {
+      "epoch": 1.9383942684704287,
+      "grad_norm": 0.30422790837511227,
+      "learning_rate": 6.1356536499159555e-06,
+      "loss": 0.4775,
+      "step": 11803
+    },
+    {
+      "epoch": 1.9385584956787716,
+      "grad_norm": 0.34677800611122794,
+      "learning_rate": 6.1352167453367305e-06,
+      "loss": 0.4449,
+      "step": 11804
+    },
+    {
+      "epoch": 1.9387227228871144,
+      "grad_norm": 0.2992127868849352,
+      "learning_rate": 6.134779822997442e-06,
+      "loss": 0.4461,
+      "step": 11805
+    },
+    {
+      "epoch": 1.938886950095457,
+      "grad_norm": 0.3231711502256461,
+      "learning_rate": 6.134342882902836e-06,
+      "loss": 0.4586,
+      "step": 11806
+    },
+    {
+      "epoch": 1.9390511773037997,
+      "grad_norm": 0.3037909014573711,
+      "learning_rate": 6.133905925057659e-06,
+      "loss": 0.4529,
+      "step": 11807
+    },
+    {
+      "epoch": 1.9392154045121426,
+      "grad_norm": 0.3337539556315541,
+      "learning_rate": 6.1334689494666564e-06,
+      "loss": 0.4604,
+      "step": 11808
+    },
+    {
+      "epoch": 1.9393796317204854,
+      "grad_norm": 0.33305602986601746,
+      "learning_rate": 6.133031956134573e-06,
+      "loss": 0.4361,
+      "step": 11809
+    },
+    {
+      "epoch": 1.939543858928828,
+      "grad_norm": 0.8091816550177268,
+      "learning_rate": 6.132594945066157e-06,
+      "loss": 0.4646,
+      "step": 11810
+    },
+    {
+      "epoch": 1.9397080861371707,
+      "grad_norm": 0.32246244156970166,
+      "learning_rate": 6.132157916266152e-06,
+      "loss": 0.4639,
+      "step": 11811
+    },
+    {
+      "epoch": 1.9398723133455134,
+      "grad_norm": 0.44159454100120704,
+      "learning_rate": 6.131720869739307e-06,
+      "loss": 0.4508,
+      "step": 11812
+    },
+    {
+      "epoch": 1.9400365405538562,
+      "grad_norm": 0.2880429219605012,
+      "learning_rate": 6.131283805490368e-06,
+      "loss": 0.4475,
+      "step": 11813
+    },
+    {
+      "epoch": 1.940200767762199,
+      "grad_norm": 0.2735895117494609,
+      "learning_rate": 6.130846723524082e-06,
+      "loss": 0.4595,
+      "step": 11814
+    },
+    {
+      "epoch": 1.9403649949705417,
+      "grad_norm": 0.3123242230326185,
+      "learning_rate": 6.130409623845196e-06,
+      "loss": 0.4765,
+      "step": 11815
+    },
+    {
+      "epoch": 1.9405292221788843,
+      "grad_norm": 0.29578411751014344,
+      "learning_rate": 6.129972506458458e-06,
+      "loss": 0.4576,
+      "step": 11816
+    },
+    {
+      "epoch": 1.9406934493872272,
+      "grad_norm": 0.4095465840591172,
+      "learning_rate": 6.129535371368614e-06,
+      "loss": 0.4277,
+      "step": 11817
+    },
+    {
+      "epoch": 1.94085767659557,
+      "grad_norm": 0.27967699795557444,
+      "learning_rate": 6.129098218580414e-06,
+      "loss": 0.44,
+      "step": 11818
+    },
+    {
+      "epoch": 1.9410219038039127,
+      "grad_norm": 0.3700003102868501,
+      "learning_rate": 6.128661048098602e-06,
+      "loss": 0.4289,
+      "step": 11819
+    },
+    {
+      "epoch": 1.9411861310122553,
+      "grad_norm": 0.4692973743249203,
+      "learning_rate": 6.128223859927931e-06,
+      "loss": 0.4652,
+      "step": 11820
+    },
+    {
+      "epoch": 1.9413503582205982,
+      "grad_norm": 0.3164140915929167,
+      "learning_rate": 6.1277866540731465e-06,
+      "loss": 0.4383,
+      "step": 11821
+    },
+    {
+      "epoch": 1.941514585428941,
+      "grad_norm": 0.4703336243325058,
+      "learning_rate": 6.1273494305389956e-06,
+      "loss": 0.4446,
+      "step": 11822
+    },
+    {
+      "epoch": 1.9416788126372837,
+      "grad_norm": 0.35750823869899867,
+      "learning_rate": 6.126912189330231e-06,
+      "loss": 0.4488,
+      "step": 11823
+    },
+    {
+      "epoch": 1.9418430398456263,
+      "grad_norm": 0.3118945660735331,
+      "learning_rate": 6.126474930451599e-06,
+      "loss": 0.4787,
+      "step": 11824
+    },
+    {
+      "epoch": 1.9420072670539692,
+      "grad_norm": 0.2687496340376281,
+      "learning_rate": 6.126037653907848e-06,
+      "loss": 0.4405,
+      "step": 11825
+    },
+    {
+      "epoch": 1.942171494262312,
+      "grad_norm": 0.591042238910758,
+      "learning_rate": 6.125600359703728e-06,
+      "loss": 0.4355,
+      "step": 11826
+    },
+    {
+      "epoch": 1.9423357214706547,
+      "grad_norm": 0.32423012638815046,
+      "learning_rate": 6.125163047843991e-06,
+      "loss": 0.4647,
+      "step": 11827
+    },
+    {
+      "epoch": 1.9424999486789973,
+      "grad_norm": 0.448756546678214,
+      "learning_rate": 6.124725718333383e-06,
+      "loss": 0.4245,
+      "step": 11828
+    },
+    {
+      "epoch": 1.94266417588734,
+      "grad_norm": 0.30948247325201406,
+      "learning_rate": 6.124288371176655e-06,
+      "loss": 0.4461,
+      "step": 11829
+    },
+    {
+      "epoch": 1.9428284030956828,
+      "grad_norm": 0.2882913994343183,
+      "learning_rate": 6.123851006378556e-06,
+      "loss": 0.4506,
+      "step": 11830
+    },
+    {
+      "epoch": 1.9429926303040257,
+      "grad_norm": 0.3698476194070314,
+      "learning_rate": 6.123413623943839e-06,
+      "loss": 0.4417,
+      "step": 11831
+    },
+    {
+      "epoch": 1.9431568575123683,
+      "grad_norm": 0.39572904633456146,
+      "learning_rate": 6.122976223877253e-06,
+      "loss": 0.4455,
+      "step": 11832
+    },
+    {
+      "epoch": 1.943321084720711,
+      "grad_norm": 0.31198435687290366,
+      "learning_rate": 6.122538806183548e-06,
+      "loss": 0.4525,
+      "step": 11833
+    },
+    {
+      "epoch": 1.9434853119290538,
+      "grad_norm": 0.3316692403626222,
+      "learning_rate": 6.122101370867475e-06,
+      "loss": 0.462,
+      "step": 11834
+    },
+    {
+      "epoch": 1.9436495391373967,
+      "grad_norm": 0.33372374349452755,
+      "learning_rate": 6.121663917933784e-06,
+      "loss": 0.4522,
+      "step": 11835
+    },
+    {
+      "epoch": 1.9438137663457393,
+      "grad_norm": 0.3229007237330423,
+      "learning_rate": 6.121226447387229e-06,
+      "loss": 0.4402,
+      "step": 11836
+    },
+    {
+      "epoch": 1.943977993554082,
+      "grad_norm": 0.34876032942863944,
+      "learning_rate": 6.12078895923256e-06,
+      "loss": 0.4379,
+      "step": 11837
+    },
+    {
+      "epoch": 1.9441422207624248,
+      "grad_norm": 0.3025199224775647,
+      "learning_rate": 6.120351453474528e-06,
+      "loss": 0.4424,
+      "step": 11838
+    },
+    {
+      "epoch": 1.9443064479707677,
+      "grad_norm": 0.3042076390572115,
+      "learning_rate": 6.119913930117884e-06,
+      "loss": 0.422,
+      "step": 11839
+    },
+    {
+      "epoch": 1.9444706751791103,
+      "grad_norm": 0.3464106773619736,
+      "learning_rate": 6.119476389167382e-06,
+      "loss": 0.4286,
+      "step": 11840
+    },
+    {
+      "epoch": 1.944634902387453,
+      "grad_norm": 0.31756238332376324,
+      "learning_rate": 6.119038830627772e-06,
+      "loss": 0.4447,
+      "step": 11841
+    },
+    {
+      "epoch": 1.9447991295957958,
+      "grad_norm": 0.3361628110899518,
+      "learning_rate": 6.118601254503809e-06,
+      "loss": 0.4364,
+      "step": 11842
+    },
+    {
+      "epoch": 1.9449633568041387,
+      "grad_norm": 0.4516933395464814,
+      "learning_rate": 6.118163660800243e-06,
+      "loss": 0.4539,
+      "step": 11843
+    },
+    {
+      "epoch": 1.9451275840124813,
+      "grad_norm": 0.3930468527746109,
+      "learning_rate": 6.117726049521826e-06,
+      "loss": 0.47,
+      "step": 11844
+    },
+    {
+      "epoch": 1.945291811220824,
+      "grad_norm": 0.35826140019934466,
+      "learning_rate": 6.117288420673315e-06,
+      "loss": 0.4543,
+      "step": 11845
+    },
+    {
+      "epoch": 1.9454560384291666,
+      "grad_norm": 0.32262749277511915,
+      "learning_rate": 6.116850774259458e-06,
+      "loss": 0.4575,
+      "step": 11846
+    },
+    {
+      "epoch": 1.9456202656375094,
+      "grad_norm": 0.42017179470753757,
+      "learning_rate": 6.116413110285014e-06,
+      "loss": 0.4259,
+      "step": 11847
+    },
+    {
+      "epoch": 1.9457844928458523,
+      "grad_norm": 0.3091060953354457,
+      "learning_rate": 6.115975428754731e-06,
+      "loss": 0.4354,
+      "step": 11848
+    },
+    {
+      "epoch": 1.945948720054195,
+      "grad_norm": 0.449447054482981,
+      "learning_rate": 6.115537729673366e-06,
+      "loss": 0.4313,
+      "step": 11849
+    },
+    {
+      "epoch": 1.9461129472625376,
+      "grad_norm": 0.3242192526893855,
+      "learning_rate": 6.115100013045671e-06,
+      "loss": 0.4504,
+      "step": 11850
+    },
+    {
+      "epoch": 1.9462771744708804,
+      "grad_norm": 0.33839339236288085,
+      "learning_rate": 6.1146622788764e-06,
+      "loss": 0.4586,
+      "step": 11851
+    },
+    {
+      "epoch": 1.9464414016792233,
+      "grad_norm": 0.41077135290200856,
+      "learning_rate": 6.11422452717031e-06,
+      "loss": 0.45,
+      "step": 11852
+    },
+    {
+      "epoch": 1.946605628887566,
+      "grad_norm": 0.46154500843898666,
+      "learning_rate": 6.113786757932154e-06,
+      "loss": 0.4687,
+      "step": 11853
+    },
+    {
+      "epoch": 1.9467698560959086,
+      "grad_norm": 0.38772725882298875,
+      "learning_rate": 6.113348971166684e-06,
+      "loss": 0.4565,
+      "step": 11854
+    },
+    {
+      "epoch": 1.9469340833042514,
+      "grad_norm": 0.34951381136472154,
+      "learning_rate": 6.1129111668786565e-06,
+      "loss": 0.433,
+      "step": 11855
+    },
+    {
+      "epoch": 1.9470983105125943,
+      "grad_norm": 0.3058343506675485,
+      "learning_rate": 6.112473345072829e-06,
+      "loss": 0.4443,
+      "step": 11856
+    },
+    {
+      "epoch": 1.947262537720937,
+      "grad_norm": 0.458504542424256,
+      "learning_rate": 6.112035505753952e-06,
+      "loss": 0.4228,
+      "step": 11857
+    },
+    {
+      "epoch": 1.9474267649292796,
+      "grad_norm": 0.3648480272272111,
+      "learning_rate": 6.111597648926786e-06,
+      "loss": 0.4657,
+      "step": 11858
+    },
+    {
+      "epoch": 1.9475909921376224,
+      "grad_norm": 0.3315564115633353,
+      "learning_rate": 6.1111597745960825e-06,
+      "loss": 0.4344,
+      "step": 11859
+    },
+    {
+      "epoch": 1.9477552193459653,
+      "grad_norm": 0.3203259202134563,
+      "learning_rate": 6.110721882766598e-06,
+      "loss": 0.4507,
+      "step": 11860
+    },
+    {
+      "epoch": 1.947919446554308,
+      "grad_norm": 0.32526518490626355,
+      "learning_rate": 6.1102839734430905e-06,
+      "loss": 0.4483,
+      "step": 11861
+    },
+    {
+      "epoch": 1.9480836737626506,
+      "grad_norm": 0.30568500186568304,
+      "learning_rate": 6.109846046630315e-06,
+      "loss": 0.4251,
+      "step": 11862
+    },
+    {
+      "epoch": 1.9482479009709932,
+      "grad_norm": 0.34335284427598955,
+      "learning_rate": 6.1094081023330265e-06,
+      "loss": 0.442,
+      "step": 11863
+    },
+    {
+      "epoch": 1.948412128179336,
+      "grad_norm": 0.3000840673867567,
+      "learning_rate": 6.108970140555982e-06,
+      "loss": 0.4462,
+      "step": 11864
+    },
+    {
+      "epoch": 1.948576355387679,
+      "grad_norm": 0.3536142230185942,
+      "learning_rate": 6.10853216130394e-06,
+      "loss": 0.4583,
+      "step": 11865
+    },
+    {
+      "epoch": 1.9487405825960216,
+      "grad_norm": 0.31933514094763754,
+      "learning_rate": 6.108094164581656e-06,
+      "loss": 0.4355,
+      "step": 11866
+    },
+    {
+      "epoch": 1.9489048098043642,
+      "grad_norm": 0.35391874407088647,
+      "learning_rate": 6.107656150393888e-06,
+      "loss": 0.4603,
+      "step": 11867
+    },
+    {
+      "epoch": 1.949069037012707,
+      "grad_norm": 0.37510286288830225,
+      "learning_rate": 6.10721811874539e-06,
+      "loss": 0.4357,
+      "step": 11868
+    },
+    {
+      "epoch": 1.94923326422105,
+      "grad_norm": 0.3732690817187727,
+      "learning_rate": 6.106780069640924e-06,
+      "loss": 0.4551,
+      "step": 11869
+    },
+    {
+      "epoch": 1.9493974914293926,
+      "grad_norm": 0.3714339876825457,
+      "learning_rate": 6.106342003085246e-06,
+      "loss": 0.4611,
+      "step": 11870
+    },
+    {
+      "epoch": 1.9495617186377352,
+      "grad_norm": 0.2837021911002435,
+      "learning_rate": 6.1059039190831115e-06,
+      "loss": 0.4401,
+      "step": 11871
+    },
+    {
+      "epoch": 1.949725945846078,
+      "grad_norm": 0.4639579970164344,
+      "learning_rate": 6.105465817639281e-06,
+      "loss": 0.4399,
+      "step": 11872
+    },
+    {
+      "epoch": 1.949890173054421,
+      "grad_norm": 0.3320134698090167,
+      "learning_rate": 6.105027698758512e-06,
+      "loss": 0.4232,
+      "step": 11873
+    },
+    {
+      "epoch": 1.9500544002627636,
+      "grad_norm": 0.28217258687959057,
+      "learning_rate": 6.104589562445565e-06,
+      "loss": 0.4613,
+      "step": 11874
+    },
+    {
+      "epoch": 1.9502186274711062,
+      "grad_norm": 0.3642163892311067,
+      "learning_rate": 6.104151408705195e-06,
+      "loss": 0.4611,
+      "step": 11875
+    },
+    {
+      "epoch": 1.950382854679449,
+      "grad_norm": 0.7508595149691321,
+      "learning_rate": 6.103713237542163e-06,
+      "loss": 0.4404,
+      "step": 11876
+    },
+    {
+      "epoch": 1.950547081887792,
+      "grad_norm": 0.33888086777812015,
+      "learning_rate": 6.103275048961227e-06,
+      "loss": 0.4326,
+      "step": 11877
+    },
+    {
+      "epoch": 1.9507113090961345,
+      "grad_norm": 0.29969210211511377,
+      "learning_rate": 6.102836842967146e-06,
+      "loss": 0.4595,
+      "step": 11878
+    },
+    {
+      "epoch": 1.9508755363044772,
+      "grad_norm": 0.3898798633549321,
+      "learning_rate": 6.102398619564681e-06,
+      "loss": 0.4445,
+      "step": 11879
+    },
+    {
+      "epoch": 1.9510397635128198,
+      "grad_norm": 0.33948844287237157,
+      "learning_rate": 6.101960378758589e-06,
+      "loss": 0.4441,
+      "step": 11880
+    },
+    {
+      "epoch": 1.9512039907211627,
+      "grad_norm": 0.3009932679535096,
+      "learning_rate": 6.101522120553633e-06,
+      "loss": 0.4689,
+      "step": 11881
+    },
+    {
+      "epoch": 1.9513682179295055,
+      "grad_norm": 0.40499756454781155,
+      "learning_rate": 6.10108384495457e-06,
+      "loss": 0.443,
+      "step": 11882
+    },
+    {
+      "epoch": 1.9515324451378482,
+      "grad_norm": 0.4204322798597718,
+      "learning_rate": 6.100645551966162e-06,
+      "loss": 0.4286,
+      "step": 11883
+    },
+    {
+      "epoch": 1.9516966723461908,
+      "grad_norm": 0.31539927216386543,
+      "learning_rate": 6.100207241593167e-06,
+      "loss": 0.4402,
+      "step": 11884
+    },
+    {
+      "epoch": 1.9518608995545337,
+      "grad_norm": 0.295725731524103,
+      "learning_rate": 6.099768913840348e-06,
+      "loss": 0.4323,
+      "step": 11885
+    },
+    {
+      "epoch": 1.9520251267628765,
+      "grad_norm": 0.5328706680390949,
+      "learning_rate": 6.099330568712465e-06,
+      "loss": 0.4273,
+      "step": 11886
+    },
+    {
+      "epoch": 1.9521893539712192,
+      "grad_norm": 0.42119041668735635,
+      "learning_rate": 6.098892206214278e-06,
+      "loss": 0.447,
+      "step": 11887
+    },
+    {
+      "epoch": 1.9523535811795618,
+      "grad_norm": 0.3552233220762019,
+      "learning_rate": 6.098453826350549e-06,
+      "loss": 0.4624,
+      "step": 11888
+    },
+    {
+      "epoch": 1.9525178083879047,
+      "grad_norm": 0.35988767125876403,
+      "learning_rate": 6.0980154291260375e-06,
+      "loss": 0.4427,
+      "step": 11889
+    },
+    {
+      "epoch": 1.9526820355962475,
+      "grad_norm": 0.3725865342348446,
+      "learning_rate": 6.097577014545507e-06,
+      "loss": 0.4391,
+      "step": 11890
+    },
+    {
+      "epoch": 1.9528462628045902,
+      "grad_norm": 0.31357186920814756,
+      "learning_rate": 6.0971385826137194e-06,
+      "loss": 0.4449,
+      "step": 11891
+    },
+    {
+      "epoch": 1.9530104900129328,
+      "grad_norm": 0.38542678167037026,
+      "learning_rate": 6.0967001333354335e-06,
+      "loss": 0.4367,
+      "step": 11892
+    },
+    {
+      "epoch": 1.9531747172212757,
+      "grad_norm": 0.37312227942542797,
+      "learning_rate": 6.096261666715413e-06,
+      "loss": 0.4501,
+      "step": 11893
+    },
+    {
+      "epoch": 1.9533389444296185,
+      "grad_norm": 0.34996403105268287,
+      "learning_rate": 6.095823182758422e-06,
+      "loss": 0.4327,
+      "step": 11894
+    },
+    {
+      "epoch": 1.9535031716379612,
+      "grad_norm": 0.4262377392819523,
+      "learning_rate": 6.0953846814692214e-06,
+      "loss": 0.4317,
+      "step": 11895
+    },
+    {
+      "epoch": 1.9536673988463038,
+      "grad_norm": 0.3406779984733831,
+      "learning_rate": 6.094946162852573e-06,
+      "loss": 0.445,
+      "step": 11896
+    },
+    {
+      "epoch": 1.9538316260546464,
+      "grad_norm": 0.32443708298506085,
+      "learning_rate": 6.09450762691324e-06,
+      "loss": 0.4404,
+      "step": 11897
+    },
+    {
+      "epoch": 1.9539958532629893,
+      "grad_norm": 0.3150022109795289,
+      "learning_rate": 6.094069073655984e-06,
+      "loss": 0.4403,
+      "step": 11898
+    },
+    {
+      "epoch": 1.9541600804713322,
+      "grad_norm": 0.3574477988232649,
+      "learning_rate": 6.093630503085571e-06,
+      "loss": 0.4603,
+      "step": 11899
+    },
+    {
+      "epoch": 1.9543243076796748,
+      "grad_norm": 0.2974068481729623,
+      "learning_rate": 6.093191915206762e-06,
+      "loss": 0.4318,
+      "step": 11900
+    },
+    {
+      "epoch": 1.9544885348880174,
+      "grad_norm": 0.3768435152698072,
+      "learning_rate": 6.092753310024322e-06,
+      "loss": 0.4547,
+      "step": 11901
+    },
+    {
+      "epoch": 1.9546527620963603,
+      "grad_norm": 0.38455005690972244,
+      "learning_rate": 6.092314687543014e-06,
+      "loss": 0.4469,
+      "step": 11902
+    },
+    {
+      "epoch": 1.9548169893047032,
+      "grad_norm": 0.31822901379518637,
+      "learning_rate": 6.091876047767601e-06,
+      "loss": 0.4425,
+      "step": 11903
+    },
+    {
+      "epoch": 1.9549812165130458,
+      "grad_norm": 0.28402660304659677,
+      "learning_rate": 6.091437390702849e-06,
+      "loss": 0.4461,
+      "step": 11904
+    },
+    {
+      "epoch": 1.9551454437213884,
+      "grad_norm": 0.3521752474751283,
+      "learning_rate": 6.090998716353522e-06,
+      "loss": 0.4318,
+      "step": 11905
+    },
+    {
+      "epoch": 1.9553096709297313,
+      "grad_norm": 0.29060245833153847,
+      "learning_rate": 6.090560024724381e-06,
+      "loss": 0.4418,
+      "step": 11906
+    },
+    {
+      "epoch": 1.9554738981380742,
+      "grad_norm": 0.3912999714229161,
+      "learning_rate": 6.0901213158201946e-06,
+      "loss": 0.4295,
+      "step": 11907
+    },
+    {
+      "epoch": 1.9556381253464168,
+      "grad_norm": 0.3728285545096288,
+      "learning_rate": 6.089682589645727e-06,
+      "loss": 0.4326,
+      "step": 11908
+    },
+    {
+      "epoch": 1.9558023525547594,
+      "grad_norm": 0.3202203110579774,
+      "learning_rate": 6.08924384620574e-06,
+      "loss": 0.4463,
+      "step": 11909
+    },
+    {
+      "epoch": 1.9559665797631023,
+      "grad_norm": 0.536561797356937,
+      "learning_rate": 6.088805085505004e-06,
+      "loss": 0.4376,
+      "step": 11910
+    },
+    {
+      "epoch": 1.9561308069714451,
+      "grad_norm": 0.4067293800969105,
+      "learning_rate": 6.08836630754828e-06,
+      "loss": 0.4476,
+      "step": 11911
+    },
+    {
+      "epoch": 1.9562950341797878,
+      "grad_norm": 0.3378217956415752,
+      "learning_rate": 6.087927512340336e-06,
+      "loss": 0.4475,
+      "step": 11912
+    },
+    {
+      "epoch": 1.9564592613881304,
+      "grad_norm": 0.29079037548101955,
+      "learning_rate": 6.087488699885936e-06,
+      "loss": 0.4429,
+      "step": 11913
+    },
+    {
+      "epoch": 1.956623488596473,
+      "grad_norm": 0.37519263681592546,
+      "learning_rate": 6.0870498701898465e-06,
+      "loss": 0.4377,
+      "step": 11914
+    },
+    {
+      "epoch": 1.956787715804816,
+      "grad_norm": 0.35455480132931083,
+      "learning_rate": 6.086611023256836e-06,
+      "loss": 0.4667,
+      "step": 11915
+    },
+    {
+      "epoch": 1.9569519430131588,
+      "grad_norm": 0.4006295729770789,
+      "learning_rate": 6.086172159091667e-06,
+      "loss": 0.4507,
+      "step": 11916
+    },
+    {
+      "epoch": 1.9571161702215014,
+      "grad_norm": 0.5609335201389217,
+      "learning_rate": 6.085733277699109e-06,
+      "loss": 0.4528,
+      "step": 11917
+    },
+    {
+      "epoch": 1.957280397429844,
+      "grad_norm": 0.3352582817058797,
+      "learning_rate": 6.085294379083927e-06,
+      "loss": 0.4575,
+      "step": 11918
+    },
+    {
+      "epoch": 1.957444624638187,
+      "grad_norm": 0.34092400470920575,
+      "learning_rate": 6.084855463250887e-06,
+      "loss": 0.4442,
+      "step": 11919
+    },
+    {
+      "epoch": 1.9576088518465298,
+      "grad_norm": 0.4316159655034461,
+      "learning_rate": 6.08441653020476e-06,
+      "loss": 0.448,
+      "step": 11920
+    },
+    {
+      "epoch": 1.9577730790548724,
+      "grad_norm": 0.28558420396013084,
+      "learning_rate": 6.083977579950309e-06,
+      "loss": 0.4419,
+      "step": 11921
+    },
+    {
+      "epoch": 1.957937306263215,
+      "grad_norm": 0.40953524784616846,
+      "learning_rate": 6.083538612492302e-06,
+      "loss": 0.4469,
+      "step": 11922
+    },
+    {
+      "epoch": 1.958101533471558,
+      "grad_norm": 0.29969065319404764,
+      "learning_rate": 6.083099627835508e-06,
+      "loss": 0.4432,
+      "step": 11923
+    },
+    {
+      "epoch": 1.9582657606799008,
+      "grad_norm": 0.46090963146047326,
+      "learning_rate": 6.082660625984697e-06,
+      "loss": 0.4466,
+      "step": 11924
+    },
+    {
+      "epoch": 1.9584299878882434,
+      "grad_norm": 0.29142333619321953,
+      "learning_rate": 6.082221606944633e-06,
+      "loss": 0.4449,
+      "step": 11925
+    },
+    {
+      "epoch": 1.958594215096586,
+      "grad_norm": 0.38445693912408685,
+      "learning_rate": 6.081782570720085e-06,
+      "loss": 0.4441,
+      "step": 11926
+    },
+    {
+      "epoch": 1.958758442304929,
+      "grad_norm": 0.2926899702163581,
+      "learning_rate": 6.081343517315823e-06,
+      "loss": 0.4741,
+      "step": 11927
+    },
+    {
+      "epoch": 1.9589226695132718,
+      "grad_norm": 0.3185654549866373,
+      "learning_rate": 6.080904446736613e-06,
+      "loss": 0.4683,
+      "step": 11928
+    },
+    {
+      "epoch": 1.9590868967216144,
+      "grad_norm": 0.3257148447442194,
+      "learning_rate": 6.080465358987227e-06,
+      "loss": 0.4369,
+      "step": 11929
+    },
+    {
+      "epoch": 1.959251123929957,
+      "grad_norm": 0.34144365741472404,
+      "learning_rate": 6.0800262540724314e-06,
+      "loss": 0.4504,
+      "step": 11930
+    },
+    {
+      "epoch": 1.9594153511382997,
+      "grad_norm": 0.40757690940344427,
+      "learning_rate": 6.079587131996997e-06,
+      "loss": 0.4606,
+      "step": 11931
+    },
+    {
+      "epoch": 1.9595795783466425,
+      "grad_norm": 0.30314743296026664,
+      "learning_rate": 6.079147992765691e-06,
+      "loss": 0.4745,
+      "step": 11932
+    },
+    {
+      "epoch": 1.9597438055549854,
+      "grad_norm": 0.3476291582985087,
+      "learning_rate": 6.078708836383285e-06,
+      "loss": 0.4584,
+      "step": 11933
+    },
+    {
+      "epoch": 1.959908032763328,
+      "grad_norm": 0.38181630953661133,
+      "learning_rate": 6.078269662854546e-06,
+      "loss": 0.4507,
+      "step": 11934
+    },
+    {
+      "epoch": 1.9600722599716707,
+      "grad_norm": 0.3039837980681577,
+      "learning_rate": 6.077830472184249e-06,
+      "loss": 0.4496,
+      "step": 11935
+    },
+    {
+      "epoch": 1.9602364871800135,
+      "grad_norm": 0.3243794082669827,
+      "learning_rate": 6.0773912643771585e-06,
+      "loss": 0.4291,
+      "step": 11936
+    },
+    {
+      "epoch": 1.9604007143883564,
+      "grad_norm": 0.4193973744888972,
+      "learning_rate": 6.076952039438048e-06,
+      "loss": 0.4761,
+      "step": 11937
+    },
+    {
+      "epoch": 1.960564941596699,
+      "grad_norm": 0.2894100680354276,
+      "learning_rate": 6.076512797371685e-06,
+      "loss": 0.4309,
+      "step": 11938
+    },
+    {
+      "epoch": 1.9607291688050417,
+      "grad_norm": 0.47924954345632687,
+      "learning_rate": 6.0760735381828444e-06,
+      "loss": 0.439,
+      "step": 11939
+    },
+    {
+      "epoch": 1.9608933960133845,
+      "grad_norm": 0.3215352880588443,
+      "learning_rate": 6.075634261876292e-06,
+      "loss": 0.4399,
+      "step": 11940
+    },
+    {
+      "epoch": 1.9610576232217274,
+      "grad_norm": 0.3105394115840711,
+      "learning_rate": 6.0751949684568034e-06,
+      "loss": 0.4423,
+      "step": 11941
+    },
+    {
+      "epoch": 1.96122185043007,
+      "grad_norm": 0.3256268730700029,
+      "learning_rate": 6.074755657929146e-06,
+      "loss": 0.4395,
+      "step": 11942
+    },
+    {
+      "epoch": 1.9613860776384127,
+      "grad_norm": 0.4212632534343356,
+      "learning_rate": 6.074316330298094e-06,
+      "loss": 0.4591,
+      "step": 11943
+    },
+    {
+      "epoch": 1.9615503048467555,
+      "grad_norm": 0.33236517435930046,
+      "learning_rate": 6.073876985568417e-06,
+      "loss": 0.4351,
+      "step": 11944
+    },
+    {
+      "epoch": 1.9617145320550984,
+      "grad_norm": 0.3231407977837,
+      "learning_rate": 6.073437623744888e-06,
+      "loss": 0.4242,
+      "step": 11945
+    },
+    {
+      "epoch": 1.961878759263441,
+      "grad_norm": 0.30963714146020055,
+      "learning_rate": 6.072998244832279e-06,
+      "loss": 0.4376,
+      "step": 11946
+    },
+    {
+      "epoch": 1.9620429864717837,
+      "grad_norm": 0.29128255878835513,
+      "learning_rate": 6.072558848835359e-06,
+      "loss": 0.4531,
+      "step": 11947
+    },
+    {
+      "epoch": 1.9622072136801263,
+      "grad_norm": 0.2906945038194113,
+      "learning_rate": 6.0721194357589036e-06,
+      "loss": 0.4589,
+      "step": 11948
+    },
+    {
+      "epoch": 1.9623714408884692,
+      "grad_norm": 0.28236743251721147,
+      "learning_rate": 6.071680005607686e-06,
+      "loss": 0.4651,
+      "step": 11949
+    },
+    {
+      "epoch": 1.962535668096812,
+      "grad_norm": 0.3685667178520881,
+      "learning_rate": 6.071240558386477e-06,
+      "loss": 0.4517,
+      "step": 11950
+    },
+    {
+      "epoch": 1.9626998953051547,
+      "grad_norm": 0.2718554589683119,
+      "learning_rate": 6.0708010941000485e-06,
+      "loss": 0.4562,
+      "step": 11951
+    },
+    {
+      "epoch": 1.9628641225134973,
+      "grad_norm": 0.3077384180029517,
+      "learning_rate": 6.070361612753175e-06,
+      "loss": 0.4375,
+      "step": 11952
+    },
+    {
+      "epoch": 1.9630283497218401,
+      "grad_norm": 0.32225790357731776,
+      "learning_rate": 6.069922114350629e-06,
+      "loss": 0.4628,
+      "step": 11953
+    },
+    {
+      "epoch": 1.963192576930183,
+      "grad_norm": 0.4871176270330463,
+      "learning_rate": 6.069482598897186e-06,
+      "loss": 0.4521,
+      "step": 11954
+    },
+    {
+      "epoch": 1.9633568041385256,
+      "grad_norm": 0.4126252515018033,
+      "learning_rate": 6.069043066397615e-06,
+      "loss": 0.4402,
+      "step": 11955
+    },
+    {
+      "epoch": 1.9635210313468683,
+      "grad_norm": 0.3037151550263849,
+      "learning_rate": 6.0686035168566945e-06,
+      "loss": 0.4552,
+      "step": 11956
+    },
+    {
+      "epoch": 1.9636852585552111,
+      "grad_norm": 0.3326743916440129,
+      "learning_rate": 6.068163950279195e-06,
+      "loss": 0.437,
+      "step": 11957
+    },
+    {
+      "epoch": 1.963849485763554,
+      "grad_norm": 0.3131914650921411,
+      "learning_rate": 6.067724366669895e-06,
+      "loss": 0.4454,
+      "step": 11958
+    },
+    {
+      "epoch": 1.9640137129718966,
+      "grad_norm": 0.33370325669191686,
+      "learning_rate": 6.067284766033564e-06,
+      "loss": 0.4556,
+      "step": 11959
+    },
+    {
+      "epoch": 1.9641779401802393,
+      "grad_norm": 0.2634975212310298,
+      "learning_rate": 6.066845148374978e-06,
+      "loss": 0.4439,
+      "step": 11960
+    },
+    {
+      "epoch": 1.9643421673885821,
+      "grad_norm": 0.297116985250537,
+      "learning_rate": 6.066405513698912e-06,
+      "loss": 0.4372,
+      "step": 11961
+    },
+    {
+      "epoch": 1.964506394596925,
+      "grad_norm": 0.3084785855406754,
+      "learning_rate": 6.0659658620101424e-06,
+      "loss": 0.4395,
+      "step": 11962
+    },
+    {
+      "epoch": 1.9646706218052676,
+      "grad_norm": 0.30382296858647195,
+      "learning_rate": 6.065526193313442e-06,
+      "loss": 0.4436,
+      "step": 11963
+    },
+    {
+      "epoch": 1.9648348490136103,
+      "grad_norm": 0.47659789553897186,
+      "learning_rate": 6.065086507613587e-06,
+      "loss": 0.4531,
+      "step": 11964
+    },
+    {
+      "epoch": 1.964999076221953,
+      "grad_norm": 0.32839380600621654,
+      "learning_rate": 6.064646804915353e-06,
+      "loss": 0.4427,
+      "step": 11965
+    },
+    {
+      "epoch": 1.9651633034302958,
+      "grad_norm": 0.28526974160350493,
+      "learning_rate": 6.0642070852235156e-06,
+      "loss": 0.4546,
+      "step": 11966
+    },
+    {
+      "epoch": 1.9653275306386386,
+      "grad_norm": 0.2921253976160405,
+      "learning_rate": 6.063767348542849e-06,
+      "loss": 0.4477,
+      "step": 11967
+    },
+    {
+      "epoch": 1.9654917578469813,
+      "grad_norm": 0.32245870607633376,
+      "learning_rate": 6.06332759487813e-06,
+      "loss": 0.4463,
+      "step": 11968
+    },
+    {
+      "epoch": 1.965655985055324,
+      "grad_norm": 0.265974759218518,
+      "learning_rate": 6.062887824234138e-06,
+      "loss": 0.4475,
+      "step": 11969
+    },
+    {
+      "epoch": 1.9658202122636668,
+      "grad_norm": 0.3420749303053669,
+      "learning_rate": 6.0624480366156455e-06,
+      "loss": 0.4371,
+      "step": 11970
+    },
+    {
+      "epoch": 1.9659844394720096,
+      "grad_norm": 0.5955902754876532,
+      "learning_rate": 6.062008232027429e-06,
+      "loss": 0.4318,
+      "step": 11971
+    },
+    {
+      "epoch": 1.9661486666803523,
+      "grad_norm": 0.2654119467227501,
+      "learning_rate": 6.061568410474266e-06,
+      "loss": 0.4618,
+      "step": 11972
+    },
+    {
+      "epoch": 1.966312893888695,
+      "grad_norm": 0.3256763544307515,
+      "learning_rate": 6.061128571960935e-06,
+      "loss": 0.4478,
+      "step": 11973
+    },
+    {
+      "epoch": 1.9664771210970378,
+      "grad_norm": 0.38804030892894026,
+      "learning_rate": 6.06068871649221e-06,
+      "loss": 0.4599,
+      "step": 11974
+    },
+    {
+      "epoch": 1.9666413483053806,
+      "grad_norm": 0.3603884815671852,
+      "learning_rate": 6.060248844072872e-06,
+      "loss": 0.4407,
+      "step": 11975
+    },
+    {
+      "epoch": 1.9668055755137233,
+      "grad_norm": 0.29521900289850656,
+      "learning_rate": 6.059808954707696e-06,
+      "loss": 0.4553,
+      "step": 11976
+    },
+    {
+      "epoch": 1.966969802722066,
+      "grad_norm": 0.31907625861012234,
+      "learning_rate": 6.059369048401459e-06,
+      "loss": 0.4485,
+      "step": 11977
+    },
+    {
+      "epoch": 1.9671340299304088,
+      "grad_norm": 0.2827112303883396,
+      "learning_rate": 6.058929125158942e-06,
+      "loss": 0.4542,
+      "step": 11978
+    },
+    {
+      "epoch": 1.9672982571387516,
+      "grad_norm": 0.4095160662478101,
+      "learning_rate": 6.05848918498492e-06,
+      "loss": 0.4447,
+      "step": 11979
+    },
+    {
+      "epoch": 1.9674624843470943,
+      "grad_norm": 0.30726510403041185,
+      "learning_rate": 6.058049227884171e-06,
+      "loss": 0.4373,
+      "step": 11980
+    },
+    {
+      "epoch": 1.967626711555437,
+      "grad_norm": 0.3706977126954439,
+      "learning_rate": 6.057609253861475e-06,
+      "loss": 0.4433,
+      "step": 11981
+    },
+    {
+      "epoch": 1.9677909387637795,
+      "grad_norm": 0.27819231084749946,
+      "learning_rate": 6.057169262921609e-06,
+      "loss": 0.4415,
+      "step": 11982
+    },
+    {
+      "epoch": 1.9679551659721224,
+      "grad_norm": 0.2774976321074535,
+      "learning_rate": 6.056729255069356e-06,
+      "loss": 0.4415,
+      "step": 11983
+    },
+    {
+      "epoch": 1.9681193931804652,
+      "grad_norm": 0.3394973626327754,
+      "learning_rate": 6.05628923030949e-06,
+      "loss": 0.4256,
+      "step": 11984
+    },
+    {
+      "epoch": 1.9682836203888079,
+      "grad_norm": 0.270213655988594,
+      "learning_rate": 6.055849188646791e-06,
+      "loss": 0.4301,
+      "step": 11985
+    },
+    {
+      "epoch": 1.9684478475971505,
+      "grad_norm": 0.44524049804233634,
+      "learning_rate": 6.055409130086039e-06,
+      "loss": 0.4662,
+      "step": 11986
+    },
+    {
+      "epoch": 1.9686120748054934,
+      "grad_norm": 0.32487464715932335,
+      "learning_rate": 6.054969054632015e-06,
+      "loss": 0.4474,
+      "step": 11987
+    },
+    {
+      "epoch": 1.9687763020138362,
+      "grad_norm": 0.36859889874615925,
+      "learning_rate": 6.0545289622894956e-06,
+      "loss": 0.4406,
+      "step": 11988
+    },
+    {
+      "epoch": 1.9689405292221789,
+      "grad_norm": 0.29552264416651164,
+      "learning_rate": 6.054088853063263e-06,
+      "loss": 0.4396,
+      "step": 11989
+    },
+    {
+      "epoch": 1.9691047564305215,
+      "grad_norm": 0.2740422492527266,
+      "learning_rate": 6.053648726958096e-06,
+      "loss": 0.4517,
+      "step": 11990
+    },
+    {
+      "epoch": 1.9692689836388644,
+      "grad_norm": 0.32702865970003103,
+      "learning_rate": 6.053208583978776e-06,
+      "loss": 0.4542,
+      "step": 11991
+    },
+    {
+      "epoch": 1.9694332108472072,
+      "grad_norm": 0.30746129718076604,
+      "learning_rate": 6.052768424130081e-06,
+      "loss": 0.436,
+      "step": 11992
+    },
+    {
+      "epoch": 1.9695974380555499,
+      "grad_norm": 0.3372313936554249,
+      "learning_rate": 6.052328247416795e-06,
+      "loss": 0.4574,
+      "step": 11993
+    },
+    {
+      "epoch": 1.9697616652638925,
+      "grad_norm": 0.31184738193789135,
+      "learning_rate": 6.051888053843697e-06,
+      "loss": 0.4519,
+      "step": 11994
+    },
+    {
+      "epoch": 1.9699258924722354,
+      "grad_norm": 0.2956821322627038,
+      "learning_rate": 6.051447843415567e-06,
+      "loss": 0.4409,
+      "step": 11995
+    },
+    {
+      "epoch": 1.9700901196805782,
+      "grad_norm": 0.40843040125952373,
+      "learning_rate": 6.051007616137187e-06,
+      "loss": 0.4558,
+      "step": 11996
+    },
+    {
+      "epoch": 1.9702543468889209,
+      "grad_norm": 0.32893246578214913,
+      "learning_rate": 6.050567372013338e-06,
+      "loss": 0.4265,
+      "step": 11997
+    },
+    {
+      "epoch": 1.9704185740972635,
+      "grad_norm": 0.26744505739084257,
+      "learning_rate": 6.050127111048803e-06,
+      "loss": 0.4546,
+      "step": 11998
+    },
+    {
+      "epoch": 1.9705828013056061,
+      "grad_norm": 0.2781686616464339,
+      "learning_rate": 6.049686833248362e-06,
+      "loss": 0.4448,
+      "step": 11999
+    },
+    {
+      "epoch": 1.970747028513949,
+      "grad_norm": 0.27749506883908925,
+      "learning_rate": 6.049246538616796e-06,
+      "loss": 0.4388,
+      "step": 12000
+    },
+    {
+      "epoch": 1.9709112557222919,
+      "grad_norm": 0.39291413716423657,
+      "learning_rate": 6.048806227158889e-06,
+      "loss": 0.4289,
+      "step": 12001
+    },
+    {
+      "epoch": 1.9710754829306345,
+      "grad_norm": 0.2988992716204715,
+      "learning_rate": 6.048365898879423e-06,
+      "loss": 0.4484,
+      "step": 12002
+    },
+    {
+      "epoch": 1.9712397101389771,
+      "grad_norm": 0.30423634070698163,
+      "learning_rate": 6.04792555378318e-06,
+      "loss": 0.4284,
+      "step": 12003
+    },
+    {
+      "epoch": 1.97140393734732,
+      "grad_norm": 0.3669559441068953,
+      "learning_rate": 6.047485191874944e-06,
+      "loss": 0.4717,
+      "step": 12004
+    },
+    {
+      "epoch": 1.9715681645556629,
+      "grad_norm": 0.33458514236368003,
+      "learning_rate": 6.047044813159494e-06,
+      "loss": 0.4448,
+      "step": 12005
+    },
+    {
+      "epoch": 1.9717323917640055,
+      "grad_norm": 0.29073195184717887,
+      "learning_rate": 6.046604417641616e-06,
+      "loss": 0.4541,
+      "step": 12006
+    },
+    {
+      "epoch": 1.9718966189723481,
+      "grad_norm": 0.3558324342641227,
+      "learning_rate": 6.046164005326092e-06,
+      "loss": 0.44,
+      "step": 12007
+    },
+    {
+      "epoch": 1.972060846180691,
+      "grad_norm": 0.28499414277142565,
+      "learning_rate": 6.0457235762177065e-06,
+      "loss": 0.4573,
+      "step": 12008
+    },
+    {
+      "epoch": 1.9722250733890339,
+      "grad_norm": 0.32143312180520023,
+      "learning_rate": 6.045283130321242e-06,
+      "loss": 0.4505,
+      "step": 12009
+    },
+    {
+      "epoch": 1.9723893005973765,
+      "grad_norm": 0.3139482439274381,
+      "learning_rate": 6.044842667641482e-06,
+      "loss": 0.4546,
+      "step": 12010
+    },
+    {
+      "epoch": 1.9725535278057191,
+      "grad_norm": 0.36164038349410554,
+      "learning_rate": 6.04440218818321e-06,
+      "loss": 0.4533,
+      "step": 12011
+    },
+    {
+      "epoch": 1.972717755014062,
+      "grad_norm": 0.34017728293412075,
+      "learning_rate": 6.043961691951212e-06,
+      "loss": 0.4747,
+      "step": 12012
+    },
+    {
+      "epoch": 1.9728819822224049,
+      "grad_norm": 0.28030323013934577,
+      "learning_rate": 6.04352117895027e-06,
+      "loss": 0.4353,
+      "step": 12013
+    },
+    {
+      "epoch": 1.9730462094307475,
+      "grad_norm": 0.36739679215647425,
+      "learning_rate": 6.043080649185171e-06,
+      "loss": 0.4448,
+      "step": 12014
+    },
+    {
+      "epoch": 1.9732104366390901,
+      "grad_norm": 0.3699289824111152,
+      "learning_rate": 6.042640102660695e-06,
+      "loss": 0.442,
+      "step": 12015
+    },
+    {
+      "epoch": 1.9733746638474328,
+      "grad_norm": 0.29453155207049314,
+      "learning_rate": 6.042199539381633e-06,
+      "loss": 0.4512,
+      "step": 12016
+    },
+    {
+      "epoch": 1.9735388910557756,
+      "grad_norm": 0.3369125096658498,
+      "learning_rate": 6.041758959352764e-06,
+      "loss": 0.4753,
+      "step": 12017
+    },
+    {
+      "epoch": 1.9737031182641185,
+      "grad_norm": 0.359560199304719,
+      "learning_rate": 6.041318362578878e-06,
+      "loss": 0.4441,
+      "step": 12018
+    },
+    {
+      "epoch": 1.9738673454724611,
+      "grad_norm": 0.4074514051910816,
+      "learning_rate": 6.040877749064757e-06,
+      "loss": 0.4534,
+      "step": 12019
+    },
+    {
+      "epoch": 1.9740315726808038,
+      "grad_norm": 0.3280832276122813,
+      "learning_rate": 6.040437118815187e-06,
+      "loss": 0.4385,
+      "step": 12020
+    },
+    {
+      "epoch": 1.9741957998891466,
+      "grad_norm": 0.29592394541807415,
+      "learning_rate": 6.039996471834956e-06,
+      "loss": 0.4272,
+      "step": 12021
+    },
+    {
+      "epoch": 1.9743600270974895,
+      "grad_norm": 0.4068099660831585,
+      "learning_rate": 6.039555808128848e-06,
+      "loss": 0.4497,
+      "step": 12022
+    },
+    {
+      "epoch": 1.9745242543058321,
+      "grad_norm": 0.40547477133364795,
+      "learning_rate": 6.039115127701649e-06,
+      "loss": 0.4596,
+      "step": 12023
+    },
+    {
+      "epoch": 1.9746884815141748,
+      "grad_norm": 0.33863995062257035,
+      "learning_rate": 6.038674430558144e-06,
+      "loss": 0.4527,
+      "step": 12024
+    },
+    {
+      "epoch": 1.9748527087225176,
+      "grad_norm": 0.2859546787151384,
+      "learning_rate": 6.038233716703122e-06,
+      "loss": 0.4227,
+      "step": 12025
+    },
+    {
+      "epoch": 1.9750169359308605,
+      "grad_norm": 0.43978961671594363,
+      "learning_rate": 6.037792986141368e-06,
+      "loss": 0.4293,
+      "step": 12026
+    },
+    {
+      "epoch": 1.975181163139203,
+      "grad_norm": 0.32042395347082325,
+      "learning_rate": 6.037352238877669e-06,
+      "loss": 0.4436,
+      "step": 12027
+    },
+    {
+      "epoch": 1.9753453903475457,
+      "grad_norm": 0.45653170556439265,
+      "learning_rate": 6.036911474916813e-06,
+      "loss": 0.4355,
+      "step": 12028
+    },
+    {
+      "epoch": 1.9755096175558886,
+      "grad_norm": 0.31606632423937275,
+      "learning_rate": 6.036470694263585e-06,
+      "loss": 0.4571,
+      "step": 12029
+    },
+    {
+      "epoch": 1.9756738447642315,
+      "grad_norm": 0.36776567749481,
+      "learning_rate": 6.036029896922774e-06,
+      "loss": 0.4325,
+      "step": 12030
+    },
+    {
+      "epoch": 1.975838071972574,
+      "grad_norm": 0.2897838666344959,
+      "learning_rate": 6.035589082899168e-06,
+      "loss": 0.4512,
+      "step": 12031
+    },
+    {
+      "epoch": 1.9760022991809167,
+      "grad_norm": 0.34939504492231116,
+      "learning_rate": 6.035148252197554e-06,
+      "loss": 0.4403,
+      "step": 12032
+    },
+    {
+      "epoch": 1.9761665263892594,
+      "grad_norm": 0.37178263437531295,
+      "learning_rate": 6.034707404822718e-06,
+      "loss": 0.4355,
+      "step": 12033
+    },
+    {
+      "epoch": 1.9763307535976022,
+      "grad_norm": 0.3404844380988809,
+      "learning_rate": 6.034266540779451e-06,
+      "loss": 0.465,
+      "step": 12034
+    },
+    {
+      "epoch": 1.976494980805945,
+      "grad_norm": 0.35425656017271817,
+      "learning_rate": 6.033825660072538e-06,
+      "loss": 0.4466,
+      "step": 12035
+    },
+    {
+      "epoch": 1.9766592080142877,
+      "grad_norm": 0.3120757128379207,
+      "learning_rate": 6.033384762706772e-06,
+      "loss": 0.422,
+      "step": 12036
+    },
+    {
+      "epoch": 1.9768234352226304,
+      "grad_norm": 0.35814186378850754,
+      "learning_rate": 6.032943848686938e-06,
+      "loss": 0.4438,
+      "step": 12037
+    },
+    {
+      "epoch": 1.9769876624309732,
+      "grad_norm": 0.31766052787014015,
+      "learning_rate": 6.032502918017823e-06,
+      "loss": 0.4316,
+      "step": 12038
+    },
+    {
+      "epoch": 1.977151889639316,
+      "grad_norm": 0.3316287841188254,
+      "learning_rate": 6.032061970704221e-06,
+      "loss": 0.4413,
+      "step": 12039
+    },
+    {
+      "epoch": 1.9773161168476587,
+      "grad_norm": 0.29366943908554144,
+      "learning_rate": 6.031621006750918e-06,
+      "loss": 0.4524,
+      "step": 12040
+    },
+    {
+      "epoch": 1.9774803440560014,
+      "grad_norm": 0.3365778223058008,
+      "learning_rate": 6.031180026162704e-06,
+      "loss": 0.4469,
+      "step": 12041
+    },
+    {
+      "epoch": 1.9776445712643442,
+      "grad_norm": 0.273328767172499,
+      "learning_rate": 6.030739028944369e-06,
+      "loss": 0.4376,
+      "step": 12042
+    },
+    {
+      "epoch": 1.977808798472687,
+      "grad_norm": 0.34032479412588085,
+      "learning_rate": 6.030298015100702e-06,
+      "loss": 0.4383,
+      "step": 12043
+    },
+    {
+      "epoch": 1.9779730256810297,
+      "grad_norm": 0.29204890209099327,
+      "learning_rate": 6.029856984636491e-06,
+      "loss": 0.4493,
+      "step": 12044
+    },
+    {
+      "epoch": 1.9781372528893724,
+      "grad_norm": 0.31070441585017955,
+      "learning_rate": 6.02941593755653e-06,
+      "loss": 0.4339,
+      "step": 12045
+    },
+    {
+      "epoch": 1.9783014800977152,
+      "grad_norm": 0.34355821495186206,
+      "learning_rate": 6.028974873865607e-06,
+      "loss": 0.4497,
+      "step": 12046
+    },
+    {
+      "epoch": 1.978465707306058,
+      "grad_norm": 0.37841270289295065,
+      "learning_rate": 6.0285337935685125e-06,
+      "loss": 0.4667,
+      "step": 12047
+    },
+    {
+      "epoch": 1.9786299345144007,
+      "grad_norm": 0.2880315830387975,
+      "learning_rate": 6.028092696670037e-06,
+      "loss": 0.446,
+      "step": 12048
+    },
+    {
+      "epoch": 1.9787941617227434,
+      "grad_norm": 0.2860235218704819,
+      "learning_rate": 6.02765158317497e-06,
+      "loss": 0.4548,
+      "step": 12049
+    },
+    {
+      "epoch": 1.978958388931086,
+      "grad_norm": 0.6820882451984894,
+      "learning_rate": 6.027210453088106e-06,
+      "loss": 0.4264,
+      "step": 12050
+    },
+    {
+      "epoch": 1.9791226161394289,
+      "grad_norm": 0.3449763478453752,
+      "learning_rate": 6.026769306414234e-06,
+      "loss": 0.4534,
+      "step": 12051
+    },
+    {
+      "epoch": 1.9792868433477717,
+      "grad_norm": 0.31103476736921737,
+      "learning_rate": 6.026328143158143e-06,
+      "loss": 0.4565,
+      "step": 12052
+    },
+    {
+      "epoch": 1.9794510705561144,
+      "grad_norm": 0.297356778106612,
+      "learning_rate": 6.0258869633246275e-06,
+      "loss": 0.4705,
+      "step": 12053
+    },
+    {
+      "epoch": 1.979615297764457,
+      "grad_norm": 0.2549611004267377,
+      "learning_rate": 6.0254457669184795e-06,
+      "loss": 0.4406,
+      "step": 12054
+    },
+    {
+      "epoch": 1.9797795249727999,
+      "grad_norm": 0.3368960762751481,
+      "learning_rate": 6.025004553944488e-06,
+      "loss": 0.4179,
+      "step": 12055
+    },
+    {
+      "epoch": 1.9799437521811427,
+      "grad_norm": 0.28991698335346405,
+      "learning_rate": 6.0245633244074485e-06,
+      "loss": 0.4597,
+      "step": 12056
+    },
+    {
+      "epoch": 1.9801079793894854,
+      "grad_norm": 0.3258666563489521,
+      "learning_rate": 6.02412207831215e-06,
+      "loss": 0.4308,
+      "step": 12057
+    },
+    {
+      "epoch": 1.980272206597828,
+      "grad_norm": 0.3596563049725067,
+      "learning_rate": 6.023680815663386e-06,
+      "loss": 0.467,
+      "step": 12058
+    },
+    {
+      "epoch": 1.9804364338061708,
+      "grad_norm": 0.3375615413381802,
+      "learning_rate": 6.02323953646595e-06,
+      "loss": 0.4426,
+      "step": 12059
+    },
+    {
+      "epoch": 1.9806006610145137,
+      "grad_norm": 0.2701265231327295,
+      "learning_rate": 6.022798240724633e-06,
+      "loss": 0.4414,
+      "step": 12060
+    },
+    {
+      "epoch": 1.9807648882228563,
+      "grad_norm": 0.327944132720733,
+      "learning_rate": 6.0223569284442296e-06,
+      "loss": 0.46,
+      "step": 12061
+    },
+    {
+      "epoch": 1.980929115431199,
+      "grad_norm": 0.28050418382709885,
+      "learning_rate": 6.021915599629533e-06,
+      "loss": 0.4757,
+      "step": 12062
+    },
+    {
+      "epoch": 1.9810933426395418,
+      "grad_norm": 0.3169619359243772,
+      "learning_rate": 6.021474254285334e-06,
+      "loss": 0.4506,
+      "step": 12063
+    },
+    {
+      "epoch": 1.9812575698478847,
+      "grad_norm": 0.2981630396884761,
+      "learning_rate": 6.021032892416428e-06,
+      "loss": 0.4647,
+      "step": 12064
+    },
+    {
+      "epoch": 1.9814217970562273,
+      "grad_norm": 0.32351176269873505,
+      "learning_rate": 6.020591514027608e-06,
+      "loss": 0.4581,
+      "step": 12065
+    },
+    {
+      "epoch": 1.98158602426457,
+      "grad_norm": 0.4231399008420313,
+      "learning_rate": 6.02015011912367e-06,
+      "loss": 0.4458,
+      "step": 12066
+    },
+    {
+      "epoch": 1.9817502514729126,
+      "grad_norm": 0.3037909970679311,
+      "learning_rate": 6.019708707709406e-06,
+      "loss": 0.4493,
+      "step": 12067
+    },
+    {
+      "epoch": 1.9819144786812555,
+      "grad_norm": 0.27711451234107276,
+      "learning_rate": 6.019267279789607e-06,
+      "loss": 0.4367,
+      "step": 12068
+    },
+    {
+      "epoch": 1.9820787058895983,
+      "grad_norm": 0.26567383056133026,
+      "learning_rate": 6.018825835369073e-06,
+      "loss": 0.4137,
+      "step": 12069
+    },
+    {
+      "epoch": 1.982242933097941,
+      "grad_norm": 0.3748808611920089,
+      "learning_rate": 6.018384374452596e-06,
+      "loss": 0.4715,
+      "step": 12070
+    },
+    {
+      "epoch": 1.9824071603062836,
+      "grad_norm": 0.381151333801941,
+      "learning_rate": 6.017942897044971e-06,
+      "loss": 0.4562,
+      "step": 12071
+    },
+    {
+      "epoch": 1.9825713875146265,
+      "grad_norm": 0.4471531506882593,
+      "learning_rate": 6.017501403150992e-06,
+      "loss": 0.4493,
+      "step": 12072
+    },
+    {
+      "epoch": 1.9827356147229693,
+      "grad_norm": 0.3843248058077205,
+      "learning_rate": 6.017059892775455e-06,
+      "loss": 0.4458,
+      "step": 12073
+    },
+    {
+      "epoch": 1.982899841931312,
+      "grad_norm": 0.4006785253213048,
+      "learning_rate": 6.016618365923154e-06,
+      "loss": 0.4549,
+      "step": 12074
+    },
+    {
+      "epoch": 1.9830640691396546,
+      "grad_norm": 0.40463613359960876,
+      "learning_rate": 6.016176822598886e-06,
+      "loss": 0.4634,
+      "step": 12075
+    },
+    {
+      "epoch": 1.9832282963479975,
+      "grad_norm": 0.2790785754162893,
+      "learning_rate": 6.015735262807448e-06,
+      "loss": 0.4595,
+      "step": 12076
+    },
+    {
+      "epoch": 1.9833925235563403,
+      "grad_norm": 0.3065794759268286,
+      "learning_rate": 6.015293686553632e-06,
+      "loss": 0.4687,
+      "step": 12077
+    },
+    {
+      "epoch": 1.983556750764683,
+      "grad_norm": 0.4796119205886854,
+      "learning_rate": 6.014852093842236e-06,
+      "loss": 0.449,
+      "step": 12078
+    },
+    {
+      "epoch": 1.9837209779730256,
+      "grad_norm": 0.32901033246316697,
+      "learning_rate": 6.014410484678055e-06,
+      "loss": 0.444,
+      "step": 12079
+    },
+    {
+      "epoch": 1.9838852051813685,
+      "grad_norm": 0.31620813737993725,
+      "learning_rate": 6.0139688590658875e-06,
+      "loss": 0.4543,
+      "step": 12080
+    },
+    {
+      "epoch": 1.9840494323897113,
+      "grad_norm": 0.5304261684883916,
+      "learning_rate": 6.013527217010528e-06,
+      "loss": 0.4352,
+      "step": 12081
+    },
+    {
+      "epoch": 1.984213659598054,
+      "grad_norm": 0.33483212395030826,
+      "learning_rate": 6.0130855585167735e-06,
+      "loss": 0.4287,
+      "step": 12082
+    },
+    {
+      "epoch": 1.9843778868063966,
+      "grad_norm": 0.31516123536811785,
+      "learning_rate": 6.012643883589422e-06,
+      "loss": 0.4534,
+      "step": 12083
+    },
+    {
+      "epoch": 1.9845421140147392,
+      "grad_norm": 0.36557989364427224,
+      "learning_rate": 6.012202192233269e-06,
+      "loss": 0.4396,
+      "step": 12084
+    },
+    {
+      "epoch": 1.984706341223082,
+      "grad_norm": 0.3248087899223682,
+      "learning_rate": 6.011760484453113e-06,
+      "loss": 0.4308,
+      "step": 12085
+    },
+    {
+      "epoch": 1.984870568431425,
+      "grad_norm": 0.33659809991010253,
+      "learning_rate": 6.01131876025375e-06,
+      "loss": 0.4239,
+      "step": 12086
+    },
+    {
+      "epoch": 1.9850347956397676,
+      "grad_norm": 0.4750655674070637,
+      "learning_rate": 6.010877019639978e-06,
+      "loss": 0.4417,
+      "step": 12087
+    },
+    {
+      "epoch": 1.9851990228481102,
+      "grad_norm": 0.7893302234320693,
+      "learning_rate": 6.010435262616595e-06,
+      "loss": 0.4511,
+      "step": 12088
+    },
+    {
+      "epoch": 1.985363250056453,
+      "grad_norm": 0.2919750160730782,
+      "learning_rate": 6.009993489188401e-06,
+      "loss": 0.4235,
+      "step": 12089
+    },
+    {
+      "epoch": 1.985527477264796,
+      "grad_norm": 0.269083122192506,
+      "learning_rate": 6.00955169936019e-06,
+      "loss": 0.4384,
+      "step": 12090
+    },
+    {
+      "epoch": 1.9856917044731386,
+      "grad_norm": 0.3046254431436656,
+      "learning_rate": 6.009109893136764e-06,
+      "loss": 0.4439,
+      "step": 12091
+    },
+    {
+      "epoch": 1.9858559316814812,
+      "grad_norm": 0.34705997935606975,
+      "learning_rate": 6.0086680705229185e-06,
+      "loss": 0.4656,
+      "step": 12092
+    },
+    {
+      "epoch": 1.986020158889824,
+      "grad_norm": 0.3014169128468884,
+      "learning_rate": 6.008226231523454e-06,
+      "loss": 0.4425,
+      "step": 12093
+    },
+    {
+      "epoch": 1.986184386098167,
+      "grad_norm": 0.3203316019921863,
+      "learning_rate": 6.007784376143168e-06,
+      "loss": 0.4604,
+      "step": 12094
+    },
+    {
+      "epoch": 1.9863486133065096,
+      "grad_norm": 0.28211784166140474,
+      "learning_rate": 6.007342504386861e-06,
+      "loss": 0.4461,
+      "step": 12095
+    },
+    {
+      "epoch": 1.9865128405148522,
+      "grad_norm": 0.713397223611457,
+      "learning_rate": 6.0069006162593316e-06,
+      "loss": 0.4712,
+      "step": 12096
+    },
+    {
+      "epoch": 1.986677067723195,
+      "grad_norm": 0.32804549480165146,
+      "learning_rate": 6.006458711765378e-06,
+      "loss": 0.4464,
+      "step": 12097
+    },
+    {
+      "epoch": 1.986841294931538,
+      "grad_norm": 0.4390495685818857,
+      "learning_rate": 6.0060167909098005e-06,
+      "loss": 0.4532,
+      "step": 12098
+    },
+    {
+      "epoch": 1.9870055221398806,
+      "grad_norm": 0.29066432280196225,
+      "learning_rate": 6.005574853697399e-06,
+      "loss": 0.4299,
+      "step": 12099
+    },
+    {
+      "epoch": 1.9871697493482232,
+      "grad_norm": 0.27970517847861814,
+      "learning_rate": 6.005132900132976e-06,
+      "loss": 0.451,
+      "step": 12100
+    },
+    {
+      "epoch": 1.9873339765565659,
+      "grad_norm": 0.2993981409645384,
+      "learning_rate": 6.004690930221326e-06,
+      "loss": 0.449,
+      "step": 12101
+    },
+    {
+      "epoch": 1.9874982037649087,
+      "grad_norm": 0.4881082910339583,
+      "learning_rate": 6.0042489439672526e-06,
+      "loss": 0.4428,
+      "step": 12102
+    },
+    {
+      "epoch": 1.9876624309732516,
+      "grad_norm": 0.35593125919821117,
+      "learning_rate": 6.0038069413755554e-06,
+      "loss": 0.4737,
+      "step": 12103
+    },
+    {
+      "epoch": 1.9878266581815942,
+      "grad_norm": 0.3404574977066621,
+      "learning_rate": 6.003364922451035e-06,
+      "loss": 0.4371,
+      "step": 12104
+    },
+    {
+      "epoch": 1.9879908853899368,
+      "grad_norm": 0.35843795271388007,
+      "learning_rate": 6.002922887198494e-06,
+      "loss": 0.4464,
+      "step": 12105
+    },
+    {
+      "epoch": 1.9881551125982797,
+      "grad_norm": 0.278424406244817,
+      "learning_rate": 6.002480835622731e-06,
+      "loss": 0.4358,
+      "step": 12106
+    },
+    {
+      "epoch": 1.9883193398066226,
+      "grad_norm": 0.3792249206543696,
+      "learning_rate": 6.0020387677285474e-06,
+      "loss": 0.4479,
+      "step": 12107
+    },
+    {
+      "epoch": 1.9884835670149652,
+      "grad_norm": 0.3097187849705982,
+      "learning_rate": 6.001596683520746e-06,
+      "loss": 0.4455,
+      "step": 12108
+    },
+    {
+      "epoch": 1.9886477942233078,
+      "grad_norm": 0.3912751913987081,
+      "learning_rate": 6.001154583004126e-06,
+      "loss": 0.4589,
+      "step": 12109
+    },
+    {
+      "epoch": 1.9888120214316507,
+      "grad_norm": 0.36773299409138227,
+      "learning_rate": 6.000712466183492e-06,
+      "loss": 0.4546,
+      "step": 12110
+    },
+    {
+      "epoch": 1.9889762486399936,
+      "grad_norm": 0.25200881130078995,
+      "learning_rate": 6.000270333063643e-06,
+      "loss": 0.4221,
+      "step": 12111
+    },
+    {
+      "epoch": 1.9891404758483362,
+      "grad_norm": 0.41745494296022206,
+      "learning_rate": 5.999828183649382e-06,
+      "loss": 0.4375,
+      "step": 12112
+    },
+    {
+      "epoch": 1.9893047030566788,
+      "grad_norm": 0.41119193520709213,
+      "learning_rate": 5.999386017945512e-06,
+      "loss": 0.4527,
+      "step": 12113
+    },
+    {
+      "epoch": 1.9894689302650217,
+      "grad_norm": 0.3524347997142921,
+      "learning_rate": 5.998943835956833e-06,
+      "loss": 0.4477,
+      "step": 12114
+    },
+    {
+      "epoch": 1.9896331574733646,
+      "grad_norm": 0.6992056215375053,
+      "learning_rate": 5.998501637688151e-06,
+      "loss": 0.4373,
+      "step": 12115
+    },
+    {
+      "epoch": 1.9897973846817072,
+      "grad_norm": 0.35753146495246646,
+      "learning_rate": 5.998059423144266e-06,
+      "loss": 0.4239,
+      "step": 12116
+    },
+    {
+      "epoch": 1.9899616118900498,
+      "grad_norm": 0.305443312300451,
+      "learning_rate": 5.9976171923299825e-06,
+      "loss": 0.4599,
+      "step": 12117
+    },
+    {
+      "epoch": 1.9901258390983925,
+      "grad_norm": 0.3468952965866211,
+      "learning_rate": 5.997174945250102e-06,
+      "loss": 0.4367,
+      "step": 12118
+    },
+    {
+      "epoch": 1.9902900663067353,
+      "grad_norm": 0.30992685733369446,
+      "learning_rate": 5.996732681909429e-06,
+      "loss": 0.4238,
+      "step": 12119
+    },
+    {
+      "epoch": 1.9904542935150782,
+      "grad_norm": 0.3089277099150746,
+      "learning_rate": 5.9962904023127654e-06,
+      "loss": 0.4668,
+      "step": 12120
+    },
+    {
+      "epoch": 1.9906185207234208,
+      "grad_norm": 0.4060973659557495,
+      "learning_rate": 5.995848106464918e-06,
+      "loss": 0.4341,
+      "step": 12121
+    },
+    {
+      "epoch": 1.9907827479317635,
+      "grad_norm": 0.3343300894937679,
+      "learning_rate": 5.995405794370687e-06,
+      "loss": 0.4388,
+      "step": 12122
+    },
+    {
+      "epoch": 1.9909469751401063,
+      "grad_norm": 0.4022894904900554,
+      "learning_rate": 5.994963466034877e-06,
+      "loss": 0.4604,
+      "step": 12123
+    },
+    {
+      "epoch": 1.9911112023484492,
+      "grad_norm": 0.2809421394977338,
+      "learning_rate": 5.994521121462295e-06,
+      "loss": 0.4306,
+      "step": 12124
+    },
+    {
+      "epoch": 1.9912754295567918,
+      "grad_norm": 0.2999392110190069,
+      "learning_rate": 5.994078760657742e-06,
+      "loss": 0.4521,
+      "step": 12125
+    },
+    {
+      "epoch": 1.9914396567651345,
+      "grad_norm": 0.3439605067320684,
+      "learning_rate": 5.9936363836260235e-06,
+      "loss": 0.4559,
+      "step": 12126
+    },
+    {
+      "epoch": 1.9916038839734773,
+      "grad_norm": 0.35520361022230595,
+      "learning_rate": 5.993193990371945e-06,
+      "loss": 0.448,
+      "step": 12127
+    },
+    {
+      "epoch": 1.9917681111818202,
+      "grad_norm": 0.3873609826678473,
+      "learning_rate": 5.99275158090031e-06,
+      "loss": 0.4431,
+      "step": 12128
+    },
+    {
+      "epoch": 1.9919323383901628,
+      "grad_norm": 0.623525798670634,
+      "learning_rate": 5.9923091552159244e-06,
+      "loss": 0.4399,
+      "step": 12129
+    },
+    {
+      "epoch": 1.9920965655985055,
+      "grad_norm": 0.375059399085205,
+      "learning_rate": 5.9918667133235946e-06,
+      "loss": 0.4486,
+      "step": 12130
+    },
+    {
+      "epoch": 1.9922607928068483,
+      "grad_norm": 0.4259468484416419,
+      "learning_rate": 5.991424255228122e-06,
+      "loss": 0.4604,
+      "step": 12131
+    },
+    {
+      "epoch": 1.9924250200151912,
+      "grad_norm": 0.48872792746183463,
+      "learning_rate": 5.990981780934316e-06,
+      "loss": 0.4451,
+      "step": 12132
+    },
+    {
+      "epoch": 1.9925892472235338,
+      "grad_norm": 0.3400954850191665,
+      "learning_rate": 5.990539290446981e-06,
+      "loss": 0.4511,
+      "step": 12133
+    },
+    {
+      "epoch": 1.9927534744318764,
+      "grad_norm": 0.3584423281926243,
+      "learning_rate": 5.990096783770923e-06,
+      "loss": 0.4383,
+      "step": 12134
+    },
+    {
+      "epoch": 1.992917701640219,
+      "grad_norm": 0.34324978031707004,
+      "learning_rate": 5.989654260910947e-06,
+      "loss": 0.442,
+      "step": 12135
+    },
+    {
+      "epoch": 1.993081928848562,
+      "grad_norm": 0.32954134231022136,
+      "learning_rate": 5.98921172187186e-06,
+      "loss": 0.4543,
+      "step": 12136
+    },
+    {
+      "epoch": 1.9932461560569048,
+      "grad_norm": 0.392536390634084,
+      "learning_rate": 5.9887691666584685e-06,
+      "loss": 0.4678,
+      "step": 12137
+    },
+    {
+      "epoch": 1.9934103832652474,
+      "grad_norm": 0.2785571757259487,
+      "learning_rate": 5.9883265952755796e-06,
+      "loss": 0.4437,
+      "step": 12138
+    },
+    {
+      "epoch": 1.99357461047359,
+      "grad_norm": 0.3650211841002478,
+      "learning_rate": 5.987884007728001e-06,
+      "loss": 0.4457,
+      "step": 12139
+    },
+    {
+      "epoch": 1.993738837681933,
+      "grad_norm": 0.5938490366217879,
+      "learning_rate": 5.987441404020537e-06,
+      "loss": 0.4364,
+      "step": 12140
+    },
+    {
+      "epoch": 1.9939030648902758,
+      "grad_norm": 0.3330731764741555,
+      "learning_rate": 5.986998784157995e-06,
+      "loss": 0.4222,
+      "step": 12141
+    },
+    {
+      "epoch": 1.9940672920986184,
+      "grad_norm": 0.3502325402671695,
+      "learning_rate": 5.986556148145183e-06,
+      "loss": 0.435,
+      "step": 12142
+    },
+    {
+      "epoch": 1.994231519306961,
+      "grad_norm": 0.3442202918445196,
+      "learning_rate": 5.98611349598691e-06,
+      "loss": 0.4507,
+      "step": 12143
+    },
+    {
+      "epoch": 1.994395746515304,
+      "grad_norm": 0.4005337023377969,
+      "learning_rate": 5.985670827687983e-06,
+      "loss": 0.4468,
+      "step": 12144
+    },
+    {
+      "epoch": 1.9945599737236468,
+      "grad_norm": 0.29403675952349245,
+      "learning_rate": 5.985228143253207e-06,
+      "loss": 0.4617,
+      "step": 12145
+    },
+    {
+      "epoch": 1.9947242009319894,
+      "grad_norm": 0.430171744081949,
+      "learning_rate": 5.984785442687394e-06,
+      "loss": 0.4428,
+      "step": 12146
+    },
+    {
+      "epoch": 1.994888428140332,
+      "grad_norm": 0.7144987121760876,
+      "learning_rate": 5.984342725995349e-06,
+      "loss": 0.4536,
+      "step": 12147
+    },
+    {
+      "epoch": 1.995052655348675,
+      "grad_norm": 0.5575791795273514,
+      "learning_rate": 5.9838999931818816e-06,
+      "loss": 0.4447,
+      "step": 12148
+    },
+    {
+      "epoch": 1.9952168825570178,
+      "grad_norm": 0.3103253306063338,
+      "learning_rate": 5.983457244251801e-06,
+      "loss": 0.4482,
+      "step": 12149
+    },
+    {
+      "epoch": 1.9953811097653604,
+      "grad_norm": 0.306105882253974,
+      "learning_rate": 5.983014479209914e-06,
+      "loss": 0.4603,
+      "step": 12150
+    },
+    {
+      "epoch": 1.995545336973703,
+      "grad_norm": 0.3256489586607816,
+      "learning_rate": 5.982571698061033e-06,
+      "loss": 0.4502,
+      "step": 12151
+    },
+    {
+      "epoch": 1.9957095641820457,
+      "grad_norm": 0.30675263299161515,
+      "learning_rate": 5.982128900809962e-06,
+      "loss": 0.4408,
+      "step": 12152
+    },
+    {
+      "epoch": 1.9958737913903886,
+      "grad_norm": 0.3979035965681152,
+      "learning_rate": 5.981686087461514e-06,
+      "loss": 0.4338,
+      "step": 12153
+    },
+    {
+      "epoch": 1.9960380185987314,
+      "grad_norm": 0.5714962273931158,
+      "learning_rate": 5.981243258020498e-06,
+      "loss": 0.473,
+      "step": 12154
+    },
+    {
+      "epoch": 1.996202245807074,
+      "grad_norm": 0.5233640080963108,
+      "learning_rate": 5.980800412491722e-06,
+      "loss": 0.4452,
+      "step": 12155
+    },
+    {
+      "epoch": 1.9963664730154167,
+      "grad_norm": 0.5095439866164315,
+      "learning_rate": 5.980357550879997e-06,
+      "loss": 0.4404,
+      "step": 12156
+    },
+    {
+      "epoch": 1.9965307002237596,
+      "grad_norm": 0.34135588876267836,
+      "learning_rate": 5.979914673190132e-06,
+      "loss": 0.4641,
+      "step": 12157
+    },
+    {
+      "epoch": 1.9966949274321024,
+      "grad_norm": 0.3166966051628447,
+      "learning_rate": 5.979471779426938e-06,
+      "loss": 0.4463,
+      "step": 12158
+    },
+    {
+      "epoch": 1.996859154640445,
+      "grad_norm": 0.3534752492741154,
+      "learning_rate": 5.9790288695952256e-06,
+      "loss": 0.4179,
+      "step": 12159
+    },
+    {
+      "epoch": 1.9970233818487877,
+      "grad_norm": 0.2803029304863608,
+      "learning_rate": 5.9785859436998035e-06,
+      "loss": 0.438,
+      "step": 12160
+    },
+    {
+      "epoch": 1.9971876090571306,
+      "grad_norm": 0.3739910815165364,
+      "learning_rate": 5.978143001745484e-06,
+      "loss": 0.4254,
+      "step": 12161
+    },
+    {
+      "epoch": 1.9973518362654734,
+      "grad_norm": 0.33129160415655884,
+      "learning_rate": 5.977700043737075e-06,
+      "loss": 0.4323,
+      "step": 12162
+    },
+    {
+      "epoch": 1.997516063473816,
+      "grad_norm": 0.31774150789673616,
+      "learning_rate": 5.977257069679393e-06,
+      "loss": 0.443,
+      "step": 12163
+    },
+    {
+      "epoch": 1.9976802906821587,
+      "grad_norm": 0.3477407627974902,
+      "learning_rate": 5.9768140795772445e-06,
+      "loss": 0.4463,
+      "step": 12164
+    },
+    {
+      "epoch": 1.9978445178905015,
+      "grad_norm": 0.3209257252688937,
+      "learning_rate": 5.9763710734354415e-06,
+      "loss": 0.4492,
+      "step": 12165
+    },
+    {
+      "epoch": 1.9980087450988444,
+      "grad_norm": 0.32082689954088633,
+      "learning_rate": 5.9759280512587966e-06,
+      "loss": 0.4436,
+      "step": 12166
+    },
+    {
+      "epoch": 1.998172972307187,
+      "grad_norm": 0.35698917095820754,
+      "learning_rate": 5.975485013052122e-06,
+      "loss": 0.4476,
+      "step": 12167
+    },
+    {
+      "epoch": 1.9983371995155297,
+      "grad_norm": 0.3286716694844807,
+      "learning_rate": 5.975041958820227e-06,
+      "loss": 0.4326,
+      "step": 12168
+    },
+    {
+      "epoch": 1.9985014267238723,
+      "grad_norm": 0.42593190197587716,
+      "learning_rate": 5.974598888567925e-06,
+      "loss": 0.4603,
+      "step": 12169
+    },
+    {
+      "epoch": 1.9986656539322152,
+      "grad_norm": 0.3397923984316032,
+      "learning_rate": 5.974155802300027e-06,
+      "loss": 0.4391,
+      "step": 12170
+    },
+    {
+      "epoch": 1.998829881140558,
+      "grad_norm": 0.4923490616754908,
+      "learning_rate": 5.97371270002135e-06,
+      "loss": 0.464,
+      "step": 12171
+    },
+    {
+      "epoch": 1.9989941083489007,
+      "grad_norm": 0.3340119955565463,
+      "learning_rate": 5.973269581736701e-06,
+      "loss": 0.4498,
+      "step": 12172
+    },
+    {
+      "epoch": 1.9991583355572433,
+      "grad_norm": 0.4640454105520234,
+      "learning_rate": 5.972826447450896e-06,
+      "loss": 0.4515,
+      "step": 12173
+    },
+    {
+      "epoch": 1.9993225627655862,
+      "grad_norm": 0.6088139999913015,
+      "learning_rate": 5.9723832971687455e-06,
+      "loss": 0.445,
+      "step": 12174
+    },
+    {
+      "epoch": 1.999486789973929,
+      "grad_norm": 0.27415029930250745,
+      "learning_rate": 5.971940130895065e-06,
+      "loss": 0.4335,
+      "step": 12175
+    },
+    {
+      "epoch": 1.9996510171822717,
+      "grad_norm": 0.30806362684880456,
+      "learning_rate": 5.971496948634665e-06,
+      "loss": 0.4513,
+      "step": 12176
+    },
+    {
+      "epoch": 1.9998152443906143,
+      "grad_norm": 0.4787944606540653,
+      "learning_rate": 5.9710537503923605e-06,
+      "loss": 0.4689,
+      "step": 12177
+    },
+    {
+      "epoch": 1.9999794715989572,
+      "grad_norm": 0.2904555426356073,
+      "learning_rate": 5.970610536172966e-06,
+      "loss": 0.4574,
+      "step": 12178
+    },
+    {
+      "epoch": 2.0001436988073,
+      "grad_norm": 0.42320407092102436,
+      "learning_rate": 5.970167305981294e-06,
+      "loss": 0.4538,
+      "step": 12179
+    },
+    {
+      "epoch": 2.0003079260156427,
+      "grad_norm": 0.3083798483536343,
+      "learning_rate": 5.969724059822159e-06,
+      "loss": 0.4534,
+      "step": 12180
+    },
+    {
+      "epoch": 2.0004721532239853,
+      "grad_norm": 0.4283547328237905,
+      "learning_rate": 5.969280797700373e-06,
+      "loss": 0.4586,
+      "step": 12181
+    },
+    {
+      "epoch": 2.000636380432328,
+      "grad_norm": 0.4112612822032779,
+      "learning_rate": 5.968837519620753e-06,
+      "loss": 0.4597,
+      "step": 12182
+    },
+    {
+      "epoch": 2.000800607640671,
+      "grad_norm": 0.3368213650434348,
+      "learning_rate": 5.968394225588113e-06,
+      "loss": 0.4564,
+      "step": 12183
+    },
+    {
+      "epoch": 2.0009648348490137,
+      "grad_norm": 0.3492692409648053,
+      "learning_rate": 5.967950915607267e-06,
+      "loss": 0.4603,
+      "step": 12184
+    },
+    {
+      "epoch": 2.0011290620573563,
+      "grad_norm": 0.327969557551925,
+      "learning_rate": 5.967507589683027e-06,
+      "loss": 0.4592,
+      "step": 12185
+    },
+    {
+      "epoch": 2.001293289265699,
+      "grad_norm": 1.0052464626999962,
+      "learning_rate": 5.967064247820214e-06,
+      "loss": 0.447,
+      "step": 12186
+    },
+    {
+      "epoch": 2.001457516474042,
+      "grad_norm": 0.3323420980370512,
+      "learning_rate": 5.966620890023639e-06,
+      "loss": 0.4497,
+      "step": 12187
+    },
+    {
+      "epoch": 2.0016217436823847,
+      "grad_norm": 0.30221353880927326,
+      "learning_rate": 5.966177516298119e-06,
+      "loss": 0.4381,
+      "step": 12188
+    },
+    {
+      "epoch": 2.0017859708907273,
+      "grad_norm": 0.34728416679258106,
+      "learning_rate": 5.965734126648467e-06,
+      "loss": 0.4511,
+      "step": 12189
+    },
+    {
+      "epoch": 2.00195019809907,
+      "grad_norm": 0.34886261634920057,
+      "learning_rate": 5.9652907210795e-06,
+      "loss": 0.4818,
+      "step": 12190
+    },
+    {
+      "epoch": 2.002114425307413,
+      "grad_norm": 0.3436162370687873,
+      "learning_rate": 5.964847299596035e-06,
+      "loss": 0.448,
+      "step": 12191
+    },
+    {
+      "epoch": 2.0022786525157557,
+      "grad_norm": 0.4215227813011232,
+      "learning_rate": 5.964403862202888e-06,
+      "loss": 0.4549,
+      "step": 12192
+    },
+    {
+      "epoch": 2.0024428797240983,
+      "grad_norm": 0.3366921499133086,
+      "learning_rate": 5.963960408904874e-06,
+      "loss": 0.4392,
+      "step": 12193
+    },
+    {
+      "epoch": 2.002607106932441,
+      "grad_norm": 0.3245529569332661,
+      "learning_rate": 5.963516939706809e-06,
+      "loss": 0.4551,
+      "step": 12194
+    },
+    {
+      "epoch": 2.0027713341407836,
+      "grad_norm": 0.33977378407262626,
+      "learning_rate": 5.963073454613509e-06,
+      "loss": 0.4695,
+      "step": 12195
+    },
+    {
+      "epoch": 2.0029355613491266,
+      "grad_norm": 0.3069779515187422,
+      "learning_rate": 5.962629953629794e-06,
+      "loss": 0.4412,
+      "step": 12196
+    },
+    {
+      "epoch": 2.0030997885574693,
+      "grad_norm": 0.3507049844560384,
+      "learning_rate": 5.962186436760476e-06,
+      "loss": 0.4761,
+      "step": 12197
+    },
+    {
+      "epoch": 2.003264015765812,
+      "grad_norm": 0.3491589362302453,
+      "learning_rate": 5.9617429040103785e-06,
+      "loss": 0.441,
+      "step": 12198
+    },
+    {
+      "epoch": 2.0034282429741546,
+      "grad_norm": 0.3742547465374901,
+      "learning_rate": 5.961299355384311e-06,
+      "loss": 0.4281,
+      "step": 12199
+    },
+    {
+      "epoch": 2.0035924701824976,
+      "grad_norm": 0.4053926473537887,
+      "learning_rate": 5.960855790887098e-06,
+      "loss": 0.4504,
+      "step": 12200
+    },
+    {
+      "epoch": 2.0037566973908403,
+      "grad_norm": 0.35957952049354897,
+      "learning_rate": 5.960412210523552e-06,
+      "loss": 0.4252,
+      "step": 12201
+    },
+    {
+      "epoch": 2.003920924599183,
+      "grad_norm": 0.32657403088533404,
+      "learning_rate": 5.959968614298493e-06,
+      "loss": 0.4472,
+      "step": 12202
+    },
+    {
+      "epoch": 2.0040851518075256,
+      "grad_norm": 0.3244325404399533,
+      "learning_rate": 5.959525002216738e-06,
+      "loss": 0.4618,
+      "step": 12203
+    },
+    {
+      "epoch": 2.0042493790158686,
+      "grad_norm": 0.6620676419938378,
+      "learning_rate": 5.959081374283106e-06,
+      "loss": 0.4516,
+      "step": 12204
+    },
+    {
+      "epoch": 2.0044136062242113,
+      "grad_norm": 0.31677753319473856,
+      "learning_rate": 5.9586377305024145e-06,
+      "loss": 0.4485,
+      "step": 12205
+    },
+    {
+      "epoch": 2.004577833432554,
+      "grad_norm": 0.4121208836205025,
+      "learning_rate": 5.958194070879482e-06,
+      "loss": 0.4532,
+      "step": 12206
+    },
+    {
+      "epoch": 2.0047420606408966,
+      "grad_norm": 0.3116933045975636,
+      "learning_rate": 5.957750395419127e-06,
+      "loss": 0.4567,
+      "step": 12207
+    },
+    {
+      "epoch": 2.0049062878492396,
+      "grad_norm": 0.33302922228958487,
+      "learning_rate": 5.957306704126169e-06,
+      "loss": 0.4518,
+      "step": 12208
+    },
+    {
+      "epoch": 2.0050705150575823,
+      "grad_norm": 0.4198718931149218,
+      "learning_rate": 5.956862997005428e-06,
+      "loss": 0.4647,
+      "step": 12209
+    },
+    {
+      "epoch": 2.005234742265925,
+      "grad_norm": 0.34251455165151395,
+      "learning_rate": 5.956419274061719e-06,
+      "loss": 0.4512,
+      "step": 12210
+    },
+    {
+      "epoch": 2.0053989694742675,
+      "grad_norm": 0.3519346429865396,
+      "learning_rate": 5.955975535299863e-06,
+      "loss": 0.4402,
+      "step": 12211
+    },
+    {
+      "epoch": 2.00556319668261,
+      "grad_norm": 0.36308105461875656,
+      "learning_rate": 5.955531780724683e-06,
+      "loss": 0.4598,
+      "step": 12212
+    },
+    {
+      "epoch": 2.0057274238909533,
+      "grad_norm": 0.3741093286692219,
+      "learning_rate": 5.955088010340995e-06,
+      "loss": 0.4443,
+      "step": 12213
+    },
+    {
+      "epoch": 2.005891651099296,
+      "grad_norm": 0.328443035688512,
+      "learning_rate": 5.954644224153618e-06,
+      "loss": 0.4192,
+      "step": 12214
+    },
+    {
+      "epoch": 2.0060558783076385,
+      "grad_norm": 0.3324771672414727,
+      "learning_rate": 5.954200422167376e-06,
+      "loss": 0.4319,
+      "step": 12215
+    },
+    {
+      "epoch": 2.006220105515981,
+      "grad_norm": 0.29254376840142776,
+      "learning_rate": 5.953756604387085e-06,
+      "loss": 0.4662,
+      "step": 12216
+    },
+    {
+      "epoch": 2.0063843327243243,
+      "grad_norm": 0.26014344535423545,
+      "learning_rate": 5.953312770817568e-06,
+      "loss": 0.432,
+      "step": 12217
+    },
+    {
+      "epoch": 2.006548559932667,
+      "grad_norm": 0.3568735894151926,
+      "learning_rate": 5.952868921463643e-06,
+      "loss": 0.4651,
+      "step": 12218
+    },
+    {
+      "epoch": 2.0067127871410095,
+      "grad_norm": 0.3141590794217856,
+      "learning_rate": 5.952425056330134e-06,
+      "loss": 0.4567,
+      "step": 12219
+    },
+    {
+      "epoch": 2.006877014349352,
+      "grad_norm": 0.3884315355667573,
+      "learning_rate": 5.951981175421858e-06,
+      "loss": 0.4439,
+      "step": 12220
+    },
+    {
+      "epoch": 2.0070412415576953,
+      "grad_norm": 0.30058138999138845,
+      "learning_rate": 5.951537278743639e-06,
+      "loss": 0.432,
+      "step": 12221
+    },
+    {
+      "epoch": 2.007205468766038,
+      "grad_norm": 0.44575480874833284,
+      "learning_rate": 5.951093366300296e-06,
+      "loss": 0.4355,
+      "step": 12222
+    },
+    {
+      "epoch": 2.0073696959743805,
+      "grad_norm": 0.366245952203828,
+      "learning_rate": 5.950649438096653e-06,
+      "loss": 0.4535,
+      "step": 12223
+    },
+    {
+      "epoch": 2.007533923182723,
+      "grad_norm": 0.2867930542655121,
+      "learning_rate": 5.9502054941375285e-06,
+      "loss": 0.4378,
+      "step": 12224
+    },
+    {
+      "epoch": 2.0076981503910662,
+      "grad_norm": 0.47746484419606905,
+      "learning_rate": 5.949761534427746e-06,
+      "loss": 0.4115,
+      "step": 12225
+    },
+    {
+      "epoch": 2.007862377599409,
+      "grad_norm": 0.37527877564632606,
+      "learning_rate": 5.9493175589721265e-06,
+      "loss": 0.4338,
+      "step": 12226
+    },
+    {
+      "epoch": 2.0080266048077515,
+      "grad_norm": 0.3297901763313656,
+      "learning_rate": 5.948873567775493e-06,
+      "loss": 0.4139,
+      "step": 12227
+    },
+    {
+      "epoch": 2.008190832016094,
+      "grad_norm": 0.2561263798325084,
+      "learning_rate": 5.948429560842666e-06,
+      "loss": 0.447,
+      "step": 12228
+    },
+    {
+      "epoch": 2.008355059224437,
+      "grad_norm": 0.4365444130306358,
+      "learning_rate": 5.94798553817847e-06,
+      "loss": 0.458,
+      "step": 12229
+    },
+    {
+      "epoch": 2.00851928643278,
+      "grad_norm": 0.5232889294616114,
+      "learning_rate": 5.9475414997877255e-06,
+      "loss": 0.447,
+      "step": 12230
+    },
+    {
+      "epoch": 2.0086835136411225,
+      "grad_norm": 0.3411242576654733,
+      "learning_rate": 5.947097445675258e-06,
+      "loss": 0.4541,
+      "step": 12231
+    },
+    {
+      "epoch": 2.008847740849465,
+      "grad_norm": 0.3540437876499908,
+      "learning_rate": 5.946653375845887e-06,
+      "loss": 0.4528,
+      "step": 12232
+    },
+    {
+      "epoch": 2.009011968057808,
+      "grad_norm": 0.41463655700949253,
+      "learning_rate": 5.946209290304437e-06,
+      "loss": 0.4484,
+      "step": 12233
+    },
+    {
+      "epoch": 2.009176195266151,
+      "grad_norm": 0.3590231272906019,
+      "learning_rate": 5.945765189055731e-06,
+      "loss": 0.4646,
+      "step": 12234
+    },
+    {
+      "epoch": 2.0093404224744935,
+      "grad_norm": 0.5948107860227744,
+      "learning_rate": 5.945321072104593e-06,
+      "loss": 0.4482,
+      "step": 12235
+    },
+    {
+      "epoch": 2.009504649682836,
+      "grad_norm": 0.309417465568033,
+      "learning_rate": 5.944876939455848e-06,
+      "loss": 0.4331,
+      "step": 12236
+    },
+    {
+      "epoch": 2.009668876891179,
+      "grad_norm": 0.3306098908330728,
+      "learning_rate": 5.944432791114314e-06,
+      "loss": 0.4499,
+      "step": 12237
+    },
+    {
+      "epoch": 2.009833104099522,
+      "grad_norm": 0.31621060772673043,
+      "learning_rate": 5.943988627084822e-06,
+      "loss": 0.4568,
+      "step": 12238
+    },
+    {
+      "epoch": 2.0099973313078645,
+      "grad_norm": 0.2856713698078628,
+      "learning_rate": 5.943544447372191e-06,
+      "loss": 0.4255,
+      "step": 12239
+    },
+    {
+      "epoch": 2.010161558516207,
+      "grad_norm": 0.28032980432657734,
+      "learning_rate": 5.943100251981248e-06,
+      "loss": 0.45,
+      "step": 12240
+    },
+    {
+      "epoch": 2.01032578572455,
+      "grad_norm": 0.3390743841064049,
+      "learning_rate": 5.942656040916815e-06,
+      "loss": 0.464,
+      "step": 12241
+    },
+    {
+      "epoch": 2.010490012932893,
+      "grad_norm": 0.3187089522691845,
+      "learning_rate": 5.942211814183721e-06,
+      "loss": 0.4534,
+      "step": 12242
+    },
+    {
+      "epoch": 2.0106542401412355,
+      "grad_norm": 0.42615693774868557,
+      "learning_rate": 5.941767571786786e-06,
+      "loss": 0.4631,
+      "step": 12243
+    },
+    {
+      "epoch": 2.010818467349578,
+      "grad_norm": 0.36667170180078384,
+      "learning_rate": 5.941323313730836e-06,
+      "loss": 0.4236,
+      "step": 12244
+    },
+    {
+      "epoch": 2.010982694557921,
+      "grad_norm": 0.43862327105157484,
+      "learning_rate": 5.940879040020696e-06,
+      "loss": 0.4519,
+      "step": 12245
+    },
+    {
+      "epoch": 2.0111469217662634,
+      "grad_norm": 0.33579804262100527,
+      "learning_rate": 5.940434750661194e-06,
+      "loss": 0.4137,
+      "step": 12246
+    },
+    {
+      "epoch": 2.0113111489746065,
+      "grad_norm": 0.30978789940948254,
+      "learning_rate": 5.939990445657153e-06,
+      "loss": 0.4343,
+      "step": 12247
+    },
+    {
+      "epoch": 2.011475376182949,
+      "grad_norm": 0.3728929087707586,
+      "learning_rate": 5.939546125013399e-06,
+      "loss": 0.4539,
+      "step": 12248
+    },
+    {
+      "epoch": 2.0116396033912918,
+      "grad_norm": 0.3256492544812732,
+      "learning_rate": 5.939101788734757e-06,
+      "loss": 0.4551,
+      "step": 12249
+    },
+    {
+      "epoch": 2.0118038305996344,
+      "grad_norm": 0.30868824642963294,
+      "learning_rate": 5.938657436826054e-06,
+      "loss": 0.4391,
+      "step": 12250
+    },
+    {
+      "epoch": 2.0119680578079775,
+      "grad_norm": 0.28528064429757055,
+      "learning_rate": 5.938213069292117e-06,
+      "loss": 0.4352,
+      "step": 12251
+    },
+    {
+      "epoch": 2.01213228501632,
+      "grad_norm": 0.720505405909572,
+      "learning_rate": 5.93776868613777e-06,
+      "loss": 0.4582,
+      "step": 12252
+    },
+    {
+      "epoch": 2.0122965122246628,
+      "grad_norm": 0.40996297358279415,
+      "learning_rate": 5.93732428736784e-06,
+      "loss": 0.4366,
+      "step": 12253
+    },
+    {
+      "epoch": 2.0124607394330054,
+      "grad_norm": 0.36966590337145583,
+      "learning_rate": 5.936879872987155e-06,
+      "loss": 0.4422,
+      "step": 12254
+    },
+    {
+      "epoch": 2.0126249666413485,
+      "grad_norm": 0.40644461183977937,
+      "learning_rate": 5.93643544300054e-06,
+      "loss": 0.4352,
+      "step": 12255
+    },
+    {
+      "epoch": 2.012789193849691,
+      "grad_norm": 0.348232310385563,
+      "learning_rate": 5.935990997412823e-06,
+      "loss": 0.4499,
+      "step": 12256
+    },
+    {
+      "epoch": 2.0129534210580338,
+      "grad_norm": 0.30635919875987044,
+      "learning_rate": 5.9355465362288315e-06,
+      "loss": 0.4597,
+      "step": 12257
+    },
+    {
+      "epoch": 2.0131176482663764,
+      "grad_norm": 0.38023284124533757,
+      "learning_rate": 5.9351020594533914e-06,
+      "loss": 0.4341,
+      "step": 12258
+    },
+    {
+      "epoch": 2.0132818754747195,
+      "grad_norm": 0.3209364969044947,
+      "learning_rate": 5.934657567091332e-06,
+      "loss": 0.4462,
+      "step": 12259
+    },
+    {
+      "epoch": 2.013446102683062,
+      "grad_norm": 0.4138574131679595,
+      "learning_rate": 5.9342130591474785e-06,
+      "loss": 0.4343,
+      "step": 12260
+    },
+    {
+      "epoch": 2.0136103298914048,
+      "grad_norm": 0.3691427016956709,
+      "learning_rate": 5.933768535626662e-06,
+      "loss": 0.4557,
+      "step": 12261
+    },
+    {
+      "epoch": 2.0137745570997474,
+      "grad_norm": 0.3137057625519252,
+      "learning_rate": 5.933323996533708e-06,
+      "loss": 0.4639,
+      "step": 12262
+    },
+    {
+      "epoch": 2.01393878430809,
+      "grad_norm": 0.4452247347531062,
+      "learning_rate": 5.932879441873445e-06,
+      "loss": 0.4478,
+      "step": 12263
+    },
+    {
+      "epoch": 2.014103011516433,
+      "grad_norm": 0.32534561106326604,
+      "learning_rate": 5.932434871650701e-06,
+      "loss": 0.4222,
+      "step": 12264
+    },
+    {
+      "epoch": 2.0142672387247758,
+      "grad_norm": 0.4228276902347355,
+      "learning_rate": 5.931990285870306e-06,
+      "loss": 0.4505,
+      "step": 12265
+    },
+    {
+      "epoch": 2.0144314659331184,
+      "grad_norm": 0.4876125453560543,
+      "learning_rate": 5.931545684537086e-06,
+      "loss": 0.4232,
+      "step": 12266
+    },
+    {
+      "epoch": 2.014595693141461,
+      "grad_norm": 0.4172990521894343,
+      "learning_rate": 5.9311010676558724e-06,
+      "loss": 0.4313,
+      "step": 12267
+    },
+    {
+      "epoch": 2.014759920349804,
+      "grad_norm": 0.4012043012163585,
+      "learning_rate": 5.9306564352314935e-06,
+      "loss": 0.4425,
+      "step": 12268
+    },
+    {
+      "epoch": 2.0149241475581467,
+      "grad_norm": 0.31821251020135133,
+      "learning_rate": 5.930211787268777e-06,
+      "loss": 0.455,
+      "step": 12269
+    },
+    {
+      "epoch": 2.0150883747664894,
+      "grad_norm": 0.3358598549375324,
+      "learning_rate": 5.929767123772555e-06,
+      "loss": 0.4501,
+      "step": 12270
+    },
+    {
+      "epoch": 2.015252601974832,
+      "grad_norm": 0.34992544409089366,
+      "learning_rate": 5.929322444747655e-06,
+      "loss": 0.4398,
+      "step": 12271
+    },
+    {
+      "epoch": 2.015416829183175,
+      "grad_norm": 0.46600892538976924,
+      "learning_rate": 5.928877750198906e-06,
+      "loss": 0.4376,
+      "step": 12272
+    },
+    {
+      "epoch": 2.0155810563915177,
+      "grad_norm": 0.4421023504524407,
+      "learning_rate": 5.928433040131139e-06,
+      "loss": 0.4627,
+      "step": 12273
+    },
+    {
+      "epoch": 2.0157452835998604,
+      "grad_norm": 0.3466636254391006,
+      "learning_rate": 5.9279883145491835e-06,
+      "loss": 0.4657,
+      "step": 12274
+    },
+    {
+      "epoch": 2.015909510808203,
+      "grad_norm": 0.41056134961088725,
+      "learning_rate": 5.927543573457871e-06,
+      "loss": 0.4482,
+      "step": 12275
+    },
+    {
+      "epoch": 2.016073738016546,
+      "grad_norm": 0.3998987345596671,
+      "learning_rate": 5.927098816862031e-06,
+      "loss": 0.4558,
+      "step": 12276
+    },
+    {
+      "epoch": 2.0162379652248887,
+      "grad_norm": 0.3411150872881389,
+      "learning_rate": 5.926654044766493e-06,
+      "loss": 0.4253,
+      "step": 12277
+    },
+    {
+      "epoch": 2.0164021924332314,
+      "grad_norm": 0.44622183591825637,
+      "learning_rate": 5.926209257176087e-06,
+      "loss": 0.4414,
+      "step": 12278
+    },
+    {
+      "epoch": 2.016566419641574,
+      "grad_norm": 0.4301937123349158,
+      "learning_rate": 5.925764454095646e-06,
+      "loss": 0.4332,
+      "step": 12279
+    },
+    {
+      "epoch": 2.0167306468499167,
+      "grad_norm": 0.518469057403512,
+      "learning_rate": 5.925319635530003e-06,
+      "loss": 0.4605,
+      "step": 12280
+    },
+    {
+      "epoch": 2.0168948740582597,
+      "grad_norm": 0.32242423510941276,
+      "learning_rate": 5.924874801483985e-06,
+      "loss": 0.4433,
+      "step": 12281
+    },
+    {
+      "epoch": 2.0170591012666024,
+      "grad_norm": 0.3939941753019156,
+      "learning_rate": 5.924429951962424e-06,
+      "loss": 0.4509,
+      "step": 12282
+    },
+    {
+      "epoch": 2.017223328474945,
+      "grad_norm": 0.330619113684146,
+      "learning_rate": 5.9239850869701516e-06,
+      "loss": 0.4334,
+      "step": 12283
+    },
+    {
+      "epoch": 2.0173875556832876,
+      "grad_norm": 0.4240219166481088,
+      "learning_rate": 5.923540206512001e-06,
+      "loss": 0.4517,
+      "step": 12284
+    },
+    {
+      "epoch": 2.0175517828916307,
+      "grad_norm": 0.37369167937967235,
+      "learning_rate": 5.923095310592804e-06,
+      "loss": 0.4418,
+      "step": 12285
+    },
+    {
+      "epoch": 2.0177160100999734,
+      "grad_norm": 0.33052399671771765,
+      "learning_rate": 5.922650399217391e-06,
+      "loss": 0.4549,
+      "step": 12286
+    },
+    {
+      "epoch": 2.017880237308316,
+      "grad_norm": 0.3318820800605246,
+      "learning_rate": 5.922205472390594e-06,
+      "loss": 0.4641,
+      "step": 12287
+    },
+    {
+      "epoch": 2.0180444645166586,
+      "grad_norm": 0.25965853886842816,
+      "learning_rate": 5.9217605301172475e-06,
+      "loss": 0.4327,
+      "step": 12288
+    },
+    {
+      "epoch": 2.0182086917250017,
+      "grad_norm": 0.39578260130111304,
+      "learning_rate": 5.921315572402183e-06,
+      "loss": 0.4443,
+      "step": 12289
+    },
+    {
+      "epoch": 2.0183729189333444,
+      "grad_norm": 0.3019127278366922,
+      "learning_rate": 5.920870599250232e-06,
+      "loss": 0.4526,
+      "step": 12290
+    },
+    {
+      "epoch": 2.018537146141687,
+      "grad_norm": 0.31515638588112566,
+      "learning_rate": 5.92042561066623e-06,
+      "loss": 0.4341,
+      "step": 12291
+    },
+    {
+      "epoch": 2.0187013733500296,
+      "grad_norm": 0.3347035481218894,
+      "learning_rate": 5.9199806066550074e-06,
+      "loss": 0.4211,
+      "step": 12292
+    },
+    {
+      "epoch": 2.0188656005583727,
+      "grad_norm": 0.342636954069315,
+      "learning_rate": 5.919535587221398e-06,
+      "loss": 0.4528,
+      "step": 12293
+    },
+    {
+      "epoch": 2.0190298277667154,
+      "grad_norm": 0.31264338761398147,
+      "learning_rate": 5.919090552370235e-06,
+      "loss": 0.4433,
+      "step": 12294
+    },
+    {
+      "epoch": 2.019194054975058,
+      "grad_norm": 0.30817308473653016,
+      "learning_rate": 5.918645502106354e-06,
+      "loss": 0.4533,
+      "step": 12295
+    },
+    {
+      "epoch": 2.0193582821834006,
+      "grad_norm": 0.4982108932983128,
+      "learning_rate": 5.918200436434586e-06,
+      "loss": 0.4467,
+      "step": 12296
+    },
+    {
+      "epoch": 2.0195225093917433,
+      "grad_norm": 0.3278038373993306,
+      "learning_rate": 5.9177553553597665e-06,
+      "loss": 0.4554,
+      "step": 12297
+    },
+    {
+      "epoch": 2.0196867366000864,
+      "grad_norm": 0.3155972149201842,
+      "learning_rate": 5.917310258886728e-06,
+      "loss": 0.4544,
+      "step": 12298
+    },
+    {
+      "epoch": 2.019850963808429,
+      "grad_norm": 0.36100394939817493,
+      "learning_rate": 5.916865147020307e-06,
+      "loss": 0.4359,
+      "step": 12299
+    },
+    {
+      "epoch": 2.0200151910167716,
+      "grad_norm": 0.474739217853946,
+      "learning_rate": 5.916420019765336e-06,
+      "loss": 0.4442,
+      "step": 12300
+    },
+    {
+      "epoch": 2.0201794182251143,
+      "grad_norm": 0.3725941176745618,
+      "learning_rate": 5.915974877126649e-06,
+      "loss": 0.4448,
+      "step": 12301
+    },
+    {
+      "epoch": 2.0203436454334573,
+      "grad_norm": 0.3394057147474312,
+      "learning_rate": 5.915529719109083e-06,
+      "loss": 0.4326,
+      "step": 12302
+    },
+    {
+      "epoch": 2.0205078726418,
+      "grad_norm": 0.3333744225766765,
+      "learning_rate": 5.9150845457174704e-06,
+      "loss": 0.444,
+      "step": 12303
+    },
+    {
+      "epoch": 2.0206720998501426,
+      "grad_norm": 0.3984379295038264,
+      "learning_rate": 5.9146393569566485e-06,
+      "loss": 0.4323,
+      "step": 12304
+    },
+    {
+      "epoch": 2.0208363270584853,
+      "grad_norm": 0.4014873188536819,
+      "learning_rate": 5.914194152831451e-06,
+      "loss": 0.4556,
+      "step": 12305
+    },
+    {
+      "epoch": 2.0210005542668283,
+      "grad_norm": 0.38413574125456695,
+      "learning_rate": 5.913748933346714e-06,
+      "loss": 0.4335,
+      "step": 12306
+    },
+    {
+      "epoch": 2.021164781475171,
+      "grad_norm": 0.33331851521554756,
+      "learning_rate": 5.9133036985072705e-06,
+      "loss": 0.4507,
+      "step": 12307
+    },
+    {
+      "epoch": 2.0213290086835136,
+      "grad_norm": 0.3326758862247054,
+      "learning_rate": 5.91285844831796e-06,
+      "loss": 0.4528,
+      "step": 12308
+    },
+    {
+      "epoch": 2.0214932358918563,
+      "grad_norm": 0.29505514100770924,
+      "learning_rate": 5.912413182783617e-06,
+      "loss": 0.4401,
+      "step": 12309
+    },
+    {
+      "epoch": 2.0216574631001993,
+      "grad_norm": 0.34154494537786756,
+      "learning_rate": 5.911967901909078e-06,
+      "loss": 0.4341,
+      "step": 12310
+    },
+    {
+      "epoch": 2.021821690308542,
+      "grad_norm": 0.4049103391456977,
+      "learning_rate": 5.911522605699176e-06,
+      "loss": 0.4444,
+      "step": 12311
+    },
+    {
+      "epoch": 2.0219859175168846,
+      "grad_norm": 0.37915910316572177,
+      "learning_rate": 5.911077294158751e-06,
+      "loss": 0.4498,
+      "step": 12312
+    },
+    {
+      "epoch": 2.0221501447252272,
+      "grad_norm": 0.42567405204291847,
+      "learning_rate": 5.910631967292638e-06,
+      "loss": 0.4521,
+      "step": 12313
+    },
+    {
+      "epoch": 2.02231437193357,
+      "grad_norm": 0.303818370149242,
+      "learning_rate": 5.910186625105676e-06,
+      "loss": 0.4415,
+      "step": 12314
+    },
+    {
+      "epoch": 2.022478599141913,
+      "grad_norm": 0.3723931157545376,
+      "learning_rate": 5.909741267602698e-06,
+      "loss": 0.4303,
+      "step": 12315
+    },
+    {
+      "epoch": 2.0226428263502556,
+      "grad_norm": 0.34244289280435986,
+      "learning_rate": 5.909295894788541e-06,
+      "loss": 0.4294,
+      "step": 12316
+    },
+    {
+      "epoch": 2.0228070535585982,
+      "grad_norm": 0.45940405346671165,
+      "learning_rate": 5.9088505066680465e-06,
+      "loss": 0.4461,
+      "step": 12317
+    },
+    {
+      "epoch": 2.022971280766941,
+      "grad_norm": 0.5371059400275238,
+      "learning_rate": 5.908405103246049e-06,
+      "loss": 0.4659,
+      "step": 12318
+    },
+    {
+      "epoch": 2.023135507975284,
+      "grad_norm": 0.37195619324867074,
+      "learning_rate": 5.907959684527387e-06,
+      "loss": 0.4438,
+      "step": 12319
+    },
+    {
+      "epoch": 2.0232997351836266,
+      "grad_norm": 0.4444748904870266,
+      "learning_rate": 5.907514250516897e-06,
+      "loss": 0.4413,
+      "step": 12320
+    },
+    {
+      "epoch": 2.0234639623919692,
+      "grad_norm": 0.4461299536447834,
+      "learning_rate": 5.907068801219417e-06,
+      "loss": 0.4335,
+      "step": 12321
+    },
+    {
+      "epoch": 2.023628189600312,
+      "grad_norm": 0.43994447948385823,
+      "learning_rate": 5.9066233366397854e-06,
+      "loss": 0.4451,
+      "step": 12322
+    },
+    {
+      "epoch": 2.023792416808655,
+      "grad_norm": 0.3247864832152194,
+      "learning_rate": 5.90617785678284e-06,
+      "loss": 0.4544,
+      "step": 12323
+    },
+    {
+      "epoch": 2.0239566440169976,
+      "grad_norm": 0.3456630868813913,
+      "learning_rate": 5.9057323616534216e-06,
+      "loss": 0.4539,
+      "step": 12324
+    },
+    {
+      "epoch": 2.0241208712253402,
+      "grad_norm": 0.4800805237875434,
+      "learning_rate": 5.905286851256365e-06,
+      "loss": 0.4624,
+      "step": 12325
+    },
+    {
+      "epoch": 2.024285098433683,
+      "grad_norm": 0.34959063433120807,
+      "learning_rate": 5.904841325596511e-06,
+      "loss": 0.4667,
+      "step": 12326
+    },
+    {
+      "epoch": 2.024449325642026,
+      "grad_norm": 0.41151557024535956,
+      "learning_rate": 5.904395784678698e-06,
+      "loss": 0.4403,
+      "step": 12327
+    },
+    {
+      "epoch": 2.0246135528503686,
+      "grad_norm": 0.2929240886104455,
+      "learning_rate": 5.903950228507764e-06,
+      "loss": 0.4422,
+      "step": 12328
+    },
+    {
+      "epoch": 2.0247777800587112,
+      "grad_norm": 0.4225784029548028,
+      "learning_rate": 5.903504657088551e-06,
+      "loss": 0.4294,
+      "step": 12329
+    },
+    {
+      "epoch": 2.024942007267054,
+      "grad_norm": 0.3198757505033295,
+      "learning_rate": 5.903059070425895e-06,
+      "loss": 0.4427,
+      "step": 12330
+    },
+    {
+      "epoch": 2.0251062344753965,
+      "grad_norm": 0.35512213130023546,
+      "learning_rate": 5.902613468524639e-06,
+      "loss": 0.435,
+      "step": 12331
+    },
+    {
+      "epoch": 2.0252704616837396,
+      "grad_norm": 0.4014329270546415,
+      "learning_rate": 5.902167851389619e-06,
+      "loss": 0.4528,
+      "step": 12332
+    },
+    {
+      "epoch": 2.0254346888920822,
+      "grad_norm": 0.3685874338668055,
+      "learning_rate": 5.901722219025678e-06,
+      "loss": 0.4603,
+      "step": 12333
+    },
+    {
+      "epoch": 2.025598916100425,
+      "grad_norm": 0.28477749529150725,
+      "learning_rate": 5.901276571437654e-06,
+      "loss": 0.4383,
+      "step": 12334
+    },
+    {
+      "epoch": 2.0257631433087675,
+      "grad_norm": 0.3095326874609088,
+      "learning_rate": 5.9008309086303875e-06,
+      "loss": 0.4409,
+      "step": 12335
+    },
+    {
+      "epoch": 2.0259273705171106,
+      "grad_norm": 0.45871021472994483,
+      "learning_rate": 5.900385230608718e-06,
+      "loss": 0.459,
+      "step": 12336
+    },
+    {
+      "epoch": 2.026091597725453,
+      "grad_norm": 1.0720311778843719,
+      "learning_rate": 5.8999395373774885e-06,
+      "loss": 0.4294,
+      "step": 12337
+    },
+    {
+      "epoch": 2.026255824933796,
+      "grad_norm": 0.2924154206097919,
+      "learning_rate": 5.899493828941537e-06,
+      "loss": 0.4704,
+      "step": 12338
+    },
+    {
+      "epoch": 2.0264200521421385,
+      "grad_norm": 0.327283098845854,
+      "learning_rate": 5.899048105305709e-06,
+      "loss": 0.4586,
+      "step": 12339
+    },
+    {
+      "epoch": 2.0265842793504816,
+      "grad_norm": 0.349614059094532,
+      "learning_rate": 5.898602366474839e-06,
+      "loss": 0.4415,
+      "step": 12340
+    },
+    {
+      "epoch": 2.026748506558824,
+      "grad_norm": 0.34197640336286816,
+      "learning_rate": 5.898156612453772e-06,
+      "loss": 0.43,
+      "step": 12341
+    },
+    {
+      "epoch": 2.026912733767167,
+      "grad_norm": 0.4623929536887179,
+      "learning_rate": 5.897710843247348e-06,
+      "loss": 0.4478,
+      "step": 12342
+    },
+    {
+      "epoch": 2.0270769609755095,
+      "grad_norm": 0.5355379594969214,
+      "learning_rate": 5.89726505886041e-06,
+      "loss": 0.4451,
+      "step": 12343
+    },
+    {
+      "epoch": 2.0272411881838526,
+      "grad_norm": 0.3055222951108194,
+      "learning_rate": 5.896819259297799e-06,
+      "loss": 0.4359,
+      "step": 12344
+    },
+    {
+      "epoch": 2.027405415392195,
+      "grad_norm": 0.3074334568504291,
+      "learning_rate": 5.896373444564355e-06,
+      "loss": 0.4468,
+      "step": 12345
+    },
+    {
+      "epoch": 2.027569642600538,
+      "grad_norm": 0.33119785643769395,
+      "learning_rate": 5.895927614664923e-06,
+      "loss": 0.4489,
+      "step": 12346
+    },
+    {
+      "epoch": 2.0277338698088805,
+      "grad_norm": 0.5005833716890481,
+      "learning_rate": 5.895481769604343e-06,
+      "loss": 0.4142,
+      "step": 12347
+    },
+    {
+      "epoch": 2.027898097017223,
+      "grad_norm": 0.4446782973011459,
+      "learning_rate": 5.895035909387459e-06,
+      "loss": 0.4455,
+      "step": 12348
+    },
+    {
+      "epoch": 2.028062324225566,
+      "grad_norm": 0.3169868306167431,
+      "learning_rate": 5.894590034019111e-06,
+      "loss": 0.4518,
+      "step": 12349
+    },
+    {
+      "epoch": 2.028226551433909,
+      "grad_norm": 0.3679129008792082,
+      "learning_rate": 5.894144143504144e-06,
+      "loss": 0.4582,
+      "step": 12350
+    },
+    {
+      "epoch": 2.0283907786422515,
+      "grad_norm": 0.3356771628200559,
+      "learning_rate": 5.8936982378474e-06,
+      "loss": 0.455,
+      "step": 12351
+    },
+    {
+      "epoch": 2.028555005850594,
+      "grad_norm": 0.37868840952035804,
+      "learning_rate": 5.893252317053722e-06,
+      "loss": 0.4564,
+      "step": 12352
+    },
+    {
+      "epoch": 2.028719233058937,
+      "grad_norm": 0.3660224757041939,
+      "learning_rate": 5.892806381127953e-06,
+      "loss": 0.416,
+      "step": 12353
+    },
+    {
+      "epoch": 2.02888346026728,
+      "grad_norm": 0.35213115451904753,
+      "learning_rate": 5.892360430074936e-06,
+      "loss": 0.4313,
+      "step": 12354
+    },
+    {
+      "epoch": 2.0290476874756225,
+      "grad_norm": 0.3061381002851472,
+      "learning_rate": 5.891914463899515e-06,
+      "loss": 0.4464,
+      "step": 12355
+    },
+    {
+      "epoch": 2.029211914683965,
+      "grad_norm": 0.3198294060410183,
+      "learning_rate": 5.891468482606532e-06,
+      "loss": 0.4465,
+      "step": 12356
+    },
+    {
+      "epoch": 2.029376141892308,
+      "grad_norm": 0.5061425404700527,
+      "learning_rate": 5.8910224862008345e-06,
+      "loss": 0.4704,
+      "step": 12357
+    },
+    {
+      "epoch": 2.029540369100651,
+      "grad_norm": 0.2906027127123769,
+      "learning_rate": 5.890576474687264e-06,
+      "loss": 0.4539,
+      "step": 12358
+    },
+    {
+      "epoch": 2.0297045963089935,
+      "grad_norm": 0.9628300081908078,
+      "learning_rate": 5.890130448070665e-06,
+      "loss": 0.4605,
+      "step": 12359
+    },
+    {
+      "epoch": 2.029868823517336,
+      "grad_norm": 0.339842469867948,
+      "learning_rate": 5.889684406355879e-06,
+      "loss": 0.4452,
+      "step": 12360
+    },
+    {
+      "epoch": 2.030033050725679,
+      "grad_norm": 0.2898973367954881,
+      "learning_rate": 5.889238349547755e-06,
+      "loss": 0.4559,
+      "step": 12361
+    },
+    {
+      "epoch": 2.030197277934022,
+      "grad_norm": 0.32955122544839677,
+      "learning_rate": 5.8887922776511355e-06,
+      "loss": 0.4593,
+      "step": 12362
+    },
+    {
+      "epoch": 2.0303615051423645,
+      "grad_norm": 0.29610395042009074,
+      "learning_rate": 5.888346190670868e-06,
+      "loss": 0.46,
+      "step": 12363
+    },
+    {
+      "epoch": 2.030525732350707,
+      "grad_norm": 0.3076361536809495,
+      "learning_rate": 5.887900088611792e-06,
+      "loss": 0.4537,
+      "step": 12364
+    },
+    {
+      "epoch": 2.0306899595590497,
+      "grad_norm": 0.3806885935501196,
+      "learning_rate": 5.887453971478756e-06,
+      "loss": 0.4385,
+      "step": 12365
+    },
+    {
+      "epoch": 2.030854186767393,
+      "grad_norm": 0.4892710621594915,
+      "learning_rate": 5.8870078392766036e-06,
+      "loss": 0.4344,
+      "step": 12366
+    },
+    {
+      "epoch": 2.0310184139757355,
+      "grad_norm": 0.6621760845336766,
+      "learning_rate": 5.886561692010184e-06,
+      "loss": 0.4566,
+      "step": 12367
+    },
+    {
+      "epoch": 2.031182641184078,
+      "grad_norm": 0.34316954928589083,
+      "learning_rate": 5.886115529684339e-06,
+      "loss": 0.4545,
+      "step": 12368
+    },
+    {
+      "epoch": 2.0313468683924207,
+      "grad_norm": 0.29649605317323635,
+      "learning_rate": 5.8856693523039155e-06,
+      "loss": 0.4409,
+      "step": 12369
+    },
+    {
+      "epoch": 2.031511095600764,
+      "grad_norm": 0.32991559461150566,
+      "learning_rate": 5.885223159873759e-06,
+      "loss": 0.4436,
+      "step": 12370
+    },
+    {
+      "epoch": 2.0316753228091065,
+      "grad_norm": 0.3997830334051244,
+      "learning_rate": 5.884776952398717e-06,
+      "loss": 0.4325,
+      "step": 12371
+    },
+    {
+      "epoch": 2.031839550017449,
+      "grad_norm": 0.2950333162096897,
+      "learning_rate": 5.884330729883634e-06,
+      "loss": 0.455,
+      "step": 12372
+    },
+    {
+      "epoch": 2.0320037772257917,
+      "grad_norm": 0.3424041916207841,
+      "learning_rate": 5.883884492333359e-06,
+      "loss": 0.44,
+      "step": 12373
+    },
+    {
+      "epoch": 2.032168004434135,
+      "grad_norm": 0.27965686096829506,
+      "learning_rate": 5.883438239752734e-06,
+      "loss": 0.427,
+      "step": 12374
+    },
+    {
+      "epoch": 2.0323322316424774,
+      "grad_norm": 0.29382204979502846,
+      "learning_rate": 5.882991972146611e-06,
+      "loss": 0.4455,
+      "step": 12375
+    },
+    {
+      "epoch": 2.03249645885082,
+      "grad_norm": 0.32015433618862266,
+      "learning_rate": 5.882545689519834e-06,
+      "loss": 0.4558,
+      "step": 12376
+    },
+    {
+      "epoch": 2.0326606860591627,
+      "grad_norm": 0.3289590563823037,
+      "learning_rate": 5.88209939187725e-06,
+      "loss": 0.4581,
+      "step": 12377
+    },
+    {
+      "epoch": 2.032824913267506,
+      "grad_norm": 0.3202040927893558,
+      "learning_rate": 5.881653079223708e-06,
+      "loss": 0.433,
+      "step": 12378
+    },
+    {
+      "epoch": 2.0329891404758484,
+      "grad_norm": 0.4237401638627447,
+      "learning_rate": 5.881206751564053e-06,
+      "loss": 0.4468,
+      "step": 12379
+    },
+    {
+      "epoch": 2.033153367684191,
+      "grad_norm": 0.4113867464229899,
+      "learning_rate": 5.8807604089031345e-06,
+      "loss": 0.4498,
+      "step": 12380
+    },
+    {
+      "epoch": 2.0333175948925337,
+      "grad_norm": 0.3713036320098837,
+      "learning_rate": 5.880314051245799e-06,
+      "loss": 0.4636,
+      "step": 12381
+    },
+    {
+      "epoch": 2.0334818221008764,
+      "grad_norm": 0.3487669051771585,
+      "learning_rate": 5.879867678596896e-06,
+      "loss": 0.4261,
+      "step": 12382
+    },
+    {
+      "epoch": 2.0336460493092194,
+      "grad_norm": 0.4297560095357159,
+      "learning_rate": 5.8794212909612705e-06,
+      "loss": 0.4632,
+      "step": 12383
+    },
+    {
+      "epoch": 2.033810276517562,
+      "grad_norm": 0.41595656127910297,
+      "learning_rate": 5.878974888343773e-06,
+      "loss": 0.4363,
+      "step": 12384
+    },
+    {
+      "epoch": 2.0339745037259047,
+      "grad_norm": 0.31941281295240154,
+      "learning_rate": 5.878528470749252e-06,
+      "loss": 0.4616,
+      "step": 12385
+    },
+    {
+      "epoch": 2.0341387309342474,
+      "grad_norm": 0.3736947516535489,
+      "learning_rate": 5.878082038182555e-06,
+      "loss": 0.4407,
+      "step": 12386
+    },
+    {
+      "epoch": 2.0343029581425904,
+      "grad_norm": 0.3648758434688456,
+      "learning_rate": 5.8776355906485325e-06,
+      "loss": 0.4615,
+      "step": 12387
+    },
+    {
+      "epoch": 2.034467185350933,
+      "grad_norm": 0.33317524041004737,
+      "learning_rate": 5.877189128152032e-06,
+      "loss": 0.4516,
+      "step": 12388
+    },
+    {
+      "epoch": 2.0346314125592757,
+      "grad_norm": 0.34434024200356267,
+      "learning_rate": 5.876742650697902e-06,
+      "loss": 0.4687,
+      "step": 12389
+    },
+    {
+      "epoch": 2.0347956397676183,
+      "grad_norm": 0.3873243560043453,
+      "learning_rate": 5.876296158290991e-06,
+      "loss": 0.4664,
+      "step": 12390
+    },
+    {
+      "epoch": 2.0349598669759614,
+      "grad_norm": 0.6587433734412117,
+      "learning_rate": 5.87584965093615e-06,
+      "loss": 0.4384,
+      "step": 12391
+    },
+    {
+      "epoch": 2.035124094184304,
+      "grad_norm": 0.376035639840286,
+      "learning_rate": 5.87540312863823e-06,
+      "loss": 0.4448,
+      "step": 12392
+    },
+    {
+      "epoch": 2.0352883213926467,
+      "grad_norm": 0.3356221661068441,
+      "learning_rate": 5.874956591402078e-06,
+      "loss": 0.4504,
+      "step": 12393
+    },
+    {
+      "epoch": 2.0354525486009893,
+      "grad_norm": 0.3364467718112007,
+      "learning_rate": 5.874510039232544e-06,
+      "loss": 0.4287,
+      "step": 12394
+    },
+    {
+      "epoch": 2.0356167758093324,
+      "grad_norm": 0.3530682596790142,
+      "learning_rate": 5.874063472134479e-06,
+      "loss": 0.4323,
+      "step": 12395
+    },
+    {
+      "epoch": 2.035781003017675,
+      "grad_norm": 0.41389982440021067,
+      "learning_rate": 5.8736168901127325e-06,
+      "loss": 0.4536,
+      "step": 12396
+    },
+    {
+      "epoch": 2.0359452302260177,
+      "grad_norm": 0.317301755237896,
+      "learning_rate": 5.873170293172156e-06,
+      "loss": 0.4371,
+      "step": 12397
+    },
+    {
+      "epoch": 2.0361094574343603,
+      "grad_norm": 0.34272680753358126,
+      "learning_rate": 5.872723681317599e-06,
+      "loss": 0.4491,
+      "step": 12398
+    },
+    {
+      "epoch": 2.036273684642703,
+      "grad_norm": 0.4006309979835223,
+      "learning_rate": 5.87227705455391e-06,
+      "loss": 0.441,
+      "step": 12399
+    },
+    {
+      "epoch": 2.036437911851046,
+      "grad_norm": 0.5542787557823411,
+      "learning_rate": 5.871830412885944e-06,
+      "loss": 0.4296,
+      "step": 12400
+    },
+    {
+      "epoch": 2.0366021390593887,
+      "grad_norm": 0.3431090307672414,
+      "learning_rate": 5.871383756318551e-06,
+      "loss": 0.4513,
+      "step": 12401
+    },
+    {
+      "epoch": 2.0367663662677313,
+      "grad_norm": 0.3586838401808626,
+      "learning_rate": 5.87093708485658e-06,
+      "loss": 0.4485,
+      "step": 12402
+    },
+    {
+      "epoch": 2.036930593476074,
+      "grad_norm": 0.3051512780842894,
+      "learning_rate": 5.8704903985048825e-06,
+      "loss": 0.4547,
+      "step": 12403
+    },
+    {
+      "epoch": 2.037094820684417,
+      "grad_norm": 0.35974186850941714,
+      "learning_rate": 5.870043697268312e-06,
+      "loss": 0.4495,
+      "step": 12404
+    },
+    {
+      "epoch": 2.0372590478927597,
+      "grad_norm": 0.6056306895454063,
+      "learning_rate": 5.869596981151719e-06,
+      "loss": 0.467,
+      "step": 12405
+    },
+    {
+      "epoch": 2.0374232751011023,
+      "grad_norm": 0.2819644783966082,
+      "learning_rate": 5.869150250159955e-06,
+      "loss": 0.4314,
+      "step": 12406
+    },
+    {
+      "epoch": 2.037587502309445,
+      "grad_norm": 0.4286095726381181,
+      "learning_rate": 5.868703504297873e-06,
+      "loss": 0.4474,
+      "step": 12407
+    },
+    {
+      "epoch": 2.037751729517788,
+      "grad_norm": 0.32038506407005746,
+      "learning_rate": 5.868256743570323e-06,
+      "loss": 0.4422,
+      "step": 12408
+    },
+    {
+      "epoch": 2.0379159567261307,
+      "grad_norm": 0.3259096360271549,
+      "learning_rate": 5.8678099679821595e-06,
+      "loss": 0.4517,
+      "step": 12409
+    },
+    {
+      "epoch": 2.0380801839344733,
+      "grad_norm": 0.27644003099357023,
+      "learning_rate": 5.867363177538234e-06,
+      "loss": 0.4327,
+      "step": 12410
+    },
+    {
+      "epoch": 2.038244411142816,
+      "grad_norm": 0.3402320682171515,
+      "learning_rate": 5.866916372243399e-06,
+      "loss": 0.4618,
+      "step": 12411
+    },
+    {
+      "epoch": 2.038408638351159,
+      "grad_norm": 0.284775187844616,
+      "learning_rate": 5.866469552102506e-06,
+      "loss": 0.4524,
+      "step": 12412
+    },
+    {
+      "epoch": 2.0385728655595017,
+      "grad_norm": 0.4978827495915803,
+      "learning_rate": 5.866022717120411e-06,
+      "loss": 0.434,
+      "step": 12413
+    },
+    {
+      "epoch": 2.0387370927678443,
+      "grad_norm": 0.8185449778955562,
+      "learning_rate": 5.865575867301965e-06,
+      "loss": 0.4595,
+      "step": 12414
+    },
+    {
+      "epoch": 2.038901319976187,
+      "grad_norm": 0.3231777306250387,
+      "learning_rate": 5.8651290026520205e-06,
+      "loss": 0.4369,
+      "step": 12415
+    },
+    {
+      "epoch": 2.0390655471845296,
+      "grad_norm": 0.329589659019701,
+      "learning_rate": 5.864682123175433e-06,
+      "loss": 0.4475,
+      "step": 12416
+    },
+    {
+      "epoch": 2.0392297743928727,
+      "grad_norm": 1.1849958099030102,
+      "learning_rate": 5.864235228877056e-06,
+      "loss": 0.4255,
+      "step": 12417
+    },
+    {
+      "epoch": 2.0393940016012153,
+      "grad_norm": 0.33656368419001453,
+      "learning_rate": 5.86378831976174e-06,
+      "loss": 0.4348,
+      "step": 12418
+    },
+    {
+      "epoch": 2.039558228809558,
+      "grad_norm": 0.3167560065962795,
+      "learning_rate": 5.863341395834341e-06,
+      "loss": 0.4144,
+      "step": 12419
+    },
+    {
+      "epoch": 2.0397224560179006,
+      "grad_norm": 0.3695074006748326,
+      "learning_rate": 5.862894457099714e-06,
+      "loss": 0.4503,
+      "step": 12420
+    },
+    {
+      "epoch": 2.0398866832262437,
+      "grad_norm": 0.3667195996387198,
+      "learning_rate": 5.862447503562713e-06,
+      "loss": 0.4379,
+      "step": 12421
+    },
+    {
+      "epoch": 2.0400509104345863,
+      "grad_norm": 0.39032351094667883,
+      "learning_rate": 5.862000535228191e-06,
+      "loss": 0.4329,
+      "step": 12422
+    },
+    {
+      "epoch": 2.040215137642929,
+      "grad_norm": 0.3380516948636253,
+      "learning_rate": 5.861553552101003e-06,
+      "loss": 0.4586,
+      "step": 12423
+    },
+    {
+      "epoch": 2.0403793648512716,
+      "grad_norm": 0.35621909121959594,
+      "learning_rate": 5.861106554186003e-06,
+      "loss": 0.4533,
+      "step": 12424
+    },
+    {
+      "epoch": 2.0405435920596147,
+      "grad_norm": 0.3354530237084805,
+      "learning_rate": 5.860659541488048e-06,
+      "loss": 0.4391,
+      "step": 12425
+    },
+    {
+      "epoch": 2.0407078192679573,
+      "grad_norm": 0.2989633077547237,
+      "learning_rate": 5.860212514011992e-06,
+      "loss": 0.4241,
+      "step": 12426
+    },
+    {
+      "epoch": 2.0408720464763,
+      "grad_norm": 0.35862761603268756,
+      "learning_rate": 5.859765471762688e-06,
+      "loss": 0.4446,
+      "step": 12427
+    },
+    {
+      "epoch": 2.0410362736846426,
+      "grad_norm": 0.30444330966462657,
+      "learning_rate": 5.859318414744995e-06,
+      "loss": 0.474,
+      "step": 12428
+    },
+    {
+      "epoch": 2.0412005008929857,
+      "grad_norm": 0.352540677765662,
+      "learning_rate": 5.8588713429637655e-06,
+      "loss": 0.4387,
+      "step": 12429
+    },
+    {
+      "epoch": 2.0413647281013283,
+      "grad_norm": 0.3778446185553416,
+      "learning_rate": 5.8584242564238566e-06,
+      "loss": 0.4591,
+      "step": 12430
+    },
+    {
+      "epoch": 2.041528955309671,
+      "grad_norm": 0.3491850807666615,
+      "learning_rate": 5.857977155130124e-06,
+      "loss": 0.4621,
+      "step": 12431
+    },
+    {
+      "epoch": 2.0416931825180136,
+      "grad_norm": 0.38612813413715863,
+      "learning_rate": 5.857530039087423e-06,
+      "loss": 0.4732,
+      "step": 12432
+    },
+    {
+      "epoch": 2.041857409726356,
+      "grad_norm": 0.3933909296763072,
+      "learning_rate": 5.85708290830061e-06,
+      "loss": 0.4572,
+      "step": 12433
+    },
+    {
+      "epoch": 2.0420216369346993,
+      "grad_norm": 0.45460547103818244,
+      "learning_rate": 5.856635762774542e-06,
+      "loss": 0.4554,
+      "step": 12434
+    },
+    {
+      "epoch": 2.042185864143042,
+      "grad_norm": 0.4897357475104495,
+      "learning_rate": 5.856188602514075e-06,
+      "loss": 0.4549,
+      "step": 12435
+    },
+    {
+      "epoch": 2.0423500913513846,
+      "grad_norm": 0.37141597461045756,
+      "learning_rate": 5.855741427524066e-06,
+      "loss": 0.4341,
+      "step": 12436
+    },
+    {
+      "epoch": 2.042514318559727,
+      "grad_norm": 0.3152170933745343,
+      "learning_rate": 5.8552942378093694e-06,
+      "loss": 0.4383,
+      "step": 12437
+    },
+    {
+      "epoch": 2.0426785457680703,
+      "grad_norm": 0.3315546517567952,
+      "learning_rate": 5.854847033374845e-06,
+      "loss": 0.4422,
+      "step": 12438
+    },
+    {
+      "epoch": 2.042842772976413,
+      "grad_norm": 0.34849627509879655,
+      "learning_rate": 5.854399814225349e-06,
+      "loss": 0.4534,
+      "step": 12439
+    },
+    {
+      "epoch": 2.0430070001847556,
+      "grad_norm": 0.3415060059087718,
+      "learning_rate": 5.8539525803657375e-06,
+      "loss": 0.4244,
+      "step": 12440
+    },
+    {
+      "epoch": 2.043171227393098,
+      "grad_norm": 0.39313173165872756,
+      "learning_rate": 5.85350533180087e-06,
+      "loss": 0.4217,
+      "step": 12441
+    },
+    {
+      "epoch": 2.0433354546014413,
+      "grad_norm": 0.3429924779240633,
+      "learning_rate": 5.853058068535603e-06,
+      "loss": 0.4577,
+      "step": 12442
+    },
+    {
+      "epoch": 2.043499681809784,
+      "grad_norm": 0.32928596156929474,
+      "learning_rate": 5.852610790574793e-06,
+      "loss": 0.4566,
+      "step": 12443
+    },
+    {
+      "epoch": 2.0436639090181266,
+      "grad_norm": 0.47849256644842075,
+      "learning_rate": 5.8521634979232995e-06,
+      "loss": 0.4444,
+      "step": 12444
+    },
+    {
+      "epoch": 2.043828136226469,
+      "grad_norm": 0.3687608357126545,
+      "learning_rate": 5.851716190585981e-06,
+      "loss": 0.4318,
+      "step": 12445
+    },
+    {
+      "epoch": 2.0439923634348123,
+      "grad_norm": 0.32895085912600486,
+      "learning_rate": 5.851268868567694e-06,
+      "loss": 0.4203,
+      "step": 12446
+    },
+    {
+      "epoch": 2.044156590643155,
+      "grad_norm": 0.37289343076052367,
+      "learning_rate": 5.850821531873298e-06,
+      "loss": 0.4433,
+      "step": 12447
+    },
+    {
+      "epoch": 2.0443208178514976,
+      "grad_norm": 0.31859032614581423,
+      "learning_rate": 5.8503741805076496e-06,
+      "loss": 0.4338,
+      "step": 12448
+    },
+    {
+      "epoch": 2.04448504505984,
+      "grad_norm": 0.2887224225883823,
+      "learning_rate": 5.8499268144756104e-06,
+      "loss": 0.4357,
+      "step": 12449
+    },
+    {
+      "epoch": 2.044649272268183,
+      "grad_norm": 0.4697271025351198,
+      "learning_rate": 5.8494794337820375e-06,
+      "loss": 0.452,
+      "step": 12450
+    },
+    {
+      "epoch": 2.044813499476526,
+      "grad_norm": 0.3282934512860411,
+      "learning_rate": 5.849032038431792e-06,
+      "loss": 0.4449,
+      "step": 12451
+    },
+    {
+      "epoch": 2.0449777266848685,
+      "grad_norm": 0.385047656440147,
+      "learning_rate": 5.8485846284297285e-06,
+      "loss": 0.4388,
+      "step": 12452
+    },
+    {
+      "epoch": 2.045141953893211,
+      "grad_norm": 0.36439150850801005,
+      "learning_rate": 5.848137203780709e-06,
+      "loss": 0.4558,
+      "step": 12453
+    },
+    {
+      "epoch": 2.045306181101554,
+      "grad_norm": 0.40450690234416153,
+      "learning_rate": 5.847689764489595e-06,
+      "loss": 0.4337,
+      "step": 12454
+    },
+    {
+      "epoch": 2.045470408309897,
+      "grad_norm": 0.43383659347572107,
+      "learning_rate": 5.847242310561243e-06,
+      "loss": 0.444,
+      "step": 12455
+    },
+    {
+      "epoch": 2.0456346355182395,
+      "grad_norm": 0.33992573109644364,
+      "learning_rate": 5.846794842000516e-06,
+      "loss": 0.4302,
+      "step": 12456
+    },
+    {
+      "epoch": 2.045798862726582,
+      "grad_norm": 0.3404891501105816,
+      "learning_rate": 5.84634735881227e-06,
+      "loss": 0.4315,
+      "step": 12457
+    },
+    {
+      "epoch": 2.045963089934925,
+      "grad_norm": 0.3948929516683718,
+      "learning_rate": 5.845899861001367e-06,
+      "loss": 0.4352,
+      "step": 12458
+    },
+    {
+      "epoch": 2.046127317143268,
+      "grad_norm": 0.36232433689948934,
+      "learning_rate": 5.845452348572668e-06,
+      "loss": 0.4228,
+      "step": 12459
+    },
+    {
+      "epoch": 2.0462915443516105,
+      "grad_norm": 0.4879837670116479,
+      "learning_rate": 5.845004821531033e-06,
+      "loss": 0.4479,
+      "step": 12460
+    },
+    {
+      "epoch": 2.046455771559953,
+      "grad_norm": 0.43242467602156015,
+      "learning_rate": 5.844557279881321e-06,
+      "loss": 0.4338,
+      "step": 12461
+    },
+    {
+      "epoch": 2.046619998768296,
+      "grad_norm": 0.46972252076723947,
+      "learning_rate": 5.844109723628395e-06,
+      "loss": 0.4296,
+      "step": 12462
+    },
+    {
+      "epoch": 2.046784225976639,
+      "grad_norm": 0.348287060668494,
+      "learning_rate": 5.843662152777117e-06,
+      "loss": 0.4462,
+      "step": 12463
+    },
+    {
+      "epoch": 2.0469484531849815,
+      "grad_norm": 0.3554017413497341,
+      "learning_rate": 5.843214567332343e-06,
+      "loss": 0.4494,
+      "step": 12464
+    },
+    {
+      "epoch": 2.047112680393324,
+      "grad_norm": 0.7426503419751455,
+      "learning_rate": 5.842766967298939e-06,
+      "loss": 0.4375,
+      "step": 12465
+    },
+    {
+      "epoch": 2.047276907601667,
+      "grad_norm": 0.34331608608976466,
+      "learning_rate": 5.842319352681763e-06,
+      "loss": 0.4527,
+      "step": 12466
+    },
+    {
+      "epoch": 2.0474411348100094,
+      "grad_norm": 0.326253689959674,
+      "learning_rate": 5.8418717234856785e-06,
+      "loss": 0.4347,
+      "step": 12467
+    },
+    {
+      "epoch": 2.0476053620183525,
+      "grad_norm": 1.2139001651559362,
+      "learning_rate": 5.841424079715548e-06,
+      "loss": 0.4685,
+      "step": 12468
+    },
+    {
+      "epoch": 2.047769589226695,
+      "grad_norm": 1.809298572630905,
+      "learning_rate": 5.840976421376231e-06,
+      "loss": 0.4267,
+      "step": 12469
+    },
+    {
+      "epoch": 2.047933816435038,
+      "grad_norm": 0.643982392760066,
+      "learning_rate": 5.840528748472593e-06,
+      "loss": 0.4511,
+      "step": 12470
+    },
+    {
+      "epoch": 2.0480980436433804,
+      "grad_norm": 0.4200958520866591,
+      "learning_rate": 5.84008106100949e-06,
+      "loss": 0.4756,
+      "step": 12471
+    },
+    {
+      "epoch": 2.0482622708517235,
+      "grad_norm": 0.3785922765747419,
+      "learning_rate": 5.839633358991792e-06,
+      "loss": 0.4455,
+      "step": 12472
+    },
+    {
+      "epoch": 2.048426498060066,
+      "grad_norm": 0.334283585605747,
+      "learning_rate": 5.839185642424356e-06,
+      "loss": 0.4554,
+      "step": 12473
+    },
+    {
+      "epoch": 2.048590725268409,
+      "grad_norm": 0.2979365135820452,
+      "learning_rate": 5.838737911312046e-06,
+      "loss": 0.4394,
+      "step": 12474
+    },
+    {
+      "epoch": 2.0487549524767514,
+      "grad_norm": 0.3928511038792992,
+      "learning_rate": 5.838290165659726e-06,
+      "loss": 0.4309,
+      "step": 12475
+    },
+    {
+      "epoch": 2.0489191796850945,
+      "grad_norm": 0.42167196945885604,
+      "learning_rate": 5.837842405472259e-06,
+      "loss": 0.4202,
+      "step": 12476
+    },
+    {
+      "epoch": 2.049083406893437,
+      "grad_norm": 0.384941912400613,
+      "learning_rate": 5.837394630754504e-06,
+      "loss": 0.4454,
+      "step": 12477
+    },
+    {
+      "epoch": 2.04924763410178,
+      "grad_norm": 0.6027380809373334,
+      "learning_rate": 5.83694684151133e-06,
+      "loss": 0.4436,
+      "step": 12478
+    },
+    {
+      "epoch": 2.0494118613101224,
+      "grad_norm": 0.5500830911834226,
+      "learning_rate": 5.836499037747598e-06,
+      "loss": 0.4446,
+      "step": 12479
+    },
+    {
+      "epoch": 2.0495760885184655,
+      "grad_norm": 0.49515778776697733,
+      "learning_rate": 5.836051219468171e-06,
+      "loss": 0.4372,
+      "step": 12480
+    },
+    {
+      "epoch": 2.049740315726808,
+      "grad_norm": 0.43075419664571624,
+      "learning_rate": 5.835603386677913e-06,
+      "loss": 0.4429,
+      "step": 12481
+    },
+    {
+      "epoch": 2.049904542935151,
+      "grad_norm": 0.3542392734227228,
+      "learning_rate": 5.8351555393816885e-06,
+      "loss": 0.4262,
+      "step": 12482
+    },
+    {
+      "epoch": 2.0500687701434934,
+      "grad_norm": 0.38268666364407705,
+      "learning_rate": 5.8347076775843604e-06,
+      "loss": 0.4303,
+      "step": 12483
+    },
+    {
+      "epoch": 2.050232997351836,
+      "grad_norm": 0.47765234932548395,
+      "learning_rate": 5.834259801290795e-06,
+      "loss": 0.4498,
+      "step": 12484
+    },
+    {
+      "epoch": 2.050397224560179,
+      "grad_norm": 0.41345370663759506,
+      "learning_rate": 5.833811910505855e-06,
+      "loss": 0.4189,
+      "step": 12485
+    },
+    {
+      "epoch": 2.050561451768522,
+      "grad_norm": 0.4104341797134906,
+      "learning_rate": 5.833364005234404e-06,
+      "loss": 0.4538,
+      "step": 12486
+    },
+    {
+      "epoch": 2.0507256789768644,
+      "grad_norm": 0.31647737223956823,
+      "learning_rate": 5.83291608548131e-06,
+      "loss": 0.4407,
+      "step": 12487
+    },
+    {
+      "epoch": 2.050889906185207,
+      "grad_norm": 0.37302273496194177,
+      "learning_rate": 5.832468151251435e-06,
+      "loss": 0.4433,
+      "step": 12488
+    },
+    {
+      "epoch": 2.05105413339355,
+      "grad_norm": 0.3554326401961096,
+      "learning_rate": 5.832020202549644e-06,
+      "loss": 0.4432,
+      "step": 12489
+    },
+    {
+      "epoch": 2.0512183606018928,
+      "grad_norm": 0.491954272790454,
+      "learning_rate": 5.831572239380806e-06,
+      "loss": 0.4564,
+      "step": 12490
+    },
+    {
+      "epoch": 2.0513825878102354,
+      "grad_norm": 0.3789475522315765,
+      "learning_rate": 5.831124261749781e-06,
+      "loss": 0.4495,
+      "step": 12491
+    },
+    {
+      "epoch": 2.051546815018578,
+      "grad_norm": 0.647424818774864,
+      "learning_rate": 5.830676269661436e-06,
+      "loss": 0.4501,
+      "step": 12492
+    },
+    {
+      "epoch": 2.051711042226921,
+      "grad_norm": 0.5823413883449976,
+      "learning_rate": 5.830228263120641e-06,
+      "loss": 0.4213,
+      "step": 12493
+    },
+    {
+      "epoch": 2.0518752694352638,
+      "grad_norm": 0.27813242543412897,
+      "learning_rate": 5.829780242132256e-06,
+      "loss": 0.4397,
+      "step": 12494
+    },
+    {
+      "epoch": 2.0520394966436064,
+      "grad_norm": 0.3284450588600766,
+      "learning_rate": 5.829332206701149e-06,
+      "loss": 0.443,
+      "step": 12495
+    },
+    {
+      "epoch": 2.052203723851949,
+      "grad_norm": 0.3445192485646923,
+      "learning_rate": 5.828884156832186e-06,
+      "loss": 0.4518,
+      "step": 12496
+    },
+    {
+      "epoch": 2.052367951060292,
+      "grad_norm": 0.549298740102889,
+      "learning_rate": 5.828436092530235e-06,
+      "loss": 0.4444,
+      "step": 12497
+    },
+    {
+      "epoch": 2.0525321782686348,
+      "grad_norm": 0.46014439109326594,
+      "learning_rate": 5.82798801380016e-06,
+      "loss": 0.4532,
+      "step": 12498
+    },
+    {
+      "epoch": 2.0526964054769774,
+      "grad_norm": 0.28368044288710964,
+      "learning_rate": 5.8275399206468304e-06,
+      "loss": 0.441,
+      "step": 12499
+    },
+    {
+      "epoch": 2.05286063268532,
+      "grad_norm": 0.3968388641653914,
+      "learning_rate": 5.8270918130751085e-06,
+      "loss": 0.4459,
+      "step": 12500
+    },
+    {
+      "epoch": 2.0530248598936627,
+      "grad_norm": 0.3710478412314034,
+      "learning_rate": 5.8266436910898656e-06,
+      "loss": 0.463,
+      "step": 12501
+    },
+    {
+      "epoch": 2.0531890871020058,
+      "grad_norm": 0.313924593295772,
+      "learning_rate": 5.826195554695966e-06,
+      "loss": 0.4392,
+      "step": 12502
+    },
+    {
+      "epoch": 2.0533533143103484,
+      "grad_norm": 0.4580347375031826,
+      "learning_rate": 5.825747403898278e-06,
+      "loss": 0.4503,
+      "step": 12503
+    },
+    {
+      "epoch": 2.053517541518691,
+      "grad_norm": 0.4132435728401116,
+      "learning_rate": 5.825299238701669e-06,
+      "loss": 0.4559,
+      "step": 12504
+    },
+    {
+      "epoch": 2.0536817687270337,
+      "grad_norm": 0.36199062801530246,
+      "learning_rate": 5.824851059111007e-06,
+      "loss": 0.4483,
+      "step": 12505
+    },
+    {
+      "epoch": 2.0538459959353768,
+      "grad_norm": 0.4748756143679125,
+      "learning_rate": 5.824402865131159e-06,
+      "loss": 0.4575,
+      "step": 12506
+    },
+    {
+      "epoch": 2.0540102231437194,
+      "grad_norm": 0.3218339347858069,
+      "learning_rate": 5.823954656766991e-06,
+      "loss": 0.4504,
+      "step": 12507
+    },
+    {
+      "epoch": 2.054174450352062,
+      "grad_norm": 0.31580790798058256,
+      "learning_rate": 5.823506434023374e-06,
+      "loss": 0.4397,
+      "step": 12508
+    },
+    {
+      "epoch": 2.0543386775604047,
+      "grad_norm": 0.35353519486956647,
+      "learning_rate": 5.823058196905177e-06,
+      "loss": 0.4517,
+      "step": 12509
+    },
+    {
+      "epoch": 2.0545029047687478,
+      "grad_norm": 0.46529793989703433,
+      "learning_rate": 5.8226099454172644e-06,
+      "loss": 0.458,
+      "step": 12510
+    },
+    {
+      "epoch": 2.0546671319770904,
+      "grad_norm": 0.3205124315397412,
+      "learning_rate": 5.822161679564506e-06,
+      "loss": 0.4364,
+      "step": 12511
+    },
+    {
+      "epoch": 2.054831359185433,
+      "grad_norm": 0.3610849789933956,
+      "learning_rate": 5.821713399351771e-06,
+      "loss": 0.4436,
+      "step": 12512
+    },
+    {
+      "epoch": 2.0549955863937757,
+      "grad_norm": 0.3604424163085528,
+      "learning_rate": 5.821265104783929e-06,
+      "loss": 0.4686,
+      "step": 12513
+    },
+    {
+      "epoch": 2.0551598136021187,
+      "grad_norm": 0.3869786566160952,
+      "learning_rate": 5.820816795865848e-06,
+      "loss": 0.4518,
+      "step": 12514
+    },
+    {
+      "epoch": 2.0553240408104614,
+      "grad_norm": 0.31366177202073153,
+      "learning_rate": 5.8203684726023965e-06,
+      "loss": 0.4382,
+      "step": 12515
+    },
+    {
+      "epoch": 2.055488268018804,
+      "grad_norm": 0.3816416592520499,
+      "learning_rate": 5.819920134998445e-06,
+      "loss": 0.4486,
+      "step": 12516
+    },
+    {
+      "epoch": 2.0556524952271467,
+      "grad_norm": 0.36611323579211497,
+      "learning_rate": 5.819471783058861e-06,
+      "loss": 0.4636,
+      "step": 12517
+    },
+    {
+      "epoch": 2.0558167224354893,
+      "grad_norm": 0.3926812761164695,
+      "learning_rate": 5.8190234167885164e-06,
+      "loss": 0.4313,
+      "step": 12518
+    },
+    {
+      "epoch": 2.0559809496438324,
+      "grad_norm": 0.3325452728082553,
+      "learning_rate": 5.818575036192279e-06,
+      "loss": 0.4638,
+      "step": 12519
+    },
+    {
+      "epoch": 2.056145176852175,
+      "grad_norm": 0.42067798281032054,
+      "learning_rate": 5.81812664127502e-06,
+      "loss": 0.4637,
+      "step": 12520
+    },
+    {
+      "epoch": 2.0563094040605177,
+      "grad_norm": 0.318034404416098,
+      "learning_rate": 5.817678232041608e-06,
+      "loss": 0.4283,
+      "step": 12521
+    },
+    {
+      "epoch": 2.0564736312688603,
+      "grad_norm": 0.48940136510980775,
+      "learning_rate": 5.817229808496915e-06,
+      "loss": 0.4436,
+      "step": 12522
+    },
+    {
+      "epoch": 2.0566378584772034,
+      "grad_norm": 0.2882544661032187,
+      "learning_rate": 5.816781370645809e-06,
+      "loss": 0.4309,
+      "step": 12523
+    },
+    {
+      "epoch": 2.056802085685546,
+      "grad_norm": 0.5042845819900779,
+      "learning_rate": 5.816332918493164e-06,
+      "loss": 0.4497,
+      "step": 12524
+    },
+    {
+      "epoch": 2.0569663128938886,
+      "grad_norm": 0.29557746058877027,
+      "learning_rate": 5.815884452043846e-06,
+      "loss": 0.4365,
+      "step": 12525
+    },
+    {
+      "epoch": 2.0571305401022313,
+      "grad_norm": 0.33012923255531806,
+      "learning_rate": 5.81543597130273e-06,
+      "loss": 0.4569,
+      "step": 12526
+    },
+    {
+      "epoch": 2.0572947673105744,
+      "grad_norm": 0.4815314599818145,
+      "learning_rate": 5.8149874762746844e-06,
+      "loss": 0.4455,
+      "step": 12527
+    },
+    {
+      "epoch": 2.057458994518917,
+      "grad_norm": 0.30529200972710946,
+      "learning_rate": 5.814538966964581e-06,
+      "loss": 0.4355,
+      "step": 12528
+    },
+    {
+      "epoch": 2.0576232217272596,
+      "grad_norm": 0.2781313905407027,
+      "learning_rate": 5.814090443377291e-06,
+      "loss": 0.4394,
+      "step": 12529
+    },
+    {
+      "epoch": 2.0577874489356023,
+      "grad_norm": 0.33271003194262444,
+      "learning_rate": 5.813641905517687e-06,
+      "loss": 0.4605,
+      "step": 12530
+    },
+    {
+      "epoch": 2.0579516761439454,
+      "grad_norm": 0.33716505625152493,
+      "learning_rate": 5.813193353390637e-06,
+      "loss": 0.4708,
+      "step": 12531
+    },
+    {
+      "epoch": 2.058115903352288,
+      "grad_norm": 0.353896302279186,
+      "learning_rate": 5.812744787001017e-06,
+      "loss": 0.4373,
+      "step": 12532
+    },
+    {
+      "epoch": 2.0582801305606306,
+      "grad_norm": 0.58896944714371,
+      "learning_rate": 5.812296206353696e-06,
+      "loss": 0.4643,
+      "step": 12533
+    },
+    {
+      "epoch": 2.0584443577689733,
+      "grad_norm": 0.292337170150739,
+      "learning_rate": 5.811847611453549e-06,
+      "loss": 0.4405,
+      "step": 12534
+    },
+    {
+      "epoch": 2.058608584977316,
+      "grad_norm": 0.29791135402795793,
+      "learning_rate": 5.811399002305445e-06,
+      "loss": 0.4583,
+      "step": 12535
+    },
+    {
+      "epoch": 2.058772812185659,
+      "grad_norm": 0.29900219174218134,
+      "learning_rate": 5.810950378914256e-06,
+      "loss": 0.4568,
+      "step": 12536
+    },
+    {
+      "epoch": 2.0589370393940016,
+      "grad_norm": 0.28608728842480263,
+      "learning_rate": 5.810501741284858e-06,
+      "loss": 0.4557,
+      "step": 12537
+    },
+    {
+      "epoch": 2.0591012666023443,
+      "grad_norm": 0.3254239739716752,
+      "learning_rate": 5.8100530894221215e-06,
+      "loss": 0.4335,
+      "step": 12538
+    },
+    {
+      "epoch": 2.059265493810687,
+      "grad_norm": 1.5228037555250937,
+      "learning_rate": 5.80960442333092e-06,
+      "loss": 0.4402,
+      "step": 12539
+    },
+    {
+      "epoch": 2.05942972101903,
+      "grad_norm": 0.4226075001082411,
+      "learning_rate": 5.809155743016125e-06,
+      "loss": 0.4543,
+      "step": 12540
+    },
+    {
+      "epoch": 2.0595939482273726,
+      "grad_norm": 0.3498474294927714,
+      "learning_rate": 5.80870704848261e-06,
+      "loss": 0.4519,
+      "step": 12541
+    },
+    {
+      "epoch": 2.0597581754357153,
+      "grad_norm": 0.31951437366502333,
+      "learning_rate": 5.808258339735251e-06,
+      "loss": 0.4289,
+      "step": 12542
+    },
+    {
+      "epoch": 2.059922402644058,
+      "grad_norm": 0.3069067592827343,
+      "learning_rate": 5.807809616778918e-06,
+      "loss": 0.448,
+      "step": 12543
+    },
+    {
+      "epoch": 2.060086629852401,
+      "grad_norm": 0.34492891136792414,
+      "learning_rate": 5.807360879618486e-06,
+      "loss": 0.4223,
+      "step": 12544
+    },
+    {
+      "epoch": 2.0602508570607436,
+      "grad_norm": 0.37131805354396374,
+      "learning_rate": 5.806912128258828e-06,
+      "loss": 0.4387,
+      "step": 12545
+    },
+    {
+      "epoch": 2.0604150842690863,
+      "grad_norm": 0.34516089438021247,
+      "learning_rate": 5.806463362704819e-06,
+      "loss": 0.4268,
+      "step": 12546
+    },
+    {
+      "epoch": 2.060579311477429,
+      "grad_norm": 0.34676052883678987,
+      "learning_rate": 5.806014582961333e-06,
+      "loss": 0.4409,
+      "step": 12547
+    },
+    {
+      "epoch": 2.060743538685772,
+      "grad_norm": 0.32879785289901325,
+      "learning_rate": 5.805565789033244e-06,
+      "loss": 0.4446,
+      "step": 12548
+    },
+    {
+      "epoch": 2.0609077658941146,
+      "grad_norm": 0.33290467087799636,
+      "learning_rate": 5.805116980925425e-06,
+      "loss": 0.4405,
+      "step": 12549
+    },
+    {
+      "epoch": 2.0610719931024573,
+      "grad_norm": 0.37078533186888135,
+      "learning_rate": 5.80466815864275e-06,
+      "loss": 0.4466,
+      "step": 12550
+    },
+    {
+      "epoch": 2.0612362203108,
+      "grad_norm": 0.2745242092923762,
+      "learning_rate": 5.804219322190098e-06,
+      "loss": 0.4486,
+      "step": 12551
+    },
+    {
+      "epoch": 2.0614004475191425,
+      "grad_norm": 0.36904796252237204,
+      "learning_rate": 5.80377047157234e-06,
+      "loss": 0.4519,
+      "step": 12552
+    },
+    {
+      "epoch": 2.0615646747274856,
+      "grad_norm": 0.5335812773510453,
+      "learning_rate": 5.8033216067943515e-06,
+      "loss": 0.4558,
+      "step": 12553
+    },
+    {
+      "epoch": 2.0617289019358283,
+      "grad_norm": 0.3053953391869156,
+      "learning_rate": 5.802872727861009e-06,
+      "loss": 0.4586,
+      "step": 12554
+    },
+    {
+      "epoch": 2.061893129144171,
+      "grad_norm": 0.40520225583447983,
+      "learning_rate": 5.802423834777186e-06,
+      "loss": 0.4351,
+      "step": 12555
+    },
+    {
+      "epoch": 2.0620573563525135,
+      "grad_norm": 0.3988187826185149,
+      "learning_rate": 5.801974927547758e-06,
+      "loss": 0.4356,
+      "step": 12556
+    },
+    {
+      "epoch": 2.0622215835608566,
+      "grad_norm": 0.3380322407657435,
+      "learning_rate": 5.8015260061776024e-06,
+      "loss": 0.4604,
+      "step": 12557
+    },
+    {
+      "epoch": 2.0623858107691992,
+      "grad_norm": 0.3348430293607122,
+      "learning_rate": 5.801077070671595e-06,
+      "loss": 0.4293,
+      "step": 12558
+    },
+    {
+      "epoch": 2.062550037977542,
+      "grad_norm": 0.3842683264535958,
+      "learning_rate": 5.80062812103461e-06,
+      "loss": 0.4341,
+      "step": 12559
+    },
+    {
+      "epoch": 2.0627142651858845,
+      "grad_norm": 0.30200025238816425,
+      "learning_rate": 5.800179157271522e-06,
+      "loss": 0.4701,
+      "step": 12560
+    },
+    {
+      "epoch": 2.0628784923942276,
+      "grad_norm": 0.31908377463654947,
+      "learning_rate": 5.79973017938721e-06,
+      "loss": 0.4494,
+      "step": 12561
+    },
+    {
+      "epoch": 2.0630427196025702,
+      "grad_norm": 0.5005507434852071,
+      "learning_rate": 5.7992811873865496e-06,
+      "loss": 0.451,
+      "step": 12562
+    },
+    {
+      "epoch": 2.063206946810913,
+      "grad_norm": 0.366742905796703,
+      "learning_rate": 5.7988321812744175e-06,
+      "loss": 0.4317,
+      "step": 12563
+    },
+    {
+      "epoch": 2.0633711740192555,
+      "grad_norm": 0.5555018777431691,
+      "learning_rate": 5.798383161055691e-06,
+      "loss": 0.4484,
+      "step": 12564
+    },
+    {
+      "epoch": 2.0635354012275986,
+      "grad_norm": 0.3180625989819009,
+      "learning_rate": 5.797934126735244e-06,
+      "loss": 0.4381,
+      "step": 12565
+    },
+    {
+      "epoch": 2.0636996284359412,
+      "grad_norm": 0.32545313266240605,
+      "learning_rate": 5.797485078317956e-06,
+      "loss": 0.4591,
+      "step": 12566
+    },
+    {
+      "epoch": 2.063863855644284,
+      "grad_norm": 0.43023638563585087,
+      "learning_rate": 5.797036015808704e-06,
+      "loss": 0.4502,
+      "step": 12567
+    },
+    {
+      "epoch": 2.0640280828526265,
+      "grad_norm": 0.3400815349047949,
+      "learning_rate": 5.796586939212365e-06,
+      "loss": 0.458,
+      "step": 12568
+    },
+    {
+      "epoch": 2.064192310060969,
+      "grad_norm": 0.34321856939719503,
+      "learning_rate": 5.796137848533816e-06,
+      "loss": 0.4243,
+      "step": 12569
+    },
+    {
+      "epoch": 2.0643565372693122,
+      "grad_norm": 0.3930605011253571,
+      "learning_rate": 5.795688743777934e-06,
+      "loss": 0.4503,
+      "step": 12570
+    },
+    {
+      "epoch": 2.064520764477655,
+      "grad_norm": 0.3441430685584969,
+      "learning_rate": 5.795239624949597e-06,
+      "loss": 0.4525,
+      "step": 12571
+    },
+    {
+      "epoch": 2.0646849916859975,
+      "grad_norm": 0.3092235368191995,
+      "learning_rate": 5.794790492053685e-06,
+      "loss": 0.4407,
+      "step": 12572
+    },
+    {
+      "epoch": 2.06484921889434,
+      "grad_norm": 0.2956042164630155,
+      "learning_rate": 5.7943413450950745e-06,
+      "loss": 0.429,
+      "step": 12573
+    },
+    {
+      "epoch": 2.0650134461026832,
+      "grad_norm": 0.5785693723379479,
+      "learning_rate": 5.793892184078642e-06,
+      "loss": 0.4476,
+      "step": 12574
+    },
+    {
+      "epoch": 2.065177673311026,
+      "grad_norm": 0.4050586734624316,
+      "learning_rate": 5.793443009009268e-06,
+      "loss": 0.4385,
+      "step": 12575
+    },
+    {
+      "epoch": 2.0653419005193685,
+      "grad_norm": 0.44417807478608184,
+      "learning_rate": 5.792993819891831e-06,
+      "loss": 0.4655,
+      "step": 12576
+    },
+    {
+      "epoch": 2.065506127727711,
+      "grad_norm": 0.32094819856464846,
+      "learning_rate": 5.792544616731208e-06,
+      "loss": 0.4371,
+      "step": 12577
+    },
+    {
+      "epoch": 2.065670354936054,
+      "grad_norm": 0.3225322891675443,
+      "learning_rate": 5.792095399532279e-06,
+      "loss": 0.4388,
+      "step": 12578
+    },
+    {
+      "epoch": 2.065834582144397,
+      "grad_norm": 0.3611766104446534,
+      "learning_rate": 5.791646168299923e-06,
+      "loss": 0.44,
+      "step": 12579
+    },
+    {
+      "epoch": 2.0659988093527395,
+      "grad_norm": 0.41265178472084757,
+      "learning_rate": 5.791196923039019e-06,
+      "loss": 0.4262,
+      "step": 12580
+    },
+    {
+      "epoch": 2.066163036561082,
+      "grad_norm": 0.5685535710349041,
+      "learning_rate": 5.790747663754445e-06,
+      "loss": 0.4407,
+      "step": 12581
+    },
+    {
+      "epoch": 2.066327263769425,
+      "grad_norm": 0.4290513029494653,
+      "learning_rate": 5.790298390451083e-06,
+      "loss": 0.4375,
+      "step": 12582
+    },
+    {
+      "epoch": 2.066491490977768,
+      "grad_norm": 0.31193517737081866,
+      "learning_rate": 5.78984910313381e-06,
+      "loss": 0.4455,
+      "step": 12583
+    },
+    {
+      "epoch": 2.0666557181861105,
+      "grad_norm": 0.7311649893085822,
+      "learning_rate": 5.789399801807506e-06,
+      "loss": 0.4621,
+      "step": 12584
+    },
+    {
+      "epoch": 2.066819945394453,
+      "grad_norm": 0.35584981635411755,
+      "learning_rate": 5.7889504864770525e-06,
+      "loss": 0.435,
+      "step": 12585
+    },
+    {
+      "epoch": 2.0669841726027958,
+      "grad_norm": 0.34996905818049884,
+      "learning_rate": 5.788501157147328e-06,
+      "loss": 0.4436,
+      "step": 12586
+    },
+    {
+      "epoch": 2.067148399811139,
+      "grad_norm": 0.2944915229752169,
+      "learning_rate": 5.788051813823214e-06,
+      "loss": 0.4583,
+      "step": 12587
+    },
+    {
+      "epoch": 2.0673126270194815,
+      "grad_norm": 0.34794445160376564,
+      "learning_rate": 5.78760245650959e-06,
+      "loss": 0.4608,
+      "step": 12588
+    },
+    {
+      "epoch": 2.067476854227824,
+      "grad_norm": 0.31060770589110037,
+      "learning_rate": 5.787153085211336e-06,
+      "loss": 0.4636,
+      "step": 12589
+    },
+    {
+      "epoch": 2.0676410814361668,
+      "grad_norm": 0.36268150169000246,
+      "learning_rate": 5.786703699933333e-06,
+      "loss": 0.4428,
+      "step": 12590
+    },
+    {
+      "epoch": 2.06780530864451,
+      "grad_norm": 0.4948601709280069,
+      "learning_rate": 5.786254300680463e-06,
+      "loss": 0.4552,
+      "step": 12591
+    },
+    {
+      "epoch": 2.0679695358528525,
+      "grad_norm": 0.8228867505642475,
+      "learning_rate": 5.785804887457604e-06,
+      "loss": 0.4567,
+      "step": 12592
+    },
+    {
+      "epoch": 2.068133763061195,
+      "grad_norm": 0.33416355670733,
+      "learning_rate": 5.78535546026964e-06,
+      "loss": 0.4259,
+      "step": 12593
+    },
+    {
+      "epoch": 2.0682979902695378,
+      "grad_norm": 0.4689217739481361,
+      "learning_rate": 5.784906019121451e-06,
+      "loss": 0.447,
+      "step": 12594
+    },
+    {
+      "epoch": 2.068462217477881,
+      "grad_norm": 0.32276567150550434,
+      "learning_rate": 5.784456564017918e-06,
+      "loss": 0.4617,
+      "step": 12595
+    },
+    {
+      "epoch": 2.0686264446862235,
+      "grad_norm": 0.3707358256052318,
+      "learning_rate": 5.784007094963924e-06,
+      "loss": 0.4451,
+      "step": 12596
+    },
+    {
+      "epoch": 2.068790671894566,
+      "grad_norm": 0.7082352154662992,
+      "learning_rate": 5.783557611964349e-06,
+      "loss": 0.457,
+      "step": 12597
+    },
+    {
+      "epoch": 2.0689548991029088,
+      "grad_norm": 0.3352628121247371,
+      "learning_rate": 5.783108115024076e-06,
+      "loss": 0.4379,
+      "step": 12598
+    },
+    {
+      "epoch": 2.069119126311252,
+      "grad_norm": 0.33373071703006074,
+      "learning_rate": 5.782658604147985e-06,
+      "loss": 0.4269,
+      "step": 12599
+    },
+    {
+      "epoch": 2.0692833535195945,
+      "grad_norm": 0.3890245550176744,
+      "learning_rate": 5.78220907934096e-06,
+      "loss": 0.4238,
+      "step": 12600
+    },
+    {
+      "epoch": 2.069447580727937,
+      "grad_norm": 0.42551587495552057,
+      "learning_rate": 5.781759540607884e-06,
+      "loss": 0.4379,
+      "step": 12601
+    },
+    {
+      "epoch": 2.0696118079362797,
+      "grad_norm": 0.3848246045704332,
+      "learning_rate": 5.781309987953638e-06,
+      "loss": 0.4355,
+      "step": 12602
+    },
+    {
+      "epoch": 2.0697760351446224,
+      "grad_norm": 0.39510346983773065,
+      "learning_rate": 5.780860421383105e-06,
+      "loss": 0.445,
+      "step": 12603
+    },
+    {
+      "epoch": 2.0699402623529655,
+      "grad_norm": 0.5293818907890222,
+      "learning_rate": 5.780410840901166e-06,
+      "loss": 0.4388,
+      "step": 12604
+    },
+    {
+      "epoch": 2.070104489561308,
+      "grad_norm": 0.43186625564440984,
+      "learning_rate": 5.779961246512707e-06,
+      "loss": 0.4424,
+      "step": 12605
+    },
+    {
+      "epoch": 2.0702687167696507,
+      "grad_norm": 0.35612676312621716,
+      "learning_rate": 5.77951163822261e-06,
+      "loss": 0.4324,
+      "step": 12606
+    },
+    {
+      "epoch": 2.0704329439779934,
+      "grad_norm": 0.44096187888292104,
+      "learning_rate": 5.779062016035756e-06,
+      "loss": 0.459,
+      "step": 12607
+    },
+    {
+      "epoch": 2.0705971711863365,
+      "grad_norm": 0.3641439511112035,
+      "learning_rate": 5.7786123799570305e-06,
+      "loss": 0.4621,
+      "step": 12608
+    },
+    {
+      "epoch": 2.070761398394679,
+      "grad_norm": 0.48210480440260733,
+      "learning_rate": 5.778162729991317e-06,
+      "loss": 0.4358,
+      "step": 12609
+    },
+    {
+      "epoch": 2.0709256256030217,
+      "grad_norm": 0.40744258619042256,
+      "learning_rate": 5.7777130661435004e-06,
+      "loss": 0.4298,
+      "step": 12610
+    },
+    {
+      "epoch": 2.0710898528113644,
+      "grad_norm": 0.2950203895635282,
+      "learning_rate": 5.777263388418461e-06,
+      "loss": 0.4276,
+      "step": 12611
+    },
+    {
+      "epoch": 2.0712540800197075,
+      "grad_norm": 0.3186235372237503,
+      "learning_rate": 5.776813696821085e-06,
+      "loss": 0.4577,
+      "step": 12612
+    },
+    {
+      "epoch": 2.07141830722805,
+      "grad_norm": 0.39173521730263494,
+      "learning_rate": 5.776363991356255e-06,
+      "loss": 0.4817,
+      "step": 12613
+    },
+    {
+      "epoch": 2.0715825344363927,
+      "grad_norm": 0.37345701908829565,
+      "learning_rate": 5.7759142720288586e-06,
+      "loss": 0.4396,
+      "step": 12614
+    },
+    {
+      "epoch": 2.0717467616447354,
+      "grad_norm": 0.6131889982299739,
+      "learning_rate": 5.775464538843775e-06,
+      "loss": 0.4282,
+      "step": 12615
+    },
+    {
+      "epoch": 2.0719109888530785,
+      "grad_norm": 0.4123409725560399,
+      "learning_rate": 5.775014791805894e-06,
+      "loss": 0.4483,
+      "step": 12616
+    },
+    {
+      "epoch": 2.072075216061421,
+      "grad_norm": 0.3181460531036635,
+      "learning_rate": 5.7745650309200965e-06,
+      "loss": 0.4546,
+      "step": 12617
+    },
+    {
+      "epoch": 2.0722394432697637,
+      "grad_norm": 0.34628966447613285,
+      "learning_rate": 5.77411525619127e-06,
+      "loss": 0.4405,
+      "step": 12618
+    },
+    {
+      "epoch": 2.0724036704781064,
+      "grad_norm": 0.32934812441671757,
+      "learning_rate": 5.773665467624296e-06,
+      "loss": 0.4236,
+      "step": 12619
+    },
+    {
+      "epoch": 2.072567897686449,
+      "grad_norm": 0.3320241576727553,
+      "learning_rate": 5.7732156652240635e-06,
+      "loss": 0.4419,
+      "step": 12620
+    },
+    {
+      "epoch": 2.072732124894792,
+      "grad_norm": 0.3751110180170982,
+      "learning_rate": 5.772765848995457e-06,
+      "loss": 0.42,
+      "step": 12621
+    },
+    {
+      "epoch": 2.0728963521031347,
+      "grad_norm": 0.3094726411805009,
+      "learning_rate": 5.772316018943361e-06,
+      "loss": 0.452,
+      "step": 12622
+    },
+    {
+      "epoch": 2.0730605793114774,
+      "grad_norm": 0.38697336585851455,
+      "learning_rate": 5.771866175072659e-06,
+      "loss": 0.4264,
+      "step": 12623
+    },
+    {
+      "epoch": 2.07322480651982,
+      "grad_norm": 0.304288316654996,
+      "learning_rate": 5.77141631738824e-06,
+      "loss": 0.4484,
+      "step": 12624
+    },
+    {
+      "epoch": 2.073389033728163,
+      "grad_norm": 0.482532511700392,
+      "learning_rate": 5.770966445894991e-06,
+      "loss": 0.4545,
+      "step": 12625
+    },
+    {
+      "epoch": 2.0735532609365057,
+      "grad_norm": 0.32262406085428247,
+      "learning_rate": 5.770516560597794e-06,
+      "loss": 0.4462,
+      "step": 12626
+    },
+    {
+      "epoch": 2.0737174881448484,
+      "grad_norm": 0.3325404739646796,
+      "learning_rate": 5.770066661501538e-06,
+      "loss": 0.4239,
+      "step": 12627
+    },
+    {
+      "epoch": 2.073881715353191,
+      "grad_norm": 1.9006644872008762,
+      "learning_rate": 5.769616748611106e-06,
+      "loss": 0.4709,
+      "step": 12628
+    },
+    {
+      "epoch": 2.074045942561534,
+      "grad_norm": 0.3423420126583943,
+      "learning_rate": 5.769166821931389e-06,
+      "loss": 0.4424,
+      "step": 12629
+    },
+    {
+      "epoch": 2.0742101697698767,
+      "grad_norm": 0.29731610057754654,
+      "learning_rate": 5.7687168814672726e-06,
+      "loss": 0.4483,
+      "step": 12630
+    },
+    {
+      "epoch": 2.0743743969782193,
+      "grad_norm": 0.3874440450920287,
+      "learning_rate": 5.768266927223642e-06,
+      "loss": 0.4681,
+      "step": 12631
+    },
+    {
+      "epoch": 2.074538624186562,
+      "grad_norm": 0.3297323496352712,
+      "learning_rate": 5.767816959205384e-06,
+      "loss": 0.4278,
+      "step": 12632
+    },
+    {
+      "epoch": 2.074702851394905,
+      "grad_norm": 0.3816910179675114,
+      "learning_rate": 5.767366977417386e-06,
+      "loss": 0.4577,
+      "step": 12633
+    },
+    {
+      "epoch": 2.0748670786032477,
+      "grad_norm": 0.4387489677520473,
+      "learning_rate": 5.766916981864536e-06,
+      "loss": 0.4406,
+      "step": 12634
+    },
+    {
+      "epoch": 2.0750313058115903,
+      "grad_norm": 0.3323500901032443,
+      "learning_rate": 5.7664669725517215e-06,
+      "loss": 0.4474,
+      "step": 12635
+    },
+    {
+      "epoch": 2.075195533019933,
+      "grad_norm": 0.4094691211689768,
+      "learning_rate": 5.766016949483831e-06,
+      "loss": 0.4354,
+      "step": 12636
+    },
+    {
+      "epoch": 2.0753597602282756,
+      "grad_norm": 0.3247612685603506,
+      "learning_rate": 5.765566912665748e-06,
+      "loss": 0.4453,
+      "step": 12637
+    },
+    {
+      "epoch": 2.0755239874366187,
+      "grad_norm": 0.2992283474155525,
+      "learning_rate": 5.765116862102365e-06,
+      "loss": 0.4508,
+      "step": 12638
+    },
+    {
+      "epoch": 2.0756882146449613,
+      "grad_norm": 0.2833987945361159,
+      "learning_rate": 5.764666797798569e-06,
+      "loss": 0.4493,
+      "step": 12639
+    },
+    {
+      "epoch": 2.075852441853304,
+      "grad_norm": 0.36076810908372076,
+      "learning_rate": 5.764216719759246e-06,
+      "loss": 0.4424,
+      "step": 12640
+    },
+    {
+      "epoch": 2.0760166690616466,
+      "grad_norm": 0.33496405013392627,
+      "learning_rate": 5.763766627989285e-06,
+      "loss": 0.4376,
+      "step": 12641
+    },
+    {
+      "epoch": 2.0761808962699897,
+      "grad_norm": 0.3664945473785977,
+      "learning_rate": 5.763316522493576e-06,
+      "loss": 0.4511,
+      "step": 12642
+    },
+    {
+      "epoch": 2.0763451234783323,
+      "grad_norm": 0.3160831448540927,
+      "learning_rate": 5.7628664032770066e-06,
+      "loss": 0.4547,
+      "step": 12643
+    },
+    {
+      "epoch": 2.076509350686675,
+      "grad_norm": 0.3800779021900539,
+      "learning_rate": 5.7624162703444655e-06,
+      "loss": 0.4571,
+      "step": 12644
+    },
+    {
+      "epoch": 2.0766735778950176,
+      "grad_norm": 0.3096148612282082,
+      "learning_rate": 5.761966123700843e-06,
+      "loss": 0.4465,
+      "step": 12645
+    },
+    {
+      "epoch": 2.0768378051033607,
+      "grad_norm": 0.4700276065412499,
+      "learning_rate": 5.761515963351024e-06,
+      "loss": 0.4371,
+      "step": 12646
+    },
+    {
+      "epoch": 2.0770020323117033,
+      "grad_norm": 0.5717112387339418,
+      "learning_rate": 5.761065789299902e-06,
+      "loss": 0.4733,
+      "step": 12647
+    },
+    {
+      "epoch": 2.077166259520046,
+      "grad_norm": 0.3478262531191078,
+      "learning_rate": 5.760615601552365e-06,
+      "loss": 0.4252,
+      "step": 12648
+    },
+    {
+      "epoch": 2.0773304867283886,
+      "grad_norm": 0.4208026301767626,
+      "learning_rate": 5.760165400113301e-06,
+      "loss": 0.4344,
+      "step": 12649
+    },
+    {
+      "epoch": 2.0774947139367317,
+      "grad_norm": 0.3241596365519559,
+      "learning_rate": 5.759715184987602e-06,
+      "loss": 0.4352,
+      "step": 12650
+    },
+    {
+      "epoch": 2.0776589411450743,
+      "grad_norm": 0.40095490723714017,
+      "learning_rate": 5.7592649561801576e-06,
+      "loss": 0.4414,
+      "step": 12651
+    },
+    {
+      "epoch": 2.077823168353417,
+      "grad_norm": 0.34814818074051035,
+      "learning_rate": 5.7588147136958555e-06,
+      "loss": 0.4656,
+      "step": 12652
+    },
+    {
+      "epoch": 2.0779873955617596,
+      "grad_norm": 0.5480012497939601,
+      "learning_rate": 5.758364457539587e-06,
+      "loss": 0.4658,
+      "step": 12653
+    },
+    {
+      "epoch": 2.0781516227701022,
+      "grad_norm": 0.4022849269106388,
+      "learning_rate": 5.757914187716242e-06,
+      "loss": 0.4481,
+      "step": 12654
+    },
+    {
+      "epoch": 2.0783158499784453,
+      "grad_norm": 0.3296491558960112,
+      "learning_rate": 5.757463904230713e-06,
+      "loss": 0.4312,
+      "step": 12655
+    },
+    {
+      "epoch": 2.078480077186788,
+      "grad_norm": 0.291526171977682,
+      "learning_rate": 5.757013607087888e-06,
+      "loss": 0.4658,
+      "step": 12656
+    },
+    {
+      "epoch": 2.0786443043951306,
+      "grad_norm": 0.3151420850285067,
+      "learning_rate": 5.756563296292658e-06,
+      "loss": 0.4162,
+      "step": 12657
+    },
+    {
+      "epoch": 2.0788085316034732,
+      "grad_norm": 0.7250952171515853,
+      "learning_rate": 5.756112971849915e-06,
+      "loss": 0.4579,
+      "step": 12658
+    },
+    {
+      "epoch": 2.0789727588118163,
+      "grad_norm": 0.323930778535311,
+      "learning_rate": 5.755662633764549e-06,
+      "loss": 0.4332,
+      "step": 12659
+    },
+    {
+      "epoch": 2.079136986020159,
+      "grad_norm": 0.3319897590687082,
+      "learning_rate": 5.755212282041452e-06,
+      "loss": 0.4634,
+      "step": 12660
+    },
+    {
+      "epoch": 2.0793012132285016,
+      "grad_norm": 0.2965645629673636,
+      "learning_rate": 5.754761916685515e-06,
+      "loss": 0.4494,
+      "step": 12661
+    },
+    {
+      "epoch": 2.0794654404368442,
+      "grad_norm": 0.4789465039629421,
+      "learning_rate": 5.754311537701626e-06,
+      "loss": 0.4645,
+      "step": 12662
+    },
+    {
+      "epoch": 2.0796296676451873,
+      "grad_norm": 0.3051466720049543,
+      "learning_rate": 5.753861145094682e-06,
+      "loss": 0.4367,
+      "step": 12663
+    },
+    {
+      "epoch": 2.07979389485353,
+      "grad_norm": 0.34364344780338235,
+      "learning_rate": 5.753410738869573e-06,
+      "loss": 0.4287,
+      "step": 12664
+    },
+    {
+      "epoch": 2.0799581220618726,
+      "grad_norm": 0.2942069240384221,
+      "learning_rate": 5.75296031903119e-06,
+      "loss": 0.4331,
+      "step": 12665
+    },
+    {
+      "epoch": 2.080122349270215,
+      "grad_norm": 0.4048541793023706,
+      "learning_rate": 5.752509885584423e-06,
+      "loss": 0.4461,
+      "step": 12666
+    },
+    {
+      "epoch": 2.0802865764785583,
+      "grad_norm": 0.3473334256844807,
+      "learning_rate": 5.752059438534168e-06,
+      "loss": 0.4515,
+      "step": 12667
+    },
+    {
+      "epoch": 2.080450803686901,
+      "grad_norm": 0.33238636058550425,
+      "learning_rate": 5.751608977885315e-06,
+      "loss": 0.4528,
+      "step": 12668
+    },
+    {
+      "epoch": 2.0806150308952436,
+      "grad_norm": 0.3399610070481964,
+      "learning_rate": 5.751158503642758e-06,
+      "loss": 0.4663,
+      "step": 12669
+    },
+    {
+      "epoch": 2.080779258103586,
+      "grad_norm": 0.44808401171722956,
+      "learning_rate": 5.750708015811389e-06,
+      "loss": 0.4313,
+      "step": 12670
+    },
+    {
+      "epoch": 2.080943485311929,
+      "grad_norm": 0.30559942252118855,
+      "learning_rate": 5.7502575143960985e-06,
+      "loss": 0.4576,
+      "step": 12671
+    },
+    {
+      "epoch": 2.081107712520272,
+      "grad_norm": 0.29963645365849745,
+      "learning_rate": 5.749806999401783e-06,
+      "loss": 0.4201,
+      "step": 12672
+    },
+    {
+      "epoch": 2.0812719397286146,
+      "grad_norm": 0.30932198493791446,
+      "learning_rate": 5.7493564708333324e-06,
+      "loss": 0.4507,
+      "step": 12673
+    },
+    {
+      "epoch": 2.081436166936957,
+      "grad_norm": 0.38662130188894195,
+      "learning_rate": 5.748905928695643e-06,
+      "loss": 0.4598,
+      "step": 12674
+    },
+    {
+      "epoch": 2.0816003941453,
+      "grad_norm": 0.3130379513175994,
+      "learning_rate": 5.748455372993606e-06,
+      "loss": 0.4262,
+      "step": 12675
+    },
+    {
+      "epoch": 2.081764621353643,
+      "grad_norm": 0.28607888291252975,
+      "learning_rate": 5.748004803732115e-06,
+      "loss": 0.4517,
+      "step": 12676
+    },
+    {
+      "epoch": 2.0819288485619856,
+      "grad_norm": 0.2895532108578058,
+      "learning_rate": 5.747554220916065e-06,
+      "loss": 0.4261,
+      "step": 12677
+    },
+    {
+      "epoch": 2.082093075770328,
+      "grad_norm": 0.4363820417166602,
+      "learning_rate": 5.747103624550347e-06,
+      "loss": 0.4455,
+      "step": 12678
+    },
+    {
+      "epoch": 2.082257302978671,
+      "grad_norm": 0.35167816630302146,
+      "learning_rate": 5.7466530146398576e-06,
+      "loss": 0.4309,
+      "step": 12679
+    },
+    {
+      "epoch": 2.082421530187014,
+      "grad_norm": 0.3577392213525,
+      "learning_rate": 5.746202391189488e-06,
+      "loss": 0.4519,
+      "step": 12680
+    },
+    {
+      "epoch": 2.0825857573953566,
+      "grad_norm": 0.33848902980111123,
+      "learning_rate": 5.745751754204137e-06,
+      "loss": 0.4325,
+      "step": 12681
+    },
+    {
+      "epoch": 2.082749984603699,
+      "grad_norm": 0.3605086606372533,
+      "learning_rate": 5.7453011036886955e-06,
+      "loss": 0.4356,
+      "step": 12682
+    },
+    {
+      "epoch": 2.082914211812042,
+      "grad_norm": 0.35835761281442624,
+      "learning_rate": 5.744850439648058e-06,
+      "loss": 0.4476,
+      "step": 12683
+    },
+    {
+      "epoch": 2.083078439020385,
+      "grad_norm": 0.29445262431486496,
+      "learning_rate": 5.744399762087121e-06,
+      "loss": 0.4385,
+      "step": 12684
+    },
+    {
+      "epoch": 2.0832426662287276,
+      "grad_norm": 0.2838943444170344,
+      "learning_rate": 5.7439490710107785e-06,
+      "loss": 0.4305,
+      "step": 12685
+    },
+    {
+      "epoch": 2.08340689343707,
+      "grad_norm": 0.39643038537134845,
+      "learning_rate": 5.743498366423923e-06,
+      "loss": 0.4353,
+      "step": 12686
+    },
+    {
+      "epoch": 2.083571120645413,
+      "grad_norm": 0.26822456161375324,
+      "learning_rate": 5.743047648331453e-06,
+      "loss": 0.4378,
+      "step": 12687
+    },
+    {
+      "epoch": 2.0837353478537555,
+      "grad_norm": 0.3759392038585772,
+      "learning_rate": 5.742596916738263e-06,
+      "loss": 0.4605,
+      "step": 12688
+    },
+    {
+      "epoch": 2.0838995750620986,
+      "grad_norm": 0.39420292932984524,
+      "learning_rate": 5.742146171649249e-06,
+      "loss": 0.4342,
+      "step": 12689
+    },
+    {
+      "epoch": 2.084063802270441,
+      "grad_norm": 0.4549535925788061,
+      "learning_rate": 5.741695413069304e-06,
+      "loss": 0.4289,
+      "step": 12690
+    },
+    {
+      "epoch": 2.084228029478784,
+      "grad_norm": 0.3253546959983401,
+      "learning_rate": 5.741244641003325e-06,
+      "loss": 0.439,
+      "step": 12691
+    },
+    {
+      "epoch": 2.0843922566871265,
+      "grad_norm": 0.2965903307460332,
+      "learning_rate": 5.740793855456207e-06,
+      "loss": 0.4389,
+      "step": 12692
+    },
+    {
+      "epoch": 2.0845564838954695,
+      "grad_norm": 0.32124029350309213,
+      "learning_rate": 5.74034305643285e-06,
+      "loss": 0.4189,
+      "step": 12693
+    },
+    {
+      "epoch": 2.084720711103812,
+      "grad_norm": 0.3386294509403138,
+      "learning_rate": 5.7398922439381455e-06,
+      "loss": 0.4544,
+      "step": 12694
+    },
+    {
+      "epoch": 2.084884938312155,
+      "grad_norm": 0.316581751906967,
+      "learning_rate": 5.739441417976991e-06,
+      "loss": 0.4422,
+      "step": 12695
+    },
+    {
+      "epoch": 2.0850491655204975,
+      "grad_norm": 0.3995425057420119,
+      "learning_rate": 5.738990578554283e-06,
+      "loss": 0.435,
+      "step": 12696
+    },
+    {
+      "epoch": 2.0852133927288405,
+      "grad_norm": 0.41720524884760174,
+      "learning_rate": 5.73853972567492e-06,
+      "loss": 0.4638,
+      "step": 12697
+    },
+    {
+      "epoch": 2.085377619937183,
+      "grad_norm": 0.34690086468609344,
+      "learning_rate": 5.738088859343795e-06,
+      "loss": 0.4625,
+      "step": 12698
+    },
+    {
+      "epoch": 2.085541847145526,
+      "grad_norm": 0.3311064860378474,
+      "learning_rate": 5.737637979565808e-06,
+      "loss": 0.4471,
+      "step": 12699
+    },
+    {
+      "epoch": 2.0857060743538685,
+      "grad_norm": 0.3264983757330153,
+      "learning_rate": 5.737187086345854e-06,
+      "loss": 0.4497,
+      "step": 12700
+    },
+    {
+      "epoch": 2.0858703015622115,
+      "grad_norm": 0.44699065895405937,
+      "learning_rate": 5.736736179688833e-06,
+      "loss": 0.4485,
+      "step": 12701
+    },
+    {
+      "epoch": 2.086034528770554,
+      "grad_norm": 0.34545500806939633,
+      "learning_rate": 5.736285259599639e-06,
+      "loss": 0.4479,
+      "step": 12702
+    },
+    {
+      "epoch": 2.086198755978897,
+      "grad_norm": 0.39482652743473867,
+      "learning_rate": 5.73583432608317e-06,
+      "loss": 0.4422,
+      "step": 12703
+    },
+    {
+      "epoch": 2.0863629831872395,
+      "grad_norm": 0.341075659467438,
+      "learning_rate": 5.735383379144325e-06,
+      "loss": 0.4414,
+      "step": 12704
+    },
+    {
+      "epoch": 2.086527210395582,
+      "grad_norm": 0.3309139894092436,
+      "learning_rate": 5.734932418788001e-06,
+      "loss": 0.4444,
+      "step": 12705
+    },
+    {
+      "epoch": 2.086691437603925,
+      "grad_norm": 0.3247661891625629,
+      "learning_rate": 5.734481445019097e-06,
+      "loss": 0.4396,
+      "step": 12706
+    },
+    {
+      "epoch": 2.086855664812268,
+      "grad_norm": 0.326532593618532,
+      "learning_rate": 5.734030457842508e-06,
+      "loss": 0.4546,
+      "step": 12707
+    },
+    {
+      "epoch": 2.0870198920206104,
+      "grad_norm": 0.3922216807175495,
+      "learning_rate": 5.733579457263135e-06,
+      "loss": 0.4629,
+      "step": 12708
+    },
+    {
+      "epoch": 2.087184119228953,
+      "grad_norm": 0.394354893553275,
+      "learning_rate": 5.7331284432858755e-06,
+      "loss": 0.4522,
+      "step": 12709
+    },
+    {
+      "epoch": 2.087348346437296,
+      "grad_norm": 0.402018882323647,
+      "learning_rate": 5.7326774159156275e-06,
+      "loss": 0.4489,
+      "step": 12710
+    },
+    {
+      "epoch": 2.087512573645639,
+      "grad_norm": 0.33228228472187143,
+      "learning_rate": 5.73222637515729e-06,
+      "loss": 0.4263,
+      "step": 12711
+    },
+    {
+      "epoch": 2.0876768008539814,
+      "grad_norm": 0.4159735965654368,
+      "learning_rate": 5.731775321015762e-06,
+      "loss": 0.4548,
+      "step": 12712
+    },
+    {
+      "epoch": 2.087841028062324,
+      "grad_norm": 0.4843775098777932,
+      "learning_rate": 5.731324253495942e-06,
+      "loss": 0.4442,
+      "step": 12713
+    },
+    {
+      "epoch": 2.088005255270667,
+      "grad_norm": 0.33755449261044895,
+      "learning_rate": 5.730873172602731e-06,
+      "loss": 0.4554,
+      "step": 12714
+    },
+    {
+      "epoch": 2.08816948247901,
+      "grad_norm": 0.4117674329873463,
+      "learning_rate": 5.730422078341024e-06,
+      "loss": 0.4238,
+      "step": 12715
+    },
+    {
+      "epoch": 2.0883337096873524,
+      "grad_norm": 0.3001693197707603,
+      "learning_rate": 5.729970970715722e-06,
+      "loss": 0.4425,
+      "step": 12716
+    },
+    {
+      "epoch": 2.088497936895695,
+      "grad_norm": 0.3338517787117876,
+      "learning_rate": 5.729519849731726e-06,
+      "loss": 0.4456,
+      "step": 12717
+    },
+    {
+      "epoch": 2.088662164104038,
+      "grad_norm": 0.42389755185789696,
+      "learning_rate": 5.729068715393936e-06,
+      "loss": 0.4574,
+      "step": 12718
+    },
+    {
+      "epoch": 2.088826391312381,
+      "grad_norm": 0.44123544041308665,
+      "learning_rate": 5.72861756770725e-06,
+      "loss": 0.4652,
+      "step": 12719
+    },
+    {
+      "epoch": 2.0889906185207234,
+      "grad_norm": 0.35369014153572237,
+      "learning_rate": 5.7281664066765675e-06,
+      "loss": 0.4426,
+      "step": 12720
+    },
+    {
+      "epoch": 2.089154845729066,
+      "grad_norm": 0.3249641567404478,
+      "learning_rate": 5.727715232306789e-06,
+      "loss": 0.4383,
+      "step": 12721
+    },
+    {
+      "epoch": 2.0893190729374087,
+      "grad_norm": 0.32121212235146046,
+      "learning_rate": 5.727264044602817e-06,
+      "loss": 0.4364,
+      "step": 12722
+    },
+    {
+      "epoch": 2.089483300145752,
+      "grad_norm": 0.3503459484904393,
+      "learning_rate": 5.72681284356955e-06,
+      "loss": 0.454,
+      "step": 12723
+    },
+    {
+      "epoch": 2.0896475273540944,
+      "grad_norm": 0.3292701887707263,
+      "learning_rate": 5.726361629211887e-06,
+      "loss": 0.4371,
+      "step": 12724
+    },
+    {
+      "epoch": 2.089811754562437,
+      "grad_norm": 0.29747395088568335,
+      "learning_rate": 5.7259104015347315e-06,
+      "loss": 0.4453,
+      "step": 12725
+    },
+    {
+      "epoch": 2.0899759817707797,
+      "grad_norm": 0.34403589806070617,
+      "learning_rate": 5.725459160542981e-06,
+      "loss": 0.433,
+      "step": 12726
+    },
+    {
+      "epoch": 2.090140208979123,
+      "grad_norm": 0.3846893582906354,
+      "learning_rate": 5.72500790624154e-06,
+      "loss": 0.4345,
+      "step": 12727
+    },
+    {
+      "epoch": 2.0903044361874654,
+      "grad_norm": 0.3406838298169791,
+      "learning_rate": 5.724556638635308e-06,
+      "loss": 0.4279,
+      "step": 12728
+    },
+    {
+      "epoch": 2.090468663395808,
+      "grad_norm": 0.4079214046800453,
+      "learning_rate": 5.724105357729185e-06,
+      "loss": 0.4468,
+      "step": 12729
+    },
+    {
+      "epoch": 2.0906328906041507,
+      "grad_norm": 0.3143000270751584,
+      "learning_rate": 5.723654063528074e-06,
+      "loss": 0.4315,
+      "step": 12730
+    },
+    {
+      "epoch": 2.0907971178124938,
+      "grad_norm": 0.32459926896482116,
+      "learning_rate": 5.723202756036876e-06,
+      "loss": 0.4371,
+      "step": 12731
+    },
+    {
+      "epoch": 2.0909613450208364,
+      "grad_norm": 0.31723359855263167,
+      "learning_rate": 5.722751435260493e-06,
+      "loss": 0.4486,
+      "step": 12732
+    },
+    {
+      "epoch": 2.091125572229179,
+      "grad_norm": 0.3074426863744088,
+      "learning_rate": 5.7223001012038254e-06,
+      "loss": 0.4255,
+      "step": 12733
+    },
+    {
+      "epoch": 2.0912897994375217,
+      "grad_norm": 0.3509154461452414,
+      "learning_rate": 5.721848753871777e-06,
+      "loss": 0.4373,
+      "step": 12734
+    },
+    {
+      "epoch": 2.0914540266458648,
+      "grad_norm": 0.35720034447561816,
+      "learning_rate": 5.721397393269249e-06,
+      "loss": 0.4243,
+      "step": 12735
+    },
+    {
+      "epoch": 2.0916182538542074,
+      "grad_norm": 0.3614472189732083,
+      "learning_rate": 5.720946019401143e-06,
+      "loss": 0.4608,
+      "step": 12736
+    },
+    {
+      "epoch": 2.09178248106255,
+      "grad_norm": 0.332457995357455,
+      "learning_rate": 5.720494632272363e-06,
+      "loss": 0.4511,
+      "step": 12737
+    },
+    {
+      "epoch": 2.0919467082708927,
+      "grad_norm": 0.4137143471345643,
+      "learning_rate": 5.720043231887808e-06,
+      "loss": 0.4487,
+      "step": 12738
+    },
+    {
+      "epoch": 2.0921109354792353,
+      "grad_norm": 0.30603264772356137,
+      "learning_rate": 5.719591818252387e-06,
+      "loss": 0.4734,
+      "step": 12739
+    },
+    {
+      "epoch": 2.0922751626875784,
+      "grad_norm": 0.40363108857332247,
+      "learning_rate": 5.719140391370996e-06,
+      "loss": 0.4526,
+      "step": 12740
+    },
+    {
+      "epoch": 2.092439389895921,
+      "grad_norm": 0.38789927188689594,
+      "learning_rate": 5.718688951248541e-06,
+      "loss": 0.4449,
+      "step": 12741
+    },
+    {
+      "epoch": 2.0926036171042637,
+      "grad_norm": 0.3309222451571833,
+      "learning_rate": 5.718237497889926e-06,
+      "loss": 0.4547,
+      "step": 12742
+    },
+    {
+      "epoch": 2.0927678443126063,
+      "grad_norm": 0.4855079010515181,
+      "learning_rate": 5.717786031300054e-06,
+      "loss": 0.4249,
+      "step": 12743
+    },
+    {
+      "epoch": 2.0929320715209494,
+      "grad_norm": 0.38233922938650783,
+      "learning_rate": 5.717334551483825e-06,
+      "loss": 0.4391,
+      "step": 12744
+    },
+    {
+      "epoch": 2.093096298729292,
+      "grad_norm": 0.3406979729307253,
+      "learning_rate": 5.716883058446147e-06,
+      "loss": 0.4369,
+      "step": 12745
+    },
+    {
+      "epoch": 2.0932605259376347,
+      "grad_norm": 0.30944525643888987,
+      "learning_rate": 5.716431552191921e-06,
+      "loss": 0.4251,
+      "step": 12746
+    },
+    {
+      "epoch": 2.0934247531459773,
+      "grad_norm": 0.31735897928397955,
+      "learning_rate": 5.7159800327260525e-06,
+      "loss": 0.4425,
+      "step": 12747
+    },
+    {
+      "epoch": 2.0935889803543204,
+      "grad_norm": 0.5216526947925395,
+      "learning_rate": 5.715528500053444e-06,
+      "loss": 0.442,
+      "step": 12748
+    },
+    {
+      "epoch": 2.093753207562663,
+      "grad_norm": 0.47185098877306997,
+      "learning_rate": 5.715076954178999e-06,
+      "loss": 0.4323,
+      "step": 12749
+    },
+    {
+      "epoch": 2.0939174347710057,
+      "grad_norm": 0.4193173428927055,
+      "learning_rate": 5.714625395107623e-06,
+      "loss": 0.4698,
+      "step": 12750
+    },
+    {
+      "epoch": 2.0940816619793483,
+      "grad_norm": 0.6124497113365915,
+      "learning_rate": 5.714173822844221e-06,
+      "loss": 0.4431,
+      "step": 12751
+    },
+    {
+      "epoch": 2.0942458891876914,
+      "grad_norm": 0.293266143258921,
+      "learning_rate": 5.713722237393696e-06,
+      "loss": 0.4421,
+      "step": 12752
+    },
+    {
+      "epoch": 2.094410116396034,
+      "grad_norm": 0.6165608268152939,
+      "learning_rate": 5.713270638760955e-06,
+      "loss": 0.4347,
+      "step": 12753
+    },
+    {
+      "epoch": 2.0945743436043767,
+      "grad_norm": 0.3047436508071504,
+      "learning_rate": 5.7128190269508995e-06,
+      "loss": 0.457,
+      "step": 12754
+    },
+    {
+      "epoch": 2.0947385708127193,
+      "grad_norm": 0.3668284014045656,
+      "learning_rate": 5.712367401968436e-06,
+      "loss": 0.4328,
+      "step": 12755
+    },
+    {
+      "epoch": 2.094902798021062,
+      "grad_norm": 0.33986561289705924,
+      "learning_rate": 5.711915763818472e-06,
+      "loss": 0.4687,
+      "step": 12756
+    },
+    {
+      "epoch": 2.095067025229405,
+      "grad_norm": 0.3706800130778407,
+      "learning_rate": 5.711464112505909e-06,
+      "loss": 0.4489,
+      "step": 12757
+    },
+    {
+      "epoch": 2.0952312524377477,
+      "grad_norm": 0.39079281718805936,
+      "learning_rate": 5.711012448035652e-06,
+      "loss": 0.4633,
+      "step": 12758
+    },
+    {
+      "epoch": 2.0953954796460903,
+      "grad_norm": 0.32449038886318793,
+      "learning_rate": 5.710560770412611e-06,
+      "loss": 0.4216,
+      "step": 12759
+    },
+    {
+      "epoch": 2.095559706854433,
+      "grad_norm": 0.3466183057361125,
+      "learning_rate": 5.710109079641688e-06,
+      "loss": 0.4291,
+      "step": 12760
+    },
+    {
+      "epoch": 2.095723934062776,
+      "grad_norm": 0.3446100118332191,
+      "learning_rate": 5.70965737572779e-06,
+      "loss": 0.4605,
+      "step": 12761
+    },
+    {
+      "epoch": 2.0958881612711187,
+      "grad_norm": 0.3495938088505321,
+      "learning_rate": 5.7092056586758225e-06,
+      "loss": 0.4442,
+      "step": 12762
+    },
+    {
+      "epoch": 2.0960523884794613,
+      "grad_norm": 0.36872695332068844,
+      "learning_rate": 5.708753928490691e-06,
+      "loss": 0.459,
+      "step": 12763
+    },
+    {
+      "epoch": 2.096216615687804,
+      "grad_norm": 0.372112203602689,
+      "learning_rate": 5.708302185177304e-06,
+      "loss": 0.4351,
+      "step": 12764
+    },
+    {
+      "epoch": 2.096380842896147,
+      "grad_norm": 0.31392433890046545,
+      "learning_rate": 5.707850428740565e-06,
+      "loss": 0.4495,
+      "step": 12765
+    },
+    {
+      "epoch": 2.0965450701044896,
+      "grad_norm": 0.307439405862085,
+      "learning_rate": 5.707398659185383e-06,
+      "loss": 0.4327,
+      "step": 12766
+    },
+    {
+      "epoch": 2.0967092973128323,
+      "grad_norm": 0.41439290769866416,
+      "learning_rate": 5.706946876516664e-06,
+      "loss": 0.4503,
+      "step": 12767
+    },
+    {
+      "epoch": 2.096873524521175,
+      "grad_norm": 0.3074612197500893,
+      "learning_rate": 5.706495080739314e-06,
+      "loss": 0.4367,
+      "step": 12768
+    },
+    {
+      "epoch": 2.097037751729518,
+      "grad_norm": 0.5222822521040957,
+      "learning_rate": 5.706043271858241e-06,
+      "loss": 0.416,
+      "step": 12769
+    },
+    {
+      "epoch": 2.0972019789378606,
+      "grad_norm": 0.34805519039750665,
+      "learning_rate": 5.70559144987835e-06,
+      "loss": 0.4527,
+      "step": 12770
+    },
+    {
+      "epoch": 2.0973662061462033,
+      "grad_norm": 0.3251807292612414,
+      "learning_rate": 5.7051396148045514e-06,
+      "loss": 0.4297,
+      "step": 12771
+    },
+    {
+      "epoch": 2.097530433354546,
+      "grad_norm": 0.30016213451186896,
+      "learning_rate": 5.70468776664175e-06,
+      "loss": 0.4301,
+      "step": 12772
+    },
+    {
+      "epoch": 2.0976946605628886,
+      "grad_norm": 0.3500767429575023,
+      "learning_rate": 5.704235905394855e-06,
+      "loss": 0.4569,
+      "step": 12773
+    },
+    {
+      "epoch": 2.0978588877712316,
+      "grad_norm": 0.44306698843867504,
+      "learning_rate": 5.703784031068771e-06,
+      "loss": 0.4474,
+      "step": 12774
+    },
+    {
+      "epoch": 2.0980231149795743,
+      "grad_norm": 0.35216126579915613,
+      "learning_rate": 5.703332143668409e-06,
+      "loss": 0.4296,
+      "step": 12775
+    },
+    {
+      "epoch": 2.098187342187917,
+      "grad_norm": 0.3069250697540145,
+      "learning_rate": 5.702880243198678e-06,
+      "loss": 0.4239,
+      "step": 12776
+    },
+    {
+      "epoch": 2.0983515693962596,
+      "grad_norm": 0.3491742333738244,
+      "learning_rate": 5.7024283296644825e-06,
+      "loss": 0.4424,
+      "step": 12777
+    },
+    {
+      "epoch": 2.0985157966046026,
+      "grad_norm": 0.3532745513440168,
+      "learning_rate": 5.701976403070732e-06,
+      "loss": 0.4547,
+      "step": 12778
+    },
+    {
+      "epoch": 2.0986800238129453,
+      "grad_norm": 0.6853319161198199,
+      "learning_rate": 5.701524463422336e-06,
+      "loss": 0.442,
+      "step": 12779
+    },
+    {
+      "epoch": 2.098844251021288,
+      "grad_norm": 0.31164913944888184,
+      "learning_rate": 5.701072510724201e-06,
+      "loss": 0.4422,
+      "step": 12780
+    },
+    {
+      "epoch": 2.0990084782296305,
+      "grad_norm": 0.33442415700253836,
+      "learning_rate": 5.700620544981238e-06,
+      "loss": 0.4329,
+      "step": 12781
+    },
+    {
+      "epoch": 2.0991727054379736,
+      "grad_norm": 0.29691650447931556,
+      "learning_rate": 5.7001685661983545e-06,
+      "loss": 0.4464,
+      "step": 12782
+    },
+    {
+      "epoch": 2.0993369326463163,
+      "grad_norm": 0.4273272208437839,
+      "learning_rate": 5.699716574380459e-06,
+      "loss": 0.4387,
+      "step": 12783
+    },
+    {
+      "epoch": 2.099501159854659,
+      "grad_norm": 0.3664218978422905,
+      "learning_rate": 5.699264569532461e-06,
+      "loss": 0.4407,
+      "step": 12784
+    },
+    {
+      "epoch": 2.0996653870630015,
+      "grad_norm": 0.2972060184677094,
+      "learning_rate": 5.698812551659271e-06,
+      "loss": 0.431,
+      "step": 12785
+    },
+    {
+      "epoch": 2.0998296142713446,
+      "grad_norm": 0.3639133252374928,
+      "learning_rate": 5.698360520765798e-06,
+      "loss": 0.4439,
+      "step": 12786
+    },
+    {
+      "epoch": 2.0999938414796873,
+      "grad_norm": 0.30413518112204113,
+      "learning_rate": 5.697908476856948e-06,
+      "loss": 0.4441,
+      "step": 12787
+    },
+    {
+      "epoch": 2.10015806868803,
+      "grad_norm": 0.34623216367586335,
+      "learning_rate": 5.697456419937635e-06,
+      "loss": 0.4718,
+      "step": 12788
+    },
+    {
+      "epoch": 2.1003222958963725,
+      "grad_norm": 0.4154051288309667,
+      "learning_rate": 5.697004350012767e-06,
+      "loss": 0.4382,
+      "step": 12789
+    },
+    {
+      "epoch": 2.100486523104715,
+      "grad_norm": 0.33461676750805575,
+      "learning_rate": 5.696552267087253e-06,
+      "loss": 0.4585,
+      "step": 12790
+    },
+    {
+      "epoch": 2.1006507503130583,
+      "grad_norm": 0.4078581820722512,
+      "learning_rate": 5.696100171166006e-06,
+      "loss": 0.4295,
+      "step": 12791
+    },
+    {
+      "epoch": 2.100814977521401,
+      "grad_norm": 0.28825390150215113,
+      "learning_rate": 5.695648062253933e-06,
+      "loss": 0.4289,
+      "step": 12792
+    },
+    {
+      "epoch": 2.1009792047297435,
+      "grad_norm": 0.43816165937186574,
+      "learning_rate": 5.695195940355946e-06,
+      "loss": 0.4544,
+      "step": 12793
+    },
+    {
+      "epoch": 2.101143431938086,
+      "grad_norm": 0.5703413364880195,
+      "learning_rate": 5.694743805476955e-06,
+      "loss": 0.4567,
+      "step": 12794
+    },
+    {
+      "epoch": 2.1013076591464293,
+      "grad_norm": 0.35355716963073625,
+      "learning_rate": 5.69429165762187e-06,
+      "loss": 0.4656,
+      "step": 12795
+    },
+    {
+      "epoch": 2.101471886354772,
+      "grad_norm": 1.0919983683018712,
+      "learning_rate": 5.693839496795605e-06,
+      "loss": 0.4431,
+      "step": 12796
+    },
+    {
+      "epoch": 2.1016361135631145,
+      "grad_norm": 0.33052976505924586,
+      "learning_rate": 5.6933873230030665e-06,
+      "loss": 0.4374,
+      "step": 12797
+    },
+    {
+      "epoch": 2.101800340771457,
+      "grad_norm": 1.522094268884421,
+      "learning_rate": 5.692935136249169e-06,
+      "loss": 0.4613,
+      "step": 12798
+    },
+    {
+      "epoch": 2.1019645679798002,
+      "grad_norm": 0.35010018307879287,
+      "learning_rate": 5.692482936538821e-06,
+      "loss": 0.4227,
+      "step": 12799
+    },
+    {
+      "epoch": 2.102128795188143,
+      "grad_norm": 0.3341888208412005,
+      "learning_rate": 5.692030723876934e-06,
+      "loss": 0.4534,
+      "step": 12800
+    },
+    {
+      "epoch": 2.1022930223964855,
+      "grad_norm": 0.4399329112161016,
+      "learning_rate": 5.691578498268423e-06,
+      "loss": 0.4605,
+      "step": 12801
+    },
+    {
+      "epoch": 2.102457249604828,
+      "grad_norm": 0.4055705345962353,
+      "learning_rate": 5.691126259718197e-06,
+      "loss": 0.458,
+      "step": 12802
+    },
+    {
+      "epoch": 2.1026214768131712,
+      "grad_norm": 0.312902913896997,
+      "learning_rate": 5.690674008231166e-06,
+      "loss": 0.4495,
+      "step": 12803
+    },
+    {
+      "epoch": 2.102785704021514,
+      "grad_norm": 0.3318988567653643,
+      "learning_rate": 5.690221743812244e-06,
+      "loss": 0.4626,
+      "step": 12804
+    },
+    {
+      "epoch": 2.1029499312298565,
+      "grad_norm": 0.34101958993527215,
+      "learning_rate": 5.689769466466344e-06,
+      "loss": 0.4202,
+      "step": 12805
+    },
+    {
+      "epoch": 2.103114158438199,
+      "grad_norm": 0.30553517046334017,
+      "learning_rate": 5.689317176198377e-06,
+      "loss": 0.448,
+      "step": 12806
+    },
+    {
+      "epoch": 2.103278385646542,
+      "grad_norm": 0.29574474153044406,
+      "learning_rate": 5.688864873013256e-06,
+      "loss": 0.4267,
+      "step": 12807
+    },
+    {
+      "epoch": 2.103442612854885,
+      "grad_norm": 0.4147066715116571,
+      "learning_rate": 5.688412556915891e-06,
+      "loss": 0.4351,
+      "step": 12808
+    },
+    {
+      "epoch": 2.1036068400632275,
+      "grad_norm": 0.3180074558597533,
+      "learning_rate": 5.687960227911197e-06,
+      "loss": 0.434,
+      "step": 12809
+    },
+    {
+      "epoch": 2.10377106727157,
+      "grad_norm": 0.3715894218391581,
+      "learning_rate": 5.6875078860040854e-06,
+      "loss": 0.4319,
+      "step": 12810
+    },
+    {
+      "epoch": 2.103935294479913,
+      "grad_norm": 0.3526111909409907,
+      "learning_rate": 5.687055531199473e-06,
+      "loss": 0.4302,
+      "step": 12811
+    },
+    {
+      "epoch": 2.104099521688256,
+      "grad_norm": 0.3747430339445163,
+      "learning_rate": 5.6866031635022664e-06,
+      "loss": 0.4376,
+      "step": 12812
+    },
+    {
+      "epoch": 2.1042637488965985,
+      "grad_norm": 0.4540276918521262,
+      "learning_rate": 5.686150782917382e-06,
+      "loss": 0.4383,
+      "step": 12813
+    },
+    {
+      "epoch": 2.104427976104941,
+      "grad_norm": 0.3795862762748715,
+      "learning_rate": 5.685698389449735e-06,
+      "loss": 0.4659,
+      "step": 12814
+    },
+    {
+      "epoch": 2.104592203313284,
+      "grad_norm": 0.41673579474541966,
+      "learning_rate": 5.685245983104235e-06,
+      "loss": 0.4393,
+      "step": 12815
+    },
+    {
+      "epoch": 2.104756430521627,
+      "grad_norm": 0.3116492369443997,
+      "learning_rate": 5.684793563885799e-06,
+      "loss": 0.4574,
+      "step": 12816
+    },
+    {
+      "epoch": 2.1049206577299695,
+      "grad_norm": 0.3399815985478469,
+      "learning_rate": 5.684341131799338e-06,
+      "loss": 0.4369,
+      "step": 12817
+    },
+    {
+      "epoch": 2.105084884938312,
+      "grad_norm": 0.5281393777848307,
+      "learning_rate": 5.683888686849769e-06,
+      "loss": 0.4143,
+      "step": 12818
+    },
+    {
+      "epoch": 2.1052491121466548,
+      "grad_norm": 0.3429045769350071,
+      "learning_rate": 5.6834362290420015e-06,
+      "loss": 0.4436,
+      "step": 12819
+    },
+    {
+      "epoch": 2.105413339354998,
+      "grad_norm": 0.4486855426294437,
+      "learning_rate": 5.682983758380955e-06,
+      "loss": 0.438,
+      "step": 12820
+    },
+    {
+      "epoch": 2.1055775665633405,
+      "grad_norm": 0.29769448844466256,
+      "learning_rate": 5.682531274871538e-06,
+      "loss": 0.4387,
+      "step": 12821
+    },
+    {
+      "epoch": 2.105741793771683,
+      "grad_norm": 0.35663890431486234,
+      "learning_rate": 5.68207877851867e-06,
+      "loss": 0.4351,
+      "step": 12822
+    },
+    {
+      "epoch": 2.1059060209800258,
+      "grad_norm": 0.3455155676747153,
+      "learning_rate": 5.6816262693272625e-06,
+      "loss": 0.4459,
+      "step": 12823
+    },
+    {
+      "epoch": 2.1060702481883684,
+      "grad_norm": 0.3476365547950273,
+      "learning_rate": 5.681173747302231e-06,
+      "loss": 0.437,
+      "step": 12824
+    },
+    {
+      "epoch": 2.1062344753967115,
+      "grad_norm": 0.31649578705530995,
+      "learning_rate": 5.680721212448492e-06,
+      "loss": 0.4493,
+      "step": 12825
+    },
+    {
+      "epoch": 2.106398702605054,
+      "grad_norm": 0.3246652992220023,
+      "learning_rate": 5.680268664770957e-06,
+      "loss": 0.4527,
+      "step": 12826
+    },
+    {
+      "epoch": 2.1065629298133968,
+      "grad_norm": 0.3578087030598692,
+      "learning_rate": 5.679816104274546e-06,
+      "loss": 0.4339,
+      "step": 12827
+    },
+    {
+      "epoch": 2.1067271570217394,
+      "grad_norm": 0.4092237618364479,
+      "learning_rate": 5.679363530964167e-06,
+      "loss": 0.447,
+      "step": 12828
+    },
+    {
+      "epoch": 2.1068913842300825,
+      "grad_norm": 0.3361602929137724,
+      "learning_rate": 5.678910944844742e-06,
+      "loss": 0.4472,
+      "step": 12829
+    },
+    {
+      "epoch": 2.107055611438425,
+      "grad_norm": 0.38061197908341743,
+      "learning_rate": 5.6784583459211855e-06,
+      "loss": 0.437,
+      "step": 12830
+    },
+    {
+      "epoch": 2.1072198386467678,
+      "grad_norm": 0.33944842725725227,
+      "learning_rate": 5.678005734198412e-06,
+      "loss": 0.4142,
+      "step": 12831
+    },
+    {
+      "epoch": 2.1073840658551104,
+      "grad_norm": 0.27748281160012517,
+      "learning_rate": 5.677553109681335e-06,
+      "loss": 0.4437,
+      "step": 12832
+    },
+    {
+      "epoch": 2.1075482930634535,
+      "grad_norm": 0.493159606649079,
+      "learning_rate": 5.677100472374873e-06,
+      "loss": 0.4537,
+      "step": 12833
+    },
+    {
+      "epoch": 2.107712520271796,
+      "grad_norm": 0.33143414634234014,
+      "learning_rate": 5.676647822283942e-06,
+      "loss": 0.4473,
+      "step": 12834
+    },
+    {
+      "epoch": 2.1078767474801388,
+      "grad_norm": 0.33655801079061937,
+      "learning_rate": 5.67619515941346e-06,
+      "loss": 0.439,
+      "step": 12835
+    },
+    {
+      "epoch": 2.1080409746884814,
+      "grad_norm": 0.30323548739001926,
+      "learning_rate": 5.675742483768339e-06,
+      "loss": 0.4321,
+      "step": 12836
+    },
+    {
+      "epoch": 2.1082052018968245,
+      "grad_norm": 0.3308812641204237,
+      "learning_rate": 5.675289795353498e-06,
+      "loss": 0.4344,
+      "step": 12837
+    },
+    {
+      "epoch": 2.108369429105167,
+      "grad_norm": 0.3358438917250515,
+      "learning_rate": 5.674837094173854e-06,
+      "loss": 0.4495,
+      "step": 12838
+    },
+    {
+      "epoch": 2.1085336563135098,
+      "grad_norm": 0.4482835836763625,
+      "learning_rate": 5.674384380234323e-06,
+      "loss": 0.4488,
+      "step": 12839
+    },
+    {
+      "epoch": 2.1086978835218524,
+      "grad_norm": 0.29048868571265907,
+      "learning_rate": 5.673931653539824e-06,
+      "loss": 0.4486,
+      "step": 12840
+    },
+    {
+      "epoch": 2.108862110730195,
+      "grad_norm": 0.4323304845732,
+      "learning_rate": 5.6734789140952695e-06,
+      "loss": 0.4321,
+      "step": 12841
+    },
+    {
+      "epoch": 2.109026337938538,
+      "grad_norm": 0.30474831684058123,
+      "learning_rate": 5.673026161905581e-06,
+      "loss": 0.4372,
+      "step": 12842
+    },
+    {
+      "epoch": 2.1091905651468807,
+      "grad_norm": 0.38957900370384485,
+      "learning_rate": 5.672573396975674e-06,
+      "loss": 0.4331,
+      "step": 12843
+    },
+    {
+      "epoch": 2.1093547923552234,
+      "grad_norm": 0.33247541581000345,
+      "learning_rate": 5.672120619310466e-06,
+      "loss": 0.4318,
+      "step": 12844
+    },
+    {
+      "epoch": 2.109519019563566,
+      "grad_norm": 0.366253574432419,
+      "learning_rate": 5.671667828914876e-06,
+      "loss": 0.4543,
+      "step": 12845
+    },
+    {
+      "epoch": 2.109683246771909,
+      "grad_norm": 0.28846496806996375,
+      "learning_rate": 5.6712150257938196e-06,
+      "loss": 0.4418,
+      "step": 12846
+    },
+    {
+      "epoch": 2.1098474739802517,
+      "grad_norm": 0.37577112889588415,
+      "learning_rate": 5.670762209952215e-06,
+      "loss": 0.4423,
+      "step": 12847
+    },
+    {
+      "epoch": 2.1100117011885944,
+      "grad_norm": 0.327766941899745,
+      "learning_rate": 5.670309381394982e-06,
+      "loss": 0.4452,
+      "step": 12848
+    },
+    {
+      "epoch": 2.110175928396937,
+      "grad_norm": 0.3545541182125981,
+      "learning_rate": 5.669856540127037e-06,
+      "loss": 0.4251,
+      "step": 12849
+    },
+    {
+      "epoch": 2.11034015560528,
+      "grad_norm": 0.3924250514834875,
+      "learning_rate": 5.6694036861533e-06,
+      "loss": 0.4482,
+      "step": 12850
+    },
+    {
+      "epoch": 2.1105043828136227,
+      "grad_norm": 0.3583855348922854,
+      "learning_rate": 5.668950819478688e-06,
+      "loss": 0.4352,
+      "step": 12851
+    },
+    {
+      "epoch": 2.1106686100219654,
+      "grad_norm": 0.3968481426345676,
+      "learning_rate": 5.66849794010812e-06,
+      "loss": 0.4337,
+      "step": 12852
+    },
+    {
+      "epoch": 2.110832837230308,
+      "grad_norm": 0.3224990826352743,
+      "learning_rate": 5.6680450480465145e-06,
+      "loss": 0.4443,
+      "step": 12853
+    },
+    {
+      "epoch": 2.110997064438651,
+      "grad_norm": 0.39144802472570384,
+      "learning_rate": 5.667592143298791e-06,
+      "loss": 0.4289,
+      "step": 12854
+    },
+    {
+      "epoch": 2.1111612916469937,
+      "grad_norm": 0.3718451290345085,
+      "learning_rate": 5.667139225869867e-06,
+      "loss": 0.4506,
+      "step": 12855
+    },
+    {
+      "epoch": 2.1113255188553364,
+      "grad_norm": 0.3297370105246588,
+      "learning_rate": 5.666686295764665e-06,
+      "loss": 0.4506,
+      "step": 12856
+    },
+    {
+      "epoch": 2.111489746063679,
+      "grad_norm": 0.5862997461169424,
+      "learning_rate": 5.6662333529880994e-06,
+      "loss": 0.4494,
+      "step": 12857
+    },
+    {
+      "epoch": 2.1116539732720216,
+      "grad_norm": 0.42111008173194764,
+      "learning_rate": 5.665780397545093e-06,
+      "loss": 0.4436,
+      "step": 12858
+    },
+    {
+      "epoch": 2.1118182004803647,
+      "grad_norm": 0.33023195184528636,
+      "learning_rate": 5.665327429440566e-06,
+      "loss": 0.4374,
+      "step": 12859
+    },
+    {
+      "epoch": 2.1119824276887074,
+      "grad_norm": 0.39047191719718477,
+      "learning_rate": 5.664874448679434e-06,
+      "loss": 0.4404,
+      "step": 12860
+    },
+    {
+      "epoch": 2.11214665489705,
+      "grad_norm": 0.36364259425271644,
+      "learning_rate": 5.6644214552666205e-06,
+      "loss": 0.4453,
+      "step": 12861
+    },
+    {
+      "epoch": 2.1123108821053926,
+      "grad_norm": 0.47280712222169236,
+      "learning_rate": 5.663968449207044e-06,
+      "loss": 0.442,
+      "step": 12862
+    },
+    {
+      "epoch": 2.1124751093137357,
+      "grad_norm": 0.34497358185281746,
+      "learning_rate": 5.663515430505626e-06,
+      "loss": 0.4632,
+      "step": 12863
+    },
+    {
+      "epoch": 2.1126393365220784,
+      "grad_norm": 0.35565401681845893,
+      "learning_rate": 5.663062399167285e-06,
+      "loss": 0.4572,
+      "step": 12864
+    },
+    {
+      "epoch": 2.112803563730421,
+      "grad_norm": 0.3225578882958658,
+      "learning_rate": 5.662609355196944e-06,
+      "loss": 0.4578,
+      "step": 12865
+    },
+    {
+      "epoch": 2.1129677909387636,
+      "grad_norm": 0.3207304288053452,
+      "learning_rate": 5.662156298599518e-06,
+      "loss": 0.4319,
+      "step": 12866
+    },
+    {
+      "epoch": 2.1131320181471067,
+      "grad_norm": 0.6186662136471093,
+      "learning_rate": 5.661703229379933e-06,
+      "loss": 0.4241,
+      "step": 12867
+    },
+    {
+      "epoch": 2.1132962453554494,
+      "grad_norm": 0.3004416227688589,
+      "learning_rate": 5.661250147543107e-06,
+      "loss": 0.4552,
+      "step": 12868
+    },
+    {
+      "epoch": 2.113460472563792,
+      "grad_norm": 0.3337040866259995,
+      "learning_rate": 5.660797053093965e-06,
+      "loss": 0.4431,
+      "step": 12869
+    },
+    {
+      "epoch": 2.1136246997721346,
+      "grad_norm": 0.30755631851279613,
+      "learning_rate": 5.6603439460374226e-06,
+      "loss": 0.4349,
+      "step": 12870
+    },
+    {
+      "epoch": 2.1137889269804777,
+      "grad_norm": 0.35471867913214444,
+      "learning_rate": 5.659890826378403e-06,
+      "loss": 0.4443,
+      "step": 12871
+    },
+    {
+      "epoch": 2.1139531541888203,
+      "grad_norm": 0.29714877631996933,
+      "learning_rate": 5.659437694121827e-06,
+      "loss": 0.4423,
+      "step": 12872
+    },
+    {
+      "epoch": 2.114117381397163,
+      "grad_norm": 0.2906786915109368,
+      "learning_rate": 5.658984549272619e-06,
+      "loss": 0.4479,
+      "step": 12873
+    },
+    {
+      "epoch": 2.1142816086055056,
+      "grad_norm": 0.32187424124789155,
+      "learning_rate": 5.658531391835699e-06,
+      "loss": 0.4357,
+      "step": 12874
+    },
+    {
+      "epoch": 2.1144458358138483,
+      "grad_norm": 0.29924613569598135,
+      "learning_rate": 5.658078221815986e-06,
+      "loss": 0.4509,
+      "step": 12875
+    },
+    {
+      "epoch": 2.1146100630221913,
+      "grad_norm": 0.30270006577595426,
+      "learning_rate": 5.657625039218405e-06,
+      "loss": 0.4335,
+      "step": 12876
+    },
+    {
+      "epoch": 2.114774290230534,
+      "grad_norm": 0.3369789030042965,
+      "learning_rate": 5.6571718440478774e-06,
+      "loss": 0.4334,
+      "step": 12877
+    },
+    {
+      "epoch": 2.1149385174388766,
+      "grad_norm": 0.6159556214342785,
+      "learning_rate": 5.656718636309324e-06,
+      "loss": 0.4382,
+      "step": 12878
+    },
+    {
+      "epoch": 2.1151027446472193,
+      "grad_norm": 0.3931772247618362,
+      "learning_rate": 5.65626541600767e-06,
+      "loss": 0.4486,
+      "step": 12879
+    },
+    {
+      "epoch": 2.1152669718555623,
+      "grad_norm": 0.3023065547371581,
+      "learning_rate": 5.655812183147834e-06,
+      "loss": 0.4478,
+      "step": 12880
+    },
+    {
+      "epoch": 2.115431199063905,
+      "grad_norm": 0.46112021793977825,
+      "learning_rate": 5.655358937734742e-06,
+      "loss": 0.4535,
+      "step": 12881
+    },
+    {
+      "epoch": 2.1155954262722476,
+      "grad_norm": 0.31824756429454715,
+      "learning_rate": 5.654905679773315e-06,
+      "loss": 0.4336,
+      "step": 12882
+    },
+    {
+      "epoch": 2.1157596534805903,
+      "grad_norm": 0.45968576013510043,
+      "learning_rate": 5.654452409268476e-06,
+      "loss": 0.4484,
+      "step": 12883
+    },
+    {
+      "epoch": 2.1159238806889333,
+      "grad_norm": 0.32054414238138046,
+      "learning_rate": 5.653999126225148e-06,
+      "loss": 0.4297,
+      "step": 12884
+    },
+    {
+      "epoch": 2.116088107897276,
+      "grad_norm": 1.3758552124140686,
+      "learning_rate": 5.653545830648254e-06,
+      "loss": 0.4522,
+      "step": 12885
+    },
+    {
+      "epoch": 2.1162523351056186,
+      "grad_norm": 0.3020933416951688,
+      "learning_rate": 5.653092522542717e-06,
+      "loss": 0.4419,
+      "step": 12886
+    },
+    {
+      "epoch": 2.1164165623139612,
+      "grad_norm": 0.29168531190257574,
+      "learning_rate": 5.652639201913461e-06,
+      "loss": 0.4563,
+      "step": 12887
+    },
+    {
+      "epoch": 2.1165807895223043,
+      "grad_norm": 0.39239511980870456,
+      "learning_rate": 5.652185868765409e-06,
+      "loss": 0.4533,
+      "step": 12888
+    },
+    {
+      "epoch": 2.116745016730647,
+      "grad_norm": 0.328826136046041,
+      "learning_rate": 5.651732523103485e-06,
+      "loss": 0.4404,
+      "step": 12889
+    },
+    {
+      "epoch": 2.1169092439389896,
+      "grad_norm": 0.3431931638495928,
+      "learning_rate": 5.6512791649326136e-06,
+      "loss": 0.4611,
+      "step": 12890
+    },
+    {
+      "epoch": 2.1170734711473322,
+      "grad_norm": 0.28812679712227424,
+      "learning_rate": 5.650825794257716e-06,
+      "loss": 0.438,
+      "step": 12891
+    },
+    {
+      "epoch": 2.117237698355675,
+      "grad_norm": 0.2902608751383587,
+      "learning_rate": 5.650372411083718e-06,
+      "loss": 0.4342,
+      "step": 12892
+    },
+    {
+      "epoch": 2.117401925564018,
+      "grad_norm": 0.37581087709899136,
+      "learning_rate": 5.649919015415546e-06,
+      "loss": 0.4447,
+      "step": 12893
+    },
+    {
+      "epoch": 2.1175661527723606,
+      "grad_norm": 0.32938488910246017,
+      "learning_rate": 5.649465607258122e-06,
+      "loss": 0.4396,
+      "step": 12894
+    },
+    {
+      "epoch": 2.1177303799807032,
+      "grad_norm": 0.4706196191524123,
+      "learning_rate": 5.649012186616368e-06,
+      "loss": 0.4348,
+      "step": 12895
+    },
+    {
+      "epoch": 2.117894607189046,
+      "grad_norm": 0.41697900587103304,
+      "learning_rate": 5.648558753495212e-06,
+      "loss": 0.4366,
+      "step": 12896
+    },
+    {
+      "epoch": 2.118058834397389,
+      "grad_norm": 0.31704559733383314,
+      "learning_rate": 5.648105307899579e-06,
+      "loss": 0.4481,
+      "step": 12897
+    },
+    {
+      "epoch": 2.1182230616057316,
+      "grad_norm": 0.47313061714381116,
+      "learning_rate": 5.647651849834392e-06,
+      "loss": 0.4421,
+      "step": 12898
+    },
+    {
+      "epoch": 2.1183872888140742,
+      "grad_norm": 0.3885098414913026,
+      "learning_rate": 5.647198379304578e-06,
+      "loss": 0.4371,
+      "step": 12899
+    },
+    {
+      "epoch": 2.118551516022417,
+      "grad_norm": 0.35465476613932184,
+      "learning_rate": 5.646744896315059e-06,
+      "loss": 0.453,
+      "step": 12900
+    },
+    {
+      "epoch": 2.11871574323076,
+      "grad_norm": 0.33812434114988693,
+      "learning_rate": 5.646291400870763e-06,
+      "loss": 0.4476,
+      "step": 12901
+    },
+    {
+      "epoch": 2.1188799704391026,
+      "grad_norm": 0.2907497876750063,
+      "learning_rate": 5.645837892976615e-06,
+      "loss": 0.4291,
+      "step": 12902
+    },
+    {
+      "epoch": 2.1190441976474452,
+      "grad_norm": 0.2562377614658291,
+      "learning_rate": 5.6453843726375395e-06,
+      "loss": 0.4528,
+      "step": 12903
+    },
+    {
+      "epoch": 2.119208424855788,
+      "grad_norm": 0.4183063196730311,
+      "learning_rate": 5.644930839858463e-06,
+      "loss": 0.4566,
+      "step": 12904
+    },
+    {
+      "epoch": 2.119372652064131,
+      "grad_norm": 0.3360083856619884,
+      "learning_rate": 5.644477294644312e-06,
+      "loss": 0.4612,
+      "step": 12905
+    },
+    {
+      "epoch": 2.1195368792724736,
+      "grad_norm": 0.3550177847074318,
+      "learning_rate": 5.644023737000011e-06,
+      "loss": 0.413,
+      "step": 12906
+    },
+    {
+      "epoch": 2.119701106480816,
+      "grad_norm": 0.3168948460799242,
+      "learning_rate": 5.643570166930485e-06,
+      "loss": 0.4267,
+      "step": 12907
+    },
+    {
+      "epoch": 2.119865333689159,
+      "grad_norm": 0.41299478080841584,
+      "learning_rate": 5.643116584440665e-06,
+      "loss": 0.4389,
+      "step": 12908
+    },
+    {
+      "epoch": 2.1200295608975015,
+      "grad_norm": 0.285914803678043,
+      "learning_rate": 5.642662989535472e-06,
+      "loss": 0.4204,
+      "step": 12909
+    },
+    {
+      "epoch": 2.1201937881058446,
+      "grad_norm": 0.3225596692159442,
+      "learning_rate": 5.642209382219836e-06,
+      "loss": 0.4421,
+      "step": 12910
+    },
+    {
+      "epoch": 2.120358015314187,
+      "grad_norm": 0.3499417079033318,
+      "learning_rate": 5.6417557624986815e-06,
+      "loss": 0.4402,
+      "step": 12911
+    },
+    {
+      "epoch": 2.12052224252253,
+      "grad_norm": 0.29267360443334167,
+      "learning_rate": 5.641302130376935e-06,
+      "loss": 0.4348,
+      "step": 12912
+    },
+    {
+      "epoch": 2.1206864697308725,
+      "grad_norm": 0.32426656914720464,
+      "learning_rate": 5.640848485859526e-06,
+      "loss": 0.4419,
+      "step": 12913
+    },
+    {
+      "epoch": 2.1208506969392156,
+      "grad_norm": 0.3281027482639434,
+      "learning_rate": 5.6403948289513795e-06,
+      "loss": 0.4561,
+      "step": 12914
+    },
+    {
+      "epoch": 2.121014924147558,
+      "grad_norm": 0.3220048109109444,
+      "learning_rate": 5.6399411596574245e-06,
+      "loss": 0.4575,
+      "step": 12915
+    },
+    {
+      "epoch": 2.121179151355901,
+      "grad_norm": 0.5050677501725908,
+      "learning_rate": 5.639487477982585e-06,
+      "loss": 0.4443,
+      "step": 12916
+    },
+    {
+      "epoch": 2.1213433785642435,
+      "grad_norm": 0.33091551603894176,
+      "learning_rate": 5.639033783931792e-06,
+      "loss": 0.438,
+      "step": 12917
+    },
+    {
+      "epoch": 2.1215076057725866,
+      "grad_norm": 0.5370997172267648,
+      "learning_rate": 5.63858007750997e-06,
+      "loss": 0.442,
+      "step": 12918
+    },
+    {
+      "epoch": 2.121671832980929,
+      "grad_norm": 0.3195961149692968,
+      "learning_rate": 5.638126358722049e-06,
+      "loss": 0.4289,
+      "step": 12919
+    },
+    {
+      "epoch": 2.121836060189272,
+      "grad_norm": 0.29324147405114565,
+      "learning_rate": 5.637672627572955e-06,
+      "loss": 0.4474,
+      "step": 12920
+    },
+    {
+      "epoch": 2.1220002873976145,
+      "grad_norm": 0.2773735461045715,
+      "learning_rate": 5.637218884067618e-06,
+      "loss": 0.4613,
+      "step": 12921
+    },
+    {
+      "epoch": 2.1221645146059576,
+      "grad_norm": 0.3540915443199416,
+      "learning_rate": 5.636765128210965e-06,
+      "loss": 0.4309,
+      "step": 12922
+    },
+    {
+      "epoch": 2.1223287418143,
+      "grad_norm": 0.3887095467543331,
+      "learning_rate": 5.636311360007924e-06,
+      "loss": 0.4282,
+      "step": 12923
+    },
+    {
+      "epoch": 2.122492969022643,
+      "grad_norm": 0.3299693991544904,
+      "learning_rate": 5.635857579463423e-06,
+      "loss": 0.4328,
+      "step": 12924
+    },
+    {
+      "epoch": 2.1226571962309855,
+      "grad_norm": 0.37227913141766533,
+      "learning_rate": 5.635403786582392e-06,
+      "loss": 0.4406,
+      "step": 12925
+    },
+    {
+      "epoch": 2.122821423439328,
+      "grad_norm": 0.3635459484822922,
+      "learning_rate": 5.634949981369758e-06,
+      "loss": 0.4633,
+      "step": 12926
+    },
+    {
+      "epoch": 2.122985650647671,
+      "grad_norm": 0.34839992097364925,
+      "learning_rate": 5.634496163830452e-06,
+      "loss": 0.4488,
+      "step": 12927
+    },
+    {
+      "epoch": 2.123149877856014,
+      "grad_norm": 0.333284343465869,
+      "learning_rate": 5.634042333969401e-06,
+      "loss": 0.4568,
+      "step": 12928
+    },
+    {
+      "epoch": 2.1233141050643565,
+      "grad_norm": 0.3805554693548058,
+      "learning_rate": 5.633588491791533e-06,
+      "loss": 0.4514,
+      "step": 12929
+    },
+    {
+      "epoch": 2.123478332272699,
+      "grad_norm": 0.33782011930290967,
+      "learning_rate": 5.63313463730178e-06,
+      "loss": 0.4413,
+      "step": 12930
+    },
+    {
+      "epoch": 2.123642559481042,
+      "grad_norm": 0.3804235495657651,
+      "learning_rate": 5.63268077050507e-06,
+      "loss": 0.4547,
+      "step": 12931
+    },
+    {
+      "epoch": 2.123806786689385,
+      "grad_norm": 0.34878407931263294,
+      "learning_rate": 5.632226891406332e-06,
+      "loss": 0.4447,
+      "step": 12932
+    },
+    {
+      "epoch": 2.1239710138977275,
+      "grad_norm": 0.2980431569749951,
+      "learning_rate": 5.631773000010497e-06,
+      "loss": 0.4501,
+      "step": 12933
+    },
+    {
+      "epoch": 2.12413524110607,
+      "grad_norm": 0.2758512424001063,
+      "learning_rate": 5.631319096322493e-06,
+      "loss": 0.4375,
+      "step": 12934
+    },
+    {
+      "epoch": 2.124299468314413,
+      "grad_norm": 0.31581503124840343,
+      "learning_rate": 5.6308651803472505e-06,
+      "loss": 0.4398,
+      "step": 12935
+    },
+    {
+      "epoch": 2.124463695522756,
+      "grad_norm": 0.4350563326476109,
+      "learning_rate": 5.630411252089699e-06,
+      "loss": 0.4385,
+      "step": 12936
+    },
+    {
+      "epoch": 2.1246279227310985,
+      "grad_norm": 0.32043068856204204,
+      "learning_rate": 5.62995731155477e-06,
+      "loss": 0.4418,
+      "step": 12937
+    },
+    {
+      "epoch": 2.124792149939441,
+      "grad_norm": 0.3878698223491185,
+      "learning_rate": 5.629503358747392e-06,
+      "loss": 0.4548,
+      "step": 12938
+    },
+    {
+      "epoch": 2.124956377147784,
+      "grad_norm": 0.395039731182331,
+      "learning_rate": 5.6290493936724965e-06,
+      "loss": 0.4387,
+      "step": 12939
+    },
+    {
+      "epoch": 2.125120604356127,
+      "grad_norm": 0.3315098767767503,
+      "learning_rate": 5.628595416335014e-06,
+      "loss": 0.4409,
+      "step": 12940
+    },
+    {
+      "epoch": 2.1252848315644695,
+      "grad_norm": 0.3856410406992659,
+      "learning_rate": 5.628141426739875e-06,
+      "loss": 0.4578,
+      "step": 12941
+    },
+    {
+      "epoch": 2.125449058772812,
+      "grad_norm": 0.2973728241039122,
+      "learning_rate": 5.627687424892011e-06,
+      "loss": 0.4412,
+      "step": 12942
+    },
+    {
+      "epoch": 2.1256132859811547,
+      "grad_norm": 0.7612320769649644,
+      "learning_rate": 5.62723341079635e-06,
+      "loss": 0.4356,
+      "step": 12943
+    },
+    {
+      "epoch": 2.125777513189498,
+      "grad_norm": 0.34729546541020695,
+      "learning_rate": 5.626779384457826e-06,
+      "loss": 0.4264,
+      "step": 12944
+    },
+    {
+      "epoch": 2.1259417403978405,
+      "grad_norm": 0.3043427111694188,
+      "learning_rate": 5.6263253458813706e-06,
+      "loss": 0.4497,
+      "step": 12945
+    },
+    {
+      "epoch": 2.126105967606183,
+      "grad_norm": 0.32606508097107895,
+      "learning_rate": 5.625871295071912e-06,
+      "loss": 0.4416,
+      "step": 12946
+    },
+    {
+      "epoch": 2.1262701948145257,
+      "grad_norm": 0.4127021304590552,
+      "learning_rate": 5.625417232034384e-06,
+      "loss": 0.45,
+      "step": 12947
+    },
+    {
+      "epoch": 2.126434422022869,
+      "grad_norm": 0.41019044971934865,
+      "learning_rate": 5.624963156773718e-06,
+      "loss": 0.4496,
+      "step": 12948
+    },
+    {
+      "epoch": 2.1265986492312114,
+      "grad_norm": 0.31596254214135494,
+      "learning_rate": 5.624509069294845e-06,
+      "loss": 0.4331,
+      "step": 12949
+    },
+    {
+      "epoch": 2.126762876439554,
+      "grad_norm": 0.5525985814660054,
+      "learning_rate": 5.6240549696026975e-06,
+      "loss": 0.4287,
+      "step": 12950
+    },
+    {
+      "epoch": 2.1269271036478967,
+      "grad_norm": 0.34763719641785923,
+      "learning_rate": 5.623600857702207e-06,
+      "loss": 0.4355,
+      "step": 12951
+    },
+    {
+      "epoch": 2.12709133085624,
+      "grad_norm": 0.3335035146638821,
+      "learning_rate": 5.6231467335983055e-06,
+      "loss": 0.4353,
+      "step": 12952
+    },
+    {
+      "epoch": 2.1272555580645824,
+      "grad_norm": 0.3217134828394556,
+      "learning_rate": 5.622692597295925e-06,
+      "loss": 0.4277,
+      "step": 12953
+    },
+    {
+      "epoch": 2.127419785272925,
+      "grad_norm": 0.3820768103292228,
+      "learning_rate": 5.622238448799999e-06,
+      "loss": 0.4487,
+      "step": 12954
+    },
+    {
+      "epoch": 2.1275840124812677,
+      "grad_norm": 0.4124972463916858,
+      "learning_rate": 5.621784288115459e-06,
+      "loss": 0.4457,
+      "step": 12955
+    },
+    {
+      "epoch": 2.127748239689611,
+      "grad_norm": 0.41009248794214526,
+      "learning_rate": 5.621330115247238e-06,
+      "loss": 0.4431,
+      "step": 12956
+    },
+    {
+      "epoch": 2.1279124668979534,
+      "grad_norm": 0.324419514380881,
+      "learning_rate": 5.620875930200269e-06,
+      "loss": 0.4171,
+      "step": 12957
+    },
+    {
+      "epoch": 2.128076694106296,
+      "grad_norm": 0.44635639840875285,
+      "learning_rate": 5.620421732979484e-06,
+      "loss": 0.4335,
+      "step": 12958
+    },
+    {
+      "epoch": 2.1282409213146387,
+      "grad_norm": 0.3072615539433608,
+      "learning_rate": 5.619967523589817e-06,
+      "loss": 0.427,
+      "step": 12959
+    },
+    {
+      "epoch": 2.1284051485229813,
+      "grad_norm": 0.4381451304638284,
+      "learning_rate": 5.619513302036201e-06,
+      "loss": 0.4552,
+      "step": 12960
+    },
+    {
+      "epoch": 2.1285693757313244,
+      "grad_norm": 0.6381434192285261,
+      "learning_rate": 5.6190590683235686e-06,
+      "loss": 0.4299,
+      "step": 12961
+    },
+    {
+      "epoch": 2.128733602939667,
+      "grad_norm": 0.45459376598759865,
+      "learning_rate": 5.618604822456854e-06,
+      "loss": 0.4414,
+      "step": 12962
+    },
+    {
+      "epoch": 2.1288978301480097,
+      "grad_norm": 0.49530451943785425,
+      "learning_rate": 5.6181505644409904e-06,
+      "loss": 0.462,
+      "step": 12963
+    },
+    {
+      "epoch": 2.1290620573563523,
+      "grad_norm": 0.3506050848447432,
+      "learning_rate": 5.617696294280911e-06,
+      "loss": 0.4587,
+      "step": 12964
+    },
+    {
+      "epoch": 2.1292262845646954,
+      "grad_norm": 0.3191382725265907,
+      "learning_rate": 5.617242011981551e-06,
+      "loss": 0.4386,
+      "step": 12965
+    },
+    {
+      "epoch": 2.129390511773038,
+      "grad_norm": 0.3174610775036302,
+      "learning_rate": 5.616787717547844e-06,
+      "loss": 0.4326,
+      "step": 12966
+    },
+    {
+      "epoch": 2.1295547389813807,
+      "grad_norm": 0.537093039177515,
+      "learning_rate": 5.616333410984723e-06,
+      "loss": 0.4605,
+      "step": 12967
+    },
+    {
+      "epoch": 2.1297189661897233,
+      "grad_norm": 0.31637109473600866,
+      "learning_rate": 5.615879092297121e-06,
+      "loss": 0.4557,
+      "step": 12968
+    },
+    {
+      "epoch": 2.1298831933980664,
+      "grad_norm": 0.3918704384575548,
+      "learning_rate": 5.615424761489978e-06,
+      "loss": 0.4314,
+      "step": 12969
+    },
+    {
+      "epoch": 2.130047420606409,
+      "grad_norm": 0.35561557929773774,
+      "learning_rate": 5.6149704185682215e-06,
+      "loss": 0.4345,
+      "step": 12970
+    },
+    {
+      "epoch": 2.1302116478147517,
+      "grad_norm": 0.4669581383106852,
+      "learning_rate": 5.614516063536791e-06,
+      "loss": 0.4602,
+      "step": 12971
+    },
+    {
+      "epoch": 2.1303758750230943,
+      "grad_norm": 0.398092381638841,
+      "learning_rate": 5.614061696400619e-06,
+      "loss": 0.4342,
+      "step": 12972
+    },
+    {
+      "epoch": 2.1305401022314374,
+      "grad_norm": 0.5613327187613648,
+      "learning_rate": 5.6136073171646404e-06,
+      "loss": 0.4566,
+      "step": 12973
+    },
+    {
+      "epoch": 2.13070432943978,
+      "grad_norm": 0.34100163249184556,
+      "learning_rate": 5.6131529258337906e-06,
+      "loss": 0.444,
+      "step": 12974
+    },
+    {
+      "epoch": 2.1308685566481227,
+      "grad_norm": 0.4985210923911517,
+      "learning_rate": 5.612698522413005e-06,
+      "loss": 0.4402,
+      "step": 12975
+    },
+    {
+      "epoch": 2.1310327838564653,
+      "grad_norm": 0.3421448979356347,
+      "learning_rate": 5.61224410690722e-06,
+      "loss": 0.4419,
+      "step": 12976
+    },
+    {
+      "epoch": 2.131197011064808,
+      "grad_norm": 0.34251735534570665,
+      "learning_rate": 5.611789679321369e-06,
+      "loss": 0.4548,
+      "step": 12977
+    },
+    {
+      "epoch": 2.131361238273151,
+      "grad_norm": 0.3535445694441817,
+      "learning_rate": 5.611335239660387e-06,
+      "loss": 0.4449,
+      "step": 12978
+    },
+    {
+      "epoch": 2.1315254654814937,
+      "grad_norm": 0.32032005716195605,
+      "learning_rate": 5.610880787929211e-06,
+      "loss": 0.454,
+      "step": 12979
+    },
+    {
+      "epoch": 2.1316896926898363,
+      "grad_norm": 0.32004352432492167,
+      "learning_rate": 5.610426324132778e-06,
+      "loss": 0.4466,
+      "step": 12980
+    },
+    {
+      "epoch": 2.131853919898179,
+      "grad_norm": 0.29014953643068214,
+      "learning_rate": 5.6099718482760235e-06,
+      "loss": 0.4714,
+      "step": 12981
+    },
+    {
+      "epoch": 2.132018147106522,
+      "grad_norm": 0.34390516599209603,
+      "learning_rate": 5.609517360363881e-06,
+      "loss": 0.4304,
+      "step": 12982
+    },
+    {
+      "epoch": 2.1321823743148647,
+      "grad_norm": 0.3213270673623128,
+      "learning_rate": 5.6090628604012875e-06,
+      "loss": 0.4484,
+      "step": 12983
+    },
+    {
+      "epoch": 2.1323466015232073,
+      "grad_norm": 0.3058927295010632,
+      "learning_rate": 5.608608348393181e-06,
+      "loss": 0.4505,
+      "step": 12984
+    },
+    {
+      "epoch": 2.13251082873155,
+      "grad_norm": 0.4679978012777797,
+      "learning_rate": 5.608153824344498e-06,
+      "loss": 0.4449,
+      "step": 12985
+    },
+    {
+      "epoch": 2.132675055939893,
+      "grad_norm": 0.3197622088020438,
+      "learning_rate": 5.607699288260174e-06,
+      "loss": 0.4454,
+      "step": 12986
+    },
+    {
+      "epoch": 2.1328392831482357,
+      "grad_norm": 0.27578517030301564,
+      "learning_rate": 5.607244740145145e-06,
+      "loss": 0.4387,
+      "step": 12987
+    },
+    {
+      "epoch": 2.1330035103565783,
+      "grad_norm": 0.395659876071687,
+      "learning_rate": 5.606790180004349e-06,
+      "loss": 0.4415,
+      "step": 12988
+    },
+    {
+      "epoch": 2.133167737564921,
+      "grad_norm": 0.3075443442701608,
+      "learning_rate": 5.6063356078427225e-06,
+      "loss": 0.4536,
+      "step": 12989
+    },
+    {
+      "epoch": 2.133331964773264,
+      "grad_norm": 0.41313177332432993,
+      "learning_rate": 5.605881023665203e-06,
+      "loss": 0.4289,
+      "step": 12990
+    },
+    {
+      "epoch": 2.1334961919816067,
+      "grad_norm": 0.3571092549071176,
+      "learning_rate": 5.605426427476729e-06,
+      "loss": 0.4664,
+      "step": 12991
+    },
+    {
+      "epoch": 2.1336604191899493,
+      "grad_norm": 0.39056888922382665,
+      "learning_rate": 5.604971819282235e-06,
+      "loss": 0.4258,
+      "step": 12992
+    },
+    {
+      "epoch": 2.133824646398292,
+      "grad_norm": 0.46537769479849345,
+      "learning_rate": 5.60451719908666e-06,
+      "loss": 0.449,
+      "step": 12993
+    },
+    {
+      "epoch": 2.1339888736066346,
+      "grad_norm": 0.32186461409923345,
+      "learning_rate": 5.604062566894941e-06,
+      "loss": 0.4595,
+      "step": 12994
+    },
+    {
+      "epoch": 2.1341531008149777,
+      "grad_norm": 0.31168989365248334,
+      "learning_rate": 5.603607922712017e-06,
+      "loss": 0.4561,
+      "step": 12995
+    },
+    {
+      "epoch": 2.1343173280233203,
+      "grad_norm": 0.31585319111399035,
+      "learning_rate": 5.603153266542826e-06,
+      "loss": 0.431,
+      "step": 12996
+    },
+    {
+      "epoch": 2.134481555231663,
+      "grad_norm": 0.30053858074909556,
+      "learning_rate": 5.602698598392304e-06,
+      "loss": 0.4168,
+      "step": 12997
+    },
+    {
+      "epoch": 2.1346457824400056,
+      "grad_norm": 0.36234676368672447,
+      "learning_rate": 5.602243918265391e-06,
+      "loss": 0.4358,
+      "step": 12998
+    },
+    {
+      "epoch": 2.1348100096483487,
+      "grad_norm": 0.33428480744979727,
+      "learning_rate": 5.601789226167023e-06,
+      "loss": 0.4185,
+      "step": 12999
+    },
+    {
+      "epoch": 2.1349742368566913,
+      "grad_norm": 0.3244302536241494,
+      "learning_rate": 5.601334522102142e-06,
+      "loss": 0.4707,
+      "step": 13000
+    },
+    {
+      "epoch": 2.135138464065034,
+      "grad_norm": 0.3502757192894509,
+      "learning_rate": 5.600879806075683e-06,
+      "loss": 0.4303,
+      "step": 13001
+    },
+    {
+      "epoch": 2.1353026912733766,
+      "grad_norm": 0.3685548422888394,
+      "learning_rate": 5.600425078092588e-06,
+      "loss": 0.436,
+      "step": 13002
+    },
+    {
+      "epoch": 2.1354669184817197,
+      "grad_norm": 0.2929653807287389,
+      "learning_rate": 5.599970338157792e-06,
+      "loss": 0.4328,
+      "step": 13003
+    },
+    {
+      "epoch": 2.1356311456900623,
+      "grad_norm": 0.28914285602883866,
+      "learning_rate": 5.599515586276236e-06,
+      "loss": 0.4346,
+      "step": 13004
+    },
+    {
+      "epoch": 2.135795372898405,
+      "grad_norm": 0.3748436051569583,
+      "learning_rate": 5.59906082245286e-06,
+      "loss": 0.4497,
+      "step": 13005
+    },
+    {
+      "epoch": 2.1359596001067476,
+      "grad_norm": 0.4301403569066467,
+      "learning_rate": 5.598606046692603e-06,
+      "loss": 0.449,
+      "step": 13006
+    },
+    {
+      "epoch": 2.1361238273150907,
+      "grad_norm": 0.2750459024303177,
+      "learning_rate": 5.598151259000401e-06,
+      "loss": 0.437,
+      "step": 13007
+    },
+    {
+      "epoch": 2.1362880545234333,
+      "grad_norm": 0.32575515399845445,
+      "learning_rate": 5.597696459381197e-06,
+      "loss": 0.4617,
+      "step": 13008
+    },
+    {
+      "epoch": 2.136452281731776,
+      "grad_norm": 0.2996017078795059,
+      "learning_rate": 5.597241647839928e-06,
+      "loss": 0.4336,
+      "step": 13009
+    },
+    {
+      "epoch": 2.1366165089401186,
+      "grad_norm": 0.3353749243950319,
+      "learning_rate": 5.596786824381538e-06,
+      "loss": 0.4374,
+      "step": 13010
+    },
+    {
+      "epoch": 2.136780736148461,
+      "grad_norm": 0.3015536802983604,
+      "learning_rate": 5.596331989010964e-06,
+      "loss": 0.4537,
+      "step": 13011
+    },
+    {
+      "epoch": 2.1369449633568043,
+      "grad_norm": 0.7102526684523022,
+      "learning_rate": 5.595877141733144e-06,
+      "loss": 0.4264,
+      "step": 13012
+    },
+    {
+      "epoch": 2.137109190565147,
+      "grad_norm": 1.0802129528539435,
+      "learning_rate": 5.595422282553021e-06,
+      "loss": 0.4365,
+      "step": 13013
+    },
+    {
+      "epoch": 2.1372734177734896,
+      "grad_norm": 0.3765904066273181,
+      "learning_rate": 5.594967411475532e-06,
+      "loss": 0.4365,
+      "step": 13014
+    },
+    {
+      "epoch": 2.137437644981832,
+      "grad_norm": 0.3296267997367895,
+      "learning_rate": 5.594512528505624e-06,
+      "loss": 0.452,
+      "step": 13015
+    },
+    {
+      "epoch": 2.1376018721901753,
+      "grad_norm": 0.32635813808393965,
+      "learning_rate": 5.59405763364823e-06,
+      "loss": 0.4335,
+      "step": 13016
+    },
+    {
+      "epoch": 2.137766099398518,
+      "grad_norm": 0.3054992350198616,
+      "learning_rate": 5.593602726908295e-06,
+      "loss": 0.441,
+      "step": 13017
+    },
+    {
+      "epoch": 2.1379303266068606,
+      "grad_norm": 0.3160707553495685,
+      "learning_rate": 5.593147808290756e-06,
+      "loss": 0.4384,
+      "step": 13018
+    },
+    {
+      "epoch": 2.138094553815203,
+      "grad_norm": 0.28928587834988345,
+      "learning_rate": 5.592692877800559e-06,
+      "loss": 0.4279,
+      "step": 13019
+    },
+    {
+      "epoch": 2.1382587810235463,
+      "grad_norm": 0.401047698366232,
+      "learning_rate": 5.592237935442642e-06,
+      "loss": 0.4183,
+      "step": 13020
+    },
+    {
+      "epoch": 2.138423008231889,
+      "grad_norm": 0.3045454687637298,
+      "learning_rate": 5.591782981221946e-06,
+      "loss": 0.4626,
+      "step": 13021
+    },
+    {
+      "epoch": 2.1385872354402315,
+      "grad_norm": 0.2681046998791144,
+      "learning_rate": 5.591328015143411e-06,
+      "loss": 0.44,
+      "step": 13022
+    },
+    {
+      "epoch": 2.138751462648574,
+      "grad_norm": 0.4218919748405687,
+      "learning_rate": 5.590873037211982e-06,
+      "loss": 0.4417,
+      "step": 13023
+    },
+    {
+      "epoch": 2.1389156898569173,
+      "grad_norm": 0.2793908490461997,
+      "learning_rate": 5.590418047432597e-06,
+      "loss": 0.4348,
+      "step": 13024
+    },
+    {
+      "epoch": 2.13907991706526,
+      "grad_norm": 0.3172246138905424,
+      "learning_rate": 5.589963045810202e-06,
+      "loss": 0.446,
+      "step": 13025
+    },
+    {
+      "epoch": 2.1392441442736025,
+      "grad_norm": 0.312775867484758,
+      "learning_rate": 5.589508032349734e-06,
+      "loss": 0.4463,
+      "step": 13026
+    },
+    {
+      "epoch": 2.139408371481945,
+      "grad_norm": 0.3568040949452443,
+      "learning_rate": 5.589053007056136e-06,
+      "loss": 0.4454,
+      "step": 13027
+    },
+    {
+      "epoch": 2.139572598690288,
+      "grad_norm": 0.3577210247324337,
+      "learning_rate": 5.588597969934353e-06,
+      "loss": 0.4448,
+      "step": 13028
+    },
+    {
+      "epoch": 2.139736825898631,
+      "grad_norm": 0.33210112769047656,
+      "learning_rate": 5.588142920989323e-06,
+      "loss": 0.443,
+      "step": 13029
+    },
+    {
+      "epoch": 2.1399010531069735,
+      "grad_norm": 0.3524122026419944,
+      "learning_rate": 5.587687860225991e-06,
+      "loss": 0.4571,
+      "step": 13030
+    },
+    {
+      "epoch": 2.140065280315316,
+      "grad_norm": 0.5054906330538123,
+      "learning_rate": 5.5872327876493e-06,
+      "loss": 0.4429,
+      "step": 13031
+    },
+    {
+      "epoch": 2.140229507523659,
+      "grad_norm": 0.32656193120293153,
+      "learning_rate": 5.58677770326419e-06,
+      "loss": 0.4489,
+      "step": 13032
+    },
+    {
+      "epoch": 2.140393734732002,
+      "grad_norm": 0.3670053673951288,
+      "learning_rate": 5.586322607075604e-06,
+      "loss": 0.4258,
+      "step": 13033
+    },
+    {
+      "epoch": 2.1405579619403445,
+      "grad_norm": 0.3701069054773225,
+      "learning_rate": 5.585867499088488e-06,
+      "loss": 0.4419,
+      "step": 13034
+    },
+    {
+      "epoch": 2.140722189148687,
+      "grad_norm": 0.34285488173179895,
+      "learning_rate": 5.5854123793077805e-06,
+      "loss": 0.4371,
+      "step": 13035
+    },
+    {
+      "epoch": 2.14088641635703,
+      "grad_norm": 0.2893085117521016,
+      "learning_rate": 5.5849572477384276e-06,
+      "loss": 0.4504,
+      "step": 13036
+    },
+    {
+      "epoch": 2.141050643565373,
+      "grad_norm": 0.4135270238157861,
+      "learning_rate": 5.584502104385371e-06,
+      "loss": 0.4536,
+      "step": 13037
+    },
+    {
+      "epoch": 2.1412148707737155,
+      "grad_norm": 0.9039359209112584,
+      "learning_rate": 5.584046949253554e-06,
+      "loss": 0.4541,
+      "step": 13038
+    },
+    {
+      "epoch": 2.141379097982058,
+      "grad_norm": 0.4604628288870824,
+      "learning_rate": 5.583591782347923e-06,
+      "loss": 0.4627,
+      "step": 13039
+    },
+    {
+      "epoch": 2.141543325190401,
+      "grad_norm": 0.361287982585956,
+      "learning_rate": 5.583136603673417e-06,
+      "loss": 0.4319,
+      "step": 13040
+    },
+    {
+      "epoch": 2.141707552398744,
+      "grad_norm": 0.28686311031123324,
+      "learning_rate": 5.582681413234982e-06,
+      "loss": 0.4349,
+      "step": 13041
+    },
+    {
+      "epoch": 2.1418717796070865,
+      "grad_norm": 0.35422337043200885,
+      "learning_rate": 5.582226211037562e-06,
+      "loss": 0.4418,
+      "step": 13042
+    },
+    {
+      "epoch": 2.142036006815429,
+      "grad_norm": 0.3200420784217416,
+      "learning_rate": 5.5817709970861e-06,
+      "loss": 0.4364,
+      "step": 13043
+    },
+    {
+      "epoch": 2.142200234023772,
+      "grad_norm": 0.35940330185959896,
+      "learning_rate": 5.581315771385542e-06,
+      "loss": 0.4302,
+      "step": 13044
+    },
+    {
+      "epoch": 2.1423644612321144,
+      "grad_norm": 0.288818054045889,
+      "learning_rate": 5.580860533940831e-06,
+      "loss": 0.4481,
+      "step": 13045
+    },
+    {
+      "epoch": 2.1425286884404575,
+      "grad_norm": 0.34310898496235703,
+      "learning_rate": 5.5804052847569096e-06,
+      "loss": 0.4445,
+      "step": 13046
+    },
+    {
+      "epoch": 2.1426929156488,
+      "grad_norm": 0.3507973417433449,
+      "learning_rate": 5.579950023838725e-06,
+      "loss": 0.4413,
+      "step": 13047
+    },
+    {
+      "epoch": 2.142857142857143,
+      "grad_norm": 0.40522847529414024,
+      "learning_rate": 5.57949475119122e-06,
+      "loss": 0.4522,
+      "step": 13048
+    },
+    {
+      "epoch": 2.1430213700654854,
+      "grad_norm": 0.3351266446909605,
+      "learning_rate": 5.579039466819341e-06,
+      "loss": 0.4465,
+      "step": 13049
+    },
+    {
+      "epoch": 2.1431855972738285,
+      "grad_norm": 0.3245795390439118,
+      "learning_rate": 5.578584170728031e-06,
+      "loss": 0.4415,
+      "step": 13050
+    },
+    {
+      "epoch": 2.143349824482171,
+      "grad_norm": 0.2817301746950823,
+      "learning_rate": 5.578128862922235e-06,
+      "loss": 0.443,
+      "step": 13051
+    },
+    {
+      "epoch": 2.143514051690514,
+      "grad_norm": 0.3772109800765201,
+      "learning_rate": 5.5776735434069e-06,
+      "loss": 0.452,
+      "step": 13052
+    },
+    {
+      "epoch": 2.1436782788988564,
+      "grad_norm": 0.41514420047183903,
+      "learning_rate": 5.577218212186968e-06,
+      "loss": 0.4514,
+      "step": 13053
+    },
+    {
+      "epoch": 2.1438425061071995,
+      "grad_norm": 0.34853574492762074,
+      "learning_rate": 5.576762869267388e-06,
+      "loss": 0.4421,
+      "step": 13054
+    },
+    {
+      "epoch": 2.144006733315542,
+      "grad_norm": 0.37626637398815654,
+      "learning_rate": 5.576307514653103e-06,
+      "loss": 0.4602,
+      "step": 13055
+    },
+    {
+      "epoch": 2.144170960523885,
+      "grad_norm": 0.3908924960128689,
+      "learning_rate": 5.5758521483490605e-06,
+      "loss": 0.4251,
+      "step": 13056
+    },
+    {
+      "epoch": 2.1443351877322274,
+      "grad_norm": 0.3037037346436557,
+      "learning_rate": 5.575396770360205e-06,
+      "loss": 0.4376,
+      "step": 13057
+    },
+    {
+      "epoch": 2.1444994149405705,
+      "grad_norm": 0.39166416556086264,
+      "learning_rate": 5.5749413806914825e-06,
+      "loss": 0.4551,
+      "step": 13058
+    },
+    {
+      "epoch": 2.144663642148913,
+      "grad_norm": 0.2758285721291899,
+      "learning_rate": 5.57448597934784e-06,
+      "loss": 0.4538,
+      "step": 13059
+    },
+    {
+      "epoch": 2.144827869357256,
+      "grad_norm": 0.30518189424114334,
+      "learning_rate": 5.57403056633422e-06,
+      "loss": 0.4432,
+      "step": 13060
+    },
+    {
+      "epoch": 2.1449920965655984,
+      "grad_norm": 0.3021189985917047,
+      "learning_rate": 5.573575141655574e-06,
+      "loss": 0.4256,
+      "step": 13061
+    },
+    {
+      "epoch": 2.145156323773941,
+      "grad_norm": 0.34734420320468,
+      "learning_rate": 5.573119705316844e-06,
+      "loss": 0.4478,
+      "step": 13062
+    },
+    {
+      "epoch": 2.145320550982284,
+      "grad_norm": 0.48479298084851663,
+      "learning_rate": 5.572664257322978e-06,
+      "loss": 0.4456,
+      "step": 13063
+    },
+    {
+      "epoch": 2.1454847781906268,
+      "grad_norm": 0.9737469501992383,
+      "learning_rate": 5.572208797678923e-06,
+      "loss": 0.4519,
+      "step": 13064
+    },
+    {
+      "epoch": 2.1456490053989694,
+      "grad_norm": 0.373845968709997,
+      "learning_rate": 5.571753326389628e-06,
+      "loss": 0.4225,
+      "step": 13065
+    },
+    {
+      "epoch": 2.145813232607312,
+      "grad_norm": 0.32952687545014325,
+      "learning_rate": 5.571297843460035e-06,
+      "loss": 0.4551,
+      "step": 13066
+    },
+    {
+      "epoch": 2.145977459815655,
+      "grad_norm": 0.28185179231913293,
+      "learning_rate": 5.570842348895093e-06,
+      "loss": 0.4522,
+      "step": 13067
+    },
+    {
+      "epoch": 2.1461416870239978,
+      "grad_norm": 0.3750226506331507,
+      "learning_rate": 5.570386842699751e-06,
+      "loss": 0.4388,
+      "step": 13068
+    },
+    {
+      "epoch": 2.1463059142323404,
+      "grad_norm": 0.363792534235299,
+      "learning_rate": 5.569931324878955e-06,
+      "loss": 0.4566,
+      "step": 13069
+    },
+    {
+      "epoch": 2.146470141440683,
+      "grad_norm": 0.4173978857711169,
+      "learning_rate": 5.56947579543765e-06,
+      "loss": 0.4456,
+      "step": 13070
+    },
+    {
+      "epoch": 2.146634368649026,
+      "grad_norm": 0.3617383410012619,
+      "learning_rate": 5.5690202543807866e-06,
+      "loss": 0.4536,
+      "step": 13071
+    },
+    {
+      "epoch": 2.1467985958573688,
+      "grad_norm": 0.31683670414164145,
+      "learning_rate": 5.568564701713312e-06,
+      "loss": 0.4339,
+      "step": 13072
+    },
+    {
+      "epoch": 2.1469628230657114,
+      "grad_norm": 0.6722978519914278,
+      "learning_rate": 5.568109137440174e-06,
+      "loss": 0.4453,
+      "step": 13073
+    },
+    {
+      "epoch": 2.147127050274054,
+      "grad_norm": 0.31359460500682335,
+      "learning_rate": 5.567653561566319e-06,
+      "loss": 0.4406,
+      "step": 13074
+    },
+    {
+      "epoch": 2.147291277482397,
+      "grad_norm": 0.3756857719735049,
+      "learning_rate": 5.567197974096695e-06,
+      "loss": 0.4512,
+      "step": 13075
+    },
+    {
+      "epoch": 2.1474555046907398,
+      "grad_norm": 0.28206587584500525,
+      "learning_rate": 5.566742375036252e-06,
+      "loss": 0.4431,
+      "step": 13076
+    },
+    {
+      "epoch": 2.1476197318990824,
+      "grad_norm": 0.3034882846676163,
+      "learning_rate": 5.566286764389937e-06,
+      "loss": 0.4387,
+      "step": 13077
+    },
+    {
+      "epoch": 2.147783959107425,
+      "grad_norm": 0.33707031344715366,
+      "learning_rate": 5.5658311421627e-06,
+      "loss": 0.4486,
+      "step": 13078
+    },
+    {
+      "epoch": 2.1479481863157677,
+      "grad_norm": 0.43035245643608405,
+      "learning_rate": 5.5653755083594865e-06,
+      "loss": 0.4515,
+      "step": 13079
+    },
+    {
+      "epoch": 2.1481124135241108,
+      "grad_norm": 0.3180318636784757,
+      "learning_rate": 5.564919862985248e-06,
+      "loss": 0.4323,
+      "step": 13080
+    },
+    {
+      "epoch": 2.1482766407324534,
+      "grad_norm": 0.7311184003595356,
+      "learning_rate": 5.564464206044931e-06,
+      "loss": 0.462,
+      "step": 13081
+    },
+    {
+      "epoch": 2.148440867940796,
+      "grad_norm": 0.3498635077663437,
+      "learning_rate": 5.5640085375434855e-06,
+      "loss": 0.4308,
+      "step": 13082
+    },
+    {
+      "epoch": 2.1486050951491387,
+      "grad_norm": 0.9829592808567664,
+      "learning_rate": 5.5635528574858614e-06,
+      "loss": 0.4572,
+      "step": 13083
+    },
+    {
+      "epoch": 2.1487693223574817,
+      "grad_norm": 0.3950510374950475,
+      "learning_rate": 5.563097165877006e-06,
+      "loss": 0.4555,
+      "step": 13084
+    },
+    {
+      "epoch": 2.1489335495658244,
+      "grad_norm": 1.013249951682912,
+      "learning_rate": 5.562641462721869e-06,
+      "loss": 0.4385,
+      "step": 13085
+    },
+    {
+      "epoch": 2.149097776774167,
+      "grad_norm": 0.3005075649363099,
+      "learning_rate": 5.562185748025402e-06,
+      "loss": 0.4302,
+      "step": 13086
+    },
+    {
+      "epoch": 2.1492620039825097,
+      "grad_norm": 0.7128528813599668,
+      "learning_rate": 5.561730021792551e-06,
+      "loss": 0.4462,
+      "step": 13087
+    },
+    {
+      "epoch": 2.1494262311908527,
+      "grad_norm": 0.34082831138362135,
+      "learning_rate": 5.561274284028269e-06,
+      "loss": 0.4103,
+      "step": 13088
+    },
+    {
+      "epoch": 2.1495904583991954,
+      "grad_norm": 0.27912374758491476,
+      "learning_rate": 5.560818534737502e-06,
+      "loss": 0.4734,
+      "step": 13089
+    },
+    {
+      "epoch": 2.149754685607538,
+      "grad_norm": 0.41419567612684344,
+      "learning_rate": 5.560362773925204e-06,
+      "loss": 0.4371,
+      "step": 13090
+    },
+    {
+      "epoch": 2.1499189128158807,
+      "grad_norm": 0.533304143338659,
+      "learning_rate": 5.559907001596322e-06,
+      "loss": 0.4488,
+      "step": 13091
+    },
+    {
+      "epoch": 2.1500831400242237,
+      "grad_norm": 0.3287216931610325,
+      "learning_rate": 5.559451217755807e-06,
+      "loss": 0.423,
+      "step": 13092
+    },
+    {
+      "epoch": 2.1502473672325664,
+      "grad_norm": 0.2987444710279113,
+      "learning_rate": 5.55899542240861e-06,
+      "loss": 0.4284,
+      "step": 13093
+    },
+    {
+      "epoch": 2.150411594440909,
+      "grad_norm": 0.3915364216911103,
+      "learning_rate": 5.558539615559681e-06,
+      "loss": 0.435,
+      "step": 13094
+    },
+    {
+      "epoch": 2.1505758216492517,
+      "grad_norm": 0.34923159544476823,
+      "learning_rate": 5.55808379721397e-06,
+      "loss": 0.4442,
+      "step": 13095
+    },
+    {
+      "epoch": 2.1507400488575943,
+      "grad_norm": 0.3603631249694652,
+      "learning_rate": 5.557627967376427e-06,
+      "loss": 0.4224,
+      "step": 13096
+    },
+    {
+      "epoch": 2.1509042760659374,
+      "grad_norm": 0.47614439422632243,
+      "learning_rate": 5.557172126052005e-06,
+      "loss": 0.4412,
+      "step": 13097
+    },
+    {
+      "epoch": 2.15106850327428,
+      "grad_norm": 0.3048139485131453,
+      "learning_rate": 5.556716273245654e-06,
+      "loss": 0.4444,
+      "step": 13098
+    },
+    {
+      "epoch": 2.1512327304826226,
+      "grad_norm": 0.3342919021364949,
+      "learning_rate": 5.556260408962323e-06,
+      "loss": 0.4726,
+      "step": 13099
+    },
+    {
+      "epoch": 2.1513969576909653,
+      "grad_norm": 0.29465856880907676,
+      "learning_rate": 5.5558045332069645e-06,
+      "loss": 0.4508,
+      "step": 13100
+    },
+    {
+      "epoch": 2.1515611848993084,
+      "grad_norm": 0.2961227755518122,
+      "learning_rate": 5.555348645984531e-06,
+      "loss": 0.4494,
+      "step": 13101
+    },
+    {
+      "epoch": 2.151725412107651,
+      "grad_norm": 0.27934800568638346,
+      "learning_rate": 5.554892747299973e-06,
+      "loss": 0.4391,
+      "step": 13102
+    },
+    {
+      "epoch": 2.1518896393159936,
+      "grad_norm": 0.33111901066285837,
+      "learning_rate": 5.554436837158242e-06,
+      "loss": 0.4348,
+      "step": 13103
+    },
+    {
+      "epoch": 2.1520538665243363,
+      "grad_norm": 0.3684615911145497,
+      "learning_rate": 5.553980915564289e-06,
+      "loss": 0.4521,
+      "step": 13104
+    },
+    {
+      "epoch": 2.1522180937326794,
+      "grad_norm": 0.4292233032330315,
+      "learning_rate": 5.553524982523065e-06,
+      "loss": 0.4528,
+      "step": 13105
+    },
+    {
+      "epoch": 2.152382320941022,
+      "grad_norm": 0.43574829372998947,
+      "learning_rate": 5.553069038039525e-06,
+      "loss": 0.4537,
+      "step": 13106
+    },
+    {
+      "epoch": 2.1525465481493646,
+      "grad_norm": 0.294165085934928,
+      "learning_rate": 5.55261308211862e-06,
+      "loss": 0.4509,
+      "step": 13107
+    },
+    {
+      "epoch": 2.1527107753577073,
+      "grad_norm": 0.31222154571936955,
+      "learning_rate": 5.5521571147653e-06,
+      "loss": 0.4509,
+      "step": 13108
+    },
+    {
+      "epoch": 2.1528750025660504,
+      "grad_norm": 0.27790905680021066,
+      "learning_rate": 5.551701135984519e-06,
+      "loss": 0.4542,
+      "step": 13109
+    },
+    {
+      "epoch": 2.153039229774393,
+      "grad_norm": 0.4239281111220191,
+      "learning_rate": 5.551245145781228e-06,
+      "loss": 0.4555,
+      "step": 13110
+    },
+    {
+      "epoch": 2.1532034569827356,
+      "grad_norm": 0.33405406926044195,
+      "learning_rate": 5.550789144160381e-06,
+      "loss": 0.4255,
+      "step": 13111
+    },
+    {
+      "epoch": 2.1533676841910783,
+      "grad_norm": 0.3045577111345963,
+      "learning_rate": 5.550333131126931e-06,
+      "loss": 0.4337,
+      "step": 13112
+    },
+    {
+      "epoch": 2.153531911399421,
+      "grad_norm": 0.3667448605274694,
+      "learning_rate": 5.549877106685829e-06,
+      "loss": 0.4398,
+      "step": 13113
+    },
+    {
+      "epoch": 2.153696138607764,
+      "grad_norm": 0.3587313899783721,
+      "learning_rate": 5.549421070842028e-06,
+      "loss": 0.4337,
+      "step": 13114
+    },
+    {
+      "epoch": 2.1538603658161066,
+      "grad_norm": 0.31611137498570135,
+      "learning_rate": 5.548965023600482e-06,
+      "loss": 0.43,
+      "step": 13115
+    },
+    {
+      "epoch": 2.1540245930244493,
+      "grad_norm": 0.41761280354962393,
+      "learning_rate": 5.548508964966144e-06,
+      "loss": 0.4389,
+      "step": 13116
+    },
+    {
+      "epoch": 2.154188820232792,
+      "grad_norm": 0.3276668465800213,
+      "learning_rate": 5.548052894943968e-06,
+      "loss": 0.4386,
+      "step": 13117
+    },
+    {
+      "epoch": 2.154353047441135,
+      "grad_norm": 0.3058632740019484,
+      "learning_rate": 5.547596813538905e-06,
+      "loss": 0.4275,
+      "step": 13118
+    },
+    {
+      "epoch": 2.1545172746494776,
+      "grad_norm": 0.3134859102629066,
+      "learning_rate": 5.547140720755911e-06,
+      "loss": 0.4368,
+      "step": 13119
+    },
+    {
+      "epoch": 2.1546815018578203,
+      "grad_norm": 1.9136277979976022,
+      "learning_rate": 5.546684616599937e-06,
+      "loss": 0.4645,
+      "step": 13120
+    },
+    {
+      "epoch": 2.154845729066163,
+      "grad_norm": 0.3081164038829003,
+      "learning_rate": 5.5462285010759385e-06,
+      "loss": 0.4404,
+      "step": 13121
+    },
+    {
+      "epoch": 2.155009956274506,
+      "grad_norm": 0.7771526801527265,
+      "learning_rate": 5.545772374188871e-06,
+      "loss": 0.4397,
+      "step": 13122
+    },
+    {
+      "epoch": 2.1551741834828486,
+      "grad_norm": 0.2715873990185136,
+      "learning_rate": 5.545316235943686e-06,
+      "loss": 0.4275,
+      "step": 13123
+    },
+    {
+      "epoch": 2.1553384106911913,
+      "grad_norm": 0.33280980643281594,
+      "learning_rate": 5.544860086345337e-06,
+      "loss": 0.4512,
+      "step": 13124
+    },
+    {
+      "epoch": 2.155502637899534,
+      "grad_norm": 0.3851568549443537,
+      "learning_rate": 5.54440392539878e-06,
+      "loss": 0.4319,
+      "step": 13125
+    },
+    {
+      "epoch": 2.155666865107877,
+      "grad_norm": 0.3063502288189881,
+      "learning_rate": 5.5439477531089685e-06,
+      "loss": 0.4368,
+      "step": 13126
+    },
+    {
+      "epoch": 2.1558310923162196,
+      "grad_norm": 0.32417940549514135,
+      "learning_rate": 5.543491569480859e-06,
+      "loss": 0.4548,
+      "step": 13127
+    },
+    {
+      "epoch": 2.1559953195245622,
+      "grad_norm": 0.2819287851853112,
+      "learning_rate": 5.543035374519403e-06,
+      "loss": 0.4314,
+      "step": 13128
+    },
+    {
+      "epoch": 2.156159546732905,
+      "grad_norm": 0.3343017480828349,
+      "learning_rate": 5.542579168229557e-06,
+      "loss": 0.4413,
+      "step": 13129
+    },
+    {
+      "epoch": 2.1563237739412475,
+      "grad_norm": 0.30548747472069426,
+      "learning_rate": 5.542122950616274e-06,
+      "loss": 0.443,
+      "step": 13130
+    },
+    {
+      "epoch": 2.1564880011495906,
+      "grad_norm": 0.27993439186124225,
+      "learning_rate": 5.5416667216845124e-06,
+      "loss": 0.4386,
+      "step": 13131
+    },
+    {
+      "epoch": 2.1566522283579332,
+      "grad_norm": 0.32577795398226056,
+      "learning_rate": 5.541210481439225e-06,
+      "loss": 0.4435,
+      "step": 13132
+    },
+    {
+      "epoch": 2.156816455566276,
+      "grad_norm": 0.32823780085029325,
+      "learning_rate": 5.540754229885367e-06,
+      "loss": 0.4507,
+      "step": 13133
+    },
+    {
+      "epoch": 2.1569806827746185,
+      "grad_norm": 0.3047427313061152,
+      "learning_rate": 5.5402979670278946e-06,
+      "loss": 0.43,
+      "step": 13134
+    },
+    {
+      "epoch": 2.1571449099829616,
+      "grad_norm": 0.340813084378149,
+      "learning_rate": 5.539841692871761e-06,
+      "loss": 0.4279,
+      "step": 13135
+    },
+    {
+      "epoch": 2.1573091371913042,
+      "grad_norm": 0.46663307065196113,
+      "learning_rate": 5.539385407421925e-06,
+      "loss": 0.4634,
+      "step": 13136
+    },
+    {
+      "epoch": 2.157473364399647,
+      "grad_norm": 0.37566933247851336,
+      "learning_rate": 5.538929110683342e-06,
+      "loss": 0.4472,
+      "step": 13137
+    },
+    {
+      "epoch": 2.1576375916079895,
+      "grad_norm": 0.30868472477797043,
+      "learning_rate": 5.538472802660965e-06,
+      "loss": 0.4586,
+      "step": 13138
+    },
+    {
+      "epoch": 2.1578018188163326,
+      "grad_norm": 0.34066658358138063,
+      "learning_rate": 5.538016483359751e-06,
+      "loss": 0.4448,
+      "step": 13139
+    },
+    {
+      "epoch": 2.1579660460246752,
+      "grad_norm": 0.4073911081151271,
+      "learning_rate": 5.537560152784659e-06,
+      "loss": 0.4563,
+      "step": 13140
+    },
+    {
+      "epoch": 2.158130273233018,
+      "grad_norm": 0.29822675903780166,
+      "learning_rate": 5.537103810940641e-06,
+      "loss": 0.4621,
+      "step": 13141
+    },
+    {
+      "epoch": 2.1582945004413605,
+      "grad_norm": 0.31971502442009747,
+      "learning_rate": 5.536647457832656e-06,
+      "loss": 0.4553,
+      "step": 13142
+    },
+    {
+      "epoch": 2.1584587276497036,
+      "grad_norm": 0.3426810765747357,
+      "learning_rate": 5.53619109346566e-06,
+      "loss": 0.4477,
+      "step": 13143
+    },
+    {
+      "epoch": 2.1586229548580462,
+      "grad_norm": 0.945124020682834,
+      "learning_rate": 5.5357347178446086e-06,
+      "loss": 0.4475,
+      "step": 13144
+    },
+    {
+      "epoch": 2.158787182066389,
+      "grad_norm": 0.4918830844122075,
+      "learning_rate": 5.535278330974459e-06,
+      "loss": 0.4576,
+      "step": 13145
+    },
+    {
+      "epoch": 2.1589514092747315,
+      "grad_norm": 0.3058756489796773,
+      "learning_rate": 5.534821932860169e-06,
+      "loss": 0.4337,
+      "step": 13146
+    },
+    {
+      "epoch": 2.159115636483074,
+      "grad_norm": 0.3120707299763528,
+      "learning_rate": 5.534365523506694e-06,
+      "loss": 0.4735,
+      "step": 13147
+    },
+    {
+      "epoch": 2.1592798636914172,
+      "grad_norm": 0.2846250730751281,
+      "learning_rate": 5.5339091029189925e-06,
+      "loss": 0.437,
+      "step": 13148
+    },
+    {
+      "epoch": 2.15944409089976,
+      "grad_norm": 0.3201064753442044,
+      "learning_rate": 5.53345267110202e-06,
+      "loss": 0.469,
+      "step": 13149
+    },
+    {
+      "epoch": 2.1596083181081025,
+      "grad_norm": 0.29766138661669544,
+      "learning_rate": 5.532996228060735e-06,
+      "loss": 0.429,
+      "step": 13150
+    },
+    {
+      "epoch": 2.159772545316445,
+      "grad_norm": 0.2846032186777436,
+      "learning_rate": 5.532539773800095e-06,
+      "loss": 0.4401,
+      "step": 13151
+    },
+    {
+      "epoch": 2.159936772524788,
+      "grad_norm": 0.3129947601427393,
+      "learning_rate": 5.5320833083250565e-06,
+      "loss": 0.446,
+      "step": 13152
+    },
+    {
+      "epoch": 2.160100999733131,
+      "grad_norm": 0.3144839696313304,
+      "learning_rate": 5.531626831640578e-06,
+      "loss": 0.452,
+      "step": 13153
+    },
+    {
+      "epoch": 2.1602652269414735,
+      "grad_norm": 0.2959377033395708,
+      "learning_rate": 5.531170343751617e-06,
+      "loss": 0.4373,
+      "step": 13154
+    },
+    {
+      "epoch": 2.160429454149816,
+      "grad_norm": 0.40203476934860716,
+      "learning_rate": 5.530713844663132e-06,
+      "loss": 0.4332,
+      "step": 13155
+    },
+    {
+      "epoch": 2.160593681358159,
+      "grad_norm": 0.34219461068539536,
+      "learning_rate": 5.530257334380081e-06,
+      "loss": 0.4348,
+      "step": 13156
+    },
+    {
+      "epoch": 2.160757908566502,
+      "grad_norm": 0.33771282218080734,
+      "learning_rate": 5.529800812907421e-06,
+      "loss": 0.4639,
+      "step": 13157
+    },
+    {
+      "epoch": 2.1609221357748445,
+      "grad_norm": 0.31856610527953794,
+      "learning_rate": 5.529344280250111e-06,
+      "loss": 0.4474,
+      "step": 13158
+    },
+    {
+      "epoch": 2.161086362983187,
+      "grad_norm": 0.3171282603923975,
+      "learning_rate": 5.528887736413109e-06,
+      "loss": 0.457,
+      "step": 13159
+    },
+    {
+      "epoch": 2.16125059019153,
+      "grad_norm": 0.35217117196778364,
+      "learning_rate": 5.528431181401375e-06,
+      "loss": 0.4619,
+      "step": 13160
+    },
+    {
+      "epoch": 2.161414817399873,
+      "grad_norm": 0.6060833146178379,
+      "learning_rate": 5.527974615219866e-06,
+      "loss": 0.4392,
+      "step": 13161
+    },
+    {
+      "epoch": 2.1615790446082155,
+      "grad_norm": 0.589550773106566,
+      "learning_rate": 5.527518037873542e-06,
+      "loss": 0.4371,
+      "step": 13162
+    },
+    {
+      "epoch": 2.161743271816558,
+      "grad_norm": 0.3145128961142026,
+      "learning_rate": 5.527061449367359e-06,
+      "loss": 0.4506,
+      "step": 13163
+    },
+    {
+      "epoch": 2.1619074990249008,
+      "grad_norm": 0.28483661690939105,
+      "learning_rate": 5.52660484970628e-06,
+      "loss": 0.4407,
+      "step": 13164
+    },
+    {
+      "epoch": 2.162071726233244,
+      "grad_norm": 0.31160106147746397,
+      "learning_rate": 5.526148238895262e-06,
+      "loss": 0.4325,
+      "step": 13165
+    },
+    {
+      "epoch": 2.1622359534415865,
+      "grad_norm": 0.8214018457380634,
+      "learning_rate": 5.525691616939266e-06,
+      "loss": 0.449,
+      "step": 13166
+    },
+    {
+      "epoch": 2.162400180649929,
+      "grad_norm": 0.42579552944763127,
+      "learning_rate": 5.525234983843247e-06,
+      "loss": 0.4375,
+      "step": 13167
+    },
+    {
+      "epoch": 2.1625644078582718,
+      "grad_norm": 0.3172437304419483,
+      "learning_rate": 5.524778339612168e-06,
+      "loss": 0.4406,
+      "step": 13168
+    },
+    {
+      "epoch": 2.162728635066615,
+      "grad_norm": 0.3581226513019574,
+      "learning_rate": 5.5243216842509895e-06,
+      "loss": 0.4547,
+      "step": 13169
+    },
+    {
+      "epoch": 2.1628928622749575,
+      "grad_norm": 1.0570246497495848,
+      "learning_rate": 5.523865017764668e-06,
+      "loss": 0.4544,
+      "step": 13170
+    },
+    {
+      "epoch": 2.1630570894833,
+      "grad_norm": 0.3390380057126517,
+      "learning_rate": 5.523408340158167e-06,
+      "loss": 0.4214,
+      "step": 13171
+    },
+    {
+      "epoch": 2.1632213166916427,
+      "grad_norm": 0.36776919352984283,
+      "learning_rate": 5.5229516514364426e-06,
+      "loss": 0.4381,
+      "step": 13172
+    },
+    {
+      "epoch": 2.163385543899986,
+      "grad_norm": 0.30982836563580607,
+      "learning_rate": 5.522494951604457e-06,
+      "loss": 0.4489,
+      "step": 13173
+    },
+    {
+      "epoch": 2.1635497711083285,
+      "grad_norm": 0.44372920909803376,
+      "learning_rate": 5.522038240667172e-06,
+      "loss": 0.455,
+      "step": 13174
+    },
+    {
+      "epoch": 2.163713998316671,
+      "grad_norm": 0.3466451063504817,
+      "learning_rate": 5.521581518629544e-06,
+      "loss": 0.4525,
+      "step": 13175
+    },
+    {
+      "epoch": 2.1638782255250137,
+      "grad_norm": 0.3106973226661059,
+      "learning_rate": 5.521124785496538e-06,
+      "loss": 0.4483,
+      "step": 13176
+    },
+    {
+      "epoch": 2.164042452733357,
+      "grad_norm": 0.28579331113902645,
+      "learning_rate": 5.52066804127311e-06,
+      "loss": 0.445,
+      "step": 13177
+    },
+    {
+      "epoch": 2.1642066799416995,
+      "grad_norm": 0.26832241525859035,
+      "learning_rate": 5.5202112859642245e-06,
+      "loss": 0.4222,
+      "step": 13178
+    },
+    {
+      "epoch": 2.164370907150042,
+      "grad_norm": 0.32216668813831933,
+      "learning_rate": 5.51975451957484e-06,
+      "loss": 0.4538,
+      "step": 13179
+    },
+    {
+      "epoch": 2.1645351343583847,
+      "grad_norm": 0.34484591491438377,
+      "learning_rate": 5.519297742109918e-06,
+      "loss": 0.4371,
+      "step": 13180
+    },
+    {
+      "epoch": 2.1646993615667274,
+      "grad_norm": 0.328862402254495,
+      "learning_rate": 5.518840953574418e-06,
+      "loss": 0.4318,
+      "step": 13181
+    },
+    {
+      "epoch": 2.1648635887750705,
+      "grad_norm": 0.27757721400634844,
+      "learning_rate": 5.518384153973306e-06,
+      "loss": 0.4412,
+      "step": 13182
+    },
+    {
+      "epoch": 2.165027815983413,
+      "grad_norm": 0.32631979517754256,
+      "learning_rate": 5.517927343311538e-06,
+      "loss": 0.4317,
+      "step": 13183
+    },
+    {
+      "epoch": 2.1651920431917557,
+      "grad_norm": 0.27432476169664666,
+      "learning_rate": 5.517470521594078e-06,
+      "loss": 0.4293,
+      "step": 13184
+    },
+    {
+      "epoch": 2.1653562704000984,
+      "grad_norm": 0.28966912344757134,
+      "learning_rate": 5.517013688825888e-06,
+      "loss": 0.4549,
+      "step": 13185
+    },
+    {
+      "epoch": 2.1655204976084415,
+      "grad_norm": 0.6730067083934345,
+      "learning_rate": 5.516556845011929e-06,
+      "loss": 0.4474,
+      "step": 13186
+    },
+    {
+      "epoch": 2.165684724816784,
+      "grad_norm": 0.29913936874253383,
+      "learning_rate": 5.516099990157161e-06,
+      "loss": 0.4675,
+      "step": 13187
+    },
+    {
+      "epoch": 2.1658489520251267,
+      "grad_norm": 0.3424981637374919,
+      "learning_rate": 5.515643124266546e-06,
+      "loss": 0.4373,
+      "step": 13188
+    },
+    {
+      "epoch": 2.1660131792334694,
+      "grad_norm": 0.30630459205370697,
+      "learning_rate": 5.51518624734505e-06,
+      "loss": 0.4465,
+      "step": 13189
+    },
+    {
+      "epoch": 2.1661774064418124,
+      "grad_norm": 0.34883759198472747,
+      "learning_rate": 5.514729359397632e-06,
+      "loss": 0.4404,
+      "step": 13190
+    },
+    {
+      "epoch": 2.166341633650155,
+      "grad_norm": 0.302842026651347,
+      "learning_rate": 5.5142724604292555e-06,
+      "loss": 0.4529,
+      "step": 13191
+    },
+    {
+      "epoch": 2.1665058608584977,
+      "grad_norm": 0.3322886428250413,
+      "learning_rate": 5.513815550444881e-06,
+      "loss": 0.4315,
+      "step": 13192
+    },
+    {
+      "epoch": 2.1666700880668404,
+      "grad_norm": 0.3008715144635831,
+      "learning_rate": 5.513358629449472e-06,
+      "loss": 0.4503,
+      "step": 13193
+    },
+    {
+      "epoch": 2.1668343152751834,
+      "grad_norm": 0.31224920669877515,
+      "learning_rate": 5.512901697447992e-06,
+      "loss": 0.4351,
+      "step": 13194
+    },
+    {
+      "epoch": 2.166998542483526,
+      "grad_norm": 0.28463358286996604,
+      "learning_rate": 5.512444754445403e-06,
+      "loss": 0.4423,
+      "step": 13195
+    },
+    {
+      "epoch": 2.1671627696918687,
+      "grad_norm": 0.4034444644429901,
+      "learning_rate": 5.511987800446668e-06,
+      "loss": 0.4404,
+      "step": 13196
+    },
+    {
+      "epoch": 2.1673269969002114,
+      "grad_norm": 0.38665482565892567,
+      "learning_rate": 5.511530835456749e-06,
+      "loss": 0.4366,
+      "step": 13197
+    },
+    {
+      "epoch": 2.167491224108554,
+      "grad_norm": 0.3494804160464828,
+      "learning_rate": 5.51107385948061e-06,
+      "loss": 0.4525,
+      "step": 13198
+    },
+    {
+      "epoch": 2.167655451316897,
+      "grad_norm": 0.3841231205093899,
+      "learning_rate": 5.510616872523214e-06,
+      "loss": 0.4509,
+      "step": 13199
+    },
+    {
+      "epoch": 2.1678196785252397,
+      "grad_norm": 0.28514366579375827,
+      "learning_rate": 5.510159874589527e-06,
+      "loss": 0.4426,
+      "step": 13200
+    },
+    {
+      "epoch": 2.1679839057335824,
+      "grad_norm": 0.3620023664594434,
+      "learning_rate": 5.5097028656845065e-06,
+      "loss": 0.4358,
+      "step": 13201
+    },
+    {
+      "epoch": 2.168148132941925,
+      "grad_norm": 0.32201321607882394,
+      "learning_rate": 5.509245845813121e-06,
+      "loss": 0.4403,
+      "step": 13202
+    },
+    {
+      "epoch": 2.168312360150268,
+      "grad_norm": 0.3210763284964427,
+      "learning_rate": 5.508788814980333e-06,
+      "loss": 0.4568,
+      "step": 13203
+    },
+    {
+      "epoch": 2.1684765873586107,
+      "grad_norm": 0.3317228382026454,
+      "learning_rate": 5.508331773191104e-06,
+      "loss": 0.432,
+      "step": 13204
+    },
+    {
+      "epoch": 2.1686408145669533,
+      "grad_norm": 0.30859052454192315,
+      "learning_rate": 5.507874720450403e-06,
+      "loss": 0.4565,
+      "step": 13205
+    },
+    {
+      "epoch": 2.168805041775296,
+      "grad_norm": 0.27373704128356613,
+      "learning_rate": 5.507417656763189e-06,
+      "loss": 0.4408,
+      "step": 13206
+    },
+    {
+      "epoch": 2.168969268983639,
+      "grad_norm": 0.390830204231793,
+      "learning_rate": 5.506960582134428e-06,
+      "loss": 0.4389,
+      "step": 13207
+    },
+    {
+      "epoch": 2.1691334961919817,
+      "grad_norm": 0.316536079441099,
+      "learning_rate": 5.506503496569085e-06,
+      "loss": 0.4676,
+      "step": 13208
+    },
+    {
+      "epoch": 2.1692977234003243,
+      "grad_norm": 0.30981856257732066,
+      "learning_rate": 5.506046400072122e-06,
+      "loss": 0.4464,
+      "step": 13209
+    },
+    {
+      "epoch": 2.169461950608667,
+      "grad_norm": 0.8040089066478628,
+      "learning_rate": 5.505589292648508e-06,
+      "loss": 0.4571,
+      "step": 13210
+    },
+    {
+      "epoch": 2.16962617781701,
+      "grad_norm": 0.29944619080582335,
+      "learning_rate": 5.505132174303204e-06,
+      "loss": 0.4429,
+      "step": 13211
+    },
+    {
+      "epoch": 2.1697904050253527,
+      "grad_norm": 0.2931380112774569,
+      "learning_rate": 5.504675045041174e-06,
+      "loss": 0.432,
+      "step": 13212
+    },
+    {
+      "epoch": 2.1699546322336953,
+      "grad_norm": 0.32723713071086563,
+      "learning_rate": 5.504217904867386e-06,
+      "loss": 0.4673,
+      "step": 13213
+    },
+    {
+      "epoch": 2.170118859442038,
+      "grad_norm": 0.2680010066570768,
+      "learning_rate": 5.503760753786804e-06,
+      "loss": 0.4677,
+      "step": 13214
+    },
+    {
+      "epoch": 2.1702830866503806,
+      "grad_norm": 0.29538207235909225,
+      "learning_rate": 5.503303591804392e-06,
+      "loss": 0.4396,
+      "step": 13215
+    },
+    {
+      "epoch": 2.1704473138587237,
+      "grad_norm": 0.3072993424577965,
+      "learning_rate": 5.5028464189251155e-06,
+      "loss": 0.4505,
+      "step": 13216
+    },
+    {
+      "epoch": 2.1706115410670663,
+      "grad_norm": 0.311688439811001,
+      "learning_rate": 5.502389235153941e-06,
+      "loss": 0.4313,
+      "step": 13217
+    },
+    {
+      "epoch": 2.170775768275409,
+      "grad_norm": 0.35361331964980885,
+      "learning_rate": 5.501932040495832e-06,
+      "loss": 0.4646,
+      "step": 13218
+    },
+    {
+      "epoch": 2.1709399954837516,
+      "grad_norm": 0.339514386407374,
+      "learning_rate": 5.501474834955756e-06,
+      "loss": 0.4427,
+      "step": 13219
+    },
+    {
+      "epoch": 2.1711042226920947,
+      "grad_norm": 0.3359321760848037,
+      "learning_rate": 5.501017618538679e-06,
+      "loss": 0.4365,
+      "step": 13220
+    },
+    {
+      "epoch": 2.1712684499004373,
+      "grad_norm": 0.29013968578222354,
+      "learning_rate": 5.500560391249565e-06,
+      "loss": 0.4308,
+      "step": 13221
+    },
+    {
+      "epoch": 2.17143267710878,
+      "grad_norm": 0.33718799981468217,
+      "learning_rate": 5.5001031530933794e-06,
+      "loss": 0.4393,
+      "step": 13222
+    },
+    {
+      "epoch": 2.1715969043171226,
+      "grad_norm": 0.3303170855751071,
+      "learning_rate": 5.499645904075091e-06,
+      "loss": 0.4309,
+      "step": 13223
+    },
+    {
+      "epoch": 2.1717611315254657,
+      "grad_norm": 0.3150015557314923,
+      "learning_rate": 5.499188644199664e-06,
+      "loss": 0.4339,
+      "step": 13224
+    },
+    {
+      "epoch": 2.1719253587338083,
+      "grad_norm": 0.3075108331038909,
+      "learning_rate": 5.4987313734720665e-06,
+      "loss": 0.451,
+      "step": 13225
+    },
+    {
+      "epoch": 2.172089585942151,
+      "grad_norm": 0.33633278099740577,
+      "learning_rate": 5.4982740918972625e-06,
+      "loss": 0.431,
+      "step": 13226
+    },
+    {
+      "epoch": 2.1722538131504936,
+      "grad_norm": 0.2829229596881379,
+      "learning_rate": 5.497816799480219e-06,
+      "loss": 0.4352,
+      "step": 13227
+    },
+    {
+      "epoch": 2.1724180403588367,
+      "grad_norm": 0.2738157895536735,
+      "learning_rate": 5.497359496225905e-06,
+      "loss": 0.4403,
+      "step": 13228
+    },
+    {
+      "epoch": 2.1725822675671793,
+      "grad_norm": 0.3134244993684418,
+      "learning_rate": 5.496902182139286e-06,
+      "loss": 0.4459,
+      "step": 13229
+    },
+    {
+      "epoch": 2.172746494775522,
+      "grad_norm": 0.3357126734986133,
+      "learning_rate": 5.496444857225326e-06,
+      "loss": 0.4443,
+      "step": 13230
+    },
+    {
+      "epoch": 2.1729107219838646,
+      "grad_norm": 0.3924307054131407,
+      "learning_rate": 5.495987521488996e-06,
+      "loss": 0.437,
+      "step": 13231
+    },
+    {
+      "epoch": 2.1730749491922072,
+      "grad_norm": 0.3104051211706088,
+      "learning_rate": 5.495530174935261e-06,
+      "loss": 0.446,
+      "step": 13232
+    },
+    {
+      "epoch": 2.1732391764005503,
+      "grad_norm": 0.38846412274853226,
+      "learning_rate": 5.49507281756909e-06,
+      "loss": 0.4555,
+      "step": 13233
+    },
+    {
+      "epoch": 2.173403403608893,
+      "grad_norm": 0.3313264905531258,
+      "learning_rate": 5.4946154493954495e-06,
+      "loss": 0.4344,
+      "step": 13234
+    },
+    {
+      "epoch": 2.1735676308172356,
+      "grad_norm": 0.3243910241314963,
+      "learning_rate": 5.494158070419304e-06,
+      "loss": 0.4423,
+      "step": 13235
+    },
+    {
+      "epoch": 2.1737318580255782,
+      "grad_norm": 0.5319929603313633,
+      "learning_rate": 5.493700680645626e-06,
+      "loss": 0.4391,
+      "step": 13236
+    },
+    {
+      "epoch": 2.1738960852339213,
+      "grad_norm": 0.26307427359923147,
+      "learning_rate": 5.49324328007938e-06,
+      "loss": 0.4479,
+      "step": 13237
+    },
+    {
+      "epoch": 2.174060312442264,
+      "grad_norm": 0.3178794271208854,
+      "learning_rate": 5.492785868725535e-06,
+      "loss": 0.4577,
+      "step": 13238
+    },
+    {
+      "epoch": 2.1742245396506066,
+      "grad_norm": 0.2809221309088171,
+      "learning_rate": 5.49232844658906e-06,
+      "loss": 0.469,
+      "step": 13239
+    },
+    {
+      "epoch": 2.174388766858949,
+      "grad_norm": 0.3735418362737004,
+      "learning_rate": 5.491871013674921e-06,
+      "loss": 0.4233,
+      "step": 13240
+    },
+    {
+      "epoch": 2.1745529940672923,
+      "grad_norm": 0.2725105093139469,
+      "learning_rate": 5.491413569988085e-06,
+      "loss": 0.4543,
+      "step": 13241
+    },
+    {
+      "epoch": 2.174717221275635,
+      "grad_norm": 0.3151526891984052,
+      "learning_rate": 5.490956115533523e-06,
+      "loss": 0.4607,
+      "step": 13242
+    },
+    {
+      "epoch": 2.1748814484839776,
+      "grad_norm": 0.36699330490616217,
+      "learning_rate": 5.4904986503162035e-06,
+      "loss": 0.4554,
+      "step": 13243
+    },
+    {
+      "epoch": 2.17504567569232,
+      "grad_norm": 0.3380410448393794,
+      "learning_rate": 5.490041174341094e-06,
+      "loss": 0.4351,
+      "step": 13244
+    },
+    {
+      "epoch": 2.1752099029006633,
+      "grad_norm": 0.3373211564238718,
+      "learning_rate": 5.489583687613164e-06,
+      "loss": 0.4429,
+      "step": 13245
+    },
+    {
+      "epoch": 2.175374130109006,
+      "grad_norm": 0.3425286568997985,
+      "learning_rate": 5.48912619013738e-06,
+      "loss": 0.4482,
+      "step": 13246
+    },
+    {
+      "epoch": 2.1755383573173486,
+      "grad_norm": 0.30984839575858375,
+      "learning_rate": 5.488668681918712e-06,
+      "loss": 0.4354,
+      "step": 13247
+    },
+    {
+      "epoch": 2.175702584525691,
+      "grad_norm": 0.2957698078978961,
+      "learning_rate": 5.488211162962132e-06,
+      "loss": 0.4426,
+      "step": 13248
+    },
+    {
+      "epoch": 2.175866811734034,
+      "grad_norm": 0.3545053186690607,
+      "learning_rate": 5.487753633272605e-06,
+      "loss": 0.4529,
+      "step": 13249
+    },
+    {
+      "epoch": 2.176031038942377,
+      "grad_norm": 0.345155396866791,
+      "learning_rate": 5.4872960928551015e-06,
+      "loss": 0.4319,
+      "step": 13250
+    },
+    {
+      "epoch": 2.1761952661507196,
+      "grad_norm": 0.3096685753746912,
+      "learning_rate": 5.4868385417145905e-06,
+      "loss": 0.4251,
+      "step": 13251
+    },
+    {
+      "epoch": 2.176359493359062,
+      "grad_norm": 0.34819121870325165,
+      "learning_rate": 5.486380979856042e-06,
+      "loss": 0.439,
+      "step": 13252
+    },
+    {
+      "epoch": 2.176523720567405,
+      "grad_norm": 0.365145070586753,
+      "learning_rate": 5.485923407284428e-06,
+      "loss": 0.4552,
+      "step": 13253
+    },
+    {
+      "epoch": 2.176687947775748,
+      "grad_norm": 0.302409582339075,
+      "learning_rate": 5.4854658240047145e-06,
+      "loss": 0.4576,
+      "step": 13254
+    },
+    {
+      "epoch": 2.1768521749840906,
+      "grad_norm": 0.3099827305656743,
+      "learning_rate": 5.4850082300218725e-06,
+      "loss": 0.4182,
+      "step": 13255
+    },
+    {
+      "epoch": 2.177016402192433,
+      "grad_norm": 0.3303933405482676,
+      "learning_rate": 5.4845506253408705e-06,
+      "loss": 0.4438,
+      "step": 13256
+    },
+    {
+      "epoch": 2.177180629400776,
+      "grad_norm": 0.32728970434710136,
+      "learning_rate": 5.484093009966682e-06,
+      "loss": 0.4346,
+      "step": 13257
+    },
+    {
+      "epoch": 2.177344856609119,
+      "grad_norm": 0.32484025588411,
+      "learning_rate": 5.483635383904273e-06,
+      "loss": 0.4507,
+      "step": 13258
+    },
+    {
+      "epoch": 2.1775090838174616,
+      "grad_norm": 0.3801502598555356,
+      "learning_rate": 5.483177747158619e-06,
+      "loss": 0.4434,
+      "step": 13259
+    },
+    {
+      "epoch": 2.177673311025804,
+      "grad_norm": 0.2890146137698474,
+      "learning_rate": 5.482720099734686e-06,
+      "loss": 0.4617,
+      "step": 13260
+    },
+    {
+      "epoch": 2.177837538234147,
+      "grad_norm": 0.3036280477508943,
+      "learning_rate": 5.482262441637445e-06,
+      "loss": 0.4692,
+      "step": 13261
+    },
+    {
+      "epoch": 2.17800176544249,
+      "grad_norm": 0.2767654696097153,
+      "learning_rate": 5.481804772871868e-06,
+      "loss": 0.4392,
+      "step": 13262
+    },
+    {
+      "epoch": 2.1781659926508325,
+      "grad_norm": 0.3103322912436335,
+      "learning_rate": 5.481347093442926e-06,
+      "loss": 0.4351,
+      "step": 13263
+    },
+    {
+      "epoch": 2.178330219859175,
+      "grad_norm": 0.3730970128678796,
+      "learning_rate": 5.480889403355589e-06,
+      "loss": 0.446,
+      "step": 13264
+    },
+    {
+      "epoch": 2.178494447067518,
+      "grad_norm": 0.4406758847281105,
+      "learning_rate": 5.4804317026148274e-06,
+      "loss": 0.449,
+      "step": 13265
+    },
+    {
+      "epoch": 2.1786586742758605,
+      "grad_norm": 0.31286994077533276,
+      "learning_rate": 5.4799739912256126e-06,
+      "loss": 0.4337,
+      "step": 13266
+    },
+    {
+      "epoch": 2.1788229014842035,
+      "grad_norm": 0.31232970447969627,
+      "learning_rate": 5.479516269192915e-06,
+      "loss": 0.4538,
+      "step": 13267
+    },
+    {
+      "epoch": 2.178987128692546,
+      "grad_norm": 0.39424467153664117,
+      "learning_rate": 5.47905853652171e-06,
+      "loss": 0.4461,
+      "step": 13268
+    },
+    {
+      "epoch": 2.179151355900889,
+      "grad_norm": 0.2763369646076082,
+      "learning_rate": 5.4786007932169634e-06,
+      "loss": 0.4412,
+      "step": 13269
+    },
+    {
+      "epoch": 2.1793155831092315,
+      "grad_norm": 0.3549322006277845,
+      "learning_rate": 5.478143039283651e-06,
+      "loss": 0.4395,
+      "step": 13270
+    },
+    {
+      "epoch": 2.1794798103175745,
+      "grad_norm": 0.3806948554055574,
+      "learning_rate": 5.477685274726741e-06,
+      "loss": 0.4377,
+      "step": 13271
+    },
+    {
+      "epoch": 2.179644037525917,
+      "grad_norm": 0.36432326690094835,
+      "learning_rate": 5.477227499551208e-06,
+      "loss": 0.4354,
+      "step": 13272
+    },
+    {
+      "epoch": 2.17980826473426,
+      "grad_norm": 0.4599377738684147,
+      "learning_rate": 5.476769713762024e-06,
+      "loss": 0.4674,
+      "step": 13273
+    },
+    {
+      "epoch": 2.1799724919426025,
+      "grad_norm": 0.3250357559341009,
+      "learning_rate": 5.47631191736416e-06,
+      "loss": 0.4174,
+      "step": 13274
+    },
+    {
+      "epoch": 2.1801367191509455,
+      "grad_norm": 0.32868619559098194,
+      "learning_rate": 5.475854110362586e-06,
+      "loss": 0.4413,
+      "step": 13275
+    },
+    {
+      "epoch": 2.180300946359288,
+      "grad_norm": 0.34100178065605913,
+      "learning_rate": 5.475396292762278e-06,
+      "loss": 0.4561,
+      "step": 13276
+    },
+    {
+      "epoch": 2.180465173567631,
+      "grad_norm": 0.5793751621914814,
+      "learning_rate": 5.4749384645682054e-06,
+      "loss": 0.4586,
+      "step": 13277
+    },
+    {
+      "epoch": 2.1806294007759734,
+      "grad_norm": 0.4069008528073191,
+      "learning_rate": 5.474480625785343e-06,
+      "loss": 0.4164,
+      "step": 13278
+    },
+    {
+      "epoch": 2.1807936279843165,
+      "grad_norm": 0.3734986648935865,
+      "learning_rate": 5.474022776418661e-06,
+      "loss": 0.4493,
+      "step": 13279
+    },
+    {
+      "epoch": 2.180957855192659,
+      "grad_norm": 0.3441849653741825,
+      "learning_rate": 5.473564916473134e-06,
+      "loss": 0.4106,
+      "step": 13280
+    },
+    {
+      "epoch": 2.181122082401002,
+      "grad_norm": 0.29951787954088177,
+      "learning_rate": 5.473107045953734e-06,
+      "loss": 0.4416,
+      "step": 13281
+    },
+    {
+      "epoch": 2.1812863096093444,
+      "grad_norm": 0.4575444456948254,
+      "learning_rate": 5.472649164865434e-06,
+      "loss": 0.4501,
+      "step": 13282
+    },
+    {
+      "epoch": 2.181450536817687,
+      "grad_norm": 0.3720739585342384,
+      "learning_rate": 5.472191273213208e-06,
+      "loss": 0.433,
+      "step": 13283
+    },
+    {
+      "epoch": 2.18161476402603,
+      "grad_norm": 0.3481388997776785,
+      "learning_rate": 5.471733371002027e-06,
+      "loss": 0.4431,
+      "step": 13284
+    },
+    {
+      "epoch": 2.181778991234373,
+      "grad_norm": 0.31569298160947046,
+      "learning_rate": 5.471275458236865e-06,
+      "loss": 0.4709,
+      "step": 13285
+    },
+    {
+      "epoch": 2.1819432184427154,
+      "grad_norm": 0.3675943375619008,
+      "learning_rate": 5.470817534922698e-06,
+      "loss": 0.4618,
+      "step": 13286
+    },
+    {
+      "epoch": 2.182107445651058,
+      "grad_norm": 0.2889262086312248,
+      "learning_rate": 5.470359601064495e-06,
+      "loss": 0.4383,
+      "step": 13287
+    },
+    {
+      "epoch": 2.182271672859401,
+      "grad_norm": 0.38329425216994023,
+      "learning_rate": 5.469901656667235e-06,
+      "loss": 0.437,
+      "step": 13288
+    },
+    {
+      "epoch": 2.182435900067744,
+      "grad_norm": 0.3052468220993611,
+      "learning_rate": 5.469443701735887e-06,
+      "loss": 0.4634,
+      "step": 13289
+    },
+    {
+      "epoch": 2.1826001272760864,
+      "grad_norm": 0.3603417523141972,
+      "learning_rate": 5.468985736275426e-06,
+      "loss": 0.4242,
+      "step": 13290
+    },
+    {
+      "epoch": 2.182764354484429,
+      "grad_norm": 0.27351967521952214,
+      "learning_rate": 5.468527760290828e-06,
+      "loss": 0.4479,
+      "step": 13291
+    },
+    {
+      "epoch": 2.182928581692772,
+      "grad_norm": 0.7082922791523382,
+      "learning_rate": 5.468069773787066e-06,
+      "loss": 0.4306,
+      "step": 13292
+    },
+    {
+      "epoch": 2.183092808901115,
+      "grad_norm": 0.3798345765622769,
+      "learning_rate": 5.467611776769112e-06,
+      "loss": 0.4247,
+      "step": 13293
+    },
+    {
+      "epoch": 2.1832570361094574,
+      "grad_norm": 0.532693372970332,
+      "learning_rate": 5.467153769241942e-06,
+      "loss": 0.4451,
+      "step": 13294
+    },
+    {
+      "epoch": 2.1834212633178,
+      "grad_norm": 0.37091825250503574,
+      "learning_rate": 5.466695751210532e-06,
+      "loss": 0.4563,
+      "step": 13295
+    },
+    {
+      "epoch": 2.183585490526143,
+      "grad_norm": 0.3210104071626388,
+      "learning_rate": 5.466237722679854e-06,
+      "loss": 0.435,
+      "step": 13296
+    },
+    {
+      "epoch": 2.183749717734486,
+      "grad_norm": 0.280191494056349,
+      "learning_rate": 5.465779683654884e-06,
+      "loss": 0.4162,
+      "step": 13297
+    },
+    {
+      "epoch": 2.1839139449428284,
+      "grad_norm": 0.3062510831224376,
+      "learning_rate": 5.465321634140597e-06,
+      "loss": 0.4377,
+      "step": 13298
+    },
+    {
+      "epoch": 2.184078172151171,
+      "grad_norm": 0.5031795325550483,
+      "learning_rate": 5.464863574141968e-06,
+      "loss": 0.4578,
+      "step": 13299
+    },
+    {
+      "epoch": 2.1842423993595137,
+      "grad_norm": 0.3211714977323895,
+      "learning_rate": 5.464405503663969e-06,
+      "loss": 0.436,
+      "step": 13300
+    },
+    {
+      "epoch": 2.184406626567857,
+      "grad_norm": 0.3563243829030005,
+      "learning_rate": 5.463947422711578e-06,
+      "loss": 0.4407,
+      "step": 13301
+    },
+    {
+      "epoch": 2.1845708537761994,
+      "grad_norm": 0.5806060878516683,
+      "learning_rate": 5.46348933128977e-06,
+      "loss": 0.4444,
+      "step": 13302
+    },
+    {
+      "epoch": 2.184735080984542,
+      "grad_norm": 0.3560099487190159,
+      "learning_rate": 5.463031229403521e-06,
+      "loss": 0.4552,
+      "step": 13303
+    },
+    {
+      "epoch": 2.1848993081928847,
+      "grad_norm": 0.27206041929266034,
+      "learning_rate": 5.462573117057804e-06,
+      "loss": 0.4567,
+      "step": 13304
+    },
+    {
+      "epoch": 2.1850635354012278,
+      "grad_norm": 0.3331834230070897,
+      "learning_rate": 5.462114994257596e-06,
+      "loss": 0.4334,
+      "step": 13305
+    },
+    {
+      "epoch": 2.1852277626095704,
+      "grad_norm": 0.31765346174580844,
+      "learning_rate": 5.461656861007872e-06,
+      "loss": 0.4502,
+      "step": 13306
+    },
+    {
+      "epoch": 2.185391989817913,
+      "grad_norm": 0.38588394716998564,
+      "learning_rate": 5.461198717313611e-06,
+      "loss": 0.4341,
+      "step": 13307
+    },
+    {
+      "epoch": 2.1855562170262557,
+      "grad_norm": 0.37336262040504764,
+      "learning_rate": 5.460740563179784e-06,
+      "loss": 0.4542,
+      "step": 13308
+    },
+    {
+      "epoch": 2.1857204442345988,
+      "grad_norm": 0.30693822439100865,
+      "learning_rate": 5.46028239861137e-06,
+      "loss": 0.4285,
+      "step": 13309
+    },
+    {
+      "epoch": 2.1858846714429414,
+      "grad_norm": 0.35159606342901895,
+      "learning_rate": 5.4598242236133434e-06,
+      "loss": 0.439,
+      "step": 13310
+    },
+    {
+      "epoch": 2.186048898651284,
+      "grad_norm": 0.5059138852284416,
+      "learning_rate": 5.459366038190682e-06,
+      "loss": 0.4463,
+      "step": 13311
+    },
+    {
+      "epoch": 2.1862131258596267,
+      "grad_norm": 0.42008476646526727,
+      "learning_rate": 5.458907842348362e-06,
+      "loss": 0.4468,
+      "step": 13312
+    },
+    {
+      "epoch": 2.1863773530679698,
+      "grad_norm": 0.4795580884704337,
+      "learning_rate": 5.458449636091359e-06,
+      "loss": 0.4393,
+      "step": 13313
+    },
+    {
+      "epoch": 2.1865415802763124,
+      "grad_norm": 0.2752909964148785,
+      "learning_rate": 5.457991419424649e-06,
+      "loss": 0.4263,
+      "step": 13314
+    },
+    {
+      "epoch": 2.186705807484655,
+      "grad_norm": 0.2727469759578638,
+      "learning_rate": 5.45753319235321e-06,
+      "loss": 0.4538,
+      "step": 13315
+    },
+    {
+      "epoch": 2.1868700346929977,
+      "grad_norm": 0.4963736171783304,
+      "learning_rate": 5.45707495488202e-06,
+      "loss": 0.4378,
+      "step": 13316
+    },
+    {
+      "epoch": 2.1870342619013403,
+      "grad_norm": 0.3257932535539851,
+      "learning_rate": 5.456616707016054e-06,
+      "loss": 0.46,
+      "step": 13317
+    },
+    {
+      "epoch": 2.1871984891096834,
+      "grad_norm": 0.30857930741888084,
+      "learning_rate": 5.456158448760289e-06,
+      "loss": 0.4387,
+      "step": 13318
+    },
+    {
+      "epoch": 2.187362716318026,
+      "grad_norm": 0.3837863937799594,
+      "learning_rate": 5.455700180119701e-06,
+      "loss": 0.4317,
+      "step": 13319
+    },
+    {
+      "epoch": 2.1875269435263687,
+      "grad_norm": 0.46225930270367893,
+      "learning_rate": 5.45524190109927e-06,
+      "loss": 0.4528,
+      "step": 13320
+    },
+    {
+      "epoch": 2.1876911707347113,
+      "grad_norm": 0.302233763714421,
+      "learning_rate": 5.454783611703972e-06,
+      "loss": 0.4556,
+      "step": 13321
+    },
+    {
+      "epoch": 2.1878553979430544,
+      "grad_norm": 0.34696738770294017,
+      "learning_rate": 5.454325311938786e-06,
+      "loss": 0.4173,
+      "step": 13322
+    },
+    {
+      "epoch": 2.188019625151397,
+      "grad_norm": 0.28421677391663946,
+      "learning_rate": 5.453867001808686e-06,
+      "loss": 0.4347,
+      "step": 13323
+    },
+    {
+      "epoch": 2.1881838523597397,
+      "grad_norm": 0.2761714620499444,
+      "learning_rate": 5.453408681318653e-06,
+      "loss": 0.4571,
+      "step": 13324
+    },
+    {
+      "epoch": 2.1883480795680823,
+      "grad_norm": 0.28369005370351114,
+      "learning_rate": 5.452950350473663e-06,
+      "loss": 0.4234,
+      "step": 13325
+    },
+    {
+      "epoch": 2.1885123067764254,
+      "grad_norm": 0.3172399933133409,
+      "learning_rate": 5.452492009278697e-06,
+      "loss": 0.4257,
+      "step": 13326
+    },
+    {
+      "epoch": 2.188676533984768,
+      "grad_norm": 0.38036218554406415,
+      "learning_rate": 5.452033657738727e-06,
+      "loss": 0.4449,
+      "step": 13327
+    },
+    {
+      "epoch": 2.1888407611931107,
+      "grad_norm": 0.2983968064999166,
+      "learning_rate": 5.4515752958587376e-06,
+      "loss": 0.4528,
+      "step": 13328
+    },
+    {
+      "epoch": 2.1890049884014533,
+      "grad_norm": 0.4085831022949406,
+      "learning_rate": 5.4511169236437026e-06,
+      "loss": 0.4323,
+      "step": 13329
+    },
+    {
+      "epoch": 2.1891692156097964,
+      "grad_norm": 0.3073148198837579,
+      "learning_rate": 5.450658541098603e-06,
+      "loss": 0.4251,
+      "step": 13330
+    },
+    {
+      "epoch": 2.189333442818139,
+      "grad_norm": 0.3552304549244225,
+      "learning_rate": 5.450200148228416e-06,
+      "loss": 0.4332,
+      "step": 13331
+    },
+    {
+      "epoch": 2.1894976700264817,
+      "grad_norm": 0.4129208416334564,
+      "learning_rate": 5.449741745038121e-06,
+      "loss": 0.4513,
+      "step": 13332
+    },
+    {
+      "epoch": 2.1896618972348243,
+      "grad_norm": 0.45681481402253127,
+      "learning_rate": 5.449283331532696e-06,
+      "loss": 0.4273,
+      "step": 13333
+    },
+    {
+      "epoch": 2.189826124443167,
+      "grad_norm": 0.28285821802334854,
+      "learning_rate": 5.4488249077171185e-06,
+      "loss": 0.4364,
+      "step": 13334
+    },
+    {
+      "epoch": 2.18999035165151,
+      "grad_norm": 0.3035145508717711,
+      "learning_rate": 5.44836647359637e-06,
+      "loss": 0.4488,
+      "step": 13335
+    },
+    {
+      "epoch": 2.1901545788598527,
+      "grad_norm": 0.44449304721783334,
+      "learning_rate": 5.447908029175429e-06,
+      "loss": 0.437,
+      "step": 13336
+    },
+    {
+      "epoch": 2.1903188060681953,
+      "grad_norm": 0.3076759098785597,
+      "learning_rate": 5.447449574459275e-06,
+      "loss": 0.4598,
+      "step": 13337
+    },
+    {
+      "epoch": 2.190483033276538,
+      "grad_norm": 0.35226795756882684,
+      "learning_rate": 5.446991109452884e-06,
+      "loss": 0.4451,
+      "step": 13338
+    },
+    {
+      "epoch": 2.190647260484881,
+      "grad_norm": 0.3400446784915832,
+      "learning_rate": 5.44653263416124e-06,
+      "loss": 0.4341,
+      "step": 13339
+    },
+    {
+      "epoch": 2.1908114876932236,
+      "grad_norm": 0.3271085565370917,
+      "learning_rate": 5.446074148589319e-06,
+      "loss": 0.4204,
+      "step": 13340
+    },
+    {
+      "epoch": 2.1909757149015663,
+      "grad_norm": 0.31458599529445347,
+      "learning_rate": 5.445615652742105e-06,
+      "loss": 0.4514,
+      "step": 13341
+    },
+    {
+      "epoch": 2.191139942109909,
+      "grad_norm": 0.3009846308571748,
+      "learning_rate": 5.445157146624571e-06,
+      "loss": 0.4541,
+      "step": 13342
+    },
+    {
+      "epoch": 2.191304169318252,
+      "grad_norm": 0.31103393177917293,
+      "learning_rate": 5.444698630241701e-06,
+      "loss": 0.4652,
+      "step": 13343
+    },
+    {
+      "epoch": 2.1914683965265946,
+      "grad_norm": 0.283057430214247,
+      "learning_rate": 5.444240103598475e-06,
+      "loss": 0.4532,
+      "step": 13344
+    },
+    {
+      "epoch": 2.1916326237349373,
+      "grad_norm": 0.29297163882420746,
+      "learning_rate": 5.4437815666998725e-06,
+      "loss": 0.4457,
+      "step": 13345
+    },
+    {
+      "epoch": 2.19179685094328,
+      "grad_norm": 0.3287114100302679,
+      "learning_rate": 5.4433230195508744e-06,
+      "loss": 0.4414,
+      "step": 13346
+    },
+    {
+      "epoch": 2.191961078151623,
+      "grad_norm": 0.33619588189444394,
+      "learning_rate": 5.442864462156459e-06,
+      "loss": 0.4293,
+      "step": 13347
+    },
+    {
+      "epoch": 2.1921253053599656,
+      "grad_norm": 0.30645737669849693,
+      "learning_rate": 5.442405894521608e-06,
+      "loss": 0.4648,
+      "step": 13348
+    },
+    {
+      "epoch": 2.1922895325683083,
+      "grad_norm": 0.29798132461815596,
+      "learning_rate": 5.441947316651303e-06,
+      "loss": 0.4462,
+      "step": 13349
+    },
+    {
+      "epoch": 2.192453759776651,
+      "grad_norm": 0.3376869196756182,
+      "learning_rate": 5.441488728550522e-06,
+      "loss": 0.4192,
+      "step": 13350
+    },
+    {
+      "epoch": 2.1926179869849935,
+      "grad_norm": 0.3381138888087342,
+      "learning_rate": 5.4410301302242485e-06,
+      "loss": 0.4398,
+      "step": 13351
+    },
+    {
+      "epoch": 2.1927822141933366,
+      "grad_norm": 0.3274713175485761,
+      "learning_rate": 5.440571521677461e-06,
+      "loss": 0.4326,
+      "step": 13352
+    },
+    {
+      "epoch": 2.1929464414016793,
+      "grad_norm": 0.3378498619220711,
+      "learning_rate": 5.440112902915141e-06,
+      "loss": 0.4504,
+      "step": 13353
+    },
+    {
+      "epoch": 2.193110668610022,
+      "grad_norm": 0.3158528798529663,
+      "learning_rate": 5.439654273942271e-06,
+      "loss": 0.4214,
+      "step": 13354
+    },
+    {
+      "epoch": 2.1932748958183645,
+      "grad_norm": 0.3505857824282715,
+      "learning_rate": 5.439195634763829e-06,
+      "loss": 0.4612,
+      "step": 13355
+    },
+    {
+      "epoch": 2.1934391230267076,
+      "grad_norm": 0.28799065331510204,
+      "learning_rate": 5.438736985384801e-06,
+      "loss": 0.4298,
+      "step": 13356
+    },
+    {
+      "epoch": 2.1936033502350503,
+      "grad_norm": 0.6200882777286091,
+      "learning_rate": 5.438278325810165e-06,
+      "loss": 0.4536,
+      "step": 13357
+    },
+    {
+      "epoch": 2.193767577443393,
+      "grad_norm": 0.270771121171262,
+      "learning_rate": 5.437819656044903e-06,
+      "loss": 0.4481,
+      "step": 13358
+    },
+    {
+      "epoch": 2.1939318046517355,
+      "grad_norm": 0.42757081634472144,
+      "learning_rate": 5.437360976093996e-06,
+      "loss": 0.4518,
+      "step": 13359
+    },
+    {
+      "epoch": 2.1940960318600786,
+      "grad_norm": 0.365960276967426,
+      "learning_rate": 5.436902285962429e-06,
+      "loss": 0.4292,
+      "step": 13360
+    },
+    {
+      "epoch": 2.1942602590684213,
+      "grad_norm": 0.3204066220765556,
+      "learning_rate": 5.436443585655178e-06,
+      "loss": 0.4467,
+      "step": 13361
+    },
+    {
+      "epoch": 2.194424486276764,
+      "grad_norm": 0.623873347080686,
+      "learning_rate": 5.435984875177231e-06,
+      "loss": 0.4261,
+      "step": 13362
+    },
+    {
+      "epoch": 2.1945887134851065,
+      "grad_norm": 0.338318483075795,
+      "learning_rate": 5.435526154533565e-06,
+      "loss": 0.4425,
+      "step": 13363
+    },
+    {
+      "epoch": 2.1947529406934496,
+      "grad_norm": 0.2930560099271347,
+      "learning_rate": 5.4350674237291666e-06,
+      "loss": 0.4622,
+      "step": 13364
+    },
+    {
+      "epoch": 2.1949171679017923,
+      "grad_norm": 0.4904425915483045,
+      "learning_rate": 5.434608682769016e-06,
+      "loss": 0.4576,
+      "step": 13365
+    },
+    {
+      "epoch": 2.195081395110135,
+      "grad_norm": 0.36543303457059684,
+      "learning_rate": 5.434149931658095e-06,
+      "loss": 0.4477,
+      "step": 13366
+    },
+    {
+      "epoch": 2.1952456223184775,
+      "grad_norm": 0.373200542196656,
+      "learning_rate": 5.433691170401385e-06,
+      "loss": 0.4362,
+      "step": 13367
+    },
+    {
+      "epoch": 2.19540984952682,
+      "grad_norm": 0.3640609524375743,
+      "learning_rate": 5.433232399003872e-06,
+      "loss": 0.424,
+      "step": 13368
+    },
+    {
+      "epoch": 2.1955740767351632,
+      "grad_norm": 0.31082790107927044,
+      "learning_rate": 5.4327736174705355e-06,
+      "loss": 0.4329,
+      "step": 13369
+    },
+    {
+      "epoch": 2.195738303943506,
+      "grad_norm": 0.5255387391693609,
+      "learning_rate": 5.432314825806362e-06,
+      "loss": 0.4538,
+      "step": 13370
+    },
+    {
+      "epoch": 2.1959025311518485,
+      "grad_norm": 0.3212060850181198,
+      "learning_rate": 5.431856024016333e-06,
+      "loss": 0.4353,
+      "step": 13371
+    },
+    {
+      "epoch": 2.196066758360191,
+      "grad_norm": 0.313522628120879,
+      "learning_rate": 5.4313972121054275e-06,
+      "loss": 0.4384,
+      "step": 13372
+    },
+    {
+      "epoch": 2.1962309855685342,
+      "grad_norm": 0.3749338646774257,
+      "learning_rate": 5.430938390078634e-06,
+      "loss": 0.4467,
+      "step": 13373
+    },
+    {
+      "epoch": 2.196395212776877,
+      "grad_norm": 0.2977863779812921,
+      "learning_rate": 5.430479557940933e-06,
+      "loss": 0.4495,
+      "step": 13374
+    },
+    {
+      "epoch": 2.1965594399852195,
+      "grad_norm": 0.28315765122651704,
+      "learning_rate": 5.430020715697309e-06,
+      "loss": 0.4448,
+      "step": 13375
+    },
+    {
+      "epoch": 2.196723667193562,
+      "grad_norm": 0.30598231606697035,
+      "learning_rate": 5.429561863352744e-06,
+      "loss": 0.4422,
+      "step": 13376
+    },
+    {
+      "epoch": 2.1968878944019052,
+      "grad_norm": 0.35302107028478596,
+      "learning_rate": 5.429103000912222e-06,
+      "loss": 0.4544,
+      "step": 13377
+    },
+    {
+      "epoch": 2.197052121610248,
+      "grad_norm": 0.8169332006027017,
+      "learning_rate": 5.42864412838073e-06,
+      "loss": 0.4185,
+      "step": 13378
+    },
+    {
+      "epoch": 2.1972163488185905,
+      "grad_norm": 0.3353800209447468,
+      "learning_rate": 5.4281852457632475e-06,
+      "loss": 0.4456,
+      "step": 13379
+    },
+    {
+      "epoch": 2.197380576026933,
+      "grad_norm": 0.29430933701926504,
+      "learning_rate": 5.42772635306476e-06,
+      "loss": 0.4505,
+      "step": 13380
+    },
+    {
+      "epoch": 2.1975448032352762,
+      "grad_norm": 0.448133699112051,
+      "learning_rate": 5.427267450290251e-06,
+      "loss": 0.4529,
+      "step": 13381
+    },
+    {
+      "epoch": 2.197709030443619,
+      "grad_norm": 0.32960691354841337,
+      "learning_rate": 5.426808537444707e-06,
+      "loss": 0.4342,
+      "step": 13382
+    },
+    {
+      "epoch": 2.1978732576519615,
+      "grad_norm": 0.33870856955694467,
+      "learning_rate": 5.42634961453311e-06,
+      "loss": 0.4419,
+      "step": 13383
+    },
+    {
+      "epoch": 2.198037484860304,
+      "grad_norm": 0.34599685380848416,
+      "learning_rate": 5.425890681560443e-06,
+      "loss": 0.4368,
+      "step": 13384
+    },
+    {
+      "epoch": 2.198201712068647,
+      "grad_norm": 0.5129549940134407,
+      "learning_rate": 5.425431738531693e-06,
+      "loss": 0.4605,
+      "step": 13385
+    },
+    {
+      "epoch": 2.19836593927699,
+      "grad_norm": 0.3833834807966303,
+      "learning_rate": 5.424972785451844e-06,
+      "loss": 0.4247,
+      "step": 13386
+    },
+    {
+      "epoch": 2.1985301664853325,
+      "grad_norm": 0.3204893316070029,
+      "learning_rate": 5.424513822325881e-06,
+      "loss": 0.4422,
+      "step": 13387
+    },
+    {
+      "epoch": 2.198694393693675,
+      "grad_norm": 0.36698443042704193,
+      "learning_rate": 5.4240548491587885e-06,
+      "loss": 0.413,
+      "step": 13388
+    },
+    {
+      "epoch": 2.198858620902018,
+      "grad_norm": 0.3811017115206296,
+      "learning_rate": 5.4235958659555495e-06,
+      "loss": 0.4781,
+      "step": 13389
+    },
+    {
+      "epoch": 2.199022848110361,
+      "grad_norm": 0.32255326893783004,
+      "learning_rate": 5.4231368727211526e-06,
+      "loss": 0.4523,
+      "step": 13390
+    },
+    {
+      "epoch": 2.1991870753187035,
+      "grad_norm": 0.572634516628206,
+      "learning_rate": 5.422677869460581e-06,
+      "loss": 0.4524,
+      "step": 13391
+    },
+    {
+      "epoch": 2.199351302527046,
+      "grad_norm": 0.3633323388041515,
+      "learning_rate": 5.422218856178818e-06,
+      "loss": 0.422,
+      "step": 13392
+    },
+    {
+      "epoch": 2.1995155297353888,
+      "grad_norm": 0.34242897491608737,
+      "learning_rate": 5.421759832880852e-06,
+      "loss": 0.4436,
+      "step": 13393
+    },
+    {
+      "epoch": 2.199679756943732,
+      "grad_norm": 0.36406147142191303,
+      "learning_rate": 5.421300799571668e-06,
+      "loss": 0.4312,
+      "step": 13394
+    },
+    {
+      "epoch": 2.1998439841520745,
+      "grad_norm": 0.2844107960510927,
+      "learning_rate": 5.420841756256251e-06,
+      "loss": 0.4572,
+      "step": 13395
+    },
+    {
+      "epoch": 2.200008211360417,
+      "grad_norm": 0.33568246893123604,
+      "learning_rate": 5.420382702939585e-06,
+      "loss": 0.4307,
+      "step": 13396
+    },
+    {
+      "epoch": 2.2001724385687598,
+      "grad_norm": 0.310811054875232,
+      "learning_rate": 5.419923639626657e-06,
+      "loss": 0.4317,
+      "step": 13397
+    },
+    {
+      "epoch": 2.200336665777103,
+      "grad_norm": 0.3290758135242331,
+      "learning_rate": 5.419464566322454e-06,
+      "loss": 0.4512,
+      "step": 13398
+    },
+    {
+      "epoch": 2.2005008929854455,
+      "grad_norm": 0.30047132711866265,
+      "learning_rate": 5.419005483031963e-06,
+      "loss": 0.4323,
+      "step": 13399
+    },
+    {
+      "epoch": 2.200665120193788,
+      "grad_norm": 0.2852061143153248,
+      "learning_rate": 5.4185463897601675e-06,
+      "loss": 0.4317,
+      "step": 13400
+    },
+    {
+      "epoch": 2.2008293474021308,
+      "grad_norm": 0.3378154655198677,
+      "learning_rate": 5.418087286512053e-06,
+      "loss": 0.4562,
+      "step": 13401
+    },
+    {
+      "epoch": 2.2009935746104734,
+      "grad_norm": 0.3812354025796814,
+      "learning_rate": 5.4176281732926076e-06,
+      "loss": 0.4594,
+      "step": 13402
+    },
+    {
+      "epoch": 2.2011578018188165,
+      "grad_norm": 0.41651687708085594,
+      "learning_rate": 5.417169050106818e-06,
+      "loss": 0.4339,
+      "step": 13403
+    },
+    {
+      "epoch": 2.201322029027159,
+      "grad_norm": 0.2878243822875417,
+      "learning_rate": 5.41670991695967e-06,
+      "loss": 0.4233,
+      "step": 13404
+    },
+    {
+      "epoch": 2.2014862562355018,
+      "grad_norm": 0.32510137346731693,
+      "learning_rate": 5.416250773856151e-06,
+      "loss": 0.4496,
+      "step": 13405
+    },
+    {
+      "epoch": 2.2016504834438444,
+      "grad_norm": 0.2746385162503503,
+      "learning_rate": 5.415791620801245e-06,
+      "loss": 0.4438,
+      "step": 13406
+    },
+    {
+      "epoch": 2.2018147106521875,
+      "grad_norm": 0.34944524205436805,
+      "learning_rate": 5.415332457799944e-06,
+      "loss": 0.45,
+      "step": 13407
+    },
+    {
+      "epoch": 2.20197893786053,
+      "grad_norm": 0.35040689222504817,
+      "learning_rate": 5.41487328485723e-06,
+      "loss": 0.4436,
+      "step": 13408
+    },
+    {
+      "epoch": 2.2021431650688728,
+      "grad_norm": 0.2910862319373962,
+      "learning_rate": 5.414414101978092e-06,
+      "loss": 0.4252,
+      "step": 13409
+    },
+    {
+      "epoch": 2.2023073922772154,
+      "grad_norm": 0.28937276374288046,
+      "learning_rate": 5.413954909167518e-06,
+      "loss": 0.4517,
+      "step": 13410
+    },
+    {
+      "epoch": 2.2024716194855585,
+      "grad_norm": 0.5219373842745609,
+      "learning_rate": 5.413495706430494e-06,
+      "loss": 0.4383,
+      "step": 13411
+    },
+    {
+      "epoch": 2.202635846693901,
+      "grad_norm": 8.93823507260522,
+      "learning_rate": 5.4130364937720085e-06,
+      "loss": 0.4431,
+      "step": 13412
+    },
+    {
+      "epoch": 2.2028000739022437,
+      "grad_norm": 0.31732089025480203,
+      "learning_rate": 5.412577271197047e-06,
+      "loss": 0.4258,
+      "step": 13413
+    },
+    {
+      "epoch": 2.2029643011105864,
+      "grad_norm": 0.3292152186596089,
+      "learning_rate": 5.4121180387105995e-06,
+      "loss": 0.4336,
+      "step": 13414
+    },
+    {
+      "epoch": 2.2031285283189295,
+      "grad_norm": 0.3087280941473258,
+      "learning_rate": 5.411658796317653e-06,
+      "loss": 0.4587,
+      "step": 13415
+    },
+    {
+      "epoch": 2.203292755527272,
+      "grad_norm": 0.3653836232230834,
+      "learning_rate": 5.411199544023195e-06,
+      "loss": 0.4342,
+      "step": 13416
+    },
+    {
+      "epoch": 2.2034569827356147,
+      "grad_norm": 0.27311329098667947,
+      "learning_rate": 5.410740281832212e-06,
+      "loss": 0.4418,
+      "step": 13417
+    },
+    {
+      "epoch": 2.2036212099439574,
+      "grad_norm": 0.49616160219727334,
+      "learning_rate": 5.410281009749694e-06,
+      "loss": 0.4256,
+      "step": 13418
+    },
+    {
+      "epoch": 2.2037854371523,
+      "grad_norm": 0.4126504742938762,
+      "learning_rate": 5.40982172778063e-06,
+      "loss": 0.4317,
+      "step": 13419
+    },
+    {
+      "epoch": 2.203949664360643,
+      "grad_norm": 0.5683373155278401,
+      "learning_rate": 5.409362435930006e-06,
+      "loss": 0.446,
+      "step": 13420
+    },
+    {
+      "epoch": 2.2041138915689857,
+      "grad_norm": 0.5772482685621865,
+      "learning_rate": 5.408903134202812e-06,
+      "loss": 0.4537,
+      "step": 13421
+    },
+    {
+      "epoch": 2.2042781187773284,
+      "grad_norm": 0.9024243798707035,
+      "learning_rate": 5.408443822604033e-06,
+      "loss": 0.4324,
+      "step": 13422
+    },
+    {
+      "epoch": 2.204442345985671,
+      "grad_norm": 0.432771176343393,
+      "learning_rate": 5.407984501138664e-06,
+      "loss": 0.4397,
+      "step": 13423
+    },
+    {
+      "epoch": 2.204606573194014,
+      "grad_norm": 0.6234869815136771,
+      "learning_rate": 5.407525169811689e-06,
+      "loss": 0.4374,
+      "step": 13424
+    },
+    {
+      "epoch": 2.2047708004023567,
+      "grad_norm": 0.5594566730793743,
+      "learning_rate": 5.4070658286280965e-06,
+      "loss": 0.4396,
+      "step": 13425
+    },
+    {
+      "epoch": 2.2049350276106994,
+      "grad_norm": 0.5564798666728442,
+      "learning_rate": 5.406606477592876e-06,
+      "loss": 0.4682,
+      "step": 13426
+    },
+    {
+      "epoch": 2.205099254819042,
+      "grad_norm": 0.4843442564867565,
+      "learning_rate": 5.406147116711019e-06,
+      "loss": 0.437,
+      "step": 13427
+    },
+    {
+      "epoch": 2.205263482027385,
+      "grad_norm": 0.36792018386639114,
+      "learning_rate": 5.405687745987512e-06,
+      "loss": 0.4345,
+      "step": 13428
+    },
+    {
+      "epoch": 2.2054277092357277,
+      "grad_norm": 0.3879908470046889,
+      "learning_rate": 5.405228365427346e-06,
+      "loss": 0.4384,
+      "step": 13429
+    },
+    {
+      "epoch": 2.2055919364440704,
+      "grad_norm": 0.45374078326582784,
+      "learning_rate": 5.404768975035508e-06,
+      "loss": 0.4356,
+      "step": 13430
+    },
+    {
+      "epoch": 2.205756163652413,
+      "grad_norm": 0.3407757604959021,
+      "learning_rate": 5.404309574816988e-06,
+      "loss": 0.4366,
+      "step": 13431
+    },
+    {
+      "epoch": 2.205920390860756,
+      "grad_norm": 0.3141407896315137,
+      "learning_rate": 5.4038501647767785e-06,
+      "loss": 0.4357,
+      "step": 13432
+    },
+    {
+      "epoch": 2.2060846180690987,
+      "grad_norm": 0.3267762123383814,
+      "learning_rate": 5.403390744919866e-06,
+      "loss": 0.4614,
+      "step": 13433
+    },
+    {
+      "epoch": 2.2062488452774414,
+      "grad_norm": 0.37945900480305744,
+      "learning_rate": 5.4029313152512424e-06,
+      "loss": 0.4551,
+      "step": 13434
+    },
+    {
+      "epoch": 2.206413072485784,
+      "grad_norm": 0.4124331195609138,
+      "learning_rate": 5.402471875775894e-06,
+      "loss": 0.4538,
+      "step": 13435
+    },
+    {
+      "epoch": 2.2065772996941266,
+      "grad_norm": 0.6909163990609498,
+      "learning_rate": 5.402012426498814e-06,
+      "loss": 0.4542,
+      "step": 13436
+    },
+    {
+      "epoch": 2.2067415269024697,
+      "grad_norm": 0.4938727070001304,
+      "learning_rate": 5.401552967424993e-06,
+      "loss": 0.4599,
+      "step": 13437
+    },
+    {
+      "epoch": 2.2069057541108124,
+      "grad_norm": 0.4031025804524951,
+      "learning_rate": 5.401093498559418e-06,
+      "loss": 0.4301,
+      "step": 13438
+    },
+    {
+      "epoch": 2.207069981319155,
+      "grad_norm": 0.3031333079133587,
+      "learning_rate": 5.400634019907082e-06,
+      "loss": 0.429,
+      "step": 13439
+    },
+    {
+      "epoch": 2.2072342085274976,
+      "grad_norm": 0.4760746129352209,
+      "learning_rate": 5.400174531472973e-06,
+      "loss": 0.4522,
+      "step": 13440
+    },
+    {
+      "epoch": 2.2073984357358407,
+      "grad_norm": 0.37261624908788993,
+      "learning_rate": 5.3997150332620855e-06,
+      "loss": 0.448,
+      "step": 13441
+    },
+    {
+      "epoch": 2.2075626629441834,
+      "grad_norm": 0.3805111962488991,
+      "learning_rate": 5.399255525279405e-06,
+      "loss": 0.449,
+      "step": 13442
+    },
+    {
+      "epoch": 2.207726890152526,
+      "grad_norm": 0.36182810691107475,
+      "learning_rate": 5.398796007529926e-06,
+      "loss": 0.4328,
+      "step": 13443
+    },
+    {
+      "epoch": 2.2078911173608686,
+      "grad_norm": 0.34366558370047795,
+      "learning_rate": 5.398336480018638e-06,
+      "loss": 0.4285,
+      "step": 13444
+    },
+    {
+      "epoch": 2.2080553445692117,
+      "grad_norm": 0.3058709969084702,
+      "learning_rate": 5.397876942750532e-06,
+      "loss": 0.4481,
+      "step": 13445
+    },
+    {
+      "epoch": 2.2082195717775543,
+      "grad_norm": 0.33400468632077346,
+      "learning_rate": 5.397417395730599e-06,
+      "loss": 0.449,
+      "step": 13446
+    },
+    {
+      "epoch": 2.208383798985897,
+      "grad_norm": 0.4026208309464146,
+      "learning_rate": 5.396957838963829e-06,
+      "loss": 0.4412,
+      "step": 13447
+    },
+    {
+      "epoch": 2.2085480261942396,
+      "grad_norm": 0.35216660485768597,
+      "learning_rate": 5.396498272455217e-06,
+      "loss": 0.4534,
+      "step": 13448
+    },
+    {
+      "epoch": 2.2087122534025827,
+      "grad_norm": 0.4944410426286778,
+      "learning_rate": 5.39603869620975e-06,
+      "loss": 0.451,
+      "step": 13449
+    },
+    {
+      "epoch": 2.2088764806109253,
+      "grad_norm": 0.35289336830677026,
+      "learning_rate": 5.395579110232421e-06,
+      "loss": 0.443,
+      "step": 13450
+    },
+    {
+      "epoch": 2.209040707819268,
+      "grad_norm": 0.3493194840023415,
+      "learning_rate": 5.395119514528222e-06,
+      "loss": 0.4473,
+      "step": 13451
+    },
+    {
+      "epoch": 2.2092049350276106,
+      "grad_norm": 0.2907115504572143,
+      "learning_rate": 5.394659909102144e-06,
+      "loss": 0.4417,
+      "step": 13452
+    },
+    {
+      "epoch": 2.2093691622359533,
+      "grad_norm": 0.3184738041292627,
+      "learning_rate": 5.39420029395918e-06,
+      "loss": 0.4483,
+      "step": 13453
+    },
+    {
+      "epoch": 2.2095333894442963,
+      "grad_norm": 0.3785358669834699,
+      "learning_rate": 5.393740669104321e-06,
+      "loss": 0.4491,
+      "step": 13454
+    },
+    {
+      "epoch": 2.209697616652639,
+      "grad_norm": 0.4155139242934717,
+      "learning_rate": 5.393281034542559e-06,
+      "loss": 0.4555,
+      "step": 13455
+    },
+    {
+      "epoch": 2.2098618438609816,
+      "grad_norm": 0.3500073461421561,
+      "learning_rate": 5.392821390278885e-06,
+      "loss": 0.4247,
+      "step": 13456
+    },
+    {
+      "epoch": 2.2100260710693242,
+      "grad_norm": 0.293089681541331,
+      "learning_rate": 5.392361736318293e-06,
+      "loss": 0.4445,
+      "step": 13457
+    },
+    {
+      "epoch": 2.2101902982776673,
+      "grad_norm": 0.37913419132543585,
+      "learning_rate": 5.3919020726657746e-06,
+      "loss": 0.4511,
+      "step": 13458
+    },
+    {
+      "epoch": 2.21035452548601,
+      "grad_norm": 0.3325337362826304,
+      "learning_rate": 5.391442399326323e-06,
+      "loss": 0.4466,
+      "step": 13459
+    },
+    {
+      "epoch": 2.2105187526943526,
+      "grad_norm": 0.30312378625958575,
+      "learning_rate": 5.390982716304928e-06,
+      "loss": 0.4444,
+      "step": 13460
+    },
+    {
+      "epoch": 2.2106829799026952,
+      "grad_norm": 0.45363941132765107,
+      "learning_rate": 5.3905230236065845e-06,
+      "loss": 0.4339,
+      "step": 13461
+    },
+    {
+      "epoch": 2.2108472071110383,
+      "grad_norm": 0.35081016919396296,
+      "learning_rate": 5.390063321236287e-06,
+      "loss": 0.4468,
+      "step": 13462
+    },
+    {
+      "epoch": 2.211011434319381,
+      "grad_norm": 0.4229490886794608,
+      "learning_rate": 5.389603609199025e-06,
+      "loss": 0.4377,
+      "step": 13463
+    },
+    {
+      "epoch": 2.2111756615277236,
+      "grad_norm": 0.5187989672492346,
+      "learning_rate": 5.389143887499791e-06,
+      "loss": 0.4276,
+      "step": 13464
+    },
+    {
+      "epoch": 2.2113398887360662,
+      "grad_norm": 0.349141035977585,
+      "learning_rate": 5.38868415614358e-06,
+      "loss": 0.4585,
+      "step": 13465
+    },
+    {
+      "epoch": 2.2115041159444093,
+      "grad_norm": 0.38330880903839587,
+      "learning_rate": 5.388224415135387e-06,
+      "loss": 0.4477,
+      "step": 13466
+    },
+    {
+      "epoch": 2.211668343152752,
+      "grad_norm": 0.3643086489980097,
+      "learning_rate": 5.387764664480201e-06,
+      "loss": 0.435,
+      "step": 13467
+    },
+    {
+      "epoch": 2.2118325703610946,
+      "grad_norm": 1.2701962885891152,
+      "learning_rate": 5.387304904183018e-06,
+      "loss": 0.4448,
+      "step": 13468
+    },
+    {
+      "epoch": 2.2119967975694372,
+      "grad_norm": 0.30489125033902276,
+      "learning_rate": 5.38684513424883e-06,
+      "loss": 0.4342,
+      "step": 13469
+    },
+    {
+      "epoch": 2.21216102477778,
+      "grad_norm": 0.31767007262386243,
+      "learning_rate": 5.386385354682632e-06,
+      "loss": 0.4261,
+      "step": 13470
+    },
+    {
+      "epoch": 2.212325251986123,
+      "grad_norm": 0.4085825381441104,
+      "learning_rate": 5.385925565489416e-06,
+      "loss": 0.4199,
+      "step": 13471
+    },
+    {
+      "epoch": 2.2124894791944656,
+      "grad_norm": 0.33488337192858,
+      "learning_rate": 5.385465766674178e-06,
+      "loss": 0.4192,
+      "step": 13472
+    },
+    {
+      "epoch": 2.2126537064028082,
+      "grad_norm": 0.3389412489135619,
+      "learning_rate": 5.385005958241911e-06,
+      "loss": 0.441,
+      "step": 13473
+    },
+    {
+      "epoch": 2.212817933611151,
+      "grad_norm": 0.3722364049933887,
+      "learning_rate": 5.384546140197607e-06,
+      "loss": 0.4543,
+      "step": 13474
+    },
+    {
+      "epoch": 2.212982160819494,
+      "grad_norm": 0.38592443024169204,
+      "learning_rate": 5.384086312546262e-06,
+      "loss": 0.4365,
+      "step": 13475
+    },
+    {
+      "epoch": 2.2131463880278366,
+      "grad_norm": 0.3146545371530525,
+      "learning_rate": 5.383626475292869e-06,
+      "loss": 0.4475,
+      "step": 13476
+    },
+    {
+      "epoch": 2.2133106152361792,
+      "grad_norm": 0.27347639969455967,
+      "learning_rate": 5.383166628442426e-06,
+      "loss": 0.4403,
+      "step": 13477
+    },
+    {
+      "epoch": 2.213474842444522,
+      "grad_norm": 0.3374210661622069,
+      "learning_rate": 5.382706771999923e-06,
+      "loss": 0.4394,
+      "step": 13478
+    },
+    {
+      "epoch": 2.213639069652865,
+      "grad_norm": 0.45448892335658087,
+      "learning_rate": 5.382246905970355e-06,
+      "loss": 0.4253,
+      "step": 13479
+    },
+    {
+      "epoch": 2.2138032968612076,
+      "grad_norm": 0.28749697984270106,
+      "learning_rate": 5.381787030358718e-06,
+      "loss": 0.4429,
+      "step": 13480
+    },
+    {
+      "epoch": 2.21396752406955,
+      "grad_norm": 0.36495583828274214,
+      "learning_rate": 5.3813271451700064e-06,
+      "loss": 0.4351,
+      "step": 13481
+    },
+    {
+      "epoch": 2.214131751277893,
+      "grad_norm": 0.4129797086459647,
+      "learning_rate": 5.380867250409214e-06,
+      "loss": 0.4359,
+      "step": 13482
+    },
+    {
+      "epoch": 2.214295978486236,
+      "grad_norm": 0.3320847636462863,
+      "learning_rate": 5.380407346081338e-06,
+      "loss": 0.4417,
+      "step": 13483
+    },
+    {
+      "epoch": 2.2144602056945786,
+      "grad_norm": 0.33595000981170686,
+      "learning_rate": 5.379947432191372e-06,
+      "loss": 0.4466,
+      "step": 13484
+    },
+    {
+      "epoch": 2.214624432902921,
+      "grad_norm": 0.3459516900532747,
+      "learning_rate": 5.379487508744311e-06,
+      "loss": 0.4249,
+      "step": 13485
+    },
+    {
+      "epoch": 2.214788660111264,
+      "grad_norm": 0.289936742009504,
+      "learning_rate": 5.3790275757451496e-06,
+      "loss": 0.4356,
+      "step": 13486
+    },
+    {
+      "epoch": 2.2149528873196065,
+      "grad_norm": 0.33670869144840854,
+      "learning_rate": 5.378567633198885e-06,
+      "loss": 0.4436,
+      "step": 13487
+    },
+    {
+      "epoch": 2.2151171145279496,
+      "grad_norm": 0.36515225989280037,
+      "learning_rate": 5.378107681110511e-06,
+      "loss": 0.4227,
+      "step": 13488
+    },
+    {
+      "epoch": 2.215281341736292,
+      "grad_norm": 0.3082061760174833,
+      "learning_rate": 5.377647719485024e-06,
+      "loss": 0.4461,
+      "step": 13489
+    },
+    {
+      "epoch": 2.215445568944635,
+      "grad_norm": 0.3461068541761683,
+      "learning_rate": 5.377187748327418e-06,
+      "loss": 0.4383,
+      "step": 13490
+    },
+    {
+      "epoch": 2.2156097961529775,
+      "grad_norm": 0.3146220360337419,
+      "learning_rate": 5.3767277676426915e-06,
+      "loss": 0.4579,
+      "step": 13491
+    },
+    {
+      "epoch": 2.2157740233613206,
+      "grad_norm": 0.2933583327660923,
+      "learning_rate": 5.376267777435838e-06,
+      "loss": 0.4662,
+      "step": 13492
+    },
+    {
+      "epoch": 2.215938250569663,
+      "grad_norm": 0.36036615968876007,
+      "learning_rate": 5.375807777711855e-06,
+      "loss": 0.4282,
+      "step": 13493
+    },
+    {
+      "epoch": 2.216102477778006,
+      "grad_norm": 0.3082561878071788,
+      "learning_rate": 5.3753477684757366e-06,
+      "loss": 0.446,
+      "step": 13494
+    },
+    {
+      "epoch": 2.2162667049863485,
+      "grad_norm": 0.2860483461505643,
+      "learning_rate": 5.374887749732482e-06,
+      "loss": 0.4432,
+      "step": 13495
+    },
+    {
+      "epoch": 2.2164309321946916,
+      "grad_norm": 0.30065915277707855,
+      "learning_rate": 5.3744277214870836e-06,
+      "loss": 0.4567,
+      "step": 13496
+    },
+    {
+      "epoch": 2.216595159403034,
+      "grad_norm": 0.28339697722762697,
+      "learning_rate": 5.3739676837445414e-06,
+      "loss": 0.4367,
+      "step": 13497
+    },
+    {
+      "epoch": 2.216759386611377,
+      "grad_norm": 0.374467941980816,
+      "learning_rate": 5.37350763650985e-06,
+      "loss": 0.4438,
+      "step": 13498
+    },
+    {
+      "epoch": 2.2169236138197195,
+      "grad_norm": 0.35613700649582364,
+      "learning_rate": 5.3730475797880066e-06,
+      "loss": 0.4382,
+      "step": 13499
+    },
+    {
+      "epoch": 2.2170878410280626,
+      "grad_norm": 0.40460989848519385,
+      "learning_rate": 5.372587513584006e-06,
+      "loss": 0.4322,
+      "step": 13500
+    },
+    {
+      "epoch": 2.217252068236405,
+      "grad_norm": 0.276120695564519,
+      "learning_rate": 5.372127437902847e-06,
+      "loss": 0.4412,
+      "step": 13501
+    },
+    {
+      "epoch": 2.217416295444748,
+      "grad_norm": 0.2981034665655008,
+      "learning_rate": 5.3716673527495275e-06,
+      "loss": 0.4262,
+      "step": 13502
+    },
+    {
+      "epoch": 2.2175805226530905,
+      "grad_norm": 0.3595831031378883,
+      "learning_rate": 5.371207258129041e-06,
+      "loss": 0.4305,
+      "step": 13503
+    },
+    {
+      "epoch": 2.217744749861433,
+      "grad_norm": 0.35041234040561664,
+      "learning_rate": 5.3707471540463884e-06,
+      "loss": 0.457,
+      "step": 13504
+    },
+    {
+      "epoch": 2.217908977069776,
+      "grad_norm": 0.3376497856717965,
+      "learning_rate": 5.370287040506563e-06,
+      "loss": 0.4454,
+      "step": 13505
+    },
+    {
+      "epoch": 2.218073204278119,
+      "grad_norm": 0.3492901426130223,
+      "learning_rate": 5.369826917514565e-06,
+      "loss": 0.4239,
+      "step": 13506
+    },
+    {
+      "epoch": 2.2182374314864615,
+      "grad_norm": 0.3528616323045615,
+      "learning_rate": 5.369366785075392e-06,
+      "loss": 0.4418,
+      "step": 13507
+    },
+    {
+      "epoch": 2.218401658694804,
+      "grad_norm": 0.5411887062016001,
+      "learning_rate": 5.368906643194039e-06,
+      "loss": 0.4467,
+      "step": 13508
+    },
+    {
+      "epoch": 2.218565885903147,
+      "grad_norm": 0.281710369398488,
+      "learning_rate": 5.3684464918755054e-06,
+      "loss": 0.4289,
+      "step": 13509
+    },
+    {
+      "epoch": 2.21873011311149,
+      "grad_norm": 0.31953154742558165,
+      "learning_rate": 5.367986331124787e-06,
+      "loss": 0.4363,
+      "step": 13510
+    },
+    {
+      "epoch": 2.2188943403198325,
+      "grad_norm": 0.27411536805995634,
+      "learning_rate": 5.367526160946885e-06,
+      "loss": 0.4382,
+      "step": 13511
+    },
+    {
+      "epoch": 2.219058567528175,
+      "grad_norm": 0.28347580644466774,
+      "learning_rate": 5.367065981346796e-06,
+      "loss": 0.4318,
+      "step": 13512
+    },
+    {
+      "epoch": 2.219222794736518,
+      "grad_norm": 0.35565024009430696,
+      "learning_rate": 5.366605792329516e-06,
+      "loss": 0.4408,
+      "step": 13513
+    },
+    {
+      "epoch": 2.219387021944861,
+      "grad_norm": 0.3173430366533922,
+      "learning_rate": 5.366145593900044e-06,
+      "loss": 0.4578,
+      "step": 13514
+    },
+    {
+      "epoch": 2.2195512491532035,
+      "grad_norm": 0.35393030124900166,
+      "learning_rate": 5.365685386063378e-06,
+      "loss": 0.4452,
+      "step": 13515
+    },
+    {
+      "epoch": 2.219715476361546,
+      "grad_norm": 0.27084181578202887,
+      "learning_rate": 5.365225168824519e-06,
+      "loss": 0.4474,
+      "step": 13516
+    },
+    {
+      "epoch": 2.219879703569889,
+      "grad_norm": 0.6068977121317608,
+      "learning_rate": 5.364764942188463e-06,
+      "loss": 0.4433,
+      "step": 13517
+    },
+    {
+      "epoch": 2.220043930778232,
+      "grad_norm": 0.2827573163324981,
+      "learning_rate": 5.364304706160208e-06,
+      "loss": 0.4488,
+      "step": 13518
+    },
+    {
+      "epoch": 2.2202081579865744,
+      "grad_norm": 0.3787643336445636,
+      "learning_rate": 5.363844460744755e-06,
+      "loss": 0.4316,
+      "step": 13519
+    },
+    {
+      "epoch": 2.220372385194917,
+      "grad_norm": 0.30874648437768004,
+      "learning_rate": 5.3633842059471e-06,
+      "loss": 0.4378,
+      "step": 13520
+    },
+    {
+      "epoch": 2.2205366124032597,
+      "grad_norm": 0.336792254197232,
+      "learning_rate": 5.3629239417722444e-06,
+      "loss": 0.4524,
+      "step": 13521
+    },
+    {
+      "epoch": 2.220700839611603,
+      "grad_norm": 0.3227196910096211,
+      "learning_rate": 5.362463668225184e-06,
+      "loss": 0.4412,
+      "step": 13522
+    },
+    {
+      "epoch": 2.2208650668199454,
+      "grad_norm": 0.32507962705134064,
+      "learning_rate": 5.362003385310921e-06,
+      "loss": 0.4536,
+      "step": 13523
+    },
+    {
+      "epoch": 2.221029294028288,
+      "grad_norm": 0.2788200062922952,
+      "learning_rate": 5.361543093034452e-06,
+      "loss": 0.4432,
+      "step": 13524
+    },
+    {
+      "epoch": 2.2211935212366307,
+      "grad_norm": 0.31509433182889934,
+      "learning_rate": 5.361082791400778e-06,
+      "loss": 0.4341,
+      "step": 13525
+    },
+    {
+      "epoch": 2.221357748444974,
+      "grad_norm": 0.3591196718424824,
+      "learning_rate": 5.360622480414898e-06,
+      "loss": 0.4329,
+      "step": 13526
+    },
+    {
+      "epoch": 2.2215219756533164,
+      "grad_norm": 0.30416959370332597,
+      "learning_rate": 5.360162160081811e-06,
+      "loss": 0.4394,
+      "step": 13527
+    },
+    {
+      "epoch": 2.221686202861659,
+      "grad_norm": 0.6191722097673911,
+      "learning_rate": 5.359701830406516e-06,
+      "loss": 0.4481,
+      "step": 13528
+    },
+    {
+      "epoch": 2.2218504300700017,
+      "grad_norm": 0.7503304964039875,
+      "learning_rate": 5.359241491394014e-06,
+      "loss": 0.4292,
+      "step": 13529
+    },
+    {
+      "epoch": 2.222014657278345,
+      "grad_norm": 0.42825504406683945,
+      "learning_rate": 5.358781143049304e-06,
+      "loss": 0.4342,
+      "step": 13530
+    },
+    {
+      "epoch": 2.2221788844866874,
+      "grad_norm": 0.29088118407198305,
+      "learning_rate": 5.358320785377386e-06,
+      "loss": 0.4375,
+      "step": 13531
+    },
+    {
+      "epoch": 2.22234311169503,
+      "grad_norm": 0.33448670629242305,
+      "learning_rate": 5.357860418383259e-06,
+      "loss": 0.432,
+      "step": 13532
+    },
+    {
+      "epoch": 2.2225073389033727,
+      "grad_norm": 0.28810069927283627,
+      "learning_rate": 5.357400042071925e-06,
+      "loss": 0.4489,
+      "step": 13533
+    },
+    {
+      "epoch": 2.222671566111716,
+      "grad_norm": 0.39361019961728927,
+      "learning_rate": 5.356939656448381e-06,
+      "loss": 0.4466,
+      "step": 13534
+    },
+    {
+      "epoch": 2.2228357933200584,
+      "grad_norm": 0.3077810757273589,
+      "learning_rate": 5.356479261517631e-06,
+      "loss": 0.4344,
+      "step": 13535
+    },
+    {
+      "epoch": 2.223000020528401,
+      "grad_norm": 0.44608262878290833,
+      "learning_rate": 5.356018857284672e-06,
+      "loss": 0.4395,
+      "step": 13536
+    },
+    {
+      "epoch": 2.2231642477367437,
+      "grad_norm": 0.45678308542254015,
+      "learning_rate": 5.355558443754508e-06,
+      "loss": 0.4452,
+      "step": 13537
+    },
+    {
+      "epoch": 2.2233284749450863,
+      "grad_norm": 0.3910797651862828,
+      "learning_rate": 5.355098020932136e-06,
+      "loss": 0.4339,
+      "step": 13538
+    },
+    {
+      "epoch": 2.2234927021534294,
+      "grad_norm": 1.0072820509471674,
+      "learning_rate": 5.354637588822559e-06,
+      "loss": 0.4631,
+      "step": 13539
+    },
+    {
+      "epoch": 2.223656929361772,
+      "grad_norm": 0.3495902620317632,
+      "learning_rate": 5.354177147430777e-06,
+      "loss": 0.4667,
+      "step": 13540
+    },
+    {
+      "epoch": 2.2238211565701147,
+      "grad_norm": 0.35133521693691183,
+      "learning_rate": 5.353716696761791e-06,
+      "loss": 0.4499,
+      "step": 13541
+    },
+    {
+      "epoch": 2.2239853837784573,
+      "grad_norm": 0.40684209571486746,
+      "learning_rate": 5.3532562368206006e-06,
+      "loss": 0.4511,
+      "step": 13542
+    },
+    {
+      "epoch": 2.2241496109868004,
+      "grad_norm": 0.652470986013063,
+      "learning_rate": 5.3527957676122085e-06,
+      "loss": 0.451,
+      "step": 13543
+    },
+    {
+      "epoch": 2.224313838195143,
+      "grad_norm": 0.32350572923745236,
+      "learning_rate": 5.352335289141614e-06,
+      "loss": 0.4437,
+      "step": 13544
+    },
+    {
+      "epoch": 2.2244780654034857,
+      "grad_norm": 0.28213502193799567,
+      "learning_rate": 5.351874801413822e-06,
+      "loss": 0.4281,
+      "step": 13545
+    },
+    {
+      "epoch": 2.2246422926118283,
+      "grad_norm": 0.34122318350658276,
+      "learning_rate": 5.351414304433832e-06,
+      "loss": 0.4425,
+      "step": 13546
+    },
+    {
+      "epoch": 2.2248065198201714,
+      "grad_norm": 0.34055176377424634,
+      "learning_rate": 5.350953798206643e-06,
+      "loss": 0.4641,
+      "step": 13547
+    },
+    {
+      "epoch": 2.224970747028514,
+      "grad_norm": 0.3056182901324451,
+      "learning_rate": 5.350493282737258e-06,
+      "loss": 0.4472,
+      "step": 13548
+    },
+    {
+      "epoch": 2.2251349742368567,
+      "grad_norm": 0.4295171215648557,
+      "learning_rate": 5.3500327580306804e-06,
+      "loss": 0.4411,
+      "step": 13549
+    },
+    {
+      "epoch": 2.2252992014451993,
+      "grad_norm": 0.31663958292820704,
+      "learning_rate": 5.349572224091911e-06,
+      "loss": 0.4465,
+      "step": 13550
+    },
+    {
+      "epoch": 2.2254634286535424,
+      "grad_norm": 0.3231332831732216,
+      "learning_rate": 5.34911168092595e-06,
+      "loss": 0.4488,
+      "step": 13551
+    },
+    {
+      "epoch": 2.225627655861885,
+      "grad_norm": 0.3500335217632181,
+      "learning_rate": 5.348651128537802e-06,
+      "loss": 0.4341,
+      "step": 13552
+    },
+    {
+      "epoch": 2.2257918830702277,
+      "grad_norm": 0.282708868398423,
+      "learning_rate": 5.348190566932467e-06,
+      "loss": 0.4248,
+      "step": 13553
+    },
+    {
+      "epoch": 2.2259561102785703,
+      "grad_norm": 0.3214462040558992,
+      "learning_rate": 5.347729996114949e-06,
+      "loss": 0.4301,
+      "step": 13554
+    },
+    {
+      "epoch": 2.226120337486913,
+      "grad_norm": 0.31372033167420005,
+      "learning_rate": 5.347269416090249e-06,
+      "loss": 0.4259,
+      "step": 13555
+    },
+    {
+      "epoch": 2.226284564695256,
+      "grad_norm": 0.3327297401900336,
+      "learning_rate": 5.3468088268633695e-06,
+      "loss": 0.4441,
+      "step": 13556
+    },
+    {
+      "epoch": 2.2264487919035987,
+      "grad_norm": 0.3294442381478894,
+      "learning_rate": 5.346348228439312e-06,
+      "loss": 0.4471,
+      "step": 13557
+    },
+    {
+      "epoch": 2.2266130191119413,
+      "grad_norm": 0.3929457500037696,
+      "learning_rate": 5.345887620823081e-06,
+      "loss": 0.446,
+      "step": 13558
+    },
+    {
+      "epoch": 2.226777246320284,
+      "grad_norm": 0.3631153626815294,
+      "learning_rate": 5.345427004019677e-06,
+      "loss": 0.4353,
+      "step": 13559
+    },
+    {
+      "epoch": 2.226941473528627,
+      "grad_norm": 0.2869505335422441,
+      "learning_rate": 5.344966378034106e-06,
+      "loss": 0.4031,
+      "step": 13560
+    },
+    {
+      "epoch": 2.2271057007369697,
+      "grad_norm": 1.0574105052879286,
+      "learning_rate": 5.3445057428713675e-06,
+      "loss": 0.4474,
+      "step": 13561
+    },
+    {
+      "epoch": 2.2272699279453123,
+      "grad_norm": 0.3579479567718406,
+      "learning_rate": 5.344045098536466e-06,
+      "loss": 0.4628,
+      "step": 13562
+    },
+    {
+      "epoch": 2.227434155153655,
+      "grad_norm": 0.3108652105716026,
+      "learning_rate": 5.343584445034403e-06,
+      "loss": 0.4389,
+      "step": 13563
+    },
+    {
+      "epoch": 2.227598382361998,
+      "grad_norm": 0.31838087198749193,
+      "learning_rate": 5.343123782370185e-06,
+      "loss": 0.4672,
+      "step": 13564
+    },
+    {
+      "epoch": 2.2277626095703407,
+      "grad_norm": 0.348928364328581,
+      "learning_rate": 5.342663110548813e-06,
+      "loss": 0.4227,
+      "step": 13565
+    },
+    {
+      "epoch": 2.2279268367786833,
+      "grad_norm": 0.375223157962436,
+      "learning_rate": 5.34220242957529e-06,
+      "loss": 0.4516,
+      "step": 13566
+    },
+    {
+      "epoch": 2.228091063987026,
+      "grad_norm": 0.36572646774424566,
+      "learning_rate": 5.3417417394546195e-06,
+      "loss": 0.4651,
+      "step": 13567
+    },
+    {
+      "epoch": 2.228255291195369,
+      "grad_norm": 0.3584240750256261,
+      "learning_rate": 5.341281040191805e-06,
+      "loss": 0.4305,
+      "step": 13568
+    },
+    {
+      "epoch": 2.2284195184037117,
+      "grad_norm": 0.3472023415751713,
+      "learning_rate": 5.340820331791852e-06,
+      "loss": 0.4582,
+      "step": 13569
+    },
+    {
+      "epoch": 2.2285837456120543,
+      "grad_norm": 0.36169487795111643,
+      "learning_rate": 5.340359614259764e-06,
+      "loss": 0.4476,
+      "step": 13570
+    },
+    {
+      "epoch": 2.228747972820397,
+      "grad_norm": 0.4990169623639773,
+      "learning_rate": 5.3398988876005415e-06,
+      "loss": 0.448,
+      "step": 13571
+    },
+    {
+      "epoch": 2.2289122000287396,
+      "grad_norm": 0.3237664127605256,
+      "learning_rate": 5.339438151819192e-06,
+      "loss": 0.44,
+      "step": 13572
+    },
+    {
+      "epoch": 2.2290764272370827,
+      "grad_norm": 0.3970367551739623,
+      "learning_rate": 5.338977406920717e-06,
+      "loss": 0.434,
+      "step": 13573
+    },
+    {
+      "epoch": 2.2292406544454253,
+      "grad_norm": 0.35942224426552793,
+      "learning_rate": 5.338516652910123e-06,
+      "loss": 0.4468,
+      "step": 13574
+    },
+    {
+      "epoch": 2.229404881653768,
+      "grad_norm": 0.27415087723762693,
+      "learning_rate": 5.338055889792414e-06,
+      "loss": 0.4583,
+      "step": 13575
+    },
+    {
+      "epoch": 2.2295691088621106,
+      "grad_norm": 0.30038302397071287,
+      "learning_rate": 5.337595117572591e-06,
+      "loss": 0.4448,
+      "step": 13576
+    },
+    {
+      "epoch": 2.2297333360704537,
+      "grad_norm": 0.6853846292788964,
+      "learning_rate": 5.337134336255663e-06,
+      "loss": 0.4408,
+      "step": 13577
+    },
+    {
+      "epoch": 2.2298975632787963,
+      "grad_norm": 0.5243657367788079,
+      "learning_rate": 5.336673545846631e-06,
+      "loss": 0.4418,
+      "step": 13578
+    },
+    {
+      "epoch": 2.230061790487139,
+      "grad_norm": 0.3239461966116936,
+      "learning_rate": 5.336212746350502e-06,
+      "loss": 0.4591,
+      "step": 13579
+    },
+    {
+      "epoch": 2.2302260176954816,
+      "grad_norm": 0.380531154323319,
+      "learning_rate": 5.33575193777228e-06,
+      "loss": 0.4608,
+      "step": 13580
+    },
+    {
+      "epoch": 2.2303902449038246,
+      "grad_norm": 0.30471466475214776,
+      "learning_rate": 5.33529112011697e-06,
+      "loss": 0.4535,
+      "step": 13581
+    },
+    {
+      "epoch": 2.2305544721121673,
+      "grad_norm": 0.42057761873332117,
+      "learning_rate": 5.334830293389576e-06,
+      "loss": 0.4575,
+      "step": 13582
+    },
+    {
+      "epoch": 2.23071869932051,
+      "grad_norm": 0.30684393851680114,
+      "learning_rate": 5.334369457595104e-06,
+      "loss": 0.4538,
+      "step": 13583
+    },
+    {
+      "epoch": 2.2308829265288526,
+      "grad_norm": 0.4505269932130537,
+      "learning_rate": 5.333908612738558e-06,
+      "loss": 0.4441,
+      "step": 13584
+    },
+    {
+      "epoch": 2.2310471537371956,
+      "grad_norm": 0.3104451390730639,
+      "learning_rate": 5.333447758824945e-06,
+      "loss": 0.4415,
+      "step": 13585
+    },
+    {
+      "epoch": 2.2312113809455383,
+      "grad_norm": 0.3042509785408218,
+      "learning_rate": 5.332986895859269e-06,
+      "loss": 0.4442,
+      "step": 13586
+    },
+    {
+      "epoch": 2.231375608153881,
+      "grad_norm": 0.33087608068349694,
+      "learning_rate": 5.332526023846536e-06,
+      "loss": 0.4431,
+      "step": 13587
+    },
+    {
+      "epoch": 2.2315398353622236,
+      "grad_norm": 0.3253474702931535,
+      "learning_rate": 5.332065142791751e-06,
+      "loss": 0.4242,
+      "step": 13588
+    },
+    {
+      "epoch": 2.231704062570566,
+      "grad_norm": 0.3869258011453058,
+      "learning_rate": 5.33160425269992e-06,
+      "loss": 0.4458,
+      "step": 13589
+    },
+    {
+      "epoch": 2.2318682897789093,
+      "grad_norm": 0.4440672206002762,
+      "learning_rate": 5.331143353576048e-06,
+      "loss": 0.4346,
+      "step": 13590
+    },
+    {
+      "epoch": 2.232032516987252,
+      "grad_norm": 0.36785815536200994,
+      "learning_rate": 5.330682445425143e-06,
+      "loss": 0.4452,
+      "step": 13591
+    },
+    {
+      "epoch": 2.2321967441955946,
+      "grad_norm": 0.44275564494307584,
+      "learning_rate": 5.330221528252207e-06,
+      "loss": 0.4502,
+      "step": 13592
+    },
+    {
+      "epoch": 2.232360971403937,
+      "grad_norm": 0.343543553199038,
+      "learning_rate": 5.3297606020622495e-06,
+      "loss": 0.4496,
+      "step": 13593
+    },
+    {
+      "epoch": 2.2325251986122803,
+      "grad_norm": 0.37180636084536056,
+      "learning_rate": 5.3292996668602765e-06,
+      "loss": 0.4493,
+      "step": 13594
+    },
+    {
+      "epoch": 2.232689425820623,
+      "grad_norm": 0.29019902675559567,
+      "learning_rate": 5.328838722651292e-06,
+      "loss": 0.441,
+      "step": 13595
+    },
+    {
+      "epoch": 2.2328536530289655,
+      "grad_norm": 0.45008975878901003,
+      "learning_rate": 5.3283777694403045e-06,
+      "loss": 0.4499,
+      "step": 13596
+    },
+    {
+      "epoch": 2.233017880237308,
+      "grad_norm": 0.34751380757528144,
+      "learning_rate": 5.327916807232318e-06,
+      "loss": 0.4276,
+      "step": 13597
+    },
+    {
+      "epoch": 2.2331821074456513,
+      "grad_norm": 0.4329884068791182,
+      "learning_rate": 5.32745583603234e-06,
+      "loss": 0.4444,
+      "step": 13598
+    },
+    {
+      "epoch": 2.233346334653994,
+      "grad_norm": 0.34680911321201413,
+      "learning_rate": 5.32699485584538e-06,
+      "loss": 0.4257,
+      "step": 13599
+    },
+    {
+      "epoch": 2.2335105618623365,
+      "grad_norm": 0.4622230443995473,
+      "learning_rate": 5.326533866676443e-06,
+      "loss": 0.4285,
+      "step": 13600
+    },
+    {
+      "epoch": 2.233674789070679,
+      "grad_norm": 0.5131870445694399,
+      "learning_rate": 5.3260728685305314e-06,
+      "loss": 0.4491,
+      "step": 13601
+    },
+    {
+      "epoch": 2.2338390162790223,
+      "grad_norm": 0.40076083507511995,
+      "learning_rate": 5.325611861412656e-06,
+      "loss": 0.4386,
+      "step": 13602
+    },
+    {
+      "epoch": 2.234003243487365,
+      "grad_norm": 0.27194120329673643,
+      "learning_rate": 5.325150845327826e-06,
+      "loss": 0.4402,
+      "step": 13603
+    },
+    {
+      "epoch": 2.2341674706957075,
+      "grad_norm": 0.3627492697365076,
+      "learning_rate": 5.3246898202810455e-06,
+      "loss": 0.4561,
+      "step": 13604
+    },
+    {
+      "epoch": 2.23433169790405,
+      "grad_norm": 0.36811125848416526,
+      "learning_rate": 5.324228786277321e-06,
+      "loss": 0.4442,
+      "step": 13605
+    },
+    {
+      "epoch": 2.234495925112393,
+      "grad_norm": 0.32740568146167454,
+      "learning_rate": 5.323767743321661e-06,
+      "loss": 0.4416,
+      "step": 13606
+    },
+    {
+      "epoch": 2.234660152320736,
+      "grad_norm": 0.2928833678123084,
+      "learning_rate": 5.323306691419072e-06,
+      "loss": 0.4526,
+      "step": 13607
+    },
+    {
+      "epoch": 2.2348243795290785,
+      "grad_norm": 0.5940056253711327,
+      "learning_rate": 5.322845630574564e-06,
+      "loss": 0.4452,
+      "step": 13608
+    },
+    {
+      "epoch": 2.234988606737421,
+      "grad_norm": 0.42021729424037163,
+      "learning_rate": 5.322384560793143e-06,
+      "loss": 0.4168,
+      "step": 13609
+    },
+    {
+      "epoch": 2.235152833945764,
+      "grad_norm": 0.3073837421956184,
+      "learning_rate": 5.321923482079815e-06,
+      "loss": 0.4486,
+      "step": 13610
+    },
+    {
+      "epoch": 2.235317061154107,
+      "grad_norm": 0.3522817208717473,
+      "learning_rate": 5.321462394439591e-06,
+      "loss": 0.4284,
+      "step": 13611
+    },
+    {
+      "epoch": 2.2354812883624495,
+      "grad_norm": 0.3598916939940968,
+      "learning_rate": 5.3210012978774765e-06,
+      "loss": 0.4464,
+      "step": 13612
+    },
+    {
+      "epoch": 2.235645515570792,
+      "grad_norm": 0.2617858004173523,
+      "learning_rate": 5.320540192398479e-06,
+      "loss": 0.4511,
+      "step": 13613
+    },
+    {
+      "epoch": 2.235809742779135,
+      "grad_norm": 0.31940286486217156,
+      "learning_rate": 5.3200790780076096e-06,
+      "loss": 0.4472,
+      "step": 13614
+    },
+    {
+      "epoch": 2.235973969987478,
+      "grad_norm": 0.2959868358849483,
+      "learning_rate": 5.319617954709873e-06,
+      "loss": 0.4249,
+      "step": 13615
+    },
+    {
+      "epoch": 2.2361381971958205,
+      "grad_norm": 0.3030407022951804,
+      "learning_rate": 5.319156822510281e-06,
+      "loss": 0.448,
+      "step": 13616
+    },
+    {
+      "epoch": 2.236302424404163,
+      "grad_norm": 0.34246789595360566,
+      "learning_rate": 5.318695681413839e-06,
+      "loss": 0.4412,
+      "step": 13617
+    },
+    {
+      "epoch": 2.236466651612506,
+      "grad_norm": 0.3426164905586456,
+      "learning_rate": 5.318234531425555e-06,
+      "loss": 0.4696,
+      "step": 13618
+    },
+    {
+      "epoch": 2.236630878820849,
+      "grad_norm": 0.32189875802648454,
+      "learning_rate": 5.317773372550441e-06,
+      "loss": 0.4516,
+      "step": 13619
+    },
+    {
+      "epoch": 2.2367951060291915,
+      "grad_norm": 0.29844986670895207,
+      "learning_rate": 5.317312204793502e-06,
+      "loss": 0.4394,
+      "step": 13620
+    },
+    {
+      "epoch": 2.236959333237534,
+      "grad_norm": 0.30412717030286257,
+      "learning_rate": 5.316851028159749e-06,
+      "loss": 0.4352,
+      "step": 13621
+    },
+    {
+      "epoch": 2.237123560445877,
+      "grad_norm": 0.3608272137967386,
+      "learning_rate": 5.3163898426541905e-06,
+      "loss": 0.4342,
+      "step": 13622
+    },
+    {
+      "epoch": 2.2372877876542194,
+      "grad_norm": 0.2920497088883626,
+      "learning_rate": 5.315928648281834e-06,
+      "loss": 0.445,
+      "step": 13623
+    },
+    {
+      "epoch": 2.2374520148625625,
+      "grad_norm": 0.3006645750815423,
+      "learning_rate": 5.315467445047691e-06,
+      "loss": 0.4137,
+      "step": 13624
+    },
+    {
+      "epoch": 2.237616242070905,
+      "grad_norm": 0.28330615816664806,
+      "learning_rate": 5.315006232956768e-06,
+      "loss": 0.4441,
+      "step": 13625
+    },
+    {
+      "epoch": 2.237780469279248,
+      "grad_norm": 0.4482961606066074,
+      "learning_rate": 5.314545012014077e-06,
+      "loss": 0.4477,
+      "step": 13626
+    },
+    {
+      "epoch": 2.2379446964875904,
+      "grad_norm": 0.5955992178112777,
+      "learning_rate": 5.314083782224624e-06,
+      "loss": 0.4501,
+      "step": 13627
+    },
+    {
+      "epoch": 2.2381089236959335,
+      "grad_norm": 0.33806720327208106,
+      "learning_rate": 5.3136225435934215e-06,
+      "loss": 0.449,
+      "step": 13628
+    },
+    {
+      "epoch": 2.238273150904276,
+      "grad_norm": 0.3012297840190392,
+      "learning_rate": 5.313161296125477e-06,
+      "loss": 0.4313,
+      "step": 13629
+    },
+    {
+      "epoch": 2.238437378112619,
+      "grad_norm": 0.34520611961056774,
+      "learning_rate": 5.312700039825801e-06,
+      "loss": 0.4413,
+      "step": 13630
+    },
+    {
+      "epoch": 2.2386016053209614,
+      "grad_norm": 0.3207154044252607,
+      "learning_rate": 5.312238774699403e-06,
+      "loss": 0.4581,
+      "step": 13631
+    },
+    {
+      "epoch": 2.2387658325293045,
+      "grad_norm": 0.3156908540345743,
+      "learning_rate": 5.311777500751293e-06,
+      "loss": 0.4465,
+      "step": 13632
+    },
+    {
+      "epoch": 2.238930059737647,
+      "grad_norm": 0.5587797292971918,
+      "learning_rate": 5.31131621798648e-06,
+      "loss": 0.4327,
+      "step": 13633
+    },
+    {
+      "epoch": 2.2390942869459898,
+      "grad_norm": 0.3836716634336738,
+      "learning_rate": 5.310854926409975e-06,
+      "loss": 0.4342,
+      "step": 13634
+    },
+    {
+      "epoch": 2.2392585141543324,
+      "grad_norm": 0.39705515751436415,
+      "learning_rate": 5.310393626026789e-06,
+      "loss": 0.4496,
+      "step": 13635
+    },
+    {
+      "epoch": 2.2394227413626755,
+      "grad_norm": 0.4116210776291705,
+      "learning_rate": 5.3099323168419276e-06,
+      "loss": 0.4644,
+      "step": 13636
+    },
+    {
+      "epoch": 2.239586968571018,
+      "grad_norm": 0.48621166225056617,
+      "learning_rate": 5.309470998860407e-06,
+      "loss": 0.428,
+      "step": 13637
+    },
+    {
+      "epoch": 2.2397511957793608,
+      "grad_norm": 0.29530363881665117,
+      "learning_rate": 5.309009672087236e-06,
+      "loss": 0.4504,
+      "step": 13638
+    },
+    {
+      "epoch": 2.2399154229877034,
+      "grad_norm": 0.3208941816493528,
+      "learning_rate": 5.308548336527421e-06,
+      "loss": 0.4145,
+      "step": 13639
+    },
+    {
+      "epoch": 2.240079650196046,
+      "grad_norm": 0.34304395384197434,
+      "learning_rate": 5.3080869921859765e-06,
+      "loss": 0.4554,
+      "step": 13640
+    },
+    {
+      "epoch": 2.240243877404389,
+      "grad_norm": 0.2997489296447226,
+      "learning_rate": 5.3076256390679136e-06,
+      "loss": 0.4306,
+      "step": 13641
+    },
+    {
+      "epoch": 2.2404081046127318,
+      "grad_norm": 0.3598332489608276,
+      "learning_rate": 5.30716427717824e-06,
+      "loss": 0.4744,
+      "step": 13642
+    },
+    {
+      "epoch": 2.2405723318210744,
+      "grad_norm": 0.3640856566145552,
+      "learning_rate": 5.306702906521969e-06,
+      "loss": 0.4426,
+      "step": 13643
+    },
+    {
+      "epoch": 2.240736559029417,
+      "grad_norm": 0.25957988968195783,
+      "learning_rate": 5.30624152710411e-06,
+      "loss": 0.4347,
+      "step": 13644
+    },
+    {
+      "epoch": 2.24090078623776,
+      "grad_norm": 0.3488760731941589,
+      "learning_rate": 5.305780138929676e-06,
+      "loss": 0.4667,
+      "step": 13645
+    },
+    {
+      "epoch": 2.2410650134461028,
+      "grad_norm": 0.4165200065998101,
+      "learning_rate": 5.305318742003677e-06,
+      "loss": 0.4525,
+      "step": 13646
+    },
+    {
+      "epoch": 2.2412292406544454,
+      "grad_norm": 0.31530256152853486,
+      "learning_rate": 5.304857336331123e-06,
+      "loss": 0.4622,
+      "step": 13647
+    },
+    {
+      "epoch": 2.241393467862788,
+      "grad_norm": 0.4446666704947999,
+      "learning_rate": 5.304395921917027e-06,
+      "loss": 0.4301,
+      "step": 13648
+    },
+    {
+      "epoch": 2.241557695071131,
+      "grad_norm": 0.4053189269354466,
+      "learning_rate": 5.303934498766399e-06,
+      "loss": 0.4412,
+      "step": 13649
+    },
+    {
+      "epoch": 2.2417219222794738,
+      "grad_norm": 0.3625864205313529,
+      "learning_rate": 5.303473066884254e-06,
+      "loss": 0.4487,
+      "step": 13650
+    },
+    {
+      "epoch": 2.2418861494878164,
+      "grad_norm": 0.3885493040306324,
+      "learning_rate": 5.303011626275599e-06,
+      "loss": 0.4359,
+      "step": 13651
+    },
+    {
+      "epoch": 2.242050376696159,
+      "grad_norm": 0.318432682101398,
+      "learning_rate": 5.3025501769454475e-06,
+      "loss": 0.4322,
+      "step": 13652
+    },
+    {
+      "epoch": 2.242214603904502,
+      "grad_norm": 0.3007696684784351,
+      "learning_rate": 5.302088718898812e-06,
+      "loss": 0.4489,
+      "step": 13653
+    },
+    {
+      "epoch": 2.2423788311128448,
+      "grad_norm": 0.3887145358064661,
+      "learning_rate": 5.3016272521407036e-06,
+      "loss": 0.4483,
+      "step": 13654
+    },
+    {
+      "epoch": 2.2425430583211874,
+      "grad_norm": 0.3409044775291692,
+      "learning_rate": 5.301165776676134e-06,
+      "loss": 0.4433,
+      "step": 13655
+    },
+    {
+      "epoch": 2.24270728552953,
+      "grad_norm": 0.3670942744271576,
+      "learning_rate": 5.300704292510116e-06,
+      "loss": 0.4526,
+      "step": 13656
+    },
+    {
+      "epoch": 2.2428715127378727,
+      "grad_norm": 0.29240835654932273,
+      "learning_rate": 5.300242799647662e-06,
+      "loss": 0.4306,
+      "step": 13657
+    },
+    {
+      "epoch": 2.2430357399462157,
+      "grad_norm": 0.39224385150203395,
+      "learning_rate": 5.299781298093785e-06,
+      "loss": 0.4398,
+      "step": 13658
+    },
+    {
+      "epoch": 2.2431999671545584,
+      "grad_norm": 0.3124724267924632,
+      "learning_rate": 5.299319787853495e-06,
+      "loss": 0.4426,
+      "step": 13659
+    },
+    {
+      "epoch": 2.243364194362901,
+      "grad_norm": 0.3118206048528069,
+      "learning_rate": 5.298858268931805e-06,
+      "loss": 0.4599,
+      "step": 13660
+    },
+    {
+      "epoch": 2.2435284215712437,
+      "grad_norm": 0.3472209576784062,
+      "learning_rate": 5.298396741333729e-06,
+      "loss": 0.4427,
+      "step": 13661
+    },
+    {
+      "epoch": 2.2436926487795867,
+      "grad_norm": 0.28816801110315016,
+      "learning_rate": 5.297935205064279e-06,
+      "loss": 0.4155,
+      "step": 13662
+    },
+    {
+      "epoch": 2.2438568759879294,
+      "grad_norm": 0.3280013408771476,
+      "learning_rate": 5.297473660128469e-06,
+      "loss": 0.4351,
+      "step": 13663
+    },
+    {
+      "epoch": 2.244021103196272,
+      "grad_norm": 0.3013157024739304,
+      "learning_rate": 5.297012106531308e-06,
+      "loss": 0.4542,
+      "step": 13664
+    },
+    {
+      "epoch": 2.2441853304046147,
+      "grad_norm": 0.2698186431923782,
+      "learning_rate": 5.296550544277813e-06,
+      "loss": 0.4358,
+      "step": 13665
+    },
+    {
+      "epoch": 2.2443495576129573,
+      "grad_norm": 0.3152377394168143,
+      "learning_rate": 5.296088973372994e-06,
+      "loss": 0.4271,
+      "step": 13666
+    },
+    {
+      "epoch": 2.2445137848213004,
+      "grad_norm": 0.332333820287266,
+      "learning_rate": 5.295627393821867e-06,
+      "loss": 0.4439,
+      "step": 13667
+    },
+    {
+      "epoch": 2.244678012029643,
+      "grad_norm": 0.3540494011551216,
+      "learning_rate": 5.295165805629443e-06,
+      "loss": 0.427,
+      "step": 13668
+    },
+    {
+      "epoch": 2.2448422392379856,
+      "grad_norm": 0.27766510632642816,
+      "learning_rate": 5.294704208800735e-06,
+      "loss": 0.4408,
+      "step": 13669
+    },
+    {
+      "epoch": 2.2450064664463287,
+      "grad_norm": 0.32517347498764476,
+      "learning_rate": 5.2942426033407585e-06,
+      "loss": 0.4429,
+      "step": 13670
+    },
+    {
+      "epoch": 2.2451706936546714,
+      "grad_norm": 0.3013033368759928,
+      "learning_rate": 5.293780989254526e-06,
+      "loss": 0.4392,
+      "step": 13671
+    },
+    {
+      "epoch": 2.245334920863014,
+      "grad_norm": 0.282745086702727,
+      "learning_rate": 5.293319366547051e-06,
+      "loss": 0.4381,
+      "step": 13672
+    },
+    {
+      "epoch": 2.2454991480713566,
+      "grad_norm": 0.3523847887648517,
+      "learning_rate": 5.292857735223346e-06,
+      "loss": 0.4364,
+      "step": 13673
+    },
+    {
+      "epoch": 2.2456633752796993,
+      "grad_norm": 0.3750636418052584,
+      "learning_rate": 5.292396095288428e-06,
+      "loss": 0.4701,
+      "step": 13674
+    },
+    {
+      "epoch": 2.2458276024880424,
+      "grad_norm": 0.6454444354941448,
+      "learning_rate": 5.2919344467473074e-06,
+      "loss": 0.4425,
+      "step": 13675
+    },
+    {
+      "epoch": 2.245991829696385,
+      "grad_norm": 0.32847604128266267,
+      "learning_rate": 5.2914727896049996e-06,
+      "loss": 0.4349,
+      "step": 13676
+    },
+    {
+      "epoch": 2.2461560569047276,
+      "grad_norm": 0.27768628742635093,
+      "learning_rate": 5.291011123866519e-06,
+      "loss": 0.4275,
+      "step": 13677
+    },
+    {
+      "epoch": 2.2463202841130703,
+      "grad_norm": 0.29142901921932585,
+      "learning_rate": 5.290549449536879e-06,
+      "loss": 0.4339,
+      "step": 13678
+    },
+    {
+      "epoch": 2.2464845113214134,
+      "grad_norm": 0.5690887619651148,
+      "learning_rate": 5.290087766621095e-06,
+      "loss": 0.429,
+      "step": 13679
+    },
+    {
+      "epoch": 2.246648738529756,
+      "grad_norm": 0.3532823427162118,
+      "learning_rate": 5.289626075124179e-06,
+      "loss": 0.444,
+      "step": 13680
+    },
+    {
+      "epoch": 2.2468129657380986,
+      "grad_norm": 0.2671035256621192,
+      "learning_rate": 5.289164375051147e-06,
+      "loss": 0.46,
+      "step": 13681
+    },
+    {
+      "epoch": 2.2469771929464413,
+      "grad_norm": 0.34637078136166244,
+      "learning_rate": 5.2887026664070154e-06,
+      "loss": 0.4498,
+      "step": 13682
+    },
+    {
+      "epoch": 2.247141420154784,
+      "grad_norm": 0.3572211705234191,
+      "learning_rate": 5.288240949196796e-06,
+      "loss": 0.4573,
+      "step": 13683
+    },
+    {
+      "epoch": 2.247305647363127,
+      "grad_norm": 0.3579960088428089,
+      "learning_rate": 5.287779223425504e-06,
+      "loss": 0.4302,
+      "step": 13684
+    },
+    {
+      "epoch": 2.2474698745714696,
+      "grad_norm": 0.34066301797241494,
+      "learning_rate": 5.287317489098154e-06,
+      "loss": 0.443,
+      "step": 13685
+    },
+    {
+      "epoch": 2.2476341017798123,
+      "grad_norm": 0.32676299112996604,
+      "learning_rate": 5.286855746219762e-06,
+      "loss": 0.4399,
+      "step": 13686
+    },
+    {
+      "epoch": 2.2477983289881553,
+      "grad_norm": 0.36409542955257634,
+      "learning_rate": 5.2863939947953435e-06,
+      "loss": 0.4546,
+      "step": 13687
+    },
+    {
+      "epoch": 2.247962556196498,
+      "grad_norm": 0.4273691550545424,
+      "learning_rate": 5.285932234829911e-06,
+      "loss": 0.4636,
+      "step": 13688
+    },
+    {
+      "epoch": 2.2481267834048406,
+      "grad_norm": 0.36166863699389046,
+      "learning_rate": 5.285470466328482e-06,
+      "loss": 0.4517,
+      "step": 13689
+    },
+    {
+      "epoch": 2.2482910106131833,
+      "grad_norm": 0.3815787960529812,
+      "learning_rate": 5.285008689296069e-06,
+      "loss": 0.4625,
+      "step": 13690
+    },
+    {
+      "epoch": 2.248455237821526,
+      "grad_norm": 0.26063558836316875,
+      "learning_rate": 5.284546903737692e-06,
+      "loss": 0.4617,
+      "step": 13691
+    },
+    {
+      "epoch": 2.248619465029869,
+      "grad_norm": 0.4614932545774242,
+      "learning_rate": 5.284085109658363e-06,
+      "loss": 0.4235,
+      "step": 13692
+    },
+    {
+      "epoch": 2.2487836922382116,
+      "grad_norm": 0.32931830311298993,
+      "learning_rate": 5.283623307063098e-06,
+      "loss": 0.427,
+      "step": 13693
+    },
+    {
+      "epoch": 2.2489479194465543,
+      "grad_norm": 2.310641384368478,
+      "learning_rate": 5.283161495956912e-06,
+      "loss": 0.4346,
+      "step": 13694
+    },
+    {
+      "epoch": 2.249112146654897,
+      "grad_norm": 0.5378084523050332,
+      "learning_rate": 5.282699676344822e-06,
+      "loss": 0.4452,
+      "step": 13695
+    },
+    {
+      "epoch": 2.24927637386324,
+      "grad_norm": 0.34680472456278405,
+      "learning_rate": 5.282237848231844e-06,
+      "loss": 0.4457,
+      "step": 13696
+    },
+    {
+      "epoch": 2.2494406010715826,
+      "grad_norm": 0.3445890634659044,
+      "learning_rate": 5.281776011622994e-06,
+      "loss": 0.4428,
+      "step": 13697
+    },
+    {
+      "epoch": 2.2496048282799253,
+      "grad_norm": 0.2858278812975166,
+      "learning_rate": 5.2813141665232865e-06,
+      "loss": 0.4467,
+      "step": 13698
+    },
+    {
+      "epoch": 2.249769055488268,
+      "grad_norm": 0.3145020721528898,
+      "learning_rate": 5.280852312937738e-06,
+      "loss": 0.433,
+      "step": 13699
+    },
+    {
+      "epoch": 2.2499332826966105,
+      "grad_norm": 0.3817643009363534,
+      "learning_rate": 5.280390450871367e-06,
+      "loss": 0.446,
+      "step": 13700
+    },
+    {
+      "epoch": 2.2500975099049536,
+      "grad_norm": 0.4631720137309753,
+      "learning_rate": 5.279928580329187e-06,
+      "loss": 0.4365,
+      "step": 13701
+    },
+    {
+      "epoch": 2.2502617371132962,
+      "grad_norm": 0.3874126600272311,
+      "learning_rate": 5.279466701316214e-06,
+      "loss": 0.4451,
+      "step": 13702
+    },
+    {
+      "epoch": 2.250425964321639,
+      "grad_norm": 0.7340837113996945,
+      "learning_rate": 5.279004813837466e-06,
+      "loss": 0.454,
+      "step": 13703
+    },
+    {
+      "epoch": 2.250590191529982,
+      "grad_norm": 0.32826893954558084,
+      "learning_rate": 5.278542917897961e-06,
+      "loss": 0.4456,
+      "step": 13704
+    },
+    {
+      "epoch": 2.2507544187383246,
+      "grad_norm": 0.30313075634072284,
+      "learning_rate": 5.2780810135027115e-06,
+      "loss": 0.467,
+      "step": 13705
+    },
+    {
+      "epoch": 2.2509186459466672,
+      "grad_norm": 0.2936344465324073,
+      "learning_rate": 5.277619100656739e-06,
+      "loss": 0.4437,
+      "step": 13706
+    },
+    {
+      "epoch": 2.25108287315501,
+      "grad_norm": 0.32540102092866996,
+      "learning_rate": 5.277157179365056e-06,
+      "loss": 0.4375,
+      "step": 13707
+    },
+    {
+      "epoch": 2.2512471003633525,
+      "grad_norm": 0.5418154782829453,
+      "learning_rate": 5.276695249632683e-06,
+      "loss": 0.428,
+      "step": 13708
+    },
+    {
+      "epoch": 2.2514113275716956,
+      "grad_norm": 0.4711780155828875,
+      "learning_rate": 5.276233311464635e-06,
+      "loss": 0.4598,
+      "step": 13709
+    },
+    {
+      "epoch": 2.2515755547800382,
+      "grad_norm": 0.4652902743526495,
+      "learning_rate": 5.275771364865929e-06,
+      "loss": 0.4705,
+      "step": 13710
+    },
+    {
+      "epoch": 2.251739781988381,
+      "grad_norm": 0.38330167114251695,
+      "learning_rate": 5.275309409841583e-06,
+      "loss": 0.4371,
+      "step": 13711
+    },
+    {
+      "epoch": 2.2519040091967235,
+      "grad_norm": 0.34378733317073845,
+      "learning_rate": 5.274847446396615e-06,
+      "loss": 0.4462,
+      "step": 13712
+    },
+    {
+      "epoch": 2.2520682364050666,
+      "grad_norm": 0.3504403391095108,
+      "learning_rate": 5.27438547453604e-06,
+      "loss": 0.4499,
+      "step": 13713
+    },
+    {
+      "epoch": 2.2522324636134092,
+      "grad_norm": 0.3535033845955811,
+      "learning_rate": 5.273923494264877e-06,
+      "loss": 0.4429,
+      "step": 13714
+    },
+    {
+      "epoch": 2.252396690821752,
+      "grad_norm": 0.33591679853041095,
+      "learning_rate": 5.273461505588142e-06,
+      "loss": 0.4497,
+      "step": 13715
+    },
+    {
+      "epoch": 2.2525609180300945,
+      "grad_norm": 0.3383253978626295,
+      "learning_rate": 5.272999508510857e-06,
+      "loss": 0.4391,
+      "step": 13716
+    },
+    {
+      "epoch": 2.252725145238437,
+      "grad_norm": 0.32385647020423586,
+      "learning_rate": 5.272537503038036e-06,
+      "loss": 0.4337,
+      "step": 13717
+    },
+    {
+      "epoch": 2.2528893724467802,
+      "grad_norm": 0.3396561220455788,
+      "learning_rate": 5.2720754891746965e-06,
+      "loss": 0.4611,
+      "step": 13718
+    },
+    {
+      "epoch": 2.253053599655123,
+      "grad_norm": 0.38140665585784617,
+      "learning_rate": 5.271613466925859e-06,
+      "loss": 0.4453,
+      "step": 13719
+    },
+    {
+      "epoch": 2.2532178268634655,
+      "grad_norm": 0.6540812656662718,
+      "learning_rate": 5.2711514362965395e-06,
+      "loss": 0.443,
+      "step": 13720
+    },
+    {
+      "epoch": 2.2533820540718086,
+      "grad_norm": 0.28094182270807777,
+      "learning_rate": 5.270689397291757e-06,
+      "loss": 0.4211,
+      "step": 13721
+    },
+    {
+      "epoch": 2.253546281280151,
+      "grad_norm": 0.425522068083951,
+      "learning_rate": 5.2702273499165286e-06,
+      "loss": 0.4424,
+      "step": 13722
+    },
+    {
+      "epoch": 2.253710508488494,
+      "grad_norm": 0.41584840792055144,
+      "learning_rate": 5.2697652941758724e-06,
+      "loss": 0.4358,
+      "step": 13723
+    },
+    {
+      "epoch": 2.2538747356968365,
+      "grad_norm": 0.3177024211408362,
+      "learning_rate": 5.26930323007481e-06,
+      "loss": 0.434,
+      "step": 13724
+    },
+    {
+      "epoch": 2.254038962905179,
+      "grad_norm": 0.3330303286606085,
+      "learning_rate": 5.2688411576183565e-06,
+      "loss": 0.4462,
+      "step": 13725
+    },
+    {
+      "epoch": 2.254203190113522,
+      "grad_norm": 0.294027547485759,
+      "learning_rate": 5.268379076811532e-06,
+      "loss": 0.4463,
+      "step": 13726
+    },
+    {
+      "epoch": 2.254367417321865,
+      "grad_norm": 0.3264056370297087,
+      "learning_rate": 5.267916987659355e-06,
+      "loss": 0.4393,
+      "step": 13727
+    },
+    {
+      "epoch": 2.2545316445302075,
+      "grad_norm": 0.29553128037110177,
+      "learning_rate": 5.267454890166842e-06,
+      "loss": 0.4719,
+      "step": 13728
+    },
+    {
+      "epoch": 2.25469587173855,
+      "grad_norm": 0.331978086815522,
+      "learning_rate": 5.266992784339016e-06,
+      "loss": 0.4409,
+      "step": 13729
+    },
+    {
+      "epoch": 2.254860098946893,
+      "grad_norm": 0.2809751050096038,
+      "learning_rate": 5.266530670180894e-06,
+      "loss": 0.4285,
+      "step": 13730
+    },
+    {
+      "epoch": 2.255024326155236,
+      "grad_norm": 0.3182632580568042,
+      "learning_rate": 5.266068547697493e-06,
+      "loss": 0.4454,
+      "step": 13731
+    },
+    {
+      "epoch": 2.2551885533635785,
+      "grad_norm": 0.32836916473275135,
+      "learning_rate": 5.265606416893835e-06,
+      "loss": 0.4307,
+      "step": 13732
+    },
+    {
+      "epoch": 2.255352780571921,
+      "grad_norm": 0.2853170249483263,
+      "learning_rate": 5.265144277774938e-06,
+      "loss": 0.4344,
+      "step": 13733
+    },
+    {
+      "epoch": 2.2555170077802638,
+      "grad_norm": 0.5178439532805238,
+      "learning_rate": 5.2646821303458215e-06,
+      "loss": 0.4399,
+      "step": 13734
+    },
+    {
+      "epoch": 2.255681234988607,
+      "grad_norm": 0.5507828085958464,
+      "learning_rate": 5.264219974611505e-06,
+      "loss": 0.4655,
+      "step": 13735
+    },
+    {
+      "epoch": 2.2558454621969495,
+      "grad_norm": 0.3382196997965826,
+      "learning_rate": 5.263757810577006e-06,
+      "loss": 0.4631,
+      "step": 13736
+    },
+    {
+      "epoch": 2.256009689405292,
+      "grad_norm": 0.28330362320821995,
+      "learning_rate": 5.263295638247347e-06,
+      "loss": 0.4392,
+      "step": 13737
+    },
+    {
+      "epoch": 2.256173916613635,
+      "grad_norm": 0.35874744100148126,
+      "learning_rate": 5.262833457627546e-06,
+      "loss": 0.4491,
+      "step": 13738
+    },
+    {
+      "epoch": 2.256338143821978,
+      "grad_norm": 0.267510615311155,
+      "learning_rate": 5.262371268722623e-06,
+      "loss": 0.4176,
+      "step": 13739
+    },
+    {
+      "epoch": 2.2565023710303205,
+      "grad_norm": 0.5092229768402633,
+      "learning_rate": 5.261909071537598e-06,
+      "loss": 0.4427,
+      "step": 13740
+    },
+    {
+      "epoch": 2.256666598238663,
+      "grad_norm": 0.2804393487003495,
+      "learning_rate": 5.261446866077491e-06,
+      "loss": 0.433,
+      "step": 13741
+    },
+    {
+      "epoch": 2.2568308254470058,
+      "grad_norm": 0.3899613005160599,
+      "learning_rate": 5.260984652347323e-06,
+      "loss": 0.4338,
+      "step": 13742
+    },
+    {
+      "epoch": 2.256995052655349,
+      "grad_norm": 0.3052554308900633,
+      "learning_rate": 5.260522430352111e-06,
+      "loss": 0.418,
+      "step": 13743
+    },
+    {
+      "epoch": 2.2571592798636915,
+      "grad_norm": 0.3072133071489857,
+      "learning_rate": 5.260060200096877e-06,
+      "loss": 0.4692,
+      "step": 13744
+    },
+    {
+      "epoch": 2.257323507072034,
+      "grad_norm": 0.2834940365753804,
+      "learning_rate": 5.259597961586644e-06,
+      "loss": 0.4443,
+      "step": 13745
+    },
+    {
+      "epoch": 2.2574877342803767,
+      "grad_norm": 0.34360106449572925,
+      "learning_rate": 5.259135714826429e-06,
+      "loss": 0.4458,
+      "step": 13746
+    },
+    {
+      "epoch": 2.25765196148872,
+      "grad_norm": 0.34200371546354,
+      "learning_rate": 5.2586734598212515e-06,
+      "loss": 0.4496,
+      "step": 13747
+    },
+    {
+      "epoch": 2.2578161886970625,
+      "grad_norm": 0.32078529673268424,
+      "learning_rate": 5.258211196576134e-06,
+      "loss": 0.4541,
+      "step": 13748
+    },
+    {
+      "epoch": 2.257980415905405,
+      "grad_norm": 0.7634754044032883,
+      "learning_rate": 5.257748925096098e-06,
+      "loss": 0.4563,
+      "step": 13749
+    },
+    {
+      "epoch": 2.2581446431137477,
+      "grad_norm": 0.341481961545268,
+      "learning_rate": 5.257286645386164e-06,
+      "loss": 0.4456,
+      "step": 13750
+    },
+    {
+      "epoch": 2.2583088703220904,
+      "grad_norm": 0.2911188279442169,
+      "learning_rate": 5.256824357451351e-06,
+      "loss": 0.4547,
+      "step": 13751
+    },
+    {
+      "epoch": 2.2584730975304335,
+      "grad_norm": 0.37741366724925646,
+      "learning_rate": 5.256362061296681e-06,
+      "loss": 0.4595,
+      "step": 13752
+    },
+    {
+      "epoch": 2.258637324738776,
+      "grad_norm": 0.3611787880484088,
+      "learning_rate": 5.255899756927174e-06,
+      "loss": 0.4582,
+      "step": 13753
+    },
+    {
+      "epoch": 2.2588015519471187,
+      "grad_norm": 0.913157147165417,
+      "learning_rate": 5.255437444347852e-06,
+      "loss": 0.4317,
+      "step": 13754
+    },
+    {
+      "epoch": 2.258965779155462,
+      "grad_norm": 0.7085119346882607,
+      "learning_rate": 5.254975123563737e-06,
+      "loss": 0.4393,
+      "step": 13755
+    },
+    {
+      "epoch": 2.2591300063638045,
+      "grad_norm": 0.4123409307715226,
+      "learning_rate": 5.25451279457985e-06,
+      "loss": 0.4388,
+      "step": 13756
+    },
+    {
+      "epoch": 2.259294233572147,
+      "grad_norm": 0.3130367328214307,
+      "learning_rate": 5.25405045740121e-06,
+      "loss": 0.461,
+      "step": 13757
+    },
+    {
+      "epoch": 2.2594584607804897,
+      "grad_norm": 0.4009373961373831,
+      "learning_rate": 5.253588112032841e-06,
+      "loss": 0.4546,
+      "step": 13758
+    },
+    {
+      "epoch": 2.2596226879888324,
+      "grad_norm": 0.3549456349719573,
+      "learning_rate": 5.253125758479763e-06,
+      "loss": 0.4515,
+      "step": 13759
+    },
+    {
+      "epoch": 2.2597869151971754,
+      "grad_norm": 0.3200803902850578,
+      "learning_rate": 5.2526633967469995e-06,
+      "loss": 0.4402,
+      "step": 13760
+    },
+    {
+      "epoch": 2.259951142405518,
+      "grad_norm": 0.3332251927527049,
+      "learning_rate": 5.25220102683957e-06,
+      "loss": 0.4173,
+      "step": 13761
+    },
+    {
+      "epoch": 2.2601153696138607,
+      "grad_norm": 0.3379245162929123,
+      "learning_rate": 5.251738648762497e-06,
+      "loss": 0.4477,
+      "step": 13762
+    },
+    {
+      "epoch": 2.2602795968222034,
+      "grad_norm": 0.4068887227418874,
+      "learning_rate": 5.251276262520804e-06,
+      "loss": 0.4476,
+      "step": 13763
+    },
+    {
+      "epoch": 2.2604438240305464,
+      "grad_norm": 0.35292379389392914,
+      "learning_rate": 5.25081386811951e-06,
+      "loss": 0.4647,
+      "step": 13764
+    },
+    {
+      "epoch": 2.260608051238889,
+      "grad_norm": 0.31261175083494874,
+      "learning_rate": 5.2503514655636405e-06,
+      "loss": 0.456,
+      "step": 13765
+    },
+    {
+      "epoch": 2.2607722784472317,
+      "grad_norm": 0.3023902044167681,
+      "learning_rate": 5.249889054858214e-06,
+      "loss": 0.4486,
+      "step": 13766
+    },
+    {
+      "epoch": 2.2609365056555744,
+      "grad_norm": 0.3510764596785331,
+      "learning_rate": 5.249426636008257e-06,
+      "loss": 0.4648,
+      "step": 13767
+    },
+    {
+      "epoch": 2.261100732863917,
+      "grad_norm": 0.29579015381658,
+      "learning_rate": 5.248964209018787e-06,
+      "loss": 0.4333,
+      "step": 13768
+    },
+    {
+      "epoch": 2.26126496007226,
+      "grad_norm": 0.3517170589499364,
+      "learning_rate": 5.24850177389483e-06,
+      "loss": 0.4315,
+      "step": 13769
+    },
+    {
+      "epoch": 2.2614291872806027,
+      "grad_norm": 0.30858214290887587,
+      "learning_rate": 5.248039330641407e-06,
+      "loss": 0.4664,
+      "step": 13770
+    },
+    {
+      "epoch": 2.2615934144889454,
+      "grad_norm": 0.6687128463622946,
+      "learning_rate": 5.247576879263542e-06,
+      "loss": 0.4483,
+      "step": 13771
+    },
+    {
+      "epoch": 2.2617576416972884,
+      "grad_norm": 0.3588713029547824,
+      "learning_rate": 5.247114419766255e-06,
+      "loss": 0.426,
+      "step": 13772
+    },
+    {
+      "epoch": 2.261921868905631,
+      "grad_norm": 0.3184520396061027,
+      "learning_rate": 5.246651952154569e-06,
+      "loss": 0.4415,
+      "step": 13773
+    },
+    {
+      "epoch": 2.2620860961139737,
+      "grad_norm": 0.3182348076386226,
+      "learning_rate": 5.246189476433512e-06,
+      "loss": 0.4154,
+      "step": 13774
+    },
+    {
+      "epoch": 2.2622503233223163,
+      "grad_norm": 0.3472059557046651,
+      "learning_rate": 5.245726992608101e-06,
+      "loss": 0.4297,
+      "step": 13775
+    },
+    {
+      "epoch": 2.262414550530659,
+      "grad_norm": 0.4028328577668929,
+      "learning_rate": 5.245264500683361e-06,
+      "loss": 0.4502,
+      "step": 13776
+    },
+    {
+      "epoch": 2.262578777739002,
+      "grad_norm": 0.30761601943596584,
+      "learning_rate": 5.244802000664314e-06,
+      "loss": 0.4434,
+      "step": 13777
+    },
+    {
+      "epoch": 2.2627430049473447,
+      "grad_norm": 0.388515157470138,
+      "learning_rate": 5.244339492555987e-06,
+      "loss": 0.4364,
+      "step": 13778
+    },
+    {
+      "epoch": 2.2629072321556873,
+      "grad_norm": 0.35450795824264086,
+      "learning_rate": 5.2438769763634e-06,
+      "loss": 0.4367,
+      "step": 13779
+    },
+    {
+      "epoch": 2.26307145936403,
+      "grad_norm": 0.3294525187120768,
+      "learning_rate": 5.243414452091578e-06,
+      "loss": 0.4377,
+      "step": 13780
+    },
+    {
+      "epoch": 2.263235686572373,
+      "grad_norm": 0.3722129129869963,
+      "learning_rate": 5.242951919745542e-06,
+      "loss": 0.4388,
+      "step": 13781
+    },
+    {
+      "epoch": 2.2633999137807157,
+      "grad_norm": 0.4507616586121919,
+      "learning_rate": 5.242489379330317e-06,
+      "loss": 0.4447,
+      "step": 13782
+    },
+    {
+      "epoch": 2.2635641409890583,
+      "grad_norm": 0.3598737035818175,
+      "learning_rate": 5.242026830850928e-06,
+      "loss": 0.4402,
+      "step": 13783
+    },
+    {
+      "epoch": 2.263728368197401,
+      "grad_norm": 0.3324765686675211,
+      "learning_rate": 5.241564274312398e-06,
+      "loss": 0.4466,
+      "step": 13784
+    },
+    {
+      "epoch": 2.2638925954057436,
+      "grad_norm": 0.4083224933688828,
+      "learning_rate": 5.241101709719749e-06,
+      "loss": 0.4498,
+      "step": 13785
+    },
+    {
+      "epoch": 2.2640568226140867,
+      "grad_norm": 0.2933933784871219,
+      "learning_rate": 5.240639137078007e-06,
+      "loss": 0.4265,
+      "step": 13786
+    },
+    {
+      "epoch": 2.2642210498224293,
+      "grad_norm": 0.27710009965559146,
+      "learning_rate": 5.240176556392194e-06,
+      "loss": 0.4314,
+      "step": 13787
+    },
+    {
+      "epoch": 2.264385277030772,
+      "grad_norm": 0.29841634796489386,
+      "learning_rate": 5.239713967667338e-06,
+      "loss": 0.4508,
+      "step": 13788
+    },
+    {
+      "epoch": 2.264549504239115,
+      "grad_norm": 0.5230942248164453,
+      "learning_rate": 5.23925137090846e-06,
+      "loss": 0.4558,
+      "step": 13789
+    },
+    {
+      "epoch": 2.2647137314474577,
+      "grad_norm": 0.5844725375386487,
+      "learning_rate": 5.238788766120583e-06,
+      "loss": 0.4491,
+      "step": 13790
+    },
+    {
+      "epoch": 2.2648779586558003,
+      "grad_norm": 0.35880413792086935,
+      "learning_rate": 5.238326153308733e-06,
+      "loss": 0.4613,
+      "step": 13791
+    },
+    {
+      "epoch": 2.265042185864143,
+      "grad_norm": 0.27176190824021323,
+      "learning_rate": 5.237863532477936e-06,
+      "loss": 0.4297,
+      "step": 13792
+    },
+    {
+      "epoch": 2.2652064130724856,
+      "grad_norm": 0.31010986715177674,
+      "learning_rate": 5.237400903633215e-06,
+      "loss": 0.4429,
+      "step": 13793
+    },
+    {
+      "epoch": 2.2653706402808287,
+      "grad_norm": 0.35101079799031804,
+      "learning_rate": 5.236938266779595e-06,
+      "loss": 0.4443,
+      "step": 13794
+    },
+    {
+      "epoch": 2.2655348674891713,
+      "grad_norm": 0.2913660569347403,
+      "learning_rate": 5.236475621922099e-06,
+      "loss": 0.4161,
+      "step": 13795
+    },
+    {
+      "epoch": 2.265699094697514,
+      "grad_norm": 0.37533689834149453,
+      "learning_rate": 5.236012969065754e-06,
+      "loss": 0.434,
+      "step": 13796
+    },
+    {
+      "epoch": 2.2658633219058566,
+      "grad_norm": 0.312108000748107,
+      "learning_rate": 5.2355503082155835e-06,
+      "loss": 0.4432,
+      "step": 13797
+    },
+    {
+      "epoch": 2.2660275491141997,
+      "grad_norm": 0.33505756371900425,
+      "learning_rate": 5.2350876393766135e-06,
+      "loss": 0.4509,
+      "step": 13798
+    },
+    {
+      "epoch": 2.2661917763225423,
+      "grad_norm": 0.2958642981997617,
+      "learning_rate": 5.2346249625538684e-06,
+      "loss": 0.4313,
+      "step": 13799
+    },
+    {
+      "epoch": 2.266356003530885,
+      "grad_norm": 0.3161571807919032,
+      "learning_rate": 5.234162277752374e-06,
+      "loss": 0.4637,
+      "step": 13800
+    },
+    {
+      "epoch": 2.2665202307392276,
+      "grad_norm": 0.2850364464418512,
+      "learning_rate": 5.233699584977154e-06,
+      "loss": 0.4434,
+      "step": 13801
+    },
+    {
+      "epoch": 2.2666844579475702,
+      "grad_norm": 0.44128845128357014,
+      "learning_rate": 5.233236884233234e-06,
+      "loss": 0.4314,
+      "step": 13802
+    },
+    {
+      "epoch": 2.2668486851559133,
+      "grad_norm": 0.2709781947969362,
+      "learning_rate": 5.232774175525642e-06,
+      "loss": 0.4356,
+      "step": 13803
+    },
+    {
+      "epoch": 2.267012912364256,
+      "grad_norm": 0.3143610695517449,
+      "learning_rate": 5.232311458859401e-06,
+      "loss": 0.452,
+      "step": 13804
+    },
+    {
+      "epoch": 2.2671771395725986,
+      "grad_norm": 0.2811302211053564,
+      "learning_rate": 5.231848734239536e-06,
+      "loss": 0.4454,
+      "step": 13805
+    },
+    {
+      "epoch": 2.2673413667809417,
+      "grad_norm": 0.29524919404369326,
+      "learning_rate": 5.231386001671074e-06,
+      "loss": 0.4252,
+      "step": 13806
+    },
+    {
+      "epoch": 2.2675055939892843,
+      "grad_norm": 0.3036902739174187,
+      "learning_rate": 5.23092326115904e-06,
+      "loss": 0.4454,
+      "step": 13807
+    },
+    {
+      "epoch": 2.267669821197627,
+      "grad_norm": 0.3115339770671733,
+      "learning_rate": 5.230460512708461e-06,
+      "loss": 0.4635,
+      "step": 13808
+    },
+    {
+      "epoch": 2.2678340484059696,
+      "grad_norm": 0.29688931339668434,
+      "learning_rate": 5.2299977563243635e-06,
+      "loss": 0.4468,
+      "step": 13809
+    },
+    {
+      "epoch": 2.267998275614312,
+      "grad_norm": 0.3085715958936347,
+      "learning_rate": 5.229534992011769e-06,
+      "loss": 0.4583,
+      "step": 13810
+    },
+    {
+      "epoch": 2.2681625028226553,
+      "grad_norm": 0.3095940496568009,
+      "learning_rate": 5.229072219775708e-06,
+      "loss": 0.4407,
+      "step": 13811
+    },
+    {
+      "epoch": 2.268326730030998,
+      "grad_norm": 0.4908325226525298,
+      "learning_rate": 5.228609439621206e-06,
+      "loss": 0.4401,
+      "step": 13812
+    },
+    {
+      "epoch": 2.2684909572393406,
+      "grad_norm": 0.44589233555087854,
+      "learning_rate": 5.2281466515532894e-06,
+      "loss": 0.4209,
+      "step": 13813
+    },
+    {
+      "epoch": 2.268655184447683,
+      "grad_norm": 0.2802820328979476,
+      "learning_rate": 5.227683855576983e-06,
+      "loss": 0.4287,
+      "step": 13814
+    },
+    {
+      "epoch": 2.2688194116560263,
+      "grad_norm": 0.4046303814982037,
+      "learning_rate": 5.227221051697313e-06,
+      "loss": 0.4517,
+      "step": 13815
+    },
+    {
+      "epoch": 2.268983638864369,
+      "grad_norm": 0.33943476882464263,
+      "learning_rate": 5.226758239919308e-06,
+      "loss": 0.4378,
+      "step": 13816
+    },
+    {
+      "epoch": 2.2691478660727116,
+      "grad_norm": 0.3936793154256377,
+      "learning_rate": 5.226295420247994e-06,
+      "loss": 0.459,
+      "step": 13817
+    },
+    {
+      "epoch": 2.269312093281054,
+      "grad_norm": 0.2697716266058795,
+      "learning_rate": 5.225832592688397e-06,
+      "loss": 0.4345,
+      "step": 13818
+    },
+    {
+      "epoch": 2.269476320489397,
+      "grad_norm": 0.3475167102931843,
+      "learning_rate": 5.2253697572455424e-06,
+      "loss": 0.4517,
+      "step": 13819
+    },
+    {
+      "epoch": 2.26964054769774,
+      "grad_norm": 0.3057864002599708,
+      "learning_rate": 5.224906913924459e-06,
+      "loss": 0.4186,
+      "step": 13820
+    },
+    {
+      "epoch": 2.2698047749060826,
+      "grad_norm": 0.3029753715957564,
+      "learning_rate": 5.224444062730174e-06,
+      "loss": 0.4541,
+      "step": 13821
+    },
+    {
+      "epoch": 2.269969002114425,
+      "grad_norm": 0.307432523810183,
+      "learning_rate": 5.2239812036677135e-06,
+      "loss": 0.4453,
+      "step": 13822
+    },
+    {
+      "epoch": 2.2701332293227683,
+      "grad_norm": 0.3060059931511143,
+      "learning_rate": 5.2235183367421055e-06,
+      "loss": 0.4426,
+      "step": 13823
+    },
+    {
+      "epoch": 2.270297456531111,
+      "grad_norm": 0.36256508614149663,
+      "learning_rate": 5.2230554619583756e-06,
+      "loss": 0.4367,
+      "step": 13824
+    },
+    {
+      "epoch": 2.2704616837394536,
+      "grad_norm": 0.3647724136370156,
+      "learning_rate": 5.222592579321552e-06,
+      "loss": 0.4479,
+      "step": 13825
+    },
+    {
+      "epoch": 2.270625910947796,
+      "grad_norm": 0.31823959283917014,
+      "learning_rate": 5.2221296888366615e-06,
+      "loss": 0.442,
+      "step": 13826
+    },
+    {
+      "epoch": 2.270790138156139,
+      "grad_norm": 0.30892461908438634,
+      "learning_rate": 5.221666790508733e-06,
+      "loss": 0.4305,
+      "step": 13827
+    },
+    {
+      "epoch": 2.270954365364482,
+      "grad_norm": 0.5498779110061026,
+      "learning_rate": 5.221203884342793e-06,
+      "loss": 0.4374,
+      "step": 13828
+    },
+    {
+      "epoch": 2.2711185925728246,
+      "grad_norm": 0.3437254626690269,
+      "learning_rate": 5.22074097034387e-06,
+      "loss": 0.4173,
+      "step": 13829
+    },
+    {
+      "epoch": 2.271282819781167,
+      "grad_norm": 0.39163066483566433,
+      "learning_rate": 5.22027804851699e-06,
+      "loss": 0.4317,
+      "step": 13830
+    },
+    {
+      "epoch": 2.27144704698951,
+      "grad_norm": 0.31973453986137745,
+      "learning_rate": 5.219815118867182e-06,
+      "loss": 0.417,
+      "step": 13831
+    },
+    {
+      "epoch": 2.271611274197853,
+      "grad_norm": 0.5012814391620765,
+      "learning_rate": 5.219352181399474e-06,
+      "loss": 0.4577,
+      "step": 13832
+    },
+    {
+      "epoch": 2.2717755014061956,
+      "grad_norm": 0.3326116761911855,
+      "learning_rate": 5.218889236118892e-06,
+      "loss": 0.4128,
+      "step": 13833
+    },
+    {
+      "epoch": 2.271939728614538,
+      "grad_norm": 0.34504752744659983,
+      "learning_rate": 5.218426283030467e-06,
+      "loss": 0.4525,
+      "step": 13834
+    },
+    {
+      "epoch": 2.272103955822881,
+      "grad_norm": 0.3250596236621934,
+      "learning_rate": 5.2179633221392255e-06,
+      "loss": 0.4545,
+      "step": 13835
+    },
+    {
+      "epoch": 2.2722681830312235,
+      "grad_norm": 0.28065194575482644,
+      "learning_rate": 5.217500353450196e-06,
+      "loss": 0.4345,
+      "step": 13836
+    },
+    {
+      "epoch": 2.2724324102395665,
+      "grad_norm": 0.7255066479196198,
+      "learning_rate": 5.217037376968406e-06,
+      "loss": 0.4438,
+      "step": 13837
+    },
+    {
+      "epoch": 2.272596637447909,
+      "grad_norm": 0.4017089397573672,
+      "learning_rate": 5.216574392698886e-06,
+      "loss": 0.4591,
+      "step": 13838
+    },
+    {
+      "epoch": 2.272760864656252,
+      "grad_norm": 0.5304928041774932,
+      "learning_rate": 5.216111400646662e-06,
+      "loss": 0.4441,
+      "step": 13839
+    },
+    {
+      "epoch": 2.272925091864595,
+      "grad_norm": 0.2609976185282584,
+      "learning_rate": 5.215648400816763e-06,
+      "loss": 0.4351,
+      "step": 13840
+    },
+    {
+      "epoch": 2.2730893190729375,
+      "grad_norm": 0.358583283765098,
+      "learning_rate": 5.215185393214219e-06,
+      "loss": 0.4385,
+      "step": 13841
+    },
+    {
+      "epoch": 2.27325354628128,
+      "grad_norm": 0.3020230812447199,
+      "learning_rate": 5.2147223778440586e-06,
+      "loss": 0.4399,
+      "step": 13842
+    },
+    {
+      "epoch": 2.273417773489623,
+      "grad_norm": 0.3367916444539525,
+      "learning_rate": 5.21425935471131e-06,
+      "loss": 0.4235,
+      "step": 13843
+    },
+    {
+      "epoch": 2.2735820006979655,
+      "grad_norm": 0.2941663693437456,
+      "learning_rate": 5.213796323821002e-06,
+      "loss": 0.4502,
+      "step": 13844
+    },
+    {
+      "epoch": 2.2737462279063085,
+      "grad_norm": 0.4697417001010032,
+      "learning_rate": 5.2133332851781624e-06,
+      "loss": 0.4267,
+      "step": 13845
+    },
+    {
+      "epoch": 2.273910455114651,
+      "grad_norm": 0.3388810367942243,
+      "learning_rate": 5.212870238787823e-06,
+      "loss": 0.4564,
+      "step": 13846
+    },
+    {
+      "epoch": 2.274074682322994,
+      "grad_norm": 0.2844059312289679,
+      "learning_rate": 5.2124071846550115e-06,
+      "loss": 0.4527,
+      "step": 13847
+    },
+    {
+      "epoch": 2.2742389095313364,
+      "grad_norm": 0.4828332149280781,
+      "learning_rate": 5.2119441227847555e-06,
+      "loss": 0.4426,
+      "step": 13848
+    },
+    {
+      "epoch": 2.2744031367396795,
+      "grad_norm": 0.33234453953277954,
+      "learning_rate": 5.2114810531820864e-06,
+      "loss": 0.456,
+      "step": 13849
+    },
+    {
+      "epoch": 2.274567363948022,
+      "grad_norm": 0.29849007460004134,
+      "learning_rate": 5.211017975852035e-06,
+      "loss": 0.4232,
+      "step": 13850
+    },
+    {
+      "epoch": 2.274731591156365,
+      "grad_norm": 0.31706913669563674,
+      "learning_rate": 5.210554890799626e-06,
+      "loss": 0.4273,
+      "step": 13851
+    },
+    {
+      "epoch": 2.2748958183647074,
+      "grad_norm": 0.4073711965734478,
+      "learning_rate": 5.210091798029894e-06,
+      "loss": 0.4376,
+      "step": 13852
+    },
+    {
+      "epoch": 2.27506004557305,
+      "grad_norm": 0.5229393755986401,
+      "learning_rate": 5.209628697547865e-06,
+      "loss": 0.4459,
+      "step": 13853
+    },
+    {
+      "epoch": 2.275224272781393,
+      "grad_norm": 0.3537211489217148,
+      "learning_rate": 5.209165589358571e-06,
+      "loss": 0.4747,
+      "step": 13854
+    },
+    {
+      "epoch": 2.275388499989736,
+      "grad_norm": 0.40748720524206644,
+      "learning_rate": 5.20870247346704e-06,
+      "loss": 0.4416,
+      "step": 13855
+    },
+    {
+      "epoch": 2.2755527271980784,
+      "grad_norm": 0.3216274797784098,
+      "learning_rate": 5.2082393498783036e-06,
+      "loss": 0.4572,
+      "step": 13856
+    },
+    {
+      "epoch": 2.2757169544064215,
+      "grad_norm": 0.35278286290399846,
+      "learning_rate": 5.207776218597391e-06,
+      "loss": 0.4452,
+      "step": 13857
+    },
+    {
+      "epoch": 2.275881181614764,
+      "grad_norm": 0.4098238080258137,
+      "learning_rate": 5.207313079629331e-06,
+      "loss": 0.432,
+      "step": 13858
+    },
+    {
+      "epoch": 2.276045408823107,
+      "grad_norm": 0.4580092831691726,
+      "learning_rate": 5.2068499329791575e-06,
+      "loss": 0.4349,
+      "step": 13859
+    },
+    {
+      "epoch": 2.2762096360314494,
+      "grad_norm": 0.38826013386276875,
+      "learning_rate": 5.206386778651896e-06,
+      "loss": 0.4141,
+      "step": 13860
+    },
+    {
+      "epoch": 2.276373863239792,
+      "grad_norm": 0.2941958339679804,
+      "learning_rate": 5.20592361665258e-06,
+      "loss": 0.4247,
+      "step": 13861
+    },
+    {
+      "epoch": 2.276538090448135,
+      "grad_norm": 0.3003495984838666,
+      "learning_rate": 5.20546044698624e-06,
+      "loss": 0.4258,
+      "step": 13862
+    },
+    {
+      "epoch": 2.276702317656478,
+      "grad_norm": 0.33772284519341567,
+      "learning_rate": 5.204997269657905e-06,
+      "loss": 0.439,
+      "step": 13863
+    },
+    {
+      "epoch": 2.2768665448648204,
+      "grad_norm": 0.35383257962280384,
+      "learning_rate": 5.2045340846726054e-06,
+      "loss": 0.4494,
+      "step": 13864
+    },
+    {
+      "epoch": 2.277030772073163,
+      "grad_norm": 0.3434125913276977,
+      "learning_rate": 5.204070892035372e-06,
+      "loss": 0.4356,
+      "step": 13865
+    },
+    {
+      "epoch": 2.277194999281506,
+      "grad_norm": 0.30071978455766885,
+      "learning_rate": 5.203607691751237e-06,
+      "loss": 0.4401,
+      "step": 13866
+    },
+    {
+      "epoch": 2.277359226489849,
+      "grad_norm": 0.29424246947418803,
+      "learning_rate": 5.203144483825231e-06,
+      "loss": 0.4518,
+      "step": 13867
+    },
+    {
+      "epoch": 2.2775234536981914,
+      "grad_norm": 0.7205369293800823,
+      "learning_rate": 5.202681268262382e-06,
+      "loss": 0.433,
+      "step": 13868
+    },
+    {
+      "epoch": 2.277687680906534,
+      "grad_norm": 0.4550218086842287,
+      "learning_rate": 5.202218045067724e-06,
+      "loss": 0.4227,
+      "step": 13869
+    },
+    {
+      "epoch": 2.2778519081148767,
+      "grad_norm": 0.5522939727022292,
+      "learning_rate": 5.201754814246286e-06,
+      "loss": 0.4329,
+      "step": 13870
+    },
+    {
+      "epoch": 2.27801613532322,
+      "grad_norm": 0.34802550224546164,
+      "learning_rate": 5.201291575803103e-06,
+      "loss": 0.4396,
+      "step": 13871
+    },
+    {
+      "epoch": 2.2781803625315624,
+      "grad_norm": 0.29412082409797263,
+      "learning_rate": 5.200828329743202e-06,
+      "loss": 0.4474,
+      "step": 13872
+    },
+    {
+      "epoch": 2.278344589739905,
+      "grad_norm": 0.3387381717897766,
+      "learning_rate": 5.200365076071616e-06,
+      "loss": 0.4344,
+      "step": 13873
+    },
+    {
+      "epoch": 2.278508816948248,
+      "grad_norm": 0.3520361563968808,
+      "learning_rate": 5.199901814793376e-06,
+      "loss": 0.4562,
+      "step": 13874
+    },
+    {
+      "epoch": 2.2786730441565908,
+      "grad_norm": 0.3288366800434973,
+      "learning_rate": 5.199438545913514e-06,
+      "loss": 0.4355,
+      "step": 13875
+    },
+    {
+      "epoch": 2.2788372713649334,
+      "grad_norm": 0.29865338535780817,
+      "learning_rate": 5.19897526943706e-06,
+      "loss": 0.4518,
+      "step": 13876
+    },
+    {
+      "epoch": 2.279001498573276,
+      "grad_norm": 0.3404207815125893,
+      "learning_rate": 5.198511985369049e-06,
+      "loss": 0.4322,
+      "step": 13877
+    },
+    {
+      "epoch": 2.2791657257816187,
+      "grad_norm": 0.31980604254782763,
+      "learning_rate": 5.198048693714509e-06,
+      "loss": 0.4281,
+      "step": 13878
+    },
+    {
+      "epoch": 2.2793299529899618,
+      "grad_norm": 0.2872471598384732,
+      "learning_rate": 5.197585394478474e-06,
+      "loss": 0.4482,
+      "step": 13879
+    },
+    {
+      "epoch": 2.2794941801983044,
+      "grad_norm": 0.3154232003073397,
+      "learning_rate": 5.1971220876659745e-06,
+      "loss": 0.4364,
+      "step": 13880
+    },
+    {
+      "epoch": 2.279658407406647,
+      "grad_norm": 0.6348972132812276,
+      "learning_rate": 5.196658773282044e-06,
+      "loss": 0.4298,
+      "step": 13881
+    },
+    {
+      "epoch": 2.2798226346149897,
+      "grad_norm": 0.28308749448615994,
+      "learning_rate": 5.196195451331715e-06,
+      "loss": 0.4315,
+      "step": 13882
+    },
+    {
+      "epoch": 2.2799868618233328,
+      "grad_norm": 0.37848023486848437,
+      "learning_rate": 5.195732121820016e-06,
+      "loss": 0.4639,
+      "step": 13883
+    },
+    {
+      "epoch": 2.2801510890316754,
+      "grad_norm": 0.40271339721462446,
+      "learning_rate": 5.1952687847519835e-06,
+      "loss": 0.4328,
+      "step": 13884
+    },
+    {
+      "epoch": 2.280315316240018,
+      "grad_norm": 0.3014707376437167,
+      "learning_rate": 5.194805440132647e-06,
+      "loss": 0.4214,
+      "step": 13885
+    },
+    {
+      "epoch": 2.2804795434483607,
+      "grad_norm": 0.3373711819342816,
+      "learning_rate": 5.19434208796704e-06,
+      "loss": 0.4473,
+      "step": 13886
+    },
+    {
+      "epoch": 2.2806437706567033,
+      "grad_norm": 0.38535680109940185,
+      "learning_rate": 5.193878728260194e-06,
+      "loss": 0.423,
+      "step": 13887
+    },
+    {
+      "epoch": 2.2808079978650464,
+      "grad_norm": 0.39323458035612957,
+      "learning_rate": 5.193415361017145e-06,
+      "loss": 0.4376,
+      "step": 13888
+    },
+    {
+      "epoch": 2.280972225073389,
+      "grad_norm": 0.40926189382068584,
+      "learning_rate": 5.19295198624292e-06,
+      "loss": 0.4268,
+      "step": 13889
+    },
+    {
+      "epoch": 2.2811364522817317,
+      "grad_norm": 0.39605642816296704,
+      "learning_rate": 5.192488603942555e-06,
+      "loss": 0.441,
+      "step": 13890
+    },
+    {
+      "epoch": 2.2813006794900748,
+      "grad_norm": 0.3321036515817321,
+      "learning_rate": 5.192025214121084e-06,
+      "loss": 0.4469,
+      "step": 13891
+    },
+    {
+      "epoch": 2.2814649066984174,
+      "grad_norm": 0.3084896280179283,
+      "learning_rate": 5.191561816783538e-06,
+      "loss": 0.4458,
+      "step": 13892
+    },
+    {
+      "epoch": 2.28162913390676,
+      "grad_norm": 0.4768889527534172,
+      "learning_rate": 5.1910984119349495e-06,
+      "loss": 0.4617,
+      "step": 13893
+    },
+    {
+      "epoch": 2.2817933611151027,
+      "grad_norm": 0.479253467678791,
+      "learning_rate": 5.190634999580352e-06,
+      "loss": 0.4393,
+      "step": 13894
+    },
+    {
+      "epoch": 2.2819575883234453,
+      "grad_norm": 0.34832956361346284,
+      "learning_rate": 5.190171579724779e-06,
+      "loss": 0.4353,
+      "step": 13895
+    },
+    {
+      "epoch": 2.2821218155317884,
+      "grad_norm": 0.3443708744422341,
+      "learning_rate": 5.189708152373266e-06,
+      "loss": 0.4213,
+      "step": 13896
+    },
+    {
+      "epoch": 2.282286042740131,
+      "grad_norm": 0.5197437799964241,
+      "learning_rate": 5.189244717530841e-06,
+      "loss": 0.4352,
+      "step": 13897
+    },
+    {
+      "epoch": 2.2824502699484737,
+      "grad_norm": 0.3786963383411827,
+      "learning_rate": 5.188781275202542e-06,
+      "loss": 0.4144,
+      "step": 13898
+    },
+    {
+      "epoch": 2.2826144971568163,
+      "grad_norm": 0.31878254563006886,
+      "learning_rate": 5.188317825393398e-06,
+      "loss": 0.4375,
+      "step": 13899
+    },
+    {
+      "epoch": 2.2827787243651594,
+      "grad_norm": 0.32371021620229423,
+      "learning_rate": 5.187854368108448e-06,
+      "loss": 0.467,
+      "step": 13900
+    },
+    {
+      "epoch": 2.282942951573502,
+      "grad_norm": 0.27065989521911527,
+      "learning_rate": 5.1873909033527225e-06,
+      "loss": 0.4494,
+      "step": 13901
+    },
+    {
+      "epoch": 2.2831071787818447,
+      "grad_norm": 0.36088684689567835,
+      "learning_rate": 5.186927431131254e-06,
+      "loss": 0.4493,
+      "step": 13902
+    },
+    {
+      "epoch": 2.2832714059901873,
+      "grad_norm": 0.3054866896938407,
+      "learning_rate": 5.186463951449079e-06,
+      "loss": 0.444,
+      "step": 13903
+    },
+    {
+      "epoch": 2.28343563319853,
+      "grad_norm": 0.9427890070184183,
+      "learning_rate": 5.1860004643112295e-06,
+      "loss": 0.4566,
+      "step": 13904
+    },
+    {
+      "epoch": 2.283599860406873,
+      "grad_norm": 0.36980738576340033,
+      "learning_rate": 5.1855369697227405e-06,
+      "loss": 0.4283,
+      "step": 13905
+    },
+    {
+      "epoch": 2.2837640876152157,
+      "grad_norm": 0.3261290341141939,
+      "learning_rate": 5.185073467688646e-06,
+      "loss": 0.4368,
+      "step": 13906
+    },
+    {
+      "epoch": 2.2839283148235583,
+      "grad_norm": 0.33320968485632513,
+      "learning_rate": 5.184609958213978e-06,
+      "loss": 0.4469,
+      "step": 13907
+    },
+    {
+      "epoch": 2.2840925420319014,
+      "grad_norm": 0.2856889531244492,
+      "learning_rate": 5.184146441303773e-06,
+      "loss": 0.4525,
+      "step": 13908
+    },
+    {
+      "epoch": 2.284256769240244,
+      "grad_norm": 0.3942354513878941,
+      "learning_rate": 5.183682916963066e-06,
+      "loss": 0.4316,
+      "step": 13909
+    },
+    {
+      "epoch": 2.2844209964485866,
+      "grad_norm": 0.3732289311915922,
+      "learning_rate": 5.183219385196887e-06,
+      "loss": 0.4307,
+      "step": 13910
+    },
+    {
+      "epoch": 2.2845852236569293,
+      "grad_norm": 0.33045744667353805,
+      "learning_rate": 5.182755846010276e-06,
+      "loss": 0.4419,
+      "step": 13911
+    },
+    {
+      "epoch": 2.284749450865272,
+      "grad_norm": 0.3139712346377227,
+      "learning_rate": 5.1822922994082636e-06,
+      "loss": 0.4438,
+      "step": 13912
+    },
+    {
+      "epoch": 2.284913678073615,
+      "grad_norm": 0.4955002940911547,
+      "learning_rate": 5.181828745395886e-06,
+      "loss": 0.4432,
+      "step": 13913
+    },
+    {
+      "epoch": 2.2850779052819576,
+      "grad_norm": 0.353124834913557,
+      "learning_rate": 5.181365183978176e-06,
+      "loss": 0.4328,
+      "step": 13914
+    },
+    {
+      "epoch": 2.2852421324903003,
+      "grad_norm": 0.33109773121135133,
+      "learning_rate": 5.180901615160172e-06,
+      "loss": 0.4238,
+      "step": 13915
+    },
+    {
+      "epoch": 2.285406359698643,
+      "grad_norm": 0.3091236114028852,
+      "learning_rate": 5.180438038946905e-06,
+      "loss": 0.4395,
+      "step": 13916
+    },
+    {
+      "epoch": 2.285570586906986,
+      "grad_norm": 0.3072253498064372,
+      "learning_rate": 5.179974455343412e-06,
+      "loss": 0.4536,
+      "step": 13917
+    },
+    {
+      "epoch": 2.2857348141153286,
+      "grad_norm": 0.36653777780275587,
+      "learning_rate": 5.179510864354727e-06,
+      "loss": 0.4406,
+      "step": 13918
+    },
+    {
+      "epoch": 2.2858990413236713,
+      "grad_norm": 0.3347389257699555,
+      "learning_rate": 5.179047265985885e-06,
+      "loss": 0.4269,
+      "step": 13919
+    },
+    {
+      "epoch": 2.286063268532014,
+      "grad_norm": 0.30147606850707453,
+      "learning_rate": 5.178583660241923e-06,
+      "loss": 0.4446,
+      "step": 13920
+    },
+    {
+      "epoch": 2.2862274957403566,
+      "grad_norm": 0.37986592730870994,
+      "learning_rate": 5.178120047127874e-06,
+      "loss": 0.4437,
+      "step": 13921
+    },
+    {
+      "epoch": 2.2863917229486996,
+      "grad_norm": 0.4007300546013079,
+      "learning_rate": 5.177656426648774e-06,
+      "loss": 0.4545,
+      "step": 13922
+    },
+    {
+      "epoch": 2.2865559501570423,
+      "grad_norm": 0.32051943336442223,
+      "learning_rate": 5.177192798809658e-06,
+      "loss": 0.4105,
+      "step": 13923
+    },
+    {
+      "epoch": 2.286720177365385,
+      "grad_norm": 0.38011102453141177,
+      "learning_rate": 5.1767291636155625e-06,
+      "loss": 0.4485,
+      "step": 13924
+    },
+    {
+      "epoch": 2.286884404573728,
+      "grad_norm": 0.35926434734994367,
+      "learning_rate": 5.176265521071523e-06,
+      "loss": 0.4362,
+      "step": 13925
+    },
+    {
+      "epoch": 2.2870486317820706,
+      "grad_norm": 0.3731633735917125,
+      "learning_rate": 5.175801871182575e-06,
+      "loss": 0.4455,
+      "step": 13926
+    },
+    {
+      "epoch": 2.2872128589904133,
+      "grad_norm": 0.29885224871406696,
+      "learning_rate": 5.175338213953752e-06,
+      "loss": 0.457,
+      "step": 13927
+    },
+    {
+      "epoch": 2.287377086198756,
+      "grad_norm": 0.3104186968740682,
+      "learning_rate": 5.174874549390092e-06,
+      "loss": 0.4183,
+      "step": 13928
+    },
+    {
+      "epoch": 2.2875413134070985,
+      "grad_norm": 0.34129434291138006,
+      "learning_rate": 5.17441087749663e-06,
+      "loss": 0.4402,
+      "step": 13929
+    },
+    {
+      "epoch": 2.2877055406154416,
+      "grad_norm": 0.31123742064990884,
+      "learning_rate": 5.173947198278405e-06,
+      "loss": 0.4332,
+      "step": 13930
+    },
+    {
+      "epoch": 2.2878697678237843,
+      "grad_norm": 0.29266834175594275,
+      "learning_rate": 5.173483511740448e-06,
+      "loss": 0.444,
+      "step": 13931
+    },
+    {
+      "epoch": 2.288033995032127,
+      "grad_norm": 0.3725237989488707,
+      "learning_rate": 5.173019817887798e-06,
+      "loss": 0.4677,
+      "step": 13932
+    },
+    {
+      "epoch": 2.2881982222404695,
+      "grad_norm": 0.31685516806527486,
+      "learning_rate": 5.1725561167254895e-06,
+      "loss": 0.4456,
+      "step": 13933
+    },
+    {
+      "epoch": 2.2883624494488126,
+      "grad_norm": 0.39825508760706196,
+      "learning_rate": 5.172092408258562e-06,
+      "loss": 0.4505,
+      "step": 13934
+    },
+    {
+      "epoch": 2.2885266766571553,
+      "grad_norm": 0.3634662074018649,
+      "learning_rate": 5.171628692492049e-06,
+      "loss": 0.4398,
+      "step": 13935
+    },
+    {
+      "epoch": 2.288690903865498,
+      "grad_norm": 0.4549972944347865,
+      "learning_rate": 5.171164969430987e-06,
+      "loss": 0.4439,
+      "step": 13936
+    },
+    {
+      "epoch": 2.2888551310738405,
+      "grad_norm": 0.40498915261239143,
+      "learning_rate": 5.1707012390804125e-06,
+      "loss": 0.4444,
+      "step": 13937
+    },
+    {
+      "epoch": 2.289019358282183,
+      "grad_norm": 0.4218955172962593,
+      "learning_rate": 5.1702375014453645e-06,
+      "loss": 0.4519,
+      "step": 13938
+    },
+    {
+      "epoch": 2.2891835854905263,
+      "grad_norm": 0.3348184465306751,
+      "learning_rate": 5.1697737565308755e-06,
+      "loss": 0.4556,
+      "step": 13939
+    },
+    {
+      "epoch": 2.289347812698869,
+      "grad_norm": 0.3186255812768499,
+      "learning_rate": 5.169310004341987e-06,
+      "loss": 0.4308,
+      "step": 13940
+    },
+    {
+      "epoch": 2.2895120399072115,
+      "grad_norm": 0.38078161937545435,
+      "learning_rate": 5.1688462448837315e-06,
+      "loss": 0.4242,
+      "step": 13941
+    },
+    {
+      "epoch": 2.2896762671155546,
+      "grad_norm": 0.2922132896482114,
+      "learning_rate": 5.168382478161149e-06,
+      "loss": 0.4328,
+      "step": 13942
+    },
+    {
+      "epoch": 2.2898404943238972,
+      "grad_norm": 0.34740487598494213,
+      "learning_rate": 5.167918704179275e-06,
+      "loss": 0.4308,
+      "step": 13943
+    },
+    {
+      "epoch": 2.29000472153224,
+      "grad_norm": 0.3537760593250267,
+      "learning_rate": 5.167454922943146e-06,
+      "loss": 0.4307,
+      "step": 13944
+    },
+    {
+      "epoch": 2.2901689487405825,
+      "grad_norm": 0.2928523746459882,
+      "learning_rate": 5.1669911344578e-06,
+      "loss": 0.4521,
+      "step": 13945
+    },
+    {
+      "epoch": 2.290333175948925,
+      "grad_norm": 0.47139678614977115,
+      "learning_rate": 5.166527338728275e-06,
+      "loss": 0.4305,
+      "step": 13946
+    },
+    {
+      "epoch": 2.2904974031572682,
+      "grad_norm": 0.2985657027832332,
+      "learning_rate": 5.166063535759606e-06,
+      "loss": 0.4491,
+      "step": 13947
+    },
+    {
+      "epoch": 2.290661630365611,
+      "grad_norm": 0.3731853926343701,
+      "learning_rate": 5.165599725556832e-06,
+      "loss": 0.4415,
+      "step": 13948
+    },
+    {
+      "epoch": 2.2908258575739535,
+      "grad_norm": 0.31532136412340483,
+      "learning_rate": 5.165135908124991e-06,
+      "loss": 0.4377,
+      "step": 13949
+    },
+    {
+      "epoch": 2.290990084782296,
+      "grad_norm": 1.126183260087053,
+      "learning_rate": 5.1646720834691185e-06,
+      "loss": 0.433,
+      "step": 13950
+    },
+    {
+      "epoch": 2.2911543119906392,
+      "grad_norm": 0.2954449428221947,
+      "learning_rate": 5.164208251594255e-06,
+      "loss": 0.4184,
+      "step": 13951
+    },
+    {
+      "epoch": 2.291318539198982,
+      "grad_norm": 0.3198516287960699,
+      "learning_rate": 5.163744412505434e-06,
+      "loss": 0.4518,
+      "step": 13952
+    },
+    {
+      "epoch": 2.2914827664073245,
+      "grad_norm": 0.417184169928783,
+      "learning_rate": 5.163280566207697e-06,
+      "loss": 0.4285,
+      "step": 13953
+    },
+    {
+      "epoch": 2.291646993615667,
+      "grad_norm": 0.3120139219686837,
+      "learning_rate": 5.162816712706081e-06,
+      "loss": 0.4569,
+      "step": 13954
+    },
+    {
+      "epoch": 2.29181122082401,
+      "grad_norm": 0.33460930796334654,
+      "learning_rate": 5.162352852005622e-06,
+      "loss": 0.4061,
+      "step": 13955
+    },
+    {
+      "epoch": 2.291975448032353,
+      "grad_norm": 0.5502547933745806,
+      "learning_rate": 5.16188898411136e-06,
+      "loss": 0.4349,
+      "step": 13956
+    },
+    {
+      "epoch": 2.2921396752406955,
+      "grad_norm": 0.3768704001174054,
+      "learning_rate": 5.161425109028332e-06,
+      "loss": 0.4496,
+      "step": 13957
+    },
+    {
+      "epoch": 2.292303902449038,
+      "grad_norm": 0.36340695623770736,
+      "learning_rate": 5.160961226761576e-06,
+      "loss": 0.4414,
+      "step": 13958
+    },
+    {
+      "epoch": 2.2924681296573812,
+      "grad_norm": 0.40539103376015434,
+      "learning_rate": 5.160497337316133e-06,
+      "loss": 0.4497,
+      "step": 13959
+    },
+    {
+      "epoch": 2.292632356865724,
+      "grad_norm": 0.3861065338987038,
+      "learning_rate": 5.160033440697038e-06,
+      "loss": 0.4512,
+      "step": 13960
+    },
+    {
+      "epoch": 2.2927965840740665,
+      "grad_norm": 0.2884253200044417,
+      "learning_rate": 5.15956953690933e-06,
+      "loss": 0.4278,
+      "step": 13961
+    },
+    {
+      "epoch": 2.292960811282409,
+      "grad_norm": 0.30611232750057144,
+      "learning_rate": 5.159105625958048e-06,
+      "loss": 0.4382,
+      "step": 13962
+    },
+    {
+      "epoch": 2.2931250384907518,
+      "grad_norm": 0.34316335987687846,
+      "learning_rate": 5.158641707848231e-06,
+      "loss": 0.4671,
+      "step": 13963
+    },
+    {
+      "epoch": 2.293289265699095,
+      "grad_norm": 0.3970977754941982,
+      "learning_rate": 5.158177782584917e-06,
+      "loss": 0.4288,
+      "step": 13964
+    },
+    {
+      "epoch": 2.2934534929074375,
+      "grad_norm": 0.3403187216231823,
+      "learning_rate": 5.1577138501731435e-06,
+      "loss": 0.4418,
+      "step": 13965
+    },
+    {
+      "epoch": 2.29361772011578,
+      "grad_norm": 0.5190818158430706,
+      "learning_rate": 5.1572499106179515e-06,
+      "loss": 0.4337,
+      "step": 13966
+    },
+    {
+      "epoch": 2.2937819473241228,
+      "grad_norm": 0.3232592487442511,
+      "learning_rate": 5.156785963924378e-06,
+      "loss": 0.4505,
+      "step": 13967
+    },
+    {
+      "epoch": 2.293946174532466,
+      "grad_norm": 0.2813019946585075,
+      "learning_rate": 5.156322010097464e-06,
+      "loss": 0.4342,
+      "step": 13968
+    },
+    {
+      "epoch": 2.2941104017408085,
+      "grad_norm": 0.3356722636096625,
+      "learning_rate": 5.155858049142247e-06,
+      "loss": 0.4353,
+      "step": 13969
+    },
+    {
+      "epoch": 2.294274628949151,
+      "grad_norm": 0.2909214828377573,
+      "learning_rate": 5.155394081063766e-06,
+      "loss": 0.4321,
+      "step": 13970
+    },
+    {
+      "epoch": 2.2944388561574938,
+      "grad_norm": 0.3733982655008716,
+      "learning_rate": 5.154930105867061e-06,
+      "loss": 0.472,
+      "step": 13971
+    },
+    {
+      "epoch": 2.2946030833658364,
+      "grad_norm": 0.4654045248653441,
+      "learning_rate": 5.154466123557169e-06,
+      "loss": 0.4429,
+      "step": 13972
+    },
+    {
+      "epoch": 2.2947673105741795,
+      "grad_norm": 0.3637665009212365,
+      "learning_rate": 5.154002134139132e-06,
+      "loss": 0.4238,
+      "step": 13973
+    },
+    {
+      "epoch": 2.294931537782522,
+      "grad_norm": 0.2909657126305831,
+      "learning_rate": 5.15353813761799e-06,
+      "loss": 0.4607,
+      "step": 13974
+    },
+    {
+      "epoch": 2.2950957649908648,
+      "grad_norm": 0.42400086471079623,
+      "learning_rate": 5.153074133998778e-06,
+      "loss": 0.4594,
+      "step": 13975
+    },
+    {
+      "epoch": 2.295259992199208,
+      "grad_norm": 0.36972331015652127,
+      "learning_rate": 5.15261012328654e-06,
+      "loss": 0.4262,
+      "step": 13976
+    },
+    {
+      "epoch": 2.2954242194075505,
+      "grad_norm": 0.4389775606377693,
+      "learning_rate": 5.152146105486313e-06,
+      "loss": 0.4249,
+      "step": 13977
+    },
+    {
+      "epoch": 2.295588446615893,
+      "grad_norm": 0.33415273938695683,
+      "learning_rate": 5.1516820806031395e-06,
+      "loss": 0.4354,
+      "step": 13978
+    },
+    {
+      "epoch": 2.2957526738242358,
+      "grad_norm": 0.4278522839527496,
+      "learning_rate": 5.151218048642055e-06,
+      "loss": 0.4476,
+      "step": 13979
+    },
+    {
+      "epoch": 2.2959169010325784,
+      "grad_norm": 0.3741320360256869,
+      "learning_rate": 5.150754009608105e-06,
+      "loss": 0.4484,
+      "step": 13980
+    },
+    {
+      "epoch": 2.2960811282409215,
+      "grad_norm": 0.311783369735317,
+      "learning_rate": 5.150289963506323e-06,
+      "loss": 0.4509,
+      "step": 13981
+    },
+    {
+      "epoch": 2.296245355449264,
+      "grad_norm": 0.5405852343668192,
+      "learning_rate": 5.149825910341753e-06,
+      "loss": 0.4455,
+      "step": 13982
+    },
+    {
+      "epoch": 2.2964095826576068,
+      "grad_norm": 0.3519118169672042,
+      "learning_rate": 5.1493618501194355e-06,
+      "loss": 0.4337,
+      "step": 13983
+    },
+    {
+      "epoch": 2.2965738098659494,
+      "grad_norm": 0.28809793111921655,
+      "learning_rate": 5.148897782844409e-06,
+      "loss": 0.4485,
+      "step": 13984
+    },
+    {
+      "epoch": 2.2967380370742925,
+      "grad_norm": 0.37024335296106775,
+      "learning_rate": 5.1484337085217134e-06,
+      "loss": 0.4317,
+      "step": 13985
+    },
+    {
+      "epoch": 2.296902264282635,
+      "grad_norm": 0.38276051392136357,
+      "learning_rate": 5.14796962715639e-06,
+      "loss": 0.4423,
+      "step": 13986
+    },
+    {
+      "epoch": 2.2970664914909777,
+      "grad_norm": 0.29217104055044624,
+      "learning_rate": 5.147505538753478e-06,
+      "loss": 0.4473,
+      "step": 13987
+    },
+    {
+      "epoch": 2.2972307186993204,
+      "grad_norm": 0.47709697511449556,
+      "learning_rate": 5.147041443318021e-06,
+      "loss": 0.4508,
+      "step": 13988
+    },
+    {
+      "epoch": 2.297394945907663,
+      "grad_norm": 0.30567178053755234,
+      "learning_rate": 5.146577340855056e-06,
+      "loss": 0.4429,
+      "step": 13989
+    },
+    {
+      "epoch": 2.297559173116006,
+      "grad_norm": 0.3211430439562128,
+      "learning_rate": 5.146113231369625e-06,
+      "loss": 0.4527,
+      "step": 13990
+    },
+    {
+      "epoch": 2.2977234003243487,
+      "grad_norm": 0.30219116042974015,
+      "learning_rate": 5.145649114866768e-06,
+      "loss": 0.457,
+      "step": 13991
+    },
+    {
+      "epoch": 2.2978876275326914,
+      "grad_norm": 0.2898625848815241,
+      "learning_rate": 5.145184991351529e-06,
+      "loss": 0.4148,
+      "step": 13992
+    },
+    {
+      "epoch": 2.2980518547410345,
+      "grad_norm": 1.1202762278001295,
+      "learning_rate": 5.144720860828944e-06,
+      "loss": 0.4187,
+      "step": 13993
+    },
+    {
+      "epoch": 2.298216081949377,
+      "grad_norm": 0.5445625866550631,
+      "learning_rate": 5.144256723304056e-06,
+      "loss": 0.4431,
+      "step": 13994
+    },
+    {
+      "epoch": 2.2983803091577197,
+      "grad_norm": 0.35936423083423646,
+      "learning_rate": 5.143792578781906e-06,
+      "loss": 0.4579,
+      "step": 13995
+    },
+    {
+      "epoch": 2.2985445363660624,
+      "grad_norm": 0.3387039272937696,
+      "learning_rate": 5.143328427267535e-06,
+      "loss": 0.4532,
+      "step": 13996
+    },
+    {
+      "epoch": 2.298708763574405,
+      "grad_norm": 0.35295839414607105,
+      "learning_rate": 5.142864268765985e-06,
+      "loss": 0.4416,
+      "step": 13997
+    },
+    {
+      "epoch": 2.298872990782748,
+      "grad_norm": 0.28125146920926414,
+      "learning_rate": 5.142400103282298e-06,
+      "loss": 0.4393,
+      "step": 13998
+    },
+    {
+      "epoch": 2.2990372179910907,
+      "grad_norm": 0.4012659500929637,
+      "learning_rate": 5.141935930821512e-06,
+      "loss": 0.4453,
+      "step": 13999
+    },
+    {
+      "epoch": 2.2992014451994334,
+      "grad_norm": 0.3300866074431469,
+      "learning_rate": 5.14147175138867e-06,
+      "loss": 0.4331,
+      "step": 14000
+    },
+    {
+      "epoch": 2.299365672407776,
+      "grad_norm": 0.4325211113663812,
+      "learning_rate": 5.1410075649888156e-06,
+      "loss": 0.4416,
+      "step": 14001
+    },
+    {
+      "epoch": 2.299529899616119,
+      "grad_norm": 0.32184158336621926,
+      "learning_rate": 5.1405433716269865e-06,
+      "loss": 0.4513,
+      "step": 14002
+    },
+    {
+      "epoch": 2.2996941268244617,
+      "grad_norm": 0.35911228396405964,
+      "learning_rate": 5.140079171308228e-06,
+      "loss": 0.4395,
+      "step": 14003
+    },
+    {
+      "epoch": 2.2998583540328044,
+      "grad_norm": 0.3489742010041331,
+      "learning_rate": 5.139614964037577e-06,
+      "loss": 0.4434,
+      "step": 14004
+    },
+    {
+      "epoch": 2.300022581241147,
+      "grad_norm": 0.37801129331195343,
+      "learning_rate": 5.139150749820081e-06,
+      "loss": 0.4453,
+      "step": 14005
+    },
+    {
+      "epoch": 2.3001868084494896,
+      "grad_norm": 0.3762648304082034,
+      "learning_rate": 5.138686528660778e-06,
+      "loss": 0.4245,
+      "step": 14006
+    },
+    {
+      "epoch": 2.3003510356578327,
+      "grad_norm": 0.33802117667366877,
+      "learning_rate": 5.1382223005647094e-06,
+      "loss": 0.4454,
+      "step": 14007
+    },
+    {
+      "epoch": 2.3005152628661754,
+      "grad_norm": 0.4124698432484217,
+      "learning_rate": 5.137758065536921e-06,
+      "loss": 0.4344,
+      "step": 14008
+    },
+    {
+      "epoch": 2.300679490074518,
+      "grad_norm": 0.4053973640415479,
+      "learning_rate": 5.137293823582452e-06,
+      "loss": 0.4372,
+      "step": 14009
+    },
+    {
+      "epoch": 2.300843717282861,
+      "grad_norm": 0.32205027479758536,
+      "learning_rate": 5.1368295747063455e-06,
+      "loss": 0.4201,
+      "step": 14010
+    },
+    {
+      "epoch": 2.3010079444912037,
+      "grad_norm": 0.3158597315788733,
+      "learning_rate": 5.136365318913641e-06,
+      "loss": 0.4489,
+      "step": 14011
+    },
+    {
+      "epoch": 2.3011721716995464,
+      "grad_norm": 0.37379440368033845,
+      "learning_rate": 5.1359010562093855e-06,
+      "loss": 0.4481,
+      "step": 14012
+    },
+    {
+      "epoch": 2.301336398907889,
+      "grad_norm": 0.3371250590690597,
+      "learning_rate": 5.135436786598619e-06,
+      "loss": 0.4418,
+      "step": 14013
+    },
+    {
+      "epoch": 2.3015006261162316,
+      "grad_norm": 0.3469445965971112,
+      "learning_rate": 5.134972510086382e-06,
+      "loss": 0.4428,
+      "step": 14014
+    },
+    {
+      "epoch": 2.3016648533245747,
+      "grad_norm": 0.35999734715615955,
+      "learning_rate": 5.134508226677719e-06,
+      "loss": 0.4387,
+      "step": 14015
+    },
+    {
+      "epoch": 2.3018290805329173,
+      "grad_norm": 0.3987865712873361,
+      "learning_rate": 5.134043936377672e-06,
+      "loss": 0.4541,
+      "step": 14016
+    },
+    {
+      "epoch": 2.30199330774126,
+      "grad_norm": 0.30289366918869837,
+      "learning_rate": 5.133579639191286e-06,
+      "loss": 0.4473,
+      "step": 14017
+    },
+    {
+      "epoch": 2.3021575349496026,
+      "grad_norm": 0.403215916390885,
+      "learning_rate": 5.133115335123601e-06,
+      "loss": 0.4545,
+      "step": 14018
+    },
+    {
+      "epoch": 2.3023217621579457,
+      "grad_norm": 0.34769404599655834,
+      "learning_rate": 5.13265102417966e-06,
+      "loss": 0.4373,
+      "step": 14019
+    },
+    {
+      "epoch": 2.3024859893662883,
+      "grad_norm": 0.4296539613249186,
+      "learning_rate": 5.132186706364507e-06,
+      "loss": 0.4341,
+      "step": 14020
+    },
+    {
+      "epoch": 2.302650216574631,
+      "grad_norm": 0.3154857891196642,
+      "learning_rate": 5.131722381683183e-06,
+      "loss": 0.4581,
+      "step": 14021
+    },
+    {
+      "epoch": 2.3028144437829736,
+      "grad_norm": 0.3455431355471438,
+      "learning_rate": 5.131258050140734e-06,
+      "loss": 0.4238,
+      "step": 14022
+    },
+    {
+      "epoch": 2.3029786709913163,
+      "grad_norm": 0.5078514368461862,
+      "learning_rate": 5.130793711742201e-06,
+      "loss": 0.4437,
+      "step": 14023
+    },
+    {
+      "epoch": 2.3031428981996593,
+      "grad_norm": 0.34871399726253866,
+      "learning_rate": 5.130329366492628e-06,
+      "loss": 0.4279,
+      "step": 14024
+    },
+    {
+      "epoch": 2.303307125408002,
+      "grad_norm": 0.4938279579093215,
+      "learning_rate": 5.129865014397057e-06,
+      "loss": 0.4359,
+      "step": 14025
+    },
+    {
+      "epoch": 2.3034713526163446,
+      "grad_norm": 0.4455533239162793,
+      "learning_rate": 5.129400655460533e-06,
+      "loss": 0.4302,
+      "step": 14026
+    },
+    {
+      "epoch": 2.3036355798246877,
+      "grad_norm": 0.317398104299277,
+      "learning_rate": 5.1289362896881e-06,
+      "loss": 0.4441,
+      "step": 14027
+    },
+    {
+      "epoch": 2.3037998070330303,
+      "grad_norm": 0.3335278163989015,
+      "learning_rate": 5.128471917084798e-06,
+      "loss": 0.4409,
+      "step": 14028
+    },
+    {
+      "epoch": 2.303964034241373,
+      "grad_norm": 0.3116408451208168,
+      "learning_rate": 5.128007537655673e-06,
+      "loss": 0.4565,
+      "step": 14029
+    },
+    {
+      "epoch": 2.3041282614497156,
+      "grad_norm": 0.3616035436306267,
+      "learning_rate": 5.1275431514057686e-06,
+      "loss": 0.4402,
+      "step": 14030
+    },
+    {
+      "epoch": 2.3042924886580582,
+      "grad_norm": 0.3156254049719913,
+      "learning_rate": 5.127078758340128e-06,
+      "loss": 0.4342,
+      "step": 14031
+    },
+    {
+      "epoch": 2.3044567158664013,
+      "grad_norm": 0.37717368073785584,
+      "learning_rate": 5.126614358463795e-06,
+      "loss": 0.4355,
+      "step": 14032
+    },
+    {
+      "epoch": 2.304620943074744,
+      "grad_norm": 0.6232263943964357,
+      "learning_rate": 5.126149951781814e-06,
+      "loss": 0.4383,
+      "step": 14033
+    },
+    {
+      "epoch": 2.3047851702830866,
+      "grad_norm": 0.3038821240087618,
+      "learning_rate": 5.1256855382992285e-06,
+      "loss": 0.4616,
+      "step": 14034
+    },
+    {
+      "epoch": 2.3049493974914292,
+      "grad_norm": 0.3415423211697857,
+      "learning_rate": 5.125221118021082e-06,
+      "loss": 0.437,
+      "step": 14035
+    },
+    {
+      "epoch": 2.3051136246997723,
+      "grad_norm": 0.273495773934862,
+      "learning_rate": 5.124756690952418e-06,
+      "loss": 0.4553,
+      "step": 14036
+    },
+    {
+      "epoch": 2.305277851908115,
+      "grad_norm": 0.32609384140101105,
+      "learning_rate": 5.124292257098284e-06,
+      "loss": 0.4358,
+      "step": 14037
+    },
+    {
+      "epoch": 2.3054420791164576,
+      "grad_norm": 0.29790082821128433,
+      "learning_rate": 5.123827816463722e-06,
+      "loss": 0.4455,
+      "step": 14038
+    },
+    {
+      "epoch": 2.3056063063248002,
+      "grad_norm": 0.4301738879695801,
+      "learning_rate": 5.123363369053774e-06,
+      "loss": 0.4351,
+      "step": 14039
+    },
+    {
+      "epoch": 2.305770533533143,
+      "grad_norm": 0.4560846085606717,
+      "learning_rate": 5.122898914873487e-06,
+      "loss": 0.4325,
+      "step": 14040
+    },
+    {
+      "epoch": 2.305934760741486,
+      "grad_norm": 0.3587378842032936,
+      "learning_rate": 5.122434453927905e-06,
+      "loss": 0.4325,
+      "step": 14041
+    },
+    {
+      "epoch": 2.3060989879498286,
+      "grad_norm": 0.4712292651053053,
+      "learning_rate": 5.121969986222074e-06,
+      "loss": 0.4323,
+      "step": 14042
+    },
+    {
+      "epoch": 2.3062632151581712,
+      "grad_norm": 0.40531640859089124,
+      "learning_rate": 5.121505511761036e-06,
+      "loss": 0.4324,
+      "step": 14043
+    },
+    {
+      "epoch": 2.3064274423665143,
+      "grad_norm": 0.28030142930937785,
+      "learning_rate": 5.121041030549835e-06,
+      "loss": 0.4453,
+      "step": 14044
+    },
+    {
+      "epoch": 2.306591669574857,
+      "grad_norm": 0.3595351679875266,
+      "learning_rate": 5.120576542593519e-06,
+      "loss": 0.4409,
+      "step": 14045
+    },
+    {
+      "epoch": 2.3067558967831996,
+      "grad_norm": 0.30426305503665396,
+      "learning_rate": 5.120112047897132e-06,
+      "loss": 0.4528,
+      "step": 14046
+    },
+    {
+      "epoch": 2.3069201239915422,
+      "grad_norm": 0.5848941502153416,
+      "learning_rate": 5.119647546465717e-06,
+      "loss": 0.4311,
+      "step": 14047
+    },
+    {
+      "epoch": 2.307084351199885,
+      "grad_norm": 0.32612350991860606,
+      "learning_rate": 5.11918303830432e-06,
+      "loss": 0.4377,
+      "step": 14048
+    },
+    {
+      "epoch": 2.307248578408228,
+      "grad_norm": 0.2763098121789989,
+      "learning_rate": 5.118718523417985e-06,
+      "loss": 0.4235,
+      "step": 14049
+    },
+    {
+      "epoch": 2.3074128056165706,
+      "grad_norm": 0.3295502056022296,
+      "learning_rate": 5.118254001811759e-06,
+      "loss": 0.4374,
+      "step": 14050
+    },
+    {
+      "epoch": 2.307577032824913,
+      "grad_norm": 0.3994712281531475,
+      "learning_rate": 5.117789473490688e-06,
+      "loss": 0.4412,
+      "step": 14051
+    },
+    {
+      "epoch": 2.307741260033256,
+      "grad_norm": 0.3115856036129265,
+      "learning_rate": 5.117324938459813e-06,
+      "loss": 0.4284,
+      "step": 14052
+    },
+    {
+      "epoch": 2.307905487241599,
+      "grad_norm": 0.9618678183501287,
+      "learning_rate": 5.116860396724183e-06,
+      "loss": 0.451,
+      "step": 14053
+    },
+    {
+      "epoch": 2.3080697144499416,
+      "grad_norm": 0.3948070348891769,
+      "learning_rate": 5.116395848288842e-06,
+      "loss": 0.4334,
+      "step": 14054
+    },
+    {
+      "epoch": 2.308233941658284,
+      "grad_norm": 0.3034880244011292,
+      "learning_rate": 5.115931293158835e-06,
+      "loss": 0.4449,
+      "step": 14055
+    },
+    {
+      "epoch": 2.308398168866627,
+      "grad_norm": 0.31571765021660086,
+      "learning_rate": 5.11546673133921e-06,
+      "loss": 0.4207,
+      "step": 14056
+    },
+    {
+      "epoch": 2.3085623960749695,
+      "grad_norm": 0.29877701295733133,
+      "learning_rate": 5.1150021628350095e-06,
+      "loss": 0.4526,
+      "step": 14057
+    },
+    {
+      "epoch": 2.3087266232833126,
+      "grad_norm": 0.34530395459750013,
+      "learning_rate": 5.114537587651279e-06,
+      "loss": 0.4274,
+      "step": 14058
+    },
+    {
+      "epoch": 2.308890850491655,
+      "grad_norm": 0.369050788922719,
+      "learning_rate": 5.114073005793068e-06,
+      "loss": 0.4483,
+      "step": 14059
+    },
+    {
+      "epoch": 2.309055077699998,
+      "grad_norm": 0.28538396109289,
+      "learning_rate": 5.113608417265419e-06,
+      "loss": 0.4338,
+      "step": 14060
+    },
+    {
+      "epoch": 2.309219304908341,
+      "grad_norm": 0.32589537320587636,
+      "learning_rate": 5.113143822073379e-06,
+      "loss": 0.4394,
+      "step": 14061
+    },
+    {
+      "epoch": 2.3093835321166836,
+      "grad_norm": 0.32042454658452446,
+      "learning_rate": 5.1126792202219936e-06,
+      "loss": 0.4452,
+      "step": 14062
+    },
+    {
+      "epoch": 2.309547759325026,
+      "grad_norm": 0.33793205302871954,
+      "learning_rate": 5.11221461171631e-06,
+      "loss": 0.4362,
+      "step": 14063
+    },
+    {
+      "epoch": 2.309711986533369,
+      "grad_norm": 0.2879793855044109,
+      "learning_rate": 5.111749996561371e-06,
+      "loss": 0.4442,
+      "step": 14064
+    },
+    {
+      "epoch": 2.3098762137417115,
+      "grad_norm": 0.3600649860400257,
+      "learning_rate": 5.111285374762227e-06,
+      "loss": 0.4214,
+      "step": 14065
+    },
+    {
+      "epoch": 2.3100404409500546,
+      "grad_norm": 0.3584892274240055,
+      "learning_rate": 5.110820746323924e-06,
+      "loss": 0.4371,
+      "step": 14066
+    },
+    {
+      "epoch": 2.310204668158397,
+      "grad_norm": 0.3154401212773581,
+      "learning_rate": 5.1103561112515035e-06,
+      "loss": 0.4408,
+      "step": 14067
+    },
+    {
+      "epoch": 2.31036889536674,
+      "grad_norm": 0.42719130845995357,
+      "learning_rate": 5.109891469550018e-06,
+      "loss": 0.4538,
+      "step": 14068
+    },
+    {
+      "epoch": 2.3105331225750825,
+      "grad_norm": 0.33674862607670436,
+      "learning_rate": 5.109426821224509e-06,
+      "loss": 0.4304,
+      "step": 14069
+    },
+    {
+      "epoch": 2.3106973497834256,
+      "grad_norm": 0.33291473916346687,
+      "learning_rate": 5.108962166280025e-06,
+      "loss": 0.4515,
+      "step": 14070
+    },
+    {
+      "epoch": 2.310861576991768,
+      "grad_norm": 0.34892866493515384,
+      "learning_rate": 5.108497504721614e-06,
+      "loss": 0.4331,
+      "step": 14071
+    },
+    {
+      "epoch": 2.311025804200111,
+      "grad_norm": 0.300162383893827,
+      "learning_rate": 5.108032836554321e-06,
+      "loss": 0.4546,
+      "step": 14072
+    },
+    {
+      "epoch": 2.3111900314084535,
+      "grad_norm": 0.28162989647215003,
+      "learning_rate": 5.107568161783193e-06,
+      "loss": 0.4533,
+      "step": 14073
+    },
+    {
+      "epoch": 2.311354258616796,
+      "grad_norm": 0.38128347717669486,
+      "learning_rate": 5.107103480413277e-06,
+      "loss": 0.4519,
+      "step": 14074
+    },
+    {
+      "epoch": 2.311518485825139,
+      "grad_norm": 0.29761447186213036,
+      "learning_rate": 5.106638792449619e-06,
+      "loss": 0.4547,
+      "step": 14075
+    },
+    {
+      "epoch": 2.311682713033482,
+      "grad_norm": 0.40785614066545356,
+      "learning_rate": 5.10617409789727e-06,
+      "loss": 0.419,
+      "step": 14076
+    },
+    {
+      "epoch": 2.3118469402418245,
+      "grad_norm": 0.35949262056041725,
+      "learning_rate": 5.105709396761271e-06,
+      "loss": 0.4461,
+      "step": 14077
+    },
+    {
+      "epoch": 2.3120111674501675,
+      "grad_norm": 0.460989553636019,
+      "learning_rate": 5.105244689046672e-06,
+      "loss": 0.4462,
+      "step": 14078
+    },
+    {
+      "epoch": 2.31217539465851,
+      "grad_norm": 0.2828943692652613,
+      "learning_rate": 5.104779974758521e-06,
+      "loss": 0.4257,
+      "step": 14079
+    },
+    {
+      "epoch": 2.312339621866853,
+      "grad_norm": 0.3003531542996423,
+      "learning_rate": 5.1043152539018645e-06,
+      "loss": 0.4605,
+      "step": 14080
+    },
+    {
+      "epoch": 2.3125038490751955,
+      "grad_norm": 0.3975382133409971,
+      "learning_rate": 5.103850526481751e-06,
+      "loss": 0.4359,
+      "step": 14081
+    },
+    {
+      "epoch": 2.312668076283538,
+      "grad_norm": 0.32455473999813356,
+      "learning_rate": 5.103385792503224e-06,
+      "loss": 0.4535,
+      "step": 14082
+    },
+    {
+      "epoch": 2.312832303491881,
+      "grad_norm": 0.3051514804701622,
+      "learning_rate": 5.102921051971335e-06,
+      "loss": 0.465,
+      "step": 14083
+    },
+    {
+      "epoch": 2.312996530700224,
+      "grad_norm": 0.3355561230778918,
+      "learning_rate": 5.102456304891131e-06,
+      "loss": 0.4458,
+      "step": 14084
+    },
+    {
+      "epoch": 2.3131607579085665,
+      "grad_norm": 0.37683145351901637,
+      "learning_rate": 5.101991551267657e-06,
+      "loss": 0.4505,
+      "step": 14085
+    },
+    {
+      "epoch": 2.313324985116909,
+      "grad_norm": 0.37557984452306886,
+      "learning_rate": 5.101526791105964e-06,
+      "loss": 0.4602,
+      "step": 14086
+    },
+    {
+      "epoch": 2.313489212325252,
+      "grad_norm": 0.3750722304582995,
+      "learning_rate": 5.101062024411098e-06,
+      "loss": 0.4527,
+      "step": 14087
+    },
+    {
+      "epoch": 2.313653439533595,
+      "grad_norm": 0.27858798055563705,
+      "learning_rate": 5.100597251188107e-06,
+      "loss": 0.4474,
+      "step": 14088
+    },
+    {
+      "epoch": 2.3138176667419375,
+      "grad_norm": 0.299630465893534,
+      "learning_rate": 5.100132471442038e-06,
+      "loss": 0.4342,
+      "step": 14089
+    },
+    {
+      "epoch": 2.31398189395028,
+      "grad_norm": 0.3387134968545231,
+      "learning_rate": 5.0996676851779405e-06,
+      "loss": 0.4392,
+      "step": 14090
+    },
+    {
+      "epoch": 2.3141461211586227,
+      "grad_norm": 0.4640134154275601,
+      "learning_rate": 5.099202892400863e-06,
+      "loss": 0.4295,
+      "step": 14091
+    },
+    {
+      "epoch": 2.314310348366966,
+      "grad_norm": 0.33222568022393845,
+      "learning_rate": 5.098738093115851e-06,
+      "loss": 0.4542,
+      "step": 14092
+    },
+    {
+      "epoch": 2.3144745755753084,
+      "grad_norm": 0.4566089439660268,
+      "learning_rate": 5.098273287327954e-06,
+      "loss": 0.4321,
+      "step": 14093
+    },
+    {
+      "epoch": 2.314638802783651,
+      "grad_norm": 0.3340150445386211,
+      "learning_rate": 5.097808475042221e-06,
+      "loss": 0.4348,
+      "step": 14094
+    },
+    {
+      "epoch": 2.314803029991994,
+      "grad_norm": 0.31587059691726316,
+      "learning_rate": 5.097343656263701e-06,
+      "loss": 0.44,
+      "step": 14095
+    },
+    {
+      "epoch": 2.314967257200337,
+      "grad_norm": 0.29369202918284504,
+      "learning_rate": 5.0968788309974405e-06,
+      "loss": 0.4436,
+      "step": 14096
+    },
+    {
+      "epoch": 2.3151314844086794,
+      "grad_norm": 0.3455351361430038,
+      "learning_rate": 5.096413999248489e-06,
+      "loss": 0.4494,
+      "step": 14097
+    },
+    {
+      "epoch": 2.315295711617022,
+      "grad_norm": 0.29585478735310294,
+      "learning_rate": 5.095949161021894e-06,
+      "loss": 0.4273,
+      "step": 14098
+    },
+    {
+      "epoch": 2.3154599388253647,
+      "grad_norm": 0.26622562749246065,
+      "learning_rate": 5.0954843163227035e-06,
+      "loss": 0.435,
+      "step": 14099
+    },
+    {
+      "epoch": 2.315624166033708,
+      "grad_norm": 0.2716855705390083,
+      "learning_rate": 5.09501946515597e-06,
+      "loss": 0.4576,
+      "step": 14100
+    },
+    {
+      "epoch": 2.3157883932420504,
+      "grad_norm": 0.38750734835384804,
+      "learning_rate": 5.094554607526738e-06,
+      "loss": 0.4542,
+      "step": 14101
+    },
+    {
+      "epoch": 2.315952620450393,
+      "grad_norm": 0.39359876643785513,
+      "learning_rate": 5.094089743440059e-06,
+      "loss": 0.4345,
+      "step": 14102
+    },
+    {
+      "epoch": 2.3161168476587357,
+      "grad_norm": 0.2808952967455288,
+      "learning_rate": 5.09362487290098e-06,
+      "loss": 0.4414,
+      "step": 14103
+    },
+    {
+      "epoch": 2.316281074867079,
+      "grad_norm": 0.3194014628944141,
+      "learning_rate": 5.093159995914551e-06,
+      "loss": 0.4424,
+      "step": 14104
+    },
+    {
+      "epoch": 2.3164453020754214,
+      "grad_norm": 0.4332979821627784,
+      "learning_rate": 5.092695112485822e-06,
+      "loss": 0.4527,
+      "step": 14105
+    },
+    {
+      "epoch": 2.316609529283764,
+      "grad_norm": 0.5310827151398166,
+      "learning_rate": 5.09223022261984e-06,
+      "loss": 0.446,
+      "step": 14106
+    },
+    {
+      "epoch": 2.3167737564921067,
+      "grad_norm": 0.2786846133670859,
+      "learning_rate": 5.091765326321655e-06,
+      "loss": 0.4418,
+      "step": 14107
+    },
+    {
+      "epoch": 2.3169379837004493,
+      "grad_norm": 0.329155169416799,
+      "learning_rate": 5.091300423596316e-06,
+      "loss": 0.4484,
+      "step": 14108
+    },
+    {
+      "epoch": 2.3171022109087924,
+      "grad_norm": 0.3980959735763232,
+      "learning_rate": 5.0908355144488736e-06,
+      "loss": 0.4526,
+      "step": 14109
+    },
+    {
+      "epoch": 2.317266438117135,
+      "grad_norm": 0.32127663985472377,
+      "learning_rate": 5.090370598884376e-06,
+      "loss": 0.4498,
+      "step": 14110
+    },
+    {
+      "epoch": 2.3174306653254777,
+      "grad_norm": 0.34683145841006363,
+      "learning_rate": 5.089905676907873e-06,
+      "loss": 0.4459,
+      "step": 14111
+    },
+    {
+      "epoch": 2.317594892533821,
+      "grad_norm": 0.43147739958741066,
+      "learning_rate": 5.0894407485244124e-06,
+      "loss": 0.4443,
+      "step": 14112
+    },
+    {
+      "epoch": 2.3177591197421634,
+      "grad_norm": 0.4648840760522442,
+      "learning_rate": 5.0889758137390466e-06,
+      "loss": 0.4456,
+      "step": 14113
+    },
+    {
+      "epoch": 2.317923346950506,
+      "grad_norm": 0.31100822351587076,
+      "learning_rate": 5.0885108725568235e-06,
+      "loss": 0.4501,
+      "step": 14114
+    },
+    {
+      "epoch": 2.3180875741588487,
+      "grad_norm": 0.31343052539393956,
+      "learning_rate": 5.088045924982794e-06,
+      "loss": 0.4409,
+      "step": 14115
+    },
+    {
+      "epoch": 2.3182518013671913,
+      "grad_norm": 0.34227482246820434,
+      "learning_rate": 5.0875809710220075e-06,
+      "loss": 0.4308,
+      "step": 14116
+    },
+    {
+      "epoch": 2.3184160285755344,
+      "grad_norm": 0.27676355153912735,
+      "learning_rate": 5.087116010679511e-06,
+      "loss": 0.4558,
+      "step": 14117
+    },
+    {
+      "epoch": 2.318580255783877,
+      "grad_norm": 0.3073291035116375,
+      "learning_rate": 5.08665104396036e-06,
+      "loss": 0.4345,
+      "step": 14118
+    },
+    {
+      "epoch": 2.3187444829922197,
+      "grad_norm": 0.32193629097359744,
+      "learning_rate": 5.086186070869601e-06,
+      "loss": 0.439,
+      "step": 14119
+    },
+    {
+      "epoch": 2.3189087102005623,
+      "grad_norm": 0.281479804541597,
+      "learning_rate": 5.085721091412284e-06,
+      "loss": 0.4354,
+      "step": 14120
+    },
+    {
+      "epoch": 2.3190729374089054,
+      "grad_norm": 0.35718102691551273,
+      "learning_rate": 5.085256105593459e-06,
+      "loss": 0.424,
+      "step": 14121
+    },
+    {
+      "epoch": 2.319237164617248,
+      "grad_norm": 0.28737312989718716,
+      "learning_rate": 5.084791113418178e-06,
+      "loss": 0.4309,
+      "step": 14122
+    },
+    {
+      "epoch": 2.3194013918255907,
+      "grad_norm": 0.26492875726681686,
+      "learning_rate": 5.08432611489149e-06,
+      "loss": 0.4475,
+      "step": 14123
+    },
+    {
+      "epoch": 2.3195656190339333,
+      "grad_norm": 0.3803862478716756,
+      "learning_rate": 5.083861110018444e-06,
+      "loss": 0.4277,
+      "step": 14124
+    },
+    {
+      "epoch": 2.319729846242276,
+      "grad_norm": 0.3684425492831474,
+      "learning_rate": 5.083396098804093e-06,
+      "loss": 0.4391,
+      "step": 14125
+    },
+    {
+      "epoch": 2.319894073450619,
+      "grad_norm": 0.2797802642574338,
+      "learning_rate": 5.082931081253487e-06,
+      "loss": 0.4504,
+      "step": 14126
+    },
+    {
+      "epoch": 2.3200583006589617,
+      "grad_norm": 0.31374455966117754,
+      "learning_rate": 5.0824660573716756e-06,
+      "loss": 0.432,
+      "step": 14127
+    },
+    {
+      "epoch": 2.3202225278673043,
+      "grad_norm": 0.33890373708963784,
+      "learning_rate": 5.082001027163708e-06,
+      "loss": 0.4218,
+      "step": 14128
+    },
+    {
+      "epoch": 2.3203867550756474,
+      "grad_norm": 0.3227194866548745,
+      "learning_rate": 5.081535990634639e-06,
+      "loss": 0.4614,
+      "step": 14129
+    },
+    {
+      "epoch": 2.32055098228399,
+      "grad_norm": 0.4470375783159691,
+      "learning_rate": 5.081070947789517e-06,
+      "loss": 0.4233,
+      "step": 14130
+    },
+    {
+      "epoch": 2.3207152094923327,
+      "grad_norm": 0.32100958956086817,
+      "learning_rate": 5.080605898633392e-06,
+      "loss": 0.4519,
+      "step": 14131
+    },
+    {
+      "epoch": 2.3208794367006753,
+      "grad_norm": 0.3035182537145447,
+      "learning_rate": 5.080140843171315e-06,
+      "loss": 0.4415,
+      "step": 14132
+    },
+    {
+      "epoch": 2.321043663909018,
+      "grad_norm": 0.38367426741253585,
+      "learning_rate": 5.079675781408337e-06,
+      "loss": 0.4198,
+      "step": 14133
+    },
+    {
+      "epoch": 2.321207891117361,
+      "grad_norm": 0.28865208882220506,
+      "learning_rate": 5.079210713349512e-06,
+      "loss": 0.4461,
+      "step": 14134
+    },
+    {
+      "epoch": 2.3213721183257037,
+      "grad_norm": 0.36621689632654064,
+      "learning_rate": 5.078745638999888e-06,
+      "loss": 0.4345,
+      "step": 14135
+    },
+    {
+      "epoch": 2.3215363455340463,
+      "grad_norm": 0.4429051331999256,
+      "learning_rate": 5.0782805583645165e-06,
+      "loss": 0.4532,
+      "step": 14136
+    },
+    {
+      "epoch": 2.321700572742389,
+      "grad_norm": 0.31515172267910496,
+      "learning_rate": 5.077815471448449e-06,
+      "loss": 0.4311,
+      "step": 14137
+    },
+    {
+      "epoch": 2.321864799950732,
+      "grad_norm": 0.3330602157170481,
+      "learning_rate": 5.077350378256737e-06,
+      "loss": 0.4416,
+      "step": 14138
+    },
+    {
+      "epoch": 2.3220290271590747,
+      "grad_norm": 0.31717514726939133,
+      "learning_rate": 5.076885278794433e-06,
+      "loss": 0.4261,
+      "step": 14139
+    },
+    {
+      "epoch": 2.3221932543674173,
+      "grad_norm": 0.3302356919478289,
+      "learning_rate": 5.0764201730665866e-06,
+      "loss": 0.4234,
+      "step": 14140
+    },
+    {
+      "epoch": 2.32235748157576,
+      "grad_norm": 0.33608792819627553,
+      "learning_rate": 5.0759550610782494e-06,
+      "loss": 0.439,
+      "step": 14141
+    },
+    {
+      "epoch": 2.3225217087841026,
+      "grad_norm": 0.36175993010131097,
+      "learning_rate": 5.075489942834474e-06,
+      "loss": 0.4306,
+      "step": 14142
+    },
+    {
+      "epoch": 2.3226859359924457,
+      "grad_norm": 0.2664133792406407,
+      "learning_rate": 5.075024818340312e-06,
+      "loss": 0.4309,
+      "step": 14143
+    },
+    {
+      "epoch": 2.3228501632007883,
+      "grad_norm": 0.30003761719849853,
+      "learning_rate": 5.0745596876008145e-06,
+      "loss": 0.4443,
+      "step": 14144
+    },
+    {
+      "epoch": 2.323014390409131,
+      "grad_norm": 0.3464197651882214,
+      "learning_rate": 5.074094550621033e-06,
+      "loss": 0.4303,
+      "step": 14145
+    },
+    {
+      "epoch": 2.323178617617474,
+      "grad_norm": 0.304019749705254,
+      "learning_rate": 5.07362940740602e-06,
+      "loss": 0.4499,
+      "step": 14146
+    },
+    {
+      "epoch": 2.3233428448258167,
+      "grad_norm": 0.3001217873675227,
+      "learning_rate": 5.073164257960828e-06,
+      "loss": 0.4441,
+      "step": 14147
+    },
+    {
+      "epoch": 2.3235070720341593,
+      "grad_norm": 0.46064457005809356,
+      "learning_rate": 5.072699102290509e-06,
+      "loss": 0.4646,
+      "step": 14148
+    },
+    {
+      "epoch": 2.323671299242502,
+      "grad_norm": 0.3779222203657981,
+      "learning_rate": 5.072233940400112e-06,
+      "loss": 0.4595,
+      "step": 14149
+    },
+    {
+      "epoch": 2.3238355264508446,
+      "grad_norm": 0.41772714934031113,
+      "learning_rate": 5.071768772294692e-06,
+      "loss": 0.4357,
+      "step": 14150
+    },
+    {
+      "epoch": 2.3239997536591877,
+      "grad_norm": 0.3716025171384937,
+      "learning_rate": 5.0713035979793025e-06,
+      "loss": 0.4315,
+      "step": 14151
+    },
+    {
+      "epoch": 2.3241639808675303,
+      "grad_norm": 0.305323170356014,
+      "learning_rate": 5.070838417458992e-06,
+      "loss": 0.4386,
+      "step": 14152
+    },
+    {
+      "epoch": 2.324328208075873,
+      "grad_norm": 0.39248809891274733,
+      "learning_rate": 5.070373230738815e-06,
+      "loss": 0.4287,
+      "step": 14153
+    },
+    {
+      "epoch": 2.3244924352842156,
+      "grad_norm": 0.38032600364508357,
+      "learning_rate": 5.069908037823823e-06,
+      "loss": 0.4473,
+      "step": 14154
+    },
+    {
+      "epoch": 2.3246566624925586,
+      "grad_norm": 0.33848005664175346,
+      "learning_rate": 5.069442838719071e-06,
+      "loss": 0.449,
+      "step": 14155
+    },
+    {
+      "epoch": 2.3248208897009013,
+      "grad_norm": 0.3314219058532578,
+      "learning_rate": 5.068977633429607e-06,
+      "loss": 0.4281,
+      "step": 14156
+    },
+    {
+      "epoch": 2.324985116909244,
+      "grad_norm": 0.33796106931331144,
+      "learning_rate": 5.068512421960487e-06,
+      "loss": 0.438,
+      "step": 14157
+    },
+    {
+      "epoch": 2.3251493441175866,
+      "grad_norm": 0.3271594956649946,
+      "learning_rate": 5.068047204316763e-06,
+      "loss": 0.4406,
+      "step": 14158
+    },
+    {
+      "epoch": 2.325313571325929,
+      "grad_norm": 0.4212547832783069,
+      "learning_rate": 5.067581980503489e-06,
+      "loss": 0.4389,
+      "step": 14159
+    },
+    {
+      "epoch": 2.3254777985342723,
+      "grad_norm": 0.31153221402017683,
+      "learning_rate": 5.067116750525714e-06,
+      "loss": 0.4556,
+      "step": 14160
+    },
+    {
+      "epoch": 2.325642025742615,
+      "grad_norm": 0.29817177042555487,
+      "learning_rate": 5.066651514388493e-06,
+      "loss": 0.411,
+      "step": 14161
+    },
+    {
+      "epoch": 2.3258062529509576,
+      "grad_norm": 0.5029510933292733,
+      "learning_rate": 5.06618627209688e-06,
+      "loss": 0.4246,
+      "step": 14162
+    },
+    {
+      "epoch": 2.3259704801593006,
+      "grad_norm": 0.3514442344266123,
+      "learning_rate": 5.065721023655927e-06,
+      "loss": 0.4491,
+      "step": 14163
+    },
+    {
+      "epoch": 2.3261347073676433,
+      "grad_norm": 0.4915963134364145,
+      "learning_rate": 5.065255769070687e-06,
+      "loss": 0.4187,
+      "step": 14164
+    },
+    {
+      "epoch": 2.326298934575986,
+      "grad_norm": 0.2865691070834308,
+      "learning_rate": 5.064790508346213e-06,
+      "loss": 0.4314,
+      "step": 14165
+    },
+    {
+      "epoch": 2.3264631617843285,
+      "grad_norm": 0.3247883587286542,
+      "learning_rate": 5.064325241487559e-06,
+      "loss": 0.4283,
+      "step": 14166
+    },
+    {
+      "epoch": 2.326627388992671,
+      "grad_norm": 0.35794421758596046,
+      "learning_rate": 5.063859968499777e-06,
+      "loss": 0.4566,
+      "step": 14167
+    },
+    {
+      "epoch": 2.3267916162010143,
+      "grad_norm": 0.46948198967138616,
+      "learning_rate": 5.063394689387921e-06,
+      "loss": 0.4318,
+      "step": 14168
+    },
+    {
+      "epoch": 2.326955843409357,
+      "grad_norm": 0.30464816469968276,
+      "learning_rate": 5.062929404157046e-06,
+      "loss": 0.4245,
+      "step": 14169
+    },
+    {
+      "epoch": 2.3271200706176995,
+      "grad_norm": 0.4028945435079524,
+      "learning_rate": 5.062464112812202e-06,
+      "loss": 0.416,
+      "step": 14170
+    },
+    {
+      "epoch": 2.327284297826042,
+      "grad_norm": 0.3313330258097169,
+      "learning_rate": 5.061998815358444e-06,
+      "loss": 0.4552,
+      "step": 14171
+    },
+    {
+      "epoch": 2.3274485250343853,
+      "grad_norm": 0.3077777867179962,
+      "learning_rate": 5.061533511800827e-06,
+      "loss": 0.4567,
+      "step": 14172
+    },
+    {
+      "epoch": 2.327612752242728,
+      "grad_norm": 0.29356991243950614,
+      "learning_rate": 5.061068202144404e-06,
+      "loss": 0.4314,
+      "step": 14173
+    },
+    {
+      "epoch": 2.3277769794510705,
+      "grad_norm": 0.3028690657624948,
+      "learning_rate": 5.060602886394227e-06,
+      "loss": 0.4399,
+      "step": 14174
+    },
+    {
+      "epoch": 2.327941206659413,
+      "grad_norm": 0.3859579578351086,
+      "learning_rate": 5.060137564555352e-06,
+      "loss": 0.4357,
+      "step": 14175
+    },
+    {
+      "epoch": 2.328105433867756,
+      "grad_norm": 0.4652567023941955,
+      "learning_rate": 5.0596722366328316e-06,
+      "loss": 0.4401,
+      "step": 14176
+    },
+    {
+      "epoch": 2.328269661076099,
+      "grad_norm": 0.3837296610403499,
+      "learning_rate": 5.059206902631719e-06,
+      "loss": 0.4507,
+      "step": 14177
+    },
+    {
+      "epoch": 2.3284338882844415,
+      "grad_norm": 0.2869324518141024,
+      "learning_rate": 5.0587415625570725e-06,
+      "loss": 0.4319,
+      "step": 14178
+    },
+    {
+      "epoch": 2.328598115492784,
+      "grad_norm": 0.3169258499686373,
+      "learning_rate": 5.05827621641394e-06,
+      "loss": 0.4254,
+      "step": 14179
+    },
+    {
+      "epoch": 2.3287623427011273,
+      "grad_norm": 0.6569998287700313,
+      "learning_rate": 5.057810864207379e-06,
+      "loss": 0.4293,
+      "step": 14180
+    },
+    {
+      "epoch": 2.32892656990947,
+      "grad_norm": 0.4219660136497421,
+      "learning_rate": 5.057345505942444e-06,
+      "loss": 0.4211,
+      "step": 14181
+    },
+    {
+      "epoch": 2.3290907971178125,
+      "grad_norm": 0.3104954360549804,
+      "learning_rate": 5.056880141624187e-06,
+      "loss": 0.4445,
+      "step": 14182
+    },
+    {
+      "epoch": 2.329255024326155,
+      "grad_norm": 0.27808873262914974,
+      "learning_rate": 5.056414771257665e-06,
+      "loss": 0.415,
+      "step": 14183
+    },
+    {
+      "epoch": 2.329419251534498,
+      "grad_norm": 0.38868025370533194,
+      "learning_rate": 5.055949394847932e-06,
+      "loss": 0.4325,
+      "step": 14184
+    },
+    {
+      "epoch": 2.329583478742841,
+      "grad_norm": 0.34073141928086287,
+      "learning_rate": 5.05548401240004e-06,
+      "loss": 0.4338,
+      "step": 14185
+    },
+    {
+      "epoch": 2.3297477059511835,
+      "grad_norm": 0.3401716194953938,
+      "learning_rate": 5.0550186239190445e-06,
+      "loss": 0.4423,
+      "step": 14186
+    },
+    {
+      "epoch": 2.329911933159526,
+      "grad_norm": 0.3208427891596354,
+      "learning_rate": 5.0545532294100016e-06,
+      "loss": 0.4463,
+      "step": 14187
+    },
+    {
+      "epoch": 2.330076160367869,
+      "grad_norm": 0.36160948952617955,
+      "learning_rate": 5.054087828877966e-06,
+      "loss": 0.4596,
+      "step": 14188
+    },
+    {
+      "epoch": 2.330240387576212,
+      "grad_norm": 0.31086819709984803,
+      "learning_rate": 5.053622422327991e-06,
+      "loss": 0.4191,
+      "step": 14189
+    },
+    {
+      "epoch": 2.3304046147845545,
+      "grad_norm": 0.34567113619909157,
+      "learning_rate": 5.053157009765131e-06,
+      "loss": 0.4426,
+      "step": 14190
+    },
+    {
+      "epoch": 2.330568841992897,
+      "grad_norm": 0.3963209834997856,
+      "learning_rate": 5.052691591194442e-06,
+      "loss": 0.4531,
+      "step": 14191
+    },
+    {
+      "epoch": 2.33073306920124,
+      "grad_norm": 0.3103594508662998,
+      "learning_rate": 5.052226166620978e-06,
+      "loss": 0.4437,
+      "step": 14192
+    },
+    {
+      "epoch": 2.3308972964095824,
+      "grad_norm": 0.6589932994119513,
+      "learning_rate": 5.051760736049797e-06,
+      "loss": 0.4587,
+      "step": 14193
+    },
+    {
+      "epoch": 2.3310615236179255,
+      "grad_norm": 0.3181522531554116,
+      "learning_rate": 5.051295299485949e-06,
+      "loss": 0.4471,
+      "step": 14194
+    },
+    {
+      "epoch": 2.331225750826268,
+      "grad_norm": 0.5168257697310911,
+      "learning_rate": 5.0508298569344915e-06,
+      "loss": 0.4496,
+      "step": 14195
+    },
+    {
+      "epoch": 2.331389978034611,
+      "grad_norm": 0.39529469305292186,
+      "learning_rate": 5.05036440840048e-06,
+      "loss": 0.4473,
+      "step": 14196
+    },
+    {
+      "epoch": 2.331554205242954,
+      "grad_norm": 0.9181809301881988,
+      "learning_rate": 5.049898953888971e-06,
+      "loss": 0.4445,
+      "step": 14197
+    },
+    {
+      "epoch": 2.3317184324512965,
+      "grad_norm": 0.324874626375649,
+      "learning_rate": 5.049433493405018e-06,
+      "loss": 0.4317,
+      "step": 14198
+    },
+    {
+      "epoch": 2.331882659659639,
+      "grad_norm": 0.3660678533136915,
+      "learning_rate": 5.048968026953676e-06,
+      "loss": 0.4453,
+      "step": 14199
+    },
+    {
+      "epoch": 2.332046886867982,
+      "grad_norm": 0.29991457677255184,
+      "learning_rate": 5.048502554540001e-06,
+      "loss": 0.4349,
+      "step": 14200
+    },
+    {
+      "epoch": 2.3322111140763244,
+      "grad_norm": 0.41932979749369553,
+      "learning_rate": 5.048037076169049e-06,
+      "loss": 0.4427,
+      "step": 14201
+    },
+    {
+      "epoch": 2.3323753412846675,
+      "grad_norm": 0.3987618004851579,
+      "learning_rate": 5.047571591845875e-06,
+      "loss": 0.4478,
+      "step": 14202
+    },
+    {
+      "epoch": 2.33253956849301,
+      "grad_norm": 0.3376552578233765,
+      "learning_rate": 5.047106101575535e-06,
+      "loss": 0.4224,
+      "step": 14203
+    },
+    {
+      "epoch": 2.332703795701353,
+      "grad_norm": 0.3926530091665801,
+      "learning_rate": 5.046640605363084e-06,
+      "loss": 0.4436,
+      "step": 14204
+    },
+    {
+      "epoch": 2.3328680229096954,
+      "grad_norm": 0.6240595384014374,
+      "learning_rate": 5.046175103213579e-06,
+      "loss": 0.4306,
+      "step": 14205
+    },
+    {
+      "epoch": 2.3330322501180385,
+      "grad_norm": 0.4986024189108446,
+      "learning_rate": 5.045709595132074e-06,
+      "loss": 0.4649,
+      "step": 14206
+    },
+    {
+      "epoch": 2.333196477326381,
+      "grad_norm": 0.2973310236981245,
+      "learning_rate": 5.045244081123627e-06,
+      "loss": 0.4206,
+      "step": 14207
+    },
+    {
+      "epoch": 2.3333607045347238,
+      "grad_norm": 0.5038718583325013,
+      "learning_rate": 5.044778561193291e-06,
+      "loss": 0.425,
+      "step": 14208
+    },
+    {
+      "epoch": 2.3335249317430664,
+      "grad_norm": 0.32869318551080595,
+      "learning_rate": 5.044313035346126e-06,
+      "loss": 0.4238,
+      "step": 14209
+    },
+    {
+      "epoch": 2.333689158951409,
+      "grad_norm": 0.7844142557053408,
+      "learning_rate": 5.043847503587184e-06,
+      "loss": 0.4586,
+      "step": 14210
+    },
+    {
+      "epoch": 2.333853386159752,
+      "grad_norm": 0.39747748879408473,
+      "learning_rate": 5.043381965921524e-06,
+      "loss": 0.425,
+      "step": 14211
+    },
+    {
+      "epoch": 2.3340176133680948,
+      "grad_norm": 0.3712658915797102,
+      "learning_rate": 5.042916422354202e-06,
+      "loss": 0.4311,
+      "step": 14212
+    },
+    {
+      "epoch": 2.3341818405764374,
+      "grad_norm": 0.35805916644196095,
+      "learning_rate": 5.042450872890272e-06,
+      "loss": 0.4497,
+      "step": 14213
+    },
+    {
+      "epoch": 2.3343460677847805,
+      "grad_norm": 0.3489653733566996,
+      "learning_rate": 5.041985317534793e-06,
+      "loss": 0.4386,
+      "step": 14214
+    },
+    {
+      "epoch": 2.334510294993123,
+      "grad_norm": 1.552344521578485,
+      "learning_rate": 5.04151975629282e-06,
+      "loss": 0.4399,
+      "step": 14215
+    },
+    {
+      "epoch": 2.3346745222014658,
+      "grad_norm": 0.4330518043111234,
+      "learning_rate": 5.041054189169409e-06,
+      "loss": 0.4358,
+      "step": 14216
+    },
+    {
+      "epoch": 2.3348387494098084,
+      "grad_norm": 0.32002591932723284,
+      "learning_rate": 5.040588616169618e-06,
+      "loss": 0.419,
+      "step": 14217
+    },
+    {
+      "epoch": 2.335002976618151,
+      "grad_norm": 0.31767776648707025,
+      "learning_rate": 5.040123037298503e-06,
+      "loss": 0.4407,
+      "step": 14218
+    },
+    {
+      "epoch": 2.335167203826494,
+      "grad_norm": 0.366803435400509,
+      "learning_rate": 5.039657452561119e-06,
+      "loss": 0.4371,
+      "step": 14219
+    },
+    {
+      "epoch": 2.3353314310348368,
+      "grad_norm": 0.3296447167812067,
+      "learning_rate": 5.039191861962524e-06,
+      "loss": 0.4202,
+      "step": 14220
+    },
+    {
+      "epoch": 2.3354956582431794,
+      "grad_norm": 0.3349276761092498,
+      "learning_rate": 5.0387262655077755e-06,
+      "loss": 0.4325,
+      "step": 14221
+    },
+    {
+      "epoch": 2.335659885451522,
+      "grad_norm": 0.3980378072812971,
+      "learning_rate": 5.0382606632019325e-06,
+      "loss": 0.4229,
+      "step": 14222
+    },
+    {
+      "epoch": 2.335824112659865,
+      "grad_norm": 0.3202898785116211,
+      "learning_rate": 5.037795055050046e-06,
+      "loss": 0.4743,
+      "step": 14223
+    },
+    {
+      "epoch": 2.3359883398682078,
+      "grad_norm": 0.31778100820349403,
+      "learning_rate": 5.037329441057176e-06,
+      "loss": 0.4217,
+      "step": 14224
+    },
+    {
+      "epoch": 2.3361525670765504,
+      "grad_norm": 0.32080705990736375,
+      "learning_rate": 5.0368638212283795e-06,
+      "loss": 0.4228,
+      "step": 14225
+    },
+    {
+      "epoch": 2.336316794284893,
+      "grad_norm": 0.3102344359741497,
+      "learning_rate": 5.036398195568716e-06,
+      "loss": 0.4358,
+      "step": 14226
+    },
+    {
+      "epoch": 2.3364810214932357,
+      "grad_norm": 0.29092912120353454,
+      "learning_rate": 5.035932564083238e-06,
+      "loss": 0.4297,
+      "step": 14227
+    },
+    {
+      "epoch": 2.3366452487015787,
+      "grad_norm": 0.27798039965725674,
+      "learning_rate": 5.035466926777007e-06,
+      "loss": 0.4143,
+      "step": 14228
+    },
+    {
+      "epoch": 2.3368094759099214,
+      "grad_norm": 0.44537914816795954,
+      "learning_rate": 5.035001283655076e-06,
+      "loss": 0.4425,
+      "step": 14229
+    },
+    {
+      "epoch": 2.336973703118264,
+      "grad_norm": 0.3381259638195042,
+      "learning_rate": 5.0345356347225065e-06,
+      "loss": 0.4383,
+      "step": 14230
+    },
+    {
+      "epoch": 2.337137930326607,
+      "grad_norm": 0.36687226092077213,
+      "learning_rate": 5.034069979984353e-06,
+      "loss": 0.4437,
+      "step": 14231
+    },
+    {
+      "epoch": 2.3373021575349497,
+      "grad_norm": 0.46311466060791356,
+      "learning_rate": 5.033604319445676e-06,
+      "loss": 0.4397,
+      "step": 14232
+    },
+    {
+      "epoch": 2.3374663847432924,
+      "grad_norm": 0.3617763925092038,
+      "learning_rate": 5.0331386531115285e-06,
+      "loss": 0.4663,
+      "step": 14233
+    },
+    {
+      "epoch": 2.337630611951635,
+      "grad_norm": 0.44891382965557786,
+      "learning_rate": 5.032672980986972e-06,
+      "loss": 0.4556,
+      "step": 14234
+    },
+    {
+      "epoch": 2.3377948391599777,
+      "grad_norm": 0.36788493675452716,
+      "learning_rate": 5.032207303077063e-06,
+      "loss": 0.4325,
+      "step": 14235
+    },
+    {
+      "epoch": 2.3379590663683207,
+      "grad_norm": 0.3044382854335584,
+      "learning_rate": 5.031741619386858e-06,
+      "loss": 0.4566,
+      "step": 14236
+    },
+    {
+      "epoch": 2.3381232935766634,
+      "grad_norm": 0.3440239656824285,
+      "learning_rate": 5.0312759299214175e-06,
+      "loss": 0.4456,
+      "step": 14237
+    },
+    {
+      "epoch": 2.338287520785006,
+      "grad_norm": 0.3696861896699254,
+      "learning_rate": 5.030810234685796e-06,
+      "loss": 0.4308,
+      "step": 14238
+    },
+    {
+      "epoch": 2.3384517479933487,
+      "grad_norm": 0.4248817510969493,
+      "learning_rate": 5.030344533685054e-06,
+      "loss": 0.4313,
+      "step": 14239
+    },
+    {
+      "epoch": 2.3386159752016917,
+      "grad_norm": 0.3077063680986963,
+      "learning_rate": 5.02987882692425e-06,
+      "loss": 0.4457,
+      "step": 14240
+    },
+    {
+      "epoch": 2.3387802024100344,
+      "grad_norm": 0.337582640428433,
+      "learning_rate": 5.02941311440844e-06,
+      "loss": 0.4135,
+      "step": 14241
+    },
+    {
+      "epoch": 2.338944429618377,
+      "grad_norm": 0.3671872733120523,
+      "learning_rate": 5.028947396142681e-06,
+      "loss": 0.4155,
+      "step": 14242
+    },
+    {
+      "epoch": 2.3391086568267196,
+      "grad_norm": 0.3805787562666042,
+      "learning_rate": 5.028481672132034e-06,
+      "loss": 0.4486,
+      "step": 14243
+    },
+    {
+      "epoch": 2.3392728840350623,
+      "grad_norm": 0.5836936584296454,
+      "learning_rate": 5.028015942381555e-06,
+      "loss": 0.4465,
+      "step": 14244
+    },
+    {
+      "epoch": 2.3394371112434054,
+      "grad_norm": 0.36682613928154184,
+      "learning_rate": 5.027550206896304e-06,
+      "loss": 0.4313,
+      "step": 14245
+    },
+    {
+      "epoch": 2.339601338451748,
+      "grad_norm": 0.3211669947248319,
+      "learning_rate": 5.027084465681339e-06,
+      "loss": 0.4331,
+      "step": 14246
+    },
+    {
+      "epoch": 2.3397655656600906,
+      "grad_norm": 0.33075234650746704,
+      "learning_rate": 5.026618718741719e-06,
+      "loss": 0.4381,
+      "step": 14247
+    },
+    {
+      "epoch": 2.3399297928684337,
+      "grad_norm": 0.3224871130791294,
+      "learning_rate": 5.0261529660824994e-06,
+      "loss": 0.441,
+      "step": 14248
+    },
+    {
+      "epoch": 2.3400940200767764,
+      "grad_norm": 0.5106742651433831,
+      "learning_rate": 5.025687207708743e-06,
+      "loss": 0.4349,
+      "step": 14249
+    },
+    {
+      "epoch": 2.340258247285119,
+      "grad_norm": 0.3670987534944953,
+      "learning_rate": 5.025221443625504e-06,
+      "loss": 0.43,
+      "step": 14250
+    },
+    {
+      "epoch": 2.3404224744934616,
+      "grad_norm": 0.3152223583268576,
+      "learning_rate": 5.024755673837845e-06,
+      "loss": 0.4203,
+      "step": 14251
+    },
+    {
+      "epoch": 2.3405867017018043,
+      "grad_norm": 0.3104109942911494,
+      "learning_rate": 5.024289898350825e-06,
+      "loss": 0.4453,
+      "step": 14252
+    },
+    {
+      "epoch": 2.3407509289101474,
+      "grad_norm": 0.3354252677596092,
+      "learning_rate": 5.0238241171694974e-06,
+      "loss": 0.426,
+      "step": 14253
+    },
+    {
+      "epoch": 2.34091515611849,
+      "grad_norm": 0.3096400471568136,
+      "learning_rate": 5.023358330298925e-06,
+      "loss": 0.4414,
+      "step": 14254
+    },
+    {
+      "epoch": 2.3410793833268326,
+      "grad_norm": 0.4627023481999118,
+      "learning_rate": 5.022892537744167e-06,
+      "loss": 0.4296,
+      "step": 14255
+    },
+    {
+      "epoch": 2.3412436105351753,
+      "grad_norm": 0.47554609569492284,
+      "learning_rate": 5.022426739510283e-06,
+      "loss": 0.4319,
+      "step": 14256
+    },
+    {
+      "epoch": 2.3414078377435183,
+      "grad_norm": 0.33244191992096805,
+      "learning_rate": 5.021960935602329e-06,
+      "loss": 0.4346,
+      "step": 14257
+    },
+    {
+      "epoch": 2.341572064951861,
+      "grad_norm": 0.3671414791598005,
+      "learning_rate": 5.021495126025366e-06,
+      "loss": 0.4197,
+      "step": 14258
+    },
+    {
+      "epoch": 2.3417362921602036,
+      "grad_norm": 0.3001605538598186,
+      "learning_rate": 5.021029310784453e-06,
+      "loss": 0.435,
+      "step": 14259
+    },
+    {
+      "epoch": 2.3419005193685463,
+      "grad_norm": 0.41049676168709537,
+      "learning_rate": 5.020563489884649e-06,
+      "loss": 0.4452,
+      "step": 14260
+    },
+    {
+      "epoch": 2.342064746576889,
+      "grad_norm": 0.39958421223807394,
+      "learning_rate": 5.020097663331014e-06,
+      "loss": 0.4378,
+      "step": 14261
+    },
+    {
+      "epoch": 2.342228973785232,
+      "grad_norm": 0.3526260350919074,
+      "learning_rate": 5.019631831128605e-06,
+      "loss": 0.4508,
+      "step": 14262
+    },
+    {
+      "epoch": 2.3423932009935746,
+      "grad_norm": 0.36019790514805233,
+      "learning_rate": 5.019165993282483e-06,
+      "loss": 0.4508,
+      "step": 14263
+    },
+    {
+      "epoch": 2.3425574282019173,
+      "grad_norm": 0.39219392948370696,
+      "learning_rate": 5.018700149797709e-06,
+      "loss": 0.4285,
+      "step": 14264
+    },
+    {
+      "epoch": 2.3427216554102603,
+      "grad_norm": 0.34781631591555273,
+      "learning_rate": 5.018234300679341e-06,
+      "loss": 0.4335,
+      "step": 14265
+    },
+    {
+      "epoch": 2.342885882618603,
+      "grad_norm": 0.33544889426458624,
+      "learning_rate": 5.017768445932438e-06,
+      "loss": 0.4483,
+      "step": 14266
+    },
+    {
+      "epoch": 2.3430501098269456,
+      "grad_norm": 0.3154253538619397,
+      "learning_rate": 5.01730258556206e-06,
+      "loss": 0.4427,
+      "step": 14267
+    },
+    {
+      "epoch": 2.3432143370352883,
+      "grad_norm": 0.33192751785748736,
+      "learning_rate": 5.016836719573268e-06,
+      "loss": 0.4133,
+      "step": 14268
+    },
+    {
+      "epoch": 2.343378564243631,
+      "grad_norm": 0.313437034515309,
+      "learning_rate": 5.01637084797112e-06,
+      "loss": 0.4341,
+      "step": 14269
+    },
+    {
+      "epoch": 2.343542791451974,
+      "grad_norm": 0.31116640811452134,
+      "learning_rate": 5.015904970760677e-06,
+      "loss": 0.4455,
+      "step": 14270
+    },
+    {
+      "epoch": 2.3437070186603166,
+      "grad_norm": 0.44806362836241137,
+      "learning_rate": 5.015439087946998e-06,
+      "loss": 0.4522,
+      "step": 14271
+    },
+    {
+      "epoch": 2.3438712458686592,
+      "grad_norm": 0.34147456503692797,
+      "learning_rate": 5.0149731995351445e-06,
+      "loss": 0.4347,
+      "step": 14272
+    },
+    {
+      "epoch": 2.344035473077002,
+      "grad_norm": 0.31152642014698967,
+      "learning_rate": 5.014507305530173e-06,
+      "loss": 0.4205,
+      "step": 14273
+    },
+    {
+      "epoch": 2.344199700285345,
+      "grad_norm": 0.34475814543629774,
+      "learning_rate": 5.014041405937147e-06,
+      "loss": 0.4342,
+      "step": 14274
+    },
+    {
+      "epoch": 2.3443639274936876,
+      "grad_norm": 0.4015184961878321,
+      "learning_rate": 5.013575500761127e-06,
+      "loss": 0.4257,
+      "step": 14275
+    },
+    {
+      "epoch": 2.3445281547020302,
+      "grad_norm": 0.3033020240845556,
+      "learning_rate": 5.013109590007172e-06,
+      "loss": 0.4382,
+      "step": 14276
+    },
+    {
+      "epoch": 2.344692381910373,
+      "grad_norm": 0.338000873883839,
+      "learning_rate": 5.012643673680339e-06,
+      "loss": 0.4404,
+      "step": 14277
+    },
+    {
+      "epoch": 2.3448566091187155,
+      "grad_norm": 0.31550507098643704,
+      "learning_rate": 5.012177751785694e-06,
+      "loss": 0.4527,
+      "step": 14278
+    },
+    {
+      "epoch": 2.3450208363270586,
+      "grad_norm": 0.2923251975366749,
+      "learning_rate": 5.0117118243282925e-06,
+      "loss": 0.4509,
+      "step": 14279
+    },
+    {
+      "epoch": 2.3451850635354012,
+      "grad_norm": 0.4025916966292508,
+      "learning_rate": 5.011245891313199e-06,
+      "loss": 0.4418,
+      "step": 14280
+    },
+    {
+      "epoch": 2.345349290743744,
+      "grad_norm": 0.3495439251380644,
+      "learning_rate": 5.010779952745472e-06,
+      "loss": 0.4466,
+      "step": 14281
+    },
+    {
+      "epoch": 2.345513517952087,
+      "grad_norm": 0.3030842299624673,
+      "learning_rate": 5.010314008630171e-06,
+      "loss": 0.451,
+      "step": 14282
+    },
+    {
+      "epoch": 2.3456777451604296,
+      "grad_norm": 0.25459738553951666,
+      "learning_rate": 5.009848058972359e-06,
+      "loss": 0.4288,
+      "step": 14283
+    },
+    {
+      "epoch": 2.3458419723687722,
+      "grad_norm": 0.34130323628216985,
+      "learning_rate": 5.009382103777093e-06,
+      "loss": 0.4415,
+      "step": 14284
+    },
+    {
+      "epoch": 2.346006199577115,
+      "grad_norm": 0.31855931505260354,
+      "learning_rate": 5.008916143049439e-06,
+      "loss": 0.4409,
+      "step": 14285
+    },
+    {
+      "epoch": 2.3461704267854575,
+      "grad_norm": 0.529587397164376,
+      "learning_rate": 5.008450176794455e-06,
+      "loss": 0.4339,
+      "step": 14286
+    },
+    {
+      "epoch": 2.3463346539938006,
+      "grad_norm": 0.30583306940272903,
+      "learning_rate": 5.0079842050172e-06,
+      "loss": 0.4358,
+      "step": 14287
+    },
+    {
+      "epoch": 2.3464988812021432,
+      "grad_norm": 0.3376306795943431,
+      "learning_rate": 5.007518227722738e-06,
+      "loss": 0.4744,
+      "step": 14288
+    },
+    {
+      "epoch": 2.346663108410486,
+      "grad_norm": 0.3799763220578188,
+      "learning_rate": 5.007052244916129e-06,
+      "loss": 0.4431,
+      "step": 14289
+    },
+    {
+      "epoch": 2.3468273356188285,
+      "grad_norm": 0.2766437623030712,
+      "learning_rate": 5.006586256602433e-06,
+      "loss": 0.4062,
+      "step": 14290
+    },
+    {
+      "epoch": 2.3469915628271716,
+      "grad_norm": 0.3172427330468487,
+      "learning_rate": 5.006120262786712e-06,
+      "loss": 0.4506,
+      "step": 14291
+    },
+    {
+      "epoch": 2.3471557900355142,
+      "grad_norm": 0.370368306164017,
+      "learning_rate": 5.005654263474026e-06,
+      "loss": 0.4444,
+      "step": 14292
+    },
+    {
+      "epoch": 2.347320017243857,
+      "grad_norm": 0.5424344754056988,
+      "learning_rate": 5.0051882586694384e-06,
+      "loss": 0.4411,
+      "step": 14293
+    },
+    {
+      "epoch": 2.3474842444521995,
+      "grad_norm": 0.37539884285858194,
+      "learning_rate": 5.004722248378009e-06,
+      "loss": 0.4245,
+      "step": 14294
+    },
+    {
+      "epoch": 2.347648471660542,
+      "grad_norm": 0.317396319175305,
+      "learning_rate": 5.0042562326048e-06,
+      "loss": 0.4444,
+      "step": 14295
+    },
+    {
+      "epoch": 2.347812698868885,
+      "grad_norm": 0.28224187924061256,
+      "learning_rate": 5.003790211354872e-06,
+      "loss": 0.4551,
+      "step": 14296
+    },
+    {
+      "epoch": 2.347976926077228,
+      "grad_norm": 0.3263896166098868,
+      "learning_rate": 5.003324184633286e-06,
+      "loss": 0.4499,
+      "step": 14297
+    },
+    {
+      "epoch": 2.3481411532855705,
+      "grad_norm": 0.3283510564766459,
+      "learning_rate": 5.002858152445104e-06,
+      "loss": 0.439,
+      "step": 14298
+    },
+    {
+      "epoch": 2.3483053804939136,
+      "grad_norm": 0.34528668929370976,
+      "learning_rate": 5.002392114795388e-06,
+      "loss": 0.4474,
+      "step": 14299
+    },
+    {
+      "epoch": 2.348469607702256,
+      "grad_norm": 0.3675402895714062,
+      "learning_rate": 5.0019260716892e-06,
+      "loss": 0.466,
+      "step": 14300
+    },
+    {
+      "epoch": 2.348633834910599,
+      "grad_norm": 0.32888646075463274,
+      "learning_rate": 5.001460023131601e-06,
+      "loss": 0.4308,
+      "step": 14301
+    },
+    {
+      "epoch": 2.3487980621189415,
+      "grad_norm": 0.34194545071114996,
+      "learning_rate": 5.000993969127652e-06,
+      "loss": 0.4381,
+      "step": 14302
+    },
+    {
+      "epoch": 2.348962289327284,
+      "grad_norm": 0.3481120091692677,
+      "learning_rate": 5.000527909682415e-06,
+      "loss": 0.4281,
+      "step": 14303
+    },
+    {
+      "epoch": 2.349126516535627,
+      "grad_norm": 0.30426600918544544,
+      "learning_rate": 5.000061844800953e-06,
+      "loss": 0.4242,
+      "step": 14304
+    },
+    {
+      "epoch": 2.34929074374397,
+      "grad_norm": 0.33042610605754225,
+      "learning_rate": 4.9995957744883286e-06,
+      "loss": 0.4362,
+      "step": 14305
+    },
+    {
+      "epoch": 2.3494549709523125,
+      "grad_norm": 0.6139370838276929,
+      "learning_rate": 4.999129698749602e-06,
+      "loss": 0.4475,
+      "step": 14306
+    },
+    {
+      "epoch": 2.349619198160655,
+      "grad_norm": 0.3091523983404541,
+      "learning_rate": 4.998663617589835e-06,
+      "loss": 0.4586,
+      "step": 14307
+    },
+    {
+      "epoch": 2.349783425368998,
+      "grad_norm": 0.3671883433413459,
+      "learning_rate": 4.998197531014091e-06,
+      "loss": 0.4205,
+      "step": 14308
+    },
+    {
+      "epoch": 2.349947652577341,
+      "grad_norm": 0.28508625616699385,
+      "learning_rate": 4.997731439027432e-06,
+      "loss": 0.4259,
+      "step": 14309
+    },
+    {
+      "epoch": 2.3501118797856835,
+      "grad_norm": 0.33190313156947665,
+      "learning_rate": 4.99726534163492e-06,
+      "loss": 0.4272,
+      "step": 14310
+    },
+    {
+      "epoch": 2.350276106994026,
+      "grad_norm": 0.3684075426938841,
+      "learning_rate": 4.996799238841616e-06,
+      "loss": 0.4357,
+      "step": 14311
+    },
+    {
+      "epoch": 2.3504403342023688,
+      "grad_norm": 0.29104654696695337,
+      "learning_rate": 4.9963331306525834e-06,
+      "loss": 0.4507,
+      "step": 14312
+    },
+    {
+      "epoch": 2.350604561410712,
+      "grad_norm": 0.45342829692727055,
+      "learning_rate": 4.995867017072885e-06,
+      "loss": 0.4348,
+      "step": 14313
+    },
+    {
+      "epoch": 2.3507687886190545,
+      "grad_norm": 0.28462488186154083,
+      "learning_rate": 4.995400898107584e-06,
+      "loss": 0.4408,
+      "step": 14314
+    },
+    {
+      "epoch": 2.350933015827397,
+      "grad_norm": 0.32996492095223623,
+      "learning_rate": 4.994934773761742e-06,
+      "loss": 0.455,
+      "step": 14315
+    },
+    {
+      "epoch": 2.35109724303574,
+      "grad_norm": 0.4413132936257363,
+      "learning_rate": 4.994468644040419e-06,
+      "loss": 0.4453,
+      "step": 14316
+    },
+    {
+      "epoch": 2.351261470244083,
+      "grad_norm": 0.287439219083887,
+      "learning_rate": 4.994002508948682e-06,
+      "loss": 0.45,
+      "step": 14317
+    },
+    {
+      "epoch": 2.3514256974524255,
+      "grad_norm": 0.6687016558497426,
+      "learning_rate": 4.993536368491592e-06,
+      "loss": 0.4322,
+      "step": 14318
+    },
+    {
+      "epoch": 2.351589924660768,
+      "grad_norm": 0.31833933331295167,
+      "learning_rate": 4.99307022267421e-06,
+      "loss": 0.4233,
+      "step": 14319
+    },
+    {
+      "epoch": 2.3517541518691107,
+      "grad_norm": 0.2936523952712435,
+      "learning_rate": 4.992604071501601e-06,
+      "loss": 0.424,
+      "step": 14320
+    },
+    {
+      "epoch": 2.351918379077454,
+      "grad_norm": 0.29199165743680827,
+      "learning_rate": 4.992137914978827e-06,
+      "loss": 0.4367,
+      "step": 14321
+    },
+    {
+      "epoch": 2.3520826062857965,
+      "grad_norm": 0.38049347369291886,
+      "learning_rate": 4.991671753110952e-06,
+      "loss": 0.4562,
+      "step": 14322
+    },
+    {
+      "epoch": 2.352246833494139,
+      "grad_norm": 0.28514980201093715,
+      "learning_rate": 4.991205585903037e-06,
+      "loss": 0.4403,
+      "step": 14323
+    },
+    {
+      "epoch": 2.3524110607024817,
+      "grad_norm": 0.29063242226173736,
+      "learning_rate": 4.990739413360147e-06,
+      "loss": 0.4173,
+      "step": 14324
+    },
+    {
+      "epoch": 2.352575287910825,
+      "grad_norm": 0.33956667259069695,
+      "learning_rate": 4.990273235487343e-06,
+      "loss": 0.4395,
+      "step": 14325
+    },
+    {
+      "epoch": 2.3527395151191675,
+      "grad_norm": 0.29213608629516113,
+      "learning_rate": 4.9898070522896885e-06,
+      "loss": 0.436,
+      "step": 14326
+    },
+    {
+      "epoch": 2.35290374232751,
+      "grad_norm": 0.3184200244453,
+      "learning_rate": 4.9893408637722504e-06,
+      "loss": 0.4421,
+      "step": 14327
+    },
+    {
+      "epoch": 2.3530679695358527,
+      "grad_norm": 0.9054970444536435,
+      "learning_rate": 4.988874669940086e-06,
+      "loss": 0.4359,
+      "step": 14328
+    },
+    {
+      "epoch": 2.3532321967441954,
+      "grad_norm": 0.8009512337096192,
+      "learning_rate": 4.988408470798264e-06,
+      "loss": 0.4652,
+      "step": 14329
+    },
+    {
+      "epoch": 2.3533964239525385,
+      "grad_norm": 0.3000048361872114,
+      "learning_rate": 4.987942266351845e-06,
+      "loss": 0.4389,
+      "step": 14330
+    },
+    {
+      "epoch": 2.353560651160881,
+      "grad_norm": 0.4435834381481093,
+      "learning_rate": 4.987476056605892e-06,
+      "loss": 0.4462,
+      "step": 14331
+    },
+    {
+      "epoch": 2.3537248783692237,
+      "grad_norm": 0.31409223317024465,
+      "learning_rate": 4.987009841565469e-06,
+      "loss": 0.4288,
+      "step": 14332
+    },
+    {
+      "epoch": 2.353889105577567,
+      "grad_norm": 0.3775543340666963,
+      "learning_rate": 4.986543621235641e-06,
+      "loss": 0.4687,
+      "step": 14333
+    },
+    {
+      "epoch": 2.3540533327859094,
+      "grad_norm": 0.3154684698397416,
+      "learning_rate": 4.986077395621471e-06,
+      "loss": 0.4511,
+      "step": 14334
+    },
+    {
+      "epoch": 2.354217559994252,
+      "grad_norm": 0.33377923891010136,
+      "learning_rate": 4.985611164728022e-06,
+      "loss": 0.4495,
+      "step": 14335
+    },
+    {
+      "epoch": 2.3543817872025947,
+      "grad_norm": 0.38798308412223753,
+      "learning_rate": 4.9851449285603575e-06,
+      "loss": 0.4527,
+      "step": 14336
+    },
+    {
+      "epoch": 2.3545460144109374,
+      "grad_norm": 0.3595666147784764,
+      "learning_rate": 4.984678687123542e-06,
+      "loss": 0.4423,
+      "step": 14337
+    },
+    {
+      "epoch": 2.3547102416192804,
+      "grad_norm": 0.35797627990956515,
+      "learning_rate": 4.984212440422639e-06,
+      "loss": 0.4325,
+      "step": 14338
+    },
+    {
+      "epoch": 2.354874468827623,
+      "grad_norm": 0.29402622902463466,
+      "learning_rate": 4.9837461884627134e-06,
+      "loss": 0.4498,
+      "step": 14339
+    },
+    {
+      "epoch": 2.3550386960359657,
+      "grad_norm": 0.3619177067733234,
+      "learning_rate": 4.983279931248827e-06,
+      "loss": 0.4209,
+      "step": 14340
+    },
+    {
+      "epoch": 2.3552029232443084,
+      "grad_norm": 0.3469217107065268,
+      "learning_rate": 4.982813668786045e-06,
+      "loss": 0.4301,
+      "step": 14341
+    },
+    {
+      "epoch": 2.3553671504526514,
+      "grad_norm": 0.2869461199173222,
+      "learning_rate": 4.982347401079432e-06,
+      "loss": 0.4353,
+      "step": 14342
+    },
+    {
+      "epoch": 2.355531377660994,
+      "grad_norm": 0.3310294155726151,
+      "learning_rate": 4.981881128134052e-06,
+      "loss": 0.4306,
+      "step": 14343
+    },
+    {
+      "epoch": 2.3556956048693367,
+      "grad_norm": 0.44544791162175035,
+      "learning_rate": 4.981414849954969e-06,
+      "loss": 0.4457,
+      "step": 14344
+    },
+    {
+      "epoch": 2.3558598320776793,
+      "grad_norm": 0.2821848497782474,
+      "learning_rate": 4.980948566547246e-06,
+      "loss": 0.4283,
+      "step": 14345
+    },
+    {
+      "epoch": 2.356024059286022,
+      "grad_norm": 0.2899346904371948,
+      "learning_rate": 4.980482277915948e-06,
+      "loss": 0.4384,
+      "step": 14346
+    },
+    {
+      "epoch": 2.356188286494365,
+      "grad_norm": 0.28895582908267986,
+      "learning_rate": 4.9800159840661416e-06,
+      "loss": 0.4425,
+      "step": 14347
+    },
+    {
+      "epoch": 2.3563525137027077,
+      "grad_norm": 0.32261166832480315,
+      "learning_rate": 4.979549685002888e-06,
+      "loss": 0.4473,
+      "step": 14348
+    },
+    {
+      "epoch": 2.3565167409110503,
+      "grad_norm": 0.41278684584248443,
+      "learning_rate": 4.979083380731254e-06,
+      "loss": 0.4489,
+      "step": 14349
+    },
+    {
+      "epoch": 2.3566809681193934,
+      "grad_norm": 0.36717664735562605,
+      "learning_rate": 4.978617071256302e-06,
+      "loss": 0.4333,
+      "step": 14350
+    },
+    {
+      "epoch": 2.356845195327736,
+      "grad_norm": 0.2871702500281055,
+      "learning_rate": 4.978150756583098e-06,
+      "loss": 0.4314,
+      "step": 14351
+    },
+    {
+      "epoch": 2.3570094225360787,
+      "grad_norm": 0.296831622491759,
+      "learning_rate": 4.977684436716707e-06,
+      "loss": 0.4459,
+      "step": 14352
+    },
+    {
+      "epoch": 2.3571736497444213,
+      "grad_norm": 0.3152784713613363,
+      "learning_rate": 4.977218111662193e-06,
+      "loss": 0.4313,
+      "step": 14353
+    },
+    {
+      "epoch": 2.357337876952764,
+      "grad_norm": 0.3179123907686836,
+      "learning_rate": 4.97675178142462e-06,
+      "loss": 0.4383,
+      "step": 14354
+    },
+    {
+      "epoch": 2.357502104161107,
+      "grad_norm": 0.33097652269284605,
+      "learning_rate": 4.976285446009053e-06,
+      "loss": 0.4608,
+      "step": 14355
+    },
+    {
+      "epoch": 2.3576663313694497,
+      "grad_norm": 0.26370984506069994,
+      "learning_rate": 4.9758191054205595e-06,
+      "loss": 0.4194,
+      "step": 14356
+    },
+    {
+      "epoch": 2.3578305585777923,
+      "grad_norm": 0.3503675341705753,
+      "learning_rate": 4.975352759664201e-06,
+      "loss": 0.4318,
+      "step": 14357
+    },
+    {
+      "epoch": 2.357994785786135,
+      "grad_norm": 0.26075769429226026,
+      "learning_rate": 4.974886408745045e-06,
+      "loss": 0.4337,
+      "step": 14358
+    },
+    {
+      "epoch": 2.358159012994478,
+      "grad_norm": 0.3167737749473058,
+      "learning_rate": 4.9744200526681545e-06,
+      "loss": 0.4472,
+      "step": 14359
+    },
+    {
+      "epoch": 2.3583232402028207,
+      "grad_norm": 0.3267289682222068,
+      "learning_rate": 4.973953691438595e-06,
+      "loss": 0.4489,
+      "step": 14360
+    },
+    {
+      "epoch": 2.3584874674111633,
+      "grad_norm": 0.48223262088179,
+      "learning_rate": 4.973487325061433e-06,
+      "loss": 0.4413,
+      "step": 14361
+    },
+    {
+      "epoch": 2.358651694619506,
+      "grad_norm": 0.29004525308874013,
+      "learning_rate": 4.973020953541732e-06,
+      "loss": 0.4247,
+      "step": 14362
+    },
+    {
+      "epoch": 2.3588159218278486,
+      "grad_norm": 0.30160171868034047,
+      "learning_rate": 4.972554576884559e-06,
+      "loss": 0.448,
+      "step": 14363
+    },
+    {
+      "epoch": 2.3589801490361917,
+      "grad_norm": 0.41944901184902755,
+      "learning_rate": 4.972088195094978e-06,
+      "loss": 0.4512,
+      "step": 14364
+    },
+    {
+      "epoch": 2.3591443762445343,
+      "grad_norm": 0.28947942760816986,
+      "learning_rate": 4.971621808178054e-06,
+      "loss": 0.4422,
+      "step": 14365
+    },
+    {
+      "epoch": 2.359308603452877,
+      "grad_norm": 0.3238543328135968,
+      "learning_rate": 4.971155416138853e-06,
+      "loss": 0.4397,
+      "step": 14366
+    },
+    {
+      "epoch": 2.35947283066122,
+      "grad_norm": 0.3703052518219256,
+      "learning_rate": 4.970689018982442e-06,
+      "loss": 0.4297,
+      "step": 14367
+    },
+    {
+      "epoch": 2.3596370578695627,
+      "grad_norm": 0.32595968322039665,
+      "learning_rate": 4.9702226167138855e-06,
+      "loss": 0.4132,
+      "step": 14368
+    },
+    {
+      "epoch": 2.3598012850779053,
+      "grad_norm": 0.36003670628504997,
+      "learning_rate": 4.9697562093382475e-06,
+      "loss": 0.4519,
+      "step": 14369
+    },
+    {
+      "epoch": 2.359965512286248,
+      "grad_norm": 0.35066351988470934,
+      "learning_rate": 4.969289796860595e-06,
+      "loss": 0.4557,
+      "step": 14370
+    },
+    {
+      "epoch": 2.3601297394945906,
+      "grad_norm": 0.28461795043565574,
+      "learning_rate": 4.968823379285993e-06,
+      "loss": 0.4465,
+      "step": 14371
+    },
+    {
+      "epoch": 2.3602939667029337,
+      "grad_norm": 0.3349794950571723,
+      "learning_rate": 4.96835695661951e-06,
+      "loss": 0.4298,
+      "step": 14372
+    },
+    {
+      "epoch": 2.3604581939112763,
+      "grad_norm": 0.4623462596162397,
+      "learning_rate": 4.967890528866209e-06,
+      "loss": 0.4376,
+      "step": 14373
+    },
+    {
+      "epoch": 2.360622421119619,
+      "grad_norm": 0.3045711511063072,
+      "learning_rate": 4.967424096031155e-06,
+      "loss": 0.4281,
+      "step": 14374
+    },
+    {
+      "epoch": 2.3607866483279616,
+      "grad_norm": 0.28236555126313645,
+      "learning_rate": 4.966957658119415e-06,
+      "loss": 0.4293,
+      "step": 14375
+    },
+    {
+      "epoch": 2.3609508755363047,
+      "grad_norm": 0.329845338092337,
+      "learning_rate": 4.966491215136056e-06,
+      "loss": 0.4448,
+      "step": 14376
+    },
+    {
+      "epoch": 2.3611151027446473,
+      "grad_norm": 0.2780889848515211,
+      "learning_rate": 4.966024767086145e-06,
+      "loss": 0.4521,
+      "step": 14377
+    },
+    {
+      "epoch": 2.36127932995299,
+      "grad_norm": 0.7223244736692684,
+      "learning_rate": 4.965558313974746e-06,
+      "loss": 0.4267,
+      "step": 14378
+    },
+    {
+      "epoch": 2.3614435571613326,
+      "grad_norm": 0.29452548364476305,
+      "learning_rate": 4.965091855806925e-06,
+      "loss": 0.4378,
+      "step": 14379
+    },
+    {
+      "epoch": 2.3616077843696752,
+      "grad_norm": 0.34535556397656414,
+      "learning_rate": 4.964625392587749e-06,
+      "loss": 0.4374,
+      "step": 14380
+    },
+    {
+      "epoch": 2.3617720115780183,
+      "grad_norm": 0.28214696649413246,
+      "learning_rate": 4.9641589243222845e-06,
+      "loss": 0.4398,
+      "step": 14381
+    },
+    {
+      "epoch": 2.361936238786361,
+      "grad_norm": 0.9998617351395512,
+      "learning_rate": 4.963692451015597e-06,
+      "loss": 0.4421,
+      "step": 14382
+    },
+    {
+      "epoch": 2.3621004659947036,
+      "grad_norm": 0.31297145796805065,
+      "learning_rate": 4.963225972672753e-06,
+      "loss": 0.4216,
+      "step": 14383
+    },
+    {
+      "epoch": 2.3622646932030467,
+      "grad_norm": 0.2939100139357004,
+      "learning_rate": 4.96275948929882e-06,
+      "loss": 0.4648,
+      "step": 14384
+    },
+    {
+      "epoch": 2.3624289204113893,
+      "grad_norm": 0.530158610934547,
+      "learning_rate": 4.962293000898864e-06,
+      "loss": 0.4314,
+      "step": 14385
+    },
+    {
+      "epoch": 2.362593147619732,
+      "grad_norm": 0.40901577717616516,
+      "learning_rate": 4.9618265074779496e-06,
+      "loss": 0.428,
+      "step": 14386
+    },
+    {
+      "epoch": 2.3627573748280746,
+      "grad_norm": 0.27364877261831405,
+      "learning_rate": 4.961360009041146e-06,
+      "loss": 0.44,
+      "step": 14387
+    },
+    {
+      "epoch": 2.362921602036417,
+      "grad_norm": 0.30209721956900093,
+      "learning_rate": 4.9608935055935175e-06,
+      "loss": 0.4211,
+      "step": 14388
+    },
+    {
+      "epoch": 2.3630858292447603,
+      "grad_norm": 0.39013537477111404,
+      "learning_rate": 4.960426997140134e-06,
+      "loss": 0.4411,
+      "step": 14389
+    },
+    {
+      "epoch": 2.363250056453103,
+      "grad_norm": 0.5139156781012,
+      "learning_rate": 4.959960483686059e-06,
+      "loss": 0.4373,
+      "step": 14390
+    },
+    {
+      "epoch": 2.3634142836614456,
+      "grad_norm": 0.3278145517121621,
+      "learning_rate": 4.959493965236361e-06,
+      "loss": 0.4433,
+      "step": 14391
+    },
+    {
+      "epoch": 2.363578510869788,
+      "grad_norm": 0.3005983703147699,
+      "learning_rate": 4.959027441796107e-06,
+      "loss": 0.4535,
+      "step": 14392
+    },
+    {
+      "epoch": 2.3637427380781313,
+      "grad_norm": 0.4589025460002175,
+      "learning_rate": 4.958560913370363e-06,
+      "loss": 0.4363,
+      "step": 14393
+    },
+    {
+      "epoch": 2.363906965286474,
+      "grad_norm": 0.504401185692525,
+      "learning_rate": 4.958094379964196e-06,
+      "loss": 0.4448,
+      "step": 14394
+    },
+    {
+      "epoch": 2.3640711924948166,
+      "grad_norm": 0.29769659114726227,
+      "learning_rate": 4.9576278415826725e-06,
+      "loss": 0.4592,
+      "step": 14395
+    },
+    {
+      "epoch": 2.364235419703159,
+      "grad_norm": 0.29314450078630233,
+      "learning_rate": 4.957161298230861e-06,
+      "loss": 0.4562,
+      "step": 14396
+    },
+    {
+      "epoch": 2.364399646911502,
+      "grad_norm": 0.33090458937336886,
+      "learning_rate": 4.956694749913829e-06,
+      "loss": 0.4701,
+      "step": 14397
+    },
+    {
+      "epoch": 2.364563874119845,
+      "grad_norm": 0.2934499635019847,
+      "learning_rate": 4.956228196636643e-06,
+      "loss": 0.4375,
+      "step": 14398
+    },
+    {
+      "epoch": 2.3647281013281876,
+      "grad_norm": 0.3996609817040107,
+      "learning_rate": 4.9557616384043685e-06,
+      "loss": 0.4257,
+      "step": 14399
+    },
+    {
+      "epoch": 2.36489232853653,
+      "grad_norm": 0.32228752842073993,
+      "learning_rate": 4.955295075222074e-06,
+      "loss": 0.4337,
+      "step": 14400
+    },
+    {
+      "epoch": 2.3650565557448733,
+      "grad_norm": 0.3656720599456349,
+      "learning_rate": 4.954828507094828e-06,
+      "loss": 0.4355,
+      "step": 14401
+    },
+    {
+      "epoch": 2.365220782953216,
+      "grad_norm": 0.31683870295473615,
+      "learning_rate": 4.954361934027697e-06,
+      "loss": 0.4442,
+      "step": 14402
+    },
+    {
+      "epoch": 2.3653850101615586,
+      "grad_norm": 0.3107408972347142,
+      "learning_rate": 4.953895356025748e-06,
+      "loss": 0.4561,
+      "step": 14403
+    },
+    {
+      "epoch": 2.365549237369901,
+      "grad_norm": 0.32621837651394164,
+      "learning_rate": 4.9534287730940486e-06,
+      "loss": 0.4338,
+      "step": 14404
+    },
+    {
+      "epoch": 2.365713464578244,
+      "grad_norm": 0.29890532466434366,
+      "learning_rate": 4.952962185237667e-06,
+      "loss": 0.4527,
+      "step": 14405
+    },
+    {
+      "epoch": 2.365877691786587,
+      "grad_norm": 0.5040022006465182,
+      "learning_rate": 4.952495592461671e-06,
+      "loss": 0.4354,
+      "step": 14406
+    },
+    {
+      "epoch": 2.3660419189949295,
+      "grad_norm": 0.31470278545685143,
+      "learning_rate": 4.952028994771127e-06,
+      "loss": 0.4347,
+      "step": 14407
+    },
+    {
+      "epoch": 2.366206146203272,
+      "grad_norm": 0.3173507245055464,
+      "learning_rate": 4.951562392171103e-06,
+      "loss": 0.4437,
+      "step": 14408
+    },
+    {
+      "epoch": 2.366370373411615,
+      "grad_norm": 0.3154547169516024,
+      "learning_rate": 4.9510957846666665e-06,
+      "loss": 0.4601,
+      "step": 14409
+    },
+    {
+      "epoch": 2.366534600619958,
+      "grad_norm": 0.38068921310440645,
+      "learning_rate": 4.950629172262888e-06,
+      "loss": 0.4304,
+      "step": 14410
+    },
+    {
+      "epoch": 2.3666988278283005,
+      "grad_norm": 0.4921138013809312,
+      "learning_rate": 4.9501625549648315e-06,
+      "loss": 0.4571,
+      "step": 14411
+    },
+    {
+      "epoch": 2.366863055036643,
+      "grad_norm": 0.3367350539856582,
+      "learning_rate": 4.949695932777568e-06,
+      "loss": 0.4411,
+      "step": 14412
+    },
+    {
+      "epoch": 2.367027282244986,
+      "grad_norm": 0.2890413534719377,
+      "learning_rate": 4.949229305706163e-06,
+      "loss": 0.4387,
+      "step": 14413
+    },
+    {
+      "epoch": 2.3671915094533285,
+      "grad_norm": 0.29384243153256306,
+      "learning_rate": 4.948762673755688e-06,
+      "loss": 0.4309,
+      "step": 14414
+    },
+    {
+      "epoch": 2.3673557366616715,
+      "grad_norm": 0.30973836082355344,
+      "learning_rate": 4.948296036931206e-06,
+      "loss": 0.4332,
+      "step": 14415
+    },
+    {
+      "epoch": 2.367519963870014,
+      "grad_norm": 0.3154947522688878,
+      "learning_rate": 4.947829395237789e-06,
+      "loss": 0.4325,
+      "step": 14416
+    },
+    {
+      "epoch": 2.367684191078357,
+      "grad_norm": 0.3468663049157948,
+      "learning_rate": 4.947362748680506e-06,
+      "loss": 0.4446,
+      "step": 14417
+    },
+    {
+      "epoch": 2.3678484182867,
+      "grad_norm": 0.6625064405457736,
+      "learning_rate": 4.946896097264421e-06,
+      "loss": 0.4359,
+      "step": 14418
+    },
+    {
+      "epoch": 2.3680126454950425,
+      "grad_norm": 0.36755003431616734,
+      "learning_rate": 4.946429440994606e-06,
+      "loss": 0.4501,
+      "step": 14419
+    },
+    {
+      "epoch": 2.368176872703385,
+      "grad_norm": 0.3180908616830341,
+      "learning_rate": 4.945962779876127e-06,
+      "loss": 0.4367,
+      "step": 14420
+    },
+    {
+      "epoch": 2.368341099911728,
+      "grad_norm": 0.2969125348286323,
+      "learning_rate": 4.945496113914055e-06,
+      "loss": 0.4397,
+      "step": 14421
+    },
+    {
+      "epoch": 2.3685053271200704,
+      "grad_norm": 0.3484937349702599,
+      "learning_rate": 4.945029443113455e-06,
+      "loss": 0.4477,
+      "step": 14422
+    },
+    {
+      "epoch": 2.3686695543284135,
+      "grad_norm": 0.3364151217982416,
+      "learning_rate": 4.9445627674794e-06,
+      "loss": 0.4408,
+      "step": 14423
+    },
+    {
+      "epoch": 2.368833781536756,
+      "grad_norm": 0.27632261732359437,
+      "learning_rate": 4.944096087016953e-06,
+      "loss": 0.4327,
+      "step": 14424
+    },
+    {
+      "epoch": 2.368998008745099,
+      "grad_norm": 0.33643617461694814,
+      "learning_rate": 4.943629401731187e-06,
+      "loss": 0.4327,
+      "step": 14425
+    },
+    {
+      "epoch": 2.3691622359534414,
+      "grad_norm": 0.251371038402946,
+      "learning_rate": 4.94316271162717e-06,
+      "loss": 0.4295,
+      "step": 14426
+    },
+    {
+      "epoch": 2.3693264631617845,
+      "grad_norm": 0.273291389563221,
+      "learning_rate": 4.942696016709969e-06,
+      "loss": 0.4571,
+      "step": 14427
+    },
+    {
+      "epoch": 2.369490690370127,
+      "grad_norm": 0.3027602849916757,
+      "learning_rate": 4.942229316984654e-06,
+      "loss": 0.4303,
+      "step": 14428
+    },
+    {
+      "epoch": 2.36965491757847,
+      "grad_norm": 0.29359944961044193,
+      "learning_rate": 4.941762612456292e-06,
+      "loss": 0.447,
+      "step": 14429
+    },
+    {
+      "epoch": 2.3698191447868124,
+      "grad_norm": 0.30796589557839593,
+      "learning_rate": 4.941295903129954e-06,
+      "loss": 0.4462,
+      "step": 14430
+    },
+    {
+      "epoch": 2.369983371995155,
+      "grad_norm": 0.3321031226177786,
+      "learning_rate": 4.94082918901071e-06,
+      "loss": 0.442,
+      "step": 14431
+    },
+    {
+      "epoch": 2.370147599203498,
+      "grad_norm": 0.28573543373725724,
+      "learning_rate": 4.940362470103627e-06,
+      "loss": 0.4416,
+      "step": 14432
+    },
+    {
+      "epoch": 2.370311826411841,
+      "grad_norm": 0.34980867774696045,
+      "learning_rate": 4.939895746413773e-06,
+      "loss": 0.4405,
+      "step": 14433
+    },
+    {
+      "epoch": 2.3704760536201834,
+      "grad_norm": 0.38809888263301406,
+      "learning_rate": 4.939429017946218e-06,
+      "loss": 0.4441,
+      "step": 14434
+    },
+    {
+      "epoch": 2.3706402808285265,
+      "grad_norm": 0.2776636961242803,
+      "learning_rate": 4.938962284706034e-06,
+      "loss": 0.4449,
+      "step": 14435
+    },
+    {
+      "epoch": 2.370804508036869,
+      "grad_norm": 0.31336049708314334,
+      "learning_rate": 4.9384955466982855e-06,
+      "loss": 0.4198,
+      "step": 14436
+    },
+    {
+      "epoch": 2.370968735245212,
+      "grad_norm": 0.2897266594367285,
+      "learning_rate": 4.938028803928044e-06,
+      "loss": 0.4437,
+      "step": 14437
+    },
+    {
+      "epoch": 2.3711329624535544,
+      "grad_norm": 0.3676946069702938,
+      "learning_rate": 4.93756205640038e-06,
+      "loss": 0.4314,
+      "step": 14438
+    },
+    {
+      "epoch": 2.371297189661897,
+      "grad_norm": 0.29396140174936525,
+      "learning_rate": 4.937095304120362e-06,
+      "loss": 0.4426,
+      "step": 14439
+    },
+    {
+      "epoch": 2.37146141687024,
+      "grad_norm": 0.3154282077143417,
+      "learning_rate": 4.936628547093057e-06,
+      "loss": 0.4423,
+      "step": 14440
+    },
+    {
+      "epoch": 2.371625644078583,
+      "grad_norm": 0.5470785694493393,
+      "learning_rate": 4.936161785323538e-06,
+      "loss": 0.4369,
+      "step": 14441
+    },
+    {
+      "epoch": 2.3717898712869254,
+      "grad_norm": 0.2680266468948304,
+      "learning_rate": 4.935695018816872e-06,
+      "loss": 0.4333,
+      "step": 14442
+    },
+    {
+      "epoch": 2.371954098495268,
+      "grad_norm": 0.310046436229158,
+      "learning_rate": 4.935228247578129e-06,
+      "loss": 0.4326,
+      "step": 14443
+    },
+    {
+      "epoch": 2.372118325703611,
+      "grad_norm": 0.4633103707884507,
+      "learning_rate": 4.93476147161238e-06,
+      "loss": 0.4542,
+      "step": 14444
+    },
+    {
+      "epoch": 2.372282552911954,
+      "grad_norm": 0.4442241811118408,
+      "learning_rate": 4.9342946909246935e-06,
+      "loss": 0.439,
+      "step": 14445
+    },
+    {
+      "epoch": 2.3724467801202964,
+      "grad_norm": 0.3754637769732102,
+      "learning_rate": 4.933827905520139e-06,
+      "loss": 0.4583,
+      "step": 14446
+    },
+    {
+      "epoch": 2.372611007328639,
+      "grad_norm": 0.2746592828941314,
+      "learning_rate": 4.933361115403787e-06,
+      "loss": 0.4348,
+      "step": 14447
+    },
+    {
+      "epoch": 2.3727752345369817,
+      "grad_norm": 0.48162906768990915,
+      "learning_rate": 4.932894320580707e-06,
+      "loss": 0.4588,
+      "step": 14448
+    },
+    {
+      "epoch": 2.3729394617453248,
+      "grad_norm": 0.27825006803186564,
+      "learning_rate": 4.9324275210559675e-06,
+      "loss": 0.4456,
+      "step": 14449
+    },
+    {
+      "epoch": 2.3731036889536674,
+      "grad_norm": 0.6205579010136238,
+      "learning_rate": 4.931960716834641e-06,
+      "loss": 0.4453,
+      "step": 14450
+    },
+    {
+      "epoch": 2.37326791616201,
+      "grad_norm": 0.3124084196903969,
+      "learning_rate": 4.931493907921796e-06,
+      "loss": 0.4307,
+      "step": 14451
+    },
+    {
+      "epoch": 2.373432143370353,
+      "grad_norm": 0.3028093135665966,
+      "learning_rate": 4.931027094322503e-06,
+      "loss": 0.436,
+      "step": 14452
+    },
+    {
+      "epoch": 2.3735963705786958,
+      "grad_norm": 0.3348730023292204,
+      "learning_rate": 4.93056027604183e-06,
+      "loss": 0.4355,
+      "step": 14453
+    },
+    {
+      "epoch": 2.3737605977870384,
+      "grad_norm": 0.33042531193006286,
+      "learning_rate": 4.93009345308485e-06,
+      "loss": 0.4485,
+      "step": 14454
+    },
+    {
+      "epoch": 2.373924824995381,
+      "grad_norm": 0.28904118773468906,
+      "learning_rate": 4.929626625456633e-06,
+      "loss": 0.437,
+      "step": 14455
+    },
+    {
+      "epoch": 2.3740890522037237,
+      "grad_norm": 0.5665194365436337,
+      "learning_rate": 4.929159793162247e-06,
+      "loss": 0.4356,
+      "step": 14456
+    },
+    {
+      "epoch": 2.3742532794120668,
+      "grad_norm": 0.3139144028595357,
+      "learning_rate": 4.9286929562067635e-06,
+      "loss": 0.4295,
+      "step": 14457
+    },
+    {
+      "epoch": 2.3744175066204094,
+      "grad_norm": 0.3337755379913973,
+      "learning_rate": 4.928226114595252e-06,
+      "loss": 0.4425,
+      "step": 14458
+    },
+    {
+      "epoch": 2.374581733828752,
+      "grad_norm": 0.6960305232843772,
+      "learning_rate": 4.927759268332783e-06,
+      "loss": 0.4401,
+      "step": 14459
+    },
+    {
+      "epoch": 2.3747459610370947,
+      "grad_norm": 0.3636555112169848,
+      "learning_rate": 4.927292417424429e-06,
+      "loss": 0.4307,
+      "step": 14460
+    },
+    {
+      "epoch": 2.3749101882454378,
+      "grad_norm": 0.4151227793689786,
+      "learning_rate": 4.92682556187526e-06,
+      "loss": 0.4312,
+      "step": 14461
+    },
+    {
+      "epoch": 2.3750744154537804,
+      "grad_norm": 0.3043352830410617,
+      "learning_rate": 4.926358701690343e-06,
+      "loss": 0.4435,
+      "step": 14462
+    },
+    {
+      "epoch": 2.375238642662123,
+      "grad_norm": 0.41972681854749694,
+      "learning_rate": 4.925891836874751e-06,
+      "loss": 0.4381,
+      "step": 14463
+    },
+    {
+      "epoch": 2.3754028698704657,
+      "grad_norm": 0.3183437051748933,
+      "learning_rate": 4.925424967433557e-06,
+      "loss": 0.4492,
+      "step": 14464
+    },
+    {
+      "epoch": 2.3755670970788083,
+      "grad_norm": 0.3825905561550247,
+      "learning_rate": 4.924958093371828e-06,
+      "loss": 0.4457,
+      "step": 14465
+    },
+    {
+      "epoch": 2.3757313242871514,
+      "grad_norm": 0.3127106608442882,
+      "learning_rate": 4.924491214694636e-06,
+      "loss": 0.4438,
+      "step": 14466
+    },
+    {
+      "epoch": 2.375895551495494,
+      "grad_norm": 0.30704550035120054,
+      "learning_rate": 4.924024331407051e-06,
+      "loss": 0.4409,
+      "step": 14467
+    },
+    {
+      "epoch": 2.3760597787038367,
+      "grad_norm": 0.5557837596183262,
+      "learning_rate": 4.923557443514145e-06,
+      "loss": 0.4268,
+      "step": 14468
+    },
+    {
+      "epoch": 2.3762240059121797,
+      "grad_norm": 0.31056296034006503,
+      "learning_rate": 4.92309055102099e-06,
+      "loss": 0.4226,
+      "step": 14469
+    },
+    {
+      "epoch": 2.3763882331205224,
+      "grad_norm": 0.3614259977141636,
+      "learning_rate": 4.922623653932655e-06,
+      "loss": 0.437,
+      "step": 14470
+    },
+    {
+      "epoch": 2.376552460328865,
+      "grad_norm": 0.32688326087771474,
+      "learning_rate": 4.92215675225421e-06,
+      "loss": 0.4247,
+      "step": 14471
+    },
+    {
+      "epoch": 2.3767166875372077,
+      "grad_norm": 0.29109959383026757,
+      "learning_rate": 4.921689845990726e-06,
+      "loss": 0.4389,
+      "step": 14472
+    },
+    {
+      "epoch": 2.3768809147455503,
+      "grad_norm": 0.3201231594951292,
+      "learning_rate": 4.921222935147279e-06,
+      "loss": 0.4474,
+      "step": 14473
+    },
+    {
+      "epoch": 2.3770451419538934,
+      "grad_norm": 0.30953462437114765,
+      "learning_rate": 4.920756019728934e-06,
+      "loss": 0.4296,
+      "step": 14474
+    },
+    {
+      "epoch": 2.377209369162236,
+      "grad_norm": 0.33172813843029597,
+      "learning_rate": 4.9202890997407656e-06,
+      "loss": 0.4333,
+      "step": 14475
+    },
+    {
+      "epoch": 2.3773735963705787,
+      "grad_norm": 0.2837245728085035,
+      "learning_rate": 4.9198221751878435e-06,
+      "loss": 0.4304,
+      "step": 14476
+    },
+    {
+      "epoch": 2.3775378235789213,
+      "grad_norm": 0.36359957078420707,
+      "learning_rate": 4.919355246075241e-06,
+      "loss": 0.4302,
+      "step": 14477
+    },
+    {
+      "epoch": 2.3777020507872644,
+      "grad_norm": 0.3465557420820277,
+      "learning_rate": 4.918888312408026e-06,
+      "loss": 0.4385,
+      "step": 14478
+    },
+    {
+      "epoch": 2.377866277995607,
+      "grad_norm": 0.28285182301127754,
+      "learning_rate": 4.918421374191272e-06,
+      "loss": 0.4361,
+      "step": 14479
+    },
+    {
+      "epoch": 2.3780305052039497,
+      "grad_norm": 0.2794892656742042,
+      "learning_rate": 4.917954431430051e-06,
+      "loss": 0.4353,
+      "step": 14480
+    },
+    {
+      "epoch": 2.3781947324122923,
+      "grad_norm": 0.9462095011637344,
+      "learning_rate": 4.917487484129434e-06,
+      "loss": 0.442,
+      "step": 14481
+    },
+    {
+      "epoch": 2.378358959620635,
+      "grad_norm": 0.34605196349121137,
+      "learning_rate": 4.917020532294491e-06,
+      "loss": 0.4431,
+      "step": 14482
+    },
+    {
+      "epoch": 2.378523186828978,
+      "grad_norm": 0.29326419873734,
+      "learning_rate": 4.916553575930295e-06,
+      "loss": 0.429,
+      "step": 14483
+    },
+    {
+      "epoch": 2.3786874140373206,
+      "grad_norm": 0.32590839117850195,
+      "learning_rate": 4.9160866150419185e-06,
+      "loss": 0.4518,
+      "step": 14484
+    },
+    {
+      "epoch": 2.3788516412456633,
+      "grad_norm": 0.3073546322282407,
+      "learning_rate": 4.9156196496344315e-06,
+      "loss": 0.4276,
+      "step": 14485
+    },
+    {
+      "epoch": 2.3790158684540064,
+      "grad_norm": 0.36654859711759064,
+      "learning_rate": 4.915152679712905e-06,
+      "loss": 0.4439,
+      "step": 14486
+    },
+    {
+      "epoch": 2.379180095662349,
+      "grad_norm": 0.3038320472398642,
+      "learning_rate": 4.914685705282413e-06,
+      "loss": 0.4282,
+      "step": 14487
+    },
+    {
+      "epoch": 2.3793443228706916,
+      "grad_norm": 0.34343378873985786,
+      "learning_rate": 4.914218726348026e-06,
+      "loss": 0.4538,
+      "step": 14488
+    },
+    {
+      "epoch": 2.3795085500790343,
+      "grad_norm": 0.30341262594748736,
+      "learning_rate": 4.913751742914817e-06,
+      "loss": 0.4309,
+      "step": 14489
+    },
+    {
+      "epoch": 2.379672777287377,
+      "grad_norm": 0.29250488553478204,
+      "learning_rate": 4.913284754987856e-06,
+      "loss": 0.4135,
+      "step": 14490
+    },
+    {
+      "epoch": 2.37983700449572,
+      "grad_norm": 0.27498243268008893,
+      "learning_rate": 4.912817762572216e-06,
+      "loss": 0.4461,
+      "step": 14491
+    },
+    {
+      "epoch": 2.3800012317040626,
+      "grad_norm": 0.34557233743510063,
+      "learning_rate": 4.912350765672968e-06,
+      "loss": 0.419,
+      "step": 14492
+    },
+    {
+      "epoch": 2.3801654589124053,
+      "grad_norm": 0.2647786611846677,
+      "learning_rate": 4.911883764295186e-06,
+      "loss": 0.4275,
+      "step": 14493
+    },
+    {
+      "epoch": 2.380329686120748,
+      "grad_norm": 0.3375971662260146,
+      "learning_rate": 4.911416758443941e-06,
+      "loss": 0.4321,
+      "step": 14494
+    },
+    {
+      "epoch": 2.380493913329091,
+      "grad_norm": 0.29484030285466734,
+      "learning_rate": 4.910949748124306e-06,
+      "loss": 0.4441,
+      "step": 14495
+    },
+    {
+      "epoch": 2.3806581405374336,
+      "grad_norm": 0.3515144629906051,
+      "learning_rate": 4.9104827333413515e-06,
+      "loss": 0.4307,
+      "step": 14496
+    },
+    {
+      "epoch": 2.3808223677457763,
+      "grad_norm": 0.2737921671106952,
+      "learning_rate": 4.91001571410015e-06,
+      "loss": 0.4262,
+      "step": 14497
+    },
+    {
+      "epoch": 2.380986594954119,
+      "grad_norm": 0.3125825530773969,
+      "learning_rate": 4.909548690405777e-06,
+      "loss": 0.4474,
+      "step": 14498
+    },
+    {
+      "epoch": 2.3811508221624615,
+      "grad_norm": 0.3069349138335879,
+      "learning_rate": 4.909081662263299e-06,
+      "loss": 0.4183,
+      "step": 14499
+    },
+    {
+      "epoch": 2.3813150493708046,
+      "grad_norm": 0.31701743882648536,
+      "learning_rate": 4.908614629677794e-06,
+      "loss": 0.4235,
+      "step": 14500
+    },
+    {
+      "epoch": 2.3814792765791473,
+      "grad_norm": 0.34108602787338077,
+      "learning_rate": 4.908147592654332e-06,
+      "loss": 0.4231,
+      "step": 14501
+    },
+    {
+      "epoch": 2.38164350378749,
+      "grad_norm": 0.3935501178826395,
+      "learning_rate": 4.9076805511979845e-06,
+      "loss": 0.4539,
+      "step": 14502
+    },
+    {
+      "epoch": 2.381807730995833,
+      "grad_norm": 0.3131609275378316,
+      "learning_rate": 4.907213505313825e-06,
+      "loss": 0.4473,
+      "step": 14503
+    },
+    {
+      "epoch": 2.3819719582041756,
+      "grad_norm": 0.3002432702582786,
+      "learning_rate": 4.9067464550069275e-06,
+      "loss": 0.442,
+      "step": 14504
+    },
+    {
+      "epoch": 2.3821361854125183,
+      "grad_norm": 0.3096378279868265,
+      "learning_rate": 4.906279400282362e-06,
+      "loss": 0.4402,
+      "step": 14505
+    },
+    {
+      "epoch": 2.382300412620861,
+      "grad_norm": 0.2763473957324622,
+      "learning_rate": 4.905812341145204e-06,
+      "loss": 0.4298,
+      "step": 14506
+    },
+    {
+      "epoch": 2.3824646398292035,
+      "grad_norm": 0.26518791364843874,
+      "learning_rate": 4.905345277600524e-06,
+      "loss": 0.4342,
+      "step": 14507
+    },
+    {
+      "epoch": 2.3826288670375466,
+      "grad_norm": 0.2906595974916181,
+      "learning_rate": 4.904878209653394e-06,
+      "loss": 0.4429,
+      "step": 14508
+    },
+    {
+      "epoch": 2.3827930942458893,
+      "grad_norm": 0.3391787092663081,
+      "learning_rate": 4.904411137308889e-06,
+      "loss": 0.4355,
+      "step": 14509
+    },
+    {
+      "epoch": 2.382957321454232,
+      "grad_norm": 0.300445441947826,
+      "learning_rate": 4.9039440605720834e-06,
+      "loss": 0.4338,
+      "step": 14510
+    },
+    {
+      "epoch": 2.3831215486625745,
+      "grad_norm": 0.26723588598079895,
+      "learning_rate": 4.903476979448045e-06,
+      "loss": 0.4346,
+      "step": 14511
+    },
+    {
+      "epoch": 2.3832857758709176,
+      "grad_norm": 0.2838691958214916,
+      "learning_rate": 4.903009893941851e-06,
+      "loss": 0.4504,
+      "step": 14512
+    },
+    {
+      "epoch": 2.3834500030792602,
+      "grad_norm": 0.36319610087457616,
+      "learning_rate": 4.902542804058573e-06,
+      "loss": 0.4413,
+      "step": 14513
+    },
+    {
+      "epoch": 2.383614230287603,
+      "grad_norm": 0.4271662136062928,
+      "learning_rate": 4.902075709803284e-06,
+      "loss": 0.4456,
+      "step": 14514
+    },
+    {
+      "epoch": 2.3837784574959455,
+      "grad_norm": 0.5264160203486998,
+      "learning_rate": 4.901608611181057e-06,
+      "loss": 0.4405,
+      "step": 14515
+    },
+    {
+      "epoch": 2.383942684704288,
+      "grad_norm": 0.28902641518996575,
+      "learning_rate": 4.901141508196965e-06,
+      "loss": 0.4243,
+      "step": 14516
+    },
+    {
+      "epoch": 2.3841069119126312,
+      "grad_norm": 0.28998258912696195,
+      "learning_rate": 4.900674400856082e-06,
+      "loss": 0.4389,
+      "step": 14517
+    },
+    {
+      "epoch": 2.384271139120974,
+      "grad_norm": 0.297485830160264,
+      "learning_rate": 4.900207289163482e-06,
+      "loss": 0.4473,
+      "step": 14518
+    },
+    {
+      "epoch": 2.3844353663293165,
+      "grad_norm": 0.32857178706650425,
+      "learning_rate": 4.899740173124236e-06,
+      "loss": 0.437,
+      "step": 14519
+    },
+    {
+      "epoch": 2.3845995935376596,
+      "grad_norm": 0.33976806970867174,
+      "learning_rate": 4.899273052743418e-06,
+      "loss": 0.439,
+      "step": 14520
+    },
+    {
+      "epoch": 2.3847638207460022,
+      "grad_norm": 0.2981178461359611,
+      "learning_rate": 4.898805928026102e-06,
+      "loss": 0.4224,
+      "step": 14521
+    },
+    {
+      "epoch": 2.384928047954345,
+      "grad_norm": 0.40828126605470355,
+      "learning_rate": 4.8983387989773605e-06,
+      "loss": 0.4499,
+      "step": 14522
+    },
+    {
+      "epoch": 2.3850922751626875,
+      "grad_norm": 0.28611327371513295,
+      "learning_rate": 4.8978716656022686e-06,
+      "loss": 0.4221,
+      "step": 14523
+    },
+    {
+      "epoch": 2.38525650237103,
+      "grad_norm": 0.39824063682909056,
+      "learning_rate": 4.8974045279059e-06,
+      "loss": 0.4388,
+      "step": 14524
+    },
+    {
+      "epoch": 2.3854207295793732,
+      "grad_norm": 0.5609924737582214,
+      "learning_rate": 4.896937385893327e-06,
+      "loss": 0.4526,
+      "step": 14525
+    },
+    {
+      "epoch": 2.385584956787716,
+      "grad_norm": 0.31889911235677965,
+      "learning_rate": 4.896470239569622e-06,
+      "loss": 0.4425,
+      "step": 14526
+    },
+    {
+      "epoch": 2.3857491839960585,
+      "grad_norm": 0.302289456244296,
+      "learning_rate": 4.8960030889398605e-06,
+      "loss": 0.4016,
+      "step": 14527
+    },
+    {
+      "epoch": 2.385913411204401,
+      "grad_norm": 0.3105265703326025,
+      "learning_rate": 4.895535934009116e-06,
+      "loss": 0.4446,
+      "step": 14528
+    },
+    {
+      "epoch": 2.3860776384127442,
+      "grad_norm": 0.29037798400142933,
+      "learning_rate": 4.895068774782463e-06,
+      "loss": 0.4385,
+      "step": 14529
+    },
+    {
+      "epoch": 2.386241865621087,
+      "grad_norm": 0.3254765727629465,
+      "learning_rate": 4.894601611264973e-06,
+      "loss": 0.4645,
+      "step": 14530
+    },
+    {
+      "epoch": 2.3864060928294295,
+      "grad_norm": 0.28695733641227344,
+      "learning_rate": 4.894134443461723e-06,
+      "loss": 0.4496,
+      "step": 14531
+    },
+    {
+      "epoch": 2.386570320037772,
+      "grad_norm": 0.31091703634070156,
+      "learning_rate": 4.893667271377783e-06,
+      "loss": 0.4481,
+      "step": 14532
+    },
+    {
+      "epoch": 2.386734547246115,
+      "grad_norm": 0.3169028926886554,
+      "learning_rate": 4.8932000950182316e-06,
+      "loss": 0.4358,
+      "step": 14533
+    },
+    {
+      "epoch": 2.386898774454458,
+      "grad_norm": 0.32378659995173825,
+      "learning_rate": 4.892732914388138e-06,
+      "loss": 0.4216,
+      "step": 14534
+    },
+    {
+      "epoch": 2.3870630016628005,
+      "grad_norm": 0.3424600596134925,
+      "learning_rate": 4.89226572949258e-06,
+      "loss": 0.4347,
+      "step": 14535
+    },
+    {
+      "epoch": 2.387227228871143,
+      "grad_norm": 0.4083489609261254,
+      "learning_rate": 4.891798540336628e-06,
+      "loss": 0.4335,
+      "step": 14536
+    },
+    {
+      "epoch": 2.387391456079486,
+      "grad_norm": 0.31418734811658605,
+      "learning_rate": 4.891331346925361e-06,
+      "loss": 0.4284,
+      "step": 14537
+    },
+    {
+      "epoch": 2.387555683287829,
+      "grad_norm": 0.3118148361256724,
+      "learning_rate": 4.89086414926385e-06,
+      "loss": 0.4376,
+      "step": 14538
+    },
+    {
+      "epoch": 2.3877199104961715,
+      "grad_norm": 0.3205728376077856,
+      "learning_rate": 4.890396947357169e-06,
+      "loss": 0.4489,
+      "step": 14539
+    },
+    {
+      "epoch": 2.387884137704514,
+      "grad_norm": 0.3037510041274979,
+      "learning_rate": 4.889929741210394e-06,
+      "loss": 0.4459,
+      "step": 14540
+    },
+    {
+      "epoch": 2.3880483649128568,
+      "grad_norm": 0.29716183603278645,
+      "learning_rate": 4.889462530828597e-06,
+      "loss": 0.4372,
+      "step": 14541
+    },
+    {
+      "epoch": 2.3882125921212,
+      "grad_norm": 0.29372871219067426,
+      "learning_rate": 4.888995316216855e-06,
+      "loss": 0.4338,
+      "step": 14542
+    },
+    {
+      "epoch": 2.3883768193295425,
+      "grad_norm": 0.31376207464816375,
+      "learning_rate": 4.888528097380241e-06,
+      "loss": 0.4438,
+      "step": 14543
+    },
+    {
+      "epoch": 2.388541046537885,
+      "grad_norm": 0.3638794588880243,
+      "learning_rate": 4.88806087432383e-06,
+      "loss": 0.4374,
+      "step": 14544
+    },
+    {
+      "epoch": 2.3887052737462278,
+      "grad_norm": 0.4157192070799767,
+      "learning_rate": 4.8875936470526956e-06,
+      "loss": 0.4471,
+      "step": 14545
+    },
+    {
+      "epoch": 2.388869500954571,
+      "grad_norm": 0.4629312247969477,
+      "learning_rate": 4.887126415571912e-06,
+      "loss": 0.4498,
+      "step": 14546
+    },
+    {
+      "epoch": 2.3890337281629135,
+      "grad_norm": 0.35641292496732957,
+      "learning_rate": 4.886659179886555e-06,
+      "loss": 0.4136,
+      "step": 14547
+    },
+    {
+      "epoch": 2.389197955371256,
+      "grad_norm": 0.29823085621147055,
+      "learning_rate": 4.886191940001701e-06,
+      "loss": 0.4372,
+      "step": 14548
+    },
+    {
+      "epoch": 2.3893621825795988,
+      "grad_norm": 0.3648444197028791,
+      "learning_rate": 4.88572469592242e-06,
+      "loss": 0.4408,
+      "step": 14549
+    },
+    {
+      "epoch": 2.3895264097879414,
+      "grad_norm": 0.3131996192986303,
+      "learning_rate": 4.88525744765379e-06,
+      "loss": 0.4271,
+      "step": 14550
+    },
+    {
+      "epoch": 2.3896906369962845,
+      "grad_norm": 0.9393732568333811,
+      "learning_rate": 4.884790195200884e-06,
+      "loss": 0.4277,
+      "step": 14551
+    },
+    {
+      "epoch": 2.389854864204627,
+      "grad_norm": 0.49807294959646065,
+      "learning_rate": 4.88432293856878e-06,
+      "loss": 0.4137,
+      "step": 14552
+    },
+    {
+      "epoch": 2.3900190914129698,
+      "grad_norm": 0.2795714012023261,
+      "learning_rate": 4.88385567776255e-06,
+      "loss": 0.426,
+      "step": 14553
+    },
+    {
+      "epoch": 2.390183318621313,
+      "grad_norm": 0.3680107524133055,
+      "learning_rate": 4.883388412787269e-06,
+      "loss": 0.4374,
+      "step": 14554
+    },
+    {
+      "epoch": 2.3903475458296555,
+      "grad_norm": 0.2576176856027463,
+      "learning_rate": 4.882921143648013e-06,
+      "loss": 0.4269,
+      "step": 14555
+    },
+    {
+      "epoch": 2.390511773037998,
+      "grad_norm": 0.33334827288723884,
+      "learning_rate": 4.882453870349858e-06,
+      "loss": 0.4424,
+      "step": 14556
+    },
+    {
+      "epoch": 2.3906760002463407,
+      "grad_norm": 0.32855118523344423,
+      "learning_rate": 4.881986592897875e-06,
+      "loss": 0.4258,
+      "step": 14557
+    },
+    {
+      "epoch": 2.3908402274546834,
+      "grad_norm": 0.2641498222991579,
+      "learning_rate": 4.881519311297145e-06,
+      "loss": 0.4335,
+      "step": 14558
+    },
+    {
+      "epoch": 2.3910044546630265,
+      "grad_norm": 0.30007217089842214,
+      "learning_rate": 4.881052025552737e-06,
+      "loss": 0.4448,
+      "step": 14559
+    },
+    {
+      "epoch": 2.391168681871369,
+      "grad_norm": 0.2652220391149274,
+      "learning_rate": 4.880584735669731e-06,
+      "loss": 0.4102,
+      "step": 14560
+    },
+    {
+      "epoch": 2.3913329090797117,
+      "grad_norm": 0.7023406690762278,
+      "learning_rate": 4.880117441653199e-06,
+      "loss": 0.4472,
+      "step": 14561
+    },
+    {
+      "epoch": 2.3914971362880544,
+      "grad_norm": 0.3832103540577619,
+      "learning_rate": 4.879650143508217e-06,
+      "loss": 0.443,
+      "step": 14562
+    },
+    {
+      "epoch": 2.3916613634963975,
+      "grad_norm": 0.3768401551082294,
+      "learning_rate": 4.879182841239863e-06,
+      "loss": 0.4443,
+      "step": 14563
+    },
+    {
+      "epoch": 2.39182559070474,
+      "grad_norm": 0.529999664933227,
+      "learning_rate": 4.87871553485321e-06,
+      "loss": 0.4524,
+      "step": 14564
+    },
+    {
+      "epoch": 2.3919898179130827,
+      "grad_norm": 0.5284450836711834,
+      "learning_rate": 4.878248224353334e-06,
+      "loss": 0.4441,
+      "step": 14565
+    },
+    {
+      "epoch": 2.3921540451214254,
+      "grad_norm": 0.32265336090808,
+      "learning_rate": 4.877780909745308e-06,
+      "loss": 0.4271,
+      "step": 14566
+    },
+    {
+      "epoch": 2.392318272329768,
+      "grad_norm": 0.30649852459360966,
+      "learning_rate": 4.8773135910342105e-06,
+      "loss": 0.4307,
+      "step": 14567
+    },
+    {
+      "epoch": 2.392482499538111,
+      "grad_norm": 0.34600620366711066,
+      "learning_rate": 4.876846268225117e-06,
+      "loss": 0.426,
+      "step": 14568
+    },
+    {
+      "epoch": 2.3926467267464537,
+      "grad_norm": 0.2898991988895646,
+      "learning_rate": 4.876378941323102e-06,
+      "loss": 0.4361,
+      "step": 14569
+    },
+    {
+      "epoch": 2.3928109539547964,
+      "grad_norm": 0.31657080709066865,
+      "learning_rate": 4.87591161033324e-06,
+      "loss": 0.4468,
+      "step": 14570
+    },
+    {
+      "epoch": 2.392975181163139,
+      "grad_norm": 0.3682597438220124,
+      "learning_rate": 4.875444275260609e-06,
+      "loss": 0.4413,
+      "step": 14571
+    },
+    {
+      "epoch": 2.393139408371482,
+      "grad_norm": 0.2818996138587926,
+      "learning_rate": 4.8749769361102855e-06,
+      "loss": 0.4281,
+      "step": 14572
+    },
+    {
+      "epoch": 2.3933036355798247,
+      "grad_norm": 0.2760481393412478,
+      "learning_rate": 4.874509592887342e-06,
+      "loss": 0.4633,
+      "step": 14573
+    },
+    {
+      "epoch": 2.3934678627881674,
+      "grad_norm": 0.3090199962692466,
+      "learning_rate": 4.874042245596856e-06,
+      "loss": 0.4345,
+      "step": 14574
+    },
+    {
+      "epoch": 2.39363208999651,
+      "grad_norm": 0.3753236126182619,
+      "learning_rate": 4.873574894243902e-06,
+      "loss": 0.4334,
+      "step": 14575
+    },
+    {
+      "epoch": 2.393796317204853,
+      "grad_norm": 0.49666749402603405,
+      "learning_rate": 4.873107538833558e-06,
+      "loss": 0.4138,
+      "step": 14576
+    },
+    {
+      "epoch": 2.3939605444131957,
+      "grad_norm": 0.44873693999064757,
+      "learning_rate": 4.8726401793709e-06,
+      "loss": 0.4385,
+      "step": 14577
+    },
+    {
+      "epoch": 2.3941247716215384,
+      "grad_norm": 0.5624188771958173,
+      "learning_rate": 4.872172815861003e-06,
+      "loss": 0.4535,
+      "step": 14578
+    },
+    {
+      "epoch": 2.394288998829881,
+      "grad_norm": 0.36776524540715916,
+      "learning_rate": 4.871705448308942e-06,
+      "loss": 0.4431,
+      "step": 14579
+    },
+    {
+      "epoch": 2.394453226038224,
+      "grad_norm": 0.40112504999237575,
+      "learning_rate": 4.871238076719794e-06,
+      "loss": 0.4004,
+      "step": 14580
+    },
+    {
+      "epoch": 2.3946174532465667,
+      "grad_norm": 0.38816955768952405,
+      "learning_rate": 4.8707707010986365e-06,
+      "loss": 0.444,
+      "step": 14581
+    },
+    {
+      "epoch": 2.3947816804549094,
+      "grad_norm": 0.40890090055450345,
+      "learning_rate": 4.870303321450544e-06,
+      "loss": 0.4244,
+      "step": 14582
+    },
+    {
+      "epoch": 2.394945907663252,
+      "grad_norm": 0.347335496686221,
+      "learning_rate": 4.869835937780592e-06,
+      "loss": 0.4351,
+      "step": 14583
+    },
+    {
+      "epoch": 2.3951101348715946,
+      "grad_norm": 0.37816795611763476,
+      "learning_rate": 4.869368550093859e-06,
+      "loss": 0.4281,
+      "step": 14584
+    },
+    {
+      "epoch": 2.3952743620799377,
+      "grad_norm": 0.31915528359769213,
+      "learning_rate": 4.868901158395418e-06,
+      "loss": 0.4397,
+      "step": 14585
+    },
+    {
+      "epoch": 2.3954385892882804,
+      "grad_norm": 0.33747799228179265,
+      "learning_rate": 4.86843376269035e-06,
+      "loss": 0.4453,
+      "step": 14586
+    },
+    {
+      "epoch": 2.395602816496623,
+      "grad_norm": 0.3983997690577079,
+      "learning_rate": 4.867966362983728e-06,
+      "loss": 0.4484,
+      "step": 14587
+    },
+    {
+      "epoch": 2.3957670437049656,
+      "grad_norm": 0.475030543908353,
+      "learning_rate": 4.86749895928063e-06,
+      "loss": 0.428,
+      "step": 14588
+    },
+    {
+      "epoch": 2.3959312709133087,
+      "grad_norm": 0.3346654030306633,
+      "learning_rate": 4.86703155158613e-06,
+      "loss": 0.4342,
+      "step": 14589
+    },
+    {
+      "epoch": 2.3960954981216513,
+      "grad_norm": 0.286620871577916,
+      "learning_rate": 4.866564139905308e-06,
+      "loss": 0.4243,
+      "step": 14590
+    },
+    {
+      "epoch": 2.396259725329994,
+      "grad_norm": 0.3188961526696404,
+      "learning_rate": 4.866096724243238e-06,
+      "loss": 0.4302,
+      "step": 14591
+    },
+    {
+      "epoch": 2.3964239525383366,
+      "grad_norm": 0.4410066533907783,
+      "learning_rate": 4.8656293046049976e-06,
+      "loss": 0.4246,
+      "step": 14592
+    },
+    {
+      "epoch": 2.3965881797466797,
+      "grad_norm": 0.30105566284819324,
+      "learning_rate": 4.865161880995663e-06,
+      "loss": 0.4263,
+      "step": 14593
+    },
+    {
+      "epoch": 2.3967524069550223,
+      "grad_norm": 0.30798326084830907,
+      "learning_rate": 4.864694453420312e-06,
+      "loss": 0.447,
+      "step": 14594
+    },
+    {
+      "epoch": 2.396916634163365,
+      "grad_norm": 0.32150354987246904,
+      "learning_rate": 4.864227021884018e-06,
+      "loss": 0.4431,
+      "step": 14595
+    },
+    {
+      "epoch": 2.3970808613717076,
+      "grad_norm": 0.42606120355256905,
+      "learning_rate": 4.863759586391862e-06,
+      "loss": 0.4228,
+      "step": 14596
+    },
+    {
+      "epoch": 2.3972450885800507,
+      "grad_norm": 0.3043593996027058,
+      "learning_rate": 4.863292146948919e-06,
+      "loss": 0.4576,
+      "step": 14597
+    },
+    {
+      "epoch": 2.3974093157883933,
+      "grad_norm": 0.24487691882867033,
+      "learning_rate": 4.862824703560266e-06,
+      "loss": 0.4333,
+      "step": 14598
+    },
+    {
+      "epoch": 2.397573542996736,
+      "grad_norm": 0.4344466671080413,
+      "learning_rate": 4.862357256230979e-06,
+      "loss": 0.4571,
+      "step": 14599
+    },
+    {
+      "epoch": 2.3977377702050786,
+      "grad_norm": 0.4551198744757983,
+      "learning_rate": 4.861889804966136e-06,
+      "loss": 0.4434,
+      "step": 14600
+    },
+    {
+      "epoch": 2.3979019974134212,
+      "grad_norm": 0.3458326007684332,
+      "learning_rate": 4.861422349770814e-06,
+      "loss": 0.413,
+      "step": 14601
+    },
+    {
+      "epoch": 2.3980662246217643,
+      "grad_norm": 0.28749146184798846,
+      "learning_rate": 4.8609548906500895e-06,
+      "loss": 0.4352,
+      "step": 14602
+    },
+    {
+      "epoch": 2.398230451830107,
+      "grad_norm": 0.3190568605296688,
+      "learning_rate": 4.860487427609039e-06,
+      "loss": 0.412,
+      "step": 14603
+    },
+    {
+      "epoch": 2.3983946790384496,
+      "grad_norm": 0.30187148550272713,
+      "learning_rate": 4.860019960652741e-06,
+      "loss": 0.438,
+      "step": 14604
+    },
+    {
+      "epoch": 2.3985589062467922,
+      "grad_norm": 0.3017561582254539,
+      "learning_rate": 4.859552489786272e-06,
+      "loss": 0.4505,
+      "step": 14605
+    },
+    {
+      "epoch": 2.3987231334551353,
+      "grad_norm": 1.293891871086484,
+      "learning_rate": 4.85908501501471e-06,
+      "loss": 0.469,
+      "step": 14606
+    },
+    {
+      "epoch": 2.398887360663478,
+      "grad_norm": 0.5743909815937656,
+      "learning_rate": 4.858617536343131e-06,
+      "loss": 0.4301,
+      "step": 14607
+    },
+    {
+      "epoch": 2.3990515878718206,
+      "grad_norm": 0.30495791593097715,
+      "learning_rate": 4.858150053776612e-06,
+      "loss": 0.4623,
+      "step": 14608
+    },
+    {
+      "epoch": 2.3992158150801632,
+      "grad_norm": 0.34780932489427824,
+      "learning_rate": 4.857682567320231e-06,
+      "loss": 0.4429,
+      "step": 14609
+    },
+    {
+      "epoch": 2.3993800422885063,
+      "grad_norm": 0.3017423370784942,
+      "learning_rate": 4.857215076979065e-06,
+      "loss": 0.4366,
+      "step": 14610
+    },
+    {
+      "epoch": 2.399544269496849,
+      "grad_norm": 0.49134203031797363,
+      "learning_rate": 4.856747582758193e-06,
+      "loss": 0.4387,
+      "step": 14611
+    },
+    {
+      "epoch": 2.3997084967051916,
+      "grad_norm": 0.36562730185281356,
+      "learning_rate": 4.856280084662692e-06,
+      "loss": 0.4644,
+      "step": 14612
+    },
+    {
+      "epoch": 2.3998727239135342,
+      "grad_norm": 0.3097029681533732,
+      "learning_rate": 4.855812582697637e-06,
+      "loss": 0.4408,
+      "step": 14613
+    },
+    {
+      "epoch": 2.4000369511218773,
+      "grad_norm": 0.38873300419101187,
+      "learning_rate": 4.8553450768681075e-06,
+      "loss": 0.4431,
+      "step": 14614
+    },
+    {
+      "epoch": 2.40020117833022,
+      "grad_norm": 0.31761409622079567,
+      "learning_rate": 4.854877567179182e-06,
+      "loss": 0.4001,
+      "step": 14615
+    },
+    {
+      "epoch": 2.4003654055385626,
+      "grad_norm": 0.2777570280076414,
+      "learning_rate": 4.8544100536359375e-06,
+      "loss": 0.4245,
+      "step": 14616
+    },
+    {
+      "epoch": 2.4005296327469052,
+      "grad_norm": 0.31920044070247267,
+      "learning_rate": 4.853942536243449e-06,
+      "loss": 0.4412,
+      "step": 14617
+    },
+    {
+      "epoch": 2.400693859955248,
+      "grad_norm": 0.37106495638871273,
+      "learning_rate": 4.8534750150067965e-06,
+      "loss": 0.4423,
+      "step": 14618
+    },
+    {
+      "epoch": 2.400858087163591,
+      "grad_norm": 0.30136802472197904,
+      "learning_rate": 4.853007489931059e-06,
+      "loss": 0.4284,
+      "step": 14619
+    },
+    {
+      "epoch": 2.4010223143719336,
+      "grad_norm": 0.34743551515882315,
+      "learning_rate": 4.8525399610213115e-06,
+      "loss": 0.4357,
+      "step": 14620
+    },
+    {
+      "epoch": 2.4011865415802762,
+      "grad_norm": 0.4630662332483221,
+      "learning_rate": 4.852072428282635e-06,
+      "loss": 0.4507,
+      "step": 14621
+    },
+    {
+      "epoch": 2.401350768788619,
+      "grad_norm": 0.34421273907801864,
+      "learning_rate": 4.851604891720104e-06,
+      "loss": 0.4261,
+      "step": 14622
+    },
+    {
+      "epoch": 2.401514995996962,
+      "grad_norm": 0.30429866073920575,
+      "learning_rate": 4.851137351338798e-06,
+      "loss": 0.4375,
+      "step": 14623
+    },
+    {
+      "epoch": 2.4016792232053046,
+      "grad_norm": 0.30710285649408753,
+      "learning_rate": 4.850669807143795e-06,
+      "loss": 0.4438,
+      "step": 14624
+    },
+    {
+      "epoch": 2.401843450413647,
+      "grad_norm": 0.3111504370306694,
+      "learning_rate": 4.850202259140173e-06,
+      "loss": 0.4482,
+      "step": 14625
+    },
+    {
+      "epoch": 2.40200767762199,
+      "grad_norm": 0.2772359140975464,
+      "learning_rate": 4.84973470733301e-06,
+      "loss": 0.4459,
+      "step": 14626
+    },
+    {
+      "epoch": 2.402171904830333,
+      "grad_norm": 0.34639818656776106,
+      "learning_rate": 4.849267151727385e-06,
+      "loss": 0.4418,
+      "step": 14627
+    },
+    {
+      "epoch": 2.4023361320386756,
+      "grad_norm": 0.2964347781930256,
+      "learning_rate": 4.848799592328374e-06,
+      "loss": 0.4443,
+      "step": 14628
+    },
+    {
+      "epoch": 2.402500359247018,
+      "grad_norm": 0.33595929206815733,
+      "learning_rate": 4.848332029141055e-06,
+      "loss": 0.4301,
+      "step": 14629
+    },
+    {
+      "epoch": 2.402664586455361,
+      "grad_norm": 0.4015804960132361,
+      "learning_rate": 4.847864462170509e-06,
+      "loss": 0.4376,
+      "step": 14630
+    },
+    {
+      "epoch": 2.402828813663704,
+      "grad_norm": 0.36583852333102235,
+      "learning_rate": 4.847396891421814e-06,
+      "loss": 0.4408,
+      "step": 14631
+    },
+    {
+      "epoch": 2.4029930408720466,
+      "grad_norm": 0.3243934301895329,
+      "learning_rate": 4.8469293169000455e-06,
+      "loss": 0.432,
+      "step": 14632
+    },
+    {
+      "epoch": 2.403157268080389,
+      "grad_norm": 1.1242199526297239,
+      "learning_rate": 4.846461738610282e-06,
+      "loss": 0.4221,
+      "step": 14633
+    },
+    {
+      "epoch": 2.403321495288732,
+      "grad_norm": 0.6124803657632618,
+      "learning_rate": 4.845994156557604e-06,
+      "loss": 0.4373,
+      "step": 14634
+    },
+    {
+      "epoch": 2.4034857224970745,
+      "grad_norm": 0.31726548703950835,
+      "learning_rate": 4.84552657074709e-06,
+      "loss": 0.4681,
+      "step": 14635
+    },
+    {
+      "epoch": 2.4036499497054176,
+      "grad_norm": 0.3402830482709174,
+      "learning_rate": 4.845058981183817e-06,
+      "loss": 0.4427,
+      "step": 14636
+    },
+    {
+      "epoch": 2.40381417691376,
+      "grad_norm": 0.41415711718620096,
+      "learning_rate": 4.8445913878728644e-06,
+      "loss": 0.4509,
+      "step": 14637
+    },
+    {
+      "epoch": 2.403978404122103,
+      "grad_norm": 0.30062268597490205,
+      "learning_rate": 4.844123790819309e-06,
+      "loss": 0.4277,
+      "step": 14638
+    },
+    {
+      "epoch": 2.4041426313304455,
+      "grad_norm": 0.3777055667239158,
+      "learning_rate": 4.84365619002823e-06,
+      "loss": 0.4244,
+      "step": 14639
+    },
+    {
+      "epoch": 2.4043068585387886,
+      "grad_norm": 0.35823953526949825,
+      "learning_rate": 4.84318858550471e-06,
+      "loss": 0.4382,
+      "step": 14640
+    },
+    {
+      "epoch": 2.404471085747131,
+      "grad_norm": 0.29672629622495195,
+      "learning_rate": 4.842720977253822e-06,
+      "loss": 0.4321,
+      "step": 14641
+    },
+    {
+      "epoch": 2.404635312955474,
+      "grad_norm": 0.4199981609070766,
+      "learning_rate": 4.842253365280647e-06,
+      "loss": 0.4396,
+      "step": 14642
+    },
+    {
+      "epoch": 2.4047995401638165,
+      "grad_norm": 0.3244624973865936,
+      "learning_rate": 4.841785749590264e-06,
+      "loss": 0.4583,
+      "step": 14643
+    },
+    {
+      "epoch": 2.4049637673721596,
+      "grad_norm": 0.30771806534999885,
+      "learning_rate": 4.841318130187752e-06,
+      "loss": 0.4359,
+      "step": 14644
+    },
+    {
+      "epoch": 2.405127994580502,
+      "grad_norm": 0.29862090519592716,
+      "learning_rate": 4.840850507078189e-06,
+      "loss": 0.4414,
+      "step": 14645
+    },
+    {
+      "epoch": 2.405292221788845,
+      "grad_norm": 0.32810428176479883,
+      "learning_rate": 4.840382880266654e-06,
+      "loss": 0.4249,
+      "step": 14646
+    },
+    {
+      "epoch": 2.4054564489971875,
+      "grad_norm": 0.36740941744035893,
+      "learning_rate": 4.8399152497582255e-06,
+      "loss": 0.4569,
+      "step": 14647
+    },
+    {
+      "epoch": 2.4056206762055306,
+      "grad_norm": 0.2634679213019516,
+      "learning_rate": 4.839447615557984e-06,
+      "loss": 0.4461,
+      "step": 14648
+    },
+    {
+      "epoch": 2.405784903413873,
+      "grad_norm": 0.29881940701178017,
+      "learning_rate": 4.838979977671007e-06,
+      "loss": 0.4378,
+      "step": 14649
+    },
+    {
+      "epoch": 2.405949130622216,
+      "grad_norm": 0.29574009739461327,
+      "learning_rate": 4.838512336102374e-06,
+      "loss": 0.4135,
+      "step": 14650
+    },
+    {
+      "epoch": 2.4061133578305585,
+      "grad_norm": 0.30183737594772203,
+      "learning_rate": 4.838044690857163e-06,
+      "loss": 0.4418,
+      "step": 14651
+    },
+    {
+      "epoch": 2.406277585038901,
+      "grad_norm": 0.3247926430668289,
+      "learning_rate": 4.8375770419404566e-06,
+      "loss": 0.4605,
+      "step": 14652
+    },
+    {
+      "epoch": 2.406441812247244,
+      "grad_norm": 0.2958133357958267,
+      "learning_rate": 4.837109389357329e-06,
+      "loss": 0.4294,
+      "step": 14653
+    },
+    {
+      "epoch": 2.406606039455587,
+      "grad_norm": 0.36391652836693766,
+      "learning_rate": 4.836641733112861e-06,
+      "loss": 0.4277,
+      "step": 14654
+    },
+    {
+      "epoch": 2.4067702666639295,
+      "grad_norm": 0.332486098628157,
+      "learning_rate": 4.836174073212136e-06,
+      "loss": 0.4281,
+      "step": 14655
+    },
+    {
+      "epoch": 2.406934493872272,
+      "grad_norm": 0.31970043225608835,
+      "learning_rate": 4.835706409660227e-06,
+      "loss": 0.4152,
+      "step": 14656
+    },
+    {
+      "epoch": 2.407098721080615,
+      "grad_norm": 0.38834894755127186,
+      "learning_rate": 4.8352387424622174e-06,
+      "loss": 0.4234,
+      "step": 14657
+    },
+    {
+      "epoch": 2.407262948288958,
+      "grad_norm": 0.4903854226706532,
+      "learning_rate": 4.834771071623184e-06,
+      "loss": 0.4418,
+      "step": 14658
+    },
+    {
+      "epoch": 2.4074271754973005,
+      "grad_norm": 0.36363960384397537,
+      "learning_rate": 4.834303397148208e-06,
+      "loss": 0.4549,
+      "step": 14659
+    },
+    {
+      "epoch": 2.407591402705643,
+      "grad_norm": 0.29685954160347816,
+      "learning_rate": 4.8338357190423684e-06,
+      "loss": 0.4276,
+      "step": 14660
+    },
+    {
+      "epoch": 2.407755629913986,
+      "grad_norm": 0.29239727092336887,
+      "learning_rate": 4.833368037310746e-06,
+      "loss": 0.4477,
+      "step": 14661
+    },
+    {
+      "epoch": 2.407919857122329,
+      "grad_norm": 0.25791649206692546,
+      "learning_rate": 4.832900351958416e-06,
+      "loss": 0.4322,
+      "step": 14662
+    },
+    {
+      "epoch": 2.4080840843306714,
+      "grad_norm": 0.3088996194131839,
+      "learning_rate": 4.832432662990462e-06,
+      "loss": 0.4291,
+      "step": 14663
+    },
+    {
+      "epoch": 2.408248311539014,
+      "grad_norm": 0.41234587740782985,
+      "learning_rate": 4.831964970411962e-06,
+      "loss": 0.4531,
+      "step": 14664
+    },
+    {
+      "epoch": 2.408412538747357,
+      "grad_norm": 0.28658592688301243,
+      "learning_rate": 4.831497274227996e-06,
+      "loss": 0.4164,
+      "step": 14665
+    },
+    {
+      "epoch": 2.4085767659557,
+      "grad_norm": 0.340510853960804,
+      "learning_rate": 4.831029574443644e-06,
+      "loss": 0.4475,
+      "step": 14666
+    },
+    {
+      "epoch": 2.4087409931640424,
+      "grad_norm": 0.28144928044547207,
+      "learning_rate": 4.830561871063983e-06,
+      "loss": 0.4171,
+      "step": 14667
+    },
+    {
+      "epoch": 2.408905220372385,
+      "grad_norm": 0.28529132479570535,
+      "learning_rate": 4.830094164094096e-06,
+      "loss": 0.4457,
+      "step": 14668
+    },
+    {
+      "epoch": 2.4090694475807277,
+      "grad_norm": 0.3332582267058933,
+      "learning_rate": 4.829626453539062e-06,
+      "loss": 0.4128,
+      "step": 14669
+    },
+    {
+      "epoch": 2.409233674789071,
+      "grad_norm": 0.3058023319027211,
+      "learning_rate": 4.829158739403962e-06,
+      "loss": 0.4351,
+      "step": 14670
+    },
+    {
+      "epoch": 2.4093979019974134,
+      "grad_norm": 0.294349833850524,
+      "learning_rate": 4.828691021693872e-06,
+      "loss": 0.4613,
+      "step": 14671
+    },
+    {
+      "epoch": 2.409562129205756,
+      "grad_norm": 0.45108829225684566,
+      "learning_rate": 4.828223300413873e-06,
+      "loss": 0.4235,
+      "step": 14672
+    },
+    {
+      "epoch": 2.4097263564140987,
+      "grad_norm": 0.29688811430073514,
+      "learning_rate": 4.8277555755690495e-06,
+      "loss": 0.45,
+      "step": 14673
+    },
+    {
+      "epoch": 2.409890583622442,
+      "grad_norm": 0.3621360599742629,
+      "learning_rate": 4.827287847164475e-06,
+      "loss": 0.4473,
+      "step": 14674
+    },
+    {
+      "epoch": 2.4100548108307844,
+      "grad_norm": 0.27637854458418865,
+      "learning_rate": 4.826820115205234e-06,
+      "loss": 0.4417,
+      "step": 14675
+    },
+    {
+      "epoch": 2.410219038039127,
+      "grad_norm": 0.5931964755976143,
+      "learning_rate": 4.8263523796964045e-06,
+      "loss": 0.4292,
+      "step": 14676
+    },
+    {
+      "epoch": 2.4103832652474697,
+      "grad_norm": 0.27390405093577597,
+      "learning_rate": 4.825884640643067e-06,
+      "loss": 0.4545,
+      "step": 14677
+    },
+    {
+      "epoch": 2.410547492455813,
+      "grad_norm": 0.397743748353218,
+      "learning_rate": 4.825416898050302e-06,
+      "loss": 0.4375,
+      "step": 14678
+    },
+    {
+      "epoch": 2.4107117196641554,
+      "grad_norm": 0.31255784691419425,
+      "learning_rate": 4.82494915192319e-06,
+      "loss": 0.4449,
+      "step": 14679
+    },
+    {
+      "epoch": 2.410875946872498,
+      "grad_norm": 1.4403416299574465,
+      "learning_rate": 4.824481402266809e-06,
+      "loss": 0.442,
+      "step": 14680
+    },
+    {
+      "epoch": 2.4110401740808407,
+      "grad_norm": 0.33922366749113997,
+      "learning_rate": 4.82401364908624e-06,
+      "loss": 0.4288,
+      "step": 14681
+    },
+    {
+      "epoch": 2.411204401289184,
+      "grad_norm": 0.31757509275868,
+      "learning_rate": 4.823545892386567e-06,
+      "loss": 0.4422,
+      "step": 14682
+    },
+    {
+      "epoch": 2.4113686284975264,
+      "grad_norm": 0.304788723585256,
+      "learning_rate": 4.823078132172865e-06,
+      "loss": 0.4255,
+      "step": 14683
+    },
+    {
+      "epoch": 2.411532855705869,
+      "grad_norm": 0.27335390554437755,
+      "learning_rate": 4.822610368450218e-06,
+      "loss": 0.4283,
+      "step": 14684
+    },
+    {
+      "epoch": 2.4116970829142117,
+      "grad_norm": 0.3469198590453585,
+      "learning_rate": 4.822142601223704e-06,
+      "loss": 0.4274,
+      "step": 14685
+    },
+    {
+      "epoch": 2.4118613101225543,
+      "grad_norm": 0.3807364827960831,
+      "learning_rate": 4.8216748304984055e-06,
+      "loss": 0.4338,
+      "step": 14686
+    },
+    {
+      "epoch": 2.4120255373308974,
+      "grad_norm": 0.3371271371858553,
+      "learning_rate": 4.8212070562794e-06,
+      "loss": 0.435,
+      "step": 14687
+    },
+    {
+      "epoch": 2.41218976453924,
+      "grad_norm": 0.43622776633348503,
+      "learning_rate": 4.820739278571771e-06,
+      "loss": 0.4447,
+      "step": 14688
+    },
+    {
+      "epoch": 2.4123539917475827,
+      "grad_norm": 0.334119379770926,
+      "learning_rate": 4.820271497380598e-06,
+      "loss": 0.4353,
+      "step": 14689
+    },
+    {
+      "epoch": 2.4125182189559253,
+      "grad_norm": 0.41180417575499706,
+      "learning_rate": 4.819803712710961e-06,
+      "loss": 0.4331,
+      "step": 14690
+    },
+    {
+      "epoch": 2.4126824461642684,
+      "grad_norm": 0.31571247354022247,
+      "learning_rate": 4.819335924567942e-06,
+      "loss": 0.4396,
+      "step": 14691
+    },
+    {
+      "epoch": 2.412846673372611,
+      "grad_norm": 0.34680247785096713,
+      "learning_rate": 4.818868132956619e-06,
+      "loss": 0.4331,
+      "step": 14692
+    },
+    {
+      "epoch": 2.4130109005809537,
+      "grad_norm": 0.3080440473620233,
+      "learning_rate": 4.818400337882075e-06,
+      "loss": 0.43,
+      "step": 14693
+    },
+    {
+      "epoch": 2.4131751277892963,
+      "grad_norm": 0.2817006501283312,
+      "learning_rate": 4.8179325393493906e-06,
+      "loss": 0.449,
+      "step": 14694
+    },
+    {
+      "epoch": 2.4133393549976394,
+      "grad_norm": 0.3112562541286652,
+      "learning_rate": 4.817464737363646e-06,
+      "loss": 0.4373,
+      "step": 14695
+    },
+    {
+      "epoch": 2.413503582205982,
+      "grad_norm": 0.3130396013840511,
+      "learning_rate": 4.816996931929922e-06,
+      "loss": 0.4451,
+      "step": 14696
+    },
+    {
+      "epoch": 2.4136678094143247,
+      "grad_norm": 0.32321218750431113,
+      "learning_rate": 4.816529123053298e-06,
+      "loss": 0.4302,
+      "step": 14697
+    },
+    {
+      "epoch": 2.4138320366226673,
+      "grad_norm": 0.3698459892322004,
+      "learning_rate": 4.816061310738859e-06,
+      "loss": 0.4359,
+      "step": 14698
+    },
+    {
+      "epoch": 2.4139962638310104,
+      "grad_norm": 0.42413402293263597,
+      "learning_rate": 4.815593494991681e-06,
+      "loss": 0.4245,
+      "step": 14699
+    },
+    {
+      "epoch": 2.414160491039353,
+      "grad_norm": 0.30761605557733,
+      "learning_rate": 4.815125675816848e-06,
+      "loss": 0.4256,
+      "step": 14700
+    },
+    {
+      "epoch": 2.4143247182476957,
+      "grad_norm": 0.24302570900623616,
+      "learning_rate": 4.81465785321944e-06,
+      "loss": 0.4306,
+      "step": 14701
+    },
+    {
+      "epoch": 2.4144889454560383,
+      "grad_norm": 0.2927736876723968,
+      "learning_rate": 4.814190027204536e-06,
+      "loss": 0.4504,
+      "step": 14702
+    },
+    {
+      "epoch": 2.414653172664381,
+      "grad_norm": 0.38682221463914745,
+      "learning_rate": 4.813722197777223e-06,
+      "loss": 0.4254,
+      "step": 14703
+    },
+    {
+      "epoch": 2.414817399872724,
+      "grad_norm": 0.29190540997776415,
+      "learning_rate": 4.813254364942578e-06,
+      "loss": 0.4435,
+      "step": 14704
+    },
+    {
+      "epoch": 2.4149816270810667,
+      "grad_norm": 0.300206389002567,
+      "learning_rate": 4.81278652870568e-06,
+      "loss": 0.4327,
+      "step": 14705
+    },
+    {
+      "epoch": 2.4151458542894093,
+      "grad_norm": 0.39121257578811,
+      "learning_rate": 4.812318689071613e-06,
+      "loss": 0.4254,
+      "step": 14706
+    },
+    {
+      "epoch": 2.415310081497752,
+      "grad_norm": 0.2755867218892501,
+      "learning_rate": 4.811850846045459e-06,
+      "loss": 0.4484,
+      "step": 14707
+    },
+    {
+      "epoch": 2.415474308706095,
+      "grad_norm": 0.3882210233624939,
+      "learning_rate": 4.811382999632297e-06,
+      "loss": 0.4283,
+      "step": 14708
+    },
+    {
+      "epoch": 2.4156385359144377,
+      "grad_norm": 0.33194414181235643,
+      "learning_rate": 4.81091514983721e-06,
+      "loss": 0.4329,
+      "step": 14709
+    },
+    {
+      "epoch": 2.4158027631227803,
+      "grad_norm": 0.5485778450887616,
+      "learning_rate": 4.810447296665278e-06,
+      "loss": 0.4398,
+      "step": 14710
+    },
+    {
+      "epoch": 2.415966990331123,
+      "grad_norm": 0.2714483130812466,
+      "learning_rate": 4.809979440121583e-06,
+      "loss": 0.4438,
+      "step": 14711
+    },
+    {
+      "epoch": 2.416131217539466,
+      "grad_norm": 0.442827826944641,
+      "learning_rate": 4.8095115802112055e-06,
+      "loss": 0.4292,
+      "step": 14712
+    },
+    {
+      "epoch": 2.4162954447478087,
+      "grad_norm": 0.2997106074840441,
+      "learning_rate": 4.809043716939229e-06,
+      "loss": 0.43,
+      "step": 14713
+    },
+    {
+      "epoch": 2.4164596719561513,
+      "grad_norm": 0.3101144702754857,
+      "learning_rate": 4.8085758503107335e-06,
+      "loss": 0.4637,
+      "step": 14714
+    },
+    {
+      "epoch": 2.416623899164494,
+      "grad_norm": 0.4345488387599415,
+      "learning_rate": 4.8081079803308006e-06,
+      "loss": 0.441,
+      "step": 14715
+    },
+    {
+      "epoch": 2.416788126372837,
+      "grad_norm": 0.31524736528031577,
+      "learning_rate": 4.807640107004511e-06,
+      "loss": 0.4385,
+      "step": 14716
+    },
+    {
+      "epoch": 2.4169523535811797,
+      "grad_norm": 0.9910231052912541,
+      "learning_rate": 4.807172230336947e-06,
+      "loss": 0.451,
+      "step": 14717
+    },
+    {
+      "epoch": 2.4171165807895223,
+      "grad_norm": 0.353751030084181,
+      "learning_rate": 4.806704350333191e-06,
+      "loss": 0.4298,
+      "step": 14718
+    },
+    {
+      "epoch": 2.417280807997865,
+      "grad_norm": 0.2960552918752241,
+      "learning_rate": 4.8062364669983255e-06,
+      "loss": 0.4353,
+      "step": 14719
+    },
+    {
+      "epoch": 2.4174450352062076,
+      "grad_norm": 0.44033970780659726,
+      "learning_rate": 4.80576858033743e-06,
+      "loss": 0.4103,
+      "step": 14720
+    },
+    {
+      "epoch": 2.4176092624145507,
+      "grad_norm": 0.37219487130298756,
+      "learning_rate": 4.8053006903555846e-06,
+      "loss": 0.4263,
+      "step": 14721
+    },
+    {
+      "epoch": 2.4177734896228933,
+      "grad_norm": 0.30714372461167905,
+      "learning_rate": 4.804832797057875e-06,
+      "loss": 0.4256,
+      "step": 14722
+    },
+    {
+      "epoch": 2.417937716831236,
+      "grad_norm": 0.31272508721749476,
+      "learning_rate": 4.804364900449382e-06,
+      "loss": 0.4506,
+      "step": 14723
+    },
+    {
+      "epoch": 2.4181019440395786,
+      "grad_norm": 0.4030508319923359,
+      "learning_rate": 4.803897000535186e-06,
+      "loss": 0.4464,
+      "step": 14724
+    },
+    {
+      "epoch": 2.4182661712479216,
+      "grad_norm": 0.3141823854058452,
+      "learning_rate": 4.8034290973203695e-06,
+      "loss": 0.4415,
+      "step": 14725
+    },
+    {
+      "epoch": 2.4184303984562643,
+      "grad_norm": 0.33834956229051,
+      "learning_rate": 4.802961190810014e-06,
+      "loss": 0.4135,
+      "step": 14726
+    },
+    {
+      "epoch": 2.418594625664607,
+      "grad_norm": 0.33520231537538203,
+      "learning_rate": 4.802493281009202e-06,
+      "loss": 0.4389,
+      "step": 14727
+    },
+    {
+      "epoch": 2.4187588528729496,
+      "grad_norm": 0.289427403704886,
+      "learning_rate": 4.802025367923017e-06,
+      "loss": 0.4206,
+      "step": 14728
+    },
+    {
+      "epoch": 2.4189230800812926,
+      "grad_norm": 0.31902463019262406,
+      "learning_rate": 4.801557451556538e-06,
+      "loss": 0.4171,
+      "step": 14729
+    },
+    {
+      "epoch": 2.4190873072896353,
+      "grad_norm": 0.34344005937323363,
+      "learning_rate": 4.801089531914847e-06,
+      "loss": 0.4319,
+      "step": 14730
+    },
+    {
+      "epoch": 2.419251534497978,
+      "grad_norm": 0.3326226745260404,
+      "learning_rate": 4.80062160900303e-06,
+      "loss": 0.4359,
+      "step": 14731
+    },
+    {
+      "epoch": 2.4194157617063206,
+      "grad_norm": 0.34417344186097304,
+      "learning_rate": 4.800153682826166e-06,
+      "loss": 0.4375,
+      "step": 14732
+    },
+    {
+      "epoch": 2.4195799889146636,
+      "grad_norm": 0.2824968263701154,
+      "learning_rate": 4.799685753389338e-06,
+      "loss": 0.4317,
+      "step": 14733
+    },
+    {
+      "epoch": 2.4197442161230063,
+      "grad_norm": 0.4136191893113875,
+      "learning_rate": 4.799217820697626e-06,
+      "loss": 0.4159,
+      "step": 14734
+    },
+    {
+      "epoch": 2.419908443331349,
+      "grad_norm": 0.31603338795216673,
+      "learning_rate": 4.7987498847561146e-06,
+      "loss": 0.4325,
+      "step": 14735
+    },
+    {
+      "epoch": 2.4200726705396916,
+      "grad_norm": 0.34601978001013006,
+      "learning_rate": 4.798281945569887e-06,
+      "loss": 0.4193,
+      "step": 14736
+    },
+    {
+      "epoch": 2.420236897748034,
+      "grad_norm": 0.31763167069909876,
+      "learning_rate": 4.797814003144023e-06,
+      "loss": 0.4432,
+      "step": 14737
+    },
+    {
+      "epoch": 2.4204011249563773,
+      "grad_norm": 0.47743166016080013,
+      "learning_rate": 4.797346057483606e-06,
+      "loss": 0.4225,
+      "step": 14738
+    },
+    {
+      "epoch": 2.42056535216472,
+      "grad_norm": 0.36547044513866017,
+      "learning_rate": 4.796878108593718e-06,
+      "loss": 0.4388,
+      "step": 14739
+    },
+    {
+      "epoch": 2.4207295793730625,
+      "grad_norm": 0.3777551355717173,
+      "learning_rate": 4.796410156479443e-06,
+      "loss": 0.4363,
+      "step": 14740
+    },
+    {
+      "epoch": 2.420893806581405,
+      "grad_norm": 0.35697617648628965,
+      "learning_rate": 4.79594220114586e-06,
+      "loss": 0.4224,
+      "step": 14741
+    },
+    {
+      "epoch": 2.4210580337897483,
+      "grad_norm": 0.3862378017848256,
+      "learning_rate": 4.795474242598054e-06,
+      "loss": 0.4375,
+      "step": 14742
+    },
+    {
+      "epoch": 2.421222260998091,
+      "grad_norm": 0.30417337392006183,
+      "learning_rate": 4.7950062808411085e-06,
+      "loss": 0.4333,
+      "step": 14743
+    },
+    {
+      "epoch": 2.4213864882064335,
+      "grad_norm": 0.42732171860737683,
+      "learning_rate": 4.794538315880103e-06,
+      "loss": 0.4333,
+      "step": 14744
+    },
+    {
+      "epoch": 2.421550715414776,
+      "grad_norm": 0.31445632031102666,
+      "learning_rate": 4.7940703477201225e-06,
+      "loss": 0.4304,
+      "step": 14745
+    },
+    {
+      "epoch": 2.4217149426231193,
+      "grad_norm": 0.6221104220450133,
+      "learning_rate": 4.793602376366248e-06,
+      "loss": 0.4356,
+      "step": 14746
+    },
+    {
+      "epoch": 2.421879169831462,
+      "grad_norm": 0.32177085663706345,
+      "learning_rate": 4.793134401823564e-06,
+      "loss": 0.4323,
+      "step": 14747
+    },
+    {
+      "epoch": 2.4220433970398045,
+      "grad_norm": 0.410211944772348,
+      "learning_rate": 4.792666424097151e-06,
+      "loss": 0.4331,
+      "step": 14748
+    },
+    {
+      "epoch": 2.422207624248147,
+      "grad_norm": 0.3075597976329248,
+      "learning_rate": 4.7921984431920935e-06,
+      "loss": 0.43,
+      "step": 14749
+    },
+    {
+      "epoch": 2.4223718514564903,
+      "grad_norm": 0.35261253135593046,
+      "learning_rate": 4.791730459113472e-06,
+      "loss": 0.4492,
+      "step": 14750
+    },
+    {
+      "epoch": 2.422536078664833,
+      "grad_norm": 0.36899376715600435,
+      "learning_rate": 4.791262471866372e-06,
+      "loss": 0.4108,
+      "step": 14751
+    },
+    {
+      "epoch": 2.4227003058731755,
+      "grad_norm": 0.37434397764969857,
+      "learning_rate": 4.790794481455874e-06,
+      "loss": 0.4432,
+      "step": 14752
+    },
+    {
+      "epoch": 2.422864533081518,
+      "grad_norm": 0.31972500121132164,
+      "learning_rate": 4.7903264878870635e-06,
+      "loss": 0.4082,
+      "step": 14753
+    },
+    {
+      "epoch": 2.423028760289861,
+      "grad_norm": 0.3123339025336729,
+      "learning_rate": 4.78985849116502e-06,
+      "loss": 0.4427,
+      "step": 14754
+    },
+    {
+      "epoch": 2.423192987498204,
+      "grad_norm": 0.2795028420999253,
+      "learning_rate": 4.789390491294827e-06,
+      "loss": 0.433,
+      "step": 14755
+    },
+    {
+      "epoch": 2.4233572147065465,
+      "grad_norm": 0.328024548668862,
+      "learning_rate": 4.78892248828157e-06,
+      "loss": 0.4298,
+      "step": 14756
+    },
+    {
+      "epoch": 2.423521441914889,
+      "grad_norm": 0.31231298140414665,
+      "learning_rate": 4.788454482130332e-06,
+      "loss": 0.4279,
+      "step": 14757
+    },
+    {
+      "epoch": 2.423685669123232,
+      "grad_norm": 0.3172952986153162,
+      "learning_rate": 4.787986472846193e-06,
+      "loss": 0.4454,
+      "step": 14758
+    },
+    {
+      "epoch": 2.423849896331575,
+      "grad_norm": 0.35895962136883086,
+      "learning_rate": 4.787518460434238e-06,
+      "loss": 0.4469,
+      "step": 14759
+    },
+    {
+      "epoch": 2.4240141235399175,
+      "grad_norm": 0.25783531582906527,
+      "learning_rate": 4.7870504448995495e-06,
+      "loss": 0.4376,
+      "step": 14760
+    },
+    {
+      "epoch": 2.42417835074826,
+      "grad_norm": 0.31120025248246125,
+      "learning_rate": 4.78658242624721e-06,
+      "loss": 0.4412,
+      "step": 14761
+    },
+    {
+      "epoch": 2.424342577956603,
+      "grad_norm": 0.28799008577143,
+      "learning_rate": 4.786114404482305e-06,
+      "loss": 0.4224,
+      "step": 14762
+    },
+    {
+      "epoch": 2.424506805164946,
+      "grad_norm": 0.32125630055880444,
+      "learning_rate": 4.7856463796099156e-06,
+      "loss": 0.4404,
+      "step": 14763
+    },
+    {
+      "epoch": 2.4246710323732885,
+      "grad_norm": 0.32524375153824536,
+      "learning_rate": 4.785178351635124e-06,
+      "loss": 0.4425,
+      "step": 14764
+    },
+    {
+      "epoch": 2.424835259581631,
+      "grad_norm": 0.43420083943177323,
+      "learning_rate": 4.784710320563016e-06,
+      "loss": 0.4355,
+      "step": 14765
+    },
+    {
+      "epoch": 2.424999486789974,
+      "grad_norm": 0.34655407456659604,
+      "learning_rate": 4.784242286398674e-06,
+      "loss": 0.4326,
+      "step": 14766
+    },
+    {
+      "epoch": 2.425163713998317,
+      "grad_norm": 0.2747114594008047,
+      "learning_rate": 4.78377424914718e-06,
+      "loss": 0.426,
+      "step": 14767
+    },
+    {
+      "epoch": 2.4253279412066595,
+      "grad_norm": 0.34923290616290603,
+      "learning_rate": 4.78330620881362e-06,
+      "loss": 0.4456,
+      "step": 14768
+    },
+    {
+      "epoch": 2.425492168415002,
+      "grad_norm": 0.26275802173147667,
+      "learning_rate": 4.782838165403076e-06,
+      "loss": 0.4401,
+      "step": 14769
+    },
+    {
+      "epoch": 2.425656395623345,
+      "grad_norm": 0.3533306405127362,
+      "learning_rate": 4.7823701189206295e-06,
+      "loss": 0.4389,
+      "step": 14770
+    },
+    {
+      "epoch": 2.4258206228316874,
+      "grad_norm": 0.3336264849677881,
+      "learning_rate": 4.781902069371367e-06,
+      "loss": 0.4475,
+      "step": 14771
+    },
+    {
+      "epoch": 2.4259848500400305,
+      "grad_norm": 0.496776075438404,
+      "learning_rate": 4.78143401676037e-06,
+      "loss": 0.4364,
+      "step": 14772
+    },
+    {
+      "epoch": 2.426149077248373,
+      "grad_norm": 0.3384395539167162,
+      "learning_rate": 4.780965961092722e-06,
+      "loss": 0.4439,
+      "step": 14773
+    },
+    {
+      "epoch": 2.426313304456716,
+      "grad_norm": 0.33369923606988255,
+      "learning_rate": 4.78049790237351e-06,
+      "loss": 0.4317,
+      "step": 14774
+    },
+    {
+      "epoch": 2.4264775316650584,
+      "grad_norm": 0.29929761882085637,
+      "learning_rate": 4.780029840607812e-06,
+      "loss": 0.4413,
+      "step": 14775
+    },
+    {
+      "epoch": 2.4266417588734015,
+      "grad_norm": 0.30729134782795475,
+      "learning_rate": 4.7795617758007145e-06,
+      "loss": 0.429,
+      "step": 14776
+    },
+    {
+      "epoch": 2.426805986081744,
+      "grad_norm": 0.28668811524646165,
+      "learning_rate": 4.779093707957303e-06,
+      "loss": 0.4459,
+      "step": 14777
+    },
+    {
+      "epoch": 2.4269702132900868,
+      "grad_norm": 0.9480953824797685,
+      "learning_rate": 4.7786256370826586e-06,
+      "loss": 0.4403,
+      "step": 14778
+    },
+    {
+      "epoch": 2.4271344404984294,
+      "grad_norm": 0.2905434139198909,
+      "learning_rate": 4.7781575631818645e-06,
+      "loss": 0.44,
+      "step": 14779
+    },
+    {
+      "epoch": 2.4272986677067725,
+      "grad_norm": 0.36488714133691247,
+      "learning_rate": 4.777689486260006e-06,
+      "loss": 0.4242,
+      "step": 14780
+    },
+    {
+      "epoch": 2.427462894915115,
+      "grad_norm": 0.5288134862615336,
+      "learning_rate": 4.777221406322168e-06,
+      "loss": 0.4292,
+      "step": 14781
+    },
+    {
+      "epoch": 2.4276271221234578,
+      "grad_norm": 0.2905834476410685,
+      "learning_rate": 4.776753323373432e-06,
+      "loss": 0.4505,
+      "step": 14782
+    },
+    {
+      "epoch": 2.4277913493318004,
+      "grad_norm": 0.33826065646395237,
+      "learning_rate": 4.7762852374188815e-06,
+      "loss": 0.4299,
+      "step": 14783
+    },
+    {
+      "epoch": 2.4279555765401435,
+      "grad_norm": 0.40177622398477075,
+      "learning_rate": 4.7758171484636015e-06,
+      "loss": 0.4552,
+      "step": 14784
+    },
+    {
+      "epoch": 2.428119803748486,
+      "grad_norm": 0.27650968021480343,
+      "learning_rate": 4.775349056512676e-06,
+      "loss": 0.431,
+      "step": 14785
+    },
+    {
+      "epoch": 2.4282840309568288,
+      "grad_norm": 0.4170100154683804,
+      "learning_rate": 4.77488096157119e-06,
+      "loss": 0.4399,
+      "step": 14786
+    },
+    {
+      "epoch": 2.4284482581651714,
+      "grad_norm": 0.3064746958240771,
+      "learning_rate": 4.774412863644227e-06,
+      "loss": 0.4325,
+      "step": 14787
+    },
+    {
+      "epoch": 2.428612485373514,
+      "grad_norm": 0.6095324612409638,
+      "learning_rate": 4.773944762736868e-06,
+      "loss": 0.4377,
+      "step": 14788
+    },
+    {
+      "epoch": 2.428776712581857,
+      "grad_norm": 0.2843106785607604,
+      "learning_rate": 4.7734766588542005e-06,
+      "loss": 0.4241,
+      "step": 14789
+    },
+    {
+      "epoch": 2.4289409397901998,
+      "grad_norm": 0.3693598879437371,
+      "learning_rate": 4.773008552001308e-06,
+      "loss": 0.4398,
+      "step": 14790
+    },
+    {
+      "epoch": 2.4291051669985424,
+      "grad_norm": 0.2911478294752035,
+      "learning_rate": 4.772540442183273e-06,
+      "loss": 0.431,
+      "step": 14791
+    },
+    {
+      "epoch": 2.429269394206885,
+      "grad_norm": 0.27983158639217526,
+      "learning_rate": 4.772072329405182e-06,
+      "loss": 0.432,
+      "step": 14792
+    },
+    {
+      "epoch": 2.429433621415228,
+      "grad_norm": 0.35636279705662716,
+      "learning_rate": 4.771604213672116e-06,
+      "loss": 0.4412,
+      "step": 14793
+    },
+    {
+      "epoch": 2.4295978486235708,
+      "grad_norm": 0.3399555636216479,
+      "learning_rate": 4.771136094989162e-06,
+      "loss": 0.435,
+      "step": 14794
+    },
+    {
+      "epoch": 2.4297620758319134,
+      "grad_norm": 0.34618491205707147,
+      "learning_rate": 4.770667973361403e-06,
+      "loss": 0.4566,
+      "step": 14795
+    },
+    {
+      "epoch": 2.429926303040256,
+      "grad_norm": 0.28967649101482,
+      "learning_rate": 4.770199848793924e-06,
+      "loss": 0.4202,
+      "step": 14796
+    },
+    {
+      "epoch": 2.430090530248599,
+      "grad_norm": 0.2779166934004358,
+      "learning_rate": 4.7697317212918075e-06,
+      "loss": 0.4289,
+      "step": 14797
+    },
+    {
+      "epoch": 2.4302547574569417,
+      "grad_norm": 0.2812164799099954,
+      "learning_rate": 4.76926359086014e-06,
+      "loss": 0.4519,
+      "step": 14798
+    },
+    {
+      "epoch": 2.4304189846652844,
+      "grad_norm": 0.32263573432382836,
+      "learning_rate": 4.768795457504005e-06,
+      "loss": 0.4667,
+      "step": 14799
+    },
+    {
+      "epoch": 2.430583211873627,
+      "grad_norm": 0.28947225570522306,
+      "learning_rate": 4.768327321228487e-06,
+      "loss": 0.4348,
+      "step": 14800
+    },
+    {
+      "epoch": 2.43074743908197,
+      "grad_norm": 0.2764048458026499,
+      "learning_rate": 4.7678591820386705e-06,
+      "loss": 0.4371,
+      "step": 14801
+    },
+    {
+      "epoch": 2.4309116662903127,
+      "grad_norm": 0.2854607261745386,
+      "learning_rate": 4.7673910399396396e-06,
+      "loss": 0.4262,
+      "step": 14802
+    },
+    {
+      "epoch": 2.4310758934986554,
+      "grad_norm": 0.26922749726743767,
+      "learning_rate": 4.766922894936479e-06,
+      "loss": 0.4454,
+      "step": 14803
+    },
+    {
+      "epoch": 2.431240120706998,
+      "grad_norm": 0.2646681350023495,
+      "learning_rate": 4.766454747034273e-06,
+      "loss": 0.4247,
+      "step": 14804
+    },
+    {
+      "epoch": 2.4314043479153407,
+      "grad_norm": 0.34550230812725563,
+      "learning_rate": 4.765986596238106e-06,
+      "loss": 0.4341,
+      "step": 14805
+    },
+    {
+      "epoch": 2.4315685751236837,
+      "grad_norm": 0.2869108230808502,
+      "learning_rate": 4.765518442553063e-06,
+      "loss": 0.417,
+      "step": 14806
+    },
+    {
+      "epoch": 2.4317328023320264,
+      "grad_norm": 0.2849337599012945,
+      "learning_rate": 4.765050285984229e-06,
+      "loss": 0.4495,
+      "step": 14807
+    },
+    {
+      "epoch": 2.431897029540369,
+      "grad_norm": 0.4105925836148604,
+      "learning_rate": 4.7645821265366875e-06,
+      "loss": 0.425,
+      "step": 14808
+    },
+    {
+      "epoch": 2.4320612567487117,
+      "grad_norm": 0.3121318395626099,
+      "learning_rate": 4.764113964215523e-06,
+      "loss": 0.4407,
+      "step": 14809
+    },
+    {
+      "epoch": 2.4322254839570547,
+      "grad_norm": 0.28767812785841906,
+      "learning_rate": 4.763645799025822e-06,
+      "loss": 0.4481,
+      "step": 14810
+    },
+    {
+      "epoch": 2.4323897111653974,
+      "grad_norm": 0.3154937609679195,
+      "learning_rate": 4.763177630972669e-06,
+      "loss": 0.4548,
+      "step": 14811
+    },
+    {
+      "epoch": 2.43255393837374,
+      "grad_norm": 0.3992594414402425,
+      "learning_rate": 4.762709460061147e-06,
+      "loss": 0.4427,
+      "step": 14812
+    },
+    {
+      "epoch": 2.4327181655820826,
+      "grad_norm": 0.3513545926571055,
+      "learning_rate": 4.762241286296342e-06,
+      "loss": 0.47,
+      "step": 14813
+    },
+    {
+      "epoch": 2.4328823927904257,
+      "grad_norm": 0.29704178929318303,
+      "learning_rate": 4.761773109683338e-06,
+      "loss": 0.4328,
+      "step": 14814
+    },
+    {
+      "epoch": 2.4330466199987684,
+      "grad_norm": 0.31782862402263434,
+      "learning_rate": 4.761304930227222e-06,
+      "loss": 0.4252,
+      "step": 14815
+    },
+    {
+      "epoch": 2.433210847207111,
+      "grad_norm": 0.28176998645123574,
+      "learning_rate": 4.760836747933077e-06,
+      "loss": 0.4255,
+      "step": 14816
+    },
+    {
+      "epoch": 2.4333750744154536,
+      "grad_norm": 0.3323179661972225,
+      "learning_rate": 4.760368562805988e-06,
+      "loss": 0.4447,
+      "step": 14817
+    },
+    {
+      "epoch": 2.4335393016237967,
+      "grad_norm": 0.3144062587438668,
+      "learning_rate": 4.759900374851038e-06,
+      "loss": 0.431,
+      "step": 14818
+    },
+    {
+      "epoch": 2.4337035288321394,
+      "grad_norm": 0.31873388546233744,
+      "learning_rate": 4.759432184073317e-06,
+      "loss": 0.43,
+      "step": 14819
+    },
+    {
+      "epoch": 2.433867756040482,
+      "grad_norm": 0.35986862702714584,
+      "learning_rate": 4.758963990477906e-06,
+      "loss": 0.4175,
+      "step": 14820
+    },
+    {
+      "epoch": 2.4340319832488246,
+      "grad_norm": 0.4216160669476413,
+      "learning_rate": 4.758495794069893e-06,
+      "loss": 0.4186,
+      "step": 14821
+    },
+    {
+      "epoch": 2.4341962104571673,
+      "grad_norm": 0.2656357025434144,
+      "learning_rate": 4.758027594854359e-06,
+      "loss": 0.4245,
+      "step": 14822
+    },
+    {
+      "epoch": 2.4343604376655104,
+      "grad_norm": 0.27275238122045936,
+      "learning_rate": 4.757559392836393e-06,
+      "loss": 0.4171,
+      "step": 14823
+    },
+    {
+      "epoch": 2.434524664873853,
+      "grad_norm": 0.4642214461339467,
+      "learning_rate": 4.757091188021078e-06,
+      "loss": 0.4447,
+      "step": 14824
+    },
+    {
+      "epoch": 2.4346888920821956,
+      "grad_norm": 0.32736316829012824,
+      "learning_rate": 4.7566229804135e-06,
+      "loss": 0.4379,
+      "step": 14825
+    },
+    {
+      "epoch": 2.4348531192905383,
+      "grad_norm": 0.3461060149654547,
+      "learning_rate": 4.7561547700187435e-06,
+      "loss": 0.4493,
+      "step": 14826
+    },
+    {
+      "epoch": 2.4350173464988814,
+      "grad_norm": 0.3414152378529399,
+      "learning_rate": 4.755686556841894e-06,
+      "loss": 0.4467,
+      "step": 14827
+    },
+    {
+      "epoch": 2.435181573707224,
+      "grad_norm": 0.3408110403504788,
+      "learning_rate": 4.755218340888038e-06,
+      "loss": 0.436,
+      "step": 14828
+    },
+    {
+      "epoch": 2.4353458009155666,
+      "grad_norm": 0.2892055813816026,
+      "learning_rate": 4.754750122162258e-06,
+      "loss": 0.4512,
+      "step": 14829
+    },
+    {
+      "epoch": 2.4355100281239093,
+      "grad_norm": 0.6586018155691247,
+      "learning_rate": 4.754281900669644e-06,
+      "loss": 0.4342,
+      "step": 14830
+    },
+    {
+      "epoch": 2.4356742553322523,
+      "grad_norm": 0.3309761857325862,
+      "learning_rate": 4.753813676415275e-06,
+      "loss": 0.435,
+      "step": 14831
+    },
+    {
+      "epoch": 2.435838482540595,
+      "grad_norm": 0.47816462488938366,
+      "learning_rate": 4.753345449404242e-06,
+      "loss": 0.4482,
+      "step": 14832
+    },
+    {
+      "epoch": 2.4360027097489376,
+      "grad_norm": 0.4051812130197788,
+      "learning_rate": 4.752877219641628e-06,
+      "loss": 0.4585,
+      "step": 14833
+    },
+    {
+      "epoch": 2.4361669369572803,
+      "grad_norm": 0.3283393580462784,
+      "learning_rate": 4.752408987132517e-06,
+      "loss": 0.4471,
+      "step": 14834
+    },
+    {
+      "epoch": 2.4363311641656233,
+      "grad_norm": 0.29286177768403737,
+      "learning_rate": 4.751940751881998e-06,
+      "loss": 0.4231,
+      "step": 14835
+    },
+    {
+      "epoch": 2.436495391373966,
+      "grad_norm": 0.2906969999573708,
+      "learning_rate": 4.751472513895154e-06,
+      "loss": 0.4326,
+      "step": 14836
+    },
+    {
+      "epoch": 2.4366596185823086,
+      "grad_norm": 0.39150745671413895,
+      "learning_rate": 4.75100427317707e-06,
+      "loss": 0.4434,
+      "step": 14837
+    },
+    {
+      "epoch": 2.4368238457906513,
+      "grad_norm": 0.27745309394386025,
+      "learning_rate": 4.750536029732834e-06,
+      "loss": 0.4194,
+      "step": 14838
+    },
+    {
+      "epoch": 2.436988072998994,
+      "grad_norm": 0.7360407236143521,
+      "learning_rate": 4.750067783567528e-06,
+      "loss": 0.4238,
+      "step": 14839
+    },
+    {
+      "epoch": 2.437152300207337,
+      "grad_norm": 0.4518500230554605,
+      "learning_rate": 4.749599534686242e-06,
+      "loss": 0.4224,
+      "step": 14840
+    },
+    {
+      "epoch": 2.4373165274156796,
+      "grad_norm": 0.34881458565898915,
+      "learning_rate": 4.74913128309406e-06,
+      "loss": 0.4303,
+      "step": 14841
+    },
+    {
+      "epoch": 2.4374807546240222,
+      "grad_norm": 0.2615419271331008,
+      "learning_rate": 4.748663028796065e-06,
+      "loss": 0.4413,
+      "step": 14842
+    },
+    {
+      "epoch": 2.437644981832365,
+      "grad_norm": 0.31938658995749575,
+      "learning_rate": 4.748194771797346e-06,
+      "loss": 0.4383,
+      "step": 14843
+    },
+    {
+      "epoch": 2.437809209040708,
+      "grad_norm": 0.34806719512201467,
+      "learning_rate": 4.747726512102988e-06,
+      "loss": 0.4449,
+      "step": 14844
+    },
+    {
+      "epoch": 2.4379734362490506,
+      "grad_norm": 0.2952998266610244,
+      "learning_rate": 4.747258249718077e-06,
+      "loss": 0.4133,
+      "step": 14845
+    },
+    {
+      "epoch": 2.4381376634573932,
+      "grad_norm": 0.4080042462948629,
+      "learning_rate": 4.746789984647696e-06,
+      "loss": 0.4174,
+      "step": 14846
+    },
+    {
+      "epoch": 2.438301890665736,
+      "grad_norm": 0.2617233550855106,
+      "learning_rate": 4.7463217168969335e-06,
+      "loss": 0.4516,
+      "step": 14847
+    },
+    {
+      "epoch": 2.438466117874079,
+      "grad_norm": 0.3167837063772311,
+      "learning_rate": 4.7458534464708746e-06,
+      "loss": 0.4207,
+      "step": 14848
+    },
+    {
+      "epoch": 2.4386303450824216,
+      "grad_norm": 0.34189588475116073,
+      "learning_rate": 4.745385173374608e-06,
+      "loss": 0.4362,
+      "step": 14849
+    },
+    {
+      "epoch": 2.4387945722907642,
+      "grad_norm": 0.37358037512844017,
+      "learning_rate": 4.7449168976132145e-06,
+      "loss": 0.4224,
+      "step": 14850
+    },
+    {
+      "epoch": 2.438958799499107,
+      "grad_norm": 0.3431725832259083,
+      "learning_rate": 4.744448619191783e-06,
+      "loss": 0.4262,
+      "step": 14851
+    },
+    {
+      "epoch": 2.43912302670745,
+      "grad_norm": 0.314791986594665,
+      "learning_rate": 4.743980338115398e-06,
+      "loss": 0.4557,
+      "step": 14852
+    },
+    {
+      "epoch": 2.4392872539157926,
+      "grad_norm": 0.5013463893524989,
+      "learning_rate": 4.743512054389148e-06,
+      "loss": 0.4394,
+      "step": 14853
+    },
+    {
+      "epoch": 2.4394514811241352,
+      "grad_norm": 0.2725217761544163,
+      "learning_rate": 4.743043768018117e-06,
+      "loss": 0.4114,
+      "step": 14854
+    },
+    {
+      "epoch": 2.439615708332478,
+      "grad_norm": 0.37105193624079863,
+      "learning_rate": 4.742575479007393e-06,
+      "loss": 0.4172,
+      "step": 14855
+    },
+    {
+      "epoch": 2.4397799355408205,
+      "grad_norm": 0.4803472831557481,
+      "learning_rate": 4.742107187362058e-06,
+      "loss": 0.4385,
+      "step": 14856
+    },
+    {
+      "epoch": 2.4399441627491636,
+      "grad_norm": 0.2943878838182107,
+      "learning_rate": 4.741638893087203e-06,
+      "loss": 0.4354,
+      "step": 14857
+    },
+    {
+      "epoch": 2.4401083899575062,
+      "grad_norm": 0.2980819471686045,
+      "learning_rate": 4.74117059618791e-06,
+      "loss": 0.4249,
+      "step": 14858
+    },
+    {
+      "epoch": 2.440272617165849,
+      "grad_norm": 0.7392078808610509,
+      "learning_rate": 4.740702296669269e-06,
+      "loss": 0.4452,
+      "step": 14859
+    },
+    {
+      "epoch": 2.4404368443741915,
+      "grad_norm": 0.335397733043894,
+      "learning_rate": 4.740233994536363e-06,
+      "loss": 0.4384,
+      "step": 14860
+    },
+    {
+      "epoch": 2.4406010715825346,
+      "grad_norm": 0.3189866960758374,
+      "learning_rate": 4.73976568979428e-06,
+      "loss": 0.4306,
+      "step": 14861
+    },
+    {
+      "epoch": 2.4407652987908772,
+      "grad_norm": 0.2752736569035358,
+      "learning_rate": 4.739297382448105e-06,
+      "loss": 0.4227,
+      "step": 14862
+    },
+    {
+      "epoch": 2.44092952599922,
+      "grad_norm": 0.37664773038821614,
+      "learning_rate": 4.738829072502925e-06,
+      "loss": 0.4311,
+      "step": 14863
+    },
+    {
+      "epoch": 2.4410937532075625,
+      "grad_norm": 0.3309783756164121,
+      "learning_rate": 4.738360759963827e-06,
+      "loss": 0.4323,
+      "step": 14864
+    },
+    {
+      "epoch": 2.4412579804159056,
+      "grad_norm": 0.3763551459442455,
+      "learning_rate": 4.737892444835896e-06,
+      "loss": 0.4403,
+      "step": 14865
+    },
+    {
+      "epoch": 2.441422207624248,
+      "grad_norm": 0.35097168113421806,
+      "learning_rate": 4.737424127124219e-06,
+      "loss": 0.4442,
+      "step": 14866
+    },
+    {
+      "epoch": 2.441586434832591,
+      "grad_norm": 0.3712543226418984,
+      "learning_rate": 4.7369558068338825e-06,
+      "loss": 0.4372,
+      "step": 14867
+    },
+    {
+      "epoch": 2.4417506620409335,
+      "grad_norm": 0.32506729393422107,
+      "learning_rate": 4.736487483969972e-06,
+      "loss": 0.4395,
+      "step": 14868
+    },
+    {
+      "epoch": 2.4419148892492766,
+      "grad_norm": 0.3064457747338616,
+      "learning_rate": 4.7360191585375765e-06,
+      "loss": 0.4452,
+      "step": 14869
+    },
+    {
+      "epoch": 2.442079116457619,
+      "grad_norm": 0.44702514644854086,
+      "learning_rate": 4.735550830541781e-06,
+      "loss": 0.4159,
+      "step": 14870
+    },
+    {
+      "epoch": 2.442243343665962,
+      "grad_norm": 0.3616030848631678,
+      "learning_rate": 4.73508249998767e-06,
+      "loss": 0.4256,
+      "step": 14871
+    },
+    {
+      "epoch": 2.4424075708743045,
+      "grad_norm": 0.28232285830936266,
+      "learning_rate": 4.734614166880332e-06,
+      "loss": 0.4406,
+      "step": 14872
+    },
+    {
+      "epoch": 2.442571798082647,
+      "grad_norm": 0.2805074286554534,
+      "learning_rate": 4.734145831224853e-06,
+      "loss": 0.4344,
+      "step": 14873
+    },
+    {
+      "epoch": 2.44273602529099,
+      "grad_norm": 0.2956435605595372,
+      "learning_rate": 4.733677493026323e-06,
+      "loss": 0.429,
+      "step": 14874
+    },
+    {
+      "epoch": 2.442900252499333,
+      "grad_norm": 0.2989704174696099,
+      "learning_rate": 4.7332091522898225e-06,
+      "loss": 0.4158,
+      "step": 14875
+    },
+    {
+      "epoch": 2.4430644797076755,
+      "grad_norm": 0.26915929922268195,
+      "learning_rate": 4.73274080902044e-06,
+      "loss": 0.4651,
+      "step": 14876
+    },
+    {
+      "epoch": 2.443228706916018,
+      "grad_norm": 0.3133310818964482,
+      "learning_rate": 4.732272463223265e-06,
+      "loss": 0.4208,
+      "step": 14877
+    },
+    {
+      "epoch": 2.443392934124361,
+      "grad_norm": 0.2808808633684279,
+      "learning_rate": 4.731804114903384e-06,
+      "loss": 0.4278,
+      "step": 14878
+    },
+    {
+      "epoch": 2.443557161332704,
+      "grad_norm": 0.9563998333839273,
+      "learning_rate": 4.731335764065881e-06,
+      "loss": 0.4451,
+      "step": 14879
+    },
+    {
+      "epoch": 2.4437213885410465,
+      "grad_norm": 0.30893292698775016,
+      "learning_rate": 4.730867410715844e-06,
+      "loss": 0.4448,
+      "step": 14880
+    },
+    {
+      "epoch": 2.443885615749389,
+      "grad_norm": 0.2708141085292231,
+      "learning_rate": 4.7303990548583584e-06,
+      "loss": 0.4341,
+      "step": 14881
+    },
+    {
+      "epoch": 2.444049842957732,
+      "grad_norm": 0.43372099717902635,
+      "learning_rate": 4.729930696498515e-06,
+      "loss": 0.441,
+      "step": 14882
+    },
+    {
+      "epoch": 2.444214070166075,
+      "grad_norm": 0.2895325076784331,
+      "learning_rate": 4.729462335641396e-06,
+      "loss": 0.426,
+      "step": 14883
+    },
+    {
+      "epoch": 2.4443782973744175,
+      "grad_norm": 0.33145935474776,
+      "learning_rate": 4.728993972292091e-06,
+      "loss": 0.437,
+      "step": 14884
+    },
+    {
+      "epoch": 2.44454252458276,
+      "grad_norm": 0.3935456848771356,
+      "learning_rate": 4.728525606455686e-06,
+      "loss": 0.4555,
+      "step": 14885
+    },
+    {
+      "epoch": 2.444706751791103,
+      "grad_norm": 0.2952308962107992,
+      "learning_rate": 4.728057238137269e-06,
+      "loss": 0.438,
+      "step": 14886
+    },
+    {
+      "epoch": 2.444870978999446,
+      "grad_norm": 0.3062234555252227,
+      "learning_rate": 4.727588867341925e-06,
+      "loss": 0.4143,
+      "step": 14887
+    },
+    {
+      "epoch": 2.4450352062077885,
+      "grad_norm": 0.32180242990583197,
+      "learning_rate": 4.727120494074741e-06,
+      "loss": 0.4416,
+      "step": 14888
+    },
+    {
+      "epoch": 2.445199433416131,
+      "grad_norm": 0.32843161134064053,
+      "learning_rate": 4.726652118340808e-06,
+      "loss": 0.418,
+      "step": 14889
+    },
+    {
+      "epoch": 2.4453636606244737,
+      "grad_norm": 0.3390374896278502,
+      "learning_rate": 4.726183740145208e-06,
+      "loss": 0.4518,
+      "step": 14890
+    },
+    {
+      "epoch": 2.445527887832817,
+      "grad_norm": 0.28347479023720434,
+      "learning_rate": 4.725715359493031e-06,
+      "loss": 0.4379,
+      "step": 14891
+    },
+    {
+      "epoch": 2.4456921150411595,
+      "grad_norm": 0.30168778550602693,
+      "learning_rate": 4.725246976389361e-06,
+      "loss": 0.4447,
+      "step": 14892
+    },
+    {
+      "epoch": 2.445856342249502,
+      "grad_norm": 0.3182228860270116,
+      "learning_rate": 4.724778590839289e-06,
+      "loss": 0.4436,
+      "step": 14893
+    },
+    {
+      "epoch": 2.4460205694578447,
+      "grad_norm": 0.26929248249368787,
+      "learning_rate": 4.7243102028479e-06,
+      "loss": 0.4283,
+      "step": 14894
+    },
+    {
+      "epoch": 2.446184796666188,
+      "grad_norm": 0.3502740325919489,
+      "learning_rate": 4.723841812420282e-06,
+      "loss": 0.441,
+      "step": 14895
+    },
+    {
+      "epoch": 2.4463490238745305,
+      "grad_norm": 0.4357248159824678,
+      "learning_rate": 4.723373419561521e-06,
+      "loss": 0.449,
+      "step": 14896
+    },
+    {
+      "epoch": 2.446513251082873,
+      "grad_norm": 0.2956091582918993,
+      "learning_rate": 4.722905024276704e-06,
+      "loss": 0.4464,
+      "step": 14897
+    },
+    {
+      "epoch": 2.4466774782912157,
+      "grad_norm": 0.9827935727181517,
+      "learning_rate": 4.72243662657092e-06,
+      "loss": 0.4575,
+      "step": 14898
+    },
+    {
+      "epoch": 2.446841705499559,
+      "grad_norm": 0.8911376698958159,
+      "learning_rate": 4.7219682264492565e-06,
+      "loss": 0.4177,
+      "step": 14899
+    },
+    {
+      "epoch": 2.4470059327079015,
+      "grad_norm": 0.4204288106361778,
+      "learning_rate": 4.721499823916798e-06,
+      "loss": 0.4407,
+      "step": 14900
+    },
+    {
+      "epoch": 2.447170159916244,
+      "grad_norm": 0.4214831462365801,
+      "learning_rate": 4.7210314189786335e-06,
+      "loss": 0.4385,
+      "step": 14901
+    },
+    {
+      "epoch": 2.4473343871245867,
+      "grad_norm": 0.2856166432697078,
+      "learning_rate": 4.72056301163985e-06,
+      "loss": 0.4083,
+      "step": 14902
+    },
+    {
+      "epoch": 2.44749861433293,
+      "grad_norm": 0.39927665904922716,
+      "learning_rate": 4.720094601905538e-06,
+      "loss": 0.4393,
+      "step": 14903
+    },
+    {
+      "epoch": 2.4476628415412724,
+      "grad_norm": 0.27247089834361105,
+      "learning_rate": 4.719626189780781e-06,
+      "loss": 0.4343,
+      "step": 14904
+    },
+    {
+      "epoch": 2.447827068749615,
+      "grad_norm": 0.3379655021876495,
+      "learning_rate": 4.719157775270666e-06,
+      "loss": 0.4368,
+      "step": 14905
+    },
+    {
+      "epoch": 2.4479912959579577,
+      "grad_norm": 0.5812707202881656,
+      "learning_rate": 4.718689358380282e-06,
+      "loss": 0.427,
+      "step": 14906
+    },
+    {
+      "epoch": 2.4481555231663004,
+      "grad_norm": 0.288175714574787,
+      "learning_rate": 4.718220939114718e-06,
+      "loss": 0.4393,
+      "step": 14907
+    },
+    {
+      "epoch": 2.4483197503746434,
+      "grad_norm": 0.31558613912668265,
+      "learning_rate": 4.717752517479059e-06,
+      "loss": 0.4263,
+      "step": 14908
+    },
+    {
+      "epoch": 2.448483977582986,
+      "grad_norm": 0.3122999644429297,
+      "learning_rate": 4.7172840934783935e-06,
+      "loss": 0.4497,
+      "step": 14909
+    },
+    {
+      "epoch": 2.4486482047913287,
+      "grad_norm": 0.3856354602008668,
+      "learning_rate": 4.716815667117808e-06,
+      "loss": 0.4209,
+      "step": 14910
+    },
+    {
+      "epoch": 2.4488124319996714,
+      "grad_norm": 0.3183549391768059,
+      "learning_rate": 4.716347238402393e-06,
+      "loss": 0.4361,
+      "step": 14911
+    },
+    {
+      "epoch": 2.4489766592080144,
+      "grad_norm": 0.27665517262562817,
+      "learning_rate": 4.715878807337233e-06,
+      "loss": 0.407,
+      "step": 14912
+    },
+    {
+      "epoch": 2.449140886416357,
+      "grad_norm": 0.5377088495858262,
+      "learning_rate": 4.715410373927417e-06,
+      "loss": 0.4374,
+      "step": 14913
+    },
+    {
+      "epoch": 2.4493051136246997,
+      "grad_norm": 0.30183317730928727,
+      "learning_rate": 4.714941938178032e-06,
+      "loss": 0.4332,
+      "step": 14914
+    },
+    {
+      "epoch": 2.4494693408330424,
+      "grad_norm": 0.3264467535903471,
+      "learning_rate": 4.714473500094166e-06,
+      "loss": 0.4374,
+      "step": 14915
+    },
+    {
+      "epoch": 2.4496335680413854,
+      "grad_norm": 0.33462011748221,
+      "learning_rate": 4.714005059680908e-06,
+      "loss": 0.4584,
+      "step": 14916
+    },
+    {
+      "epoch": 2.449797795249728,
+      "grad_norm": 0.3117046569026179,
+      "learning_rate": 4.713536616943342e-06,
+      "loss": 0.446,
+      "step": 14917
+    },
+    {
+      "epoch": 2.4499620224580707,
+      "grad_norm": 0.3103767396060371,
+      "learning_rate": 4.71306817188656e-06,
+      "loss": 0.4251,
+      "step": 14918
+    },
+    {
+      "epoch": 2.4501262496664133,
+      "grad_norm": 0.37468063922930944,
+      "learning_rate": 4.712599724515649e-06,
+      "loss": 0.4316,
+      "step": 14919
+    },
+    {
+      "epoch": 2.4502904768747564,
+      "grad_norm": 0.2992503309515945,
+      "learning_rate": 4.712131274835694e-06,
+      "loss": 0.4332,
+      "step": 14920
+    },
+    {
+      "epoch": 2.450454704083099,
+      "grad_norm": 0.2914771504239795,
+      "learning_rate": 4.711662822851785e-06,
+      "loss": 0.4394,
+      "step": 14921
+    },
+    {
+      "epoch": 2.4506189312914417,
+      "grad_norm": 0.4249734365242678,
+      "learning_rate": 4.711194368569009e-06,
+      "loss": 0.448,
+      "step": 14922
+    },
+    {
+      "epoch": 2.4507831584997843,
+      "grad_norm": 0.2915814818062091,
+      "learning_rate": 4.710725911992456e-06,
+      "loss": 0.4271,
+      "step": 14923
+    },
+    {
+      "epoch": 2.450947385708127,
+      "grad_norm": 0.5198274597992836,
+      "learning_rate": 4.710257453127212e-06,
+      "loss": 0.4353,
+      "step": 14924
+    },
+    {
+      "epoch": 2.45111161291647,
+      "grad_norm": 0.38566805754977934,
+      "learning_rate": 4.709788991978364e-06,
+      "loss": 0.4318,
+      "step": 14925
+    },
+    {
+      "epoch": 2.4512758401248127,
+      "grad_norm": 0.33962611959497585,
+      "learning_rate": 4.7093205285510024e-06,
+      "loss": 0.4317,
+      "step": 14926
+    },
+    {
+      "epoch": 2.4514400673331553,
+      "grad_norm": 0.3431469761470781,
+      "learning_rate": 4.7088520628502134e-06,
+      "loss": 0.4209,
+      "step": 14927
+    },
+    {
+      "epoch": 2.451604294541498,
+      "grad_norm": 0.2557889173861677,
+      "learning_rate": 4.708383594881086e-06,
+      "loss": 0.444,
+      "step": 14928
+    },
+    {
+      "epoch": 2.451768521749841,
+      "grad_norm": 0.3034543392005885,
+      "learning_rate": 4.707915124648707e-06,
+      "loss": 0.4489,
+      "step": 14929
+    },
+    {
+      "epoch": 2.4519327489581837,
+      "grad_norm": 0.3355927507316989,
+      "learning_rate": 4.707446652158164e-06,
+      "loss": 0.4363,
+      "step": 14930
+    },
+    {
+      "epoch": 2.4520969761665263,
+      "grad_norm": 0.43993410862279914,
+      "learning_rate": 4.706978177414548e-06,
+      "loss": 0.4439,
+      "step": 14931
+    },
+    {
+      "epoch": 2.452261203374869,
+      "grad_norm": 0.309306723684051,
+      "learning_rate": 4.7065097004229445e-06,
+      "loss": 0.459,
+      "step": 14932
+    },
+    {
+      "epoch": 2.452425430583212,
+      "grad_norm": 0.36210017974478553,
+      "learning_rate": 4.706041221188444e-06,
+      "loss": 0.458,
+      "step": 14933
+    },
+    {
+      "epoch": 2.4525896577915547,
+      "grad_norm": 0.44914379194192844,
+      "learning_rate": 4.705572739716132e-06,
+      "loss": 0.4297,
+      "step": 14934
+    },
+    {
+      "epoch": 2.4527538849998973,
+      "grad_norm": 0.4199867608706427,
+      "learning_rate": 4.705104256011097e-06,
+      "loss": 0.4251,
+      "step": 14935
+    },
+    {
+      "epoch": 2.45291811220824,
+      "grad_norm": 0.2960066711495447,
+      "learning_rate": 4.70463577007843e-06,
+      "loss": 0.4098,
+      "step": 14936
+    },
+    {
+      "epoch": 2.453082339416583,
+      "grad_norm": 0.29052926852285893,
+      "learning_rate": 4.704167281923215e-06,
+      "loss": 0.4391,
+      "step": 14937
+    },
+    {
+      "epoch": 2.4532465666249257,
+      "grad_norm": 0.3345783118974667,
+      "learning_rate": 4.703698791550544e-06,
+      "loss": 0.4445,
+      "step": 14938
+    },
+    {
+      "epoch": 2.4534107938332683,
+      "grad_norm": 0.31411562694600764,
+      "learning_rate": 4.703230298965503e-06,
+      "loss": 0.4374,
+      "step": 14939
+    },
+    {
+      "epoch": 2.453575021041611,
+      "grad_norm": 0.39657920815497005,
+      "learning_rate": 4.702761804173181e-06,
+      "loss": 0.4297,
+      "step": 14940
+    },
+    {
+      "epoch": 2.4537392482499536,
+      "grad_norm": 0.26347298494219035,
+      "learning_rate": 4.7022933071786674e-06,
+      "loss": 0.4503,
+      "step": 14941
+    },
+    {
+      "epoch": 2.4539034754582967,
+      "grad_norm": 0.4001769996258484,
+      "learning_rate": 4.701824807987049e-06,
+      "loss": 0.4404,
+      "step": 14942
+    },
+    {
+      "epoch": 2.4540677026666393,
+      "grad_norm": 0.3490323283984835,
+      "learning_rate": 4.701356306603414e-06,
+      "loss": 0.4108,
+      "step": 14943
+    },
+    {
+      "epoch": 2.454231929874982,
+      "grad_norm": 0.32937590189847393,
+      "learning_rate": 4.700887803032851e-06,
+      "loss": 0.4331,
+      "step": 14944
+    },
+    {
+      "epoch": 2.4543961570833246,
+      "grad_norm": 0.3413660682534813,
+      "learning_rate": 4.70041929728045e-06,
+      "loss": 0.4379,
+      "step": 14945
+    },
+    {
+      "epoch": 2.4545603842916677,
+      "grad_norm": 0.5002206756837491,
+      "learning_rate": 4.699950789351297e-06,
+      "loss": 0.4483,
+      "step": 14946
+    },
+    {
+      "epoch": 2.4547246115000103,
+      "grad_norm": 0.3417412630888464,
+      "learning_rate": 4.699482279250482e-06,
+      "loss": 0.4233,
+      "step": 14947
+    },
+    {
+      "epoch": 2.454888838708353,
+      "grad_norm": 0.2913854518469454,
+      "learning_rate": 4.699013766983093e-06,
+      "loss": 0.4401,
+      "step": 14948
+    },
+    {
+      "epoch": 2.4550530659166956,
+      "grad_norm": 0.28411465528321456,
+      "learning_rate": 4.698545252554221e-06,
+      "loss": 0.4299,
+      "step": 14949
+    },
+    {
+      "epoch": 2.4552172931250387,
+      "grad_norm": 0.33716547921552187,
+      "learning_rate": 4.69807673596895e-06,
+      "loss": 0.4505,
+      "step": 14950
+    },
+    {
+      "epoch": 2.4553815203333813,
+      "grad_norm": 0.2993655726406468,
+      "learning_rate": 4.69760821723237e-06,
+      "loss": 0.4476,
+      "step": 14951
+    },
+    {
+      "epoch": 2.455545747541724,
+      "grad_norm": 0.30247931387572397,
+      "learning_rate": 4.6971396963495725e-06,
+      "loss": 0.4362,
+      "step": 14952
+    },
+    {
+      "epoch": 2.4557099747500666,
+      "grad_norm": 0.2859025796264969,
+      "learning_rate": 4.696671173325643e-06,
+      "loss": 0.4473,
+      "step": 14953
+    },
+    {
+      "epoch": 2.4558742019584097,
+      "grad_norm": 0.31087008805358723,
+      "learning_rate": 4.6962026481656715e-06,
+      "loss": 0.4323,
+      "step": 14954
+    },
+    {
+      "epoch": 2.4560384291667523,
+      "grad_norm": 0.31108806233509195,
+      "learning_rate": 4.695734120874745e-06,
+      "loss": 0.4363,
+      "step": 14955
+    },
+    {
+      "epoch": 2.456202656375095,
+      "grad_norm": 0.3713261946065773,
+      "learning_rate": 4.6952655914579526e-06,
+      "loss": 0.4582,
+      "step": 14956
+    },
+    {
+      "epoch": 2.4563668835834376,
+      "grad_norm": 0.3066338892215127,
+      "learning_rate": 4.694797059920388e-06,
+      "loss": 0.4296,
+      "step": 14957
+    },
+    {
+      "epoch": 2.45653111079178,
+      "grad_norm": 0.2810188832318168,
+      "learning_rate": 4.694328526267133e-06,
+      "loss": 0.4197,
+      "step": 14958
+    },
+    {
+      "epoch": 2.4566953380001233,
+      "grad_norm": 0.29244327846057233,
+      "learning_rate": 4.693859990503277e-06,
+      "loss": 0.4436,
+      "step": 14959
+    },
+    {
+      "epoch": 2.456859565208466,
+      "grad_norm": 0.3156738030477161,
+      "learning_rate": 4.693391452633913e-06,
+      "loss": 0.4539,
+      "step": 14960
+    },
+    {
+      "epoch": 2.4570237924168086,
+      "grad_norm": 0.3456900198117342,
+      "learning_rate": 4.692922912664128e-06,
+      "loss": 0.4107,
+      "step": 14961
+    },
+    {
+      "epoch": 2.457188019625151,
+      "grad_norm": 0.3750918202689833,
+      "learning_rate": 4.69245437059901e-06,
+      "loss": 0.4304,
+      "step": 14962
+    },
+    {
+      "epoch": 2.4573522468334943,
+      "grad_norm": 0.42389798412654855,
+      "learning_rate": 4.691985826443647e-06,
+      "loss": 0.4171,
+      "step": 14963
+    },
+    {
+      "epoch": 2.457516474041837,
+      "grad_norm": 0.28392710882294936,
+      "learning_rate": 4.69151728020313e-06,
+      "loss": 0.4347,
+      "step": 14964
+    },
+    {
+      "epoch": 2.4576807012501796,
+      "grad_norm": 0.28639762426572174,
+      "learning_rate": 4.691048731882546e-06,
+      "loss": 0.4443,
+      "step": 14965
+    },
+    {
+      "epoch": 2.457844928458522,
+      "grad_norm": 0.25684977537668713,
+      "learning_rate": 4.690580181486986e-06,
+      "loss": 0.4359,
+      "step": 14966
+    },
+    {
+      "epoch": 2.4580091556668653,
+      "grad_norm": 0.36644612008467237,
+      "learning_rate": 4.690111629021538e-06,
+      "loss": 0.4653,
+      "step": 14967
+    },
+    {
+      "epoch": 2.458173382875208,
+      "grad_norm": 0.3646384093642947,
+      "learning_rate": 4.68964307449129e-06,
+      "loss": 0.4397,
+      "step": 14968
+    },
+    {
+      "epoch": 2.4583376100835506,
+      "grad_norm": 0.3382367193847496,
+      "learning_rate": 4.689174517901331e-06,
+      "loss": 0.4406,
+      "step": 14969
+    },
+    {
+      "epoch": 2.458501837291893,
+      "grad_norm": 0.3140018837661301,
+      "learning_rate": 4.688705959256752e-06,
+      "loss": 0.4469,
+      "step": 14970
+    },
+    {
+      "epoch": 2.4586660645002363,
+      "grad_norm": 0.35977465400020864,
+      "learning_rate": 4.688237398562639e-06,
+      "loss": 0.4258,
+      "step": 14971
+    },
+    {
+      "epoch": 2.458830291708579,
+      "grad_norm": 0.3943690831857817,
+      "learning_rate": 4.687768835824083e-06,
+      "loss": 0.4461,
+      "step": 14972
+    },
+    {
+      "epoch": 2.4589945189169216,
+      "grad_norm": 0.5226466463828919,
+      "learning_rate": 4.687300271046173e-06,
+      "loss": 0.4528,
+      "step": 14973
+    },
+    {
+      "epoch": 2.459158746125264,
+      "grad_norm": 0.2765055483214686,
+      "learning_rate": 4.686831704233999e-06,
+      "loss": 0.4521,
+      "step": 14974
+    },
+    {
+      "epoch": 2.459322973333607,
+      "grad_norm": 0.29749033942911196,
+      "learning_rate": 4.6863631353926476e-06,
+      "loss": 0.4468,
+      "step": 14975
+    },
+    {
+      "epoch": 2.45948720054195,
+      "grad_norm": 0.2902094231503712,
+      "learning_rate": 4.68589456452721e-06,
+      "loss": 0.4278,
+      "step": 14976
+    },
+    {
+      "epoch": 2.4596514277502926,
+      "grad_norm": 0.3078796587727116,
+      "learning_rate": 4.6854259916427735e-06,
+      "loss": 0.4159,
+      "step": 14977
+    },
+    {
+      "epoch": 2.459815654958635,
+      "grad_norm": 0.2985730494760598,
+      "learning_rate": 4.684957416744429e-06,
+      "loss": 0.4503,
+      "step": 14978
+    },
+    {
+      "epoch": 2.459979882166978,
+      "grad_norm": 0.3189836021672542,
+      "learning_rate": 4.684488839837265e-06,
+      "loss": 0.442,
+      "step": 14979
+    },
+    {
+      "epoch": 2.460144109375321,
+      "grad_norm": 0.4842946241587704,
+      "learning_rate": 4.684020260926369e-06,
+      "loss": 0.4377,
+      "step": 14980
+    },
+    {
+      "epoch": 2.4603083365836635,
+      "grad_norm": 0.373334795378634,
+      "learning_rate": 4.683551680016834e-06,
+      "loss": 0.4456,
+      "step": 14981
+    },
+    {
+      "epoch": 2.460472563792006,
+      "grad_norm": 0.44875805856112194,
+      "learning_rate": 4.683083097113748e-06,
+      "loss": 0.445,
+      "step": 14982
+    },
+    {
+      "epoch": 2.460636791000349,
+      "grad_norm": 0.37397571261106627,
+      "learning_rate": 4.682614512222197e-06,
+      "loss": 0.4335,
+      "step": 14983
+    },
+    {
+      "epoch": 2.460801018208692,
+      "grad_norm": 0.3465551968819545,
+      "learning_rate": 4.682145925347273e-06,
+      "loss": 0.4215,
+      "step": 14984
+    },
+    {
+      "epoch": 2.4609652454170345,
+      "grad_norm": 0.29430252055096734,
+      "learning_rate": 4.681677336494065e-06,
+      "loss": 0.4457,
+      "step": 14985
+    },
+    {
+      "epoch": 2.461129472625377,
+      "grad_norm": 0.32286831624943985,
+      "learning_rate": 4.6812087456676644e-06,
+      "loss": 0.4154,
+      "step": 14986
+    },
+    {
+      "epoch": 2.46129369983372,
+      "grad_norm": 0.3793808917437682,
+      "learning_rate": 4.680740152873157e-06,
+      "loss": 0.4303,
+      "step": 14987
+    },
+    {
+      "epoch": 2.461457927042063,
+      "grad_norm": 0.34371076774506265,
+      "learning_rate": 4.680271558115635e-06,
+      "loss": 0.4444,
+      "step": 14988
+    },
+    {
+      "epoch": 2.4616221542504055,
+      "grad_norm": 0.2800293307566269,
+      "learning_rate": 4.679802961400184e-06,
+      "loss": 0.4263,
+      "step": 14989
+    },
+    {
+      "epoch": 2.461786381458748,
+      "grad_norm": 0.4105776277011872,
+      "learning_rate": 4.679334362731898e-06,
+      "loss": 0.442,
+      "step": 14990
+    },
+    {
+      "epoch": 2.461950608667091,
+      "grad_norm": 0.29117182265697783,
+      "learning_rate": 4.678865762115864e-06,
+      "loss": 0.4155,
+      "step": 14991
+    },
+    {
+      "epoch": 2.4621148358754334,
+      "grad_norm": 0.30870836843997373,
+      "learning_rate": 4.678397159557172e-06,
+      "loss": 0.4156,
+      "step": 14992
+    },
+    {
+      "epoch": 2.4622790630837765,
+      "grad_norm": 0.2643732894140815,
+      "learning_rate": 4.67792855506091e-06,
+      "loss": 0.4354,
+      "step": 14993
+    },
+    {
+      "epoch": 2.462443290292119,
+      "grad_norm": 0.2758373427519078,
+      "learning_rate": 4.67745994863217e-06,
+      "loss": 0.4397,
+      "step": 14994
+    },
+    {
+      "epoch": 2.462607517500462,
+      "grad_norm": 0.2879361585632679,
+      "learning_rate": 4.676991340276039e-06,
+      "loss": 0.4245,
+      "step": 14995
+    },
+    {
+      "epoch": 2.4627717447088044,
+      "grad_norm": 0.3407813698263583,
+      "learning_rate": 4.676522729997611e-06,
+      "loss": 0.424,
+      "step": 14996
+    },
+    {
+      "epoch": 2.4629359719171475,
+      "grad_norm": 0.3317369355754718,
+      "learning_rate": 4.676054117801969e-06,
+      "loss": 0.445,
+      "step": 14997
+    },
+    {
+      "epoch": 2.46310019912549,
+      "grad_norm": 0.6803115759149154,
+      "learning_rate": 4.675585503694208e-06,
+      "loss": 0.4384,
+      "step": 14998
+    },
+    {
+      "epoch": 2.463264426333833,
+      "grad_norm": 0.3831499501518102,
+      "learning_rate": 4.6751168876794164e-06,
+      "loss": 0.447,
+      "step": 14999
+    },
+    {
+      "epoch": 2.4634286535421754,
+      "grad_norm": 0.8359767289627413,
+      "learning_rate": 4.674648269762681e-06,
+      "loss": 0.4318,
+      "step": 15000
+    },
+    {
+      "epoch": 2.4635928807505185,
+      "grad_norm": 0.2966619920680067,
+      "learning_rate": 4.674179649949095e-06,
+      "loss": 0.4424,
+      "step": 15001
+    },
+    {
+      "epoch": 2.463757107958861,
+      "grad_norm": 0.2732458183573598,
+      "learning_rate": 4.673711028243746e-06,
+      "loss": 0.4509,
+      "step": 15002
+    },
+    {
+      "epoch": 2.463921335167204,
+      "grad_norm": 0.29770838372139535,
+      "learning_rate": 4.673242404651725e-06,
+      "loss": 0.4195,
+      "step": 15003
+    },
+    {
+      "epoch": 2.4640855623755464,
+      "grad_norm": 0.2502454178475141,
+      "learning_rate": 4.67277377917812e-06,
+      "loss": 0.4325,
+      "step": 15004
+    },
+    {
+      "epoch": 2.4642497895838895,
+      "grad_norm": 0.27202883419282115,
+      "learning_rate": 4.672305151828022e-06,
+      "loss": 0.4389,
+      "step": 15005
+    },
+    {
+      "epoch": 2.464414016792232,
+      "grad_norm": 0.316940999995647,
+      "learning_rate": 4.671836522606521e-06,
+      "loss": 0.4244,
+      "step": 15006
+    },
+    {
+      "epoch": 2.464578244000575,
+      "grad_norm": 0.3077409686336453,
+      "learning_rate": 4.671367891518707e-06,
+      "loss": 0.4431,
+      "step": 15007
+    },
+    {
+      "epoch": 2.4647424712089174,
+      "grad_norm": 0.2991331964822248,
+      "learning_rate": 4.670899258569668e-06,
+      "loss": 0.4186,
+      "step": 15008
+    },
+    {
+      "epoch": 2.46490669841726,
+      "grad_norm": 0.30160160597116736,
+      "learning_rate": 4.670430623764495e-06,
+      "loss": 0.4209,
+      "step": 15009
+    },
+    {
+      "epoch": 2.465070925625603,
+      "grad_norm": 1.2057903308954052,
+      "learning_rate": 4.6699619871082775e-06,
+      "loss": 0.4507,
+      "step": 15010
+    },
+    {
+      "epoch": 2.465235152833946,
+      "grad_norm": 0.36722155486617347,
+      "learning_rate": 4.669493348606105e-06,
+      "loss": 0.4246,
+      "step": 15011
+    },
+    {
+      "epoch": 2.4653993800422884,
+      "grad_norm": 0.2732815069756437,
+      "learning_rate": 4.669024708263071e-06,
+      "loss": 0.4399,
+      "step": 15012
+    },
+    {
+      "epoch": 2.465563607250631,
+      "grad_norm": 0.3759797860911508,
+      "learning_rate": 4.668556066084259e-06,
+      "loss": 0.4429,
+      "step": 15013
+    },
+    {
+      "epoch": 2.465727834458974,
+      "grad_norm": 0.29556366049368965,
+      "learning_rate": 4.668087422074763e-06,
+      "loss": 0.4399,
+      "step": 15014
+    },
+    {
+      "epoch": 2.465892061667317,
+      "grad_norm": 0.33310926302971894,
+      "learning_rate": 4.667618776239674e-06,
+      "loss": 0.4282,
+      "step": 15015
+    },
+    {
+      "epoch": 2.4660562888756594,
+      "grad_norm": 0.28561588310789915,
+      "learning_rate": 4.6671501285840795e-06,
+      "loss": 0.4207,
+      "step": 15016
+    },
+    {
+      "epoch": 2.466220516084002,
+      "grad_norm": 0.31983473101658594,
+      "learning_rate": 4.666681479113069e-06,
+      "loss": 0.4324,
+      "step": 15017
+    },
+    {
+      "epoch": 2.466384743292345,
+      "grad_norm": 0.3606193430776502,
+      "learning_rate": 4.666212827831733e-06,
+      "loss": 0.4311,
+      "step": 15018
+    },
+    {
+      "epoch": 2.4665489705006878,
+      "grad_norm": 0.3478030075169639,
+      "learning_rate": 4.665744174745164e-06,
+      "loss": 0.4234,
+      "step": 15019
+    },
+    {
+      "epoch": 2.4667131977090304,
+      "grad_norm": 0.2944069325615738,
+      "learning_rate": 4.66527551985845e-06,
+      "loss": 0.4405,
+      "step": 15020
+    },
+    {
+      "epoch": 2.466877424917373,
+      "grad_norm": 0.4024844019253324,
+      "learning_rate": 4.6648068631766816e-06,
+      "loss": 0.4406,
+      "step": 15021
+    },
+    {
+      "epoch": 2.467041652125716,
+      "grad_norm": 0.2907487570514868,
+      "learning_rate": 4.664338204704947e-06,
+      "loss": 0.4362,
+      "step": 15022
+    },
+    {
+      "epoch": 2.4672058793340588,
+      "grad_norm": 0.3599279459135872,
+      "learning_rate": 4.663869544448338e-06,
+      "loss": 0.4249,
+      "step": 15023
+    },
+    {
+      "epoch": 2.4673701065424014,
+      "grad_norm": 0.4021341046173282,
+      "learning_rate": 4.663400882411946e-06,
+      "loss": 0.4577,
+      "step": 15024
+    },
+    {
+      "epoch": 2.467534333750744,
+      "grad_norm": 0.29310096697738297,
+      "learning_rate": 4.662932218600859e-06,
+      "loss": 0.4242,
+      "step": 15025
+    },
+    {
+      "epoch": 2.4676985609590867,
+      "grad_norm": 0.3008874854202207,
+      "learning_rate": 4.662463553020167e-06,
+      "loss": 0.4483,
+      "step": 15026
+    },
+    {
+      "epoch": 2.4678627881674298,
+      "grad_norm": 0.31891265100092614,
+      "learning_rate": 4.661994885674962e-06,
+      "loss": 0.4422,
+      "step": 15027
+    },
+    {
+      "epoch": 2.4680270153757724,
+      "grad_norm": 0.4460526086221128,
+      "learning_rate": 4.661526216570332e-06,
+      "loss": 0.4276,
+      "step": 15028
+    },
+    {
+      "epoch": 2.468191242584115,
+      "grad_norm": 0.3133604060641469,
+      "learning_rate": 4.661057545711369e-06,
+      "loss": 0.4528,
+      "step": 15029
+    },
+    {
+      "epoch": 2.4683554697924577,
+      "grad_norm": 0.28146024737318615,
+      "learning_rate": 4.660588873103164e-06,
+      "loss": 0.4345,
+      "step": 15030
+    },
+    {
+      "epoch": 2.4685196970008008,
+      "grad_norm": 0.528866055449549,
+      "learning_rate": 4.6601201987508035e-06,
+      "loss": 0.4364,
+      "step": 15031
+    },
+    {
+      "epoch": 2.4686839242091434,
+      "grad_norm": 0.2872816059039179,
+      "learning_rate": 4.659651522659382e-06,
+      "loss": 0.4372,
+      "step": 15032
+    },
+    {
+      "epoch": 2.468848151417486,
+      "grad_norm": 0.34378625498393467,
+      "learning_rate": 4.659182844833987e-06,
+      "loss": 0.4253,
+      "step": 15033
+    },
+    {
+      "epoch": 2.4690123786258287,
+      "grad_norm": 0.3196887697710157,
+      "learning_rate": 4.6587141652797085e-06,
+      "loss": 0.44,
+      "step": 15034
+    },
+    {
+      "epoch": 2.4691766058341718,
+      "grad_norm": 0.2937573874985924,
+      "learning_rate": 4.658245484001641e-06,
+      "loss": 0.4304,
+      "step": 15035
+    },
+    {
+      "epoch": 2.4693408330425144,
+      "grad_norm": 0.30747920832229947,
+      "learning_rate": 4.6577768010048685e-06,
+      "loss": 0.434,
+      "step": 15036
+    },
+    {
+      "epoch": 2.469505060250857,
+      "grad_norm": 0.3798041029245597,
+      "learning_rate": 4.657308116294488e-06,
+      "loss": 0.4494,
+      "step": 15037
+    },
+    {
+      "epoch": 2.4696692874591997,
+      "grad_norm": 0.30374441085047227,
+      "learning_rate": 4.656839429875584e-06,
+      "loss": 0.4268,
+      "step": 15038
+    },
+    {
+      "epoch": 2.4698335146675428,
+      "grad_norm": 0.3892278187656976,
+      "learning_rate": 4.656370741753251e-06,
+      "loss": 0.4299,
+      "step": 15039
+    },
+    {
+      "epoch": 2.4699977418758854,
+      "grad_norm": 0.2855739563377692,
+      "learning_rate": 4.655902051932576e-06,
+      "loss": 0.433,
+      "step": 15040
+    },
+    {
+      "epoch": 2.470161969084228,
+      "grad_norm": 0.3673913461509335,
+      "learning_rate": 4.655433360418654e-06,
+      "loss": 0.4347,
+      "step": 15041
+    },
+    {
+      "epoch": 2.4703261962925707,
+      "grad_norm": 0.6614279208112778,
+      "learning_rate": 4.65496466721657e-06,
+      "loss": 0.4454,
+      "step": 15042
+    },
+    {
+      "epoch": 2.4704904235009133,
+      "grad_norm": 0.27867910155523756,
+      "learning_rate": 4.654495972331418e-06,
+      "loss": 0.4296,
+      "step": 15043
+    },
+    {
+      "epoch": 2.4706546507092564,
+      "grad_norm": 0.41239281456362475,
+      "learning_rate": 4.65402727576829e-06,
+      "loss": 0.436,
+      "step": 15044
+    },
+    {
+      "epoch": 2.470818877917599,
+      "grad_norm": 0.3045078974339018,
+      "learning_rate": 4.653558577532274e-06,
+      "loss": 0.431,
+      "step": 15045
+    },
+    {
+      "epoch": 2.4709831051259417,
+      "grad_norm": 0.3162849530056356,
+      "learning_rate": 4.653089877628458e-06,
+      "loss": 0.4205,
+      "step": 15046
+    },
+    {
+      "epoch": 2.4711473323342843,
+      "grad_norm": 0.29902768192013507,
+      "learning_rate": 4.652621176061936e-06,
+      "loss": 0.4433,
+      "step": 15047
+    },
+    {
+      "epoch": 2.4713115595426274,
+      "grad_norm": 0.31846117954541137,
+      "learning_rate": 4.652152472837798e-06,
+      "loss": 0.4314,
+      "step": 15048
+    },
+    {
+      "epoch": 2.47147578675097,
+      "grad_norm": 0.2710159655321698,
+      "learning_rate": 4.651683767961136e-06,
+      "loss": 0.4532,
+      "step": 15049
+    },
+    {
+      "epoch": 2.4716400139593127,
+      "grad_norm": 0.3420916679093024,
+      "learning_rate": 4.651215061437038e-06,
+      "loss": 0.4279,
+      "step": 15050
+    },
+    {
+      "epoch": 2.4718042411676553,
+      "grad_norm": 0.27750671662652043,
+      "learning_rate": 4.650746353270595e-06,
+      "loss": 0.4153,
+      "step": 15051
+    },
+    {
+      "epoch": 2.4719684683759984,
+      "grad_norm": 0.3031546521062631,
+      "learning_rate": 4.650277643466899e-06,
+      "loss": 0.42,
+      "step": 15052
+    },
+    {
+      "epoch": 2.472132695584341,
+      "grad_norm": 0.3337669951335687,
+      "learning_rate": 4.64980893203104e-06,
+      "loss": 0.4325,
+      "step": 15053
+    },
+    {
+      "epoch": 2.4722969227926836,
+      "grad_norm": 0.2946419765296823,
+      "learning_rate": 4.64934021896811e-06,
+      "loss": 0.4389,
+      "step": 15054
+    },
+    {
+      "epoch": 2.4724611500010263,
+      "grad_norm": 0.6566714479694089,
+      "learning_rate": 4.648871504283196e-06,
+      "loss": 0.4273,
+      "step": 15055
+    },
+    {
+      "epoch": 2.4726253772093694,
+      "grad_norm": 0.4074657555880319,
+      "learning_rate": 4.6484027879813905e-06,
+      "loss": 0.4383,
+      "step": 15056
+    },
+    {
+      "epoch": 2.472789604417712,
+      "grad_norm": 0.27201573725819184,
+      "learning_rate": 4.6479340700677865e-06,
+      "loss": 0.4318,
+      "step": 15057
+    },
+    {
+      "epoch": 2.4729538316260546,
+      "grad_norm": 0.26289819417609633,
+      "learning_rate": 4.647465350547473e-06,
+      "loss": 0.4327,
+      "step": 15058
+    },
+    {
+      "epoch": 2.4731180588343973,
+      "grad_norm": 0.3174931715785728,
+      "learning_rate": 4.64699662942554e-06,
+      "loss": 0.4239,
+      "step": 15059
+    },
+    {
+      "epoch": 2.47328228604274,
+      "grad_norm": 0.2919191884008475,
+      "learning_rate": 4.64652790670708e-06,
+      "loss": 0.457,
+      "step": 15060
+    },
+    {
+      "epoch": 2.473446513251083,
+      "grad_norm": 0.31143441804325606,
+      "learning_rate": 4.646059182397181e-06,
+      "loss": 0.4143,
+      "step": 15061
+    },
+    {
+      "epoch": 2.4736107404594256,
+      "grad_norm": 0.3528192752492953,
+      "learning_rate": 4.645590456500937e-06,
+      "loss": 0.4373,
+      "step": 15062
+    },
+    {
+      "epoch": 2.4737749676677683,
+      "grad_norm": 0.334288162317713,
+      "learning_rate": 4.645121729023436e-06,
+      "loss": 0.4261,
+      "step": 15063
+    },
+    {
+      "epoch": 2.473939194876111,
+      "grad_norm": 0.49576018154807827,
+      "learning_rate": 4.644652999969772e-06,
+      "loss": 0.4409,
+      "step": 15064
+    },
+    {
+      "epoch": 2.474103422084454,
+      "grad_norm": 0.2739197448585547,
+      "learning_rate": 4.644184269345033e-06,
+      "loss": 0.4462,
+      "step": 15065
+    },
+    {
+      "epoch": 2.4742676492927966,
+      "grad_norm": 0.3100637983651266,
+      "learning_rate": 4.643715537154312e-06,
+      "loss": 0.4469,
+      "step": 15066
+    },
+    {
+      "epoch": 2.4744318765011393,
+      "grad_norm": 0.3359245159861427,
+      "learning_rate": 4.643246803402698e-06,
+      "loss": 0.4606,
+      "step": 15067
+    },
+    {
+      "epoch": 2.474596103709482,
+      "grad_norm": 0.384617259495692,
+      "learning_rate": 4.642778068095282e-06,
+      "loss": 0.4261,
+      "step": 15068
+    },
+    {
+      "epoch": 2.474760330917825,
+      "grad_norm": 0.29517197642369547,
+      "learning_rate": 4.642309331237157e-06,
+      "loss": 0.429,
+      "step": 15069
+    },
+    {
+      "epoch": 2.4749245581261676,
+      "grad_norm": 0.32753736877117223,
+      "learning_rate": 4.641840592833413e-06,
+      "loss": 0.4431,
+      "step": 15070
+    },
+    {
+      "epoch": 2.4750887853345103,
+      "grad_norm": 0.2791454602503895,
+      "learning_rate": 4.641371852889139e-06,
+      "loss": 0.4071,
+      "step": 15071
+    },
+    {
+      "epoch": 2.475253012542853,
+      "grad_norm": 0.37888047503031486,
+      "learning_rate": 4.640903111409428e-06,
+      "loss": 0.4417,
+      "step": 15072
+    },
+    {
+      "epoch": 2.475417239751196,
+      "grad_norm": 0.366354730985903,
+      "learning_rate": 4.640434368399371e-06,
+      "loss": 0.4554,
+      "step": 15073
+    },
+    {
+      "epoch": 2.4755814669595386,
+      "grad_norm": 0.31104150908583245,
+      "learning_rate": 4.63996562386406e-06,
+      "loss": 0.4289,
+      "step": 15074
+    },
+    {
+      "epoch": 2.4757456941678813,
+      "grad_norm": 0.37956424179769677,
+      "learning_rate": 4.6394968778085825e-06,
+      "loss": 0.4193,
+      "step": 15075
+    },
+    {
+      "epoch": 2.475909921376224,
+      "grad_norm": 0.335801656073786,
+      "learning_rate": 4.6390281302380325e-06,
+      "loss": 0.4467,
+      "step": 15076
+    },
+    {
+      "epoch": 2.4760741485845665,
+      "grad_norm": 0.2643865236365438,
+      "learning_rate": 4.6385593811575e-06,
+      "loss": 0.4419,
+      "step": 15077
+    },
+    {
+      "epoch": 2.4762383757929096,
+      "grad_norm": 0.30764530125320155,
+      "learning_rate": 4.638090630572076e-06,
+      "loss": 0.4202,
+      "step": 15078
+    },
+    {
+      "epoch": 2.4764026030012523,
+      "grad_norm": 0.32420305420903595,
+      "learning_rate": 4.637621878486853e-06,
+      "loss": 0.4522,
+      "step": 15079
+    },
+    {
+      "epoch": 2.476566830209595,
+      "grad_norm": 0.3363197644758546,
+      "learning_rate": 4.63715312490692e-06,
+      "loss": 0.4223,
+      "step": 15080
+    },
+    {
+      "epoch": 2.4767310574179375,
+      "grad_norm": 0.3091369151927837,
+      "learning_rate": 4.636684369837368e-06,
+      "loss": 0.4365,
+      "step": 15081
+    },
+    {
+      "epoch": 2.4768952846262806,
+      "grad_norm": 0.29132422243990463,
+      "learning_rate": 4.636215613283291e-06,
+      "loss": 0.4294,
+      "step": 15082
+    },
+    {
+      "epoch": 2.4770595118346233,
+      "grad_norm": 0.32071462618152996,
+      "learning_rate": 4.635746855249779e-06,
+      "loss": 0.4581,
+      "step": 15083
+    },
+    {
+      "epoch": 2.477223739042966,
+      "grad_norm": 0.39568575870858036,
+      "learning_rate": 4.635278095741922e-06,
+      "loss": 0.4408,
+      "step": 15084
+    },
+    {
+      "epoch": 2.4773879662513085,
+      "grad_norm": 0.2705314149096496,
+      "learning_rate": 4.634809334764811e-06,
+      "loss": 0.4407,
+      "step": 15085
+    },
+    {
+      "epoch": 2.4775521934596516,
+      "grad_norm": 0.4562209852816167,
+      "learning_rate": 4.634340572323538e-06,
+      "loss": 0.4238,
+      "step": 15086
+    },
+    {
+      "epoch": 2.4777164206679942,
+      "grad_norm": 0.35692718672318324,
+      "learning_rate": 4.633871808423195e-06,
+      "loss": 0.4366,
+      "step": 15087
+    },
+    {
+      "epoch": 2.477880647876337,
+      "grad_norm": 0.353983621673839,
+      "learning_rate": 4.633403043068873e-06,
+      "loss": 0.4546,
+      "step": 15088
+    },
+    {
+      "epoch": 2.4780448750846795,
+      "grad_norm": 0.42116045761854776,
+      "learning_rate": 4.632934276265661e-06,
+      "loss": 0.4482,
+      "step": 15089
+    },
+    {
+      "epoch": 2.4782091022930226,
+      "grad_norm": 0.3206255592467727,
+      "learning_rate": 4.6324655080186524e-06,
+      "loss": 0.4248,
+      "step": 15090
+    },
+    {
+      "epoch": 2.4783733295013652,
+      "grad_norm": 0.2792993323261359,
+      "learning_rate": 4.6319967383329395e-06,
+      "loss": 0.425,
+      "step": 15091
+    },
+    {
+      "epoch": 2.478537556709708,
+      "grad_norm": 0.2843568794714767,
+      "learning_rate": 4.631527967213611e-06,
+      "loss": 0.4416,
+      "step": 15092
+    },
+    {
+      "epoch": 2.4787017839180505,
+      "grad_norm": 0.27163007119109167,
+      "learning_rate": 4.631059194665759e-06,
+      "loss": 0.4263,
+      "step": 15093
+    },
+    {
+      "epoch": 2.478866011126393,
+      "grad_norm": 0.285132682533487,
+      "learning_rate": 4.630590420694475e-06,
+      "loss": 0.451,
+      "step": 15094
+    },
+    {
+      "epoch": 2.4790302383347362,
+      "grad_norm": 0.2962666477282085,
+      "learning_rate": 4.630121645304853e-06,
+      "loss": 0.4149,
+      "step": 15095
+    },
+    {
+      "epoch": 2.479194465543079,
+      "grad_norm": 0.3436820673601864,
+      "learning_rate": 4.629652868501979e-06,
+      "loss": 0.4354,
+      "step": 15096
+    },
+    {
+      "epoch": 2.4793586927514215,
+      "grad_norm": 0.30779119066240046,
+      "learning_rate": 4.629184090290948e-06,
+      "loss": 0.4039,
+      "step": 15097
+    },
+    {
+      "epoch": 2.479522919959764,
+      "grad_norm": 0.30109729627360404,
+      "learning_rate": 4.628715310676851e-06,
+      "loss": 0.4364,
+      "step": 15098
+    },
+    {
+      "epoch": 2.4796871471681072,
+      "grad_norm": 0.34213472664573985,
+      "learning_rate": 4.62824652966478e-06,
+      "loss": 0.4024,
+      "step": 15099
+    },
+    {
+      "epoch": 2.47985137437645,
+      "grad_norm": 0.3312074651939912,
+      "learning_rate": 4.627777747259825e-06,
+      "loss": 0.4494,
+      "step": 15100
+    },
+    {
+      "epoch": 2.4800156015847925,
+      "grad_norm": 0.3918133083512165,
+      "learning_rate": 4.627308963467076e-06,
+      "loss": 0.4383,
+      "step": 15101
+    },
+    {
+      "epoch": 2.480179828793135,
+      "grad_norm": 0.40037398856188533,
+      "learning_rate": 4.6268401782916285e-06,
+      "loss": 0.4496,
+      "step": 15102
+    },
+    {
+      "epoch": 2.4803440560014782,
+      "grad_norm": 0.29368723208387454,
+      "learning_rate": 4.6263713917385725e-06,
+      "loss": 0.4546,
+      "step": 15103
+    },
+    {
+      "epoch": 2.480508283209821,
+      "grad_norm": 0.39442925356351577,
+      "learning_rate": 4.625902603812998e-06,
+      "loss": 0.4396,
+      "step": 15104
+    },
+    {
+      "epoch": 2.4806725104181635,
+      "grad_norm": 0.49391288466102,
+      "learning_rate": 4.625433814519997e-06,
+      "loss": 0.4461,
+      "step": 15105
+    },
+    {
+      "epoch": 2.480836737626506,
+      "grad_norm": 0.6507872211415915,
+      "learning_rate": 4.624965023864661e-06,
+      "loss": 0.4257,
+      "step": 15106
+    },
+    {
+      "epoch": 2.481000964834849,
+      "grad_norm": 0.3170404081913076,
+      "learning_rate": 4.624496231852082e-06,
+      "loss": 0.4303,
+      "step": 15107
+    },
+    {
+      "epoch": 2.481165192043192,
+      "grad_norm": 0.8010381914422225,
+      "learning_rate": 4.624027438487352e-06,
+      "loss": 0.4287,
+      "step": 15108
+    },
+    {
+      "epoch": 2.4813294192515345,
+      "grad_norm": 0.3175103487015811,
+      "learning_rate": 4.623558643775561e-06,
+      "loss": 0.4396,
+      "step": 15109
+    },
+    {
+      "epoch": 2.481493646459877,
+      "grad_norm": 0.2780316156388087,
+      "learning_rate": 4.623089847721803e-06,
+      "loss": 0.4386,
+      "step": 15110
+    },
+    {
+      "epoch": 2.4816578736682198,
+      "grad_norm": 0.9957906981950299,
+      "learning_rate": 4.622621050331167e-06,
+      "loss": 0.4382,
+      "step": 15111
+    },
+    {
+      "epoch": 2.481822100876563,
+      "grad_norm": 0.3119134050343224,
+      "learning_rate": 4.622152251608747e-06,
+      "loss": 0.41,
+      "step": 15112
+    },
+    {
+      "epoch": 2.4819863280849055,
+      "grad_norm": 0.3165098283105381,
+      "learning_rate": 4.621683451559633e-06,
+      "loss": 0.4261,
+      "step": 15113
+    },
+    {
+      "epoch": 2.482150555293248,
+      "grad_norm": 0.3120583440078844,
+      "learning_rate": 4.621214650188916e-06,
+      "loss": 0.4349,
+      "step": 15114
+    },
+    {
+      "epoch": 2.4823147825015908,
+      "grad_norm": 0.2738497164372407,
+      "learning_rate": 4.620745847501689e-06,
+      "loss": 0.4321,
+      "step": 15115
+    },
+    {
+      "epoch": 2.482479009709934,
+      "grad_norm": 0.36008005991934383,
+      "learning_rate": 4.620277043503044e-06,
+      "loss": 0.4417,
+      "step": 15116
+    },
+    {
+      "epoch": 2.4826432369182765,
+      "grad_norm": 0.31989619467996416,
+      "learning_rate": 4.619808238198072e-06,
+      "loss": 0.446,
+      "step": 15117
+    },
+    {
+      "epoch": 2.482807464126619,
+      "grad_norm": 0.6824453394065131,
+      "learning_rate": 4.619339431591864e-06,
+      "loss": 0.4383,
+      "step": 15118
+    },
+    {
+      "epoch": 2.4829716913349618,
+      "grad_norm": 0.37095029373258626,
+      "learning_rate": 4.618870623689512e-06,
+      "loss": 0.4583,
+      "step": 15119
+    },
+    {
+      "epoch": 2.483135918543305,
+      "grad_norm": 0.33314170061608483,
+      "learning_rate": 4.618401814496109e-06,
+      "loss": 0.4348,
+      "step": 15120
+    },
+    {
+      "epoch": 2.4833001457516475,
+      "grad_norm": 0.33636320733646125,
+      "learning_rate": 4.617933004016744e-06,
+      "loss": 0.4554,
+      "step": 15121
+    },
+    {
+      "epoch": 2.48346437295999,
+      "grad_norm": 1.0196784103267829,
+      "learning_rate": 4.617464192256513e-06,
+      "loss": 0.4434,
+      "step": 15122
+    },
+    {
+      "epoch": 2.4836286001683328,
+      "grad_norm": 0.3193716312140261,
+      "learning_rate": 4.616995379220504e-06,
+      "loss": 0.43,
+      "step": 15123
+    },
+    {
+      "epoch": 2.483792827376676,
+      "grad_norm": 0.288516069215822,
+      "learning_rate": 4.616526564913811e-06,
+      "loss": 0.434,
+      "step": 15124
+    },
+    {
+      "epoch": 2.4839570545850185,
+      "grad_norm": 0.29912771431081314,
+      "learning_rate": 4.616057749341524e-06,
+      "loss": 0.4475,
+      "step": 15125
+    },
+    {
+      "epoch": 2.484121281793361,
+      "grad_norm": 0.2733627283094918,
+      "learning_rate": 4.615588932508735e-06,
+      "loss": 0.4382,
+      "step": 15126
+    },
+    {
+      "epoch": 2.4842855090017038,
+      "grad_norm": 0.3603048759548791,
+      "learning_rate": 4.615120114420538e-06,
+      "loss": 0.4283,
+      "step": 15127
+    },
+    {
+      "epoch": 2.4844497362100464,
+      "grad_norm": 0.3743687907784263,
+      "learning_rate": 4.6146512950820225e-06,
+      "loss": 0.448,
+      "step": 15128
+    },
+    {
+      "epoch": 2.4846139634183895,
+      "grad_norm": 0.3108138141975132,
+      "learning_rate": 4.614182474498282e-06,
+      "loss": 0.4537,
+      "step": 15129
+    },
+    {
+      "epoch": 2.484778190626732,
+      "grad_norm": 0.3197077167691703,
+      "learning_rate": 4.613713652674406e-06,
+      "loss": 0.438,
+      "step": 15130
+    },
+    {
+      "epoch": 2.4849424178350747,
+      "grad_norm": 0.4391802841487158,
+      "learning_rate": 4.613244829615488e-06,
+      "loss": 0.4311,
+      "step": 15131
+    },
+    {
+      "epoch": 2.4851066450434174,
+      "grad_norm": 0.2920981537029958,
+      "learning_rate": 4.612776005326621e-06,
+      "loss": 0.4237,
+      "step": 15132
+    },
+    {
+      "epoch": 2.4852708722517605,
+      "grad_norm": 0.29646873635485227,
+      "learning_rate": 4.612307179812896e-06,
+      "loss": 0.4341,
+      "step": 15133
+    },
+    {
+      "epoch": 2.485435099460103,
+      "grad_norm": 0.29841283343957575,
+      "learning_rate": 4.611838353079403e-06,
+      "loss": 0.4303,
+      "step": 15134
+    },
+    {
+      "epoch": 2.4855993266684457,
+      "grad_norm": 0.3250609621900752,
+      "learning_rate": 4.611369525131235e-06,
+      "loss": 0.4257,
+      "step": 15135
+    },
+    {
+      "epoch": 2.4857635538767884,
+      "grad_norm": 0.35478496086877537,
+      "learning_rate": 4.610900695973485e-06,
+      "loss": 0.4379,
+      "step": 15136
+    },
+    {
+      "epoch": 2.4859277810851315,
+      "grad_norm": 0.4670594900199308,
+      "learning_rate": 4.610431865611247e-06,
+      "loss": 0.4517,
+      "step": 15137
+    },
+    {
+      "epoch": 2.486092008293474,
+      "grad_norm": 0.31415101552518854,
+      "learning_rate": 4.609963034049607e-06,
+      "loss": 0.4298,
+      "step": 15138
+    },
+    {
+      "epoch": 2.4862562355018167,
+      "grad_norm": 0.38284691709415497,
+      "learning_rate": 4.609494201293661e-06,
+      "loss": 0.43,
+      "step": 15139
+    },
+    {
+      "epoch": 2.4864204627101594,
+      "grad_norm": 0.3092641007937,
+      "learning_rate": 4.6090253673484995e-06,
+      "loss": 0.421,
+      "step": 15140
+    },
+    {
+      "epoch": 2.4865846899185025,
+      "grad_norm": 0.35972316348470473,
+      "learning_rate": 4.608556532219216e-06,
+      "loss": 0.4597,
+      "step": 15141
+    },
+    {
+      "epoch": 2.486748917126845,
+      "grad_norm": 0.368530624076118,
+      "learning_rate": 4.608087695910903e-06,
+      "loss": 0.4258,
+      "step": 15142
+    },
+    {
+      "epoch": 2.4869131443351877,
+      "grad_norm": 0.303106049447321,
+      "learning_rate": 4.6076188584286505e-06,
+      "loss": 0.4249,
+      "step": 15143
+    },
+    {
+      "epoch": 2.4870773715435304,
+      "grad_norm": 0.48973722250178475,
+      "learning_rate": 4.607150019777551e-06,
+      "loss": 0.4328,
+      "step": 15144
+    },
+    {
+      "epoch": 2.487241598751873,
+      "grad_norm": 0.314524236592061,
+      "learning_rate": 4.606681179962697e-06,
+      "loss": 0.4424,
+      "step": 15145
+    },
+    {
+      "epoch": 2.487405825960216,
+      "grad_norm": 0.5027401543850439,
+      "learning_rate": 4.60621233898918e-06,
+      "loss": 0.4372,
+      "step": 15146
+    },
+    {
+      "epoch": 2.4875700531685587,
+      "grad_norm": 0.29423642278198925,
+      "learning_rate": 4.605743496862093e-06,
+      "loss": 0.4217,
+      "step": 15147
+    },
+    {
+      "epoch": 2.4877342803769014,
+      "grad_norm": 0.31679517006738683,
+      "learning_rate": 4.605274653586526e-06,
+      "loss": 0.4495,
+      "step": 15148
+    },
+    {
+      "epoch": 2.487898507585244,
+      "grad_norm": 0.26487858401307063,
+      "learning_rate": 4.604805809167574e-06,
+      "loss": 0.4323,
+      "step": 15149
+    },
+    {
+      "epoch": 2.488062734793587,
+      "grad_norm": 0.32293708597952736,
+      "learning_rate": 4.604336963610328e-06,
+      "loss": 0.4254,
+      "step": 15150
+    },
+    {
+      "epoch": 2.4882269620019297,
+      "grad_norm": 0.28820496702549153,
+      "learning_rate": 4.60386811691988e-06,
+      "loss": 0.4368,
+      "step": 15151
+    },
+    {
+      "epoch": 2.4883911892102724,
+      "grad_norm": 0.27518127880599397,
+      "learning_rate": 4.6033992691013225e-06,
+      "loss": 0.4453,
+      "step": 15152
+    },
+    {
+      "epoch": 2.488555416418615,
+      "grad_norm": 0.3493994813337365,
+      "learning_rate": 4.6029304201597456e-06,
+      "loss": 0.4322,
+      "step": 15153
+    },
+    {
+      "epoch": 2.488719643626958,
+      "grad_norm": 0.3251525156968644,
+      "learning_rate": 4.602461570100246e-06,
+      "loss": 0.4246,
+      "step": 15154
+    },
+    {
+      "epoch": 2.4888838708353007,
+      "grad_norm": 0.4157430492921045,
+      "learning_rate": 4.6019927189279096e-06,
+      "loss": 0.4371,
+      "step": 15155
+    },
+    {
+      "epoch": 2.4890480980436434,
+      "grad_norm": 0.30758658627784186,
+      "learning_rate": 4.601523866647834e-06,
+      "loss": 0.4267,
+      "step": 15156
+    },
+    {
+      "epoch": 2.489212325251986,
+      "grad_norm": 0.33772718364386484,
+      "learning_rate": 4.601055013265109e-06,
+      "loss": 0.4342,
+      "step": 15157
+    },
+    {
+      "epoch": 2.489376552460329,
+      "grad_norm": 0.36400549487339234,
+      "learning_rate": 4.6005861587848264e-06,
+      "loss": 0.4364,
+      "step": 15158
+    },
+    {
+      "epoch": 2.4895407796686717,
+      "grad_norm": 0.4213966493331531,
+      "learning_rate": 4.600117303212079e-06,
+      "loss": 0.3879,
+      "step": 15159
+    },
+    {
+      "epoch": 2.4897050068770143,
+      "grad_norm": 0.3207069643365226,
+      "learning_rate": 4.5996484465519594e-06,
+      "loss": 0.4352,
+      "step": 15160
+    },
+    {
+      "epoch": 2.489869234085357,
+      "grad_norm": 0.30520069096946645,
+      "learning_rate": 4.599179588809561e-06,
+      "loss": 0.4323,
+      "step": 15161
+    },
+    {
+      "epoch": 2.4900334612936996,
+      "grad_norm": 0.4020668168700281,
+      "learning_rate": 4.598710729989974e-06,
+      "loss": 0.4535,
+      "step": 15162
+    },
+    {
+      "epoch": 2.4901976885020427,
+      "grad_norm": 0.42245364077029657,
+      "learning_rate": 4.5982418700982905e-06,
+      "loss": 0.4525,
+      "step": 15163
+    },
+    {
+      "epoch": 2.4903619157103853,
+      "grad_norm": 0.8362459834605354,
+      "learning_rate": 4.597773009139604e-06,
+      "loss": 0.4362,
+      "step": 15164
+    },
+    {
+      "epoch": 2.490526142918728,
+      "grad_norm": 0.3449582767880722,
+      "learning_rate": 4.597304147119006e-06,
+      "loss": 0.4339,
+      "step": 15165
+    },
+    {
+      "epoch": 2.4906903701270706,
+      "grad_norm": 0.3066916301852268,
+      "learning_rate": 4.5968352840415904e-06,
+      "loss": 0.4309,
+      "step": 15166
+    },
+    {
+      "epoch": 2.4908545973354137,
+      "grad_norm": 0.5106999824766285,
+      "learning_rate": 4.596366419912448e-06,
+      "loss": 0.4531,
+      "step": 15167
+    },
+    {
+      "epoch": 2.4910188245437563,
+      "grad_norm": 0.34081483839278837,
+      "learning_rate": 4.595897554736671e-06,
+      "loss": 0.4411,
+      "step": 15168
+    },
+    {
+      "epoch": 2.491183051752099,
+      "grad_norm": 0.3703165743951746,
+      "learning_rate": 4.5954286885193514e-06,
+      "loss": 0.4255,
+      "step": 15169
+    },
+    {
+      "epoch": 2.4913472789604416,
+      "grad_norm": 0.2788333157420013,
+      "learning_rate": 4.5949598212655845e-06,
+      "loss": 0.4364,
+      "step": 15170
+    },
+    {
+      "epoch": 2.4915115061687847,
+      "grad_norm": 0.35393728188332935,
+      "learning_rate": 4.594490952980459e-06,
+      "loss": 0.4183,
+      "step": 15171
+    },
+    {
+      "epoch": 2.4916757333771273,
+      "grad_norm": 0.6014347064530057,
+      "learning_rate": 4.594022083669069e-06,
+      "loss": 0.4263,
+      "step": 15172
+    },
+    {
+      "epoch": 2.49183996058547,
+      "grad_norm": 0.3243854776951079,
+      "learning_rate": 4.593553213336507e-06,
+      "loss": 0.4318,
+      "step": 15173
+    },
+    {
+      "epoch": 2.4920041877938126,
+      "grad_norm": 0.392814323494369,
+      "learning_rate": 4.593084341987864e-06,
+      "loss": 0.4218,
+      "step": 15174
+    },
+    {
+      "epoch": 2.4921684150021557,
+      "grad_norm": 0.34098247165368967,
+      "learning_rate": 4.592615469628235e-06,
+      "loss": 0.4641,
+      "step": 15175
+    },
+    {
+      "epoch": 2.4923326422104983,
+      "grad_norm": 0.35972202448556473,
+      "learning_rate": 4.59214659626271e-06,
+      "loss": 0.4248,
+      "step": 15176
+    },
+    {
+      "epoch": 2.492496869418841,
+      "grad_norm": 0.2892343843339758,
+      "learning_rate": 4.591677721896382e-06,
+      "loss": 0.415,
+      "step": 15177
+    },
+    {
+      "epoch": 2.4926610966271836,
+      "grad_norm": 0.3099738318685391,
+      "learning_rate": 4.591208846534344e-06,
+      "loss": 0.4426,
+      "step": 15178
+    },
+    {
+      "epoch": 2.4928253238355262,
+      "grad_norm": 0.3571184083633324,
+      "learning_rate": 4.590739970181689e-06,
+      "loss": 0.4301,
+      "step": 15179
+    },
+    {
+      "epoch": 2.4929895510438693,
+      "grad_norm": 0.33853164613700226,
+      "learning_rate": 4.590271092843507e-06,
+      "loss": 0.449,
+      "step": 15180
+    },
+    {
+      "epoch": 2.493153778252212,
+      "grad_norm": 0.30508015812979095,
+      "learning_rate": 4.589802214524896e-06,
+      "loss": 0.4246,
+      "step": 15181
+    },
+    {
+      "epoch": 2.4933180054605546,
+      "grad_norm": 0.3546083751853678,
+      "learning_rate": 4.58933333523094e-06,
+      "loss": 0.4307,
+      "step": 15182
+    },
+    {
+      "epoch": 2.4934822326688972,
+      "grad_norm": 0.34676750407362766,
+      "learning_rate": 4.5888644549667384e-06,
+      "loss": 0.4382,
+      "step": 15183
+    },
+    {
+      "epoch": 2.4936464598772403,
+      "grad_norm": 0.33336637439983086,
+      "learning_rate": 4.58839557373738e-06,
+      "loss": 0.4363,
+      "step": 15184
+    },
+    {
+      "epoch": 2.493810687085583,
+      "grad_norm": 0.3562136027603175,
+      "learning_rate": 4.58792669154796e-06,
+      "loss": 0.4254,
+      "step": 15185
+    },
+    {
+      "epoch": 2.4939749142939256,
+      "grad_norm": 0.30767491608825387,
+      "learning_rate": 4.587457808403569e-06,
+      "loss": 0.4312,
+      "step": 15186
+    },
+    {
+      "epoch": 2.4941391415022682,
+      "grad_norm": 0.3717157302565822,
+      "learning_rate": 4.586988924309302e-06,
+      "loss": 0.4342,
+      "step": 15187
+    },
+    {
+      "epoch": 2.4943033687106113,
+      "grad_norm": 0.31073508635900254,
+      "learning_rate": 4.586520039270247e-06,
+      "loss": 0.4286,
+      "step": 15188
+    },
+    {
+      "epoch": 2.494467595918954,
+      "grad_norm": 0.300935391463759,
+      "learning_rate": 4.5860511532915e-06,
+      "loss": 0.429,
+      "step": 15189
+    },
+    {
+      "epoch": 2.4946318231272966,
+      "grad_norm": 0.2891538888734539,
+      "learning_rate": 4.585582266378153e-06,
+      "loss": 0.4273,
+      "step": 15190
+    },
+    {
+      "epoch": 2.4947960503356392,
+      "grad_norm": 0.3636855265250306,
+      "learning_rate": 4.5851133785353e-06,
+      "loss": 0.4359,
+      "step": 15191
+    },
+    {
+      "epoch": 2.4949602775439823,
+      "grad_norm": 0.31238609339074075,
+      "learning_rate": 4.584644489768029e-06,
+      "loss": 0.4496,
+      "step": 15192
+    },
+    {
+      "epoch": 2.495124504752325,
+      "grad_norm": 0.3466335457330843,
+      "learning_rate": 4.584175600081438e-06,
+      "loss": 0.4251,
+      "step": 15193
+    },
+    {
+      "epoch": 2.4952887319606676,
+      "grad_norm": 0.28513115942321365,
+      "learning_rate": 4.583706709480615e-06,
+      "loss": 0.4303,
+      "step": 15194
+    },
+    {
+      "epoch": 2.49545295916901,
+      "grad_norm": 0.3313604508733959,
+      "learning_rate": 4.5832378179706564e-06,
+      "loss": 0.4239,
+      "step": 15195
+    },
+    {
+      "epoch": 2.495617186377353,
+      "grad_norm": 0.2735952782125291,
+      "learning_rate": 4.582768925556653e-06,
+      "loss": 0.4314,
+      "step": 15196
+    },
+    {
+      "epoch": 2.495781413585696,
+      "grad_norm": 0.3199061503551554,
+      "learning_rate": 4.582300032243698e-06,
+      "loss": 0.4437,
+      "step": 15197
+    },
+    {
+      "epoch": 2.4959456407940386,
+      "grad_norm": 0.2941447415940853,
+      "learning_rate": 4.581831138036882e-06,
+      "loss": 0.4434,
+      "step": 15198
+    },
+    {
+      "epoch": 2.496109868002381,
+      "grad_norm": 0.3355786824721521,
+      "learning_rate": 4.5813622429413e-06,
+      "loss": 0.4277,
+      "step": 15199
+    },
+    {
+      "epoch": 2.496274095210724,
+      "grad_norm": 0.28103705919862076,
+      "learning_rate": 4.580893346962045e-06,
+      "loss": 0.4314,
+      "step": 15200
+    },
+    {
+      "epoch": 2.496438322419067,
+      "grad_norm": 0.3227157508902223,
+      "learning_rate": 4.5804244501042085e-06,
+      "loss": 0.4493,
+      "step": 15201
+    },
+    {
+      "epoch": 2.4966025496274096,
+      "grad_norm": 0.2970871193039058,
+      "learning_rate": 4.5799555523728824e-06,
+      "loss": 0.42,
+      "step": 15202
+    },
+    {
+      "epoch": 2.496766776835752,
+      "grad_norm": 0.4342254727787912,
+      "learning_rate": 4.57948665377316e-06,
+      "loss": 0.4333,
+      "step": 15203
+    },
+    {
+      "epoch": 2.496931004044095,
+      "grad_norm": 0.3422407517356136,
+      "learning_rate": 4.579017754310136e-06,
+      "loss": 0.4373,
+      "step": 15204
+    },
+    {
+      "epoch": 2.497095231252438,
+      "grad_norm": 0.32443689801201475,
+      "learning_rate": 4.578548853988901e-06,
+      "loss": 0.4196,
+      "step": 15205
+    },
+    {
+      "epoch": 2.4972594584607806,
+      "grad_norm": 0.3213615230691932,
+      "learning_rate": 4.578079952814547e-06,
+      "loss": 0.4284,
+      "step": 15206
+    },
+    {
+      "epoch": 2.497423685669123,
+      "grad_norm": 0.30105959294998685,
+      "learning_rate": 4.577611050792169e-06,
+      "loss": 0.4037,
+      "step": 15207
+    },
+    {
+      "epoch": 2.497587912877466,
+      "grad_norm": 0.29610221993478114,
+      "learning_rate": 4.577142147926859e-06,
+      "loss": 0.448,
+      "step": 15208
+    },
+    {
+      "epoch": 2.497752140085809,
+      "grad_norm": 0.300053251348721,
+      "learning_rate": 4.576673244223709e-06,
+      "loss": 0.4284,
+      "step": 15209
+    },
+    {
+      "epoch": 2.4979163672941516,
+      "grad_norm": 0.4108333131032182,
+      "learning_rate": 4.576204339687812e-06,
+      "loss": 0.4393,
+      "step": 15210
+    },
+    {
+      "epoch": 2.498080594502494,
+      "grad_norm": 0.42413555863246477,
+      "learning_rate": 4.57573543432426e-06,
+      "loss": 0.4404,
+      "step": 15211
+    },
+    {
+      "epoch": 2.498244821710837,
+      "grad_norm": 0.3744832314644596,
+      "learning_rate": 4.5752665281381474e-06,
+      "loss": 0.427,
+      "step": 15212
+    },
+    {
+      "epoch": 2.4984090489191795,
+      "grad_norm": 0.5036243765745887,
+      "learning_rate": 4.574797621134566e-06,
+      "loss": 0.4456,
+      "step": 15213
+    },
+    {
+      "epoch": 2.4985732761275226,
+      "grad_norm": 0.39270734558141457,
+      "learning_rate": 4.574328713318609e-06,
+      "loss": 0.4252,
+      "step": 15214
+    },
+    {
+      "epoch": 2.498737503335865,
+      "grad_norm": 0.29570628940704496,
+      "learning_rate": 4.57385980469537e-06,
+      "loss": 0.4527,
+      "step": 15215
+    },
+    {
+      "epoch": 2.498901730544208,
+      "grad_norm": 0.7676602433712431,
+      "learning_rate": 4.573390895269941e-06,
+      "loss": 0.44,
+      "step": 15216
+    },
+    {
+      "epoch": 2.4990659577525505,
+      "grad_norm": 0.6201610602230498,
+      "learning_rate": 4.572921985047413e-06,
+      "loss": 0.4306,
+      "step": 15217
+    },
+    {
+      "epoch": 2.4992301849608936,
+      "grad_norm": 1.715363184871039,
+      "learning_rate": 4.572453074032881e-06,
+      "loss": 0.4368,
+      "step": 15218
+    },
+    {
+      "epoch": 2.499394412169236,
+      "grad_norm": 0.3603458735228863,
+      "learning_rate": 4.571984162231437e-06,
+      "loss": 0.4428,
+      "step": 15219
+    },
+    {
+      "epoch": 2.499558639377579,
+      "grad_norm": 0.29670745837822965,
+      "learning_rate": 4.571515249648174e-06,
+      "loss": 0.408,
+      "step": 15220
+    },
+    {
+      "epoch": 2.4997228665859215,
+      "grad_norm": 0.28863059782458106,
+      "learning_rate": 4.571046336288186e-06,
+      "loss": 0.4034,
+      "step": 15221
+    },
+    {
+      "epoch": 2.4998870937942645,
+      "grad_norm": 0.3231624496999091,
+      "learning_rate": 4.570577422156564e-06,
+      "loss": 0.4262,
+      "step": 15222
+    },
+    {
+      "epoch": 2.500051321002607,
+      "grad_norm": 0.28074500386612955,
+      "learning_rate": 4.570108507258403e-06,
+      "loss": 0.4477,
+      "step": 15223
+    },
+    {
+      "epoch": 2.50021554821095,
+      "grad_norm": 0.2709352897613838,
+      "learning_rate": 4.569639591598794e-06,
+      "loss": 0.4358,
+      "step": 15224
+    },
+    {
+      "epoch": 2.5003797754192925,
+      "grad_norm": 0.28514705073201974,
+      "learning_rate": 4.569170675182831e-06,
+      "loss": 0.4236,
+      "step": 15225
+    },
+    {
+      "epoch": 2.5005440026276355,
+      "grad_norm": 0.2873450628834571,
+      "learning_rate": 4.5687017580156055e-06,
+      "loss": 0.4324,
+      "step": 15226
+    },
+    {
+      "epoch": 2.500708229835978,
+      "grad_norm": 0.29247613976893794,
+      "learning_rate": 4.568232840102211e-06,
+      "loss": 0.4417,
+      "step": 15227
+    },
+    {
+      "epoch": 2.500872457044321,
+      "grad_norm": 0.3039572928244555,
+      "learning_rate": 4.567763921447741e-06,
+      "loss": 0.424,
+      "step": 15228
+    },
+    {
+      "epoch": 2.5010366842526635,
+      "grad_norm": 0.3498139299117389,
+      "learning_rate": 4.56729500205729e-06,
+      "loss": 0.4579,
+      "step": 15229
+    },
+    {
+      "epoch": 2.501200911461006,
+      "grad_norm": 0.35502698847985975,
+      "learning_rate": 4.566826081935947e-06,
+      "loss": 0.4223,
+      "step": 15230
+    },
+    {
+      "epoch": 2.501365138669349,
+      "grad_norm": 0.3317335064782728,
+      "learning_rate": 4.566357161088808e-06,
+      "loss": 0.4263,
+      "step": 15231
+    },
+    {
+      "epoch": 2.501529365877692,
+      "grad_norm": 0.3393369552996207,
+      "learning_rate": 4.565888239520963e-06,
+      "loss": 0.4415,
+      "step": 15232
+    },
+    {
+      "epoch": 2.5016935930860345,
+      "grad_norm": 0.40649164473995997,
+      "learning_rate": 4.56541931723751e-06,
+      "loss": 0.4196,
+      "step": 15233
+    },
+    {
+      "epoch": 2.5018578202943775,
+      "grad_norm": 0.5651714899939198,
+      "learning_rate": 4.564950394243538e-06,
+      "loss": 0.43,
+      "step": 15234
+    },
+    {
+      "epoch": 2.50202204750272,
+      "grad_norm": 0.2903708190570917,
+      "learning_rate": 4.564481470544139e-06,
+      "loss": 0.4101,
+      "step": 15235
+    },
+    {
+      "epoch": 2.502186274711063,
+      "grad_norm": 0.407942080989367,
+      "learning_rate": 4.564012546144409e-06,
+      "loss": 0.4586,
+      "step": 15236
+    },
+    {
+      "epoch": 2.5023505019194054,
+      "grad_norm": 0.30899819635242237,
+      "learning_rate": 4.56354362104944e-06,
+      "loss": 0.4397,
+      "step": 15237
+    },
+    {
+      "epoch": 2.502514729127748,
+      "grad_norm": 0.3462923651598003,
+      "learning_rate": 4.563074695264324e-06,
+      "loss": 0.4532,
+      "step": 15238
+    },
+    {
+      "epoch": 2.5026789563360907,
+      "grad_norm": 0.3220893341552208,
+      "learning_rate": 4.562605768794156e-06,
+      "loss": 0.4317,
+      "step": 15239
+    },
+    {
+      "epoch": 2.502843183544434,
+      "grad_norm": 0.31180577802122095,
+      "learning_rate": 4.562136841644027e-06,
+      "loss": 0.4425,
+      "step": 15240
+    },
+    {
+      "epoch": 2.5030074107527764,
+      "grad_norm": 0.2764546692709857,
+      "learning_rate": 4.561667913819031e-06,
+      "loss": 0.4437,
+      "step": 15241
+    },
+    {
+      "epoch": 2.503171637961119,
+      "grad_norm": 0.31019314339615184,
+      "learning_rate": 4.56119898532426e-06,
+      "loss": 0.4204,
+      "step": 15242
+    },
+    {
+      "epoch": 2.503335865169462,
+      "grad_norm": 0.3931365921905194,
+      "learning_rate": 4.560730056164808e-06,
+      "loss": 0.4276,
+      "step": 15243
+    },
+    {
+      "epoch": 2.503500092377805,
+      "grad_norm": 0.3767687885431783,
+      "learning_rate": 4.560261126345769e-06,
+      "loss": 0.4344,
+      "step": 15244
+    },
+    {
+      "epoch": 2.5036643195861474,
+      "grad_norm": 0.35056231533364646,
+      "learning_rate": 4.5597921958722336e-06,
+      "loss": 0.439,
+      "step": 15245
+    },
+    {
+      "epoch": 2.50382854679449,
+      "grad_norm": 0.34796478913559775,
+      "learning_rate": 4.559323264749297e-06,
+      "loss": 0.4379,
+      "step": 15246
+    },
+    {
+      "epoch": 2.5039927740028327,
+      "grad_norm": 0.364929157109696,
+      "learning_rate": 4.5588543329820504e-06,
+      "loss": 0.435,
+      "step": 15247
+    },
+    {
+      "epoch": 2.504157001211176,
+      "grad_norm": 0.30504356991755455,
+      "learning_rate": 4.5583854005755875e-06,
+      "loss": 0.4206,
+      "step": 15248
+    },
+    {
+      "epoch": 2.5043212284195184,
+      "grad_norm": 0.33326552576505075,
+      "learning_rate": 4.5579164675350035e-06,
+      "loss": 0.4366,
+      "step": 15249
+    },
+    {
+      "epoch": 2.504485455627861,
+      "grad_norm": 0.30634993507320807,
+      "learning_rate": 4.557447533865389e-06,
+      "loss": 0.424,
+      "step": 15250
+    },
+    {
+      "epoch": 2.504649682836204,
+      "grad_norm": 0.3015603465660569,
+      "learning_rate": 4.556978599571838e-06,
+      "loss": 0.4437,
+      "step": 15251
+    },
+    {
+      "epoch": 2.504813910044547,
+      "grad_norm": 0.404288094881335,
+      "learning_rate": 4.556509664659441e-06,
+      "loss": 0.4333,
+      "step": 15252
+    },
+    {
+      "epoch": 2.5049781372528894,
+      "grad_norm": 0.26762791113011236,
+      "learning_rate": 4.556040729133297e-06,
+      "loss": 0.4433,
+      "step": 15253
+    },
+    {
+      "epoch": 2.505142364461232,
+      "grad_norm": 0.3917995761448291,
+      "learning_rate": 4.5555717929984945e-06,
+      "loss": 0.42,
+      "step": 15254
+    },
+    {
+      "epoch": 2.5053065916695747,
+      "grad_norm": 0.43586340257632006,
+      "learning_rate": 4.5551028562601255e-06,
+      "loss": 0.441,
+      "step": 15255
+    },
+    {
+      "epoch": 2.5054708188779173,
+      "grad_norm": 0.5715050248374042,
+      "learning_rate": 4.554633918923287e-06,
+      "loss": 0.4354,
+      "step": 15256
+    },
+    {
+      "epoch": 2.5056350460862604,
+      "grad_norm": 0.363662712363827,
+      "learning_rate": 4.554164980993069e-06,
+      "loss": 0.4175,
+      "step": 15257
+    },
+    {
+      "epoch": 2.505799273294603,
+      "grad_norm": 0.3369993889608383,
+      "learning_rate": 4.553696042474569e-06,
+      "loss": 0.4322,
+      "step": 15258
+    },
+    {
+      "epoch": 2.5059635005029457,
+      "grad_norm": 0.3051343689476861,
+      "learning_rate": 4.5532271033728745e-06,
+      "loss": 0.4337,
+      "step": 15259
+    },
+    {
+      "epoch": 2.5061277277112888,
+      "grad_norm": 0.34664385637290773,
+      "learning_rate": 4.552758163693083e-06,
+      "loss": 0.4378,
+      "step": 15260
+    },
+    {
+      "epoch": 2.5062919549196314,
+      "grad_norm": 0.38369436861216527,
+      "learning_rate": 4.552289223440284e-06,
+      "loss": 0.425,
+      "step": 15261
+    },
+    {
+      "epoch": 2.506456182127974,
+      "grad_norm": 0.31098976283094065,
+      "learning_rate": 4.551820282619575e-06,
+      "loss": 0.4392,
+      "step": 15262
+    },
+    {
+      "epoch": 2.5066204093363167,
+      "grad_norm": 0.2924212913716923,
+      "learning_rate": 4.551351341236044e-06,
+      "loss": 0.4387,
+      "step": 15263
+    },
+    {
+      "epoch": 2.5067846365446593,
+      "grad_norm": 0.3614952525688465,
+      "learning_rate": 4.55088239929479e-06,
+      "loss": 0.4477,
+      "step": 15264
+    },
+    {
+      "epoch": 2.5069488637530024,
+      "grad_norm": 0.31854411737094895,
+      "learning_rate": 4.5504134568009e-06,
+      "loss": 0.447,
+      "step": 15265
+    },
+    {
+      "epoch": 2.507113090961345,
+      "grad_norm": 0.33338819590292956,
+      "learning_rate": 4.549944513759473e-06,
+      "loss": 0.4543,
+      "step": 15266
+    },
+    {
+      "epoch": 2.5072773181696877,
+      "grad_norm": 0.37005363688745724,
+      "learning_rate": 4.549475570175597e-06,
+      "loss": 0.4504,
+      "step": 15267
+    },
+    {
+      "epoch": 2.5074415453780308,
+      "grad_norm": 0.31931176261796645,
+      "learning_rate": 4.5490066260543694e-06,
+      "loss": 0.4491,
+      "step": 15268
+    },
+    {
+      "epoch": 2.5076057725863734,
+      "grad_norm": 0.4028200604365174,
+      "learning_rate": 4.548537681400881e-06,
+      "loss": 0.4354,
+      "step": 15269
+    },
+    {
+      "epoch": 2.507769999794716,
+      "grad_norm": 0.34431494305601784,
+      "learning_rate": 4.548068736220224e-06,
+      "loss": 0.4327,
+      "step": 15270
+    },
+    {
+      "epoch": 2.5079342270030587,
+      "grad_norm": 0.3359105652446257,
+      "learning_rate": 4.547599790517496e-06,
+      "loss": 0.4452,
+      "step": 15271
+    },
+    {
+      "epoch": 2.5080984542114013,
+      "grad_norm": 0.36606184117765694,
+      "learning_rate": 4.547130844297786e-06,
+      "loss": 0.4168,
+      "step": 15272
+    },
+    {
+      "epoch": 2.508262681419744,
+      "grad_norm": 0.46539092867000703,
+      "learning_rate": 4.546661897566189e-06,
+      "loss": 0.43,
+      "step": 15273
+    },
+    {
+      "epoch": 2.508426908628087,
+      "grad_norm": 0.3260456914724407,
+      "learning_rate": 4.546192950327797e-06,
+      "loss": 0.4333,
+      "step": 15274
+    },
+    {
+      "epoch": 2.5085911358364297,
+      "grad_norm": 0.431282609159347,
+      "learning_rate": 4.545724002587706e-06,
+      "loss": 0.4328,
+      "step": 15275
+    },
+    {
+      "epoch": 2.5087553630447723,
+      "grad_norm": 0.31874399961113387,
+      "learning_rate": 4.5452550543510055e-06,
+      "loss": 0.4409,
+      "step": 15276
+    },
+    {
+      "epoch": 2.5089195902531154,
+      "grad_norm": 0.36712748209494483,
+      "learning_rate": 4.544786105622791e-06,
+      "loss": 0.4079,
+      "step": 15277
+    },
+    {
+      "epoch": 2.509083817461458,
+      "grad_norm": 0.2955054728370622,
+      "learning_rate": 4.544317156408157e-06,
+      "loss": 0.433,
+      "step": 15278
+    },
+    {
+      "epoch": 2.5092480446698007,
+      "grad_norm": 0.31795398888195725,
+      "learning_rate": 4.543848206712193e-06,
+      "loss": 0.4335,
+      "step": 15279
+    },
+    {
+      "epoch": 2.5094122718781433,
+      "grad_norm": 0.327166579049544,
+      "learning_rate": 4.543379256539995e-06,
+      "loss": 0.4405,
+      "step": 15280
+    },
+    {
+      "epoch": 2.509576499086486,
+      "grad_norm": 0.29009950534624457,
+      "learning_rate": 4.542910305896655e-06,
+      "loss": 0.4215,
+      "step": 15281
+    },
+    {
+      "epoch": 2.509740726294829,
+      "grad_norm": 0.5040405319999883,
+      "learning_rate": 4.542441354787268e-06,
+      "loss": 0.4146,
+      "step": 15282
+    },
+    {
+      "epoch": 2.5099049535031717,
+      "grad_norm": 0.3522597372682301,
+      "learning_rate": 4.541972403216927e-06,
+      "loss": 0.4387,
+      "step": 15283
+    },
+    {
+      "epoch": 2.5100691807115143,
+      "grad_norm": 0.2726311387047777,
+      "learning_rate": 4.5415034511907226e-06,
+      "loss": 0.4354,
+      "step": 15284
+    },
+    {
+      "epoch": 2.5102334079198574,
+      "grad_norm": 0.302638717120387,
+      "learning_rate": 4.5410344987137496e-06,
+      "loss": 0.435,
+      "step": 15285
+    },
+    {
+      "epoch": 2.5103976351282,
+      "grad_norm": 0.2967323251347567,
+      "learning_rate": 4.5405655457911026e-06,
+      "loss": 0.4407,
+      "step": 15286
+    },
+    {
+      "epoch": 2.5105618623365427,
+      "grad_norm": 0.31331167849939856,
+      "learning_rate": 4.540096592427874e-06,
+      "loss": 0.4412,
+      "step": 15287
+    },
+    {
+      "epoch": 2.5107260895448853,
+      "grad_norm": 0.2815771015808261,
+      "learning_rate": 4.539627638629157e-06,
+      "loss": 0.4126,
+      "step": 15288
+    },
+    {
+      "epoch": 2.510890316753228,
+      "grad_norm": 0.4327857806293426,
+      "learning_rate": 4.539158684400044e-06,
+      "loss": 0.4141,
+      "step": 15289
+    },
+    {
+      "epoch": 2.5110545439615706,
+      "grad_norm": 0.33791685760518037,
+      "learning_rate": 4.538689729745629e-06,
+      "loss": 0.4324,
+      "step": 15290
+    },
+    {
+      "epoch": 2.5112187711699137,
+      "grad_norm": 0.3428719333544963,
+      "learning_rate": 4.538220774671005e-06,
+      "loss": 0.4483,
+      "step": 15291
+    },
+    {
+      "epoch": 2.5113829983782563,
+      "grad_norm": 0.30672105224945806,
+      "learning_rate": 4.537751819181268e-06,
+      "loss": 0.4369,
+      "step": 15292
+    },
+    {
+      "epoch": 2.511547225586599,
+      "grad_norm": 0.6634715199384176,
+      "learning_rate": 4.537282863281509e-06,
+      "loss": 0.428,
+      "step": 15293
+    },
+    {
+      "epoch": 2.511711452794942,
+      "grad_norm": 0.31872231581495836,
+      "learning_rate": 4.536813906976819e-06,
+      "loss": 0.433,
+      "step": 15294
+    },
+    {
+      "epoch": 2.5118756800032846,
+      "grad_norm": 0.3231534958659712,
+      "learning_rate": 4.536344950272295e-06,
+      "loss": 0.4355,
+      "step": 15295
+    },
+    {
+      "epoch": 2.5120399072116273,
+      "grad_norm": 0.2757083960884838,
+      "learning_rate": 4.535875993173029e-06,
+      "loss": 0.424,
+      "step": 15296
+    },
+    {
+      "epoch": 2.51220413441997,
+      "grad_norm": 0.3142120617565392,
+      "learning_rate": 4.535407035684115e-06,
+      "loss": 0.4487,
+      "step": 15297
+    },
+    {
+      "epoch": 2.5123683616283126,
+      "grad_norm": 0.34568721423469645,
+      "learning_rate": 4.534938077810646e-06,
+      "loss": 0.4384,
+      "step": 15298
+    },
+    {
+      "epoch": 2.5125325888366556,
+      "grad_norm": 0.3676976562163103,
+      "learning_rate": 4.534469119557714e-06,
+      "loss": 0.4517,
+      "step": 15299
+    },
+    {
+      "epoch": 2.5126968160449983,
+      "grad_norm": 0.4372151567267689,
+      "learning_rate": 4.534000160930414e-06,
+      "loss": 0.4505,
+      "step": 15300
+    },
+    {
+      "epoch": 2.512861043253341,
+      "grad_norm": 0.38459047394944434,
+      "learning_rate": 4.5335312019338385e-06,
+      "loss": 0.4281,
+      "step": 15301
+    },
+    {
+      "epoch": 2.513025270461684,
+      "grad_norm": 0.2934325348012631,
+      "learning_rate": 4.533062242573082e-06,
+      "loss": 0.4287,
+      "step": 15302
+    },
+    {
+      "epoch": 2.5131894976700266,
+      "grad_norm": 0.3307628319172223,
+      "learning_rate": 4.532593282853236e-06,
+      "loss": 0.4387,
+      "step": 15303
+    },
+    {
+      "epoch": 2.5133537248783693,
+      "grad_norm": 0.5152030404954767,
+      "learning_rate": 4.532124322779395e-06,
+      "loss": 0.4185,
+      "step": 15304
+    },
+    {
+      "epoch": 2.513517952086712,
+      "grad_norm": 0.29601355750979563,
+      "learning_rate": 4.531655362356652e-06,
+      "loss": 0.4424,
+      "step": 15305
+    },
+    {
+      "epoch": 2.5136821792950546,
+      "grad_norm": 0.33875656452253916,
+      "learning_rate": 4.531186401590102e-06,
+      "loss": 0.4533,
+      "step": 15306
+    },
+    {
+      "epoch": 2.513846406503397,
+      "grad_norm": 0.2905312012969298,
+      "learning_rate": 4.530717440484836e-06,
+      "loss": 0.4417,
+      "step": 15307
+    },
+    {
+      "epoch": 2.5140106337117403,
+      "grad_norm": 0.45733470068768756,
+      "learning_rate": 4.530248479045949e-06,
+      "loss": 0.4504,
+      "step": 15308
+    },
+    {
+      "epoch": 2.514174860920083,
+      "grad_norm": 0.3232055665647335,
+      "learning_rate": 4.529779517278533e-06,
+      "loss": 0.4336,
+      "step": 15309
+    },
+    {
+      "epoch": 2.5143390881284255,
+      "grad_norm": 0.32595716015001,
+      "learning_rate": 4.529310555187681e-06,
+      "loss": 0.4239,
+      "step": 15310
+    },
+    {
+      "epoch": 2.5145033153367686,
+      "grad_norm": 0.32457344849632785,
+      "learning_rate": 4.528841592778489e-06,
+      "loss": 0.4606,
+      "step": 15311
+    },
+    {
+      "epoch": 2.5146675425451113,
+      "grad_norm": 0.4409322797051764,
+      "learning_rate": 4.528372630056049e-06,
+      "loss": 0.4443,
+      "step": 15312
+    },
+    {
+      "epoch": 2.514831769753454,
+      "grad_norm": 0.3427318010419413,
+      "learning_rate": 4.527903667025455e-06,
+      "loss": 0.4069,
+      "step": 15313
+    },
+    {
+      "epoch": 2.5149959969617965,
+      "grad_norm": 0.3094199668034755,
+      "learning_rate": 4.527434703691799e-06,
+      "loss": 0.4165,
+      "step": 15314
+    },
+    {
+      "epoch": 2.515160224170139,
+      "grad_norm": 0.3129531860016772,
+      "learning_rate": 4.526965740060174e-06,
+      "loss": 0.4453,
+      "step": 15315
+    },
+    {
+      "epoch": 2.5153244513784823,
+      "grad_norm": 0.36141679423305373,
+      "learning_rate": 4.526496776135675e-06,
+      "loss": 0.458,
+      "step": 15316
+    },
+    {
+      "epoch": 2.515488678586825,
+      "grad_norm": 0.32103349571056183,
+      "learning_rate": 4.526027811923398e-06,
+      "loss": 0.4344,
+      "step": 15317
+    },
+    {
+      "epoch": 2.5156529057951675,
+      "grad_norm": 0.35520418639316653,
+      "learning_rate": 4.52555884742843e-06,
+      "loss": 0.4542,
+      "step": 15318
+    },
+    {
+      "epoch": 2.5158171330035106,
+      "grad_norm": 0.38140314268757286,
+      "learning_rate": 4.525089882655868e-06,
+      "loss": 0.4279,
+      "step": 15319
+    },
+    {
+      "epoch": 2.5159813602118533,
+      "grad_norm": 0.3587596025920061,
+      "learning_rate": 4.524620917610805e-06,
+      "loss": 0.4487,
+      "step": 15320
+    },
+    {
+      "epoch": 2.516145587420196,
+      "grad_norm": 0.31918676746353725,
+      "learning_rate": 4.524151952298336e-06,
+      "loss": 0.4269,
+      "step": 15321
+    },
+    {
+      "epoch": 2.5163098146285385,
+      "grad_norm": 0.3006162767702321,
+      "learning_rate": 4.523682986723553e-06,
+      "loss": 0.4487,
+      "step": 15322
+    },
+    {
+      "epoch": 2.516474041836881,
+      "grad_norm": 0.35412734411765195,
+      "learning_rate": 4.523214020891549e-06,
+      "loss": 0.4191,
+      "step": 15323
+    },
+    {
+      "epoch": 2.516638269045224,
+      "grad_norm": 0.5091874206900988,
+      "learning_rate": 4.5227450548074165e-06,
+      "loss": 0.4503,
+      "step": 15324
+    },
+    {
+      "epoch": 2.516802496253567,
+      "grad_norm": 0.2851578874900258,
+      "learning_rate": 4.522276088476253e-06,
+      "loss": 0.4352,
+      "step": 15325
+    },
+    {
+      "epoch": 2.5169667234619095,
+      "grad_norm": 0.9997601491535643,
+      "learning_rate": 4.5218071219031476e-06,
+      "loss": 0.4468,
+      "step": 15326
+    },
+    {
+      "epoch": 2.517130950670252,
+      "grad_norm": 0.29348842306959855,
+      "learning_rate": 4.5213381550931955e-06,
+      "loss": 0.4525,
+      "step": 15327
+    },
+    {
+      "epoch": 2.5172951778785952,
+      "grad_norm": 0.2674552211461208,
+      "learning_rate": 4.52086918805149e-06,
+      "loss": 0.4416,
+      "step": 15328
+    },
+    {
+      "epoch": 2.517459405086938,
+      "grad_norm": 0.2949016922161546,
+      "learning_rate": 4.5204002207831255e-06,
+      "loss": 0.4262,
+      "step": 15329
+    },
+    {
+      "epoch": 2.5176236322952805,
+      "grad_norm": 0.32887885088298996,
+      "learning_rate": 4.519931253293194e-06,
+      "loss": 0.4439,
+      "step": 15330
+    },
+    {
+      "epoch": 2.517787859503623,
+      "grad_norm": 0.44021383142814,
+      "learning_rate": 4.519462285586789e-06,
+      "loss": 0.444,
+      "step": 15331
+    },
+    {
+      "epoch": 2.517952086711966,
+      "grad_norm": 0.3943727299935503,
+      "learning_rate": 4.518993317669005e-06,
+      "loss": 0.4211,
+      "step": 15332
+    },
+    {
+      "epoch": 2.518116313920309,
+      "grad_norm": 0.46339240270561693,
+      "learning_rate": 4.5185243495449346e-06,
+      "loss": 0.4441,
+      "step": 15333
+    },
+    {
+      "epoch": 2.5182805411286515,
+      "grad_norm": 0.2996537058649584,
+      "learning_rate": 4.518055381219671e-06,
+      "loss": 0.4177,
+      "step": 15334
+    },
+    {
+      "epoch": 2.518444768336994,
+      "grad_norm": 0.38842685502754193,
+      "learning_rate": 4.517586412698308e-06,
+      "loss": 0.4356,
+      "step": 15335
+    },
+    {
+      "epoch": 2.5186089955453372,
+      "grad_norm": 0.45335596462941113,
+      "learning_rate": 4.517117443985942e-06,
+      "loss": 0.4299,
+      "step": 15336
+    },
+    {
+      "epoch": 2.51877322275368,
+      "grad_norm": 0.3111046050312757,
+      "learning_rate": 4.516648475087662e-06,
+      "loss": 0.4326,
+      "step": 15337
+    },
+    {
+      "epoch": 2.5189374499620225,
+      "grad_norm": 0.3004520651914907,
+      "learning_rate": 4.516179506008563e-06,
+      "loss": 0.4291,
+      "step": 15338
+    },
+    {
+      "epoch": 2.519101677170365,
+      "grad_norm": 0.30623244552605644,
+      "learning_rate": 4.515710536753737e-06,
+      "loss": 0.4452,
+      "step": 15339
+    },
+    {
+      "epoch": 2.519265904378708,
+      "grad_norm": 0.29111646298893507,
+      "learning_rate": 4.515241567328281e-06,
+      "loss": 0.4212,
+      "step": 15340
+    },
+    {
+      "epoch": 2.5194301315870504,
+      "grad_norm": 0.292066653111252,
+      "learning_rate": 4.514772597737286e-06,
+      "loss": 0.4405,
+      "step": 15341
+    },
+    {
+      "epoch": 2.5195943587953935,
+      "grad_norm": 0.45941378411880673,
+      "learning_rate": 4.514303627985848e-06,
+      "loss": 0.4301,
+      "step": 15342
+    },
+    {
+      "epoch": 2.519758586003736,
+      "grad_norm": 0.36884625842137847,
+      "learning_rate": 4.513834658079056e-06,
+      "loss": 0.4311,
+      "step": 15343
+    },
+    {
+      "epoch": 2.519922813212079,
+      "grad_norm": 0.3267743781596869,
+      "learning_rate": 4.513365688022006e-06,
+      "loss": 0.4263,
+      "step": 15344
+    },
+    {
+      "epoch": 2.520087040420422,
+      "grad_norm": 0.4746571514823164,
+      "learning_rate": 4.512896717819792e-06,
+      "loss": 0.4373,
+      "step": 15345
+    },
+    {
+      "epoch": 2.5202512676287645,
+      "grad_norm": 0.3153602181102855,
+      "learning_rate": 4.512427747477508e-06,
+      "loss": 0.4146,
+      "step": 15346
+    },
+    {
+      "epoch": 2.520415494837107,
+      "grad_norm": 0.46102400221620937,
+      "learning_rate": 4.511958777000246e-06,
+      "loss": 0.4219,
+      "step": 15347
+    },
+    {
+      "epoch": 2.5205797220454498,
+      "grad_norm": 0.35545244939339854,
+      "learning_rate": 4.5114898063931e-06,
+      "loss": 0.4453,
+      "step": 15348
+    },
+    {
+      "epoch": 2.5207439492537924,
+      "grad_norm": 0.3778700125325818,
+      "learning_rate": 4.5110208356611625e-06,
+      "loss": 0.4593,
+      "step": 15349
+    },
+    {
+      "epoch": 2.5209081764621355,
+      "grad_norm": 0.43565446974417477,
+      "learning_rate": 4.510551864809529e-06,
+      "loss": 0.4214,
+      "step": 15350
+    },
+    {
+      "epoch": 2.521072403670478,
+      "grad_norm": 0.3153237572414692,
+      "learning_rate": 4.5100828938432915e-06,
+      "loss": 0.4309,
+      "step": 15351
+    },
+    {
+      "epoch": 2.5212366308788208,
+      "grad_norm": 0.3206707578049655,
+      "learning_rate": 4.509613922767543e-06,
+      "loss": 0.4325,
+      "step": 15352
+    },
+    {
+      "epoch": 2.521400858087164,
+      "grad_norm": 0.2925015730081957,
+      "learning_rate": 4.509144951587378e-06,
+      "loss": 0.4275,
+      "step": 15353
+    },
+    {
+      "epoch": 2.5215650852955065,
+      "grad_norm": 0.29391592396077754,
+      "learning_rate": 4.508675980307891e-06,
+      "loss": 0.4336,
+      "step": 15354
+    },
+    {
+      "epoch": 2.521729312503849,
+      "grad_norm": 0.3576499616385502,
+      "learning_rate": 4.508207008934173e-06,
+      "loss": 0.4186,
+      "step": 15355
+    },
+    {
+      "epoch": 2.5218935397121918,
+      "grad_norm": 0.2790704744061716,
+      "learning_rate": 4.50773803747132e-06,
+      "loss": 0.4364,
+      "step": 15356
+    },
+    {
+      "epoch": 2.5220577669205344,
+      "grad_norm": 0.2741415789587045,
+      "learning_rate": 4.507269065924424e-06,
+      "loss": 0.419,
+      "step": 15357
+    },
+    {
+      "epoch": 2.522221994128877,
+      "grad_norm": 0.31922742530373754,
+      "learning_rate": 4.506800094298579e-06,
+      "loss": 0.426,
+      "step": 15358
+    },
+    {
+      "epoch": 2.52238622133722,
+      "grad_norm": 0.30382675560648514,
+      "learning_rate": 4.506331122598877e-06,
+      "loss": 0.4343,
+      "step": 15359
+    },
+    {
+      "epoch": 2.5225504485455628,
+      "grad_norm": 0.37896547303248035,
+      "learning_rate": 4.505862150830413e-06,
+      "loss": 0.4466,
+      "step": 15360
+    },
+    {
+      "epoch": 2.5227146757539054,
+      "grad_norm": 0.3183774699805361,
+      "learning_rate": 4.505393178998282e-06,
+      "loss": 0.4527,
+      "step": 15361
+    },
+    {
+      "epoch": 2.5228789029622485,
+      "grad_norm": 0.2846751786737191,
+      "learning_rate": 4.504924207107573e-06,
+      "loss": 0.4336,
+      "step": 15362
+    },
+    {
+      "epoch": 2.523043130170591,
+      "grad_norm": 0.3478034808054333,
+      "learning_rate": 4.504455235163385e-06,
+      "loss": 0.4453,
+      "step": 15363
+    },
+    {
+      "epoch": 2.5232073573789338,
+      "grad_norm": 0.2946001247345606,
+      "learning_rate": 4.503986263170807e-06,
+      "loss": 0.4382,
+      "step": 15364
+    },
+    {
+      "epoch": 2.5233715845872764,
+      "grad_norm": 0.28990216086748805,
+      "learning_rate": 4.503517291134936e-06,
+      "loss": 0.4231,
+      "step": 15365
+    },
+    {
+      "epoch": 2.523535811795619,
+      "grad_norm": 0.3804646638287557,
+      "learning_rate": 4.503048319060862e-06,
+      "loss": 0.4286,
+      "step": 15366
+    },
+    {
+      "epoch": 2.523700039003962,
+      "grad_norm": 2.0413952323858604,
+      "learning_rate": 4.502579346953682e-06,
+      "loss": 0.445,
+      "step": 15367
+    },
+    {
+      "epoch": 2.5238642662123048,
+      "grad_norm": 0.5064912268763407,
+      "learning_rate": 4.5021103748184865e-06,
+      "loss": 0.4368,
+      "step": 15368
+    },
+    {
+      "epoch": 2.5240284934206474,
+      "grad_norm": 0.3052118739719438,
+      "learning_rate": 4.50164140266037e-06,
+      "loss": 0.4349,
+      "step": 15369
+    },
+    {
+      "epoch": 2.5241927206289905,
+      "grad_norm": 0.3198249052491719,
+      "learning_rate": 4.501172430484426e-06,
+      "loss": 0.4264,
+      "step": 15370
+    },
+    {
+      "epoch": 2.524356947837333,
+      "grad_norm": 0.5187552386652747,
+      "learning_rate": 4.5007034582957505e-06,
+      "loss": 0.4362,
+      "step": 15371
+    },
+    {
+      "epoch": 2.5245211750456757,
+      "grad_norm": 0.2570604115758852,
+      "learning_rate": 4.500234486099433e-06,
+      "loss": 0.4435,
+      "step": 15372
+    },
+    {
+      "epoch": 2.5246854022540184,
+      "grad_norm": 0.27968517151321504,
+      "learning_rate": 4.499765513900568e-06,
+      "loss": 0.4406,
+      "step": 15373
+    },
+    {
+      "epoch": 2.524849629462361,
+      "grad_norm": 0.33267787718052455,
+      "learning_rate": 4.499296541704251e-06,
+      "loss": 0.4247,
+      "step": 15374
+    },
+    {
+      "epoch": 2.5250138566707037,
+      "grad_norm": 0.3596710570956569,
+      "learning_rate": 4.498827569515574e-06,
+      "loss": 0.4202,
+      "step": 15375
+    },
+    {
+      "epoch": 2.5251780838790467,
+      "grad_norm": 0.3257269029831687,
+      "learning_rate": 4.49835859733963e-06,
+      "loss": 0.4417,
+      "step": 15376
+    },
+    {
+      "epoch": 2.5253423110873894,
+      "grad_norm": 0.4355993030836781,
+      "learning_rate": 4.4978896251815155e-06,
+      "loss": 0.4288,
+      "step": 15377
+    },
+    {
+      "epoch": 2.525506538295732,
+      "grad_norm": 0.32559613958151995,
+      "learning_rate": 4.49742065304632e-06,
+      "loss": 0.4051,
+      "step": 15378
+    },
+    {
+      "epoch": 2.525670765504075,
+      "grad_norm": 0.35226942817405693,
+      "learning_rate": 4.496951680939139e-06,
+      "loss": 0.4098,
+      "step": 15379
+    },
+    {
+      "epoch": 2.5258349927124177,
+      "grad_norm": 0.31375777585851555,
+      "learning_rate": 4.496482708865065e-06,
+      "loss": 0.4481,
+      "step": 15380
+    },
+    {
+      "epoch": 2.5259992199207604,
+      "grad_norm": 0.27628561903807153,
+      "learning_rate": 4.496013736829193e-06,
+      "loss": 0.4393,
+      "step": 15381
+    },
+    {
+      "epoch": 2.526163447129103,
+      "grad_norm": 0.3308393282880393,
+      "learning_rate": 4.495544764836616e-06,
+      "loss": 0.4384,
+      "step": 15382
+    },
+    {
+      "epoch": 2.5263276743374457,
+      "grad_norm": 0.3388520225022834,
+      "learning_rate": 4.495075792892426e-06,
+      "loss": 0.4407,
+      "step": 15383
+    },
+    {
+      "epoch": 2.5264919015457887,
+      "grad_norm": 0.2992347879039684,
+      "learning_rate": 4.494606821001719e-06,
+      "loss": 0.4266,
+      "step": 15384
+    },
+    {
+      "epoch": 2.5266561287541314,
+      "grad_norm": 0.3280214871252539,
+      "learning_rate": 4.4941378491695864e-06,
+      "loss": 0.4408,
+      "step": 15385
+    },
+    {
+      "epoch": 2.526820355962474,
+      "grad_norm": 0.31297700953904556,
+      "learning_rate": 4.493668877401125e-06,
+      "loss": 0.4237,
+      "step": 15386
+    },
+    {
+      "epoch": 2.526984583170817,
+      "grad_norm": 0.3584916739969418,
+      "learning_rate": 4.493199905701423e-06,
+      "loss": 0.4183,
+      "step": 15387
+    },
+    {
+      "epoch": 2.5271488103791597,
+      "grad_norm": 0.26870273735724437,
+      "learning_rate": 4.492730934075577e-06,
+      "loss": 0.4187,
+      "step": 15388
+    },
+    {
+      "epoch": 2.5273130375875024,
+      "grad_norm": 0.3480358281774671,
+      "learning_rate": 4.492261962528681e-06,
+      "loss": 0.4448,
+      "step": 15389
+    },
+    {
+      "epoch": 2.527477264795845,
+      "grad_norm": 0.3162081780616614,
+      "learning_rate": 4.491792991065828e-06,
+      "loss": 0.4443,
+      "step": 15390
+    },
+    {
+      "epoch": 2.5276414920041876,
+      "grad_norm": 0.2870245734500193,
+      "learning_rate": 4.49132401969211e-06,
+      "loss": 0.4417,
+      "step": 15391
+    },
+    {
+      "epoch": 2.5278057192125303,
+      "grad_norm": 0.2879612790355667,
+      "learning_rate": 4.490855048412621e-06,
+      "loss": 0.4515,
+      "step": 15392
+    },
+    {
+      "epoch": 2.5279699464208734,
+      "grad_norm": 0.3525720241438474,
+      "learning_rate": 4.490386077232457e-06,
+      "loss": 0.4362,
+      "step": 15393
+    },
+    {
+      "epoch": 2.528134173629216,
+      "grad_norm": 0.34886617957552496,
+      "learning_rate": 4.48991710615671e-06,
+      "loss": 0.4293,
+      "step": 15394
+    },
+    {
+      "epoch": 2.5282984008375586,
+      "grad_norm": 0.33907897924004743,
+      "learning_rate": 4.489448135190472e-06,
+      "loss": 0.4473,
+      "step": 15395
+    },
+    {
+      "epoch": 2.5284626280459017,
+      "grad_norm": 0.3035129464459638,
+      "learning_rate": 4.4889791643388385e-06,
+      "loss": 0.4349,
+      "step": 15396
+    },
+    {
+      "epoch": 2.5286268552542444,
+      "grad_norm": 0.30303534555734757,
+      "learning_rate": 4.488510193606901e-06,
+      "loss": 0.4357,
+      "step": 15397
+    },
+    {
+      "epoch": 2.528791082462587,
+      "grad_norm": 0.790153063740717,
+      "learning_rate": 4.4880412229997546e-06,
+      "loss": 0.427,
+      "step": 15398
+    },
+    {
+      "epoch": 2.5289553096709296,
+      "grad_norm": 0.3511038371128071,
+      "learning_rate": 4.487572252522493e-06,
+      "loss": 0.4357,
+      "step": 15399
+    },
+    {
+      "epoch": 2.5291195368792723,
+      "grad_norm": 0.41092611732691453,
+      "learning_rate": 4.4871032821802076e-06,
+      "loss": 0.438,
+      "step": 15400
+    },
+    {
+      "epoch": 2.5292837640876153,
+      "grad_norm": 0.3450527596392091,
+      "learning_rate": 4.4866343119779936e-06,
+      "loss": 0.456,
+      "step": 15401
+    },
+    {
+      "epoch": 2.529447991295958,
+      "grad_norm": 0.31325138329002983,
+      "learning_rate": 4.486165341920946e-06,
+      "loss": 0.4486,
+      "step": 15402
+    },
+    {
+      "epoch": 2.5296122185043006,
+      "grad_norm": 0.4049086399645973,
+      "learning_rate": 4.485696372014154e-06,
+      "loss": 0.4273,
+      "step": 15403
+    },
+    {
+      "epoch": 2.5297764457126437,
+      "grad_norm": 0.3570244695616439,
+      "learning_rate": 4.485227402262715e-06,
+      "loss": 0.4435,
+      "step": 15404
+    },
+    {
+      "epoch": 2.5299406729209863,
+      "grad_norm": 0.26325068167244875,
+      "learning_rate": 4.48475843267172e-06,
+      "loss": 0.4536,
+      "step": 15405
+    },
+    {
+      "epoch": 2.530104900129329,
+      "grad_norm": 0.4771733843902166,
+      "learning_rate": 4.484289463246263e-06,
+      "loss": 0.4402,
+      "step": 15406
+    },
+    {
+      "epoch": 2.5302691273376716,
+      "grad_norm": 0.5682448574100989,
+      "learning_rate": 4.483820493991438e-06,
+      "loss": 0.4431,
+      "step": 15407
+    },
+    {
+      "epoch": 2.5304333545460143,
+      "grad_norm": 0.4079954803323253,
+      "learning_rate": 4.483351524912339e-06,
+      "loss": 0.4283,
+      "step": 15408
+    },
+    {
+      "epoch": 2.530597581754357,
+      "grad_norm": 0.3742329415968252,
+      "learning_rate": 4.48288255601406e-06,
+      "loss": 0.4355,
+      "step": 15409
+    },
+    {
+      "epoch": 2.5307618089627,
+      "grad_norm": 0.31517860625091276,
+      "learning_rate": 4.482413587301691e-06,
+      "loss": 0.4219,
+      "step": 15410
+    },
+    {
+      "epoch": 2.5309260361710426,
+      "grad_norm": 0.32624001883317155,
+      "learning_rate": 4.48194461878033e-06,
+      "loss": 0.4355,
+      "step": 15411
+    },
+    {
+      "epoch": 2.5310902633793853,
+      "grad_norm": 0.4844251612004818,
+      "learning_rate": 4.481475650455067e-06,
+      "loss": 0.4177,
+      "step": 15412
+    },
+    {
+      "epoch": 2.5312544905877283,
+      "grad_norm": 0.41499405644931575,
+      "learning_rate": 4.481006682330996e-06,
+      "loss": 0.4381,
+      "step": 15413
+    },
+    {
+      "epoch": 2.531418717796071,
+      "grad_norm": 0.3254658251219828,
+      "learning_rate": 4.480537714413212e-06,
+      "loss": 0.4417,
+      "step": 15414
+    },
+    {
+      "epoch": 2.5315829450044136,
+      "grad_norm": 0.28061612293188354,
+      "learning_rate": 4.480068746706807e-06,
+      "loss": 0.4429,
+      "step": 15415
+    },
+    {
+      "epoch": 2.5317471722127562,
+      "grad_norm": 0.29524776058023783,
+      "learning_rate": 4.479599779216875e-06,
+      "loss": 0.436,
+      "step": 15416
+    },
+    {
+      "epoch": 2.531911399421099,
+      "grad_norm": 0.275370031459336,
+      "learning_rate": 4.47913081194851e-06,
+      "loss": 0.4402,
+      "step": 15417
+    },
+    {
+      "epoch": 2.532075626629442,
+      "grad_norm": 0.382911694822015,
+      "learning_rate": 4.478661844906805e-06,
+      "loss": 0.4273,
+      "step": 15418
+    },
+    {
+      "epoch": 2.5322398538377846,
+      "grad_norm": 0.2856386139248631,
+      "learning_rate": 4.4781928780968535e-06,
+      "loss": 0.4315,
+      "step": 15419
+    },
+    {
+      "epoch": 2.5324040810461272,
+      "grad_norm": 0.3662003334417495,
+      "learning_rate": 4.477723911523749e-06,
+      "loss": 0.4377,
+      "step": 15420
+    },
+    {
+      "epoch": 2.5325683082544703,
+      "grad_norm": 0.40571562617701123,
+      "learning_rate": 4.477254945192584e-06,
+      "loss": 0.4411,
+      "step": 15421
+    },
+    {
+      "epoch": 2.532732535462813,
+      "grad_norm": 0.33072718734849793,
+      "learning_rate": 4.4767859791084525e-06,
+      "loss": 0.4123,
+      "step": 15422
+    },
+    {
+      "epoch": 2.5328967626711556,
+      "grad_norm": 0.31030981069280167,
+      "learning_rate": 4.4763170132764474e-06,
+      "loss": 0.4415,
+      "step": 15423
+    },
+    {
+      "epoch": 2.5330609898794982,
+      "grad_norm": 0.3393961988335576,
+      "learning_rate": 4.475848047701664e-06,
+      "loss": 0.4484,
+      "step": 15424
+    },
+    {
+      "epoch": 2.533225217087841,
+      "grad_norm": 0.3355243579966591,
+      "learning_rate": 4.475379082389194e-06,
+      "loss": 0.4345,
+      "step": 15425
+    },
+    {
+      "epoch": 2.5333894442961835,
+      "grad_norm": 0.2739009867205716,
+      "learning_rate": 4.474910117344132e-06,
+      "loss": 0.4299,
+      "step": 15426
+    },
+    {
+      "epoch": 2.5335536715045266,
+      "grad_norm": 0.30376055594266194,
+      "learning_rate": 4.474441152571572e-06,
+      "loss": 0.4146,
+      "step": 15427
+    },
+    {
+      "epoch": 2.5337178987128692,
+      "grad_norm": 0.3275017172400074,
+      "learning_rate": 4.473972188076604e-06,
+      "loss": 0.421,
+      "step": 15428
+    },
+    {
+      "epoch": 2.533882125921212,
+      "grad_norm": 0.32453479680479597,
+      "learning_rate": 4.473503223864325e-06,
+      "loss": 0.4396,
+      "step": 15429
+    },
+    {
+      "epoch": 2.534046353129555,
+      "grad_norm": 0.48443152017160745,
+      "learning_rate": 4.473034259939826e-06,
+      "loss": 0.4342,
+      "step": 15430
+    },
+    {
+      "epoch": 2.5342105803378976,
+      "grad_norm": 0.30446815466227956,
+      "learning_rate": 4.472565296308202e-06,
+      "loss": 0.4375,
+      "step": 15431
+    },
+    {
+      "epoch": 2.5343748075462402,
+      "grad_norm": 0.28200424035613053,
+      "learning_rate": 4.472096332974545e-06,
+      "loss": 0.4148,
+      "step": 15432
+    },
+    {
+      "epoch": 2.534539034754583,
+      "grad_norm": 0.27325398958681874,
+      "learning_rate": 4.4716273699439515e-06,
+      "loss": 0.4212,
+      "step": 15433
+    },
+    {
+      "epoch": 2.5347032619629255,
+      "grad_norm": 0.2784740527403517,
+      "learning_rate": 4.471158407221511e-06,
+      "loss": 0.4329,
+      "step": 15434
+    },
+    {
+      "epoch": 2.5348674891712686,
+      "grad_norm": 0.33152751718058054,
+      "learning_rate": 4.470689444812318e-06,
+      "loss": 0.4278,
+      "step": 15435
+    },
+    {
+      "epoch": 2.535031716379611,
+      "grad_norm": 0.3668173661236447,
+      "learning_rate": 4.470220482721469e-06,
+      "loss": 0.4222,
+      "step": 15436
+    },
+    {
+      "epoch": 2.535195943587954,
+      "grad_norm": 0.30285348260361367,
+      "learning_rate": 4.469751520954053e-06,
+      "loss": 0.4429,
+      "step": 15437
+    },
+    {
+      "epoch": 2.535360170796297,
+      "grad_norm": 0.3455367713469772,
+      "learning_rate": 4.469282559515165e-06,
+      "loss": 0.4236,
+      "step": 15438
+    },
+    {
+      "epoch": 2.5355243980046396,
+      "grad_norm": 0.2989934245140012,
+      "learning_rate": 4.4688135984098994e-06,
+      "loss": 0.4492,
+      "step": 15439
+    },
+    {
+      "epoch": 2.535688625212982,
+      "grad_norm": 0.5199522243515112,
+      "learning_rate": 4.468344637643349e-06,
+      "loss": 0.4461,
+      "step": 15440
+    },
+    {
+      "epoch": 2.535852852421325,
+      "grad_norm": 0.3047139296928439,
+      "learning_rate": 4.467875677220605e-06,
+      "loss": 0.456,
+      "step": 15441
+    },
+    {
+      "epoch": 2.5360170796296675,
+      "grad_norm": 0.4383952993691737,
+      "learning_rate": 4.467406717146764e-06,
+      "loss": 0.4231,
+      "step": 15442
+    },
+    {
+      "epoch": 2.53618130683801,
+      "grad_norm": 0.29397560831634006,
+      "learning_rate": 4.466937757426919e-06,
+      "loss": 0.4269,
+      "step": 15443
+    },
+    {
+      "epoch": 2.536345534046353,
+      "grad_norm": 0.38538600677786994,
+      "learning_rate": 4.4664687980661635e-06,
+      "loss": 0.4527,
+      "step": 15444
+    },
+    {
+      "epoch": 2.536509761254696,
+      "grad_norm": 0.374965751531902,
+      "learning_rate": 4.465999839069588e-06,
+      "loss": 0.4519,
+      "step": 15445
+    },
+    {
+      "epoch": 2.5366739884630385,
+      "grad_norm": 0.3716477346667442,
+      "learning_rate": 4.465530880442287e-06,
+      "loss": 0.4409,
+      "step": 15446
+    },
+    {
+      "epoch": 2.5368382156713816,
+      "grad_norm": 1.7589956201460324,
+      "learning_rate": 4.4650619221893555e-06,
+      "loss": 0.431,
+      "step": 15447
+    },
+    {
+      "epoch": 2.537002442879724,
+      "grad_norm": 0.36684972649658787,
+      "learning_rate": 4.464592964315886e-06,
+      "loss": 0.4427,
+      "step": 15448
+    },
+    {
+      "epoch": 2.537166670088067,
+      "grad_norm": 0.33423643193368346,
+      "learning_rate": 4.464124006826971e-06,
+      "loss": 0.4296,
+      "step": 15449
+    },
+    {
+      "epoch": 2.5373308972964095,
+      "grad_norm": 0.3044931364167006,
+      "learning_rate": 4.463655049727705e-06,
+      "loss": 0.4352,
+      "step": 15450
+    },
+    {
+      "epoch": 2.537495124504752,
+      "grad_norm": 0.30416152178632283,
+      "learning_rate": 4.463186093023181e-06,
+      "loss": 0.437,
+      "step": 15451
+    },
+    {
+      "epoch": 2.537659351713095,
+      "grad_norm": 0.4942313735733957,
+      "learning_rate": 4.462717136718494e-06,
+      "loss": 0.4555,
+      "step": 15452
+    },
+    {
+      "epoch": 2.537823578921438,
+      "grad_norm": 0.3510260293567135,
+      "learning_rate": 4.462248180818733e-06,
+      "loss": 0.4535,
+      "step": 15453
+    },
+    {
+      "epoch": 2.5379878061297805,
+      "grad_norm": 0.3444097092072025,
+      "learning_rate": 4.461779225328995e-06,
+      "loss": 0.4235,
+      "step": 15454
+    },
+    {
+      "epoch": 2.5381520333381236,
+      "grad_norm": 0.3186023690214023,
+      "learning_rate": 4.461310270254372e-06,
+      "loss": 0.4269,
+      "step": 15455
+    },
+    {
+      "epoch": 2.538316260546466,
+      "grad_norm": 0.28283435452405786,
+      "learning_rate": 4.4608413155999574e-06,
+      "loss": 0.4461,
+      "step": 15456
+    },
+    {
+      "epoch": 2.538480487754809,
+      "grad_norm": 0.2835693384007139,
+      "learning_rate": 4.460372361370844e-06,
+      "loss": 0.4179,
+      "step": 15457
+    },
+    {
+      "epoch": 2.5386447149631515,
+      "grad_norm": 0.469588775510588,
+      "learning_rate": 4.459903407572127e-06,
+      "loss": 0.4278,
+      "step": 15458
+    },
+    {
+      "epoch": 2.538808942171494,
+      "grad_norm": 0.31968267198673356,
+      "learning_rate": 4.459434454208898e-06,
+      "loss": 0.4319,
+      "step": 15459
+    },
+    {
+      "epoch": 2.5389731693798367,
+      "grad_norm": 0.3139217810294942,
+      "learning_rate": 4.45896550128625e-06,
+      "loss": 0.4486,
+      "step": 15460
+    },
+    {
+      "epoch": 2.53913739658818,
+      "grad_norm": 0.6289982528503534,
+      "learning_rate": 4.458496548809279e-06,
+      "loss": 0.4383,
+      "step": 15461
+    },
+    {
+      "epoch": 2.5393016237965225,
+      "grad_norm": 0.43248669042514953,
+      "learning_rate": 4.458027596783075e-06,
+      "loss": 0.4298,
+      "step": 15462
+    },
+    {
+      "epoch": 2.539465851004865,
+      "grad_norm": 0.32175232275524884,
+      "learning_rate": 4.457558645212733e-06,
+      "loss": 0.4305,
+      "step": 15463
+    },
+    {
+      "epoch": 2.539630078213208,
+      "grad_norm": 0.33913465217808964,
+      "learning_rate": 4.457089694103345e-06,
+      "loss": 0.4423,
+      "step": 15464
+    },
+    {
+      "epoch": 2.539794305421551,
+      "grad_norm": 0.4217029864914242,
+      "learning_rate": 4.456620743460005e-06,
+      "loss": 0.4327,
+      "step": 15465
+    },
+    {
+      "epoch": 2.5399585326298935,
+      "grad_norm": 0.32107328028587867,
+      "learning_rate": 4.456151793287807e-06,
+      "loss": 0.4595,
+      "step": 15466
+    },
+    {
+      "epoch": 2.540122759838236,
+      "grad_norm": 0.4056818357852612,
+      "learning_rate": 4.455682843591845e-06,
+      "loss": 0.4314,
+      "step": 15467
+    },
+    {
+      "epoch": 2.5402869870465787,
+      "grad_norm": 0.3285580483437489,
+      "learning_rate": 4.455213894377208e-06,
+      "loss": 0.4394,
+      "step": 15468
+    },
+    {
+      "epoch": 2.540451214254922,
+      "grad_norm": 0.4463962845057985,
+      "learning_rate": 4.454744945648996e-06,
+      "loss": 0.4365,
+      "step": 15469
+    },
+    {
+      "epoch": 2.5406154414632645,
+      "grad_norm": 0.5759978564993642,
+      "learning_rate": 4.454275997412296e-06,
+      "loss": 0.4314,
+      "step": 15470
+    },
+    {
+      "epoch": 2.540779668671607,
+      "grad_norm": 0.30311744414915,
+      "learning_rate": 4.453807049672203e-06,
+      "loss": 0.4443,
+      "step": 15471
+    },
+    {
+      "epoch": 2.54094389587995,
+      "grad_norm": 0.48319619600914393,
+      "learning_rate": 4.453338102433812e-06,
+      "loss": 0.4312,
+      "step": 15472
+    },
+    {
+      "epoch": 2.541108123088293,
+      "grad_norm": 0.27961971412030734,
+      "learning_rate": 4.452869155702215e-06,
+      "loss": 0.439,
+      "step": 15473
+    },
+    {
+      "epoch": 2.5412723502966355,
+      "grad_norm": 0.33938593418946583,
+      "learning_rate": 4.452400209482505e-06,
+      "loss": 0.419,
+      "step": 15474
+    },
+    {
+      "epoch": 2.541436577504978,
+      "grad_norm": 0.31723679723348247,
+      "learning_rate": 4.451931263779775e-06,
+      "loss": 0.42,
+      "step": 15475
+    },
+    {
+      "epoch": 2.5416008047133207,
+      "grad_norm": 0.3900582478708769,
+      "learning_rate": 4.4514623185991195e-06,
+      "loss": 0.4435,
+      "step": 15476
+    },
+    {
+      "epoch": 2.5417650319216634,
+      "grad_norm": 0.3161165257957074,
+      "learning_rate": 4.4509933739456325e-06,
+      "loss": 0.4175,
+      "step": 15477
+    },
+    {
+      "epoch": 2.5419292591300064,
+      "grad_norm": 0.3988098521207848,
+      "learning_rate": 4.450524429824405e-06,
+      "loss": 0.4453,
+      "step": 15478
+    },
+    {
+      "epoch": 2.542093486338349,
+      "grad_norm": 0.32174868195541384,
+      "learning_rate": 4.45005548624053e-06,
+      "loss": 0.4372,
+      "step": 15479
+    },
+    {
+      "epoch": 2.5422577135466917,
+      "grad_norm": 0.3045389565255562,
+      "learning_rate": 4.449586543199101e-06,
+      "loss": 0.4247,
+      "step": 15480
+    },
+    {
+      "epoch": 2.542421940755035,
+      "grad_norm": 0.2908792710329188,
+      "learning_rate": 4.4491176007052115e-06,
+      "loss": 0.4271,
+      "step": 15481
+    },
+    {
+      "epoch": 2.5425861679633774,
+      "grad_norm": 0.29144215488845143,
+      "learning_rate": 4.448648658763957e-06,
+      "loss": 0.4374,
+      "step": 15482
+    },
+    {
+      "epoch": 2.54275039517172,
+      "grad_norm": 0.3977821028455875,
+      "learning_rate": 4.448179717380426e-06,
+      "loss": 0.4574,
+      "step": 15483
+    },
+    {
+      "epoch": 2.5429146223800627,
+      "grad_norm": 0.33925302540685487,
+      "learning_rate": 4.447710776559716e-06,
+      "loss": 0.4403,
+      "step": 15484
+    },
+    {
+      "epoch": 2.5430788495884054,
+      "grad_norm": 0.38977415792187947,
+      "learning_rate": 4.447241836306917e-06,
+      "loss": 0.4484,
+      "step": 15485
+    },
+    {
+      "epoch": 2.5432430767967484,
+      "grad_norm": 0.3444752824234094,
+      "learning_rate": 4.4467728966271265e-06,
+      "loss": 0.4188,
+      "step": 15486
+    },
+    {
+      "epoch": 2.543407304005091,
+      "grad_norm": 0.29387542378527076,
+      "learning_rate": 4.446303957525432e-06,
+      "loss": 0.4551,
+      "step": 15487
+    },
+    {
+      "epoch": 2.5435715312134337,
+      "grad_norm": 0.3891893197024633,
+      "learning_rate": 4.445835019006931e-06,
+      "loss": 0.4435,
+      "step": 15488
+    },
+    {
+      "epoch": 2.543735758421777,
+      "grad_norm": 0.5432113733163281,
+      "learning_rate": 4.445366081076714e-06,
+      "loss": 0.4286,
+      "step": 15489
+    },
+    {
+      "epoch": 2.5438999856301194,
+      "grad_norm": 0.30458472444959955,
+      "learning_rate": 4.444897143739875e-06,
+      "loss": 0.4291,
+      "step": 15490
+    },
+    {
+      "epoch": 2.544064212838462,
+      "grad_norm": 0.33335443953907484,
+      "learning_rate": 4.444428207001507e-06,
+      "loss": 0.4427,
+      "step": 15491
+    },
+    {
+      "epoch": 2.5442284400468047,
+      "grad_norm": 0.3114285780540954,
+      "learning_rate": 4.4439592708667044e-06,
+      "loss": 0.4315,
+      "step": 15492
+    },
+    {
+      "epoch": 2.5443926672551473,
+      "grad_norm": 0.33859893153267717,
+      "learning_rate": 4.443490335340558e-06,
+      "loss": 0.4292,
+      "step": 15493
+    },
+    {
+      "epoch": 2.54455689446349,
+      "grad_norm": 0.34687445804031664,
+      "learning_rate": 4.443021400428164e-06,
+      "loss": 0.4318,
+      "step": 15494
+    },
+    {
+      "epoch": 2.544721121671833,
+      "grad_norm": 0.2773802250144603,
+      "learning_rate": 4.442552466134613e-06,
+      "loss": 0.4369,
+      "step": 15495
+    },
+    {
+      "epoch": 2.5448853488801757,
+      "grad_norm": 0.29535434743502614,
+      "learning_rate": 4.4420835324649976e-06,
+      "loss": 0.4346,
+      "step": 15496
+    },
+    {
+      "epoch": 2.5450495760885183,
+      "grad_norm": 0.9360580914134653,
+      "learning_rate": 4.441614599424413e-06,
+      "loss": 0.4517,
+      "step": 15497
+    },
+    {
+      "epoch": 2.5452138032968614,
+      "grad_norm": 0.7513840159695082,
+      "learning_rate": 4.441145667017951e-06,
+      "loss": 0.4466,
+      "step": 15498
+    },
+    {
+      "epoch": 2.545378030505204,
+      "grad_norm": 0.34400842175373336,
+      "learning_rate": 4.4406767352507045e-06,
+      "loss": 0.4268,
+      "step": 15499
+    },
+    {
+      "epoch": 2.5455422577135467,
+      "grad_norm": 0.36609020722625096,
+      "learning_rate": 4.440207804127767e-06,
+      "loss": 0.4322,
+      "step": 15500
+    },
+    {
+      "epoch": 2.5457064849218893,
+      "grad_norm": 0.30316434981498774,
+      "learning_rate": 4.439738873654232e-06,
+      "loss": 0.4371,
+      "step": 15501
+    },
+    {
+      "epoch": 2.545870712130232,
+      "grad_norm": 0.2875496740502021,
+      "learning_rate": 4.439269943835192e-06,
+      "loss": 0.4177,
+      "step": 15502
+    },
+    {
+      "epoch": 2.546034939338575,
+      "grad_norm": 0.3142325124623637,
+      "learning_rate": 4.438801014675742e-06,
+      "loss": 0.4359,
+      "step": 15503
+    },
+    {
+      "epoch": 2.5461991665469177,
+      "grad_norm": 0.3137356864008128,
+      "learning_rate": 4.43833208618097e-06,
+      "loss": 0.4371,
+      "step": 15504
+    },
+    {
+      "epoch": 2.5463633937552603,
+      "grad_norm": 0.32170461352243107,
+      "learning_rate": 4.437863158355975e-06,
+      "loss": 0.4447,
+      "step": 15505
+    },
+    {
+      "epoch": 2.5465276209636034,
+      "grad_norm": 0.3802915803414638,
+      "learning_rate": 4.437394231205845e-06,
+      "loss": 0.4345,
+      "step": 15506
+    },
+    {
+      "epoch": 2.546691848171946,
+      "grad_norm": 0.31857101888559824,
+      "learning_rate": 4.436925304735677e-06,
+      "loss": 0.4309,
+      "step": 15507
+    },
+    {
+      "epoch": 2.5468560753802887,
+      "grad_norm": 0.2584963933988361,
+      "learning_rate": 4.4364563789505604e-06,
+      "loss": 0.4443,
+      "step": 15508
+    },
+    {
+      "epoch": 2.5470203025886313,
+      "grad_norm": 0.4038773729499395,
+      "learning_rate": 4.435987453855591e-06,
+      "loss": 0.4472,
+      "step": 15509
+    },
+    {
+      "epoch": 2.547184529796974,
+      "grad_norm": 0.31802899292435854,
+      "learning_rate": 4.43551852945586e-06,
+      "loss": 0.4401,
+      "step": 15510
+    },
+    {
+      "epoch": 2.5473487570053166,
+      "grad_norm": 0.3242892323716624,
+      "learning_rate": 4.435049605756464e-06,
+      "loss": 0.4332,
+      "step": 15511
+    },
+    {
+      "epoch": 2.5475129842136597,
+      "grad_norm": 0.26797913518694794,
+      "learning_rate": 4.434580682762491e-06,
+      "loss": 0.4278,
+      "step": 15512
+    },
+    {
+      "epoch": 2.5476772114220023,
+      "grad_norm": 0.44582865863931953,
+      "learning_rate": 4.434111760479037e-06,
+      "loss": 0.4533,
+      "step": 15513
+    },
+    {
+      "epoch": 2.547841438630345,
+      "grad_norm": 0.3055931723597788,
+      "learning_rate": 4.433642838911193e-06,
+      "loss": 0.4502,
+      "step": 15514
+    },
+    {
+      "epoch": 2.548005665838688,
+      "grad_norm": 0.3085446507289279,
+      "learning_rate": 4.433173918064053e-06,
+      "loss": 0.4381,
+      "step": 15515
+    },
+    {
+      "epoch": 2.5481698930470307,
+      "grad_norm": 0.2799455160677543,
+      "learning_rate": 4.432704997942711e-06,
+      "loss": 0.4293,
+      "step": 15516
+    },
+    {
+      "epoch": 2.5483341202553733,
+      "grad_norm": 0.3072559176696745,
+      "learning_rate": 4.432236078552259e-06,
+      "loss": 0.4313,
+      "step": 15517
+    },
+    {
+      "epoch": 2.548498347463716,
+      "grad_norm": 0.2794966100213563,
+      "learning_rate": 4.4317671598977885e-06,
+      "loss": 0.4265,
+      "step": 15518
+    },
+    {
+      "epoch": 2.5486625746720586,
+      "grad_norm": 0.28997197483099796,
+      "learning_rate": 4.431298241984396e-06,
+      "loss": 0.4536,
+      "step": 15519
+    },
+    {
+      "epoch": 2.5488268018804017,
+      "grad_norm": 0.4573616100826798,
+      "learning_rate": 4.430829324817171e-06,
+      "loss": 0.4291,
+      "step": 15520
+    },
+    {
+      "epoch": 2.5489910290887443,
+      "grad_norm": 0.4438472979662701,
+      "learning_rate": 4.430360408401207e-06,
+      "loss": 0.4267,
+      "step": 15521
+    },
+    {
+      "epoch": 2.549155256297087,
+      "grad_norm": 0.2577708578255245,
+      "learning_rate": 4.429891492741598e-06,
+      "loss": 0.448,
+      "step": 15522
+    },
+    {
+      "epoch": 2.54931948350543,
+      "grad_norm": 0.32911337426606396,
+      "learning_rate": 4.429422577843436e-06,
+      "loss": 0.4351,
+      "step": 15523
+    },
+    {
+      "epoch": 2.5494837107137727,
+      "grad_norm": 0.3524767093716376,
+      "learning_rate": 4.428953663711814e-06,
+      "loss": 0.4165,
+      "step": 15524
+    },
+    {
+      "epoch": 2.5496479379221153,
+      "grad_norm": 0.29196546007234,
+      "learning_rate": 4.428484750351825e-06,
+      "loss": 0.4344,
+      "step": 15525
+    },
+    {
+      "epoch": 2.549812165130458,
+      "grad_norm": 0.298449530861429,
+      "learning_rate": 4.428015837768563e-06,
+      "loss": 0.4259,
+      "step": 15526
+    },
+    {
+      "epoch": 2.5499763923388006,
+      "grad_norm": 0.6148962359432069,
+      "learning_rate": 4.42754692596712e-06,
+      "loss": 0.4403,
+      "step": 15527
+    },
+    {
+      "epoch": 2.550140619547143,
+      "grad_norm": 0.3280661317620682,
+      "learning_rate": 4.42707801495259e-06,
+      "loss": 0.4275,
+      "step": 15528
+    },
+    {
+      "epoch": 2.5503048467554863,
+      "grad_norm": 0.2837424072375426,
+      "learning_rate": 4.426609104730062e-06,
+      "loss": 0.4427,
+      "step": 15529
+    },
+    {
+      "epoch": 2.550469073963829,
+      "grad_norm": 0.28041871680203856,
+      "learning_rate": 4.426140195304631e-06,
+      "loss": 0.4635,
+      "step": 15530
+    },
+    {
+      "epoch": 2.5506333011721716,
+      "grad_norm": 0.350200009979362,
+      "learning_rate": 4.425671286681392e-06,
+      "loss": 0.4343,
+      "step": 15531
+    },
+    {
+      "epoch": 2.5507975283805147,
+      "grad_norm": 0.32454220273887197,
+      "learning_rate": 4.425202378865435e-06,
+      "loss": 0.4614,
+      "step": 15532
+    },
+    {
+      "epoch": 2.5509617555888573,
+      "grad_norm": 0.5367213364709559,
+      "learning_rate": 4.424733471861853e-06,
+      "loss": 0.4101,
+      "step": 15533
+    },
+    {
+      "epoch": 2.5511259827972,
+      "grad_norm": 0.28676733476987987,
+      "learning_rate": 4.42426456567574e-06,
+      "loss": 0.4259,
+      "step": 15534
+    },
+    {
+      "epoch": 2.5512902100055426,
+      "grad_norm": 0.2975809840115441,
+      "learning_rate": 4.423795660312189e-06,
+      "loss": 0.4426,
+      "step": 15535
+    },
+    {
+      "epoch": 2.551454437213885,
+      "grad_norm": 0.3249328928419927,
+      "learning_rate": 4.423326755776292e-06,
+      "loss": 0.4473,
+      "step": 15536
+    },
+    {
+      "epoch": 2.5516186644222283,
+      "grad_norm": 0.31184122620432747,
+      "learning_rate": 4.422857852073143e-06,
+      "loss": 0.4104,
+      "step": 15537
+    },
+    {
+      "epoch": 2.551782891630571,
+      "grad_norm": 0.30635029912032957,
+      "learning_rate": 4.422388949207831e-06,
+      "loss": 0.4345,
+      "step": 15538
+    },
+    {
+      "epoch": 2.5519471188389136,
+      "grad_norm": 0.33025971138949717,
+      "learning_rate": 4.421920047185453e-06,
+      "loss": 0.4492,
+      "step": 15539
+    },
+    {
+      "epoch": 2.5521113460472566,
+      "grad_norm": 0.3181232184610691,
+      "learning_rate": 4.421451146011099e-06,
+      "loss": 0.4229,
+      "step": 15540
+    },
+    {
+      "epoch": 2.5522755732555993,
+      "grad_norm": 0.28306483682463085,
+      "learning_rate": 4.420982245689865e-06,
+      "loss": 0.4365,
+      "step": 15541
+    },
+    {
+      "epoch": 2.552439800463942,
+      "grad_norm": 0.34612517738918935,
+      "learning_rate": 4.420513346226839e-06,
+      "loss": 0.432,
+      "step": 15542
+    },
+    {
+      "epoch": 2.5526040276722846,
+      "grad_norm": 0.30498092076014316,
+      "learning_rate": 4.420044447627118e-06,
+      "loss": 0.4353,
+      "step": 15543
+    },
+    {
+      "epoch": 2.552768254880627,
+      "grad_norm": 0.3009496686032841,
+      "learning_rate": 4.419575549895793e-06,
+      "loss": 0.4316,
+      "step": 15544
+    },
+    {
+      "epoch": 2.55293248208897,
+      "grad_norm": 0.29409546874362075,
+      "learning_rate": 4.419106653037956e-06,
+      "loss": 0.4305,
+      "step": 15545
+    },
+    {
+      "epoch": 2.553096709297313,
+      "grad_norm": 0.30258945834332523,
+      "learning_rate": 4.418637757058701e-06,
+      "loss": 0.4276,
+      "step": 15546
+    },
+    {
+      "epoch": 2.5532609365056556,
+      "grad_norm": 0.3330461796000069,
+      "learning_rate": 4.418168861963119e-06,
+      "loss": 0.4454,
+      "step": 15547
+    },
+    {
+      "epoch": 2.553425163713998,
+      "grad_norm": 0.3405066873911021,
+      "learning_rate": 4.4176999677563035e-06,
+      "loss": 0.4283,
+      "step": 15548
+    },
+    {
+      "epoch": 2.5535893909223413,
+      "grad_norm": 0.39288368153964764,
+      "learning_rate": 4.417231074443347e-06,
+      "loss": 0.4311,
+      "step": 15549
+    },
+    {
+      "epoch": 2.553753618130684,
+      "grad_norm": 0.37034601558345265,
+      "learning_rate": 4.416762182029344e-06,
+      "loss": 0.432,
+      "step": 15550
+    },
+    {
+      "epoch": 2.5539178453390265,
+      "grad_norm": 0.38462584820097667,
+      "learning_rate": 4.416293290519385e-06,
+      "loss": 0.452,
+      "step": 15551
+    },
+    {
+      "epoch": 2.554082072547369,
+      "grad_norm": 0.42051310281220755,
+      "learning_rate": 4.415824399918563e-06,
+      "loss": 0.435,
+      "step": 15552
+    },
+    {
+      "epoch": 2.554246299755712,
+      "grad_norm": 0.3166037955740742,
+      "learning_rate": 4.4153555102319725e-06,
+      "loss": 0.413,
+      "step": 15553
+    },
+    {
+      "epoch": 2.554410526964055,
+      "grad_norm": 0.31043044492793076,
+      "learning_rate": 4.414886621464702e-06,
+      "loss": 0.4347,
+      "step": 15554
+    },
+    {
+      "epoch": 2.5545747541723975,
+      "grad_norm": 0.34700798655471937,
+      "learning_rate": 4.414417733621847e-06,
+      "loss": 0.4421,
+      "step": 15555
+    },
+    {
+      "epoch": 2.55473898138074,
+      "grad_norm": 0.2892813037690508,
+      "learning_rate": 4.4139488467085004e-06,
+      "loss": 0.4446,
+      "step": 15556
+    },
+    {
+      "epoch": 2.5549032085890833,
+      "grad_norm": 0.3175899138027976,
+      "learning_rate": 4.413479960729754e-06,
+      "loss": 0.4314,
+      "step": 15557
+    },
+    {
+      "epoch": 2.555067435797426,
+      "grad_norm": 0.3149807022034962,
+      "learning_rate": 4.413011075690699e-06,
+      "loss": 0.4423,
+      "step": 15558
+    },
+    {
+      "epoch": 2.5552316630057685,
+      "grad_norm": 0.5347116578772787,
+      "learning_rate": 4.41254219159643e-06,
+      "loss": 0.4459,
+      "step": 15559
+    },
+    {
+      "epoch": 2.555395890214111,
+      "grad_norm": 0.37843617764625054,
+      "learning_rate": 4.41207330845204e-06,
+      "loss": 0.4341,
+      "step": 15560
+    },
+    {
+      "epoch": 2.555560117422454,
+      "grad_norm": 0.2762679102433087,
+      "learning_rate": 4.411604426262621e-06,
+      "loss": 0.4058,
+      "step": 15561
+    },
+    {
+      "epoch": 2.5557243446307965,
+      "grad_norm": 0.30060861946870354,
+      "learning_rate": 4.411135545033263e-06,
+      "loss": 0.4443,
+      "step": 15562
+    },
+    {
+      "epoch": 2.5558885718391395,
+      "grad_norm": 0.3800865842812888,
+      "learning_rate": 4.41066666476906e-06,
+      "loss": 0.43,
+      "step": 15563
+    },
+    {
+      "epoch": 2.556052799047482,
+      "grad_norm": 0.25946722154162527,
+      "learning_rate": 4.4101977854751055e-06,
+      "loss": 0.4419,
+      "step": 15564
+    },
+    {
+      "epoch": 2.556217026255825,
+      "grad_norm": 0.6759836164016616,
+      "learning_rate": 4.409728907156493e-06,
+      "loss": 0.4274,
+      "step": 15565
+    },
+    {
+      "epoch": 2.556381253464168,
+      "grad_norm": 0.6234744045352152,
+      "learning_rate": 4.409260029818312e-06,
+      "loss": 0.4346,
+      "step": 15566
+    },
+    {
+      "epoch": 2.5565454806725105,
+      "grad_norm": 0.38581918418779537,
+      "learning_rate": 4.408791153465655e-06,
+      "loss": 0.4232,
+      "step": 15567
+    },
+    {
+      "epoch": 2.556709707880853,
+      "grad_norm": 0.29573823645262287,
+      "learning_rate": 4.408322278103617e-06,
+      "loss": 0.4193,
+      "step": 15568
+    },
+    {
+      "epoch": 2.556873935089196,
+      "grad_norm": 0.3016718414824482,
+      "learning_rate": 4.407853403737291e-06,
+      "loss": 0.4545,
+      "step": 15569
+    },
+    {
+      "epoch": 2.5570381622975384,
+      "grad_norm": 0.3891485274119814,
+      "learning_rate": 4.407384530371766e-06,
+      "loss": 0.4365,
+      "step": 15570
+    },
+    {
+      "epoch": 2.5572023895058815,
+      "grad_norm": 0.2532739379279318,
+      "learning_rate": 4.406915658012136e-06,
+      "loss": 0.4317,
+      "step": 15571
+    },
+    {
+      "epoch": 2.557366616714224,
+      "grad_norm": 0.29871548761299727,
+      "learning_rate": 4.406446786663494e-06,
+      "loss": 0.4403,
+      "step": 15572
+    },
+    {
+      "epoch": 2.557530843922567,
+      "grad_norm": 0.35194308191729523,
+      "learning_rate": 4.405977916330931e-06,
+      "loss": 0.4441,
+      "step": 15573
+    },
+    {
+      "epoch": 2.55769507113091,
+      "grad_norm": 0.3031216695491342,
+      "learning_rate": 4.405509047019541e-06,
+      "loss": 0.4154,
+      "step": 15574
+    },
+    {
+      "epoch": 2.5578592983392525,
+      "grad_norm": 0.31930845237851674,
+      "learning_rate": 4.4050401787344165e-06,
+      "loss": 0.4454,
+      "step": 15575
+    },
+    {
+      "epoch": 2.558023525547595,
+      "grad_norm": 0.33002408879010403,
+      "learning_rate": 4.404571311480648e-06,
+      "loss": 0.4434,
+      "step": 15576
+    },
+    {
+      "epoch": 2.558187752755938,
+      "grad_norm": 0.30275352248274623,
+      "learning_rate": 4.404102445263329e-06,
+      "loss": 0.4509,
+      "step": 15577
+    },
+    {
+      "epoch": 2.5583519799642804,
+      "grad_norm": 0.35335386445372424,
+      "learning_rate": 4.4036335800875535e-06,
+      "loss": 0.448,
+      "step": 15578
+    },
+    {
+      "epoch": 2.558516207172623,
+      "grad_norm": 0.2800989999633153,
+      "learning_rate": 4.403164715958411e-06,
+      "loss": 0.425,
+      "step": 15579
+    },
+    {
+      "epoch": 2.558680434380966,
+      "grad_norm": 0.3127782022895306,
+      "learning_rate": 4.402695852880995e-06,
+      "loss": 0.4137,
+      "step": 15580
+    },
+    {
+      "epoch": 2.558844661589309,
+      "grad_norm": 0.3799146716087815,
+      "learning_rate": 4.402226990860397e-06,
+      "loss": 0.4266,
+      "step": 15581
+    },
+    {
+      "epoch": 2.5590088887976514,
+      "grad_norm": 0.3673831939441783,
+      "learning_rate": 4.401758129901711e-06,
+      "loss": 0.4064,
+      "step": 15582
+    },
+    {
+      "epoch": 2.5591731160059945,
+      "grad_norm": 0.3376274379342438,
+      "learning_rate": 4.401289270010027e-06,
+      "loss": 0.4233,
+      "step": 15583
+    },
+    {
+      "epoch": 2.559337343214337,
+      "grad_norm": 0.3998414262198734,
+      "learning_rate": 4.40082041119044e-06,
+      "loss": 0.4275,
+      "step": 15584
+    },
+    {
+      "epoch": 2.55950157042268,
+      "grad_norm": 0.3520853575209431,
+      "learning_rate": 4.40035155344804e-06,
+      "loss": 0.4616,
+      "step": 15585
+    },
+    {
+      "epoch": 2.5596657976310224,
+      "grad_norm": 0.37561224309961333,
+      "learning_rate": 4.399882696787922e-06,
+      "loss": 0.4394,
+      "step": 15586
+    },
+    {
+      "epoch": 2.559830024839365,
+      "grad_norm": 0.36754339045918716,
+      "learning_rate": 4.399413841215175e-06,
+      "loss": 0.4575,
+      "step": 15587
+    },
+    {
+      "epoch": 2.559994252047708,
+      "grad_norm": 0.33783166500512124,
+      "learning_rate": 4.398944986734892e-06,
+      "loss": 0.4164,
+      "step": 15588
+    },
+    {
+      "epoch": 2.560158479256051,
+      "grad_norm": 0.31452477254199035,
+      "learning_rate": 4.398476133352167e-06,
+      "loss": 0.4528,
+      "step": 15589
+    },
+    {
+      "epoch": 2.5603227064643934,
+      "grad_norm": 0.2787399568684325,
+      "learning_rate": 4.398007281072091e-06,
+      "loss": 0.4209,
+      "step": 15590
+    },
+    {
+      "epoch": 2.5604869336727365,
+      "grad_norm": 0.2767623218217474,
+      "learning_rate": 4.397538429899756e-06,
+      "loss": 0.414,
+      "step": 15591
+    },
+    {
+      "epoch": 2.560651160881079,
+      "grad_norm": 0.28888844049323786,
+      "learning_rate": 4.397069579840253e-06,
+      "loss": 0.4542,
+      "step": 15592
+    },
+    {
+      "epoch": 2.5608153880894218,
+      "grad_norm": 0.5864957714875929,
+      "learning_rate": 4.396600730898677e-06,
+      "loss": 0.4243,
+      "step": 15593
+    },
+    {
+      "epoch": 2.5609796152977644,
+      "grad_norm": 0.38962154546595434,
+      "learning_rate": 4.396131883080121e-06,
+      "loss": 0.4387,
+      "step": 15594
+    },
+    {
+      "epoch": 2.561143842506107,
+      "grad_norm": 0.27331680844407696,
+      "learning_rate": 4.395663036389673e-06,
+      "loss": 0.4464,
+      "step": 15595
+    },
+    {
+      "epoch": 2.5613080697144497,
+      "grad_norm": 0.3069416039408074,
+      "learning_rate": 4.395194190832426e-06,
+      "loss": 0.4388,
+      "step": 15596
+    },
+    {
+      "epoch": 2.5614722969227928,
+      "grad_norm": 0.3252988244608723,
+      "learning_rate": 4.394725346413474e-06,
+      "loss": 0.4312,
+      "step": 15597
+    },
+    {
+      "epoch": 2.5616365241311354,
+      "grad_norm": 0.556385918534084,
+      "learning_rate": 4.394256503137908e-06,
+      "loss": 0.4547,
+      "step": 15598
+    },
+    {
+      "epoch": 2.561800751339478,
+      "grad_norm": 0.3021824955470384,
+      "learning_rate": 4.393787661010821e-06,
+      "loss": 0.4296,
+      "step": 15599
+    },
+    {
+      "epoch": 2.561964978547821,
+      "grad_norm": 0.2910937592004871,
+      "learning_rate": 4.393318820037304e-06,
+      "loss": 0.4649,
+      "step": 15600
+    },
+    {
+      "epoch": 2.5621292057561638,
+      "grad_norm": 0.31504368710656216,
+      "learning_rate": 4.39284998022245e-06,
+      "loss": 0.4244,
+      "step": 15601
+    },
+    {
+      "epoch": 2.5622934329645064,
+      "grad_norm": 0.3230087906700632,
+      "learning_rate": 4.39238114157135e-06,
+      "loss": 0.4366,
+      "step": 15602
+    },
+    {
+      "epoch": 2.562457660172849,
+      "grad_norm": 0.33819297778333945,
+      "learning_rate": 4.3919123040890985e-06,
+      "loss": 0.4313,
+      "step": 15603
+    },
+    {
+      "epoch": 2.5626218873811917,
+      "grad_norm": 0.28026153833271106,
+      "learning_rate": 4.391443467780784e-06,
+      "loss": 0.4233,
+      "step": 15604
+    },
+    {
+      "epoch": 2.5627861145895348,
+      "grad_norm": 0.313441369830849,
+      "learning_rate": 4.390974632651502e-06,
+      "loss": 0.4295,
+      "step": 15605
+    },
+    {
+      "epoch": 2.5629503417978774,
+      "grad_norm": 0.8718709717848913,
+      "learning_rate": 4.3905057987063406e-06,
+      "loss": 0.4238,
+      "step": 15606
+    },
+    {
+      "epoch": 2.56311456900622,
+      "grad_norm": 0.26352651969380825,
+      "learning_rate": 4.390036965950394e-06,
+      "loss": 0.4334,
+      "step": 15607
+    },
+    {
+      "epoch": 2.563278796214563,
+      "grad_norm": 0.2994463931598903,
+      "learning_rate": 4.389568134388754e-06,
+      "loss": 0.4317,
+      "step": 15608
+    },
+    {
+      "epoch": 2.5634430234229058,
+      "grad_norm": 0.3646830270622841,
+      "learning_rate": 4.389099304026515e-06,
+      "loss": 0.4337,
+      "step": 15609
+    },
+    {
+      "epoch": 2.5636072506312484,
+      "grad_norm": 0.5393460161367439,
+      "learning_rate": 4.3886304748687644e-06,
+      "loss": 0.4356,
+      "step": 15610
+    },
+    {
+      "epoch": 2.563771477839591,
+      "grad_norm": 0.30025939172178034,
+      "learning_rate": 4.388161646920599e-06,
+      "loss": 0.4251,
+      "step": 15611
+    },
+    {
+      "epoch": 2.5639357050479337,
+      "grad_norm": 0.31959158175100244,
+      "learning_rate": 4.387692820187106e-06,
+      "loss": 0.436,
+      "step": 15612
+    },
+    {
+      "epoch": 2.5640999322562763,
+      "grad_norm": 0.29317675299532536,
+      "learning_rate": 4.38722399467338e-06,
+      "loss": 0.4383,
+      "step": 15613
+    },
+    {
+      "epoch": 2.5642641594646194,
+      "grad_norm": 0.34975265052513155,
+      "learning_rate": 4.386755170384513e-06,
+      "loss": 0.4398,
+      "step": 15614
+    },
+    {
+      "epoch": 2.564428386672962,
+      "grad_norm": 0.3412204875657687,
+      "learning_rate": 4.386286347325595e-06,
+      "loss": 0.4538,
+      "step": 15615
+    },
+    {
+      "epoch": 2.5645926138813047,
+      "grad_norm": 0.31861125143751867,
+      "learning_rate": 4.385817525501719e-06,
+      "loss": 0.4384,
+      "step": 15616
+    },
+    {
+      "epoch": 2.5647568410896477,
+      "grad_norm": 0.2895893168578073,
+      "learning_rate": 4.385348704917978e-06,
+      "loss": 0.4292,
+      "step": 15617
+    },
+    {
+      "epoch": 2.5649210682979904,
+      "grad_norm": 0.3212807387575279,
+      "learning_rate": 4.384879885579462e-06,
+      "loss": 0.4217,
+      "step": 15618
+    },
+    {
+      "epoch": 2.565085295506333,
+      "grad_norm": 0.37869276897825244,
+      "learning_rate": 4.384411067491265e-06,
+      "loss": 0.4226,
+      "step": 15619
+    },
+    {
+      "epoch": 2.5652495227146757,
+      "grad_norm": 0.32545970108276556,
+      "learning_rate": 4.383942250658478e-06,
+      "loss": 0.4182,
+      "step": 15620
+    },
+    {
+      "epoch": 2.5654137499230183,
+      "grad_norm": 0.5821317175586268,
+      "learning_rate": 4.383473435086191e-06,
+      "loss": 0.4307,
+      "step": 15621
+    },
+    {
+      "epoch": 2.5655779771313614,
+      "grad_norm": 0.3737727497282395,
+      "learning_rate": 4.383004620779497e-06,
+      "loss": 0.4254,
+      "step": 15622
+    },
+    {
+      "epoch": 2.565742204339704,
+      "grad_norm": 0.3252778528469744,
+      "learning_rate": 4.382535807743487e-06,
+      "loss": 0.4444,
+      "step": 15623
+    },
+    {
+      "epoch": 2.5659064315480467,
+      "grad_norm": 0.40545934497853325,
+      "learning_rate": 4.382066995983256e-06,
+      "loss": 0.4465,
+      "step": 15624
+    },
+    {
+      "epoch": 2.5660706587563897,
+      "grad_norm": 0.3786830883990562,
+      "learning_rate": 4.381598185503892e-06,
+      "loss": 0.4253,
+      "step": 15625
+    },
+    {
+      "epoch": 2.5662348859647324,
+      "grad_norm": 0.33683926885493226,
+      "learning_rate": 4.381129376310488e-06,
+      "loss": 0.4349,
+      "step": 15626
+    },
+    {
+      "epoch": 2.566399113173075,
+      "grad_norm": 0.9430359938687994,
+      "learning_rate": 4.380660568408136e-06,
+      "loss": 0.4134,
+      "step": 15627
+    },
+    {
+      "epoch": 2.5665633403814176,
+      "grad_norm": 0.4029380612559527,
+      "learning_rate": 4.38019176180193e-06,
+      "loss": 0.4342,
+      "step": 15628
+    },
+    {
+      "epoch": 2.5667275675897603,
+      "grad_norm": 0.3329457909755624,
+      "learning_rate": 4.379722956496958e-06,
+      "loss": 0.4387,
+      "step": 15629
+    },
+    {
+      "epoch": 2.566891794798103,
+      "grad_norm": 0.35796644332871863,
+      "learning_rate": 4.379254152498312e-06,
+      "loss": 0.4571,
+      "step": 15630
+    },
+    {
+      "epoch": 2.567056022006446,
+      "grad_norm": 0.34985740849790825,
+      "learning_rate": 4.378785349811085e-06,
+      "loss": 0.4408,
+      "step": 15631
+    },
+    {
+      "epoch": 2.5672202492147886,
+      "grad_norm": 0.3545240868647146,
+      "learning_rate": 4.378316548440369e-06,
+      "loss": 0.4361,
+      "step": 15632
+    },
+    {
+      "epoch": 2.5673844764231313,
+      "grad_norm": 0.297635939595374,
+      "learning_rate": 4.3778477483912545e-06,
+      "loss": 0.4233,
+      "step": 15633
+    },
+    {
+      "epoch": 2.5675487036314744,
+      "grad_norm": 0.2947778083461371,
+      "learning_rate": 4.377378949668833e-06,
+      "loss": 0.433,
+      "step": 15634
+    },
+    {
+      "epoch": 2.567712930839817,
+      "grad_norm": 0.395361891954945,
+      "learning_rate": 4.376910152278197e-06,
+      "loss": 0.4374,
+      "step": 15635
+    },
+    {
+      "epoch": 2.5678771580481596,
+      "grad_norm": 0.26747379087473455,
+      "learning_rate": 4.37644135622444e-06,
+      "loss": 0.4452,
+      "step": 15636
+    },
+    {
+      "epoch": 2.5680413852565023,
+      "grad_norm": 0.3537882378113586,
+      "learning_rate": 4.37597256151265e-06,
+      "loss": 0.4309,
+      "step": 15637
+    },
+    {
+      "epoch": 2.568205612464845,
+      "grad_norm": 0.3126543583966415,
+      "learning_rate": 4.375503768147918e-06,
+      "loss": 0.4475,
+      "step": 15638
+    },
+    {
+      "epoch": 2.568369839673188,
+      "grad_norm": 0.310994438653193,
+      "learning_rate": 4.375034976135341e-06,
+      "loss": 0.4532,
+      "step": 15639
+    },
+    {
+      "epoch": 2.5685340668815306,
+      "grad_norm": 0.2757312848098248,
+      "learning_rate": 4.374566185480005e-06,
+      "loss": 0.4254,
+      "step": 15640
+    },
+    {
+      "epoch": 2.5686982940898733,
+      "grad_norm": 0.27164222845895303,
+      "learning_rate": 4.374097396187003e-06,
+      "loss": 0.4184,
+      "step": 15641
+    },
+    {
+      "epoch": 2.5688625212982164,
+      "grad_norm": 0.27266040054382534,
+      "learning_rate": 4.373628608261428e-06,
+      "loss": 0.422,
+      "step": 15642
+    },
+    {
+      "epoch": 2.569026748506559,
+      "grad_norm": 0.29807414547875344,
+      "learning_rate": 4.373159821708372e-06,
+      "loss": 0.4379,
+      "step": 15643
+    },
+    {
+      "epoch": 2.5691909757149016,
+      "grad_norm": 0.4276170738492218,
+      "learning_rate": 4.372691036532923e-06,
+      "loss": 0.4514,
+      "step": 15644
+    },
+    {
+      "epoch": 2.5693552029232443,
+      "grad_norm": 0.30498004090935377,
+      "learning_rate": 4.372222252740177e-06,
+      "loss": 0.4365,
+      "step": 15645
+    },
+    {
+      "epoch": 2.569519430131587,
+      "grad_norm": 0.3839879430641964,
+      "learning_rate": 4.371753470335221e-06,
+      "loss": 0.4384,
+      "step": 15646
+    },
+    {
+      "epoch": 2.5696836573399295,
+      "grad_norm": 0.4011106422419418,
+      "learning_rate": 4.37128468932315e-06,
+      "loss": 0.4383,
+      "step": 15647
+    },
+    {
+      "epoch": 2.5698478845482726,
+      "grad_norm": 0.35855065427808513,
+      "learning_rate": 4.3708159097090536e-06,
+      "loss": 0.4398,
+      "step": 15648
+    },
+    {
+      "epoch": 2.5700121117566153,
+      "grad_norm": 0.3644533904290332,
+      "learning_rate": 4.370347131498022e-06,
+      "loss": 0.4257,
+      "step": 15649
+    },
+    {
+      "epoch": 2.570176338964958,
+      "grad_norm": 0.33253435890955046,
+      "learning_rate": 4.369878354695148e-06,
+      "loss": 0.4282,
+      "step": 15650
+    },
+    {
+      "epoch": 2.570340566173301,
+      "grad_norm": 0.26752244259463775,
+      "learning_rate": 4.369409579305525e-06,
+      "loss": 0.4401,
+      "step": 15651
+    },
+    {
+      "epoch": 2.5705047933816436,
+      "grad_norm": 0.27764725847093746,
+      "learning_rate": 4.368940805334241e-06,
+      "loss": 0.4348,
+      "step": 15652
+    },
+    {
+      "epoch": 2.5706690205899863,
+      "grad_norm": 0.46418971042458596,
+      "learning_rate": 4.3684720327863904e-06,
+      "loss": 0.411,
+      "step": 15653
+    },
+    {
+      "epoch": 2.570833247798329,
+      "grad_norm": 0.29686142198669646,
+      "learning_rate": 4.368003261667062e-06,
+      "loss": 0.4427,
+      "step": 15654
+    },
+    {
+      "epoch": 2.5709974750066715,
+      "grad_norm": 0.27340688193424023,
+      "learning_rate": 4.367534491981349e-06,
+      "loss": 0.4459,
+      "step": 15655
+    },
+    {
+      "epoch": 2.5711617022150146,
+      "grad_norm": 0.29452955602163794,
+      "learning_rate": 4.36706572373434e-06,
+      "loss": 0.4615,
+      "step": 15656
+    },
+    {
+      "epoch": 2.5713259294233572,
+      "grad_norm": 0.2902678707411273,
+      "learning_rate": 4.366596956931128e-06,
+      "loss": 0.4221,
+      "step": 15657
+    },
+    {
+      "epoch": 2.5714901566317,
+      "grad_norm": 0.47928414230720084,
+      "learning_rate": 4.366128191576806e-06,
+      "loss": 0.4417,
+      "step": 15658
+    },
+    {
+      "epoch": 2.571654383840043,
+      "grad_norm": 0.30012861208394687,
+      "learning_rate": 4.3656594276764616e-06,
+      "loss": 0.4254,
+      "step": 15659
+    },
+    {
+      "epoch": 2.5718186110483856,
+      "grad_norm": 0.3130155646681567,
+      "learning_rate": 4.365190665235189e-06,
+      "loss": 0.4422,
+      "step": 15660
+    },
+    {
+      "epoch": 2.5719828382567282,
+      "grad_norm": 0.3455326963780102,
+      "learning_rate": 4.36472190425808e-06,
+      "loss": 0.4318,
+      "step": 15661
+    },
+    {
+      "epoch": 2.572147065465071,
+      "grad_norm": 0.3041176120109691,
+      "learning_rate": 4.364253144750222e-06,
+      "loss": 0.4313,
+      "step": 15662
+    },
+    {
+      "epoch": 2.5723112926734135,
+      "grad_norm": 0.28055282368032514,
+      "learning_rate": 4.36378438671671e-06,
+      "loss": 0.4413,
+      "step": 15663
+    },
+    {
+      "epoch": 2.572475519881756,
+      "grad_norm": 0.5147635636054625,
+      "learning_rate": 4.363315630162632e-06,
+      "loss": 0.4397,
+      "step": 15664
+    },
+    {
+      "epoch": 2.5726397470900992,
+      "grad_norm": 0.2849554341120117,
+      "learning_rate": 4.362846875093081e-06,
+      "loss": 0.4344,
+      "step": 15665
+    },
+    {
+      "epoch": 2.572803974298442,
+      "grad_norm": 0.2846440291685015,
+      "learning_rate": 4.3623781215131475e-06,
+      "loss": 0.4277,
+      "step": 15666
+    },
+    {
+      "epoch": 2.5729682015067845,
+      "grad_norm": 0.3305141222386102,
+      "learning_rate": 4.3619093694279245e-06,
+      "loss": 0.4405,
+      "step": 15667
+    },
+    {
+      "epoch": 2.5731324287151276,
+      "grad_norm": 0.3596809632535296,
+      "learning_rate": 4.3614406188425005e-06,
+      "loss": 0.4677,
+      "step": 15668
+    },
+    {
+      "epoch": 2.5732966559234702,
+      "grad_norm": 0.4530597862843398,
+      "learning_rate": 4.360971869761968e-06,
+      "loss": 0.4387,
+      "step": 15669
+    },
+    {
+      "epoch": 2.573460883131813,
+      "grad_norm": 0.3251876593883486,
+      "learning_rate": 4.360503122191419e-06,
+      "loss": 0.4108,
+      "step": 15670
+    },
+    {
+      "epoch": 2.5736251103401555,
+      "grad_norm": 0.2803538758207167,
+      "learning_rate": 4.360034376135942e-06,
+      "loss": 0.4446,
+      "step": 15671
+    },
+    {
+      "epoch": 2.573789337548498,
+      "grad_norm": 0.28182246452077286,
+      "learning_rate": 4.3595656316006295e-06,
+      "loss": 0.4468,
+      "step": 15672
+    },
+    {
+      "epoch": 2.5739535647568412,
+      "grad_norm": 0.34577943342234163,
+      "learning_rate": 4.359096888590573e-06,
+      "loss": 0.4308,
+      "step": 15673
+    },
+    {
+      "epoch": 2.574117791965184,
+      "grad_norm": 0.38565717225066504,
+      "learning_rate": 4.358628147110862e-06,
+      "loss": 0.432,
+      "step": 15674
+    },
+    {
+      "epoch": 2.5742820191735265,
+      "grad_norm": 0.30459225924397265,
+      "learning_rate": 4.358159407166588e-06,
+      "loss": 0.4202,
+      "step": 15675
+    },
+    {
+      "epoch": 2.5744462463818696,
+      "grad_norm": 0.3196455829723776,
+      "learning_rate": 4.357690668762844e-06,
+      "loss": 0.4338,
+      "step": 15676
+    },
+    {
+      "epoch": 2.5746104735902122,
+      "grad_norm": 0.32731457911173506,
+      "learning_rate": 4.357221931904718e-06,
+      "loss": 0.4271,
+      "step": 15677
+    },
+    {
+      "epoch": 2.574774700798555,
+      "grad_norm": 0.3219906580343855,
+      "learning_rate": 4.356753196597304e-06,
+      "loss": 0.4437,
+      "step": 15678
+    },
+    {
+      "epoch": 2.5749389280068975,
+      "grad_norm": 0.29879707891187424,
+      "learning_rate": 4.35628446284569e-06,
+      "loss": 0.4442,
+      "step": 15679
+    },
+    {
+      "epoch": 2.57510315521524,
+      "grad_norm": 0.3494962037579369,
+      "learning_rate": 4.355815730654968e-06,
+      "loss": 0.4262,
+      "step": 15680
+    },
+    {
+      "epoch": 2.5752673824235828,
+      "grad_norm": 0.25523753461883125,
+      "learning_rate": 4.355347000030229e-06,
+      "loss": 0.426,
+      "step": 15681
+    },
+    {
+      "epoch": 2.575431609631926,
+      "grad_norm": 0.45019078924757433,
+      "learning_rate": 4.354878270976564e-06,
+      "loss": 0.441,
+      "step": 15682
+    },
+    {
+      "epoch": 2.5755958368402685,
+      "grad_norm": 0.32355129633184126,
+      "learning_rate": 4.354409543499064e-06,
+      "loss": 0.4299,
+      "step": 15683
+    },
+    {
+      "epoch": 2.575760064048611,
+      "grad_norm": 0.3285679353712852,
+      "learning_rate": 4.353940817602819e-06,
+      "loss": 0.4186,
+      "step": 15684
+    },
+    {
+      "epoch": 2.575924291256954,
+      "grad_norm": 0.27666075631853193,
+      "learning_rate": 4.35347209329292e-06,
+      "loss": 0.4338,
+      "step": 15685
+    },
+    {
+      "epoch": 2.576088518465297,
+      "grad_norm": 0.49437240302558794,
+      "learning_rate": 4.353003370574461e-06,
+      "loss": 0.4409,
+      "step": 15686
+    },
+    {
+      "epoch": 2.5762527456736395,
+      "grad_norm": 0.30204622384409346,
+      "learning_rate": 4.352534649452527e-06,
+      "loss": 0.4422,
+      "step": 15687
+    },
+    {
+      "epoch": 2.576416972881982,
+      "grad_norm": 0.273434931240618,
+      "learning_rate": 4.352065929932215e-06,
+      "loss": 0.4177,
+      "step": 15688
+    },
+    {
+      "epoch": 2.5765812000903248,
+      "grad_norm": 0.35155939309416256,
+      "learning_rate": 4.35159721201861e-06,
+      "loss": 0.4432,
+      "step": 15689
+    },
+    {
+      "epoch": 2.576745427298668,
+      "grad_norm": 0.3920495165730994,
+      "learning_rate": 4.351128495716805e-06,
+      "loss": 0.4235,
+      "step": 15690
+    },
+    {
+      "epoch": 2.5769096545070105,
+      "grad_norm": 0.30913151759937846,
+      "learning_rate": 4.350659781031891e-06,
+      "loss": 0.4201,
+      "step": 15691
+    },
+    {
+      "epoch": 2.577073881715353,
+      "grad_norm": 0.3076565225368957,
+      "learning_rate": 4.35019106796896e-06,
+      "loss": 0.4235,
+      "step": 15692
+    },
+    {
+      "epoch": 2.577238108923696,
+      "grad_norm": 0.29655321776974725,
+      "learning_rate": 4.349722356533101e-06,
+      "loss": 0.4459,
+      "step": 15693
+    },
+    {
+      "epoch": 2.577402336132039,
+      "grad_norm": 0.28892151775120634,
+      "learning_rate": 4.3492536467294044e-06,
+      "loss": 0.4179,
+      "step": 15694
+    },
+    {
+      "epoch": 2.5775665633403815,
+      "grad_norm": 0.41459237885226863,
+      "learning_rate": 4.3487849385629635e-06,
+      "loss": 0.4146,
+      "step": 15695
+    },
+    {
+      "epoch": 2.577730790548724,
+      "grad_norm": 0.2902977896042206,
+      "learning_rate": 4.348316232038865e-06,
+      "loss": 0.437,
+      "step": 15696
+    },
+    {
+      "epoch": 2.5778950177570668,
+      "grad_norm": 0.4496536761679526,
+      "learning_rate": 4.347847527162203e-06,
+      "loss": 0.4415,
+      "step": 15697
+    },
+    {
+      "epoch": 2.5780592449654094,
+      "grad_norm": 0.36641641828421784,
+      "learning_rate": 4.3473788239380645e-06,
+      "loss": 0.4223,
+      "step": 15698
+    },
+    {
+      "epoch": 2.5782234721737525,
+      "grad_norm": 0.36549153174167476,
+      "learning_rate": 4.3469101223715425e-06,
+      "loss": 0.4306,
+      "step": 15699
+    },
+    {
+      "epoch": 2.578387699382095,
+      "grad_norm": 0.3015497487218498,
+      "learning_rate": 4.3464414224677275e-06,
+      "loss": 0.419,
+      "step": 15700
+    },
+    {
+      "epoch": 2.5785519265904377,
+      "grad_norm": 0.27625426522960334,
+      "learning_rate": 4.345972724231711e-06,
+      "loss": 0.4503,
+      "step": 15701
+    },
+    {
+      "epoch": 2.578716153798781,
+      "grad_norm": 0.30417803932476767,
+      "learning_rate": 4.3455040276685805e-06,
+      "loss": 0.4298,
+      "step": 15702
+    },
+    {
+      "epoch": 2.5788803810071235,
+      "grad_norm": 0.31044685598214067,
+      "learning_rate": 4.345035332783431e-06,
+      "loss": 0.4572,
+      "step": 15703
+    },
+    {
+      "epoch": 2.579044608215466,
+      "grad_norm": 0.34029110494135095,
+      "learning_rate": 4.344566639581348e-06,
+      "loss": 0.4286,
+      "step": 15704
+    },
+    {
+      "epoch": 2.5792088354238087,
+      "grad_norm": 0.4220228863169494,
+      "learning_rate": 4.344097948067424e-06,
+      "loss": 0.4313,
+      "step": 15705
+    },
+    {
+      "epoch": 2.5793730626321514,
+      "grad_norm": 0.29592190663583623,
+      "learning_rate": 4.34362925824675e-06,
+      "loss": 0.4506,
+      "step": 15706
+    },
+    {
+      "epoch": 2.5795372898404945,
+      "grad_norm": 0.29158058048698404,
+      "learning_rate": 4.343160570124417e-06,
+      "loss": 0.4402,
+      "step": 15707
+    },
+    {
+      "epoch": 2.579701517048837,
+      "grad_norm": 0.3708781139752633,
+      "learning_rate": 4.3426918837055135e-06,
+      "loss": 0.4506,
+      "step": 15708
+    },
+    {
+      "epoch": 2.5798657442571797,
+      "grad_norm": 0.3318455295905244,
+      "learning_rate": 4.342223198995131e-06,
+      "loss": 0.4468,
+      "step": 15709
+    },
+    {
+      "epoch": 2.580029971465523,
+      "grad_norm": 0.4317230371408357,
+      "learning_rate": 4.34175451599836e-06,
+      "loss": 0.4466,
+      "step": 15710
+    },
+    {
+      "epoch": 2.5801941986738655,
+      "grad_norm": 0.3369867194655041,
+      "learning_rate": 4.341285834720292e-06,
+      "loss": 0.4115,
+      "step": 15711
+    },
+    {
+      "epoch": 2.580358425882208,
+      "grad_norm": 0.31912585151917744,
+      "learning_rate": 4.340817155166015e-06,
+      "loss": 0.4482,
+      "step": 15712
+    },
+    {
+      "epoch": 2.5805226530905507,
+      "grad_norm": 0.25662968380350065,
+      "learning_rate": 4.340348477340619e-06,
+      "loss": 0.4425,
+      "step": 15713
+    },
+    {
+      "epoch": 2.5806868802988934,
+      "grad_norm": 0.3063092669941935,
+      "learning_rate": 4.339879801249197e-06,
+      "loss": 0.4489,
+      "step": 15714
+    },
+    {
+      "epoch": 2.580851107507236,
+      "grad_norm": 0.2630713688895437,
+      "learning_rate": 4.339411126896836e-06,
+      "loss": 0.4542,
+      "step": 15715
+    },
+    {
+      "epoch": 2.581015334715579,
+      "grad_norm": 0.31208575964151564,
+      "learning_rate": 4.338942454288631e-06,
+      "loss": 0.4318,
+      "step": 15716
+    },
+    {
+      "epoch": 2.5811795619239217,
+      "grad_norm": 0.3594129107656559,
+      "learning_rate": 4.338473783429668e-06,
+      "loss": 0.4353,
+      "step": 15717
+    },
+    {
+      "epoch": 2.5813437891322644,
+      "grad_norm": 0.3693658716118475,
+      "learning_rate": 4.338005114325038e-06,
+      "loss": 0.4337,
+      "step": 15718
+    },
+    {
+      "epoch": 2.5815080163406074,
+      "grad_norm": 0.40035589840951313,
+      "learning_rate": 4.3375364469798315e-06,
+      "loss": 0.4231,
+      "step": 15719
+    },
+    {
+      "epoch": 2.58167224354895,
+      "grad_norm": 0.2942822598845297,
+      "learning_rate": 4.3370677813991425e-06,
+      "loss": 0.4365,
+      "step": 15720
+    },
+    {
+      "epoch": 2.5818364707572927,
+      "grad_norm": 0.2598983203441498,
+      "learning_rate": 4.3365991175880545e-06,
+      "loss": 0.43,
+      "step": 15721
+    },
+    {
+      "epoch": 2.5820006979656354,
+      "grad_norm": 0.3306450186865602,
+      "learning_rate": 4.336130455551662e-06,
+      "loss": 0.4488,
+      "step": 15722
+    },
+    {
+      "epoch": 2.582164925173978,
+      "grad_norm": 0.4172832201505027,
+      "learning_rate": 4.335661795295053e-06,
+      "loss": 0.4493,
+      "step": 15723
+    },
+    {
+      "epoch": 2.582329152382321,
+      "grad_norm": 0.4443225994832997,
+      "learning_rate": 4.3351931368233195e-06,
+      "loss": 0.4485,
+      "step": 15724
+    },
+    {
+      "epoch": 2.5824933795906637,
+      "grad_norm": 0.36022320421626164,
+      "learning_rate": 4.334724480141551e-06,
+      "loss": 0.4535,
+      "step": 15725
+    },
+    {
+      "epoch": 2.5826576067990064,
+      "grad_norm": 0.28358389028441683,
+      "learning_rate": 4.334255825254836e-06,
+      "loss": 0.4391,
+      "step": 15726
+    },
+    {
+      "epoch": 2.5828218340073494,
+      "grad_norm": 0.33976043099099573,
+      "learning_rate": 4.333787172168266e-06,
+      "loss": 0.4492,
+      "step": 15727
+    },
+    {
+      "epoch": 2.582986061215692,
+      "grad_norm": 0.2803123022258049,
+      "learning_rate": 4.333318520886932e-06,
+      "loss": 0.4182,
+      "step": 15728
+    },
+    {
+      "epoch": 2.5831502884240347,
+      "grad_norm": 0.2642610620741509,
+      "learning_rate": 4.332849871415922e-06,
+      "loss": 0.4329,
+      "step": 15729
+    },
+    {
+      "epoch": 2.5833145156323774,
+      "grad_norm": 0.2714676626725195,
+      "learning_rate": 4.332381223760327e-06,
+      "loss": 0.4358,
+      "step": 15730
+    },
+    {
+      "epoch": 2.58347874284072,
+      "grad_norm": 0.29556803031452,
+      "learning_rate": 4.331912577925237e-06,
+      "loss": 0.4179,
+      "step": 15731
+    },
+    {
+      "epoch": 2.5836429700490626,
+      "grad_norm": 0.3645808737464659,
+      "learning_rate": 4.3314439339157415e-06,
+      "loss": 0.4348,
+      "step": 15732
+    },
+    {
+      "epoch": 2.5838071972574057,
+      "grad_norm": 0.3772393709280388,
+      "learning_rate": 4.3309752917369305e-06,
+      "loss": 0.4406,
+      "step": 15733
+    },
+    {
+      "epoch": 2.5839714244657483,
+      "grad_norm": 0.27548331977469753,
+      "learning_rate": 4.330506651393894e-06,
+      "loss": 0.4471,
+      "step": 15734
+    },
+    {
+      "epoch": 2.584135651674091,
+      "grad_norm": 0.2826736816359112,
+      "learning_rate": 4.330038012891723e-06,
+      "loss": 0.4483,
+      "step": 15735
+    },
+    {
+      "epoch": 2.584299878882434,
+      "grad_norm": 0.4133515073580495,
+      "learning_rate": 4.329569376235506e-06,
+      "loss": 0.4419,
+      "step": 15736
+    },
+    {
+      "epoch": 2.5844641060907767,
+      "grad_norm": 0.3225643113998406,
+      "learning_rate": 4.329100741430334e-06,
+      "loss": 0.4471,
+      "step": 15737
+    },
+    {
+      "epoch": 2.5846283332991193,
+      "grad_norm": 0.3407019919418374,
+      "learning_rate": 4.3286321084812955e-06,
+      "loss": 0.4533,
+      "step": 15738
+    },
+    {
+      "epoch": 2.584792560507462,
+      "grad_norm": 0.3428604509352793,
+      "learning_rate": 4.32816347739348e-06,
+      "loss": 0.4755,
+      "step": 15739
+    },
+    {
+      "epoch": 2.5849567877158046,
+      "grad_norm": 0.32503432507541574,
+      "learning_rate": 4.327694848171979e-06,
+      "loss": 0.4341,
+      "step": 15740
+    },
+    {
+      "epoch": 2.5851210149241477,
+      "grad_norm": 0.36169010504914045,
+      "learning_rate": 4.327226220821881e-06,
+      "loss": 0.4603,
+      "step": 15741
+    },
+    {
+      "epoch": 2.5852852421324903,
+      "grad_norm": 0.2994689117279142,
+      "learning_rate": 4.326757595348276e-06,
+      "loss": 0.4367,
+      "step": 15742
+    },
+    {
+      "epoch": 2.585449469340833,
+      "grad_norm": 0.313911921117997,
+      "learning_rate": 4.326288971756254e-06,
+      "loss": 0.4324,
+      "step": 15743
+    },
+    {
+      "epoch": 2.585613696549176,
+      "grad_norm": 0.2735236587767174,
+      "learning_rate": 4.3258203500509055e-06,
+      "loss": 0.4339,
+      "step": 15744
+    },
+    {
+      "epoch": 2.5857779237575187,
+      "grad_norm": 0.3583785949380112,
+      "learning_rate": 4.32535173023732e-06,
+      "loss": 0.4228,
+      "step": 15745
+    },
+    {
+      "epoch": 2.5859421509658613,
+      "grad_norm": 0.3579906842061974,
+      "learning_rate": 4.324883112320586e-06,
+      "loss": 0.4374,
+      "step": 15746
+    },
+    {
+      "epoch": 2.586106378174204,
+      "grad_norm": 0.2920349175148648,
+      "learning_rate": 4.324414496305793e-06,
+      "loss": 0.448,
+      "step": 15747
+    },
+    {
+      "epoch": 2.5862706053825466,
+      "grad_norm": 0.3551419529001769,
+      "learning_rate": 4.323945882198031e-06,
+      "loss": 0.4296,
+      "step": 15748
+    },
+    {
+      "epoch": 2.5864348325908892,
+      "grad_norm": 0.38050189174987475,
+      "learning_rate": 4.3234772700023904e-06,
+      "loss": 0.4367,
+      "step": 15749
+    },
+    {
+      "epoch": 2.5865990597992323,
+      "grad_norm": 0.27693520786110415,
+      "learning_rate": 4.323008659723961e-06,
+      "loss": 0.4273,
+      "step": 15750
+    },
+    {
+      "epoch": 2.586763287007575,
+      "grad_norm": 0.2946787998493088,
+      "learning_rate": 4.32254005136783e-06,
+      "loss": 0.4506,
+      "step": 15751
+    },
+    {
+      "epoch": 2.5869275142159176,
+      "grad_norm": 0.300489672148383,
+      "learning_rate": 4.32207144493909e-06,
+      "loss": 0.4254,
+      "step": 15752
+    },
+    {
+      "epoch": 2.5870917414242607,
+      "grad_norm": 0.33785313952343987,
+      "learning_rate": 4.32160284044283e-06,
+      "loss": 0.4356,
+      "step": 15753
+    },
+    {
+      "epoch": 2.5872559686326033,
+      "grad_norm": 0.517376636026884,
+      "learning_rate": 4.321134237884138e-06,
+      "loss": 0.4211,
+      "step": 15754
+    },
+    {
+      "epoch": 2.587420195840946,
+      "grad_norm": 0.2898400184440624,
+      "learning_rate": 4.320665637268103e-06,
+      "loss": 0.4221,
+      "step": 15755
+    },
+    {
+      "epoch": 2.5875844230492886,
+      "grad_norm": 0.42197824385946514,
+      "learning_rate": 4.3201970385998164e-06,
+      "loss": 0.4449,
+      "step": 15756
+    },
+    {
+      "epoch": 2.5877486502576312,
+      "grad_norm": 0.31677534657011097,
+      "learning_rate": 4.319728441884366e-06,
+      "loss": 0.4403,
+      "step": 15757
+    },
+    {
+      "epoch": 2.5879128774659743,
+      "grad_norm": 0.2832694609001604,
+      "learning_rate": 4.319259847126843e-06,
+      "loss": 0.4378,
+      "step": 15758
+    },
+    {
+      "epoch": 2.588077104674317,
+      "grad_norm": 0.39187976296966304,
+      "learning_rate": 4.318791254332337e-06,
+      "loss": 0.4537,
+      "step": 15759
+    },
+    {
+      "epoch": 2.5882413318826596,
+      "grad_norm": 0.296304090284739,
+      "learning_rate": 4.318322663505934e-06,
+      "loss": 0.4382,
+      "step": 15760
+    },
+    {
+      "epoch": 2.5884055590910027,
+      "grad_norm": 0.2989425161690862,
+      "learning_rate": 4.317854074652727e-06,
+      "loss": 0.4451,
+      "step": 15761
+    },
+    {
+      "epoch": 2.5885697862993453,
+      "grad_norm": 0.32956500842867725,
+      "learning_rate": 4.317385487777805e-06,
+      "loss": 0.4252,
+      "step": 15762
+    },
+    {
+      "epoch": 2.588734013507688,
+      "grad_norm": 0.3554805266347347,
+      "learning_rate": 4.316916902886255e-06,
+      "loss": 0.4785,
+      "step": 15763
+    },
+    {
+      "epoch": 2.5888982407160306,
+      "grad_norm": 0.29784661781396565,
+      "learning_rate": 4.316448319983166e-06,
+      "loss": 0.449,
+      "step": 15764
+    },
+    {
+      "epoch": 2.5890624679243732,
+      "grad_norm": 0.3403796996552582,
+      "learning_rate": 4.315979739073631e-06,
+      "loss": 0.42,
+      "step": 15765
+    },
+    {
+      "epoch": 2.589226695132716,
+      "grad_norm": 0.2820204314916899,
+      "learning_rate": 4.315511160162736e-06,
+      "loss": 0.42,
+      "step": 15766
+    },
+    {
+      "epoch": 2.589390922341059,
+      "grad_norm": 0.34694290008039924,
+      "learning_rate": 4.315042583255571e-06,
+      "loss": 0.4405,
+      "step": 15767
+    },
+    {
+      "epoch": 2.5895551495494016,
+      "grad_norm": 0.285712751776153,
+      "learning_rate": 4.314574008357227e-06,
+      "loss": 0.4387,
+      "step": 15768
+    },
+    {
+      "epoch": 2.589719376757744,
+      "grad_norm": 0.2886681436122029,
+      "learning_rate": 4.31410543547279e-06,
+      "loss": 0.4346,
+      "step": 15769
+    },
+    {
+      "epoch": 2.5898836039660873,
+      "grad_norm": 0.26426338895128476,
+      "learning_rate": 4.3136368646073535e-06,
+      "loss": 0.465,
+      "step": 15770
+    },
+    {
+      "epoch": 2.59004783117443,
+      "grad_norm": 0.46212153735161743,
+      "learning_rate": 4.313168295766003e-06,
+      "loss": 0.423,
+      "step": 15771
+    },
+    {
+      "epoch": 2.5902120583827726,
+      "grad_norm": 0.27173188730977343,
+      "learning_rate": 4.312699728953827e-06,
+      "loss": 0.4151,
+      "step": 15772
+    },
+    {
+      "epoch": 2.590376285591115,
+      "grad_norm": 0.44061969942428547,
+      "learning_rate": 4.312231164175917e-06,
+      "loss": 0.437,
+      "step": 15773
+    },
+    {
+      "epoch": 2.590540512799458,
+      "grad_norm": 0.2805299615077187,
+      "learning_rate": 4.311762601437362e-06,
+      "loss": 0.4249,
+      "step": 15774
+    },
+    {
+      "epoch": 2.590704740007801,
+      "grad_norm": 0.3082668237862245,
+      "learning_rate": 4.3112940407432495e-06,
+      "loss": 0.4388,
+      "step": 15775
+    },
+    {
+      "epoch": 2.5908689672161436,
+      "grad_norm": 0.2950023105650061,
+      "learning_rate": 4.3108254820986685e-06,
+      "loss": 0.4588,
+      "step": 15776
+    },
+    {
+      "epoch": 2.591033194424486,
+      "grad_norm": 0.3201705793393197,
+      "learning_rate": 4.31035692550871e-06,
+      "loss": 0.4311,
+      "step": 15777
+    },
+    {
+      "epoch": 2.5911974216328293,
+      "grad_norm": 0.3640366984477735,
+      "learning_rate": 4.309888370978464e-06,
+      "loss": 0.4126,
+      "step": 15778
+    },
+    {
+      "epoch": 2.591361648841172,
+      "grad_norm": 0.2999324534104021,
+      "learning_rate": 4.309419818513014e-06,
+      "loss": 0.445,
+      "step": 15779
+    },
+    {
+      "epoch": 2.5915258760495146,
+      "grad_norm": 0.35312819820987595,
+      "learning_rate": 4.308951268117454e-06,
+      "loss": 0.4531,
+      "step": 15780
+    },
+    {
+      "epoch": 2.591690103257857,
+      "grad_norm": 0.3104622059310826,
+      "learning_rate": 4.308482719796871e-06,
+      "loss": 0.419,
+      "step": 15781
+    },
+    {
+      "epoch": 2.5918543304662,
+      "grad_norm": 0.40858664531530114,
+      "learning_rate": 4.308014173556353e-06,
+      "loss": 0.4378,
+      "step": 15782
+    },
+    {
+      "epoch": 2.5920185576745425,
+      "grad_norm": 0.3084003255005991,
+      "learning_rate": 4.3075456294009906e-06,
+      "loss": 0.4262,
+      "step": 15783
+    },
+    {
+      "epoch": 2.5921827848828856,
+      "grad_norm": 0.37756243501180997,
+      "learning_rate": 4.3070770873358725e-06,
+      "loss": 0.4435,
+      "step": 15784
+    },
+    {
+      "epoch": 2.592347012091228,
+      "grad_norm": 0.3318405143473803,
+      "learning_rate": 4.306608547366087e-06,
+      "loss": 0.4372,
+      "step": 15785
+    },
+    {
+      "epoch": 2.592511239299571,
+      "grad_norm": 0.31825262182092245,
+      "learning_rate": 4.306140009496722e-06,
+      "loss": 0.4355,
+      "step": 15786
+    },
+    {
+      "epoch": 2.592675466507914,
+      "grad_norm": 0.26950774595672156,
+      "learning_rate": 4.30567147373287e-06,
+      "loss": 0.4384,
+      "step": 15787
+    },
+    {
+      "epoch": 2.5928396937162566,
+      "grad_norm": 0.2773574713214042,
+      "learning_rate": 4.305202940079614e-06,
+      "loss": 0.4168,
+      "step": 15788
+    },
+    {
+      "epoch": 2.593003920924599,
+      "grad_norm": 0.27968928473718313,
+      "learning_rate": 4.304734408542048e-06,
+      "loss": 0.4126,
+      "step": 15789
+    },
+    {
+      "epoch": 2.593168148132942,
+      "grad_norm": 0.339631272148747,
+      "learning_rate": 4.304265879125256e-06,
+      "loss": 0.4252,
+      "step": 15790
+    },
+    {
+      "epoch": 2.5933323753412845,
+      "grad_norm": 0.4405429762051981,
+      "learning_rate": 4.30379735183433e-06,
+      "loss": 0.4386,
+      "step": 15791
+    },
+    {
+      "epoch": 2.5934966025496275,
+      "grad_norm": 0.33307488107275973,
+      "learning_rate": 4.303328826674358e-06,
+      "loss": 0.4191,
+      "step": 15792
+    },
+    {
+      "epoch": 2.59366082975797,
+      "grad_norm": 0.31976140803255354,
+      "learning_rate": 4.3028603036504286e-06,
+      "loss": 0.4373,
+      "step": 15793
+    },
+    {
+      "epoch": 2.593825056966313,
+      "grad_norm": 0.29829335474835267,
+      "learning_rate": 4.302391782767629e-06,
+      "loss": 0.4349,
+      "step": 15794
+    },
+    {
+      "epoch": 2.593989284174656,
+      "grad_norm": 0.3889999203353962,
+      "learning_rate": 4.301923264031052e-06,
+      "loss": 0.434,
+      "step": 15795
+    },
+    {
+      "epoch": 2.5941535113829985,
+      "grad_norm": 0.30571572031790084,
+      "learning_rate": 4.301454747445781e-06,
+      "loss": 0.4201,
+      "step": 15796
+    },
+    {
+      "epoch": 2.594317738591341,
+      "grad_norm": 0.3459232731873862,
+      "learning_rate": 4.300986233016907e-06,
+      "loss": 0.4514,
+      "step": 15797
+    },
+    {
+      "epoch": 2.594481965799684,
+      "grad_norm": 0.30146836874846944,
+      "learning_rate": 4.300517720749518e-06,
+      "loss": 0.4502,
+      "step": 15798
+    },
+    {
+      "epoch": 2.5946461930080265,
+      "grad_norm": 0.36184226392840596,
+      "learning_rate": 4.3000492106487035e-06,
+      "loss": 0.431,
+      "step": 15799
+    },
+    {
+      "epoch": 2.594810420216369,
+      "grad_norm": 0.31671853093359414,
+      "learning_rate": 4.299580702719551e-06,
+      "loss": 0.4438,
+      "step": 15800
+    },
+    {
+      "epoch": 2.594974647424712,
+      "grad_norm": 0.3050096609024856,
+      "learning_rate": 4.299112196967149e-06,
+      "loss": 0.4642,
+      "step": 15801
+    },
+    {
+      "epoch": 2.595138874633055,
+      "grad_norm": 0.41602891522432706,
+      "learning_rate": 4.2986436933965866e-06,
+      "loss": 0.4327,
+      "step": 15802
+    },
+    {
+      "epoch": 2.5953031018413975,
+      "grad_norm": 0.24868676436677487,
+      "learning_rate": 4.298175192012953e-06,
+      "loss": 0.4176,
+      "step": 15803
+    },
+    {
+      "epoch": 2.5954673290497405,
+      "grad_norm": 0.36640826122549386,
+      "learning_rate": 4.2977066928213345e-06,
+      "loss": 0.4296,
+      "step": 15804
+    },
+    {
+      "epoch": 2.595631556258083,
+      "grad_norm": 0.30540983951859807,
+      "learning_rate": 4.29723819582682e-06,
+      "loss": 0.4366,
+      "step": 15805
+    },
+    {
+      "epoch": 2.595795783466426,
+      "grad_norm": 0.3490310182612695,
+      "learning_rate": 4.296769701034497e-06,
+      "loss": 0.4336,
+      "step": 15806
+    },
+    {
+      "epoch": 2.5959600106747684,
+      "grad_norm": 0.2680422378112935,
+      "learning_rate": 4.296301208449456e-06,
+      "loss": 0.4255,
+      "step": 15807
+    },
+    {
+      "epoch": 2.596124237883111,
+      "grad_norm": 0.28315344448721086,
+      "learning_rate": 4.295832718076785e-06,
+      "loss": 0.4163,
+      "step": 15808
+    },
+    {
+      "epoch": 2.596288465091454,
+      "grad_norm": 0.2923550723051912,
+      "learning_rate": 4.295364229921571e-06,
+      "loss": 0.416,
+      "step": 15809
+    },
+    {
+      "epoch": 2.596452692299797,
+      "grad_norm": 0.28055270604204857,
+      "learning_rate": 4.294895743988902e-06,
+      "loss": 0.4469,
+      "step": 15810
+    },
+    {
+      "epoch": 2.5966169195081394,
+      "grad_norm": 0.2879848375564051,
+      "learning_rate": 4.294427260283868e-06,
+      "loss": 0.4399,
+      "step": 15811
+    },
+    {
+      "epoch": 2.5967811467164825,
+      "grad_norm": 0.5644459781722265,
+      "learning_rate": 4.293958778811558e-06,
+      "loss": 0.4669,
+      "step": 15812
+    },
+    {
+      "epoch": 2.596945373924825,
+      "grad_norm": 0.2848883445687235,
+      "learning_rate": 4.293490299577056e-06,
+      "loss": 0.421,
+      "step": 15813
+    },
+    {
+      "epoch": 2.597109601133168,
+      "grad_norm": 0.3855917851665769,
+      "learning_rate": 4.293021822585452e-06,
+      "loss": 0.4571,
+      "step": 15814
+    },
+    {
+      "epoch": 2.5972738283415104,
+      "grad_norm": 0.37697902958562113,
+      "learning_rate": 4.292553347841836e-06,
+      "loss": 0.4241,
+      "step": 15815
+    },
+    {
+      "epoch": 2.597438055549853,
+      "grad_norm": 0.3082437220187043,
+      "learning_rate": 4.2920848753512945e-06,
+      "loss": 0.4571,
+      "step": 15816
+    },
+    {
+      "epoch": 2.5976022827581957,
+      "grad_norm": 0.4806961940115057,
+      "learning_rate": 4.291616405118915e-06,
+      "loss": 0.4438,
+      "step": 15817
+    },
+    {
+      "epoch": 2.597766509966539,
+      "grad_norm": 0.33571584868959514,
+      "learning_rate": 4.291147937149787e-06,
+      "loss": 0.4355,
+      "step": 15818
+    },
+    {
+      "epoch": 2.5979307371748814,
+      "grad_norm": 0.8390119913907225,
+      "learning_rate": 4.290679471448998e-06,
+      "loss": 0.4369,
+      "step": 15819
+    },
+    {
+      "epoch": 2.598094964383224,
+      "grad_norm": 0.24477381682903024,
+      "learning_rate": 4.290211008021638e-06,
+      "loss": 0.4346,
+      "step": 15820
+    },
+    {
+      "epoch": 2.598259191591567,
+      "grad_norm": 0.36475684221611265,
+      "learning_rate": 4.289742546872789e-06,
+      "loss": 0.4485,
+      "step": 15821
+    },
+    {
+      "epoch": 2.59842341879991,
+      "grad_norm": 0.3489466071086147,
+      "learning_rate": 4.289274088007544e-06,
+      "loss": 0.4417,
+      "step": 15822
+    },
+    {
+      "epoch": 2.5985876460082524,
+      "grad_norm": 0.27971414760095065,
+      "learning_rate": 4.288805631430991e-06,
+      "loss": 0.441,
+      "step": 15823
+    },
+    {
+      "epoch": 2.598751873216595,
+      "grad_norm": 0.26586588593970034,
+      "learning_rate": 4.288337177148215e-06,
+      "loss": 0.4563,
+      "step": 15824
+    },
+    {
+      "epoch": 2.5989161004249377,
+      "grad_norm": 0.30542948972381184,
+      "learning_rate": 4.287868725164307e-06,
+      "loss": 0.4266,
+      "step": 15825
+    },
+    {
+      "epoch": 2.599080327633281,
+      "grad_norm": 0.31955052815461044,
+      "learning_rate": 4.287400275484351e-06,
+      "loss": 0.4498,
+      "step": 15826
+    },
+    {
+      "epoch": 2.5992445548416234,
+      "grad_norm": 0.31979104627134625,
+      "learning_rate": 4.28693182811344e-06,
+      "loss": 0.4317,
+      "step": 15827
+    },
+    {
+      "epoch": 2.599408782049966,
+      "grad_norm": 0.7336338061730221,
+      "learning_rate": 4.286463383056658e-06,
+      "loss": 0.4161,
+      "step": 15828
+    },
+    {
+      "epoch": 2.599573009258309,
+      "grad_norm": 0.26448341402895187,
+      "learning_rate": 4.285994940319094e-06,
+      "loss": 0.4546,
+      "step": 15829
+    },
+    {
+      "epoch": 2.599737236466652,
+      "grad_norm": 0.2814894747781621,
+      "learning_rate": 4.285526499905835e-06,
+      "loss": 0.4408,
+      "step": 15830
+    },
+    {
+      "epoch": 2.5999014636749944,
+      "grad_norm": 0.32304929685665024,
+      "learning_rate": 4.2850580618219685e-06,
+      "loss": 0.4495,
+      "step": 15831
+    },
+    {
+      "epoch": 2.600065690883337,
+      "grad_norm": 0.3003407465523042,
+      "learning_rate": 4.284589626072584e-06,
+      "loss": 0.4211,
+      "step": 15832
+    },
+    {
+      "epoch": 2.6002299180916797,
+      "grad_norm": 0.3126022106522328,
+      "learning_rate": 4.2841211926627685e-06,
+      "loss": 0.4459,
+      "step": 15833
+    },
+    {
+      "epoch": 2.6003941453000223,
+      "grad_norm": 0.3202251903747605,
+      "learning_rate": 4.283652761597607e-06,
+      "loss": 0.4123,
+      "step": 15834
+    },
+    {
+      "epoch": 2.6005583725083654,
+      "grad_norm": 0.3193033038388879,
+      "learning_rate": 4.283184332882192e-06,
+      "loss": 0.43,
+      "step": 15835
+    },
+    {
+      "epoch": 2.600722599716708,
+      "grad_norm": 0.4759091609533221,
+      "learning_rate": 4.282715906521607e-06,
+      "loss": 0.4351,
+      "step": 15836
+    },
+    {
+      "epoch": 2.6008868269250507,
+      "grad_norm": 0.34048263468535467,
+      "learning_rate": 4.2822474825209426e-06,
+      "loss": 0.4542,
+      "step": 15837
+    },
+    {
+      "epoch": 2.6010510541333938,
+      "grad_norm": 0.2964018813124492,
+      "learning_rate": 4.281779060885284e-06,
+      "loss": 0.4524,
+      "step": 15838
+    },
+    {
+      "epoch": 2.6012152813417364,
+      "grad_norm": 0.272121429844428,
+      "learning_rate": 4.281310641619719e-06,
+      "loss": 0.4389,
+      "step": 15839
+    },
+    {
+      "epoch": 2.601379508550079,
+      "grad_norm": 0.2930384076277509,
+      "learning_rate": 4.280842224729335e-06,
+      "loss": 0.4205,
+      "step": 15840
+    },
+    {
+      "epoch": 2.6015437357584217,
+      "grad_norm": 0.3932004190719817,
+      "learning_rate": 4.28037381021922e-06,
+      "loss": 0.4208,
+      "step": 15841
+    },
+    {
+      "epoch": 2.6017079629667643,
+      "grad_norm": 0.2916958143650516,
+      "learning_rate": 4.279905398094463e-06,
+      "loss": 0.4442,
+      "step": 15842
+    },
+    {
+      "epoch": 2.6018721901751074,
+      "grad_norm": 0.38320540985005647,
+      "learning_rate": 4.279436988360148e-06,
+      "loss": 0.4371,
+      "step": 15843
+    },
+    {
+      "epoch": 2.60203641738345,
+      "grad_norm": 0.3442761883932953,
+      "learning_rate": 4.278968581021366e-06,
+      "loss": 0.4267,
+      "step": 15844
+    },
+    {
+      "epoch": 2.6022006445917927,
+      "grad_norm": 0.3311876856923415,
+      "learning_rate": 4.278500176083204e-06,
+      "loss": 0.4491,
+      "step": 15845
+    },
+    {
+      "epoch": 2.6023648718001358,
+      "grad_norm": 0.271081882936446,
+      "learning_rate": 4.278031773550745e-06,
+      "loss": 0.443,
+      "step": 15846
+    },
+    {
+      "epoch": 2.6025290990084784,
+      "grad_norm": 0.27360586704891615,
+      "learning_rate": 4.27756337342908e-06,
+      "loss": 0.4234,
+      "step": 15847
+    },
+    {
+      "epoch": 2.602693326216821,
+      "grad_norm": 0.3028946730931956,
+      "learning_rate": 4.277094975723297e-06,
+      "loss": 0.4178,
+      "step": 15848
+    },
+    {
+      "epoch": 2.6028575534251637,
+      "grad_norm": 0.4211902298436584,
+      "learning_rate": 4.27662658043848e-06,
+      "loss": 0.4279,
+      "step": 15849
+    },
+    {
+      "epoch": 2.6030217806335063,
+      "grad_norm": 0.29390415171983797,
+      "learning_rate": 4.276158187579719e-06,
+      "loss": 0.4311,
+      "step": 15850
+    },
+    {
+      "epoch": 2.603186007841849,
+      "grad_norm": 0.2893391661327366,
+      "learning_rate": 4.275689797152101e-06,
+      "loss": 0.4442,
+      "step": 15851
+    },
+    {
+      "epoch": 2.603350235050192,
+      "grad_norm": 0.30309257216350804,
+      "learning_rate": 4.275221409160711e-06,
+      "loss": 0.4148,
+      "step": 15852
+    },
+    {
+      "epoch": 2.6035144622585347,
+      "grad_norm": 0.3694770688838367,
+      "learning_rate": 4.274753023610641e-06,
+      "loss": 0.4402,
+      "step": 15853
+    },
+    {
+      "epoch": 2.6036786894668773,
+      "grad_norm": 0.31224351318683174,
+      "learning_rate": 4.274284640506972e-06,
+      "loss": 0.4209,
+      "step": 15854
+    },
+    {
+      "epoch": 2.6038429166752204,
+      "grad_norm": 0.26476569410972106,
+      "learning_rate": 4.273816259854794e-06,
+      "loss": 0.4353,
+      "step": 15855
+    },
+    {
+      "epoch": 2.604007143883563,
+      "grad_norm": 0.2524068878432348,
+      "learning_rate": 4.273347881659193e-06,
+      "loss": 0.4341,
+      "step": 15856
+    },
+    {
+      "epoch": 2.6041713710919057,
+      "grad_norm": 0.3836716245953561,
+      "learning_rate": 4.272879505925259e-06,
+      "loss": 0.4405,
+      "step": 15857
+    },
+    {
+      "epoch": 2.6043355983002483,
+      "grad_norm": 0.31367484844860166,
+      "learning_rate": 4.272411132658076e-06,
+      "loss": 0.437,
+      "step": 15858
+    },
+    {
+      "epoch": 2.604499825508591,
+      "grad_norm": 0.24073228613682615,
+      "learning_rate": 4.271942761862731e-06,
+      "loss": 0.4263,
+      "step": 15859
+    },
+    {
+      "epoch": 2.604664052716934,
+      "grad_norm": 0.31112665903222925,
+      "learning_rate": 4.271474393544313e-06,
+      "loss": 0.4367,
+      "step": 15860
+    },
+    {
+      "epoch": 2.6048282799252767,
+      "grad_norm": 0.39513586710108894,
+      "learning_rate": 4.271006027707909e-06,
+      "loss": 0.4421,
+      "step": 15861
+    },
+    {
+      "epoch": 2.6049925071336193,
+      "grad_norm": 0.37016694695642754,
+      "learning_rate": 4.270537664358604e-06,
+      "loss": 0.4234,
+      "step": 15862
+    },
+    {
+      "epoch": 2.6051567343419624,
+      "grad_norm": 0.3234392124628261,
+      "learning_rate": 4.270069303501487e-06,
+      "loss": 0.4366,
+      "step": 15863
+    },
+    {
+      "epoch": 2.605320961550305,
+      "grad_norm": 0.3211553226048717,
+      "learning_rate": 4.269600945141642e-06,
+      "loss": 0.4139,
+      "step": 15864
+    },
+    {
+      "epoch": 2.6054851887586477,
+      "grad_norm": 0.297381168760795,
+      "learning_rate": 4.269132589284157e-06,
+      "loss": 0.4222,
+      "step": 15865
+    },
+    {
+      "epoch": 2.6056494159669903,
+      "grad_norm": 0.2684853314660215,
+      "learning_rate": 4.268664235934119e-06,
+      "loss": 0.4226,
+      "step": 15866
+    },
+    {
+      "epoch": 2.605813643175333,
+      "grad_norm": 0.24428461191927797,
+      "learning_rate": 4.268195885096617e-06,
+      "loss": 0.4055,
+      "step": 15867
+    },
+    {
+      "epoch": 2.6059778703836756,
+      "grad_norm": 0.30690138353801594,
+      "learning_rate": 4.267727536776734e-06,
+      "loss": 0.4327,
+      "step": 15868
+    },
+    {
+      "epoch": 2.6061420975920186,
+      "grad_norm": 0.309580731291792,
+      "learning_rate": 4.267259190979558e-06,
+      "loss": 0.4277,
+      "step": 15869
+    },
+    {
+      "epoch": 2.6063063248003613,
+      "grad_norm": 0.3521082341168475,
+      "learning_rate": 4.2667908477101794e-06,
+      "loss": 0.4369,
+      "step": 15870
+    },
+    {
+      "epoch": 2.606470552008704,
+      "grad_norm": 0.27827993463951545,
+      "learning_rate": 4.266322506973679e-06,
+      "loss": 0.4153,
+      "step": 15871
+    },
+    {
+      "epoch": 2.606634779217047,
+      "grad_norm": 0.29314152987116177,
+      "learning_rate": 4.265854168775148e-06,
+      "loss": 0.4383,
+      "step": 15872
+    },
+    {
+      "epoch": 2.6067990064253896,
+      "grad_norm": 0.4175886384913535,
+      "learning_rate": 4.265385833119668e-06,
+      "loss": 0.4374,
+      "step": 15873
+    },
+    {
+      "epoch": 2.6069632336337323,
+      "grad_norm": 0.3215341260226248,
+      "learning_rate": 4.264917500012331e-06,
+      "loss": 0.4481,
+      "step": 15874
+    },
+    {
+      "epoch": 2.607127460842075,
+      "grad_norm": 0.28916901461660643,
+      "learning_rate": 4.2644491694582196e-06,
+      "loss": 0.4291,
+      "step": 15875
+    },
+    {
+      "epoch": 2.6072916880504176,
+      "grad_norm": 0.5564639233489209,
+      "learning_rate": 4.263980841462424e-06,
+      "loss": 0.4434,
+      "step": 15876
+    },
+    {
+      "epoch": 2.6074559152587606,
+      "grad_norm": 0.4252859915961404,
+      "learning_rate": 4.263512516030027e-06,
+      "loss": 0.433,
+      "step": 15877
+    },
+    {
+      "epoch": 2.6076201424671033,
+      "grad_norm": 0.3004118104642622,
+      "learning_rate": 4.2630441931661195e-06,
+      "loss": 0.4334,
+      "step": 15878
+    },
+    {
+      "epoch": 2.607784369675446,
+      "grad_norm": 0.3107784239472434,
+      "learning_rate": 4.262575872875782e-06,
+      "loss": 0.429,
+      "step": 15879
+    },
+    {
+      "epoch": 2.607948596883789,
+      "grad_norm": 0.3734813433262263,
+      "learning_rate": 4.262107555164105e-06,
+      "loss": 0.4144,
+      "step": 15880
+    },
+    {
+      "epoch": 2.6081128240921316,
+      "grad_norm": 0.3201747720302578,
+      "learning_rate": 4.261639240036174e-06,
+      "loss": 0.4365,
+      "step": 15881
+    },
+    {
+      "epoch": 2.6082770513004743,
+      "grad_norm": 0.30335678042097514,
+      "learning_rate": 4.261170927497076e-06,
+      "loss": 0.4369,
+      "step": 15882
+    },
+    {
+      "epoch": 2.608441278508817,
+      "grad_norm": 0.30057906830959097,
+      "learning_rate": 4.260702617551896e-06,
+      "loss": 0.4274,
+      "step": 15883
+    },
+    {
+      "epoch": 2.6086055057171595,
+      "grad_norm": 0.6371499207725939,
+      "learning_rate": 4.260234310205721e-06,
+      "loss": 0.4348,
+      "step": 15884
+    },
+    {
+      "epoch": 2.608769732925502,
+      "grad_norm": 0.33918844373338264,
+      "learning_rate": 4.2597660054636376e-06,
+      "loss": 0.4335,
+      "step": 15885
+    },
+    {
+      "epoch": 2.6089339601338453,
+      "grad_norm": 0.3361159771638176,
+      "learning_rate": 4.259297703330732e-06,
+      "loss": 0.4276,
+      "step": 15886
+    },
+    {
+      "epoch": 2.609098187342188,
+      "grad_norm": 0.3028877725278089,
+      "learning_rate": 4.258829403812091e-06,
+      "loss": 0.4339,
+      "step": 15887
+    },
+    {
+      "epoch": 2.6092624145505305,
+      "grad_norm": 0.2717049246569981,
+      "learning_rate": 4.258361106912799e-06,
+      "loss": 0.422,
+      "step": 15888
+    },
+    {
+      "epoch": 2.6094266417588736,
+      "grad_norm": 0.29662739630039475,
+      "learning_rate": 4.257892812637942e-06,
+      "loss": 0.4485,
+      "step": 15889
+    },
+    {
+      "epoch": 2.6095908689672163,
+      "grad_norm": 0.32580420835705043,
+      "learning_rate": 4.257424520992608e-06,
+      "loss": 0.4304,
+      "step": 15890
+    },
+    {
+      "epoch": 2.609755096175559,
+      "grad_norm": 0.2789367086644554,
+      "learning_rate": 4.256956231981883e-06,
+      "loss": 0.4312,
+      "step": 15891
+    },
+    {
+      "epoch": 2.6099193233839015,
+      "grad_norm": 0.27050105126913654,
+      "learning_rate": 4.256487945610853e-06,
+      "loss": 0.4363,
+      "step": 15892
+    },
+    {
+      "epoch": 2.610083550592244,
+      "grad_norm": 0.32473112425656697,
+      "learning_rate": 4.256019661884601e-06,
+      "loss": 0.4514,
+      "step": 15893
+    },
+    {
+      "epoch": 2.6102477778005873,
+      "grad_norm": 1.2224814949738927,
+      "learning_rate": 4.255551380808217e-06,
+      "loss": 0.4475,
+      "step": 15894
+    },
+    {
+      "epoch": 2.61041200500893,
+      "grad_norm": 0.2899718094249228,
+      "learning_rate": 4.255083102386787e-06,
+      "loss": 0.4432,
+      "step": 15895
+    },
+    {
+      "epoch": 2.6105762322172725,
+      "grad_norm": 0.31770282494513413,
+      "learning_rate": 4.254614826625393e-06,
+      "loss": 0.4164,
+      "step": 15896
+    },
+    {
+      "epoch": 2.6107404594256156,
+      "grad_norm": 0.31746036994094073,
+      "learning_rate": 4.254146553529126e-06,
+      "loss": 0.4343,
+      "step": 15897
+    },
+    {
+      "epoch": 2.6109046866339582,
+      "grad_norm": 0.32599628063434466,
+      "learning_rate": 4.253678283103067e-06,
+      "loss": 0.4484,
+      "step": 15898
+    },
+    {
+      "epoch": 2.611068913842301,
+      "grad_norm": 0.29199535549958605,
+      "learning_rate": 4.2532100153523045e-06,
+      "loss": 0.424,
+      "step": 15899
+    },
+    {
+      "epoch": 2.6112331410506435,
+      "grad_norm": 0.30444024787236657,
+      "learning_rate": 4.252741750281925e-06,
+      "loss": 0.4227,
+      "step": 15900
+    },
+    {
+      "epoch": 2.611397368258986,
+      "grad_norm": 0.4412199564586546,
+      "learning_rate": 4.252273487897013e-06,
+      "loss": 0.4427,
+      "step": 15901
+    },
+    {
+      "epoch": 2.611561595467329,
+      "grad_norm": 0.5410335689281915,
+      "learning_rate": 4.251805228202654e-06,
+      "loss": 0.4292,
+      "step": 15902
+    },
+    {
+      "epoch": 2.611725822675672,
+      "grad_norm": 0.41213691877007574,
+      "learning_rate": 4.251336971203934e-06,
+      "loss": 0.4487,
+      "step": 15903
+    },
+    {
+      "epoch": 2.6118900498840145,
+      "grad_norm": 0.3330652790042675,
+      "learning_rate": 4.250868716905941e-06,
+      "loss": 0.4392,
+      "step": 15904
+    },
+    {
+      "epoch": 2.612054277092357,
+      "grad_norm": 0.3481140986138393,
+      "learning_rate": 4.250400465313758e-06,
+      "loss": 0.422,
+      "step": 15905
+    },
+    {
+      "epoch": 2.6122185043007002,
+      "grad_norm": 0.31634923265199455,
+      "learning_rate": 4.249932216432472e-06,
+      "loss": 0.4314,
+      "step": 15906
+    },
+    {
+      "epoch": 2.612382731509043,
+      "grad_norm": 0.2918750725823665,
+      "learning_rate": 4.249463970267168e-06,
+      "loss": 0.4341,
+      "step": 15907
+    },
+    {
+      "epoch": 2.6125469587173855,
+      "grad_norm": 1.0899497941084635,
+      "learning_rate": 4.24899572682293e-06,
+      "loss": 0.4252,
+      "step": 15908
+    },
+    {
+      "epoch": 2.612711185925728,
+      "grad_norm": 0.32946172413991354,
+      "learning_rate": 4.248527486104847e-06,
+      "loss": 0.4436,
+      "step": 15909
+    },
+    {
+      "epoch": 2.612875413134071,
+      "grad_norm": 0.28083680309894277,
+      "learning_rate": 4.248059248118003e-06,
+      "loss": 0.4234,
+      "step": 15910
+    },
+    {
+      "epoch": 2.613039640342414,
+      "grad_norm": 0.2696569450276403,
+      "learning_rate": 4.247591012867483e-06,
+      "loss": 0.4317,
+      "step": 15911
+    },
+    {
+      "epoch": 2.6132038675507565,
+      "grad_norm": 0.32797454999406345,
+      "learning_rate": 4.247122780358374e-06,
+      "loss": 0.4485,
+      "step": 15912
+    },
+    {
+      "epoch": 2.613368094759099,
+      "grad_norm": 0.2978012132300891,
+      "learning_rate": 4.246654550595759e-06,
+      "loss": 0.4444,
+      "step": 15913
+    },
+    {
+      "epoch": 2.6135323219674422,
+      "grad_norm": 0.3347378118305014,
+      "learning_rate": 4.246186323584725e-06,
+      "loss": 0.4491,
+      "step": 15914
+    },
+    {
+      "epoch": 2.613696549175785,
+      "grad_norm": 0.30672663933298117,
+      "learning_rate": 4.245718099330358e-06,
+      "loss": 0.4711,
+      "step": 15915
+    },
+    {
+      "epoch": 2.6138607763841275,
+      "grad_norm": 0.2780680965058183,
+      "learning_rate": 4.245249877837742e-06,
+      "loss": 0.4234,
+      "step": 15916
+    },
+    {
+      "epoch": 2.61402500359247,
+      "grad_norm": 0.31100650951866027,
+      "learning_rate": 4.2447816591119626e-06,
+      "loss": 0.4395,
+      "step": 15917
+    },
+    {
+      "epoch": 2.614189230800813,
+      "grad_norm": 0.3165837170282832,
+      "learning_rate": 4.244313443158106e-06,
+      "loss": 0.4534,
+      "step": 15918
+    },
+    {
+      "epoch": 2.6143534580091554,
+      "grad_norm": 0.2945574433533434,
+      "learning_rate": 4.243845229981256e-06,
+      "loss": 0.4322,
+      "step": 15919
+    },
+    {
+      "epoch": 2.6145176852174985,
+      "grad_norm": 0.2854056769828595,
+      "learning_rate": 4.243377019586501e-06,
+      "loss": 0.4216,
+      "step": 15920
+    },
+    {
+      "epoch": 2.614681912425841,
+      "grad_norm": 0.3449269045558118,
+      "learning_rate": 4.242908811978924e-06,
+      "loss": 0.4335,
+      "step": 15921
+    },
+    {
+      "epoch": 2.6148461396341838,
+      "grad_norm": 0.29707638186807517,
+      "learning_rate": 4.2424406071636085e-06,
+      "loss": 0.4307,
+      "step": 15922
+    },
+    {
+      "epoch": 2.615010366842527,
+      "grad_norm": 0.31842893735652705,
+      "learning_rate": 4.241972405145641e-06,
+      "loss": 0.4462,
+      "step": 15923
+    },
+    {
+      "epoch": 2.6151745940508695,
+      "grad_norm": 0.2784297794765863,
+      "learning_rate": 4.241504205930108e-06,
+      "loss": 0.4393,
+      "step": 15924
+    },
+    {
+      "epoch": 2.615338821259212,
+      "grad_norm": 0.637609011183369,
+      "learning_rate": 4.241036009522094e-06,
+      "loss": 0.4164,
+      "step": 15925
+    },
+    {
+      "epoch": 2.6155030484675548,
+      "grad_norm": 0.30614294622529886,
+      "learning_rate": 4.240567815926683e-06,
+      "loss": 0.4284,
+      "step": 15926
+    },
+    {
+      "epoch": 2.6156672756758974,
+      "grad_norm": 0.3447127775105285,
+      "learning_rate": 4.240099625148961e-06,
+      "loss": 0.4537,
+      "step": 15927
+    },
+    {
+      "epoch": 2.6158315028842405,
+      "grad_norm": 0.4970599090971395,
+      "learning_rate": 4.2396314371940125e-06,
+      "loss": 0.4371,
+      "step": 15928
+    },
+    {
+      "epoch": 2.615995730092583,
+      "grad_norm": 0.43555955002393065,
+      "learning_rate": 4.239163252066925e-06,
+      "loss": 0.4361,
+      "step": 15929
+    },
+    {
+      "epoch": 2.6161599573009258,
+      "grad_norm": 0.32903375832698184,
+      "learning_rate": 4.238695069772779e-06,
+      "loss": 0.4384,
+      "step": 15930
+    },
+    {
+      "epoch": 2.616324184509269,
+      "grad_norm": 0.33202092298407887,
+      "learning_rate": 4.238226890316663e-06,
+      "loss": 0.4481,
+      "step": 15931
+    },
+    {
+      "epoch": 2.6164884117176115,
+      "grad_norm": 0.45140111170335284,
+      "learning_rate": 4.237758713703659e-06,
+      "loss": 0.4361,
+      "step": 15932
+    },
+    {
+      "epoch": 2.616652638925954,
+      "grad_norm": 0.33571375004655346,
+      "learning_rate": 4.237290539938854e-06,
+      "loss": 0.4146,
+      "step": 15933
+    },
+    {
+      "epoch": 2.6168168661342968,
+      "grad_norm": 0.5085529476929775,
+      "learning_rate": 4.236822369027331e-06,
+      "loss": 0.4412,
+      "step": 15934
+    },
+    {
+      "epoch": 2.6169810933426394,
+      "grad_norm": 0.26527704034784494,
+      "learning_rate": 4.236354200974178e-06,
+      "loss": 0.4344,
+      "step": 15935
+    },
+    {
+      "epoch": 2.617145320550982,
+      "grad_norm": 0.2653581981143029,
+      "learning_rate": 4.235886035784477e-06,
+      "loss": 0.4255,
+      "step": 15936
+    },
+    {
+      "epoch": 2.617309547759325,
+      "grad_norm": 0.26429214414880825,
+      "learning_rate": 4.235417873463314e-06,
+      "loss": 0.4164,
+      "step": 15937
+    },
+    {
+      "epoch": 2.6174737749676678,
+      "grad_norm": 0.451089379274073,
+      "learning_rate": 4.234949714015772e-06,
+      "loss": 0.4303,
+      "step": 15938
+    },
+    {
+      "epoch": 2.6176380021760104,
+      "grad_norm": 0.613858816212356,
+      "learning_rate": 4.2344815574469376e-06,
+      "loss": 0.4503,
+      "step": 15939
+    },
+    {
+      "epoch": 2.6178022293843535,
+      "grad_norm": 0.42073771656410275,
+      "learning_rate": 4.234013403761895e-06,
+      "loss": 0.4332,
+      "step": 15940
+    },
+    {
+      "epoch": 2.617966456592696,
+      "grad_norm": 0.2764531610938458,
+      "learning_rate": 4.233545252965728e-06,
+      "loss": 0.4246,
+      "step": 15941
+    },
+    {
+      "epoch": 2.6181306838010387,
+      "grad_norm": 0.3069746988437526,
+      "learning_rate": 4.233077105063521e-06,
+      "loss": 0.44,
+      "step": 15942
+    },
+    {
+      "epoch": 2.6182949110093814,
+      "grad_norm": 0.34945204460664414,
+      "learning_rate": 4.232608960060361e-06,
+      "loss": 0.4338,
+      "step": 15943
+    },
+    {
+      "epoch": 2.618459138217724,
+      "grad_norm": 0.3304580150854837,
+      "learning_rate": 4.23214081796133e-06,
+      "loss": 0.4513,
+      "step": 15944
+    },
+    {
+      "epoch": 2.618623365426067,
+      "grad_norm": 0.4045233492426899,
+      "learning_rate": 4.231672678771513e-06,
+      "loss": 0.4383,
+      "step": 15945
+    },
+    {
+      "epoch": 2.6187875926344097,
+      "grad_norm": 0.3027881257346491,
+      "learning_rate": 4.231204542495996e-06,
+      "loss": 0.4323,
+      "step": 15946
+    },
+    {
+      "epoch": 2.6189518198427524,
+      "grad_norm": 0.3566048787535595,
+      "learning_rate": 4.230736409139861e-06,
+      "loss": 0.4433,
+      "step": 15947
+    },
+    {
+      "epoch": 2.6191160470510955,
+      "grad_norm": 0.28909857225231095,
+      "learning_rate": 4.230268278708193e-06,
+      "loss": 0.4371,
+      "step": 15948
+    },
+    {
+      "epoch": 2.619280274259438,
+      "grad_norm": 0.30328281123785367,
+      "learning_rate": 4.229800151206077e-06,
+      "loss": 0.4616,
+      "step": 15949
+    },
+    {
+      "epoch": 2.6194445014677807,
+      "grad_norm": 0.28871293493755906,
+      "learning_rate": 4.229332026638598e-06,
+      "loss": 0.4094,
+      "step": 15950
+    },
+    {
+      "epoch": 2.6196087286761234,
+      "grad_norm": 0.3015937440164125,
+      "learning_rate": 4.228863905010839e-06,
+      "loss": 0.4405,
+      "step": 15951
+    },
+    {
+      "epoch": 2.619772955884466,
+      "grad_norm": 0.33214719515041785,
+      "learning_rate": 4.2283957863278845e-06,
+      "loss": 0.4394,
+      "step": 15952
+    },
+    {
+      "epoch": 2.6199371830928087,
+      "grad_norm": 0.3584487107355222,
+      "learning_rate": 4.227927670594818e-06,
+      "loss": 0.4484,
+      "step": 15953
+    },
+    {
+      "epoch": 2.6201014103011517,
+      "grad_norm": 0.3666964485124772,
+      "learning_rate": 4.227459557816728e-06,
+      "loss": 0.4186,
+      "step": 15954
+    },
+    {
+      "epoch": 2.6202656375094944,
+      "grad_norm": 0.27626251940460655,
+      "learning_rate": 4.226991447998694e-06,
+      "loss": 0.4184,
+      "step": 15955
+    },
+    {
+      "epoch": 2.620429864717837,
+      "grad_norm": 0.3010169911353582,
+      "learning_rate": 4.226523341145801e-06,
+      "loss": 0.4435,
+      "step": 15956
+    },
+    {
+      "epoch": 2.62059409192618,
+      "grad_norm": 0.3042075192771724,
+      "learning_rate": 4.226055237263132e-06,
+      "loss": 0.4275,
+      "step": 15957
+    },
+    {
+      "epoch": 2.6207583191345227,
+      "grad_norm": 0.28337020528828455,
+      "learning_rate": 4.225587136355774e-06,
+      "loss": 0.4321,
+      "step": 15958
+    },
+    {
+      "epoch": 2.6209225463428654,
+      "grad_norm": 0.275572951052758,
+      "learning_rate": 4.22511903842881e-06,
+      "loss": 0.4293,
+      "step": 15959
+    },
+    {
+      "epoch": 2.621086773551208,
+      "grad_norm": 0.3765406983641153,
+      "learning_rate": 4.2246509434873235e-06,
+      "loss": 0.4488,
+      "step": 15960
+    },
+    {
+      "epoch": 2.6212510007595506,
+      "grad_norm": 0.3500604250595598,
+      "learning_rate": 4.224182851536398e-06,
+      "loss": 0.4265,
+      "step": 15961
+    },
+    {
+      "epoch": 2.6214152279678937,
+      "grad_norm": 0.3161889341013727,
+      "learning_rate": 4.22371476258112e-06,
+      "loss": 0.4507,
+      "step": 15962
+    },
+    {
+      "epoch": 2.6215794551762364,
+      "grad_norm": 0.2995552223403898,
+      "learning_rate": 4.223246676626571e-06,
+      "loss": 0.4271,
+      "step": 15963
+    },
+    {
+      "epoch": 2.621743682384579,
+      "grad_norm": 0.40480995619927845,
+      "learning_rate": 4.222778593677833e-06,
+      "loss": 0.4236,
+      "step": 15964
+    },
+    {
+      "epoch": 2.621907909592922,
+      "grad_norm": 0.3988167086614991,
+      "learning_rate": 4.222310513739995e-06,
+      "loss": 0.4419,
+      "step": 15965
+    },
+    {
+      "epoch": 2.6220721368012647,
+      "grad_norm": 0.32298607418392483,
+      "learning_rate": 4.221842436818136e-06,
+      "loss": 0.4397,
+      "step": 15966
+    },
+    {
+      "epoch": 2.6222363640096074,
+      "grad_norm": 0.29969031931968826,
+      "learning_rate": 4.2213743629173425e-06,
+      "loss": 0.4201,
+      "step": 15967
+    },
+    {
+      "epoch": 2.62240059121795,
+      "grad_norm": 0.27714271714246214,
+      "learning_rate": 4.220906292042697e-06,
+      "loss": 0.4431,
+      "step": 15968
+    },
+    {
+      "epoch": 2.6225648184262926,
+      "grad_norm": 0.2942175988231403,
+      "learning_rate": 4.220438224199285e-06,
+      "loss": 0.4351,
+      "step": 15969
+    },
+    {
+      "epoch": 2.6227290456346353,
+      "grad_norm": 0.27700577987436237,
+      "learning_rate": 4.21997015939219e-06,
+      "loss": 0.4394,
+      "step": 15970
+    },
+    {
+      "epoch": 2.6228932728429784,
+      "grad_norm": 0.31847153155102015,
+      "learning_rate": 4.219502097626492e-06,
+      "loss": 0.428,
+      "step": 15971
+    },
+    {
+      "epoch": 2.623057500051321,
+      "grad_norm": 0.2929640968271118,
+      "learning_rate": 4.219034038907278e-06,
+      "loss": 0.4342,
+      "step": 15972
+    },
+    {
+      "epoch": 2.6232217272596636,
+      "grad_norm": 0.36743081273337336,
+      "learning_rate": 4.21856598323963e-06,
+      "loss": 0.434,
+      "step": 15973
+    },
+    {
+      "epoch": 2.6233859544680067,
+      "grad_norm": 0.251752898125371,
+      "learning_rate": 4.218097930628634e-06,
+      "loss": 0.4176,
+      "step": 15974
+    },
+    {
+      "epoch": 2.6235501816763493,
+      "grad_norm": 0.26781368785221105,
+      "learning_rate": 4.217629881079372e-06,
+      "loss": 0.4441,
+      "step": 15975
+    },
+    {
+      "epoch": 2.623714408884692,
+      "grad_norm": 0.29308075035555686,
+      "learning_rate": 4.217161834596925e-06,
+      "loss": 0.4396,
+      "step": 15976
+    },
+    {
+      "epoch": 2.6238786360930346,
+      "grad_norm": 0.3126222728890951,
+      "learning_rate": 4.21669379118638e-06,
+      "loss": 0.4339,
+      "step": 15977
+    },
+    {
+      "epoch": 2.6240428633013773,
+      "grad_norm": 0.42222963655284285,
+      "learning_rate": 4.21622575085282e-06,
+      "loss": 0.4497,
+      "step": 15978
+    },
+    {
+      "epoch": 2.6242070905097203,
+      "grad_norm": 0.2680911148735966,
+      "learning_rate": 4.215757713601327e-06,
+      "loss": 0.4505,
+      "step": 15979
+    },
+    {
+      "epoch": 2.624371317718063,
+      "grad_norm": 0.26722916634092214,
+      "learning_rate": 4.215289679436986e-06,
+      "loss": 0.4328,
+      "step": 15980
+    },
+    {
+      "epoch": 2.6245355449264056,
+      "grad_norm": 0.28522394018656005,
+      "learning_rate": 4.214821648364877e-06,
+      "loss": 0.4208,
+      "step": 15981
+    },
+    {
+      "epoch": 2.6246997721347487,
+      "grad_norm": 0.3730121107654761,
+      "learning_rate": 4.214353620390086e-06,
+      "loss": 0.4349,
+      "step": 15982
+    },
+    {
+      "epoch": 2.6248639993430913,
+      "grad_norm": 0.34376190480566104,
+      "learning_rate": 4.2138855955176955e-06,
+      "loss": 0.4477,
+      "step": 15983
+    },
+    {
+      "epoch": 2.625028226551434,
+      "grad_norm": 0.35628550422099475,
+      "learning_rate": 4.21341757375279e-06,
+      "loss": 0.4342,
+      "step": 15984
+    },
+    {
+      "epoch": 2.6251924537597766,
+      "grad_norm": 0.3953333670118324,
+      "learning_rate": 4.212949555100451e-06,
+      "loss": 0.4378,
+      "step": 15985
+    },
+    {
+      "epoch": 2.6253566809681192,
+      "grad_norm": 0.3403985295123224,
+      "learning_rate": 4.212481539565762e-06,
+      "loss": 0.4206,
+      "step": 15986
+    },
+    {
+      "epoch": 2.625520908176462,
+      "grad_norm": 0.3700412707613751,
+      "learning_rate": 4.212013527153809e-06,
+      "loss": 0.4462,
+      "step": 15987
+    },
+    {
+      "epoch": 2.625685135384805,
+      "grad_norm": 0.3124818850646178,
+      "learning_rate": 4.211545517869669e-06,
+      "loss": 0.4382,
+      "step": 15988
+    },
+    {
+      "epoch": 2.6258493625931476,
+      "grad_norm": 0.37030254675822366,
+      "learning_rate": 4.21107751171843e-06,
+      "loss": 0.4218,
+      "step": 15989
+    },
+    {
+      "epoch": 2.6260135898014902,
+      "grad_norm": 0.3251159171177923,
+      "learning_rate": 4.210609508705173e-06,
+      "loss": 0.4342,
+      "step": 15990
+    },
+    {
+      "epoch": 2.6261778170098333,
+      "grad_norm": 0.28275761112001435,
+      "learning_rate": 4.210141508834981e-06,
+      "loss": 0.4313,
+      "step": 15991
+    },
+    {
+      "epoch": 2.626342044218176,
+      "grad_norm": 0.3372588684132263,
+      "learning_rate": 4.2096735121129385e-06,
+      "loss": 0.4427,
+      "step": 15992
+    },
+    {
+      "epoch": 2.6265062714265186,
+      "grad_norm": 0.35546215576563495,
+      "learning_rate": 4.2092055185441265e-06,
+      "loss": 0.4253,
+      "step": 15993
+    },
+    {
+      "epoch": 2.6266704986348612,
+      "grad_norm": 0.28538441490936256,
+      "learning_rate": 4.2087375281336286e-06,
+      "loss": 0.4459,
+      "step": 15994
+    },
+    {
+      "epoch": 2.626834725843204,
+      "grad_norm": 0.3705717411953057,
+      "learning_rate": 4.208269540886529e-06,
+      "loss": 0.4422,
+      "step": 15995
+    },
+    {
+      "epoch": 2.626998953051547,
+      "grad_norm": 0.26329011035550054,
+      "learning_rate": 4.207801556807909e-06,
+      "loss": 0.4442,
+      "step": 15996
+    },
+    {
+      "epoch": 2.6271631802598896,
+      "grad_norm": 0.28256898887020987,
+      "learning_rate": 4.20733357590285e-06,
+      "loss": 0.4422,
+      "step": 15997
+    },
+    {
+      "epoch": 2.6273274074682322,
+      "grad_norm": 0.31725685114307306,
+      "learning_rate": 4.206865598176437e-06,
+      "loss": 0.4472,
+      "step": 15998
+    },
+    {
+      "epoch": 2.6274916346765753,
+      "grad_norm": 1.181339168715183,
+      "learning_rate": 4.206397623633753e-06,
+      "loss": 0.4324,
+      "step": 15999
+    },
+    {
+      "epoch": 2.627655861884918,
+      "grad_norm": 0.5335499658081728,
+      "learning_rate": 4.2059296522798786e-06,
+      "loss": 0.4335,
+      "step": 16000
+    },
+    {
+      "epoch": 2.633813827817476,
+      "grad_norm": 0.3458413336772324,
+      "learning_rate": 4.188362348384085e-06,
+      "loss": 0.4379,
+      "step": 16001
+    },
+    {
+      "epoch": 2.6339784120365923,
+      "grad_norm": 0.5975672929918889,
+      "learning_rate": 4.1878934149027705e-06,
+      "loss": 0.4447,
+      "step": 16002
+    },
+    {
+      "epoch": 2.634142996255709,
+      "grad_norm": 0.31596954107529784,
+      "learning_rate": 4.18742448482704e-06,
+      "loss": 0.43,
+      "step": 16003
+    },
+    {
+      "epoch": 2.6343075804748253,
+      "grad_norm": 0.29705223444507267,
+      "learning_rate": 4.186955558162006e-06,
+      "loss": 0.4309,
+      "step": 16004
+    },
+    {
+      "epoch": 2.6344721646939417,
+      "grad_norm": 0.3042036993024349,
+      "learning_rate": 4.186486634912789e-06,
+      "loss": 0.4447,
+      "step": 16005
+    },
+    {
+      "epoch": 2.634636748913058,
+      "grad_norm": 0.38759716050242937,
+      "learning_rate": 4.1860177150845045e-06,
+      "loss": 0.4292,
+      "step": 16006
+    },
+    {
+      "epoch": 2.6348013331321747,
+      "grad_norm": 0.31390985601371674,
+      "learning_rate": 4.18554879868227e-06,
+      "loss": 0.4378,
+      "step": 16007
+    },
+    {
+      "epoch": 2.634965917351291,
+      "grad_norm": 0.5655176381737559,
+      "learning_rate": 4.1850798857112e-06,
+      "loss": 0.4382,
+      "step": 16008
+    },
+    {
+      "epoch": 2.6351305015704076,
+      "grad_norm": 0.3223395703472826,
+      "learning_rate": 4.184610976176414e-06,
+      "loss": 0.438,
+      "step": 16009
+    },
+    {
+      "epoch": 2.635295085789524,
+      "grad_norm": 0.35687197663501297,
+      "learning_rate": 4.184142070083026e-06,
+      "loss": 0.4237,
+      "step": 16010
+    },
+    {
+      "epoch": 2.6354596700086406,
+      "grad_norm": 0.44101539197692596,
+      "learning_rate": 4.183673167436155e-06,
+      "loss": 0.4341,
+      "step": 16011
+    },
+    {
+      "epoch": 2.635624254227757,
+      "grad_norm": 0.3007372895231009,
+      "learning_rate": 4.183204268240916e-06,
+      "loss": 0.4288,
+      "step": 16012
+    },
+    {
+      "epoch": 2.6357888384468735,
+      "grad_norm": 0.29204721833278224,
+      "learning_rate": 4.182735372502424e-06,
+      "loss": 0.4309,
+      "step": 16013
+    },
+    {
+      "epoch": 2.63595342266599,
+      "grad_norm": 0.31142547055058784,
+      "learning_rate": 4.1822664802258e-06,
+      "loss": 0.4211,
+      "step": 16014
+    },
+    {
+      "epoch": 2.6361180068851064,
+      "grad_norm": 0.4347133243104596,
+      "learning_rate": 4.181797591416155e-06,
+      "loss": 0.4404,
+      "step": 16015
+    },
+    {
+      "epoch": 2.636282591104223,
+      "grad_norm": 0.5475754911371223,
+      "learning_rate": 4.181328706078609e-06,
+      "loss": 0.427,
+      "step": 16016
+    },
+    {
+      "epoch": 2.6364471753233394,
+      "grad_norm": 0.33465708743300165,
+      "learning_rate": 4.180859824218275e-06,
+      "loss": 0.441,
+      "step": 16017
+    },
+    {
+      "epoch": 2.636611759542456,
+      "grad_norm": 0.4214285639676177,
+      "learning_rate": 4.180390945840273e-06,
+      "loss": 0.4257,
+      "step": 16018
+    },
+    {
+      "epoch": 2.6367763437615723,
+      "grad_norm": 0.380826150533155,
+      "learning_rate": 4.179922070949717e-06,
+      "loss": 0.4514,
+      "step": 16019
+    },
+    {
+      "epoch": 2.636940927980689,
+      "grad_norm": 0.2875791514989647,
+      "learning_rate": 4.179453199551723e-06,
+      "loss": 0.4438,
+      "step": 16020
+    },
+    {
+      "epoch": 2.6371055121998053,
+      "grad_norm": 0.308429485865092,
+      "learning_rate": 4.178984331651408e-06,
+      "loss": 0.4246,
+      "step": 16021
+    },
+    {
+      "epoch": 2.6372700964189217,
+      "grad_norm": 0.346006434138322,
+      "learning_rate": 4.178515467253888e-06,
+      "loss": 0.4488,
+      "step": 16022
+    },
+    {
+      "epoch": 2.637434680638038,
+      "grad_norm": 0.2868870426572875,
+      "learning_rate": 4.1780466063642804e-06,
+      "loss": 0.4168,
+      "step": 16023
+    },
+    {
+      "epoch": 2.6375992648571547,
+      "grad_norm": 0.3518715966594303,
+      "learning_rate": 4.177577748987697e-06,
+      "loss": 0.4306,
+      "step": 16024
+    },
+    {
+      "epoch": 2.637763849076271,
+      "grad_norm": 0.33694302289305983,
+      "learning_rate": 4.177108895129259e-06,
+      "loss": 0.4469,
+      "step": 16025
+    },
+    {
+      "epoch": 2.6379284332953876,
+      "grad_norm": 0.2835088136502032,
+      "learning_rate": 4.176640044794079e-06,
+      "loss": 0.4265,
+      "step": 16026
+    },
+    {
+      "epoch": 2.638093017514504,
+      "grad_norm": 0.3304200094771688,
+      "learning_rate": 4.176171197987273e-06,
+      "loss": 0.4282,
+      "step": 16027
+    },
+    {
+      "epoch": 2.6382576017336206,
+      "grad_norm": 0.29655360915312234,
+      "learning_rate": 4.1757023547139585e-06,
+      "loss": 0.4258,
+      "step": 16028
+    },
+    {
+      "epoch": 2.638422185952737,
+      "grad_norm": 0.34291231739564487,
+      "learning_rate": 4.175233514979249e-06,
+      "loss": 0.4276,
+      "step": 16029
+    },
+    {
+      "epoch": 2.6385867701718535,
+      "grad_norm": 0.32533256752976253,
+      "learning_rate": 4.174764678788264e-06,
+      "loss": 0.4225,
+      "step": 16030
+    },
+    {
+      "epoch": 2.63875135439097,
+      "grad_norm": 0.27444339256916256,
+      "learning_rate": 4.174295846146115e-06,
+      "loss": 0.4422,
+      "step": 16031
+    },
+    {
+      "epoch": 2.6389159386100864,
+      "grad_norm": 0.3594351017234265,
+      "learning_rate": 4.173827017057922e-06,
+      "loss": 0.4176,
+      "step": 16032
+    },
+    {
+      "epoch": 2.639080522829203,
+      "grad_norm": 0.3259773673795588,
+      "learning_rate": 4.173358191528798e-06,
+      "loss": 0.4315,
+      "step": 16033
+    },
+    {
+      "epoch": 2.6392451070483194,
+      "grad_norm": 0.33864295381147363,
+      "learning_rate": 4.172889369563857e-06,
+      "loss": 0.4419,
+      "step": 16034
+    },
+    {
+      "epoch": 2.639409691267436,
+      "grad_norm": 0.29615433686097753,
+      "learning_rate": 4.172420551168219e-06,
+      "loss": 0.4399,
+      "step": 16035
+    },
+    {
+      "epoch": 2.6395742754865523,
+      "grad_norm": 0.47785437472432807,
+      "learning_rate": 4.171951736346995e-06,
+      "loss": 0.4334,
+      "step": 16036
+    },
+    {
+      "epoch": 2.639738859705669,
+      "grad_norm": 0.4004573990222095,
+      "learning_rate": 4.171482925105304e-06,
+      "loss": 0.4471,
+      "step": 16037
+    },
+    {
+      "epoch": 2.6399034439247853,
+      "grad_norm": 0.3143272720035237,
+      "learning_rate": 4.17101411744826e-06,
+      "loss": 0.4294,
+      "step": 16038
+    },
+    {
+      "epoch": 2.6400680281439017,
+      "grad_norm": 0.35269234949253736,
+      "learning_rate": 4.170545313380978e-06,
+      "loss": 0.4045,
+      "step": 16039
+    },
+    {
+      "epoch": 2.6402326123630178,
+      "grad_norm": 0.340001284874301,
+      "learning_rate": 4.1700765129085754e-06,
+      "loss": 0.4191,
+      "step": 16040
+    },
+    {
+      "epoch": 2.6403971965821342,
+      "grad_norm": 0.41295947541357814,
+      "learning_rate": 4.169607716036165e-06,
+      "loss": 0.4088,
+      "step": 16041
+    },
+    {
+      "epoch": 2.6405617808012507,
+      "grad_norm": 0.31813500109467485,
+      "learning_rate": 4.169138922768865e-06,
+      "loss": 0.4433,
+      "step": 16042
+    },
+    {
+      "epoch": 2.640726365020367,
+      "grad_norm": 0.31553983739624203,
+      "learning_rate": 4.168670133111787e-06,
+      "loss": 0.4269,
+      "step": 16043
+    },
+    {
+      "epoch": 2.6408909492394836,
+      "grad_norm": 0.28908417796027375,
+      "learning_rate": 4.16820134707005e-06,
+      "loss": 0.4391,
+      "step": 16044
+    },
+    {
+      "epoch": 2.6410555334586,
+      "grad_norm": 0.25498288846337946,
+      "learning_rate": 4.1677325646487644e-06,
+      "loss": 0.4344,
+      "step": 16045
+    },
+    {
+      "epoch": 2.6412201176777166,
+      "grad_norm": 0.9148740161576957,
+      "learning_rate": 4.167263785853052e-06,
+      "loss": 0.4397,
+      "step": 16046
+    },
+    {
+      "epoch": 2.641384701896833,
+      "grad_norm": 0.3654250205826132,
+      "learning_rate": 4.166795010688024e-06,
+      "loss": 0.437,
+      "step": 16047
+    },
+    {
+      "epoch": 2.6415492861159495,
+      "grad_norm": 0.2839790292083236,
+      "learning_rate": 4.166326239158794e-06,
+      "loss": 0.4405,
+      "step": 16048
+    },
+    {
+      "epoch": 2.641713870335066,
+      "grad_norm": 0.2566802127035541,
+      "learning_rate": 4.16585747127048e-06,
+      "loss": 0.4223,
+      "step": 16049
+    },
+    {
+      "epoch": 2.6418784545541825,
+      "grad_norm": 0.3592073497185888,
+      "learning_rate": 4.165388707028195e-06,
+      "loss": 0.4197,
+      "step": 16050
+    },
+    {
+      "epoch": 2.642043038773299,
+      "grad_norm": 0.296672179833109,
+      "learning_rate": 4.164919946437056e-06,
+      "loss": 0.4494,
+      "step": 16051
+    },
+    {
+      "epoch": 2.6422076229924154,
+      "grad_norm": 0.26656374733013644,
+      "learning_rate": 4.164451189502176e-06,
+      "loss": 0.4348,
+      "step": 16052
+    },
+    {
+      "epoch": 2.642372207211532,
+      "grad_norm": 0.8647309996086165,
+      "learning_rate": 4.163982436228672e-06,
+      "loss": 0.424,
+      "step": 16053
+    },
+    {
+      "epoch": 2.6425367914306483,
+      "grad_norm": 0.2867680259115217,
+      "learning_rate": 4.163513686621655e-06,
+      "loss": 0.4553,
+      "step": 16054
+    },
+    {
+      "epoch": 2.642701375649765,
+      "grad_norm": 0.2975334360287579,
+      "learning_rate": 4.163044940686246e-06,
+      "loss": 0.4421,
+      "step": 16055
+    },
+    {
+      "epoch": 2.6428659598688813,
+      "grad_norm": 0.4094481182342718,
+      "learning_rate": 4.162576198427554e-06,
+      "loss": 0.4242,
+      "step": 16056
+    },
+    {
+      "epoch": 2.6430305440879978,
+      "grad_norm": 0.319781031677507,
+      "learning_rate": 4.162107459850696e-06,
+      "loss": 0.4294,
+      "step": 16057
+    },
+    {
+      "epoch": 2.6431951283071142,
+      "grad_norm": 0.29254963141456264,
+      "learning_rate": 4.1616387249607865e-06,
+      "loss": 0.4415,
+      "step": 16058
+    },
+    {
+      "epoch": 2.6433597125262307,
+      "grad_norm": 0.3342593584953292,
+      "learning_rate": 4.1611699937629395e-06,
+      "loss": 0.4362,
+      "step": 16059
+    },
+    {
+      "epoch": 2.643524296745347,
+      "grad_norm": 0.29622533476076573,
+      "learning_rate": 4.160701266262272e-06,
+      "loss": 0.4041,
+      "step": 16060
+    },
+    {
+      "epoch": 2.6436888809644636,
+      "grad_norm": 0.29560361745977387,
+      "learning_rate": 4.160232542463895e-06,
+      "loss": 0.4316,
+      "step": 16061
+    },
+    {
+      "epoch": 2.64385346518358,
+      "grad_norm": 0.34481303444827943,
+      "learning_rate": 4.159763822372926e-06,
+      "loss": 0.4565,
+      "step": 16062
+    },
+    {
+      "epoch": 2.6440180494026966,
+      "grad_norm": 0.44701790743094155,
+      "learning_rate": 4.15929510599448e-06,
+      "loss": 0.4322,
+      "step": 16063
+    },
+    {
+      "epoch": 2.644182633621813,
+      "grad_norm": 0.38441840228187624,
+      "learning_rate": 4.158826393333666e-06,
+      "loss": 0.4308,
+      "step": 16064
+    },
+    {
+      "epoch": 2.644347217840929,
+      "grad_norm": 0.27856025851449084,
+      "learning_rate": 4.158357684395606e-06,
+      "loss": 0.4557,
+      "step": 16065
+    },
+    {
+      "epoch": 2.6445118020600455,
+      "grad_norm": 0.346035538895123,
+      "learning_rate": 4.1578889791854074e-06,
+      "loss": 0.4371,
+      "step": 16066
+    },
+    {
+      "epoch": 2.644676386279162,
+      "grad_norm": 0.29522255544530657,
+      "learning_rate": 4.15742027770819e-06,
+      "loss": 0.4429,
+      "step": 16067
+    },
+    {
+      "epoch": 2.6448409704982785,
+      "grad_norm": 0.2568408699069682,
+      "learning_rate": 4.156951579969064e-06,
+      "loss": 0.4511,
+      "step": 16068
+    },
+    {
+      "epoch": 2.645005554717395,
+      "grad_norm": 0.2795226041069509,
+      "learning_rate": 4.156482885973147e-06,
+      "loss": 0.4317,
+      "step": 16069
+    },
+    {
+      "epoch": 2.6451701389365114,
+      "grad_norm": 0.41437289939438915,
+      "learning_rate": 4.156014195725552e-06,
+      "loss": 0.4471,
+      "step": 16070
+    },
+    {
+      "epoch": 2.645334723155628,
+      "grad_norm": 0.37249505618644185,
+      "learning_rate": 4.155545509231391e-06,
+      "loss": 0.4286,
+      "step": 16071
+    },
+    {
+      "epoch": 2.6454993073747444,
+      "grad_norm": 0.36242228653315584,
+      "learning_rate": 4.155076826495783e-06,
+      "loss": 0.439,
+      "step": 16072
+    },
+    {
+      "epoch": 2.645663891593861,
+      "grad_norm": 0.31756610529430523,
+      "learning_rate": 4.154608147523834e-06,
+      "loss": 0.4273,
+      "step": 16073
+    },
+    {
+      "epoch": 2.6458284758129773,
+      "grad_norm": 0.32159274867073534,
+      "learning_rate": 4.154139472320668e-06,
+      "loss": 0.4537,
+      "step": 16074
+    },
+    {
+      "epoch": 2.645993060032094,
+      "grad_norm": 0.335271183359619,
+      "learning_rate": 4.153670800891389e-06,
+      "loss": 0.4422,
+      "step": 16075
+    },
+    {
+      "epoch": 2.6461576442512103,
+      "grad_norm": 0.3287060062641479,
+      "learning_rate": 4.153202133241121e-06,
+      "loss": 0.4472,
+      "step": 16076
+    },
+    {
+      "epoch": 2.6463222284703267,
+      "grad_norm": 0.3426461142941525,
+      "learning_rate": 4.1527334693749716e-06,
+      "loss": 0.4241,
+      "step": 16077
+    },
+    {
+      "epoch": 2.646486812689443,
+      "grad_norm": 0.29177189773366047,
+      "learning_rate": 4.1522648092980546e-06,
+      "loss": 0.4293,
+      "step": 16078
+    },
+    {
+      "epoch": 2.6466513969085597,
+      "grad_norm": 0.35057317490047024,
+      "learning_rate": 4.151796153015486e-06,
+      "loss": 0.4336,
+      "step": 16079
+    },
+    {
+      "epoch": 2.646815981127676,
+      "grad_norm": 0.3733394546466849,
+      "learning_rate": 4.151327500532379e-06,
+      "loss": 0.4339,
+      "step": 16080
+    },
+    {
+      "epoch": 2.6469805653467926,
+      "grad_norm": 0.3274339998882645,
+      "learning_rate": 4.150858851853847e-06,
+      "loss": 0.4353,
+      "step": 16081
+    },
+    {
+      "epoch": 2.647145149565909,
+      "grad_norm": 0.39048201119481457,
+      "learning_rate": 4.150390206985002e-06,
+      "loss": 0.4288,
+      "step": 16082
+    },
+    {
+      "epoch": 2.6473097337850255,
+      "grad_norm": 0.2890017641382168,
+      "learning_rate": 4.149921565930962e-06,
+      "loss": 0.4285,
+      "step": 16083
+    },
+    {
+      "epoch": 2.647474318004142,
+      "grad_norm": 0.3689520037974736,
+      "learning_rate": 4.149452928696839e-06,
+      "loss": 0.447,
+      "step": 16084
+    },
+    {
+      "epoch": 2.6476389022232585,
+      "grad_norm": 0.32192762270808284,
+      "learning_rate": 4.148984295287743e-06,
+      "loss": 0.4378,
+      "step": 16085
+    },
+    {
+      "epoch": 2.647803486442375,
+      "grad_norm": 0.34372293910344526,
+      "learning_rate": 4.148515665708792e-06,
+      "loss": 0.4493,
+      "step": 16086
+    },
+    {
+      "epoch": 2.6479680706614914,
+      "grad_norm": 0.3676412358462734,
+      "learning_rate": 4.148047039965097e-06,
+      "loss": 0.4115,
+      "step": 16087
+    },
+    {
+      "epoch": 2.648132654880608,
+      "grad_norm": 0.32314400463920195,
+      "learning_rate": 4.147578418061772e-06,
+      "loss": 0.4302,
+      "step": 16088
+    },
+    {
+      "epoch": 2.6482972390997244,
+      "grad_norm": 0.32951777235338797,
+      "learning_rate": 4.147109800003931e-06,
+      "loss": 0.432,
+      "step": 16089
+    },
+    {
+      "epoch": 2.648461823318841,
+      "grad_norm": 0.4144155570458211,
+      "learning_rate": 4.146641185796687e-06,
+      "loss": 0.4526,
+      "step": 16090
+    },
+    {
+      "epoch": 2.6486264075379573,
+      "grad_norm": 0.32492391828837874,
+      "learning_rate": 4.146172575445153e-06,
+      "loss": 0.4171,
+      "step": 16091
+    },
+    {
+      "epoch": 2.6487909917570738,
+      "grad_norm": 0.32096250114278707,
+      "learning_rate": 4.1457039689544435e-06,
+      "loss": 0.4263,
+      "step": 16092
+    },
+    {
+      "epoch": 2.6489555759761902,
+      "grad_norm": 0.2931596596890427,
+      "learning_rate": 4.1452353663296715e-06,
+      "loss": 0.4177,
+      "step": 16093
+    },
+    {
+      "epoch": 2.6491201601953067,
+      "grad_norm": 0.42602171281467677,
+      "learning_rate": 4.144766767575947e-06,
+      "loss": 0.4397,
+      "step": 16094
+    },
+    {
+      "epoch": 2.649284744414423,
+      "grad_norm": 0.31803480047796434,
+      "learning_rate": 4.144298172698389e-06,
+      "loss": 0.4048,
+      "step": 16095
+    },
+    {
+      "epoch": 2.6494493286335397,
+      "grad_norm": 0.3114085359245422,
+      "learning_rate": 4.143829581702105e-06,
+      "loss": 0.4518,
+      "step": 16096
+    },
+    {
+      "epoch": 2.649613912852656,
+      "grad_norm": 0.27338831216461773,
+      "learning_rate": 4.143360994592211e-06,
+      "loss": 0.4443,
+      "step": 16097
+    },
+    {
+      "epoch": 2.6497784970717726,
+      "grad_norm": 0.3096832549208866,
+      "learning_rate": 4.14289241137382e-06,
+      "loss": 0.431,
+      "step": 16098
+    },
+    {
+      "epoch": 2.649943081290889,
+      "grad_norm": 0.570367929625992,
+      "learning_rate": 4.142423832052045e-06,
+      "loss": 0.433,
+      "step": 16099
+    },
+    {
+      "epoch": 2.6501076655100055,
+      "grad_norm": 0.3231277758544926,
+      "learning_rate": 4.141955256631997e-06,
+      "loss": 0.4274,
+      "step": 16100
+    },
+    {
+      "epoch": 2.650272249729122,
+      "grad_norm": 0.31080107211903213,
+      "learning_rate": 4.14148668511879e-06,
+      "loss": 0.4327,
+      "step": 16101
+    },
+    {
+      "epoch": 2.6504368339482385,
+      "grad_norm": 0.4623297273803156,
+      "learning_rate": 4.141018117517539e-06,
+      "loss": 0.4279,
+      "step": 16102
+    },
+    {
+      "epoch": 2.650601418167355,
+      "grad_norm": 0.26247998589542876,
+      "learning_rate": 4.1405495538333534e-06,
+      "loss": 0.4313,
+      "step": 16103
+    },
+    {
+      "epoch": 2.6507660023864714,
+      "grad_norm": 0.29532370622536497,
+      "learning_rate": 4.14008099407135e-06,
+      "loss": 0.4159,
+      "step": 16104
+    },
+    {
+      "epoch": 2.650930586605588,
+      "grad_norm": 0.41691707860407345,
+      "learning_rate": 4.139612438236636e-06,
+      "loss": 0.4391,
+      "step": 16105
+    },
+    {
+      "epoch": 2.6510951708247044,
+      "grad_norm": 0.4613190154024909,
+      "learning_rate": 4.13914388633433e-06,
+      "loss": 0.4255,
+      "step": 16106
+    },
+    {
+      "epoch": 2.6512597550438204,
+      "grad_norm": 0.2797038181019288,
+      "learning_rate": 4.138675338369541e-06,
+      "loss": 0.4285,
+      "step": 16107
+    },
+    {
+      "epoch": 2.651424339262937,
+      "grad_norm": 0.2871785210367013,
+      "learning_rate": 4.138206794347381e-06,
+      "loss": 0.4277,
+      "step": 16108
+    },
+    {
+      "epoch": 2.6515889234820533,
+      "grad_norm": 0.2833375328201389,
+      "learning_rate": 4.137738254272966e-06,
+      "loss": 0.4306,
+      "step": 16109
+    },
+    {
+      "epoch": 2.65175350770117,
+      "grad_norm": 0.2791742154687719,
+      "learning_rate": 4.137269718151405e-06,
+      "loss": 0.4182,
+      "step": 16110
+    },
+    {
+      "epoch": 2.6519180919202863,
+      "grad_norm": 0.2977754878777086,
+      "learning_rate": 4.1368011859878135e-06,
+      "loss": 0.4469,
+      "step": 16111
+    },
+    {
+      "epoch": 2.6520826761394027,
+      "grad_norm": 0.37579617600776255,
+      "learning_rate": 4.136332657787302e-06,
+      "loss": 0.4436,
+      "step": 16112
+    },
+    {
+      "epoch": 2.652247260358519,
+      "grad_norm": 0.2968274348201199,
+      "learning_rate": 4.135864133554983e-06,
+      "loss": 0.4178,
+      "step": 16113
+    },
+    {
+      "epoch": 2.6524118445776357,
+      "grad_norm": 0.2964357275876666,
+      "learning_rate": 4.135395613295971e-06,
+      "loss": 0.4285,
+      "step": 16114
+    },
+    {
+      "epoch": 2.652576428796752,
+      "grad_norm": 0.26860658490969397,
+      "learning_rate": 4.134927097015373e-06,
+      "loss": 0.4483,
+      "step": 16115
+    },
+    {
+      "epoch": 2.6527410130158686,
+      "grad_norm": 0.34189626645070365,
+      "learning_rate": 4.134458584718309e-06,
+      "loss": 0.4368,
+      "step": 16116
+    },
+    {
+      "epoch": 2.652905597234985,
+      "grad_norm": 0.2779968278720606,
+      "learning_rate": 4.133990076409884e-06,
+      "loss": 0.4249,
+      "step": 16117
+    },
+    {
+      "epoch": 2.6530701814541016,
+      "grad_norm": 0.30593249193352895,
+      "learning_rate": 4.133521572095214e-06,
+      "loss": 0.4443,
+      "step": 16118
+    },
+    {
+      "epoch": 2.653234765673218,
+      "grad_norm": 0.2854793229772029,
+      "learning_rate": 4.133053071779411e-06,
+      "loss": 0.4286,
+      "step": 16119
+    },
+    {
+      "epoch": 2.6533993498923345,
+      "grad_norm": 0.29006974883980136,
+      "learning_rate": 4.1325845754675856e-06,
+      "loss": 0.4163,
+      "step": 16120
+    },
+    {
+      "epoch": 2.653563934111451,
+      "grad_norm": 0.3077942187428335,
+      "learning_rate": 4.132116083164851e-06,
+      "loss": 0.45,
+      "step": 16121
+    },
+    {
+      "epoch": 2.6537285183305674,
+      "grad_norm": 0.3408742741710642,
+      "learning_rate": 4.131647594876319e-06,
+      "loss": 0.4492,
+      "step": 16122
+    },
+    {
+      "epoch": 2.653893102549684,
+      "grad_norm": 0.2926827150992984,
+      "learning_rate": 4.1311791106071026e-06,
+      "loss": 0.4253,
+      "step": 16123
+    },
+    {
+      "epoch": 2.6540576867688004,
+      "grad_norm": 0.3155775845434203,
+      "learning_rate": 4.13071063036231e-06,
+      "loss": 0.4121,
+      "step": 16124
+    },
+    {
+      "epoch": 2.654222270987917,
+      "grad_norm": 0.2646331437379279,
+      "learning_rate": 4.130242154147058e-06,
+      "loss": 0.4217,
+      "step": 16125
+    },
+    {
+      "epoch": 2.6543868552070333,
+      "grad_norm": 0.26323728951390457,
+      "learning_rate": 4.129773681966453e-06,
+      "loss": 0.4286,
+      "step": 16126
+    },
+    {
+      "epoch": 2.65455143942615,
+      "grad_norm": 0.4258820615094784,
+      "learning_rate": 4.129305213825614e-06,
+      "loss": 0.4105,
+      "step": 16127
+    },
+    {
+      "epoch": 2.6547160236452663,
+      "grad_norm": 0.2896312674933501,
+      "learning_rate": 4.128836749729646e-06,
+      "loss": 0.4373,
+      "step": 16128
+    },
+    {
+      "epoch": 2.6548806078643827,
+      "grad_norm": 0.31894235570171264,
+      "learning_rate": 4.128368289683663e-06,
+      "loss": 0.4397,
+      "step": 16129
+    },
+    {
+      "epoch": 2.655045192083499,
+      "grad_norm": 0.31359654504208,
+      "learning_rate": 4.127899833692778e-06,
+      "loss": 0.442,
+      "step": 16130
+    },
+    {
+      "epoch": 2.6552097763026157,
+      "grad_norm": 0.3494845795600111,
+      "learning_rate": 4.1274313817621e-06,
+      "loss": 0.4355,
+      "step": 16131
+    },
+    {
+      "epoch": 2.6553743605217317,
+      "grad_norm": 0.3489366047251084,
+      "learning_rate": 4.126962933896744e-06,
+      "loss": 0.4342,
+      "step": 16132
+    },
+    {
+      "epoch": 2.655538944740848,
+      "grad_norm": 0.34060977168449236,
+      "learning_rate": 4.126494490101818e-06,
+      "loss": 0.4374,
+      "step": 16133
+    },
+    {
+      "epoch": 2.6557035289599646,
+      "grad_norm": 0.2742126576753534,
+      "learning_rate": 4.126026050382436e-06,
+      "loss": 0.4451,
+      "step": 16134
+    },
+    {
+      "epoch": 2.655868113179081,
+      "grad_norm": 0.2992707000967294,
+      "learning_rate": 4.125557614743707e-06,
+      "loss": 0.4185,
+      "step": 16135
+    },
+    {
+      "epoch": 2.6560326973981976,
+      "grad_norm": 0.2877101811641106,
+      "learning_rate": 4.125089183190747e-06,
+      "loss": 0.4328,
+      "step": 16136
+    },
+    {
+      "epoch": 2.656197281617314,
+      "grad_norm": 0.4447724590169303,
+      "learning_rate": 4.124620755728663e-06,
+      "loss": 0.4457,
+      "step": 16137
+    },
+    {
+      "epoch": 2.6563618658364305,
+      "grad_norm": 0.28823615255434304,
+      "learning_rate": 4.124152332362565e-06,
+      "loss": 0.4357,
+      "step": 16138
+    },
+    {
+      "epoch": 2.656526450055547,
+      "grad_norm": 0.2877010704372036,
+      "learning_rate": 4.12368391309757e-06,
+      "loss": 0.4209,
+      "step": 16139
+    },
+    {
+      "epoch": 2.6566910342746635,
+      "grad_norm": 1.2348588948682877,
+      "learning_rate": 4.123215497938783e-06,
+      "loss": 0.434,
+      "step": 16140
+    },
+    {
+      "epoch": 2.65685561849378,
+      "grad_norm": 0.3176384519860354,
+      "learning_rate": 4.12274708689132e-06,
+      "loss": 0.4248,
+      "step": 16141
+    },
+    {
+      "epoch": 2.6570202027128964,
+      "grad_norm": 0.34314293250836697,
+      "learning_rate": 4.1222786799602895e-06,
+      "loss": 0.4255,
+      "step": 16142
+    },
+    {
+      "epoch": 2.657184786932013,
+      "grad_norm": 0.32065602089046813,
+      "learning_rate": 4.121810277150804e-06,
+      "loss": 0.4327,
+      "step": 16143
+    },
+    {
+      "epoch": 2.6573493711511293,
+      "grad_norm": 0.31641949004858255,
+      "learning_rate": 4.121341878467975e-06,
+      "loss": 0.4359,
+      "step": 16144
+    },
+    {
+      "epoch": 2.657513955370246,
+      "grad_norm": 0.31985622093017013,
+      "learning_rate": 4.120873483916909e-06,
+      "loss": 0.4469,
+      "step": 16145
+    },
+    {
+      "epoch": 2.6576785395893623,
+      "grad_norm": 0.4740004344214797,
+      "learning_rate": 4.1204050935027235e-06,
+      "loss": 0.4051,
+      "step": 16146
+    },
+    {
+      "epoch": 2.6578431238084788,
+      "grad_norm": 0.3047611687392972,
+      "learning_rate": 4.119936707230524e-06,
+      "loss": 0.4228,
+      "step": 16147
+    },
+    {
+      "epoch": 2.6580077080275952,
+      "grad_norm": 0.35193746954809046,
+      "learning_rate": 4.1194683251054246e-06,
+      "loss": 0.4341,
+      "step": 16148
+    },
+    {
+      "epoch": 2.6581722922467117,
+      "grad_norm": 0.4636378951303919,
+      "learning_rate": 4.118999947132533e-06,
+      "loss": 0.4312,
+      "step": 16149
+    },
+    {
+      "epoch": 2.658336876465828,
+      "grad_norm": 0.32009910134700315,
+      "learning_rate": 4.1185315733169645e-06,
+      "loss": 0.4396,
+      "step": 16150
+    },
+    {
+      "epoch": 2.6585014606849446,
+      "grad_norm": 0.32073795597582255,
+      "learning_rate": 4.1180632036638265e-06,
+      "loss": 0.4309,
+      "step": 16151
+    },
+    {
+      "epoch": 2.658666044904061,
+      "grad_norm": 0.4289503976591517,
+      "learning_rate": 4.117594838178229e-06,
+      "loss": 0.4286,
+      "step": 16152
+    },
+    {
+      "epoch": 2.6588306291231776,
+      "grad_norm": 0.3596741397311703,
+      "learning_rate": 4.1171264768652855e-06,
+      "loss": 0.4439,
+      "step": 16153
+    },
+    {
+      "epoch": 2.658995213342294,
+      "grad_norm": 0.35541758817558256,
+      "learning_rate": 4.1166581197301036e-06,
+      "loss": 0.4418,
+      "step": 16154
+    },
+    {
+      "epoch": 2.6591597975614105,
+      "grad_norm": 0.32675505943073413,
+      "learning_rate": 4.116189766777797e-06,
+      "loss": 0.4353,
+      "step": 16155
+    },
+    {
+      "epoch": 2.659324381780527,
+      "grad_norm": 0.3698642182004654,
+      "learning_rate": 4.115721418013473e-06,
+      "loss": 0.4203,
+      "step": 16156
+    },
+    {
+      "epoch": 2.6594889659996435,
+      "grad_norm": 0.6317514981473367,
+      "learning_rate": 4.115253073442245e-06,
+      "loss": 0.4277,
+      "step": 16157
+    },
+    {
+      "epoch": 2.65965355021876,
+      "grad_norm": 0.3042933450838122,
+      "learning_rate": 4.114784733069222e-06,
+      "loss": 0.4547,
+      "step": 16158
+    },
+    {
+      "epoch": 2.6598181344378764,
+      "grad_norm": 0.5519804724689971,
+      "learning_rate": 4.114316396899513e-06,
+      "loss": 0.4268,
+      "step": 16159
+    },
+    {
+      "epoch": 2.659982718656993,
+      "grad_norm": 0.3026651330178726,
+      "learning_rate": 4.11384806493823e-06,
+      "loss": 0.4267,
+      "step": 16160
+    },
+    {
+      "epoch": 2.6601473028761093,
+      "grad_norm": 0.2896548638142162,
+      "learning_rate": 4.113379737190482e-06,
+      "loss": 0.434,
+      "step": 16161
+    },
+    {
+      "epoch": 2.660311887095226,
+      "grad_norm": 0.3162773316354961,
+      "learning_rate": 4.112911413661382e-06,
+      "loss": 0.4183,
+      "step": 16162
+    },
+    {
+      "epoch": 2.6604764713143423,
+      "grad_norm": 0.30438307843708645,
+      "learning_rate": 4.112443094356036e-06,
+      "loss": 0.4222,
+      "step": 16163
+    },
+    {
+      "epoch": 2.6606410555334588,
+      "grad_norm": 0.325295396745359,
+      "learning_rate": 4.111974779279558e-06,
+      "loss": 0.4417,
+      "step": 16164
+    },
+    {
+      "epoch": 2.6608056397525752,
+      "grad_norm": 0.289759415015586,
+      "learning_rate": 4.111506468437057e-06,
+      "loss": 0.4367,
+      "step": 16165
+    },
+    {
+      "epoch": 2.6609702239716917,
+      "grad_norm": 0.2611358156919582,
+      "learning_rate": 4.11103816183364e-06,
+      "loss": 0.4349,
+      "step": 16166
+    },
+    {
+      "epoch": 2.661134808190808,
+      "grad_norm": 0.30681190236477174,
+      "learning_rate": 4.110569859474421e-06,
+      "loss": 0.4151,
+      "step": 16167
+    },
+    {
+      "epoch": 2.6612993924099246,
+      "grad_norm": 0.3179901230899138,
+      "learning_rate": 4.110101561364506e-06,
+      "loss": 0.4361,
+      "step": 16168
+    },
+    {
+      "epoch": 2.661463976629041,
+      "grad_norm": 0.3743547215570952,
+      "learning_rate": 4.109633267509009e-06,
+      "loss": 0.4457,
+      "step": 16169
+    },
+    {
+      "epoch": 2.6616285608481576,
+      "grad_norm": 0.3155628139691341,
+      "learning_rate": 4.109164977913037e-06,
+      "loss": 0.4396,
+      "step": 16170
+    },
+    {
+      "epoch": 2.661793145067274,
+      "grad_norm": 0.27701818933684635,
+      "learning_rate": 4.108696692581702e-06,
+      "loss": 0.4359,
+      "step": 16171
+    },
+    {
+      "epoch": 2.6619577292863905,
+      "grad_norm": 0.3668374170702665,
+      "learning_rate": 4.108228411520113e-06,
+      "loss": 0.4368,
+      "step": 16172
+    },
+    {
+      "epoch": 2.662122313505507,
+      "grad_norm": 0.3149660241315818,
+      "learning_rate": 4.1077601347333775e-06,
+      "loss": 0.4477,
+      "step": 16173
+    },
+    {
+      "epoch": 2.662286897724623,
+      "grad_norm": 0.36031363805868644,
+      "learning_rate": 4.107291862226608e-06,
+      "loss": 0.4386,
+      "step": 16174
+    },
+    {
+      "epoch": 2.6624514819437395,
+      "grad_norm": 0.3752687445537029,
+      "learning_rate": 4.106823594004912e-06,
+      "loss": 0.4326,
+      "step": 16175
+    },
+    {
+      "epoch": 2.662616066162856,
+      "grad_norm": 0.33165849121813185,
+      "learning_rate": 4.106355330073402e-06,
+      "loss": 0.4155,
+      "step": 16176
+    },
+    {
+      "epoch": 2.6627806503819724,
+      "grad_norm": 0.2996559896507021,
+      "learning_rate": 4.105887070437182e-06,
+      "loss": 0.4407,
+      "step": 16177
+    },
+    {
+      "epoch": 2.662945234601089,
+      "grad_norm": 0.3691651742015506,
+      "learning_rate": 4.1054188151013685e-06,
+      "loss": 0.4311,
+      "step": 16178
+    },
+    {
+      "epoch": 2.6631098188202054,
+      "grad_norm": 0.37706408757539167,
+      "learning_rate": 4.1049505640710655e-06,
+      "loss": 0.4414,
+      "step": 16179
+    },
+    {
+      "epoch": 2.663274403039322,
+      "grad_norm": 0.4580061387375138,
+      "learning_rate": 4.104482317351386e-06,
+      "loss": 0.425,
+      "step": 16180
+    },
+    {
+      "epoch": 2.6634389872584383,
+      "grad_norm": 0.3982468103731691,
+      "learning_rate": 4.104014074947436e-06,
+      "loss": 0.4176,
+      "step": 16181
+    },
+    {
+      "epoch": 2.6636035714775548,
+      "grad_norm": 0.31071724356969327,
+      "learning_rate": 4.103545836864326e-06,
+      "loss": 0.4289,
+      "step": 16182
+    },
+    {
+      "epoch": 2.6637681556966712,
+      "grad_norm": 0.3174287098929874,
+      "learning_rate": 4.103077603107167e-06,
+      "loss": 0.4249,
+      "step": 16183
+    },
+    {
+      "epoch": 2.6639327399157877,
+      "grad_norm": 0.28455613214851727,
+      "learning_rate": 4.102609373681066e-06,
+      "loss": 0.4451,
+      "step": 16184
+    },
+    {
+      "epoch": 2.664097324134904,
+      "grad_norm": 0.3196335886895791,
+      "learning_rate": 4.102141148591134e-06,
+      "loss": 0.4226,
+      "step": 16185
+    },
+    {
+      "epoch": 2.6642619083540207,
+      "grad_norm": 0.6233557455696098,
+      "learning_rate": 4.1016729278424765e-06,
+      "loss": 0.4241,
+      "step": 16186
+    },
+    {
+      "epoch": 2.664426492573137,
+      "grad_norm": 0.33200426377694864,
+      "learning_rate": 4.101204711440209e-06,
+      "loss": 0.4439,
+      "step": 16187
+    },
+    {
+      "epoch": 2.6645910767922536,
+      "grad_norm": 0.3156950309931716,
+      "learning_rate": 4.100736499389434e-06,
+      "loss": 0.4652,
+      "step": 16188
+    },
+    {
+      "epoch": 2.66475566101137,
+      "grad_norm": 0.29051265138452365,
+      "learning_rate": 4.100268291695262e-06,
+      "loss": 0.4058,
+      "step": 16189
+    },
+    {
+      "epoch": 2.6649202452304865,
+      "grad_norm": 0.3876607898144007,
+      "learning_rate": 4.099800088362805e-06,
+      "loss": 0.4336,
+      "step": 16190
+    },
+    {
+      "epoch": 2.665084829449603,
+      "grad_norm": 0.2801680526667986,
+      "learning_rate": 4.099331889397168e-06,
+      "loss": 0.4409,
+      "step": 16191
+    },
+    {
+      "epoch": 2.6652494136687195,
+      "grad_norm": 0.3928005012139743,
+      "learning_rate": 4.098863694803462e-06,
+      "loss": 0.4452,
+      "step": 16192
+    },
+    {
+      "epoch": 2.665413997887836,
+      "grad_norm": 0.31667327209953494,
+      "learning_rate": 4.098395504586795e-06,
+      "loss": 0.4314,
+      "step": 16193
+    },
+    {
+      "epoch": 2.6655785821069524,
+      "grad_norm": 0.4095252931530756,
+      "learning_rate": 4.0979273187522775e-06,
+      "loss": 0.4398,
+      "step": 16194
+    },
+    {
+      "epoch": 2.665743166326069,
+      "grad_norm": 0.3039674024879263,
+      "learning_rate": 4.097459137305016e-06,
+      "loss": 0.4303,
+      "step": 16195
+    },
+    {
+      "epoch": 2.6659077505451854,
+      "grad_norm": 0.3846388319882331,
+      "learning_rate": 4.096990960250118e-06,
+      "loss": 0.4087,
+      "step": 16196
+    },
+    {
+      "epoch": 2.666072334764302,
+      "grad_norm": 0.36933999442641624,
+      "learning_rate": 4.096522787592697e-06,
+      "loss": 0.435,
+      "step": 16197
+    },
+    {
+      "epoch": 2.6662369189834183,
+      "grad_norm": 0.26155978929311113,
+      "learning_rate": 4.0960546193378555e-06,
+      "loss": 0.4179,
+      "step": 16198
+    },
+    {
+      "epoch": 2.6664015032025343,
+      "grad_norm": 0.2824416471224022,
+      "learning_rate": 4.095586455490706e-06,
+      "loss": 0.4447,
+      "step": 16199
+    },
+    {
+      "epoch": 2.666566087421651,
+      "grad_norm": 0.3029193693644384,
+      "learning_rate": 4.095118296056355e-06,
+      "loss": 0.4314,
+      "step": 16200
+    },
+    {
+      "epoch": 2.6667306716407673,
+      "grad_norm": 0.5897728845044921,
+      "learning_rate": 4.0946501410399124e-06,
+      "loss": 0.4221,
+      "step": 16201
+    },
+    {
+      "epoch": 2.6668952558598837,
+      "grad_norm": 0.3111304963439345,
+      "learning_rate": 4.094181990446486e-06,
+      "loss": 0.4532,
+      "step": 16202
+    },
+    {
+      "epoch": 2.667059840079,
+      "grad_norm": 0.3465409814026947,
+      "learning_rate": 4.093713844281182e-06,
+      "loss": 0.4544,
+      "step": 16203
+    },
+    {
+      "epoch": 2.6672244242981167,
+      "grad_norm": 0.2985915021659544,
+      "learning_rate": 4.093245702549113e-06,
+      "loss": 0.4428,
+      "step": 16204
+    },
+    {
+      "epoch": 2.667389008517233,
+      "grad_norm": 0.3219223204177759,
+      "learning_rate": 4.092777565255381e-06,
+      "loss": 0.443,
+      "step": 16205
+    },
+    {
+      "epoch": 2.6675535927363496,
+      "grad_norm": 0.3190977768020272,
+      "learning_rate": 4.092309432405101e-06,
+      "loss": 0.4039,
+      "step": 16206
+    },
+    {
+      "epoch": 2.667718176955466,
+      "grad_norm": 0.29823725676161056,
+      "learning_rate": 4.091841304003376e-06,
+      "loss": 0.4328,
+      "step": 16207
+    },
+    {
+      "epoch": 2.6678827611745826,
+      "grad_norm": 0.3483582913118178,
+      "learning_rate": 4.091373180055317e-06,
+      "loss": 0.4572,
+      "step": 16208
+    },
+    {
+      "epoch": 2.668047345393699,
+      "grad_norm": 0.2992140663455201,
+      "learning_rate": 4.09090506056603e-06,
+      "loss": 0.4164,
+      "step": 16209
+    },
+    {
+      "epoch": 2.6682119296128155,
+      "grad_norm": 0.3210340795390415,
+      "learning_rate": 4.090436945540623e-06,
+      "loss": 0.4382,
+      "step": 16210
+    },
+    {
+      "epoch": 2.668376513831932,
+      "grad_norm": 0.3029225777699273,
+      "learning_rate": 4.089968834984206e-06,
+      "loss": 0.4334,
+      "step": 16211
+    },
+    {
+      "epoch": 2.6685410980510484,
+      "grad_norm": 0.29585427129308206,
+      "learning_rate": 4.089500728901885e-06,
+      "loss": 0.4171,
+      "step": 16212
+    },
+    {
+      "epoch": 2.668705682270165,
+      "grad_norm": 0.30808751863312334,
+      "learning_rate": 4.089032627298768e-06,
+      "loss": 0.4282,
+      "step": 16213
+    },
+    {
+      "epoch": 2.6688702664892814,
+      "grad_norm": 0.26945376664591575,
+      "learning_rate": 4.0885645301799615e-06,
+      "loss": 0.4313,
+      "step": 16214
+    },
+    {
+      "epoch": 2.669034850708398,
+      "grad_norm": 0.3248008830813408,
+      "learning_rate": 4.088096437550577e-06,
+      "loss": 0.4383,
+      "step": 16215
+    },
+    {
+      "epoch": 2.6691994349275143,
+      "grad_norm": 0.3257332301812428,
+      "learning_rate": 4.0876283494157186e-06,
+      "loss": 0.4334,
+      "step": 16216
+    },
+    {
+      "epoch": 2.669364019146631,
+      "grad_norm": 0.4606309903111496,
+      "learning_rate": 4.087160265780496e-06,
+      "loss": 0.4325,
+      "step": 16217
+    },
+    {
+      "epoch": 2.6695286033657473,
+      "grad_norm": 0.3205989020265071,
+      "learning_rate": 4.086692186650016e-06,
+      "loss": 0.4257,
+      "step": 16218
+    },
+    {
+      "epoch": 2.6696931875848637,
+      "grad_norm": 0.333101561605051,
+      "learning_rate": 4.086224112029385e-06,
+      "loss": 0.4371,
+      "step": 16219
+    },
+    {
+      "epoch": 2.66985777180398,
+      "grad_norm": 0.28771924507188223,
+      "learning_rate": 4.085756041923711e-06,
+      "loss": 0.438,
+      "step": 16220
+    },
+    {
+      "epoch": 2.6700223560230967,
+      "grad_norm": 0.27046072482072303,
+      "learning_rate": 4.085287976338102e-06,
+      "loss": 0.4127,
+      "step": 16221
+    },
+    {
+      "epoch": 2.670186940242213,
+      "grad_norm": 0.300922333424232,
+      "learning_rate": 4.084819915277665e-06,
+      "loss": 0.4319,
+      "step": 16222
+    },
+    {
+      "epoch": 2.6703515244613296,
+      "grad_norm": 0.24299807154068132,
+      "learning_rate": 4.0843518587475064e-06,
+      "loss": 0.415,
+      "step": 16223
+    },
+    {
+      "epoch": 2.670516108680446,
+      "grad_norm": 0.341735720442819,
+      "learning_rate": 4.083883806752737e-06,
+      "loss": 0.414,
+      "step": 16224
+    },
+    {
+      "epoch": 2.6706806928995626,
+      "grad_norm": 0.3393067511916409,
+      "learning_rate": 4.083415759298461e-06,
+      "loss": 0.429,
+      "step": 16225
+    },
+    {
+      "epoch": 2.670845277118679,
+      "grad_norm": 0.28674310194577535,
+      "learning_rate": 4.082947716389784e-06,
+      "loss": 0.429,
+      "step": 16226
+    },
+    {
+      "epoch": 2.6710098613377955,
+      "grad_norm": 0.5185646111855251,
+      "learning_rate": 4.082479678031817e-06,
+      "loss": 0.4454,
+      "step": 16227
+    },
+    {
+      "epoch": 2.671174445556912,
+      "grad_norm": 0.34007590718458663,
+      "learning_rate": 4.082011644229663e-06,
+      "loss": 0.4159,
+      "step": 16228
+    },
+    {
+      "epoch": 2.6713390297760284,
+      "grad_norm": 0.29528169814285576,
+      "learning_rate": 4.081543614988434e-06,
+      "loss": 0.4363,
+      "step": 16229
+    },
+    {
+      "epoch": 2.671503613995145,
+      "grad_norm": 0.3010034087122512,
+      "learning_rate": 4.0810755903132315e-06,
+      "loss": 0.4339,
+      "step": 16230
+    },
+    {
+      "epoch": 2.6716681982142614,
+      "grad_norm": 0.3420782797293866,
+      "learning_rate": 4.080607570209166e-06,
+      "loss": 0.4532,
+      "step": 16231
+    },
+    {
+      "epoch": 2.671832782433378,
+      "grad_norm": 0.3518435427751243,
+      "learning_rate": 4.080139554681343e-06,
+      "loss": 0.4374,
+      "step": 16232
+    },
+    {
+      "epoch": 2.6719973666524943,
+      "grad_norm": 0.3853923992274283,
+      "learning_rate": 4.0796715437348696e-06,
+      "loss": 0.4189,
+      "step": 16233
+    },
+    {
+      "epoch": 2.672161950871611,
+      "grad_norm": 0.39547593141325593,
+      "learning_rate": 4.079203537374853e-06,
+      "loss": 0.4322,
+      "step": 16234
+    },
+    {
+      "epoch": 2.6723265350907273,
+      "grad_norm": 0.4567047029006985,
+      "learning_rate": 4.078735535606399e-06,
+      "loss": 0.4189,
+      "step": 16235
+    },
+    {
+      "epoch": 2.6724911193098437,
+      "grad_norm": 0.30768753053570136,
+      "learning_rate": 4.078267538434616e-06,
+      "loss": 0.403,
+      "step": 16236
+    },
+    {
+      "epoch": 2.67265570352896,
+      "grad_norm": 0.37096345709738054,
+      "learning_rate": 4.077799545864607e-06,
+      "loss": 0.4071,
+      "step": 16237
+    },
+    {
+      "epoch": 2.6728202877480767,
+      "grad_norm": 0.2795107187902698,
+      "learning_rate": 4.0773315579014835e-06,
+      "loss": 0.4394,
+      "step": 16238
+    },
+    {
+      "epoch": 2.672984871967193,
+      "grad_norm": 0.40934872094739266,
+      "learning_rate": 4.076863574550349e-06,
+      "loss": 0.4526,
+      "step": 16239
+    },
+    {
+      "epoch": 2.673149456186309,
+      "grad_norm": 0.3011794778746138,
+      "learning_rate": 4.076395595816308e-06,
+      "loss": 0.4298,
+      "step": 16240
+    },
+    {
+      "epoch": 2.6733140404054256,
+      "grad_norm": 0.4588930576874614,
+      "learning_rate": 4.075927621704471e-06,
+      "loss": 0.4395,
+      "step": 16241
+    },
+    {
+      "epoch": 2.673478624624542,
+      "grad_norm": 0.2837479901390299,
+      "learning_rate": 4.075459652219941e-06,
+      "loss": 0.4207,
+      "step": 16242
+    },
+    {
+      "epoch": 2.6736432088436586,
+      "grad_norm": 0.37951879970028074,
+      "learning_rate": 4.074991687367827e-06,
+      "loss": 0.4461,
+      "step": 16243
+    },
+    {
+      "epoch": 2.673807793062775,
+      "grad_norm": 0.4577196555274362,
+      "learning_rate": 4.074523727153234e-06,
+      "loss": 0.4284,
+      "step": 16244
+    },
+    {
+      "epoch": 2.6739723772818915,
+      "grad_norm": 0.3745132841651956,
+      "learning_rate": 4.074055771581268e-06,
+      "loss": 0.4428,
+      "step": 16245
+    },
+    {
+      "epoch": 2.674136961501008,
+      "grad_norm": 0.5084261638812808,
+      "learning_rate": 4.073587820657037e-06,
+      "loss": 0.4531,
+      "step": 16246
+    },
+    {
+      "epoch": 2.6743015457201245,
+      "grad_norm": 0.39509714110382965,
+      "learning_rate": 4.073119874385644e-06,
+      "loss": 0.4279,
+      "step": 16247
+    },
+    {
+      "epoch": 2.674466129939241,
+      "grad_norm": 0.3186932320538688,
+      "learning_rate": 4.072651932772197e-06,
+      "loss": 0.4236,
+      "step": 16248
+    },
+    {
+      "epoch": 2.6746307141583574,
+      "grad_norm": 0.49111200088983387,
+      "learning_rate": 4.072183995821801e-06,
+      "loss": 0.4293,
+      "step": 16249
+    },
+    {
+      "epoch": 2.674795298377474,
+      "grad_norm": 0.349164641605376,
+      "learning_rate": 4.071716063539563e-06,
+      "loss": 0.4612,
+      "step": 16250
+    },
+    {
+      "epoch": 2.6749598825965903,
+      "grad_norm": 0.4703838182432665,
+      "learning_rate": 4.0712481359305876e-06,
+      "loss": 0.4506,
+      "step": 16251
+    },
+    {
+      "epoch": 2.675124466815707,
+      "grad_norm": 0.27681566344147845,
+      "learning_rate": 4.070780212999982e-06,
+      "loss": 0.4306,
+      "step": 16252
+    },
+    {
+      "epoch": 2.6752890510348233,
+      "grad_norm": 0.2706256207107712,
+      "learning_rate": 4.070312294752852e-06,
+      "loss": 0.4132,
+      "step": 16253
+    },
+    {
+      "epoch": 2.6754536352539398,
+      "grad_norm": 0.2942736546722622,
+      "learning_rate": 4.069844381194301e-06,
+      "loss": 0.4247,
+      "step": 16254
+    },
+    {
+      "epoch": 2.6756182194730562,
+      "grad_norm": 0.2887730169859917,
+      "learning_rate": 4.06937647232944e-06,
+      "loss": 0.4282,
+      "step": 16255
+    },
+    {
+      "epoch": 2.6757828036921727,
+      "grad_norm": 0.3249853712294799,
+      "learning_rate": 4.068908568163367e-06,
+      "loss": 0.4552,
+      "step": 16256
+    },
+    {
+      "epoch": 2.675947387911289,
+      "grad_norm": 0.42233120294813575,
+      "learning_rate": 4.068440668701195e-06,
+      "loss": 0.4494,
+      "step": 16257
+    },
+    {
+      "epoch": 2.6761119721304056,
+      "grad_norm": 0.39370678424915456,
+      "learning_rate": 4.067972773948022e-06,
+      "loss": 0.4425,
+      "step": 16258
+    },
+    {
+      "epoch": 2.676276556349522,
+      "grad_norm": 0.3351960219169127,
+      "learning_rate": 4.067504883908962e-06,
+      "loss": 0.4287,
+      "step": 16259
+    },
+    {
+      "epoch": 2.6764411405686386,
+      "grad_norm": 0.3646822270877999,
+      "learning_rate": 4.067036998589114e-06,
+      "loss": 0.4406,
+      "step": 16260
+    },
+    {
+      "epoch": 2.676605724787755,
+      "grad_norm": 0.27525514075166774,
+      "learning_rate": 4.066569117993586e-06,
+      "loss": 0.4177,
+      "step": 16261
+    },
+    {
+      "epoch": 2.6767703090068715,
+      "grad_norm": 0.33919213329865616,
+      "learning_rate": 4.066101242127483e-06,
+      "loss": 0.4459,
+      "step": 16262
+    },
+    {
+      "epoch": 2.676934893225988,
+      "grad_norm": 0.35308673235114507,
+      "learning_rate": 4.0656333709959085e-06,
+      "loss": 0.4379,
+      "step": 16263
+    },
+    {
+      "epoch": 2.6770994774451045,
+      "grad_norm": 0.30855438126161155,
+      "learning_rate": 4.065165504603971e-06,
+      "loss": 0.4431,
+      "step": 16264
+    },
+    {
+      "epoch": 2.677264061664221,
+      "grad_norm": 0.34653493149245607,
+      "learning_rate": 4.064697642956773e-06,
+      "loss": 0.4355,
+      "step": 16265
+    },
+    {
+      "epoch": 2.677428645883337,
+      "grad_norm": 0.29623996229396615,
+      "learning_rate": 4.064229786059422e-06,
+      "loss": 0.4255,
+      "step": 16266
+    },
+    {
+      "epoch": 2.6775932301024534,
+      "grad_norm": 0.3025284374334076,
+      "learning_rate": 4.06376193391702e-06,
+      "loss": 0.4453,
+      "step": 16267
+    },
+    {
+      "epoch": 2.67775781432157,
+      "grad_norm": 0.3830832942566304,
+      "learning_rate": 4.063294086534675e-06,
+      "loss": 0.4266,
+      "step": 16268
+    },
+    {
+      "epoch": 2.6779223985406864,
+      "grad_norm": 0.3344091233245314,
+      "learning_rate": 4.06282624391749e-06,
+      "loss": 0.4222,
+      "step": 16269
+    },
+    {
+      "epoch": 2.678086982759803,
+      "grad_norm": 0.3260377569625621,
+      "learning_rate": 4.06235840607057e-06,
+      "loss": 0.4139,
+      "step": 16270
+    },
+    {
+      "epoch": 2.6782515669789193,
+      "grad_norm": 0.4072985148156493,
+      "learning_rate": 4.061890572999021e-06,
+      "loss": 0.4392,
+      "step": 16271
+    },
+    {
+      "epoch": 2.6784161511980358,
+      "grad_norm": 0.3117617202916446,
+      "learning_rate": 4.061422744707947e-06,
+      "loss": 0.4456,
+      "step": 16272
+    },
+    {
+      "epoch": 2.6785807354171522,
+      "grad_norm": 0.34202950905471813,
+      "learning_rate": 4.060954921202454e-06,
+      "loss": 0.4252,
+      "step": 16273
+    },
+    {
+      "epoch": 2.6787453196362687,
+      "grad_norm": 0.3078055002171829,
+      "learning_rate": 4.060487102487644e-06,
+      "loss": 0.4427,
+      "step": 16274
+    },
+    {
+      "epoch": 2.678909903855385,
+      "grad_norm": 0.34492177812050034,
+      "learning_rate": 4.060019288568624e-06,
+      "loss": 0.4127,
+      "step": 16275
+    },
+    {
+      "epoch": 2.6790744880745017,
+      "grad_norm": 0.4479150830400735,
+      "learning_rate": 4.0595514794505e-06,
+      "loss": 0.41,
+      "step": 16276
+    },
+    {
+      "epoch": 2.679239072293618,
+      "grad_norm": 0.3027447731401161,
+      "learning_rate": 4.0590836751383704e-06,
+      "loss": 0.4125,
+      "step": 16277
+    },
+    {
+      "epoch": 2.6794036565127346,
+      "grad_norm": 0.3868369095960643,
+      "learning_rate": 4.058615875637347e-06,
+      "loss": 0.4229,
+      "step": 16278
+    },
+    {
+      "epoch": 2.679568240731851,
+      "grad_norm": 0.31252996703349806,
+      "learning_rate": 4.058148080952529e-06,
+      "loss": 0.4387,
+      "step": 16279
+    },
+    {
+      "epoch": 2.6797328249509675,
+      "grad_norm": 0.38131546502511204,
+      "learning_rate": 4.057680291089025e-06,
+      "loss": 0.4208,
+      "step": 16280
+    },
+    {
+      "epoch": 2.679897409170084,
+      "grad_norm": 0.31437886402272847,
+      "learning_rate": 4.057212506051935e-06,
+      "loss": 0.442,
+      "step": 16281
+    },
+    {
+      "epoch": 2.6800619933892005,
+      "grad_norm": 0.29705877207988296,
+      "learning_rate": 4.056744725846366e-06,
+      "loss": 0.4405,
+      "step": 16282
+    },
+    {
+      "epoch": 2.680226577608317,
+      "grad_norm": 0.3287127232403905,
+      "learning_rate": 4.0562769504774226e-06,
+      "loss": 0.4369,
+      "step": 16283
+    },
+    {
+      "epoch": 2.6803911618274334,
+      "grad_norm": 0.3461000338063579,
+      "learning_rate": 4.055809179950207e-06,
+      "loss": 0.4416,
+      "step": 16284
+    },
+    {
+      "epoch": 2.68055574604655,
+      "grad_norm": 0.3162634482779627,
+      "learning_rate": 4.055341414269825e-06,
+      "loss": 0.4496,
+      "step": 16285
+    },
+    {
+      "epoch": 2.6807203302656664,
+      "grad_norm": 0.36816052604616,
+      "learning_rate": 4.0548736534413795e-06,
+      "loss": 0.439,
+      "step": 16286
+    },
+    {
+      "epoch": 2.680884914484783,
+      "grad_norm": 0.354552042915324,
+      "learning_rate": 4.0544058974699764e-06,
+      "loss": 0.4346,
+      "step": 16287
+    },
+    {
+      "epoch": 2.6810494987038993,
+      "grad_norm": 0.38578018985135637,
+      "learning_rate": 4.053938146360715e-06,
+      "loss": 0.4294,
+      "step": 16288
+    },
+    {
+      "epoch": 2.6812140829230158,
+      "grad_norm": 0.26869839291294056,
+      "learning_rate": 4.053470400118707e-06,
+      "loss": 0.4321,
+      "step": 16289
+    },
+    {
+      "epoch": 2.6813786671421322,
+      "grad_norm": 0.5894318253041799,
+      "learning_rate": 4.053002658749049e-06,
+      "loss": 0.4472,
+      "step": 16290
+    },
+    {
+      "epoch": 2.6815432513612487,
+      "grad_norm": 0.32389992131136,
+      "learning_rate": 4.0525349222568475e-06,
+      "loss": 0.4352,
+      "step": 16291
+    },
+    {
+      "epoch": 2.681707835580365,
+      "grad_norm": 0.35372515325452797,
+      "learning_rate": 4.052067190647208e-06,
+      "loss": 0.4445,
+      "step": 16292
+    },
+    {
+      "epoch": 2.6818724197994817,
+      "grad_norm": 0.3780722530944741,
+      "learning_rate": 4.051599463925231e-06,
+      "loss": 0.4342,
+      "step": 16293
+    },
+    {
+      "epoch": 2.682037004018598,
+      "grad_norm": 0.3179532350010808,
+      "learning_rate": 4.051131742096022e-06,
+      "loss": 0.4373,
+      "step": 16294
+    },
+    {
+      "epoch": 2.6822015882377146,
+      "grad_norm": 0.28235609038119763,
+      "learning_rate": 4.050664025164684e-06,
+      "loss": 0.4266,
+      "step": 16295
+    },
+    {
+      "epoch": 2.682366172456831,
+      "grad_norm": 0.34937662996411334,
+      "learning_rate": 4.050196313136322e-06,
+      "loss": 0.4204,
+      "step": 16296
+    },
+    {
+      "epoch": 2.6825307566759475,
+      "grad_norm": 0.29543097261117546,
+      "learning_rate": 4.049728606016039e-06,
+      "loss": 0.4281,
+      "step": 16297
+    },
+    {
+      "epoch": 2.682695340895064,
+      "grad_norm": 0.2755549630399525,
+      "learning_rate": 4.049260903808936e-06,
+      "loss": 0.4285,
+      "step": 16298
+    },
+    {
+      "epoch": 2.6828599251141805,
+      "grad_norm": 0.3197657919433483,
+      "learning_rate": 4.048793206520118e-06,
+      "loss": 0.4312,
+      "step": 16299
+    },
+    {
+      "epoch": 2.683024509333297,
+      "grad_norm": 0.41965219409979476,
+      "learning_rate": 4.048325514154688e-06,
+      "loss": 0.4388,
+      "step": 16300
+    },
+    {
+      "epoch": 2.6831890935524134,
+      "grad_norm": 0.27678180863462526,
+      "learning_rate": 4.047857826717751e-06,
+      "loss": 0.4242,
+      "step": 16301
+    },
+    {
+      "epoch": 2.68335367777153,
+      "grad_norm": 0.347595027625458,
+      "learning_rate": 4.047390144214409e-06,
+      "loss": 0.4457,
+      "step": 16302
+    },
+    {
+      "epoch": 2.6835182619906464,
+      "grad_norm": 0.27207774173030125,
+      "learning_rate": 4.046922466649764e-06,
+      "loss": 0.4235,
+      "step": 16303
+    },
+    {
+      "epoch": 2.683682846209763,
+      "grad_norm": 0.33812820655438,
+      "learning_rate": 4.0464547940289205e-06,
+      "loss": 0.4282,
+      "step": 16304
+    },
+    {
+      "epoch": 2.6838474304288793,
+      "grad_norm": 0.31855684501272835,
+      "learning_rate": 4.045987126356982e-06,
+      "loss": 0.4508,
+      "step": 16305
+    },
+    {
+      "epoch": 2.6840120146479958,
+      "grad_norm": 0.2655956406794349,
+      "learning_rate": 4.045519463639052e-06,
+      "loss": 0.4403,
+      "step": 16306
+    },
+    {
+      "epoch": 2.684176598867112,
+      "grad_norm": 0.3066415099568566,
+      "learning_rate": 4.045051805880229e-06,
+      "loss": 0.4123,
+      "step": 16307
+    },
+    {
+      "epoch": 2.6843411830862283,
+      "grad_norm": 0.4262155359903081,
+      "learning_rate": 4.044584153085623e-06,
+      "loss": 0.4483,
+      "step": 16308
+    },
+    {
+      "epoch": 2.6845057673053447,
+      "grad_norm": 0.30152830675260467,
+      "learning_rate": 4.044116505260329e-06,
+      "loss": 0.4397,
+      "step": 16309
+    },
+    {
+      "epoch": 2.684670351524461,
+      "grad_norm": 0.3478555543350279,
+      "learning_rate": 4.043648862409457e-06,
+      "loss": 0.4255,
+      "step": 16310
+    },
+    {
+      "epoch": 2.6848349357435777,
+      "grad_norm": 0.4643246693630661,
+      "learning_rate": 4.043181224538104e-06,
+      "loss": 0.4371,
+      "step": 16311
+    },
+    {
+      "epoch": 2.684999519962694,
+      "grad_norm": 0.2839497220739598,
+      "learning_rate": 4.042713591651377e-06,
+      "loss": 0.4233,
+      "step": 16312
+    },
+    {
+      "epoch": 2.6851641041818106,
+      "grad_norm": 0.2903961775110076,
+      "learning_rate": 4.042245963754376e-06,
+      "loss": 0.4377,
+      "step": 16313
+    },
+    {
+      "epoch": 2.685328688400927,
+      "grad_norm": 0.2753754545659661,
+      "learning_rate": 4.041778340852204e-06,
+      "loss": 0.4058,
+      "step": 16314
+    },
+    {
+      "epoch": 2.6854932726200436,
+      "grad_norm": 0.3303682132491632,
+      "learning_rate": 4.041310722949964e-06,
+      "loss": 0.4355,
+      "step": 16315
+    },
+    {
+      "epoch": 2.68565785683916,
+      "grad_norm": 0.37255543523726975,
+      "learning_rate": 4.040843110052758e-06,
+      "loss": 0.4319,
+      "step": 16316
+    },
+    {
+      "epoch": 2.6858224410582765,
+      "grad_norm": 0.3054564467132957,
+      "learning_rate": 4.04037550216569e-06,
+      "loss": 0.4291,
+      "step": 16317
+    },
+    {
+      "epoch": 2.685987025277393,
+      "grad_norm": 0.28915295925285905,
+      "learning_rate": 4.0399078992938595e-06,
+      "loss": 0.463,
+      "step": 16318
+    },
+    {
+      "epoch": 2.6861516094965094,
+      "grad_norm": 0.4022542124139846,
+      "learning_rate": 4.0394403014423725e-06,
+      "loss": 0.4405,
+      "step": 16319
+    },
+    {
+      "epoch": 2.686316193715626,
+      "grad_norm": 0.4806554260121287,
+      "learning_rate": 4.038972708616328e-06,
+      "loss": 0.4386,
+      "step": 16320
+    },
+    {
+      "epoch": 2.6864807779347424,
+      "grad_norm": 0.4025570728681745,
+      "learning_rate": 4.038505120820829e-06,
+      "loss": 0.4217,
+      "step": 16321
+    },
+    {
+      "epoch": 2.686645362153859,
+      "grad_norm": 0.33411015406408345,
+      "learning_rate": 4.0380375380609785e-06,
+      "loss": 0.4164,
+      "step": 16322
+    },
+    {
+      "epoch": 2.6868099463729753,
+      "grad_norm": 0.38337978563252073,
+      "learning_rate": 4.037569960341877e-06,
+      "loss": 0.442,
+      "step": 16323
+    },
+    {
+      "epoch": 2.686974530592092,
+      "grad_norm": 0.3286192345753033,
+      "learning_rate": 4.0371023876686285e-06,
+      "loss": 0.4229,
+      "step": 16324
+    },
+    {
+      "epoch": 2.6871391148112083,
+      "grad_norm": 0.2887883452897136,
+      "learning_rate": 4.036634820046333e-06,
+      "loss": 0.429,
+      "step": 16325
+    },
+    {
+      "epoch": 2.6873036990303247,
+      "grad_norm": 0.6183909761308701,
+      "learning_rate": 4.036167257480095e-06,
+      "loss": 0.4306,
+      "step": 16326
+    },
+    {
+      "epoch": 2.687468283249441,
+      "grad_norm": 0.294742555568296,
+      "learning_rate": 4.0356996999750146e-06,
+      "loss": 0.4292,
+      "step": 16327
+    },
+    {
+      "epoch": 2.6876328674685577,
+      "grad_norm": 0.32687745293903564,
+      "learning_rate": 4.035232147536191e-06,
+      "loss": 0.421,
+      "step": 16328
+    },
+    {
+      "epoch": 2.687797451687674,
+      "grad_norm": 0.3037644479826378,
+      "learning_rate": 4.034764600168733e-06,
+      "loss": 0.4258,
+      "step": 16329
+    },
+    {
+      "epoch": 2.6879620359067906,
+      "grad_norm": 0.30225917798625956,
+      "learning_rate": 4.034297057877735e-06,
+      "loss": 0.4382,
+      "step": 16330
+    },
+    {
+      "epoch": 2.688126620125907,
+      "grad_norm": 0.38987761613900895,
+      "learning_rate": 4.033829520668302e-06,
+      "loss": 0.4328,
+      "step": 16331
+    },
+    {
+      "epoch": 2.688291204345023,
+      "grad_norm": 0.3210169110933282,
+      "learning_rate": 4.033361988545535e-06,
+      "loss": 0.4412,
+      "step": 16332
+    },
+    {
+      "epoch": 2.6884557885641396,
+      "grad_norm": 0.31864691654286775,
+      "learning_rate": 4.032894461514537e-06,
+      "loss": 0.4282,
+      "step": 16333
+    },
+    {
+      "epoch": 2.688620372783256,
+      "grad_norm": 0.4825974822411688,
+      "learning_rate": 4.0324269395804074e-06,
+      "loss": 0.4326,
+      "step": 16334
+    },
+    {
+      "epoch": 2.6887849570023725,
+      "grad_norm": 0.32769347237972846,
+      "learning_rate": 4.031959422748249e-06,
+      "loss": 0.4267,
+      "step": 16335
+    },
+    {
+      "epoch": 2.688949541221489,
+      "grad_norm": 0.38048815364433103,
+      "learning_rate": 4.031491911023163e-06,
+      "loss": 0.4389,
+      "step": 16336
+    },
+    {
+      "epoch": 2.6891141254406055,
+      "grad_norm": 1.0058854085830922,
+      "learning_rate": 4.031024404410248e-06,
+      "loss": 0.4429,
+      "step": 16337
+    },
+    {
+      "epoch": 2.689278709659722,
+      "grad_norm": 0.3754249272517528,
+      "learning_rate": 4.03055690291461e-06,
+      "loss": 0.4199,
+      "step": 16338
+    },
+    {
+      "epoch": 2.6894432938788384,
+      "grad_norm": 0.2947882584863735,
+      "learning_rate": 4.030089406541345e-06,
+      "loss": 0.4281,
+      "step": 16339
+    },
+    {
+      "epoch": 2.689607878097955,
+      "grad_norm": 0.3096317789054828,
+      "learning_rate": 4.029621915295561e-06,
+      "loss": 0.4298,
+      "step": 16340
+    },
+    {
+      "epoch": 2.6897724623170713,
+      "grad_norm": 0.35530437347375415,
+      "learning_rate": 4.029154429182352e-06,
+      "loss": 0.4357,
+      "step": 16341
+    },
+    {
+      "epoch": 2.689937046536188,
+      "grad_norm": 0.3012366191872031,
+      "learning_rate": 4.0286869482068215e-06,
+      "loss": 0.4405,
+      "step": 16342
+    },
+    {
+      "epoch": 2.6901016307553043,
+      "grad_norm": 0.31298119816673015,
+      "learning_rate": 4.028219472374073e-06,
+      "loss": 0.4375,
+      "step": 16343
+    },
+    {
+      "epoch": 2.6902662149744208,
+      "grad_norm": 0.32781224232259865,
+      "learning_rate": 4.027752001689203e-06,
+      "loss": 0.4406,
+      "step": 16344
+    },
+    {
+      "epoch": 2.6904307991935372,
+      "grad_norm": 0.414384595744943,
+      "learning_rate": 4.027284536157316e-06,
+      "loss": 0.4224,
+      "step": 16345
+    },
+    {
+      "epoch": 2.6905953834126537,
+      "grad_norm": 0.36499776322117494,
+      "learning_rate": 4.026817075783511e-06,
+      "loss": 0.4376,
+      "step": 16346
+    },
+    {
+      "epoch": 2.69075996763177,
+      "grad_norm": 0.32248926422427887,
+      "learning_rate": 4.02634962057289e-06,
+      "loss": 0.4432,
+      "step": 16347
+    },
+    {
+      "epoch": 2.6909245518508866,
+      "grad_norm": 0.338708489474019,
+      "learning_rate": 4.025882170530552e-06,
+      "loss": 0.459,
+      "step": 16348
+    },
+    {
+      "epoch": 2.691089136070003,
+      "grad_norm": 0.294052606229598,
+      "learning_rate": 4.025414725661601e-06,
+      "loss": 0.4451,
+      "step": 16349
+    },
+    {
+      "epoch": 2.6912537202891196,
+      "grad_norm": 0.2780252582537487,
+      "learning_rate": 4.024947285971133e-06,
+      "loss": 0.4453,
+      "step": 16350
+    },
+    {
+      "epoch": 2.691418304508236,
+      "grad_norm": 0.33408514253873084,
+      "learning_rate": 4.02447985146425e-06,
+      "loss": 0.4298,
+      "step": 16351
+    },
+    {
+      "epoch": 2.6915828887273525,
+      "grad_norm": 0.3270836296226998,
+      "learning_rate": 4.024012422146054e-06,
+      "loss": 0.4399,
+      "step": 16352
+    },
+    {
+      "epoch": 2.691747472946469,
+      "grad_norm": 0.36075527590040307,
+      "learning_rate": 4.023544998021644e-06,
+      "loss": 0.4183,
+      "step": 16353
+    },
+    {
+      "epoch": 2.6919120571655855,
+      "grad_norm": 0.3363764146010592,
+      "learning_rate": 4.023077579096121e-06,
+      "loss": 0.4425,
+      "step": 16354
+    },
+    {
+      "epoch": 2.692076641384702,
+      "grad_norm": 0.31465186885890206,
+      "learning_rate": 4.022610165374585e-06,
+      "loss": 0.4223,
+      "step": 16355
+    },
+    {
+      "epoch": 2.6922412256038184,
+      "grad_norm": 0.35284763955478143,
+      "learning_rate": 4.022142756862137e-06,
+      "loss": 0.4383,
+      "step": 16356
+    },
+    {
+      "epoch": 2.692405809822935,
+      "grad_norm": 0.3134771967967619,
+      "learning_rate": 4.021675353563877e-06,
+      "loss": 0.44,
+      "step": 16357
+    },
+    {
+      "epoch": 2.6925703940420513,
+      "grad_norm": 0.3802097153284701,
+      "learning_rate": 4.0212079554849025e-06,
+      "loss": 0.4319,
+      "step": 16358
+    },
+    {
+      "epoch": 2.692734978261168,
+      "grad_norm": 0.34680413997601384,
+      "learning_rate": 4.020740562630319e-06,
+      "loss": 0.4456,
+      "step": 16359
+    },
+    {
+      "epoch": 2.6928995624802843,
+      "grad_norm": 0.4307466925610047,
+      "learning_rate": 4.02027317500522e-06,
+      "loss": 0.4387,
+      "step": 16360
+    },
+    {
+      "epoch": 2.6930641466994008,
+      "grad_norm": 0.3429686840682955,
+      "learning_rate": 4.01980579261471e-06,
+      "loss": 0.4355,
+      "step": 16361
+    },
+    {
+      "epoch": 2.693228730918517,
+      "grad_norm": 0.2995725153651655,
+      "learning_rate": 4.019338415463888e-06,
+      "loss": 0.4388,
+      "step": 16362
+    },
+    {
+      "epoch": 2.6933933151376337,
+      "grad_norm": 0.5330069986000292,
+      "learning_rate": 4.018871043557852e-06,
+      "loss": 0.4396,
+      "step": 16363
+    },
+    {
+      "epoch": 2.69355789935675,
+      "grad_norm": 0.3288145812661788,
+      "learning_rate": 4.018403676901704e-06,
+      "loss": 0.4132,
+      "step": 16364
+    },
+    {
+      "epoch": 2.6937224835758666,
+      "grad_norm": 0.29583734843271503,
+      "learning_rate": 4.017936315500543e-06,
+      "loss": 0.451,
+      "step": 16365
+    },
+    {
+      "epoch": 2.693887067794983,
+      "grad_norm": 0.42497783015345564,
+      "learning_rate": 4.017468959359469e-06,
+      "loss": 0.4285,
+      "step": 16366
+    },
+    {
+      "epoch": 2.6940516520140996,
+      "grad_norm": 0.2844809828785978,
+      "learning_rate": 4.017001608483579e-06,
+      "loss": 0.4358,
+      "step": 16367
+    },
+    {
+      "epoch": 2.694216236233216,
+      "grad_norm": 0.3626814403818623,
+      "learning_rate": 4.016534262877978e-06,
+      "loss": 0.4234,
+      "step": 16368
+    },
+    {
+      "epoch": 2.6943808204523325,
+      "grad_norm": 0.28919512072085424,
+      "learning_rate": 4.016066922547759e-06,
+      "loss": 0.4327,
+      "step": 16369
+    },
+    {
+      "epoch": 2.694545404671449,
+      "grad_norm": 0.32003358765545853,
+      "learning_rate": 4.015599587498026e-06,
+      "loss": 0.4388,
+      "step": 16370
+    },
+    {
+      "epoch": 2.6947099888905655,
+      "grad_norm": 0.36882718405394466,
+      "learning_rate": 4.015132257733878e-06,
+      "loss": 0.4623,
+      "step": 16371
+    },
+    {
+      "epoch": 2.694874573109682,
+      "grad_norm": 0.27980427256449153,
+      "learning_rate": 4.014664933260411e-06,
+      "loss": 0.4161,
+      "step": 16372
+    },
+    {
+      "epoch": 2.6950391573287984,
+      "grad_norm": 0.41433725457500464,
+      "learning_rate": 4.0141976140827285e-06,
+      "loss": 0.4193,
+      "step": 16373
+    },
+    {
+      "epoch": 2.6952037415479144,
+      "grad_norm": 0.38174671828343304,
+      "learning_rate": 4.013730300205927e-06,
+      "loss": 0.4233,
+      "step": 16374
+    },
+    {
+      "epoch": 2.695368325767031,
+      "grad_norm": 0.37448167649323577,
+      "learning_rate": 4.0132629916351064e-06,
+      "loss": 0.4348,
+      "step": 16375
+    },
+    {
+      "epoch": 2.6955329099861474,
+      "grad_norm": 0.3227325820515218,
+      "learning_rate": 4.012795688375366e-06,
+      "loss": 0.4289,
+      "step": 16376
+    },
+    {
+      "epoch": 2.695697494205264,
+      "grad_norm": 0.34668685475852096,
+      "learning_rate": 4.012328390431804e-06,
+      "loss": 0.4328,
+      "step": 16377
+    },
+    {
+      "epoch": 2.6958620784243803,
+      "grad_norm": 0.27383474398604524,
+      "learning_rate": 4.0118610978095216e-06,
+      "loss": 0.4297,
+      "step": 16378
+    },
+    {
+      "epoch": 2.6960266626434968,
+      "grad_norm": 0.42749690459672113,
+      "learning_rate": 4.011393810513614e-06,
+      "loss": 0.4651,
+      "step": 16379
+    },
+    {
+      "epoch": 2.6961912468626132,
+      "grad_norm": 0.3957381987373052,
+      "learning_rate": 4.010926528549184e-06,
+      "loss": 0.4396,
+      "step": 16380
+    },
+    {
+      "epoch": 2.6963558310817297,
+      "grad_norm": 0.4465267398360211,
+      "learning_rate": 4.010459251921327e-06,
+      "loss": 0.4428,
+      "step": 16381
+    },
+    {
+      "epoch": 2.696520415300846,
+      "grad_norm": 0.33690374814706475,
+      "learning_rate": 4.009991980635144e-06,
+      "loss": 0.4245,
+      "step": 16382
+    },
+    {
+      "epoch": 2.6966849995199627,
+      "grad_norm": 0.264495976809934,
+      "learning_rate": 4.0095247146957325e-06,
+      "loss": 0.418,
+      "step": 16383
+    },
+    {
+      "epoch": 2.696849583739079,
+      "grad_norm": 0.28920170373751336,
+      "learning_rate": 4.009057454108191e-06,
+      "loss": 0.4269,
+      "step": 16384
+    },
+    {
+      "epoch": 2.6970141679581956,
+      "grad_norm": 1.0817155635714626,
+      "learning_rate": 4.00859019887762e-06,
+      "loss": 0.4385,
+      "step": 16385
+    },
+    {
+      "epoch": 2.697178752177312,
+      "grad_norm": 0.3838047671684187,
+      "learning_rate": 4.008122949009116e-06,
+      "loss": 0.4376,
+      "step": 16386
+    },
+    {
+      "epoch": 2.6973433363964285,
+      "grad_norm": 0.2556844042050589,
+      "learning_rate": 4.007655704507779e-06,
+      "loss": 0.419,
+      "step": 16387
+    },
+    {
+      "epoch": 2.697507920615545,
+      "grad_norm": 0.27940657185380346,
+      "learning_rate": 4.007188465378704e-06,
+      "loss": 0.4193,
+      "step": 16388
+    },
+    {
+      "epoch": 2.6976725048346615,
+      "grad_norm": 0.3339629542859908,
+      "learning_rate": 4.006721231626995e-06,
+      "loss": 0.4421,
+      "step": 16389
+    },
+    {
+      "epoch": 2.697837089053778,
+      "grad_norm": 0.3339466019931385,
+      "learning_rate": 4.006254003257744e-06,
+      "loss": 0.4305,
+      "step": 16390
+    },
+    {
+      "epoch": 2.6980016732728944,
+      "grad_norm": 0.28327025098936864,
+      "learning_rate": 4.0057867802760545e-06,
+      "loss": 0.4298,
+      "step": 16391
+    },
+    {
+      "epoch": 2.698166257492011,
+      "grad_norm": 0.33148361090437845,
+      "learning_rate": 4.005319562687021e-06,
+      "loss": 0.4252,
+      "step": 16392
+    },
+    {
+      "epoch": 2.6983308417111274,
+      "grad_norm": 0.32939786900837537,
+      "learning_rate": 4.004852350495744e-06,
+      "loss": 0.4211,
+      "step": 16393
+    },
+    {
+      "epoch": 2.698495425930244,
+      "grad_norm": 0.3725020408204437,
+      "learning_rate": 4.004385143707321e-06,
+      "loss": 0.4585,
+      "step": 16394
+    },
+    {
+      "epoch": 2.6986600101493603,
+      "grad_norm": 0.2853854565429146,
+      "learning_rate": 4.003917942326848e-06,
+      "loss": 0.4301,
+      "step": 16395
+    },
+    {
+      "epoch": 2.6988245943684768,
+      "grad_norm": 0.3735961596222982,
+      "learning_rate": 4.0034507463594254e-06,
+      "loss": 0.4314,
+      "step": 16396
+    },
+    {
+      "epoch": 2.6989891785875932,
+      "grad_norm": 0.3555759193710977,
+      "learning_rate": 4.002983555810149e-06,
+      "loss": 0.4408,
+      "step": 16397
+    },
+    {
+      "epoch": 2.6991537628067097,
+      "grad_norm": 0.3541206601416742,
+      "learning_rate": 4.00251637068412e-06,
+      "loss": 0.4236,
+      "step": 16398
+    },
+    {
+      "epoch": 2.6993183470258257,
+      "grad_norm": 0.2697075924755978,
+      "learning_rate": 4.00204919098643e-06,
+      "loss": 0.4272,
+      "step": 16399
+    },
+    {
+      "epoch": 2.699482931244942,
+      "grad_norm": 0.2698943695657363,
+      "learning_rate": 4.0015820167221844e-06,
+      "loss": 0.4351,
+      "step": 16400
+    },
+    {
+      "epoch": 2.6996475154640587,
+      "grad_norm": 0.3023080111451641,
+      "learning_rate": 4.001114847896476e-06,
+      "loss": 0.4313,
+      "step": 16401
+    },
+    {
+      "epoch": 2.699812099683175,
+      "grad_norm": 0.33379035571948806,
+      "learning_rate": 4.0006476845144015e-06,
+      "loss": 0.4331,
+      "step": 16402
+    },
+    {
+      "epoch": 2.6999766839022916,
+      "grad_norm": 0.4000340059845679,
+      "learning_rate": 4.000180526581062e-06,
+      "loss": 0.4484,
+      "step": 16403
+    },
+    {
+      "epoch": 2.700141268121408,
+      "grad_norm": 0.3095464294829794,
+      "learning_rate": 3.999713374101551e-06,
+      "loss": 0.4427,
+      "step": 16404
+    },
+    {
+      "epoch": 2.7003058523405246,
+      "grad_norm": 0.4313563968877461,
+      "learning_rate": 3.999246227080969e-06,
+      "loss": 0.4514,
+      "step": 16405
+    },
+    {
+      "epoch": 2.700470436559641,
+      "grad_norm": 0.3398078212298633,
+      "learning_rate": 3.998779085524413e-06,
+      "loss": 0.43,
+      "step": 16406
+    },
+    {
+      "epoch": 2.7006350207787575,
+      "grad_norm": 0.3391341204498403,
+      "learning_rate": 3.99831194943698e-06,
+      "loss": 0.4186,
+      "step": 16407
+    },
+    {
+      "epoch": 2.700799604997874,
+      "grad_norm": 0.39121165508808176,
+      "learning_rate": 3.997844818823767e-06,
+      "loss": 0.418,
+      "step": 16408
+    },
+    {
+      "epoch": 2.7009641892169904,
+      "grad_norm": 0.2978515677764767,
+      "learning_rate": 3.997377693689868e-06,
+      "loss": 0.4307,
+      "step": 16409
+    },
+    {
+      "epoch": 2.701128773436107,
+      "grad_norm": 0.269733507262398,
+      "learning_rate": 3.996910574040387e-06,
+      "loss": 0.4378,
+      "step": 16410
+    },
+    {
+      "epoch": 2.7012933576552234,
+      "grad_norm": 0.293333395883526,
+      "learning_rate": 3.996443459880414e-06,
+      "loss": 0.4246,
+      "step": 16411
+    },
+    {
+      "epoch": 2.70145794187434,
+      "grad_norm": 0.33065446535807685,
+      "learning_rate": 3.9959763512150505e-06,
+      "loss": 0.4373,
+      "step": 16412
+    },
+    {
+      "epoch": 2.7016225260934563,
+      "grad_norm": 0.4420974144520157,
+      "learning_rate": 3.99550924804939e-06,
+      "loss": 0.4277,
+      "step": 16413
+    },
+    {
+      "epoch": 2.701787110312573,
+      "grad_norm": 0.35927542344814284,
+      "learning_rate": 3.995042150388534e-06,
+      "loss": 0.4208,
+      "step": 16414
+    },
+    {
+      "epoch": 2.7019516945316893,
+      "grad_norm": 0.32015570904244817,
+      "learning_rate": 3.994575058237575e-06,
+      "loss": 0.4468,
+      "step": 16415
+    },
+    {
+      "epoch": 2.7021162787508057,
+      "grad_norm": 0.34575732730265396,
+      "learning_rate": 3.994107971601611e-06,
+      "loss": 0.4225,
+      "step": 16416
+    },
+    {
+      "epoch": 2.702280862969922,
+      "grad_norm": 0.30017266073248694,
+      "learning_rate": 3.9936408904857404e-06,
+      "loss": 0.4345,
+      "step": 16417
+    },
+    {
+      "epoch": 2.7024454471890387,
+      "grad_norm": 0.2841028088209639,
+      "learning_rate": 3.993173814895056e-06,
+      "loss": 0.4281,
+      "step": 16418
+    },
+    {
+      "epoch": 2.702610031408155,
+      "grad_norm": 0.3727144194184109,
+      "learning_rate": 3.9927067448346584e-06,
+      "loss": 0.4396,
+      "step": 16419
+    },
+    {
+      "epoch": 2.7027746156272716,
+      "grad_norm": 1.1118286108688196,
+      "learning_rate": 3.992239680309641e-06,
+      "loss": 0.4223,
+      "step": 16420
+    },
+    {
+      "epoch": 2.702939199846388,
+      "grad_norm": 0.33739735935023996,
+      "learning_rate": 3.991772621325103e-06,
+      "loss": 0.4294,
+      "step": 16421
+    },
+    {
+      "epoch": 2.7031037840655046,
+      "grad_norm": 0.3207517013717223,
+      "learning_rate": 3.991305567886139e-06,
+      "loss": 0.4354,
+      "step": 16422
+    },
+    {
+      "epoch": 2.703268368284621,
+      "grad_norm": 0.34552698173846835,
+      "learning_rate": 3.990838519997845e-06,
+      "loss": 0.4524,
+      "step": 16423
+    },
+    {
+      "epoch": 2.7034329525037375,
+      "grad_norm": 0.29631186943187254,
+      "learning_rate": 3.990371477665319e-06,
+      "loss": 0.432,
+      "step": 16424
+    },
+    {
+      "epoch": 2.703597536722854,
+      "grad_norm": 0.3395910937313192,
+      "learning_rate": 3.989904440893654e-06,
+      "loss": 0.43,
+      "step": 16425
+    },
+    {
+      "epoch": 2.7037621209419704,
+      "grad_norm": 0.30236926557935484,
+      "learning_rate": 3.98943740968795e-06,
+      "loss": 0.4171,
+      "step": 16426
+    },
+    {
+      "epoch": 2.703926705161087,
+      "grad_norm": 0.45268445138775065,
+      "learning_rate": 3.9889703840533e-06,
+      "loss": 0.451,
+      "step": 16427
+    },
+    {
+      "epoch": 2.7040912893802034,
+      "grad_norm": 0.2825909781347472,
+      "learning_rate": 3.9885033639948025e-06,
+      "loss": 0.4377,
+      "step": 16428
+    },
+    {
+      "epoch": 2.70425587359932,
+      "grad_norm": 0.6358393266769788,
+      "learning_rate": 3.988036349517551e-06,
+      "loss": 0.4415,
+      "step": 16429
+    },
+    {
+      "epoch": 2.7044204578184363,
+      "grad_norm": 0.3386341951236184,
+      "learning_rate": 3.987569340626644e-06,
+      "loss": 0.419,
+      "step": 16430
+    },
+    {
+      "epoch": 2.704585042037553,
+      "grad_norm": 0.43150716763894154,
+      "learning_rate": 3.987102337327176e-06,
+      "loss": 0.4085,
+      "step": 16431
+    },
+    {
+      "epoch": 2.7047496262566693,
+      "grad_norm": 0.33112369492597604,
+      "learning_rate": 3.986635339624241e-06,
+      "loss": 0.441,
+      "step": 16432
+    },
+    {
+      "epoch": 2.7049142104757857,
+      "grad_norm": 0.7055101510823322,
+      "learning_rate": 3.986168347522937e-06,
+      "loss": 0.4415,
+      "step": 16433
+    },
+    {
+      "epoch": 2.705078794694902,
+      "grad_norm": 0.3058808341878057,
+      "learning_rate": 3.985701361028358e-06,
+      "loss": 0.419,
+      "step": 16434
+    },
+    {
+      "epoch": 2.7052433789140187,
+      "grad_norm": 0.5188963271133904,
+      "learning_rate": 3.985234380145601e-06,
+      "loss": 0.4249,
+      "step": 16435
+    },
+    {
+      "epoch": 2.705407963133135,
+      "grad_norm": 0.36464984739773737,
+      "learning_rate": 3.984767404879761e-06,
+      "loss": 0.4299,
+      "step": 16436
+    },
+    {
+      "epoch": 2.7055725473522516,
+      "grad_norm": 0.3289296636196006,
+      "learning_rate": 3.9843004352359335e-06,
+      "loss": 0.4267,
+      "step": 16437
+    },
+    {
+      "epoch": 2.705737131571368,
+      "grad_norm": 0.33757350971555905,
+      "learning_rate": 3.983833471219215e-06,
+      "loss": 0.4125,
+      "step": 16438
+    },
+    {
+      "epoch": 2.7059017157904846,
+      "grad_norm": 0.31527342354316334,
+      "learning_rate": 3.983366512834697e-06,
+      "loss": 0.4364,
+      "step": 16439
+    },
+    {
+      "epoch": 2.706066300009601,
+      "grad_norm": 0.3019138030598178,
+      "learning_rate": 3.982899560087481e-06,
+      "loss": 0.4311,
+      "step": 16440
+    },
+    {
+      "epoch": 2.706230884228717,
+      "grad_norm": 0.31228858196188747,
+      "learning_rate": 3.982432612982656e-06,
+      "loss": 0.4314,
+      "step": 16441
+    },
+    {
+      "epoch": 2.7063954684478335,
+      "grad_norm": 0.3658219525437423,
+      "learning_rate": 3.98196567152532e-06,
+      "loss": 0.4518,
+      "step": 16442
+    },
+    {
+      "epoch": 2.70656005266695,
+      "grad_norm": 0.24508146841699552,
+      "learning_rate": 3.9814987357205685e-06,
+      "loss": 0.4343,
+      "step": 16443
+    },
+    {
+      "epoch": 2.7067246368860665,
+      "grad_norm": 0.31452290874256633,
+      "learning_rate": 3.981031805573496e-06,
+      "loss": 0.4207,
+      "step": 16444
+    },
+    {
+      "epoch": 2.706889221105183,
+      "grad_norm": 0.2755412575819989,
+      "learning_rate": 3.980564881089197e-06,
+      "loss": 0.4472,
+      "step": 16445
+    },
+    {
+      "epoch": 2.7070538053242994,
+      "grad_norm": 0.30094494464038807,
+      "learning_rate": 3.980097962272766e-06,
+      "loss": 0.4148,
+      "step": 16446
+    },
+    {
+      "epoch": 2.707218389543416,
+      "grad_norm": 0.3007444972335924,
+      "learning_rate": 3.9796310491293e-06,
+      "loss": 0.4177,
+      "step": 16447
+    },
+    {
+      "epoch": 2.7073829737625323,
+      "grad_norm": 0.3387548685818209,
+      "learning_rate": 3.97916414166389e-06,
+      "loss": 0.4534,
+      "step": 16448
+    },
+    {
+      "epoch": 2.707547557981649,
+      "grad_norm": 0.26601239553772493,
+      "learning_rate": 3.978697239881636e-06,
+      "loss": 0.4238,
+      "step": 16449
+    },
+    {
+      "epoch": 2.7077121422007653,
+      "grad_norm": 0.4986743261050538,
+      "learning_rate": 3.978230343787627e-06,
+      "loss": 0.4303,
+      "step": 16450
+    },
+    {
+      "epoch": 2.7078767264198818,
+      "grad_norm": 0.3267574436279148,
+      "learning_rate": 3.977763453386963e-06,
+      "loss": 0.4246,
+      "step": 16451
+    },
+    {
+      "epoch": 2.7080413106389982,
+      "grad_norm": 0.32715433919622156,
+      "learning_rate": 3.977296568684735e-06,
+      "loss": 0.4469,
+      "step": 16452
+    },
+    {
+      "epoch": 2.7082058948581147,
+      "grad_norm": 0.326640328060673,
+      "learning_rate": 3.976829689686037e-06,
+      "loss": 0.4384,
+      "step": 16453
+    },
+    {
+      "epoch": 2.708370479077231,
+      "grad_norm": 0.2731441484587725,
+      "learning_rate": 3.9763628163959646e-06,
+      "loss": 0.4244,
+      "step": 16454
+    },
+    {
+      "epoch": 2.7085350632963476,
+      "grad_norm": 0.29145356682928636,
+      "learning_rate": 3.975895948819612e-06,
+      "loss": 0.4303,
+      "step": 16455
+    },
+    {
+      "epoch": 2.708699647515464,
+      "grad_norm": 0.3826056619240171,
+      "learning_rate": 3.975429086962075e-06,
+      "loss": 0.4366,
+      "step": 16456
+    },
+    {
+      "epoch": 2.7088642317345806,
+      "grad_norm": 0.2903672346142324,
+      "learning_rate": 3.974962230828445e-06,
+      "loss": 0.43,
+      "step": 16457
+    },
+    {
+      "epoch": 2.709028815953697,
+      "grad_norm": 0.31055339133125603,
+      "learning_rate": 3.9744953804238184e-06,
+      "loss": 0.4243,
+      "step": 16458
+    },
+    {
+      "epoch": 2.7091934001728135,
+      "grad_norm": 0.30907072386314455,
+      "learning_rate": 3.97402853575329e-06,
+      "loss": 0.4558,
+      "step": 16459
+    },
+    {
+      "epoch": 2.70935798439193,
+      "grad_norm": 0.39217728225346443,
+      "learning_rate": 3.973561696821949e-06,
+      "loss": 0.4383,
+      "step": 16460
+    },
+    {
+      "epoch": 2.7095225686110465,
+      "grad_norm": 0.2949136541812381,
+      "learning_rate": 3.973094863634896e-06,
+      "loss": 0.4406,
+      "step": 16461
+    },
+    {
+      "epoch": 2.709687152830163,
+      "grad_norm": 0.2874774588586657,
+      "learning_rate": 3.972628036197219e-06,
+      "loss": 0.433,
+      "step": 16462
+    },
+    {
+      "epoch": 2.7098517370492794,
+      "grad_norm": 0.37726555702971565,
+      "learning_rate": 3.972161214514016e-06,
+      "loss": 0.4182,
+      "step": 16463
+    },
+    {
+      "epoch": 2.710016321268396,
+      "grad_norm": 0.3523421698654293,
+      "learning_rate": 3.971694398590378e-06,
+      "loss": 0.4313,
+      "step": 16464
+    },
+    {
+      "epoch": 2.7101809054875123,
+      "grad_norm": 0.35601099773408845,
+      "learning_rate": 3.9712275884314006e-06,
+      "loss": 0.4233,
+      "step": 16465
+    },
+    {
+      "epoch": 2.7103454897066284,
+      "grad_norm": 4.607132561798621,
+      "learning_rate": 3.9707607840421765e-06,
+      "loss": 0.4371,
+      "step": 16466
+    },
+    {
+      "epoch": 2.710510073925745,
+      "grad_norm": 0.32006363388751585,
+      "learning_rate": 3.9702939854277995e-06,
+      "loss": 0.4252,
+      "step": 16467
+    },
+    {
+      "epoch": 2.7106746581448613,
+      "grad_norm": 0.3703664092191253,
+      "learning_rate": 3.969827192593364e-06,
+      "loss": 0.4229,
+      "step": 16468
+    },
+    {
+      "epoch": 2.7108392423639778,
+      "grad_norm": 0.3535642572338403,
+      "learning_rate": 3.96936040554396e-06,
+      "loss": 0.4352,
+      "step": 16469
+    },
+    {
+      "epoch": 2.7110038265830942,
+      "grad_norm": 0.34953714902004956,
+      "learning_rate": 3.968893624284687e-06,
+      "loss": 0.4501,
+      "step": 16470
+    },
+    {
+      "epoch": 2.7111684108022107,
+      "grad_norm": 0.3261552636191001,
+      "learning_rate": 3.968426848820632e-06,
+      "loss": 0.4519,
+      "step": 16471
+    },
+    {
+      "epoch": 2.711332995021327,
+      "grad_norm": 0.4002886697589968,
+      "learning_rate": 3.967960079156893e-06,
+      "loss": 0.4319,
+      "step": 16472
+    },
+    {
+      "epoch": 2.7114975792404437,
+      "grad_norm": 0.43804861055115035,
+      "learning_rate": 3.96749331529856e-06,
+      "loss": 0.4346,
+      "step": 16473
+    },
+    {
+      "epoch": 2.71166216345956,
+      "grad_norm": 0.29537495233881805,
+      "learning_rate": 3.967026557250728e-06,
+      "loss": 0.4134,
+      "step": 16474
+    },
+    {
+      "epoch": 2.7118267476786766,
+      "grad_norm": 0.41638120350919683,
+      "learning_rate": 3.966559805018489e-06,
+      "loss": 0.3992,
+      "step": 16475
+    },
+    {
+      "epoch": 2.711991331897793,
+      "grad_norm": 0.3401077405746412,
+      "learning_rate": 3.966093058606936e-06,
+      "loss": 0.4158,
+      "step": 16476
+    },
+    {
+      "epoch": 2.7121559161169095,
+      "grad_norm": 0.49076143533819405,
+      "learning_rate": 3.9656263180211646e-06,
+      "loss": 0.4334,
+      "step": 16477
+    },
+    {
+      "epoch": 2.712320500336026,
+      "grad_norm": 0.35596719185727005,
+      "learning_rate": 3.965159583266263e-06,
+      "loss": 0.4405,
+      "step": 16478
+    },
+    {
+      "epoch": 2.7124850845551425,
+      "grad_norm": 0.2811140700107037,
+      "learning_rate": 3.9646928543473284e-06,
+      "loss": 0.4281,
+      "step": 16479
+    },
+    {
+      "epoch": 2.712649668774259,
+      "grad_norm": 0.36662815582599095,
+      "learning_rate": 3.964226131269451e-06,
+      "loss": 0.4426,
+      "step": 16480
+    },
+    {
+      "epoch": 2.7128142529933754,
+      "grad_norm": 0.3329050753007927,
+      "learning_rate": 3.963759414037725e-06,
+      "loss": 0.4567,
+      "step": 16481
+    },
+    {
+      "epoch": 2.712978837212492,
+      "grad_norm": 0.3967360921283001,
+      "learning_rate": 3.9632927026572415e-06,
+      "loss": 0.4515,
+      "step": 16482
+    },
+    {
+      "epoch": 2.7131434214316084,
+      "grad_norm": 0.4789891785118059,
+      "learning_rate": 3.962825997133094e-06,
+      "loss": 0.4336,
+      "step": 16483
+    },
+    {
+      "epoch": 2.713308005650725,
+      "grad_norm": 0.49288810478031103,
+      "learning_rate": 3.962359297470375e-06,
+      "loss": 0.433,
+      "step": 16484
+    },
+    {
+      "epoch": 2.7134725898698413,
+      "grad_norm": 0.374289866565762,
+      "learning_rate": 3.961892603674176e-06,
+      "loss": 0.419,
+      "step": 16485
+    },
+    {
+      "epoch": 2.7136371740889578,
+      "grad_norm": 0.339703176599694,
+      "learning_rate": 3.96142591574959e-06,
+      "loss": 0.4183,
+      "step": 16486
+    },
+    {
+      "epoch": 2.7138017583080742,
+      "grad_norm": 0.3705111284263445,
+      "learning_rate": 3.96095923370171e-06,
+      "loss": 0.4245,
+      "step": 16487
+    },
+    {
+      "epoch": 2.7139663425271907,
+      "grad_norm": 0.38620231082094414,
+      "learning_rate": 3.9604925575356285e-06,
+      "loss": 0.4077,
+      "step": 16488
+    },
+    {
+      "epoch": 2.714130926746307,
+      "grad_norm": 0.36650262469548656,
+      "learning_rate": 3.960025887256437e-06,
+      "loss": 0.4306,
+      "step": 16489
+    },
+    {
+      "epoch": 2.7142955109654237,
+      "grad_norm": 0.3655590240723453,
+      "learning_rate": 3.959559222869226e-06,
+      "loss": 0.4168,
+      "step": 16490
+    },
+    {
+      "epoch": 2.71446009518454,
+      "grad_norm": 0.5492373426547444,
+      "learning_rate": 3.959092564379091e-06,
+      "loss": 0.426,
+      "step": 16491
+    },
+    {
+      "epoch": 2.7146246794036566,
+      "grad_norm": 0.32705909412887524,
+      "learning_rate": 3.9586259117911205e-06,
+      "loss": 0.4088,
+      "step": 16492
+    },
+    {
+      "epoch": 2.714789263622773,
+      "grad_norm": 0.28925992733428907,
+      "learning_rate": 3.958159265110409e-06,
+      "loss": 0.4242,
+      "step": 16493
+    },
+    {
+      "epoch": 2.7149538478418895,
+      "grad_norm": 0.34237158057699313,
+      "learning_rate": 3.957692624342046e-06,
+      "loss": 0.4267,
+      "step": 16494
+    },
+    {
+      "epoch": 2.715118432061006,
+      "grad_norm": 0.36573089990123203,
+      "learning_rate": 3.957225989491125e-06,
+      "loss": 0.448,
+      "step": 16495
+    },
+    {
+      "epoch": 2.7152830162801225,
+      "grad_norm": 0.396777803147069,
+      "learning_rate": 3.956759360562738e-06,
+      "loss": 0.4295,
+      "step": 16496
+    },
+    {
+      "epoch": 2.715447600499239,
+      "grad_norm": 0.31646470571881147,
+      "learning_rate": 3.956292737561976e-06,
+      "loss": 0.4159,
+      "step": 16497
+    },
+    {
+      "epoch": 2.7156121847183554,
+      "grad_norm": 0.3603492209505748,
+      "learning_rate": 3.9558261204939305e-06,
+      "loss": 0.4007,
+      "step": 16498
+    },
+    {
+      "epoch": 2.715776768937472,
+      "grad_norm": 0.33269287449159657,
+      "learning_rate": 3.955359509363693e-06,
+      "loss": 0.4473,
+      "step": 16499
+    },
+    {
+      "epoch": 2.7159413531565884,
+      "grad_norm": 0.3743128484780302,
+      "learning_rate": 3.954892904176356e-06,
+      "loss": 0.4347,
+      "step": 16500
+    },
+    {
+      "epoch": 2.716105937375705,
+      "grad_norm": 0.31514688689823894,
+      "learning_rate": 3.954426304937008e-06,
+      "loss": 0.4237,
+      "step": 16501
+    },
+    {
+      "epoch": 2.7162705215948213,
+      "grad_norm": 1.7890578690514742,
+      "learning_rate": 3.953959711650745e-06,
+      "loss": 0.4232,
+      "step": 16502
+    },
+    {
+      "epoch": 2.7164351058139378,
+      "grad_norm": 0.30094650376522986,
+      "learning_rate": 3.953493124322655e-06,
+      "loss": 0.4214,
+      "step": 16503
+    },
+    {
+      "epoch": 2.7165996900330542,
+      "grad_norm": 0.3094817709581162,
+      "learning_rate": 3.953026542957829e-06,
+      "loss": 0.4319,
+      "step": 16504
+    },
+    {
+      "epoch": 2.7167642742521707,
+      "grad_norm": 0.34239085799206986,
+      "learning_rate": 3.9525599675613594e-06,
+      "loss": 0.4326,
+      "step": 16505
+    },
+    {
+      "epoch": 2.716928858471287,
+      "grad_norm": 0.3131738401297161,
+      "learning_rate": 3.952093398138336e-06,
+      "loss": 0.4406,
+      "step": 16506
+    },
+    {
+      "epoch": 2.717093442690403,
+      "grad_norm": 0.32423565735021526,
+      "learning_rate": 3.951626834693853e-06,
+      "loss": 0.4283,
+      "step": 16507
+    },
+    {
+      "epoch": 2.7172580269095197,
+      "grad_norm": 0.37070568272760424,
+      "learning_rate": 3.951160277232997e-06,
+      "loss": 0.4436,
+      "step": 16508
+    },
+    {
+      "epoch": 2.717422611128636,
+      "grad_norm": 0.2891131256851677,
+      "learning_rate": 3.950693725760863e-06,
+      "loss": 0.4357,
+      "step": 16509
+    },
+    {
+      "epoch": 2.7175871953477526,
+      "grad_norm": 0.28978087673201264,
+      "learning_rate": 3.950227180282538e-06,
+      "loss": 0.4372,
+      "step": 16510
+    },
+    {
+      "epoch": 2.717751779566869,
+      "grad_norm": 0.5355941888768845,
+      "learning_rate": 3.949760640803116e-06,
+      "loss": 0.4186,
+      "step": 16511
+    },
+    {
+      "epoch": 2.7179163637859856,
+      "grad_norm": 0.3526071429229731,
+      "learning_rate": 3.949294107327686e-06,
+      "loss": 0.4179,
+      "step": 16512
+    },
+    {
+      "epoch": 2.718080948005102,
+      "grad_norm": 0.5767881584871625,
+      "learning_rate": 3.948827579861338e-06,
+      "loss": 0.4172,
+      "step": 16513
+    },
+    {
+      "epoch": 2.7182455322242185,
+      "grad_norm": 0.7289027212814949,
+      "learning_rate": 3.948361058409165e-06,
+      "loss": 0.4297,
+      "step": 16514
+    },
+    {
+      "epoch": 2.718410116443335,
+      "grad_norm": 0.30909871256741495,
+      "learning_rate": 3.947894542976254e-06,
+      "loss": 0.436,
+      "step": 16515
+    },
+    {
+      "epoch": 2.7185747006624514,
+      "grad_norm": 0.3246466845888603,
+      "learning_rate": 3.947428033567699e-06,
+      "loss": 0.4501,
+      "step": 16516
+    },
+    {
+      "epoch": 2.718739284881568,
+      "grad_norm": 0.2665993992574491,
+      "learning_rate": 3.946961530188588e-06,
+      "loss": 0.4145,
+      "step": 16517
+    },
+    {
+      "epoch": 2.7189038691006844,
+      "grad_norm": 0.27713457592729585,
+      "learning_rate": 3.9464950328440124e-06,
+      "loss": 0.4148,
+      "step": 16518
+    },
+    {
+      "epoch": 2.719068453319801,
+      "grad_norm": 0.43622092377899335,
+      "learning_rate": 3.946028541539064e-06,
+      "loss": 0.4334,
+      "step": 16519
+    },
+    {
+      "epoch": 2.7192330375389173,
+      "grad_norm": 0.31195549073953505,
+      "learning_rate": 3.9455620562788275e-06,
+      "loss": 0.4259,
+      "step": 16520
+    },
+    {
+      "epoch": 2.719397621758034,
+      "grad_norm": 0.3053428947657628,
+      "learning_rate": 3.945095577068399e-06,
+      "loss": 0.4174,
+      "step": 16521
+    },
+    {
+      "epoch": 2.7195622059771503,
+      "grad_norm": 0.32858949379606484,
+      "learning_rate": 3.944629103912863e-06,
+      "loss": 0.44,
+      "step": 16522
+    },
+    {
+      "epoch": 2.7197267901962667,
+      "grad_norm": 0.3173134537251033,
+      "learning_rate": 3.944162636817316e-06,
+      "loss": 0.4361,
+      "step": 16523
+    },
+    {
+      "epoch": 2.719891374415383,
+      "grad_norm": 0.36359440024895223,
+      "learning_rate": 3.943696175786843e-06,
+      "loss": 0.4275,
+      "step": 16524
+    },
+    {
+      "epoch": 2.7200559586344997,
+      "grad_norm": 0.3136275602179564,
+      "learning_rate": 3.9432297208265365e-06,
+      "loss": 0.4161,
+      "step": 16525
+    },
+    {
+      "epoch": 2.720220542853616,
+      "grad_norm": 0.3560583689227809,
+      "learning_rate": 3.942763271941484e-06,
+      "loss": 0.4317,
+      "step": 16526
+    },
+    {
+      "epoch": 2.7203851270727326,
+      "grad_norm": 0.352877130841058,
+      "learning_rate": 3.942296829136776e-06,
+      "loss": 0.4453,
+      "step": 16527
+    },
+    {
+      "epoch": 2.720549711291849,
+      "grad_norm": 0.29121856925706796,
+      "learning_rate": 3.941830392417503e-06,
+      "loss": 0.4293,
+      "step": 16528
+    },
+    {
+      "epoch": 2.7207142955109656,
+      "grad_norm": 0.31819722571074455,
+      "learning_rate": 3.941363961788754e-06,
+      "loss": 0.4381,
+      "step": 16529
+    },
+    {
+      "epoch": 2.720878879730082,
+      "grad_norm": 0.27919041748825024,
+      "learning_rate": 3.940897537255619e-06,
+      "loss": 0.4351,
+      "step": 16530
+    },
+    {
+      "epoch": 2.7210434639491985,
+      "grad_norm": 0.3181876486582578,
+      "learning_rate": 3.940431118823185e-06,
+      "loss": 0.4341,
+      "step": 16531
+    },
+    {
+      "epoch": 2.7212080481683145,
+      "grad_norm": 0.4905460053945161,
+      "learning_rate": 3.939964706496546e-06,
+      "loss": 0.4485,
+      "step": 16532
+    },
+    {
+      "epoch": 2.721372632387431,
+      "grad_norm": 0.36060063288372324,
+      "learning_rate": 3.9394983002807875e-06,
+      "loss": 0.4614,
+      "step": 16533
+    },
+    {
+      "epoch": 2.7215372166065475,
+      "grad_norm": 0.2891799665529831,
+      "learning_rate": 3.939031900180999e-06,
+      "loss": 0.4301,
+      "step": 16534
+    },
+    {
+      "epoch": 2.721701800825664,
+      "grad_norm": 0.31653000551268723,
+      "learning_rate": 3.938565506202271e-06,
+      "loss": 0.4368,
+      "step": 16535
+    },
+    {
+      "epoch": 2.7218663850447804,
+      "grad_norm": 0.31446163576171354,
+      "learning_rate": 3.938099118349692e-06,
+      "loss": 0.4115,
+      "step": 16536
+    },
+    {
+      "epoch": 2.722030969263897,
+      "grad_norm": 0.4224125133332122,
+      "learning_rate": 3.9376327366283514e-06,
+      "loss": 0.4275,
+      "step": 16537
+    },
+    {
+      "epoch": 2.7221955534830133,
+      "grad_norm": 0.41937226295599567,
+      "learning_rate": 3.937166361043337e-06,
+      "loss": 0.4261,
+      "step": 16538
+    },
+    {
+      "epoch": 2.72236013770213,
+      "grad_norm": 0.32751387604214655,
+      "learning_rate": 3.93669999159974e-06,
+      "loss": 0.4322,
+      "step": 16539
+    },
+    {
+      "epoch": 2.7225247219212463,
+      "grad_norm": 0.31879617266630655,
+      "learning_rate": 3.936233628302649e-06,
+      "loss": 0.4266,
+      "step": 16540
+    },
+    {
+      "epoch": 2.7226893061403628,
+      "grad_norm": 0.2738330940521524,
+      "learning_rate": 3.935767271157148e-06,
+      "loss": 0.4184,
+      "step": 16541
+    },
+    {
+      "epoch": 2.7228538903594792,
+      "grad_norm": 0.33398045292163164,
+      "learning_rate": 3.935300920168334e-06,
+      "loss": 0.4391,
+      "step": 16542
+    },
+    {
+      "epoch": 2.7230184745785957,
+      "grad_norm": 0.35001300209930836,
+      "learning_rate": 3.934834575341287e-06,
+      "loss": 0.4128,
+      "step": 16543
+    },
+    {
+      "epoch": 2.723183058797712,
+      "grad_norm": 0.41861161950891773,
+      "learning_rate": 3.934368236681102e-06,
+      "loss": 0.4333,
+      "step": 16544
+    },
+    {
+      "epoch": 2.7233476430168286,
+      "grad_norm": 0.42298701569274116,
+      "learning_rate": 3.9339019041928625e-06,
+      "loss": 0.399,
+      "step": 16545
+    },
+    {
+      "epoch": 2.723512227235945,
+      "grad_norm": 0.30381499839029175,
+      "learning_rate": 3.9334355778816614e-06,
+      "loss": 0.4262,
+      "step": 16546
+    },
+    {
+      "epoch": 2.7236768114550616,
+      "grad_norm": 0.4276762118814055,
+      "learning_rate": 3.9329692577525854e-06,
+      "loss": 0.4185,
+      "step": 16547
+    },
+    {
+      "epoch": 2.723841395674178,
+      "grad_norm": 0.3489145625046093,
+      "learning_rate": 3.932502943810722e-06,
+      "loss": 0.4392,
+      "step": 16548
+    },
+    {
+      "epoch": 2.7240059798932945,
+      "grad_norm": 0.31725816333011786,
+      "learning_rate": 3.932036636061161e-06,
+      "loss": 0.4141,
+      "step": 16549
+    },
+    {
+      "epoch": 2.724170564112411,
+      "grad_norm": 0.6398241391454529,
+      "learning_rate": 3.931570334508987e-06,
+      "loss": 0.4138,
+      "step": 16550
+    },
+    {
+      "epoch": 2.7243351483315275,
+      "grad_norm": 0.4477747233220181,
+      "learning_rate": 3.931104039159293e-06,
+      "loss": 0.4398,
+      "step": 16551
+    },
+    {
+      "epoch": 2.724499732550644,
+      "grad_norm": 0.32522477304337494,
+      "learning_rate": 3.930637750017162e-06,
+      "loss": 0.4228,
+      "step": 16552
+    },
+    {
+      "epoch": 2.7246643167697604,
+      "grad_norm": 0.26621290472588405,
+      "learning_rate": 3.930171467087688e-06,
+      "loss": 0.4412,
+      "step": 16553
+    },
+    {
+      "epoch": 2.724828900988877,
+      "grad_norm": 0.31334048099794237,
+      "learning_rate": 3.929705190375953e-06,
+      "loss": 0.4332,
+      "step": 16554
+    },
+    {
+      "epoch": 2.7249934852079933,
+      "grad_norm": 0.4502108996851825,
+      "learning_rate": 3.929238919887049e-06,
+      "loss": 0.4309,
+      "step": 16555
+    },
+    {
+      "epoch": 2.72515806942711,
+      "grad_norm": 0.3080396753891864,
+      "learning_rate": 3.9287726556260615e-06,
+      "loss": 0.435,
+      "step": 16556
+    },
+    {
+      "epoch": 2.7253226536462263,
+      "grad_norm": 0.2974240979985078,
+      "learning_rate": 3.9283063975980785e-06,
+      "loss": 0.4384,
+      "step": 16557
+    },
+    {
+      "epoch": 2.7254872378653427,
+      "grad_norm": 0.3634519406751955,
+      "learning_rate": 3.927840145808188e-06,
+      "loss": 0.4346,
+      "step": 16558
+    },
+    {
+      "epoch": 2.725651822084459,
+      "grad_norm": 0.2961719414650674,
+      "learning_rate": 3.927373900261478e-06,
+      "loss": 0.4359,
+      "step": 16559
+    },
+    {
+      "epoch": 2.7258164063035757,
+      "grad_norm": 1.3632269864910096,
+      "learning_rate": 3.926907660963035e-06,
+      "loss": 0.4161,
+      "step": 16560
+    },
+    {
+      "epoch": 2.725980990522692,
+      "grad_norm": 0.48981983726866346,
+      "learning_rate": 3.926441427917946e-06,
+      "loss": 0.4361,
+      "step": 16561
+    },
+    {
+      "epoch": 2.7261455747418086,
+      "grad_norm": 0.3451352277184602,
+      "learning_rate": 3.925975201131302e-06,
+      "loss": 0.4428,
+      "step": 16562
+    },
+    {
+      "epoch": 2.726310158960925,
+      "grad_norm": 0.3423792475728588,
+      "learning_rate": 3.925508980608186e-06,
+      "loss": 0.4071,
+      "step": 16563
+    },
+    {
+      "epoch": 2.7264747431800416,
+      "grad_norm": 0.35651849165342947,
+      "learning_rate": 3.925042766353686e-06,
+      "loss": 0.4573,
+      "step": 16564
+    },
+    {
+      "epoch": 2.726639327399158,
+      "grad_norm": 0.33468366739995337,
+      "learning_rate": 3.9245765583728905e-06,
+      "loss": 0.4134,
+      "step": 16565
+    },
+    {
+      "epoch": 2.7268039116182745,
+      "grad_norm": 0.3264344724816696,
+      "learning_rate": 3.924110356670885e-06,
+      "loss": 0.4397,
+      "step": 16566
+    },
+    {
+      "epoch": 2.726968495837391,
+      "grad_norm": 0.2621226652285822,
+      "learning_rate": 3.923644161252759e-06,
+      "loss": 0.4345,
+      "step": 16567
+    },
+    {
+      "epoch": 2.7271330800565075,
+      "grad_norm": 0.38974860554227214,
+      "learning_rate": 3.923177972123597e-06,
+      "loss": 0.4474,
+      "step": 16568
+    },
+    {
+      "epoch": 2.727297664275624,
+      "grad_norm": 0.3402501223880213,
+      "learning_rate": 3.922711789288487e-06,
+      "loss": 0.4651,
+      "step": 16569
+    },
+    {
+      "epoch": 2.7274622484947404,
+      "grad_norm": 0.31851552409547534,
+      "learning_rate": 3.922245612752517e-06,
+      "loss": 0.432,
+      "step": 16570
+    },
+    {
+      "epoch": 2.727626832713857,
+      "grad_norm": 0.4068273023864668,
+      "learning_rate": 3.921779442520769e-06,
+      "loss": 0.4339,
+      "step": 16571
+    },
+    {
+      "epoch": 2.7277914169329733,
+      "grad_norm": 1.0664734260304998,
+      "learning_rate": 3.921313278598336e-06,
+      "loss": 0.4232,
+      "step": 16572
+    },
+    {
+      "epoch": 2.72795600115209,
+      "grad_norm": 0.3619878185276546,
+      "learning_rate": 3.920847120990299e-06,
+      "loss": 0.4427,
+      "step": 16573
+    },
+    {
+      "epoch": 2.728120585371206,
+      "grad_norm": 0.3263631633804949,
+      "learning_rate": 3.920380969701749e-06,
+      "loss": 0.4447,
+      "step": 16574
+    },
+    {
+      "epoch": 2.7282851695903223,
+      "grad_norm": 0.6487706367050764,
+      "learning_rate": 3.919914824737769e-06,
+      "loss": 0.4403,
+      "step": 16575
+    },
+    {
+      "epoch": 2.7284497538094388,
+      "grad_norm": 0.39630519759832494,
+      "learning_rate": 3.919448686103448e-06,
+      "loss": 0.4551,
+      "step": 16576
+    },
+    {
+      "epoch": 2.7286143380285552,
+      "grad_norm": 0.3550421170640398,
+      "learning_rate": 3.918982553803872e-06,
+      "loss": 0.4432,
+      "step": 16577
+    },
+    {
+      "epoch": 2.7287789222476717,
+      "grad_norm": 0.4224330935360739,
+      "learning_rate": 3.9185164278441245e-06,
+      "loss": 0.4201,
+      "step": 16578
+    },
+    {
+      "epoch": 2.728943506466788,
+      "grad_norm": 0.3213326665891005,
+      "learning_rate": 3.918050308229295e-06,
+      "loss": 0.4252,
+      "step": 16579
+    },
+    {
+      "epoch": 2.7291080906859047,
+      "grad_norm": 0.31491941534195894,
+      "learning_rate": 3.917584194964467e-06,
+      "loss": 0.4371,
+      "step": 16580
+    },
+    {
+      "epoch": 2.729272674905021,
+      "grad_norm": 0.35263803623632617,
+      "learning_rate": 3.917118088054731e-06,
+      "loss": 0.4346,
+      "step": 16581
+    },
+    {
+      "epoch": 2.7294372591241376,
+      "grad_norm": 0.33412122258213794,
+      "learning_rate": 3.916651987505166e-06,
+      "loss": 0.4307,
+      "step": 16582
+    },
+    {
+      "epoch": 2.729601843343254,
+      "grad_norm": 0.3795048520762357,
+      "learning_rate": 3.916185893320864e-06,
+      "loss": 0.4262,
+      "step": 16583
+    },
+    {
+      "epoch": 2.7297664275623705,
+      "grad_norm": 0.34055335812743365,
+      "learning_rate": 3.915719805506909e-06,
+      "loss": 0.4465,
+      "step": 16584
+    },
+    {
+      "epoch": 2.729931011781487,
+      "grad_norm": 0.30848973764445026,
+      "learning_rate": 3.915253724068384e-06,
+      "loss": 0.4478,
+      "step": 16585
+    },
+    {
+      "epoch": 2.7300955960006035,
+      "grad_norm": 0.3324198552438988,
+      "learning_rate": 3.9147876490103784e-06,
+      "loss": 0.4406,
+      "step": 16586
+    },
+    {
+      "epoch": 2.73026018021972,
+      "grad_norm": 0.2761581425577677,
+      "learning_rate": 3.914321580337975e-06,
+      "loss": 0.4362,
+      "step": 16587
+    },
+    {
+      "epoch": 2.7304247644388364,
+      "grad_norm": 0.2873771774415666,
+      "learning_rate": 3.913855518056263e-06,
+      "loss": 0.4433,
+      "step": 16588
+    },
+    {
+      "epoch": 2.730589348657953,
+      "grad_norm": 0.3365185935551332,
+      "learning_rate": 3.913389462170324e-06,
+      "loss": 0.4265,
+      "step": 16589
+    },
+    {
+      "epoch": 2.7307539328770694,
+      "grad_norm": 0.3202568834928568,
+      "learning_rate": 3.9129234126852455e-06,
+      "loss": 0.4205,
+      "step": 16590
+    },
+    {
+      "epoch": 2.730918517096186,
+      "grad_norm": 0.31466464812505096,
+      "learning_rate": 3.912457369606114e-06,
+      "loss": 0.4354,
+      "step": 16591
+    },
+    {
+      "epoch": 2.7310831013153023,
+      "grad_norm": 0.32197124336552835,
+      "learning_rate": 3.9119913329380105e-06,
+      "loss": 0.4313,
+      "step": 16592
+    },
+    {
+      "epoch": 2.7312476855344188,
+      "grad_norm": 0.3262108831936048,
+      "learning_rate": 3.911525302686025e-06,
+      "loss": 0.4475,
+      "step": 16593
+    },
+    {
+      "epoch": 2.7314122697535352,
+      "grad_norm": 0.3018341625636483,
+      "learning_rate": 3.911059278855239e-06,
+      "loss": 0.4259,
+      "step": 16594
+    },
+    {
+      "epoch": 2.7315768539726517,
+      "grad_norm": 0.5083638087103188,
+      "learning_rate": 3.9105932614507404e-06,
+      "loss": 0.4358,
+      "step": 16595
+    },
+    {
+      "epoch": 2.731741438191768,
+      "grad_norm": 0.39664415812454396,
+      "learning_rate": 3.910127250477611e-06,
+      "loss": 0.4399,
+      "step": 16596
+    },
+    {
+      "epoch": 2.7319060224108846,
+      "grad_norm": 0.30830712980996544,
+      "learning_rate": 3.90966124594094e-06,
+      "loss": 0.4512,
+      "step": 16597
+    },
+    {
+      "epoch": 2.732070606630001,
+      "grad_norm": 0.36449293175256364,
+      "learning_rate": 3.909195247845808e-06,
+      "loss": 0.4285,
+      "step": 16598
+    },
+    {
+      "epoch": 2.732235190849117,
+      "grad_norm": 0.2824279588786678,
+      "learning_rate": 3.9087292561973025e-06,
+      "loss": 0.4244,
+      "step": 16599
+    },
+    {
+      "epoch": 2.7323997750682336,
+      "grad_norm": 0.3732075192421565,
+      "learning_rate": 3.908263271000509e-06,
+      "loss": 0.4402,
+      "step": 16600
+    },
+    {
+      "epoch": 2.73256435928735,
+      "grad_norm": 0.3843866741677414,
+      "learning_rate": 3.907797292260507e-06,
+      "loss": 0.4059,
+      "step": 16601
+    },
+    {
+      "epoch": 2.7327289435064666,
+      "grad_norm": 0.5765114906730495,
+      "learning_rate": 3.907331319982388e-06,
+      "loss": 0.443,
+      "step": 16602
+    },
+    {
+      "epoch": 2.732893527725583,
+      "grad_norm": 0.3048172546078079,
+      "learning_rate": 3.9068653541712295e-06,
+      "loss": 0.4256,
+      "step": 16603
+    },
+    {
+      "epoch": 2.7330581119446995,
+      "grad_norm": 0.2729739798913893,
+      "learning_rate": 3.906399394832123e-06,
+      "loss": 0.4412,
+      "step": 16604
+    },
+    {
+      "epoch": 2.733222696163816,
+      "grad_norm": 0.28601395344284813,
+      "learning_rate": 3.905933441970147e-06,
+      "loss": 0.428,
+      "step": 16605
+    },
+    {
+      "epoch": 2.7333872803829324,
+      "grad_norm": 0.2899991559501144,
+      "learning_rate": 3.90546749559039e-06,
+      "loss": 0.4449,
+      "step": 16606
+    },
+    {
+      "epoch": 2.733551864602049,
+      "grad_norm": 0.383461257118734,
+      "learning_rate": 3.905001555697934e-06,
+      "loss": 0.4328,
+      "step": 16607
+    },
+    {
+      "epoch": 2.7337164488211654,
+      "grad_norm": 0.3344920575647727,
+      "learning_rate": 3.904535622297862e-06,
+      "loss": 0.4243,
+      "step": 16608
+    },
+    {
+      "epoch": 2.733881033040282,
+      "grad_norm": 0.311637742348854,
+      "learning_rate": 3.9040696953952615e-06,
+      "loss": 0.4287,
+      "step": 16609
+    },
+    {
+      "epoch": 2.7340456172593983,
+      "grad_norm": 0.39159452522404453,
+      "learning_rate": 3.9036037749952134e-06,
+      "loss": 0.4326,
+      "step": 16610
+    },
+    {
+      "epoch": 2.734210201478515,
+      "grad_norm": 0.3379876446766164,
+      "learning_rate": 3.903137861102804e-06,
+      "loss": 0.4384,
+      "step": 16611
+    },
+    {
+      "epoch": 2.7343747856976313,
+      "grad_norm": 0.32601922212061046,
+      "learning_rate": 3.902671953723115e-06,
+      "loss": 0.4263,
+      "step": 16612
+    },
+    {
+      "epoch": 2.7345393699167477,
+      "grad_norm": 0.3037710053756364,
+      "learning_rate": 3.902206052861233e-06,
+      "loss": 0.4291,
+      "step": 16613
+    },
+    {
+      "epoch": 2.734703954135864,
+      "grad_norm": 0.39634282364124573,
+      "learning_rate": 3.901740158522239e-06,
+      "loss": 0.4384,
+      "step": 16614
+    },
+    {
+      "epoch": 2.7348685383549807,
+      "grad_norm": 0.3083150422178533,
+      "learning_rate": 3.901274270711217e-06,
+      "loss": 0.4434,
+      "step": 16615
+    },
+    {
+      "epoch": 2.735033122574097,
+      "grad_norm": 0.2880021907018946,
+      "learning_rate": 3.900808389433251e-06,
+      "loss": 0.4219,
+      "step": 16616
+    },
+    {
+      "epoch": 2.7351977067932136,
+      "grad_norm": 0.28308455496725854,
+      "learning_rate": 3.900342514693425e-06,
+      "loss": 0.4293,
+      "step": 16617
+    },
+    {
+      "epoch": 2.73536229101233,
+      "grad_norm": 0.33136345258347305,
+      "learning_rate": 3.899876646496823e-06,
+      "loss": 0.42,
+      "step": 16618
+    },
+    {
+      "epoch": 2.7355268752314466,
+      "grad_norm": 0.30752082102442413,
+      "learning_rate": 3.899410784848526e-06,
+      "loss": 0.4239,
+      "step": 16619
+    },
+    {
+      "epoch": 2.735691459450563,
+      "grad_norm": 0.42132254252174034,
+      "learning_rate": 3.89894492975362e-06,
+      "loss": 0.4559,
+      "step": 16620
+    },
+    {
+      "epoch": 2.7358560436696795,
+      "grad_norm": 0.3928050979267358,
+      "learning_rate": 3.898479081217188e-06,
+      "loss": 0.4629,
+      "step": 16621
+    },
+    {
+      "epoch": 2.736020627888796,
+      "grad_norm": 0.4182456753045492,
+      "learning_rate": 3.89801323924431e-06,
+      "loss": 0.4231,
+      "step": 16622
+    },
+    {
+      "epoch": 2.7361852121079124,
+      "grad_norm": 0.401752609400978,
+      "learning_rate": 3.897547403840073e-06,
+      "loss": 0.4278,
+      "step": 16623
+    },
+    {
+      "epoch": 2.736349796327029,
+      "grad_norm": 0.3373797177237633,
+      "learning_rate": 3.897081575009557e-06,
+      "loss": 0.4372,
+      "step": 16624
+    },
+    {
+      "epoch": 2.7365143805461454,
+      "grad_norm": 0.31456143649250423,
+      "learning_rate": 3.896615752757847e-06,
+      "loss": 0.4234,
+      "step": 16625
+    },
+    {
+      "epoch": 2.736678964765262,
+      "grad_norm": 0.4505892755130963,
+      "learning_rate": 3.896149937090024e-06,
+      "loss": 0.4297,
+      "step": 16626
+    },
+    {
+      "epoch": 2.7368435489843783,
+      "grad_norm": 0.3310531708308449,
+      "learning_rate": 3.895684128011174e-06,
+      "loss": 0.4197,
+      "step": 16627
+    },
+    {
+      "epoch": 2.737008133203495,
+      "grad_norm": 0.32229938221905113,
+      "learning_rate": 3.895218325526376e-06,
+      "loss": 0.4275,
+      "step": 16628
+    },
+    {
+      "epoch": 2.7371727174226113,
+      "grad_norm": 0.3673752839728046,
+      "learning_rate": 3.894752529640714e-06,
+      "loss": 0.445,
+      "step": 16629
+    },
+    {
+      "epoch": 2.7373373016417277,
+      "grad_norm": 0.3024556838370898,
+      "learning_rate": 3.894286740359272e-06,
+      "loss": 0.4344,
+      "step": 16630
+    },
+    {
+      "epoch": 2.737501885860844,
+      "grad_norm": 0.355569626536372,
+      "learning_rate": 3.8938209576871305e-06,
+      "loss": 0.4331,
+      "step": 16631
+    },
+    {
+      "epoch": 2.7376664700799607,
+      "grad_norm": 0.3329281372530806,
+      "learning_rate": 3.893355181629374e-06,
+      "loss": 0.425,
+      "step": 16632
+    },
+    {
+      "epoch": 2.737831054299077,
+      "grad_norm": 0.2996473940409809,
+      "learning_rate": 3.892889412191082e-06,
+      "loss": 0.4515,
+      "step": 16633
+    },
+    {
+      "epoch": 2.7379956385181936,
+      "grad_norm": 0.31016466286381267,
+      "learning_rate": 3.8924236493773395e-06,
+      "loss": 0.431,
+      "step": 16634
+    },
+    {
+      "epoch": 2.73816022273731,
+      "grad_norm": 0.3464979108227558,
+      "learning_rate": 3.8919578931932275e-06,
+      "loss": 0.4303,
+      "step": 16635
+    },
+    {
+      "epoch": 2.7383248069564265,
+      "grad_norm": 0.31263854785111733,
+      "learning_rate": 3.891492143643829e-06,
+      "loss": 0.4384,
+      "step": 16636
+    },
+    {
+      "epoch": 2.738489391175543,
+      "grad_norm": 0.32162531400133154,
+      "learning_rate": 3.891026400734224e-06,
+      "loss": 0.4422,
+      "step": 16637
+    },
+    {
+      "epoch": 2.7386539753946595,
+      "grad_norm": 0.31048427444162374,
+      "learning_rate": 3.890560664469496e-06,
+      "loss": 0.4208,
+      "step": 16638
+    },
+    {
+      "epoch": 2.738818559613776,
+      "grad_norm": 0.3184756943376093,
+      "learning_rate": 3.890094934854727e-06,
+      "loss": 0.4388,
+      "step": 16639
+    },
+    {
+      "epoch": 2.7389831438328924,
+      "grad_norm": 0.32833610141557723,
+      "learning_rate": 3.889629211894999e-06,
+      "loss": 0.4252,
+      "step": 16640
+    },
+    {
+      "epoch": 2.7391477280520085,
+      "grad_norm": 0.32894370976124304,
+      "learning_rate": 3.889163495595393e-06,
+      "loss": 0.4323,
+      "step": 16641
+    },
+    {
+      "epoch": 2.739312312271125,
+      "grad_norm": 0.7708130390270826,
+      "learning_rate": 3.888697785960991e-06,
+      "loss": 0.4433,
+      "step": 16642
+    },
+    {
+      "epoch": 2.7394768964902414,
+      "grad_norm": 0.2399325823005569,
+      "learning_rate": 3.888232082996877e-06,
+      "loss": 0.4197,
+      "step": 16643
+    },
+    {
+      "epoch": 2.739641480709358,
+      "grad_norm": 0.700859157485469,
+      "learning_rate": 3.887766386708129e-06,
+      "loss": 0.4392,
+      "step": 16644
+    },
+    {
+      "epoch": 2.7398060649284743,
+      "grad_norm": 0.4217182459833526,
+      "learning_rate": 3.887300697099829e-06,
+      "loss": 0.4433,
+      "step": 16645
+    },
+    {
+      "epoch": 2.739970649147591,
+      "grad_norm": 0.46884546102603025,
+      "learning_rate": 3.8868350141770595e-06,
+      "loss": 0.4262,
+      "step": 16646
+    },
+    {
+      "epoch": 2.7401352333667073,
+      "grad_norm": 0.3110915519154355,
+      "learning_rate": 3.8863693379449015e-06,
+      "loss": 0.4337,
+      "step": 16647
+    },
+    {
+      "epoch": 2.7402998175858237,
+      "grad_norm": 0.3352084069388875,
+      "learning_rate": 3.8859036684084364e-06,
+      "loss": 0.4221,
+      "step": 16648
+    },
+    {
+      "epoch": 2.74046440180494,
+      "grad_norm": 0.36378279540182057,
+      "learning_rate": 3.885438005572746e-06,
+      "loss": 0.4214,
+      "step": 16649
+    },
+    {
+      "epoch": 2.7406289860240567,
+      "grad_norm": 0.3158512433778828,
+      "learning_rate": 3.884972349442911e-06,
+      "loss": 0.4154,
+      "step": 16650
+    },
+    {
+      "epoch": 2.740793570243173,
+      "grad_norm": 0.307548477999172,
+      "learning_rate": 3.8845067000240125e-06,
+      "loss": 0.4315,
+      "step": 16651
+    },
+    {
+      "epoch": 2.7409581544622896,
+      "grad_norm": 0.38812292332638837,
+      "learning_rate": 3.8840410573211286e-06,
+      "loss": 0.4481,
+      "step": 16652
+    },
+    {
+      "epoch": 2.741122738681406,
+      "grad_norm": 0.2783823676112584,
+      "learning_rate": 3.883575421339347e-06,
+      "loss": 0.4244,
+      "step": 16653
+    },
+    {
+      "epoch": 2.7412873229005226,
+      "grad_norm": 0.3557886541737026,
+      "learning_rate": 3.8831097920837395e-06,
+      "loss": 0.4439,
+      "step": 16654
+    },
+    {
+      "epoch": 2.741451907119639,
+      "grad_norm": 0.32008413498538474,
+      "learning_rate": 3.882644169559396e-06,
+      "loss": 0.4327,
+      "step": 16655
+    },
+    {
+      "epoch": 2.7416164913387555,
+      "grad_norm": 0.304571256600945,
+      "learning_rate": 3.882178553771391e-06,
+      "loss": 0.4448,
+      "step": 16656
+    },
+    {
+      "epoch": 2.741781075557872,
+      "grad_norm": 0.27985675276001454,
+      "learning_rate": 3.881712944724808e-06,
+      "loss": 0.4445,
+      "step": 16657
+    },
+    {
+      "epoch": 2.7419456597769885,
+      "grad_norm": 0.37539069739389663,
+      "learning_rate": 3.881247342424726e-06,
+      "loss": 0.4306,
+      "step": 16658
+    },
+    {
+      "epoch": 2.742110243996105,
+      "grad_norm": 0.3604991298614318,
+      "learning_rate": 3.880781746876226e-06,
+      "loss": 0.4246,
+      "step": 16659
+    },
+    {
+      "epoch": 2.7422748282152214,
+      "grad_norm": 0.34497239921676615,
+      "learning_rate": 3.880316158084388e-06,
+      "loss": 0.431,
+      "step": 16660
+    },
+    {
+      "epoch": 2.742439412434338,
+      "grad_norm": 0.32782234520476394,
+      "learning_rate": 3.879850576054293e-06,
+      "loss": 0.4266,
+      "step": 16661
+    },
+    {
+      "epoch": 2.7426039966534543,
+      "grad_norm": 0.355377901894709,
+      "learning_rate": 3.879385000791022e-06,
+      "loss": 0.4313,
+      "step": 16662
+    },
+    {
+      "epoch": 2.742768580872571,
+      "grad_norm": 0.3622834351699398,
+      "learning_rate": 3.878919432299652e-06,
+      "loss": 0.4421,
+      "step": 16663
+    },
+    {
+      "epoch": 2.7429331650916873,
+      "grad_norm": 0.3121404542685007,
+      "learning_rate": 3.878453870585268e-06,
+      "loss": 0.4405,
+      "step": 16664
+    },
+    {
+      "epoch": 2.7430977493108037,
+      "grad_norm": 0.5422525704023888,
+      "learning_rate": 3.877988315652946e-06,
+      "loss": 0.4363,
+      "step": 16665
+    },
+    {
+      "epoch": 2.7432623335299198,
+      "grad_norm": 0.33630247392790386,
+      "learning_rate": 3.877522767507767e-06,
+      "loss": 0.4226,
+      "step": 16666
+    },
+    {
+      "epoch": 2.7434269177490362,
+      "grad_norm": 0.3488597023051155,
+      "learning_rate": 3.877057226154812e-06,
+      "loss": 0.4228,
+      "step": 16667
+    },
+    {
+      "epoch": 2.7435915019681527,
+      "grad_norm": 0.5512200340488606,
+      "learning_rate": 3.876591691599159e-06,
+      "loss": 0.4241,
+      "step": 16668
+    },
+    {
+      "epoch": 2.743756086187269,
+      "grad_norm": 0.2967852679199231,
+      "learning_rate": 3.87612616384589e-06,
+      "loss": 0.451,
+      "step": 16669
+    },
+    {
+      "epoch": 2.7439206704063857,
+      "grad_norm": 0.329573071370927,
+      "learning_rate": 3.875660642900081e-06,
+      "loss": 0.4412,
+      "step": 16670
+    },
+    {
+      "epoch": 2.744085254625502,
+      "grad_norm": 0.3357303574720884,
+      "learning_rate": 3.875195128766815e-06,
+      "loss": 0.4352,
+      "step": 16671
+    },
+    {
+      "epoch": 2.7442498388446186,
+      "grad_norm": 0.3118697057421077,
+      "learning_rate": 3.874729621451172e-06,
+      "loss": 0.422,
+      "step": 16672
+    },
+    {
+      "epoch": 2.744414423063735,
+      "grad_norm": 0.324592192857279,
+      "learning_rate": 3.874264120958227e-06,
+      "loss": 0.433,
+      "step": 16673
+    },
+    {
+      "epoch": 2.7445790072828515,
+      "grad_norm": 0.356134316887983,
+      "learning_rate": 3.873798627293065e-06,
+      "loss": 0.4342,
+      "step": 16674
+    },
+    {
+      "epoch": 2.744743591501968,
+      "grad_norm": 0.5216140005545202,
+      "learning_rate": 3.873333140460761e-06,
+      "loss": 0.4235,
+      "step": 16675
+    },
+    {
+      "epoch": 2.7449081757210845,
+      "grad_norm": 0.5477538161873636,
+      "learning_rate": 3.872867660466396e-06,
+      "loss": 0.4384,
+      "step": 16676
+    },
+    {
+      "epoch": 2.745072759940201,
+      "grad_norm": 0.29347997209182203,
+      "learning_rate": 3.872402187315048e-06,
+      "loss": 0.4281,
+      "step": 16677
+    },
+    {
+      "epoch": 2.7452373441593174,
+      "grad_norm": 0.3099905198378355,
+      "learning_rate": 3.871936721011797e-06,
+      "loss": 0.4214,
+      "step": 16678
+    },
+    {
+      "epoch": 2.745401928378434,
+      "grad_norm": 0.4122803175887726,
+      "learning_rate": 3.8714712615617226e-06,
+      "loss": 0.445,
+      "step": 16679
+    },
+    {
+      "epoch": 2.7455665125975504,
+      "grad_norm": 0.9142629401607756,
+      "learning_rate": 3.8710058089699025e-06,
+      "loss": 0.4357,
+      "step": 16680
+    },
+    {
+      "epoch": 2.745731096816667,
+      "grad_norm": 0.3001133513388538,
+      "learning_rate": 3.870540363241417e-06,
+      "loss": 0.4291,
+      "step": 16681
+    },
+    {
+      "epoch": 2.7458956810357833,
+      "grad_norm": 0.4168866990618788,
+      "learning_rate": 3.8700749243813415e-06,
+      "loss": 0.4269,
+      "step": 16682
+    },
+    {
+      "epoch": 2.7460602652548998,
+      "grad_norm": 0.30645876076786355,
+      "learning_rate": 3.86960949239476e-06,
+      "loss": 0.4262,
+      "step": 16683
+    },
+    {
+      "epoch": 2.7462248494740162,
+      "grad_norm": 0.30916232701484836,
+      "learning_rate": 3.869144067286745e-06,
+      "loss": 0.4342,
+      "step": 16684
+    },
+    {
+      "epoch": 2.7463894336931327,
+      "grad_norm": 0.3309646231801186,
+      "learning_rate": 3.868678649062381e-06,
+      "loss": 0.4199,
+      "step": 16685
+    },
+    {
+      "epoch": 2.746554017912249,
+      "grad_norm": 0.36138802785749186,
+      "learning_rate": 3.868213237726742e-06,
+      "loss": 0.4446,
+      "step": 16686
+    },
+    {
+      "epoch": 2.7467186021313656,
+      "grad_norm": 0.33220801329003613,
+      "learning_rate": 3.8677478332849084e-06,
+      "loss": 0.4171,
+      "step": 16687
+    },
+    {
+      "epoch": 2.746883186350482,
+      "grad_norm": 0.30982777008522744,
+      "learning_rate": 3.867282435741959e-06,
+      "loss": 0.4282,
+      "step": 16688
+    },
+    {
+      "epoch": 2.7470477705695986,
+      "grad_norm": 0.36053955933065407,
+      "learning_rate": 3.8668170451029694e-06,
+      "loss": 0.4546,
+      "step": 16689
+    },
+    {
+      "epoch": 2.747212354788715,
+      "grad_norm": 0.3115291673408375,
+      "learning_rate": 3.8663516613730204e-06,
+      "loss": 0.4223,
+      "step": 16690
+    },
+    {
+      "epoch": 2.7473769390078315,
+      "grad_norm": 0.36712438720724067,
+      "learning_rate": 3.865886284557188e-06,
+      "loss": 0.4214,
+      "step": 16691
+    },
+    {
+      "epoch": 2.747541523226948,
+      "grad_norm": 0.4542084555280313,
+      "learning_rate": 3.865420914660553e-06,
+      "loss": 0.43,
+      "step": 16692
+    },
+    {
+      "epoch": 2.7477061074460645,
+      "grad_norm": 0.31832298830736644,
+      "learning_rate": 3.86495555168819e-06,
+      "loss": 0.4374,
+      "step": 16693
+    },
+    {
+      "epoch": 2.747870691665181,
+      "grad_norm": 0.3556434858838793,
+      "learning_rate": 3.864490195645182e-06,
+      "loss": 0.4368,
+      "step": 16694
+    },
+    {
+      "epoch": 2.7480352758842974,
+      "grad_norm": 0.3578218709637887,
+      "learning_rate": 3.864024846536602e-06,
+      "loss": 0.4255,
+      "step": 16695
+    },
+    {
+      "epoch": 2.748199860103414,
+      "grad_norm": 0.272539267466431,
+      "learning_rate": 3.863559504367527e-06,
+      "loss": 0.4247,
+      "step": 16696
+    },
+    {
+      "epoch": 2.7483644443225304,
+      "grad_norm": 0.36092119887107127,
+      "learning_rate": 3.863094169143038e-06,
+      "loss": 0.4139,
+      "step": 16697
+    },
+    {
+      "epoch": 2.748529028541647,
+      "grad_norm": 0.3138993620997971,
+      "learning_rate": 3.862628840868211e-06,
+      "loss": 0.4451,
+      "step": 16698
+    },
+    {
+      "epoch": 2.7486936127607633,
+      "grad_norm": 0.36947159209977626,
+      "learning_rate": 3.862163519548124e-06,
+      "loss": 0.4372,
+      "step": 16699
+    },
+    {
+      "epoch": 2.7488581969798798,
+      "grad_norm": 0.3174300936180199,
+      "learning_rate": 3.861698205187853e-06,
+      "loss": 0.4305,
+      "step": 16700
+    },
+    {
+      "epoch": 2.7490227811989962,
+      "grad_norm": 0.290319003808263,
+      "learning_rate": 3.861232897792478e-06,
+      "loss": 0.4328,
+      "step": 16701
+    },
+    {
+      "epoch": 2.7491873654181127,
+      "grad_norm": 0.3098957433837338,
+      "learning_rate": 3.860767597367076e-06,
+      "loss": 0.4244,
+      "step": 16702
+    },
+    {
+      "epoch": 2.749351949637229,
+      "grad_norm": 0.32129061982609347,
+      "learning_rate": 3.860302303916718e-06,
+      "loss": 0.4378,
+      "step": 16703
+    },
+    {
+      "epoch": 2.7495165338563456,
+      "grad_norm": 0.3653148104201589,
+      "learning_rate": 3.85983701744649e-06,
+      "loss": 0.4245,
+      "step": 16704
+    },
+    {
+      "epoch": 2.749681118075462,
+      "grad_norm": 0.7149367519498637,
+      "learning_rate": 3.859371737961464e-06,
+      "loss": 0.4298,
+      "step": 16705
+    },
+    {
+      "epoch": 2.7498457022945786,
+      "grad_norm": 0.36724382587372284,
+      "learning_rate": 3.858906465466718e-06,
+      "loss": 0.4278,
+      "step": 16706
+    },
+    {
+      "epoch": 2.7500102865136946,
+      "grad_norm": 0.2925452421742609,
+      "learning_rate": 3.858441199967328e-06,
+      "loss": 0.4211,
+      "step": 16707
+    },
+    {
+      "epoch": 2.750174870732811,
+      "grad_norm": 0.2683395976583217,
+      "learning_rate": 3.857975941468373e-06,
+      "loss": 0.4175,
+      "step": 16708
+    },
+    {
+      "epoch": 2.7503394549519276,
+      "grad_norm": 0.3136844198419058,
+      "learning_rate": 3.857510689974927e-06,
+      "loss": 0.447,
+      "step": 16709
+    },
+    {
+      "epoch": 2.750504039171044,
+      "grad_norm": 0.3030603614875939,
+      "learning_rate": 3.857045445492068e-06,
+      "loss": 0.4385,
+      "step": 16710
+    },
+    {
+      "epoch": 2.7506686233901605,
+      "grad_norm": 0.283294155437198,
+      "learning_rate": 3.856580208024873e-06,
+      "loss": 0.4033,
+      "step": 16711
+    },
+    {
+      "epoch": 2.750833207609277,
+      "grad_norm": 0.5659848655842254,
+      "learning_rate": 3.8561149775784174e-06,
+      "loss": 0.4346,
+      "step": 16712
+    },
+    {
+      "epoch": 2.7509977918283934,
+      "grad_norm": 0.3151413366114171,
+      "learning_rate": 3.85564975415778e-06,
+      "loss": 0.4431,
+      "step": 16713
+    },
+    {
+      "epoch": 2.75116237604751,
+      "grad_norm": 0.2758156307202911,
+      "learning_rate": 3.855184537768033e-06,
+      "loss": 0.4267,
+      "step": 16714
+    },
+    {
+      "epoch": 2.7513269602666264,
+      "grad_norm": 0.3780966181832298,
+      "learning_rate": 3.854719328414257e-06,
+      "loss": 0.4239,
+      "step": 16715
+    },
+    {
+      "epoch": 2.751491544485743,
+      "grad_norm": 0.2821931796559113,
+      "learning_rate": 3.854254126101525e-06,
+      "loss": 0.417,
+      "step": 16716
+    },
+    {
+      "epoch": 2.7516561287048593,
+      "grad_norm": 0.3266347677830889,
+      "learning_rate": 3.853788930834914e-06,
+      "loss": 0.4276,
+      "step": 16717
+    },
+    {
+      "epoch": 2.751820712923976,
+      "grad_norm": 0.312747836556435,
+      "learning_rate": 3.853323742619501e-06,
+      "loss": 0.4294,
+      "step": 16718
+    },
+    {
+      "epoch": 2.7519852971430923,
+      "grad_norm": 0.35643769445508905,
+      "learning_rate": 3.852858561460361e-06,
+      "loss": 0.4214,
+      "step": 16719
+    },
+    {
+      "epoch": 2.7521498813622087,
+      "grad_norm": 0.3929579469658608,
+      "learning_rate": 3.8523933873625695e-06,
+      "loss": 0.4401,
+      "step": 16720
+    },
+    {
+      "epoch": 2.752314465581325,
+      "grad_norm": 0.32304296447919406,
+      "learning_rate": 3.851928220331202e-06,
+      "loss": 0.4291,
+      "step": 16721
+    },
+    {
+      "epoch": 2.7524790498004417,
+      "grad_norm": 0.29631745968376494,
+      "learning_rate": 3.851463060371337e-06,
+      "loss": 0.4295,
+      "step": 16722
+    },
+    {
+      "epoch": 2.752643634019558,
+      "grad_norm": 0.31666231721168825,
+      "learning_rate": 3.850997907488048e-06,
+      "loss": 0.4301,
+      "step": 16723
+    },
+    {
+      "epoch": 2.7528082182386746,
+      "grad_norm": 0.3134924938039652,
+      "learning_rate": 3.850532761686411e-06,
+      "loss": 0.4151,
+      "step": 16724
+    },
+    {
+      "epoch": 2.752972802457791,
+      "grad_norm": 0.3341055846699733,
+      "learning_rate": 3.850067622971502e-06,
+      "loss": 0.4158,
+      "step": 16725
+    },
+    {
+      "epoch": 2.7531373866769075,
+      "grad_norm": 0.3944613390653974,
+      "learning_rate": 3.849602491348394e-06,
+      "loss": 0.4179,
+      "step": 16726
+    },
+    {
+      "epoch": 2.753301970896024,
+      "grad_norm": 0.34549310637003094,
+      "learning_rate": 3.849137366822165e-06,
+      "loss": 0.4228,
+      "step": 16727
+    },
+    {
+      "epoch": 2.7534665551151405,
+      "grad_norm": 0.48093379685876775,
+      "learning_rate": 3.848672249397888e-06,
+      "loss": 0.4329,
+      "step": 16728
+    },
+    {
+      "epoch": 2.753631139334257,
+      "grad_norm": 0.2653898037593713,
+      "learning_rate": 3.8482071390806405e-06,
+      "loss": 0.4263,
+      "step": 16729
+    },
+    {
+      "epoch": 2.7537957235533734,
+      "grad_norm": 0.3569534889161102,
+      "learning_rate": 3.8477420358754955e-06,
+      "loss": 0.4311,
+      "step": 16730
+    },
+    {
+      "epoch": 2.75396030777249,
+      "grad_norm": 0.37801092775394957,
+      "learning_rate": 3.847276939787531e-06,
+      "loss": 0.4154,
+      "step": 16731
+    },
+    {
+      "epoch": 2.754124891991606,
+      "grad_norm": 0.28751365365028175,
+      "learning_rate": 3.84681185082182e-06,
+      "loss": 0.4275,
+      "step": 16732
+    },
+    {
+      "epoch": 2.7542894762107224,
+      "grad_norm": 0.35444783230134985,
+      "learning_rate": 3.846346768983435e-06,
+      "loss": 0.4274,
+      "step": 16733
+    },
+    {
+      "epoch": 2.754454060429839,
+      "grad_norm": 0.35730086522911014,
+      "learning_rate": 3.845881694277456e-06,
+      "loss": 0.4165,
+      "step": 16734
+    },
+    {
+      "epoch": 2.7546186446489553,
+      "grad_norm": 0.34307664484838024,
+      "learning_rate": 3.845416626708952e-06,
+      "loss": 0.4241,
+      "step": 16735
+    },
+    {
+      "epoch": 2.754783228868072,
+      "grad_norm": 0.5066426180430484,
+      "learning_rate": 3.844951566283003e-06,
+      "loss": 0.4287,
+      "step": 16736
+    },
+    {
+      "epoch": 2.7549478130871883,
+      "grad_norm": 0.2992610263698568,
+      "learning_rate": 3.84448651300468e-06,
+      "loss": 0.4282,
+      "step": 16737
+    },
+    {
+      "epoch": 2.7551123973063048,
+      "grad_norm": 0.36298104637670003,
+      "learning_rate": 3.844021466879059e-06,
+      "loss": 0.4406,
+      "step": 16738
+    },
+    {
+      "epoch": 2.755276981525421,
+      "grad_norm": 0.25724806667635997,
+      "learning_rate": 3.843556427911214e-06,
+      "loss": 0.4136,
+      "step": 16739
+    },
+    {
+      "epoch": 2.7554415657445377,
+      "grad_norm": 0.29443173858665445,
+      "learning_rate": 3.843091396106218e-06,
+      "loss": 0.4388,
+      "step": 16740
+    },
+    {
+      "epoch": 2.755606149963654,
+      "grad_norm": 0.525197192040113,
+      "learning_rate": 3.842626371469148e-06,
+      "loss": 0.4149,
+      "step": 16741
+    },
+    {
+      "epoch": 2.7557707341827706,
+      "grad_norm": 0.31673011717278055,
+      "learning_rate": 3.842161354005076e-06,
+      "loss": 0.4333,
+      "step": 16742
+    },
+    {
+      "epoch": 2.755935318401887,
+      "grad_norm": 0.3159479306854577,
+      "learning_rate": 3.8416963437190776e-06,
+      "loss": 0.4299,
+      "step": 16743
+    },
+    {
+      "epoch": 2.7560999026210036,
+      "grad_norm": 0.3180567009797619,
+      "learning_rate": 3.841231340616224e-06,
+      "loss": 0.4415,
+      "step": 16744
+    },
+    {
+      "epoch": 2.75626448684012,
+      "grad_norm": 0.3046299149431181,
+      "learning_rate": 3.840766344701594e-06,
+      "loss": 0.431,
+      "step": 16745
+    },
+    {
+      "epoch": 2.7564290710592365,
+      "grad_norm": 0.32926788198908075,
+      "learning_rate": 3.840301355980257e-06,
+      "loss": 0.4296,
+      "step": 16746
+    },
+    {
+      "epoch": 2.756593655278353,
+      "grad_norm": 0.3202232197890211,
+      "learning_rate": 3.839836374457288e-06,
+      "loss": 0.4212,
+      "step": 16747
+    },
+    {
+      "epoch": 2.7567582394974695,
+      "grad_norm": 0.3292743896916637,
+      "learning_rate": 3.839371400137761e-06,
+      "loss": 0.4425,
+      "step": 16748
+    },
+    {
+      "epoch": 2.756922823716586,
+      "grad_norm": 0.3225566149924342,
+      "learning_rate": 3.838906433026749e-06,
+      "loss": 0.4294,
+      "step": 16749
+    },
+    {
+      "epoch": 2.7570874079357024,
+      "grad_norm": 0.43059850699838753,
+      "learning_rate": 3.838441473129328e-06,
+      "loss": 0.4374,
+      "step": 16750
+    },
+    {
+      "epoch": 2.757251992154819,
+      "grad_norm": 0.3935972124510023,
+      "learning_rate": 3.837976520450567e-06,
+      "loss": 0.4468,
+      "step": 16751
+    },
+    {
+      "epoch": 2.7574165763739353,
+      "grad_norm": 0.3989199727825013,
+      "learning_rate": 3.837511574995545e-06,
+      "loss": 0.4457,
+      "step": 16752
+    },
+    {
+      "epoch": 2.757581160593052,
+      "grad_norm": 0.3078040887317582,
+      "learning_rate": 3.837046636769331e-06,
+      "loss": 0.4484,
+      "step": 16753
+    },
+    {
+      "epoch": 2.7577457448121683,
+      "grad_norm": 0.4155164881385347,
+      "learning_rate": 3.836581705776998e-06,
+      "loss": 0.4235,
+      "step": 16754
+    },
+    {
+      "epoch": 2.7579103290312847,
+      "grad_norm": 0.9218484543356673,
+      "learning_rate": 3.836116782023624e-06,
+      "loss": 0.427,
+      "step": 16755
+    },
+    {
+      "epoch": 2.758074913250401,
+      "grad_norm": 0.4694404570897363,
+      "learning_rate": 3.835651865514277e-06,
+      "loss": 0.4346,
+      "step": 16756
+    },
+    {
+      "epoch": 2.7582394974695177,
+      "grad_norm": 0.3609226084141087,
+      "learning_rate": 3.835186956254031e-06,
+      "loss": 0.4463,
+      "step": 16757
+    },
+    {
+      "epoch": 2.758404081688634,
+      "grad_norm": 0.38001556363092925,
+      "learning_rate": 3.834722054247959e-06,
+      "loss": 0.4076,
+      "step": 16758
+    },
+    {
+      "epoch": 2.7585686659077506,
+      "grad_norm": 0.3146518163215466,
+      "learning_rate": 3.834257159501137e-06,
+      "loss": 0.4341,
+      "step": 16759
+    },
+    {
+      "epoch": 2.758733250126867,
+      "grad_norm": 0.3258618934792849,
+      "learning_rate": 3.8337922720186326e-06,
+      "loss": 0.4361,
+      "step": 16760
+    },
+    {
+      "epoch": 2.7588978343459836,
+      "grad_norm": 0.41980482455061724,
+      "learning_rate": 3.8333273918055226e-06,
+      "loss": 0.4028,
+      "step": 16761
+    },
+    {
+      "epoch": 2.7590624185651,
+      "grad_norm": 0.3035359627965373,
+      "learning_rate": 3.832862518866879e-06,
+      "loss": 0.4333,
+      "step": 16762
+    },
+    {
+      "epoch": 2.7592270027842165,
+      "grad_norm": 0.2792010265807511,
+      "learning_rate": 3.83239765320777e-06,
+      "loss": 0.4239,
+      "step": 16763
+    },
+    {
+      "epoch": 2.759391587003333,
+      "grad_norm": 0.3004060675697017,
+      "learning_rate": 3.8319327948332744e-06,
+      "loss": 0.4394,
+      "step": 16764
+    },
+    {
+      "epoch": 2.7595561712224494,
+      "grad_norm": 0.46446135062881594,
+      "learning_rate": 3.8314679437484594e-06,
+      "loss": 0.4369,
+      "step": 16765
+    },
+    {
+      "epoch": 2.759720755441566,
+      "grad_norm": 0.44409316368375035,
+      "learning_rate": 3.8310030999584014e-06,
+      "loss": 0.4264,
+      "step": 16766
+    },
+    {
+      "epoch": 2.7598853396606824,
+      "grad_norm": 0.3156353402627495,
+      "learning_rate": 3.830538263468169e-06,
+      "loss": 0.404,
+      "step": 16767
+    },
+    {
+      "epoch": 2.760049923879799,
+      "grad_norm": 0.5656328540429268,
+      "learning_rate": 3.830073434282837e-06,
+      "loss": 0.4191,
+      "step": 16768
+    },
+    {
+      "epoch": 2.7602145080989153,
+      "grad_norm": 0.34012852273365457,
+      "learning_rate": 3.829608612407476e-06,
+      "loss": 0.415,
+      "step": 16769
+    },
+    {
+      "epoch": 2.760379092318032,
+      "grad_norm": 0.3749393374332376,
+      "learning_rate": 3.829143797847157e-06,
+      "loss": 0.4375,
+      "step": 16770
+    },
+    {
+      "epoch": 2.7605436765371483,
+      "grad_norm": 0.35255460757698165,
+      "learning_rate": 3.828678990606955e-06,
+      "loss": 0.4348,
+      "step": 16771
+    },
+    {
+      "epoch": 2.7607082607562647,
+      "grad_norm": 0.26503899237119494,
+      "learning_rate": 3.828214190691939e-06,
+      "loss": 0.4285,
+      "step": 16772
+    },
+    {
+      "epoch": 2.760872844975381,
+      "grad_norm": 0.2901410596208873,
+      "learning_rate": 3.827749398107182e-06,
+      "loss": 0.4175,
+      "step": 16773
+    },
+    {
+      "epoch": 2.7610374291944972,
+      "grad_norm": 0.5798815384488164,
+      "learning_rate": 3.827284612857754e-06,
+      "loss": 0.4396,
+      "step": 16774
+    },
+    {
+      "epoch": 2.7612020134136137,
+      "grad_norm": 0.3729746218682766,
+      "learning_rate": 3.82681983494873e-06,
+      "loss": 0.4217,
+      "step": 16775
+    },
+    {
+      "epoch": 2.76136659763273,
+      "grad_norm": 0.3113239690103586,
+      "learning_rate": 3.826355064385179e-06,
+      "loss": 0.4367,
+      "step": 16776
+    },
+    {
+      "epoch": 2.7615311818518467,
+      "grad_norm": 0.3720192886770401,
+      "learning_rate": 3.825890301172171e-06,
+      "loss": 0.435,
+      "step": 16777
+    },
+    {
+      "epoch": 2.761695766070963,
+      "grad_norm": 0.29142275136633367,
+      "learning_rate": 3.82542554531478e-06,
+      "loss": 0.413,
+      "step": 16778
+    },
+    {
+      "epoch": 2.7618603502900796,
+      "grad_norm": 0.3192271465070386,
+      "learning_rate": 3.824960796818076e-06,
+      "loss": 0.4261,
+      "step": 16779
+    },
+    {
+      "epoch": 2.762024934509196,
+      "grad_norm": 0.3484695951419301,
+      "learning_rate": 3.82449605568713e-06,
+      "loss": 0.4347,
+      "step": 16780
+    },
+    {
+      "epoch": 2.7621895187283125,
+      "grad_norm": 0.31107231719861583,
+      "learning_rate": 3.824031321927014e-06,
+      "loss": 0.4563,
+      "step": 16781
+    },
+    {
+      "epoch": 2.762354102947429,
+      "grad_norm": 0.30327560992114966,
+      "learning_rate": 3.823566595542798e-06,
+      "loss": 0.4013,
+      "step": 16782
+    },
+    {
+      "epoch": 2.7625186871665455,
+      "grad_norm": 0.392686760340269,
+      "learning_rate": 3.823101876539556e-06,
+      "loss": 0.4263,
+      "step": 16783
+    },
+    {
+      "epoch": 2.762683271385662,
+      "grad_norm": 0.4008420205309258,
+      "learning_rate": 3.822637164922352e-06,
+      "loss": 0.4276,
+      "step": 16784
+    },
+    {
+      "epoch": 2.7628478556047784,
+      "grad_norm": 0.32373652463314856,
+      "learning_rate": 3.822172460696264e-06,
+      "loss": 0.4248,
+      "step": 16785
+    },
+    {
+      "epoch": 2.763012439823895,
+      "grad_norm": 0.32680014460348633,
+      "learning_rate": 3.821707763866358e-06,
+      "loss": 0.4492,
+      "step": 16786
+    },
+    {
+      "epoch": 2.7631770240430114,
+      "grad_norm": 0.35168595861593227,
+      "learning_rate": 3.821243074437706e-06,
+      "loss": 0.443,
+      "step": 16787
+    },
+    {
+      "epoch": 2.763341608262128,
+      "grad_norm": 0.3233521352290655,
+      "learning_rate": 3.820778392415379e-06,
+      "loss": 0.4318,
+      "step": 16788
+    },
+    {
+      "epoch": 2.7635061924812443,
+      "grad_norm": 0.5374254233166486,
+      "learning_rate": 3.820313717804448e-06,
+      "loss": 0.4281,
+      "step": 16789
+    },
+    {
+      "epoch": 2.7636707767003608,
+      "grad_norm": 0.35161183229666393,
+      "learning_rate": 3.819849050609982e-06,
+      "loss": 0.4144,
+      "step": 16790
+    },
+    {
+      "epoch": 2.7638353609194772,
+      "grad_norm": 0.46119190096611545,
+      "learning_rate": 3.81938439083705e-06,
+      "loss": 0.4266,
+      "step": 16791
+    },
+    {
+      "epoch": 2.7639999451385937,
+      "grad_norm": 0.2709732273695854,
+      "learning_rate": 3.818919738490726e-06,
+      "loss": 0.4288,
+      "step": 16792
+    },
+    {
+      "epoch": 2.76416452935771,
+      "grad_norm": 0.2927553770775444,
+      "learning_rate": 3.818455093576078e-06,
+      "loss": 0.4487,
+      "step": 16793
+    },
+    {
+      "epoch": 2.7643291135768266,
+      "grad_norm": 0.33049937178502276,
+      "learning_rate": 3.817990456098176e-06,
+      "loss": 0.4283,
+      "step": 16794
+    },
+    {
+      "epoch": 2.764493697795943,
+      "grad_norm": 0.5151043041028656,
+      "learning_rate": 3.817525826062088e-06,
+      "loss": 0.4361,
+      "step": 16795
+    },
+    {
+      "epoch": 2.7646582820150596,
+      "grad_norm": 0.44716660786155266,
+      "learning_rate": 3.8170612034728886e-06,
+      "loss": 0.4351,
+      "step": 16796
+    },
+    {
+      "epoch": 2.764822866234176,
+      "grad_norm": 0.3042095373540588,
+      "learning_rate": 3.8165965883356435e-06,
+      "loss": 0.4206,
+      "step": 16797
+    },
+    {
+      "epoch": 2.7649874504532925,
+      "grad_norm": 0.3198591642725671,
+      "learning_rate": 3.816131980655422e-06,
+      "loss": 0.446,
+      "step": 16798
+    },
+    {
+      "epoch": 2.7651520346724086,
+      "grad_norm": 0.3194612663957039,
+      "learning_rate": 3.815667380437298e-06,
+      "loss": 0.417,
+      "step": 16799
+    },
+    {
+      "epoch": 2.765316618891525,
+      "grad_norm": 0.5235493512889577,
+      "learning_rate": 3.815202787686337e-06,
+      "loss": 0.4101,
+      "step": 16800
+    },
+    {
+      "epoch": 2.7654812031106415,
+      "grad_norm": 0.2875249223023592,
+      "learning_rate": 3.8147382024076104e-06,
+      "loss": 0.4281,
+      "step": 16801
+    },
+    {
+      "epoch": 2.765645787329758,
+      "grad_norm": 0.4026276153596572,
+      "learning_rate": 3.8142736246061864e-06,
+      "loss": 0.4598,
+      "step": 16802
+    },
+    {
+      "epoch": 2.7658103715488744,
+      "grad_norm": 0.27404729719891796,
+      "learning_rate": 3.813809054287135e-06,
+      "loss": 0.4086,
+      "step": 16803
+    },
+    {
+      "epoch": 2.765974955767991,
+      "grad_norm": 0.3145397887623035,
+      "learning_rate": 3.8133444914555255e-06,
+      "loss": 0.4493,
+      "step": 16804
+    },
+    {
+      "epoch": 2.7661395399871074,
+      "grad_norm": 0.3397846067693523,
+      "learning_rate": 3.8128799361164277e-06,
+      "loss": 0.4198,
+      "step": 16805
+    },
+    {
+      "epoch": 2.766304124206224,
+      "grad_norm": 0.35722500103087296,
+      "learning_rate": 3.81241538827491e-06,
+      "loss": 0.4365,
+      "step": 16806
+    },
+    {
+      "epoch": 2.7664687084253403,
+      "grad_norm": 0.30238379013409583,
+      "learning_rate": 3.8119508479360402e-06,
+      "loss": 0.4608,
+      "step": 16807
+    },
+    {
+      "epoch": 2.766633292644457,
+      "grad_norm": 0.3440353515744379,
+      "learning_rate": 3.811486315104889e-06,
+      "loss": 0.4381,
+      "step": 16808
+    },
+    {
+      "epoch": 2.7667978768635733,
+      "grad_norm": 0.35348389908653965,
+      "learning_rate": 3.8110217897865228e-06,
+      "loss": 0.4347,
+      "step": 16809
+    },
+    {
+      "epoch": 2.7669624610826897,
+      "grad_norm": 0.5701370655291501,
+      "learning_rate": 3.810557271986013e-06,
+      "loss": 0.4434,
+      "step": 16810
+    },
+    {
+      "epoch": 2.767127045301806,
+      "grad_norm": 0.3437702405996044,
+      "learning_rate": 3.8100927617084262e-06,
+      "loss": 0.4401,
+      "step": 16811
+    },
+    {
+      "epoch": 2.7672916295209227,
+      "grad_norm": 0.3378670724201335,
+      "learning_rate": 3.8096282589588325e-06,
+      "loss": 0.4188,
+      "step": 16812
+    },
+    {
+      "epoch": 2.767456213740039,
+      "grad_norm": 0.35544705652232955,
+      "learning_rate": 3.8091637637423014e-06,
+      "loss": 0.4518,
+      "step": 16813
+    },
+    {
+      "epoch": 2.7676207979591556,
+      "grad_norm": 0.31464462787067055,
+      "learning_rate": 3.808699276063896e-06,
+      "loss": 0.4274,
+      "step": 16814
+    },
+    {
+      "epoch": 2.767785382178272,
+      "grad_norm": 0.2459535141240698,
+      "learning_rate": 3.808234795928692e-06,
+      "loss": 0.425,
+      "step": 16815
+    },
+    {
+      "epoch": 2.7679499663973886,
+      "grad_norm": 0.3086223489073096,
+      "learning_rate": 3.80777032334175e-06,
+      "loss": 0.4215,
+      "step": 16816
+    },
+    {
+      "epoch": 2.768114550616505,
+      "grad_norm": 0.26936310532731195,
+      "learning_rate": 3.807305858308145e-06,
+      "loss": 0.405,
+      "step": 16817
+    },
+    {
+      "epoch": 2.7682791348356215,
+      "grad_norm": 0.3025565113277676,
+      "learning_rate": 3.80684140083294e-06,
+      "loss": 0.437,
+      "step": 16818
+    },
+    {
+      "epoch": 2.768443719054738,
+      "grad_norm": 0.2957806754182231,
+      "learning_rate": 3.8063769509212065e-06,
+      "loss": 0.4352,
+      "step": 16819
+    },
+    {
+      "epoch": 2.7686083032738544,
+      "grad_norm": 0.33027851952704007,
+      "learning_rate": 3.8059125085780105e-06,
+      "loss": 0.4393,
+      "step": 16820
+    },
+    {
+      "epoch": 2.768772887492971,
+      "grad_norm": 0.36777594773645805,
+      "learning_rate": 3.8054480738084195e-06,
+      "loss": 0.4286,
+      "step": 16821
+    },
+    {
+      "epoch": 2.7689374717120874,
+      "grad_norm": 0.361574667511148,
+      "learning_rate": 3.804983646617503e-06,
+      "loss": 0.418,
+      "step": 16822
+    },
+    {
+      "epoch": 2.769102055931204,
+      "grad_norm": 0.36971154574152404,
+      "learning_rate": 3.804519227010326e-06,
+      "loss": 0.4427,
+      "step": 16823
+    },
+    {
+      "epoch": 2.7692666401503203,
+      "grad_norm": 0.32077205106904844,
+      "learning_rate": 3.8040548149919594e-06,
+      "loss": 0.4231,
+      "step": 16824
+    },
+    {
+      "epoch": 2.769431224369437,
+      "grad_norm": 0.2944188371284831,
+      "learning_rate": 3.8035904105674672e-06,
+      "loss": 0.4175,
+      "step": 16825
+    },
+    {
+      "epoch": 2.7695958085885533,
+      "grad_norm": 0.31903723649409144,
+      "learning_rate": 3.8031260137419207e-06,
+      "loss": 0.4325,
+      "step": 16826
+    },
+    {
+      "epoch": 2.7697603928076697,
+      "grad_norm": 0.3292297061095488,
+      "learning_rate": 3.802661624520384e-06,
+      "loss": 0.4273,
+      "step": 16827
+    },
+    {
+      "epoch": 2.769924977026786,
+      "grad_norm": 0.3541969818547337,
+      "learning_rate": 3.802197242907924e-06,
+      "loss": 0.4334,
+      "step": 16828
+    },
+    {
+      "epoch": 2.7700895612459027,
+      "grad_norm": 0.3672254462240884,
+      "learning_rate": 3.8017328689096106e-06,
+      "loss": 0.4251,
+      "step": 16829
+    },
+    {
+      "epoch": 2.770254145465019,
+      "grad_norm": 0.3457483782958529,
+      "learning_rate": 3.801268502530509e-06,
+      "loss": 0.4237,
+      "step": 16830
+    },
+    {
+      "epoch": 2.7704187296841356,
+      "grad_norm": 0.28709007258308455,
+      "learning_rate": 3.800804143775687e-06,
+      "loss": 0.4207,
+      "step": 16831
+    },
+    {
+      "epoch": 2.770583313903252,
+      "grad_norm": 0.2958632409294842,
+      "learning_rate": 3.8003397926502106e-06,
+      "loss": 0.4531,
+      "step": 16832
+    },
+    {
+      "epoch": 2.7707478981223685,
+      "grad_norm": 0.35996068001301706,
+      "learning_rate": 3.7998754491591478e-06,
+      "loss": 0.4378,
+      "step": 16833
+    },
+    {
+      "epoch": 2.770912482341485,
+      "grad_norm": 0.45725835910002477,
+      "learning_rate": 3.7994111133075653e-06,
+      "loss": 0.4312,
+      "step": 16834
+    },
+    {
+      "epoch": 2.7710770665606015,
+      "grad_norm": 0.37072152053424307,
+      "learning_rate": 3.7989467851005267e-06,
+      "loss": 0.4404,
+      "step": 16835
+    },
+    {
+      "epoch": 2.771241650779718,
+      "grad_norm": 0.38066903273037256,
+      "learning_rate": 3.7984824645431036e-06,
+      "loss": 0.4407,
+      "step": 16836
+    },
+    {
+      "epoch": 2.7714062349988344,
+      "grad_norm": 0.33648006260136515,
+      "learning_rate": 3.7980181516403584e-06,
+      "loss": 0.4347,
+      "step": 16837
+    },
+    {
+      "epoch": 2.771570819217951,
+      "grad_norm": 0.3286620878957204,
+      "learning_rate": 3.797553846397359e-06,
+      "loss": 0.4404,
+      "step": 16838
+    },
+    {
+      "epoch": 2.7717354034370674,
+      "grad_norm": 0.3275340096220997,
+      "learning_rate": 3.7970895488191717e-06,
+      "loss": 0.4444,
+      "step": 16839
+    },
+    {
+      "epoch": 2.771899987656184,
+      "grad_norm": 0.2860578424125304,
+      "learning_rate": 3.7966252589108634e-06,
+      "loss": 0.4433,
+      "step": 16840
+    },
+    {
+      "epoch": 2.7720645718753,
+      "grad_norm": 0.7913351638025952,
+      "learning_rate": 3.7961609766774994e-06,
+      "loss": 0.4224,
+      "step": 16841
+    },
+    {
+      "epoch": 2.7722291560944163,
+      "grad_norm": 0.4807048698131998,
+      "learning_rate": 3.795696702124145e-06,
+      "loss": 0.4401,
+      "step": 16842
+    },
+    {
+      "epoch": 2.772393740313533,
+      "grad_norm": 0.37526450105030956,
+      "learning_rate": 3.7952324352558676e-06,
+      "loss": 0.4269,
+      "step": 16843
+    },
+    {
+      "epoch": 2.7725583245326493,
+      "grad_norm": 0.2909990302337196,
+      "learning_rate": 3.7947681760777325e-06,
+      "loss": 0.4238,
+      "step": 16844
+    },
+    {
+      "epoch": 2.7727229087517657,
+      "grad_norm": 0.3265977697998563,
+      "learning_rate": 3.7943039245948074e-06,
+      "loss": 0.4365,
+      "step": 16845
+    },
+    {
+      "epoch": 2.772887492970882,
+      "grad_norm": 0.3940798235253495,
+      "learning_rate": 3.7938396808121525e-06,
+      "loss": 0.4415,
+      "step": 16846
+    },
+    {
+      "epoch": 2.7730520771899987,
+      "grad_norm": 0.320725253909283,
+      "learning_rate": 3.793375444734841e-06,
+      "loss": 0.4166,
+      "step": 16847
+    },
+    {
+      "epoch": 2.773216661409115,
+      "grad_norm": 0.3463614463862739,
+      "learning_rate": 3.7929112163679314e-06,
+      "loss": 0.4337,
+      "step": 16848
+    },
+    {
+      "epoch": 2.7733812456282316,
+      "grad_norm": 0.5034829550088746,
+      "learning_rate": 3.7924469957164952e-06,
+      "loss": 0.4265,
+      "step": 16849
+    },
+    {
+      "epoch": 2.773545829847348,
+      "grad_norm": 0.40248275758221413,
+      "learning_rate": 3.791982782785594e-06,
+      "loss": 0.4103,
+      "step": 16850
+    },
+    {
+      "epoch": 2.7737104140664646,
+      "grad_norm": 0.3980328594843701,
+      "learning_rate": 3.7915185775802934e-06,
+      "loss": 0.424,
+      "step": 16851
+    },
+    {
+      "epoch": 2.773874998285581,
+      "grad_norm": 0.3646532818983775,
+      "learning_rate": 3.7910543801056603e-06,
+      "loss": 0.4261,
+      "step": 16852
+    },
+    {
+      "epoch": 2.7740395825046975,
+      "grad_norm": 0.3362183017590112,
+      "learning_rate": 3.7905901903667576e-06,
+      "loss": 0.4297,
+      "step": 16853
+    },
+    {
+      "epoch": 2.774204166723814,
+      "grad_norm": 0.37016511030631144,
+      "learning_rate": 3.7901260083686523e-06,
+      "loss": 0.4294,
+      "step": 16854
+    },
+    {
+      "epoch": 2.7743687509429305,
+      "grad_norm": 0.361155081565757,
+      "learning_rate": 3.7896618341164087e-06,
+      "loss": 0.4351,
+      "step": 16855
+    },
+    {
+      "epoch": 2.774533335162047,
+      "grad_norm": 0.3273799523643519,
+      "learning_rate": 3.789197667615093e-06,
+      "loss": 0.4266,
+      "step": 16856
+    },
+    {
+      "epoch": 2.7746979193811634,
+      "grad_norm": 0.33376176425607657,
+      "learning_rate": 3.7887335088697676e-06,
+      "loss": 0.4245,
+      "step": 16857
+    },
+    {
+      "epoch": 2.77486250360028,
+      "grad_norm": 0.3355292741124985,
+      "learning_rate": 3.7882693578854975e-06,
+      "loss": 0.4354,
+      "step": 16858
+    },
+    {
+      "epoch": 2.7750270878193963,
+      "grad_norm": 0.5419348669241673,
+      "learning_rate": 3.787805214667349e-06,
+      "loss": 0.4349,
+      "step": 16859
+    },
+    {
+      "epoch": 2.775191672038513,
+      "grad_norm": 0.30543632162174206,
+      "learning_rate": 3.787341079220385e-06,
+      "loss": 0.4534,
+      "step": 16860
+    },
+    {
+      "epoch": 2.7753562562576293,
+      "grad_norm": 0.42483169974023705,
+      "learning_rate": 3.7868769515496715e-06,
+      "loss": 0.4455,
+      "step": 16861
+    },
+    {
+      "epoch": 2.7755208404767457,
+      "grad_norm": 0.28050471837453267,
+      "learning_rate": 3.7864128316602714e-06,
+      "loss": 0.4655,
+      "step": 16862
+    },
+    {
+      "epoch": 2.775685424695862,
+      "grad_norm": 0.3282592417587267,
+      "learning_rate": 3.78594871955725e-06,
+      "loss": 0.4241,
+      "step": 16863
+    },
+    {
+      "epoch": 2.7758500089149787,
+      "grad_norm": 0.2451870935424479,
+      "learning_rate": 3.7854846152456732e-06,
+      "loss": 0.4291,
+      "step": 16864
+    },
+    {
+      "epoch": 2.776014593134095,
+      "grad_norm": 0.32971710114845926,
+      "learning_rate": 3.7850205187305992e-06,
+      "loss": 0.4254,
+      "step": 16865
+    },
+    {
+      "epoch": 2.776179177353211,
+      "grad_norm": 0.3777895255987735,
+      "learning_rate": 3.7845564300170998e-06,
+      "loss": 0.4203,
+      "step": 16866
+    },
+    {
+      "epoch": 2.7763437615723277,
+      "grad_norm": 0.5279816235551633,
+      "learning_rate": 3.784092349110231e-06,
+      "loss": 0.4436,
+      "step": 16867
+    },
+    {
+      "epoch": 2.776508345791444,
+      "grad_norm": 0.2845430605160568,
+      "learning_rate": 3.7836282760150646e-06,
+      "loss": 0.4277,
+      "step": 16868
+    },
+    {
+      "epoch": 2.7766729300105606,
+      "grad_norm": 0.4033960800609004,
+      "learning_rate": 3.7831642107366587e-06,
+      "loss": 0.4309,
+      "step": 16869
+    },
+    {
+      "epoch": 2.776837514229677,
+      "grad_norm": 0.30830629475683824,
+      "learning_rate": 3.7827001532800793e-06,
+      "loss": 0.4338,
+      "step": 16870
+    },
+    {
+      "epoch": 2.7770020984487935,
+      "grad_norm": 0.27374971523185726,
+      "learning_rate": 3.7822361036503898e-06,
+      "loss": 0.4401,
+      "step": 16871
+    },
+    {
+      "epoch": 2.77716668266791,
+      "grad_norm": 0.36793554515865656,
+      "learning_rate": 3.7817720618526523e-06,
+      "loss": 0.4354,
+      "step": 16872
+    },
+    {
+      "epoch": 2.7773312668870265,
+      "grad_norm": 0.3654739408901122,
+      "learning_rate": 3.781308027891932e-06,
+      "loss": 0.4197,
+      "step": 16873
+    },
+    {
+      "epoch": 2.777495851106143,
+      "grad_norm": 0.2981515247509971,
+      "learning_rate": 3.7808440017732917e-06,
+      "loss": 0.4182,
+      "step": 16874
+    },
+    {
+      "epoch": 2.7776604353252594,
+      "grad_norm": 0.2876842090046348,
+      "learning_rate": 3.780379983501796e-06,
+      "loss": 0.4387,
+      "step": 16875
+    },
+    {
+      "epoch": 2.777825019544376,
+      "grad_norm": 0.5915380201347793,
+      "learning_rate": 3.779915973082504e-06,
+      "loss": 0.4339,
+      "step": 16876
+    },
+    {
+      "epoch": 2.7779896037634924,
+      "grad_norm": 0.47905263011333626,
+      "learning_rate": 3.779451970520484e-06,
+      "loss": 0.4493,
+      "step": 16877
+    },
+    {
+      "epoch": 2.778154187982609,
+      "grad_norm": 0.30270546075192667,
+      "learning_rate": 3.778987975820796e-06,
+      "loss": 0.4363,
+      "step": 16878
+    },
+    {
+      "epoch": 2.7783187722017253,
+      "grad_norm": 0.29825553662753534,
+      "learning_rate": 3.7785239889885024e-06,
+      "loss": 0.4263,
+      "step": 16879
+    },
+    {
+      "epoch": 2.7784833564208418,
+      "grad_norm": 0.33033442175320543,
+      "learning_rate": 3.778060010028668e-06,
+      "loss": 0.4245,
+      "step": 16880
+    },
+    {
+      "epoch": 2.7786479406399582,
+      "grad_norm": 0.36814904182310326,
+      "learning_rate": 3.7775960389463538e-06,
+      "loss": 0.443,
+      "step": 16881
+    },
+    {
+      "epoch": 2.7788125248590747,
+      "grad_norm": 0.43006987518328926,
+      "learning_rate": 3.777132075746624e-06,
+      "loss": 0.4367,
+      "step": 16882
+    },
+    {
+      "epoch": 2.778977109078191,
+      "grad_norm": 0.8368175177155547,
+      "learning_rate": 3.77666812043454e-06,
+      "loss": 0.4217,
+      "step": 16883
+    },
+    {
+      "epoch": 2.7791416932973076,
+      "grad_norm": 0.3555436805334529,
+      "learning_rate": 3.7762041730151654e-06,
+      "loss": 0.4227,
+      "step": 16884
+    },
+    {
+      "epoch": 2.779306277516424,
+      "grad_norm": 0.2839381903773783,
+      "learning_rate": 3.775740233493563e-06,
+      "loss": 0.4171,
+      "step": 16885
+    },
+    {
+      "epoch": 2.7794708617355406,
+      "grad_norm": 0.2696148806985721,
+      "learning_rate": 3.7752763018747915e-06,
+      "loss": 0.4255,
+      "step": 16886
+    },
+    {
+      "epoch": 2.779635445954657,
+      "grad_norm": 0.3281658213205818,
+      "learning_rate": 3.7748123781639183e-06,
+      "loss": 0.4477,
+      "step": 16887
+    },
+    {
+      "epoch": 2.7798000301737735,
+      "grad_norm": 0.38679247783092735,
+      "learning_rate": 3.7743484623660007e-06,
+      "loss": 0.4131,
+      "step": 16888
+    },
+    {
+      "epoch": 2.77996461439289,
+      "grad_norm": 0.33188086006927525,
+      "learning_rate": 3.7738845544861047e-06,
+      "loss": 0.4346,
+      "step": 16889
+    },
+    {
+      "epoch": 2.7801291986120065,
+      "grad_norm": 0.4873338195639668,
+      "learning_rate": 3.7734206545292896e-06,
+      "loss": 0.4251,
+      "step": 16890
+    },
+    {
+      "epoch": 2.780293782831123,
+      "grad_norm": 0.2832226392994431,
+      "learning_rate": 3.772956762500619e-06,
+      "loss": 0.448,
+      "step": 16891
+    },
+    {
+      "epoch": 2.7804583670502394,
+      "grad_norm": 0.3029700069785049,
+      "learning_rate": 3.7724928784051533e-06,
+      "loss": 0.4292,
+      "step": 16892
+    },
+    {
+      "epoch": 2.780622951269356,
+      "grad_norm": 0.30564716249735363,
+      "learning_rate": 3.772029002247956e-06,
+      "loss": 0.4342,
+      "step": 16893
+    },
+    {
+      "epoch": 2.7807875354884724,
+      "grad_norm": 0.3342439886152099,
+      "learning_rate": 3.7715651340340884e-06,
+      "loss": 0.4289,
+      "step": 16894
+    },
+    {
+      "epoch": 2.780952119707589,
+      "grad_norm": 0.3117023301162004,
+      "learning_rate": 3.7711012737686088e-06,
+      "loss": 0.4428,
+      "step": 16895
+    },
+    {
+      "epoch": 2.7811167039267053,
+      "grad_norm": 0.34039420971910117,
+      "learning_rate": 3.770637421456584e-06,
+      "loss": 0.4372,
+      "step": 16896
+    },
+    {
+      "epoch": 2.7812812881458218,
+      "grad_norm": 0.337347976138304,
+      "learning_rate": 3.7701735771030696e-06,
+      "loss": 0.4603,
+      "step": 16897
+    },
+    {
+      "epoch": 2.7814458723649382,
+      "grad_norm": 0.4024168661032073,
+      "learning_rate": 3.7697097407131326e-06,
+      "loss": 0.4501,
+      "step": 16898
+    },
+    {
+      "epoch": 2.7816104565840547,
+      "grad_norm": 0.36673311040074835,
+      "learning_rate": 3.76924591229183e-06,
+      "loss": 0.4253,
+      "step": 16899
+    },
+    {
+      "epoch": 2.781775040803171,
+      "grad_norm": 0.3466210120876975,
+      "learning_rate": 3.7687820918442248e-06,
+      "loss": 0.4126,
+      "step": 16900
+    },
+    {
+      "epoch": 2.7819396250222876,
+      "grad_norm": 0.4994253757949831,
+      "learning_rate": 3.768318279375378e-06,
+      "loss": 0.4371,
+      "step": 16901
+    },
+    {
+      "epoch": 2.782104209241404,
+      "grad_norm": 0.32390373399047606,
+      "learning_rate": 3.7678544748903487e-06,
+      "loss": 0.4394,
+      "step": 16902
+    },
+    {
+      "epoch": 2.7822687934605206,
+      "grad_norm": 0.467440952888479,
+      "learning_rate": 3.7673906783942002e-06,
+      "loss": 0.4308,
+      "step": 16903
+    },
+    {
+      "epoch": 2.782433377679637,
+      "grad_norm": 0.3679244911932761,
+      "learning_rate": 3.7669268898919917e-06,
+      "loss": 0.4397,
+      "step": 16904
+    },
+    {
+      "epoch": 2.7825979618987535,
+      "grad_norm": 0.28117900107872473,
+      "learning_rate": 3.7664631093887853e-06,
+      "loss": 0.4439,
+      "step": 16905
+    },
+    {
+      "epoch": 2.78276254611787,
+      "grad_norm": 0.3595577765149056,
+      "learning_rate": 3.765999336889639e-06,
+      "loss": 0.4472,
+      "step": 16906
+    },
+    {
+      "epoch": 2.7829271303369865,
+      "grad_norm": 0.3413085463391491,
+      "learning_rate": 3.7655355723996175e-06,
+      "loss": 0.4487,
+      "step": 16907
+    },
+    {
+      "epoch": 2.7830917145561025,
+      "grad_norm": 0.3171633221867142,
+      "learning_rate": 3.765071815923778e-06,
+      "loss": 0.4384,
+      "step": 16908
+    },
+    {
+      "epoch": 2.783256298775219,
+      "grad_norm": 0.41788203100764226,
+      "learning_rate": 3.76460806746718e-06,
+      "loss": 0.4313,
+      "step": 16909
+    },
+    {
+      "epoch": 2.7834208829943354,
+      "grad_norm": 0.33256616962038477,
+      "learning_rate": 3.7641443270348864e-06,
+      "loss": 0.4195,
+      "step": 16910
+    },
+    {
+      "epoch": 2.783585467213452,
+      "grad_norm": 0.4438550838839359,
+      "learning_rate": 3.763680594631956e-06,
+      "loss": 0.4469,
+      "step": 16911
+    },
+    {
+      "epoch": 2.7837500514325684,
+      "grad_norm": 0.3723740914529145,
+      "learning_rate": 3.763216870263449e-06,
+      "loss": 0.4182,
+      "step": 16912
+    },
+    {
+      "epoch": 2.783914635651685,
+      "grad_norm": 0.38287134229195346,
+      "learning_rate": 3.762753153934425e-06,
+      "loss": 0.4364,
+      "step": 16913
+    },
+    {
+      "epoch": 2.7840792198708013,
+      "grad_norm": 0.3048030213220403,
+      "learning_rate": 3.762289445649945e-06,
+      "loss": 0.4375,
+      "step": 16914
+    },
+    {
+      "epoch": 2.784243804089918,
+      "grad_norm": 0.318188085140894,
+      "learning_rate": 3.7618257454150693e-06,
+      "loss": 0.4243,
+      "step": 16915
+    },
+    {
+      "epoch": 2.7844083883090343,
+      "grad_norm": 0.34353950372326697,
+      "learning_rate": 3.7613620532348534e-06,
+      "loss": 0.4318,
+      "step": 16916
+    },
+    {
+      "epoch": 2.7845729725281507,
+      "grad_norm": 0.30445659026707206,
+      "learning_rate": 3.760898369114363e-06,
+      "loss": 0.417,
+      "step": 16917
+    },
+    {
+      "epoch": 2.784737556747267,
+      "grad_norm": 0.2983743769361913,
+      "learning_rate": 3.7604346930586528e-06,
+      "loss": 0.4367,
+      "step": 16918
+    },
+    {
+      "epoch": 2.7849021409663837,
+      "grad_norm": 0.30771506829920947,
+      "learning_rate": 3.7599710250727854e-06,
+      "loss": 0.4376,
+      "step": 16919
+    },
+    {
+      "epoch": 2.7850667251855,
+      "grad_norm": 0.35212283779278725,
+      "learning_rate": 3.7595073651618173e-06,
+      "loss": 0.4356,
+      "step": 16920
+    },
+    {
+      "epoch": 2.7852313094046166,
+      "grad_norm": 0.4328481200989893,
+      "learning_rate": 3.7590437133308104e-06,
+      "loss": 0.4232,
+      "step": 16921
+    },
+    {
+      "epoch": 2.785395893623733,
+      "grad_norm": 0.2879514901774684,
+      "learning_rate": 3.758580069584823e-06,
+      "loss": 0.4249,
+      "step": 16922
+    },
+    {
+      "epoch": 2.7855604778428495,
+      "grad_norm": 0.3926481862322684,
+      "learning_rate": 3.7581164339289125e-06,
+      "loss": 0.4157,
+      "step": 16923
+    },
+    {
+      "epoch": 2.785725062061966,
+      "grad_norm": 0.2779738532693352,
+      "learning_rate": 3.757652806368141e-06,
+      "loss": 0.42,
+      "step": 16924
+    },
+    {
+      "epoch": 2.7858896462810825,
+      "grad_norm": 0.2873667962660394,
+      "learning_rate": 3.7571891869075646e-06,
+      "loss": 0.434,
+      "step": 16925
+    },
+    {
+      "epoch": 2.786054230500199,
+      "grad_norm": 0.3209100044525901,
+      "learning_rate": 3.7567255755522452e-06,
+      "loss": 0.4236,
+      "step": 16926
+    },
+    {
+      "epoch": 2.7862188147193154,
+      "grad_norm": 0.2933805601306257,
+      "learning_rate": 3.7562619723072367e-06,
+      "loss": 0.4281,
+      "step": 16927
+    },
+    {
+      "epoch": 2.786383398938432,
+      "grad_norm": 0.3328129468806271,
+      "learning_rate": 3.755798377177604e-06,
+      "loss": 0.4346,
+      "step": 16928
+    },
+    {
+      "epoch": 2.7865479831575484,
+      "grad_norm": 0.3088428740042805,
+      "learning_rate": 3.7553347901683987e-06,
+      "loss": 0.4294,
+      "step": 16929
+    },
+    {
+      "epoch": 2.786712567376665,
+      "grad_norm": 0.35085409735244044,
+      "learning_rate": 3.7548712112846866e-06,
+      "loss": 0.4382,
+      "step": 16930
+    },
+    {
+      "epoch": 2.7868771515957813,
+      "grad_norm": 0.3807144512865194,
+      "learning_rate": 3.754407640531521e-06,
+      "loss": 0.4311,
+      "step": 16931
+    },
+    {
+      "epoch": 2.787041735814898,
+      "grad_norm": 0.48497774083128187,
+      "learning_rate": 3.753944077913961e-06,
+      "loss": 0.4377,
+      "step": 16932
+    },
+    {
+      "epoch": 2.787206320034014,
+      "grad_norm": 0.35729850499027854,
+      "learning_rate": 3.7534805234370668e-06,
+      "loss": 0.4393,
+      "step": 16933
+    },
+    {
+      "epoch": 2.7873709042531303,
+      "grad_norm": 0.5948625168020623,
+      "learning_rate": 3.7530169771058935e-06,
+      "loss": 0.4313,
+      "step": 16934
+    },
+    {
+      "epoch": 2.7875354884722467,
+      "grad_norm": 0.36432472655003784,
+      "learning_rate": 3.7525534389255022e-06,
+      "loss": 0.4484,
+      "step": 16935
+    },
+    {
+      "epoch": 2.787700072691363,
+      "grad_norm": 0.30984674335063966,
+      "learning_rate": 3.7520899089009486e-06,
+      "loss": 0.4332,
+      "step": 16936
+    },
+    {
+      "epoch": 2.7878646569104797,
+      "grad_norm": 0.7437331781268104,
+      "learning_rate": 3.7516263870372935e-06,
+      "loss": 0.4521,
+      "step": 16937
+    },
+    {
+      "epoch": 2.788029241129596,
+      "grad_norm": 0.4071443712642099,
+      "learning_rate": 3.7511628733395913e-06,
+      "loss": 0.4261,
+      "step": 16938
+    },
+    {
+      "epoch": 2.7881938253487126,
+      "grad_norm": 0.3537902345034585,
+      "learning_rate": 3.7506993678128995e-06,
+      "loss": 0.4321,
+      "step": 16939
+    },
+    {
+      "epoch": 2.788358409567829,
+      "grad_norm": 0.33455054322997646,
+      "learning_rate": 3.7502358704622792e-06,
+      "loss": 0.4215,
+      "step": 16940
+    },
+    {
+      "epoch": 2.7885229937869456,
+      "grad_norm": 0.3280338772128488,
+      "learning_rate": 3.7497723812927843e-06,
+      "loss": 0.4377,
+      "step": 16941
+    },
+    {
+      "epoch": 2.788687578006062,
+      "grad_norm": 0.34642732226049505,
+      "learning_rate": 3.7493089003094744e-06,
+      "loss": 0.4113,
+      "step": 16942
+    },
+    {
+      "epoch": 2.7888521622251785,
+      "grad_norm": 0.34188860854341835,
+      "learning_rate": 3.7488454275174056e-06,
+      "loss": 0.4193,
+      "step": 16943
+    },
+    {
+      "epoch": 2.789016746444295,
+      "grad_norm": 0.36554530240010263,
+      "learning_rate": 3.748381962921636e-06,
+      "loss": 0.4051,
+      "step": 16944
+    },
+    {
+      "epoch": 2.7891813306634115,
+      "grad_norm": 0.30335350273377804,
+      "learning_rate": 3.747918506527225e-06,
+      "loss": 0.4373,
+      "step": 16945
+    },
+    {
+      "epoch": 2.789345914882528,
+      "grad_norm": 0.34263542785101275,
+      "learning_rate": 3.747455058339223e-06,
+      "loss": 0.4121,
+      "step": 16946
+    },
+    {
+      "epoch": 2.7895104991016444,
+      "grad_norm": 0.5502643653929845,
+      "learning_rate": 3.7469916183626937e-06,
+      "loss": 0.4322,
+      "step": 16947
+    },
+    {
+      "epoch": 2.789675083320761,
+      "grad_norm": 0.3538609617300198,
+      "learning_rate": 3.7465281866026898e-06,
+      "loss": 0.4224,
+      "step": 16948
+    },
+    {
+      "epoch": 2.7898396675398773,
+      "grad_norm": 0.290319516274371,
+      "learning_rate": 3.7460647630642714e-06,
+      "loss": 0.4507,
+      "step": 16949
+    },
+    {
+      "epoch": 2.790004251758994,
+      "grad_norm": 0.3516323884400021,
+      "learning_rate": 3.7456013477524915e-06,
+      "loss": 0.4471,
+      "step": 16950
+    },
+    {
+      "epoch": 2.7901688359781103,
+      "grad_norm": 0.39564626942861164,
+      "learning_rate": 3.74513794067241e-06,
+      "loss": 0.4133,
+      "step": 16951
+    },
+    {
+      "epoch": 2.7903334201972267,
+      "grad_norm": 0.3694548077180987,
+      "learning_rate": 3.7446745418290816e-06,
+      "loss": 0.4272,
+      "step": 16952
+    },
+    {
+      "epoch": 2.790498004416343,
+      "grad_norm": 0.43330598934010617,
+      "learning_rate": 3.7442111512275626e-06,
+      "loss": 0.4454,
+      "step": 16953
+    },
+    {
+      "epoch": 2.7906625886354597,
+      "grad_norm": 0.29778710218876897,
+      "learning_rate": 3.7437477688729105e-06,
+      "loss": 0.4174,
+      "step": 16954
+    },
+    {
+      "epoch": 2.790827172854576,
+      "grad_norm": 0.42565006211331385,
+      "learning_rate": 3.7432843947701805e-06,
+      "loss": 0.4039,
+      "step": 16955
+    },
+    {
+      "epoch": 2.7909917570736926,
+      "grad_norm": 0.29606168440900626,
+      "learning_rate": 3.7428210289244293e-06,
+      "loss": 0.4128,
+      "step": 16956
+    },
+    {
+      "epoch": 2.791156341292809,
+      "grad_norm": 0.3511754169964407,
+      "learning_rate": 3.742357671340713e-06,
+      "loss": 0.435,
+      "step": 16957
+    },
+    {
+      "epoch": 2.7913209255119256,
+      "grad_norm": 0.4027590318484724,
+      "learning_rate": 3.7418943220240884e-06,
+      "loss": 0.4387,
+      "step": 16958
+    },
+    {
+      "epoch": 2.791485509731042,
+      "grad_norm": 0.3527348623847974,
+      "learning_rate": 3.7414309809796097e-06,
+      "loss": 0.4414,
+      "step": 16959
+    },
+    {
+      "epoch": 2.7916500939501585,
+      "grad_norm": 0.33095859367945063,
+      "learning_rate": 3.7409676482123323e-06,
+      "loss": 0.4443,
+      "step": 16960
+    },
+    {
+      "epoch": 2.791814678169275,
+      "grad_norm": 0.29026368944239717,
+      "learning_rate": 3.7405043237273143e-06,
+      "loss": 0.4319,
+      "step": 16961
+    },
+    {
+      "epoch": 2.7919792623883914,
+      "grad_norm": 0.32769594815474834,
+      "learning_rate": 3.740041007529609e-06,
+      "loss": 0.4364,
+      "step": 16962
+    },
+    {
+      "epoch": 2.792143846607508,
+      "grad_norm": 0.38324519531876516,
+      "learning_rate": 3.7395776996242737e-06,
+      "loss": 0.449,
+      "step": 16963
+    },
+    {
+      "epoch": 2.7923084308266244,
+      "grad_norm": 0.31201759520893474,
+      "learning_rate": 3.739114400016362e-06,
+      "loss": 0.4356,
+      "step": 16964
+    },
+    {
+      "epoch": 2.792473015045741,
+      "grad_norm": 0.32916106873222084,
+      "learning_rate": 3.7386511087109312e-06,
+      "loss": 0.4334,
+      "step": 16965
+    },
+    {
+      "epoch": 2.7926375992648573,
+      "grad_norm": 0.2519412393686649,
+      "learning_rate": 3.738187825713037e-06,
+      "loss": 0.4169,
+      "step": 16966
+    },
+    {
+      "epoch": 2.792802183483974,
+      "grad_norm": 0.3511877128500632,
+      "learning_rate": 3.7377245510277306e-06,
+      "loss": 0.429,
+      "step": 16967
+    },
+    {
+      "epoch": 2.7929667677030903,
+      "grad_norm": 0.2943120983708326,
+      "learning_rate": 3.7372612846600715e-06,
+      "loss": 0.4519,
+      "step": 16968
+    },
+    {
+      "epoch": 2.7931313519222067,
+      "grad_norm": 0.31550500106456264,
+      "learning_rate": 3.7367980266151117e-06,
+      "loss": 0.4291,
+      "step": 16969
+    },
+    {
+      "epoch": 2.793295936141323,
+      "grad_norm": 0.2594130714411502,
+      "learning_rate": 3.7363347768979084e-06,
+      "loss": 0.4216,
+      "step": 16970
+    },
+    {
+      "epoch": 2.7934605203604397,
+      "grad_norm": 0.3066082610071582,
+      "learning_rate": 3.7358715355135136e-06,
+      "loss": 0.4528,
+      "step": 16971
+    },
+    {
+      "epoch": 2.793625104579556,
+      "grad_norm": 0.2696462542228448,
+      "learning_rate": 3.735408302466985e-06,
+      "loss": 0.4357,
+      "step": 16972
+    },
+    {
+      "epoch": 2.7937896887986726,
+      "grad_norm": 0.3063848075313442,
+      "learning_rate": 3.734945077763375e-06,
+      "loss": 0.4342,
+      "step": 16973
+    },
+    {
+      "epoch": 2.7939542730177886,
+      "grad_norm": 0.2953028168617143,
+      "learning_rate": 3.7344818614077394e-06,
+      "loss": 0.427,
+      "step": 16974
+    },
+    {
+      "epoch": 2.794118857236905,
+      "grad_norm": 0.4258592757205851,
+      "learning_rate": 3.7340186534051324e-06,
+      "loss": 0.4474,
+      "step": 16975
+    },
+    {
+      "epoch": 2.7942834414560216,
+      "grad_norm": 0.27834239294905616,
+      "learning_rate": 3.7335554537606074e-06,
+      "loss": 0.442,
+      "step": 16976
+    },
+    {
+      "epoch": 2.794448025675138,
+      "grad_norm": 0.3230209326259235,
+      "learning_rate": 3.7330922624792216e-06,
+      "loss": 0.4276,
+      "step": 16977
+    },
+    {
+      "epoch": 2.7946126098942545,
+      "grad_norm": 0.3221333441755021,
+      "learning_rate": 3.732629079566024e-06,
+      "loss": 0.4458,
+      "step": 16978
+    },
+    {
+      "epoch": 2.794777194113371,
+      "grad_norm": 0.3676011024234717,
+      "learning_rate": 3.7321659050260737e-06,
+      "loss": 0.4368,
+      "step": 16979
+    },
+    {
+      "epoch": 2.7949417783324875,
+      "grad_norm": 0.2712662495086995,
+      "learning_rate": 3.7317027388644214e-06,
+      "loss": 0.4179,
+      "step": 16980
+    },
+    {
+      "epoch": 2.795106362551604,
+      "grad_norm": 0.30231518602792223,
+      "learning_rate": 3.731239581086123e-06,
+      "loss": 0.4042,
+      "step": 16981
+    },
+    {
+      "epoch": 2.7952709467707204,
+      "grad_norm": 1.6063668657993666,
+      "learning_rate": 3.730776431696231e-06,
+      "loss": 0.4352,
+      "step": 16982
+    },
+    {
+      "epoch": 2.795435530989837,
+      "grad_norm": 0.3833217390368779,
+      "learning_rate": 3.7303132906997994e-06,
+      "loss": 0.4282,
+      "step": 16983
+    },
+    {
+      "epoch": 2.7956001152089534,
+      "grad_norm": 0.3571528913706697,
+      "learning_rate": 3.7298501581018827e-06,
+      "loss": 0.4333,
+      "step": 16984
+    },
+    {
+      "epoch": 2.79576469942807,
+      "grad_norm": 0.4181591768839053,
+      "learning_rate": 3.729387033907533e-06,
+      "loss": 0.4391,
+      "step": 16985
+    },
+    {
+      "epoch": 2.7959292836471863,
+      "grad_norm": 0.3195424170629623,
+      "learning_rate": 3.7289239181218048e-06,
+      "loss": 0.4393,
+      "step": 16986
+    },
+    {
+      "epoch": 2.7960938678663028,
+      "grad_norm": 0.29082044594392087,
+      "learning_rate": 3.7284608107497503e-06,
+      "loss": 0.4378,
+      "step": 16987
+    },
+    {
+      "epoch": 2.7962584520854192,
+      "grad_norm": 0.3120472537805515,
+      "learning_rate": 3.727997711796426e-06,
+      "loss": 0.4211,
+      "step": 16988
+    },
+    {
+      "epoch": 2.7964230363045357,
+      "grad_norm": 0.2821648618544632,
+      "learning_rate": 3.7275346212668806e-06,
+      "loss": 0.4056,
+      "step": 16989
+    },
+    {
+      "epoch": 2.796587620523652,
+      "grad_norm": 0.46635738511994507,
+      "learning_rate": 3.727071539166169e-06,
+      "loss": 0.4369,
+      "step": 16990
+    },
+    {
+      "epoch": 2.7967522047427686,
+      "grad_norm": 0.34837292018747074,
+      "learning_rate": 3.7266084654993447e-06,
+      "loss": 0.4262,
+      "step": 16991
+    },
+    {
+      "epoch": 2.796916788961885,
+      "grad_norm": 0.3238602963355004,
+      "learning_rate": 3.7261454002714594e-06,
+      "loss": 0.4305,
+      "step": 16992
+    },
+    {
+      "epoch": 2.7970813731810016,
+      "grad_norm": 0.2700471790603966,
+      "learning_rate": 3.725682343487568e-06,
+      "loss": 0.4366,
+      "step": 16993
+    },
+    {
+      "epoch": 2.797245957400118,
+      "grad_norm": 0.32729950519911205,
+      "learning_rate": 3.72521929515272e-06,
+      "loss": 0.4444,
+      "step": 16994
+    },
+    {
+      "epoch": 2.7974105416192345,
+      "grad_norm": 0.3845240801769013,
+      "learning_rate": 3.7247562552719714e-06,
+      "loss": 0.4285,
+      "step": 16995
+    },
+    {
+      "epoch": 2.797575125838351,
+      "grad_norm": 0.303382662774618,
+      "learning_rate": 3.7242932238503742e-06,
+      "loss": 0.4372,
+      "step": 16996
+    },
+    {
+      "epoch": 2.7977397100574675,
+      "grad_norm": 0.3544453013233295,
+      "learning_rate": 3.7238302008929765e-06,
+      "loss": 0.448,
+      "step": 16997
+    },
+    {
+      "epoch": 2.797904294276584,
+      "grad_norm": 0.376369276483892,
+      "learning_rate": 3.7233671864048373e-06,
+      "loss": 0.452,
+      "step": 16998
+    },
+    {
+      "epoch": 2.7980688784957,
+      "grad_norm": 0.28156389454081876,
+      "learning_rate": 3.722904180391002e-06,
+      "loss": 0.4315,
+      "step": 16999
+    },
+    {
+      "epoch": 2.7982334627148164,
+      "grad_norm": 0.2676048084949686,
+      "learning_rate": 3.7224411828565293e-06,
+      "loss": 0.4146,
+      "step": 17000
+    },
+    {
+      "epoch": 2.798398046933933,
+      "grad_norm": 0.28636233503036546,
+      "learning_rate": 3.7219781938064657e-06,
+      "loss": 0.4321,
+      "step": 17001
+    },
+    {
+      "epoch": 2.7985626311530494,
+      "grad_norm": 0.28896095189299925,
+      "learning_rate": 3.721515213245867e-06,
+      "loss": 0.4383,
+      "step": 17002
+    },
+    {
+      "epoch": 2.798727215372166,
+      "grad_norm": 0.3737462974148848,
+      "learning_rate": 3.7210522411797837e-06,
+      "loss": 0.4433,
+      "step": 17003
+    },
+    {
+      "epoch": 2.7988917995912823,
+      "grad_norm": 0.33586658724596585,
+      "learning_rate": 3.7205892776132666e-06,
+      "loss": 0.4311,
+      "step": 17004
+    },
+    {
+      "epoch": 2.799056383810399,
+      "grad_norm": 0.3544658458659146,
+      "learning_rate": 3.720126322551368e-06,
+      "loss": 0.4241,
+      "step": 17005
+    },
+    {
+      "epoch": 2.7992209680295153,
+      "grad_norm": 0.3527511763637352,
+      "learning_rate": 3.71966337599914e-06,
+      "loss": 0.4472,
+      "step": 17006
+    },
+    {
+      "epoch": 2.7993855522486317,
+      "grad_norm": 0.3095155768711578,
+      "learning_rate": 3.7192004379616358e-06,
+      "loss": 0.4237,
+      "step": 17007
+    },
+    {
+      "epoch": 2.799550136467748,
+      "grad_norm": 1.0535976014865251,
+      "learning_rate": 3.7187375084439014e-06,
+      "loss": 0.4342,
+      "step": 17008
+    },
+    {
+      "epoch": 2.7997147206868647,
+      "grad_norm": 0.3112590515808455,
+      "learning_rate": 3.7182745874509945e-06,
+      "loss": 0.4138,
+      "step": 17009
+    },
+    {
+      "epoch": 2.799879304905981,
+      "grad_norm": 0.33246102620116796,
+      "learning_rate": 3.717811674987962e-06,
+      "loss": 0.4391,
+      "step": 17010
+    },
+    {
+      "epoch": 2.8000438891250976,
+      "grad_norm": 0.36308513014633353,
+      "learning_rate": 3.717348771059856e-06,
+      "loss": 0.4302,
+      "step": 17011
+    },
+    {
+      "epoch": 2.800208473344214,
+      "grad_norm": 0.43090264422151375,
+      "learning_rate": 3.716885875671728e-06,
+      "loss": 0.4362,
+      "step": 17012
+    },
+    {
+      "epoch": 2.8003730575633305,
+      "grad_norm": 0.3120753273054859,
+      "learning_rate": 3.7164229888286287e-06,
+      "loss": 0.412,
+      "step": 17013
+    },
+    {
+      "epoch": 2.800537641782447,
+      "grad_norm": 0.3137670501286763,
+      "learning_rate": 3.71596011053561e-06,
+      "loss": 0.4341,
+      "step": 17014
+    },
+    {
+      "epoch": 2.8007022260015635,
+      "grad_norm": 0.31210905483561246,
+      "learning_rate": 3.71549724079772e-06,
+      "loss": 0.4248,
+      "step": 17015
+    },
+    {
+      "epoch": 2.80086681022068,
+      "grad_norm": 0.3238692595647709,
+      "learning_rate": 3.7150343796200123e-06,
+      "loss": 0.427,
+      "step": 17016
+    },
+    {
+      "epoch": 2.8010313944397964,
+      "grad_norm": 0.3889252924366496,
+      "learning_rate": 3.714571527007535e-06,
+      "loss": 0.4369,
+      "step": 17017
+    },
+    {
+      "epoch": 2.801195978658913,
+      "grad_norm": 0.3874796764943105,
+      "learning_rate": 3.714108682965341e-06,
+      "loss": 0.4403,
+      "step": 17018
+    },
+    {
+      "epoch": 2.8013605628780294,
+      "grad_norm": 0.602048942349635,
+      "learning_rate": 3.7136458474984805e-06,
+      "loss": 0.4419,
+      "step": 17019
+    },
+    {
+      "epoch": 2.801525147097146,
+      "grad_norm": 0.4316973975224754,
+      "learning_rate": 3.7131830206120005e-06,
+      "loss": 0.4211,
+      "step": 17020
+    },
+    {
+      "epoch": 2.8016897313162623,
+      "grad_norm": 0.3025552469638141,
+      "learning_rate": 3.7127202023109545e-06,
+      "loss": 0.4494,
+      "step": 17021
+    },
+    {
+      "epoch": 2.801854315535379,
+      "grad_norm": 0.30613352483498824,
+      "learning_rate": 3.71225739260039e-06,
+      "loss": 0.4203,
+      "step": 17022
+    },
+    {
+      "epoch": 2.8020188997544953,
+      "grad_norm": 0.3390169431145897,
+      "learning_rate": 3.71179459148536e-06,
+      "loss": 0.431,
+      "step": 17023
+    },
+    {
+      "epoch": 2.8021834839736117,
+      "grad_norm": 0.32462568250008716,
+      "learning_rate": 3.7113317989709114e-06,
+      "loss": 0.4362,
+      "step": 17024
+    },
+    {
+      "epoch": 2.802348068192728,
+      "grad_norm": 0.3769661778823802,
+      "learning_rate": 3.710869015062097e-06,
+      "loss": 0.44,
+      "step": 17025
+    },
+    {
+      "epoch": 2.8025126524118447,
+      "grad_norm": 0.344819910321778,
+      "learning_rate": 3.7104062397639652e-06,
+      "loss": 0.4211,
+      "step": 17026
+    },
+    {
+      "epoch": 2.802677236630961,
+      "grad_norm": 0.28936962174503733,
+      "learning_rate": 3.709943473081563e-06,
+      "loss": 0.4193,
+      "step": 17027
+    },
+    {
+      "epoch": 2.8028418208500776,
+      "grad_norm": 0.3059987994253555,
+      "learning_rate": 3.7094807150199453e-06,
+      "loss": 0.4427,
+      "step": 17028
+    },
+    {
+      "epoch": 2.803006405069194,
+      "grad_norm": 0.3496518395650141,
+      "learning_rate": 3.7090179655841552e-06,
+      "loss": 0.4169,
+      "step": 17029
+    },
+    {
+      "epoch": 2.8031709892883105,
+      "grad_norm": 0.29300593577265654,
+      "learning_rate": 3.7085552247792484e-06,
+      "loss": 0.4374,
+      "step": 17030
+    },
+    {
+      "epoch": 2.803335573507427,
+      "grad_norm": 0.3291116815151797,
+      "learning_rate": 3.708092492610269e-06,
+      "loss": 0.4292,
+      "step": 17031
+    },
+    {
+      "epoch": 2.8035001577265435,
+      "grad_norm": 0.7757549504963241,
+      "learning_rate": 3.70762976908227e-06,
+      "loss": 0.4292,
+      "step": 17032
+    },
+    {
+      "epoch": 2.80366474194566,
+      "grad_norm": 0.3413843417394545,
+      "learning_rate": 3.707167054200298e-06,
+      "loss": 0.4519,
+      "step": 17033
+    },
+    {
+      "epoch": 2.8038293261647764,
+      "grad_norm": 0.3344160874415403,
+      "learning_rate": 3.706704347969402e-06,
+      "loss": 0.4169,
+      "step": 17034
+    },
+    {
+      "epoch": 2.803993910383893,
+      "grad_norm": 0.3432974498324751,
+      "learning_rate": 3.706241650394633e-06,
+      "loss": 0.4291,
+      "step": 17035
+    },
+    {
+      "epoch": 2.8041584946030094,
+      "grad_norm": 0.327641039036722,
+      "learning_rate": 3.7057789614810363e-06,
+      "loss": 0.4333,
+      "step": 17036
+    },
+    {
+      "epoch": 2.804323078822126,
+      "grad_norm": 0.3558890376868826,
+      "learning_rate": 3.7053162812336638e-06,
+      "loss": 0.4206,
+      "step": 17037
+    },
+    {
+      "epoch": 2.8044876630412423,
+      "grad_norm": 0.30115121607919093,
+      "learning_rate": 3.7048536096575627e-06,
+      "loss": 0.4199,
+      "step": 17038
+    },
+    {
+      "epoch": 2.8046522472603588,
+      "grad_norm": 1.5654426664598493,
+      "learning_rate": 3.704390946757783e-06,
+      "loss": 0.436,
+      "step": 17039
+    },
+    {
+      "epoch": 2.8048168314794752,
+      "grad_norm": 0.3369589627104296,
+      "learning_rate": 3.7039282925393706e-06,
+      "loss": 0.4327,
+      "step": 17040
+    },
+    {
+      "epoch": 2.8049814156985913,
+      "grad_norm": 0.2708386930655964,
+      "learning_rate": 3.7034656470073745e-06,
+      "loss": 0.4139,
+      "step": 17041
+    },
+    {
+      "epoch": 2.8051459999177077,
+      "grad_norm": 0.31847056505151394,
+      "learning_rate": 3.7030030101668444e-06,
+      "loss": 0.4155,
+      "step": 17042
+    },
+    {
+      "epoch": 2.805310584136824,
+      "grad_norm": 0.36442981805925195,
+      "learning_rate": 3.702540382022826e-06,
+      "loss": 0.4248,
+      "step": 17043
+    },
+    {
+      "epoch": 2.8054751683559407,
+      "grad_norm": 0.29620262128536895,
+      "learning_rate": 3.70207776258037e-06,
+      "loss": 0.4361,
+      "step": 17044
+    },
+    {
+      "epoch": 2.805639752575057,
+      "grad_norm": 0.9831257028070958,
+      "learning_rate": 3.7016151518445214e-06,
+      "loss": 0.4267,
+      "step": 17045
+    },
+    {
+      "epoch": 2.8058043367941736,
+      "grad_norm": 0.3674522650750558,
+      "learning_rate": 3.7011525498203315e-06,
+      "loss": 0.4439,
+      "step": 17046
+    },
+    {
+      "epoch": 2.80596892101329,
+      "grad_norm": 0.44696920449614114,
+      "learning_rate": 3.7006899565128467e-06,
+      "loss": 0.4549,
+      "step": 17047
+    },
+    {
+      "epoch": 2.8061335052324066,
+      "grad_norm": 0.48658489052050297,
+      "learning_rate": 3.7002273719271106e-06,
+      "loss": 0.4421,
+      "step": 17048
+    },
+    {
+      "epoch": 2.806298089451523,
+      "grad_norm": 0.28648773556330664,
+      "learning_rate": 3.699764796068178e-06,
+      "loss": 0.4522,
+      "step": 17049
+    },
+    {
+      "epoch": 2.8064626736706395,
+      "grad_norm": 0.6411059808424378,
+      "learning_rate": 3.69930222894109e-06,
+      "loss": 0.4365,
+      "step": 17050
+    },
+    {
+      "epoch": 2.806627257889756,
+      "grad_norm": 0.33897861852767236,
+      "learning_rate": 3.6988396705508977e-06,
+      "loss": 0.4445,
+      "step": 17051
+    },
+    {
+      "epoch": 2.8067918421088724,
+      "grad_norm": 0.3062163499760631,
+      "learning_rate": 3.6983771209026466e-06,
+      "loss": 0.4368,
+      "step": 17052
+    },
+    {
+      "epoch": 2.806956426327989,
+      "grad_norm": 0.29910086203118597,
+      "learning_rate": 3.6979145800013855e-06,
+      "loss": 0.4156,
+      "step": 17053
+    },
+    {
+      "epoch": 2.8071210105471054,
+      "grad_norm": 0.33507309223228005,
+      "learning_rate": 3.6974520478521587e-06,
+      "loss": 0.4207,
+      "step": 17054
+    },
+    {
+      "epoch": 2.807285594766222,
+      "grad_norm": 0.35606412297007517,
+      "learning_rate": 3.696989524460016e-06,
+      "loss": 0.411,
+      "step": 17055
+    },
+    {
+      "epoch": 2.8074501789853383,
+      "grad_norm": 0.4319969872987501,
+      "learning_rate": 3.6965270098300035e-06,
+      "loss": 0.4558,
+      "step": 17056
+    },
+    {
+      "epoch": 2.807614763204455,
+      "grad_norm": 0.3174102870346743,
+      "learning_rate": 3.6960645039671665e-06,
+      "loss": 0.4363,
+      "step": 17057
+    },
+    {
+      "epoch": 2.8077793474235713,
+      "grad_norm": 0.26966269521273795,
+      "learning_rate": 3.6956020068765547e-06,
+      "loss": 0.4319,
+      "step": 17058
+    },
+    {
+      "epoch": 2.8079439316426877,
+      "grad_norm": 0.3015423359547079,
+      "learning_rate": 3.6951395185632096e-06,
+      "loss": 0.4293,
+      "step": 17059
+    },
+    {
+      "epoch": 2.808108515861804,
+      "grad_norm": 0.39543701173947554,
+      "learning_rate": 3.6946770390321837e-06,
+      "loss": 0.4475,
+      "step": 17060
+    },
+    {
+      "epoch": 2.8082731000809207,
+      "grad_norm": 0.29558063112252037,
+      "learning_rate": 3.694214568288518e-06,
+      "loss": 0.4326,
+      "step": 17061
+    },
+    {
+      "epoch": 2.808437684300037,
+      "grad_norm": 0.29241953332047743,
+      "learning_rate": 3.693752106337264e-06,
+      "loss": 0.4463,
+      "step": 17062
+    },
+    {
+      "epoch": 2.8086022685191536,
+      "grad_norm": 0.36922966517747435,
+      "learning_rate": 3.6932896531834643e-06,
+      "loss": 0.4424,
+      "step": 17063
+    },
+    {
+      "epoch": 2.80876685273827,
+      "grad_norm": 0.34813899240461527,
+      "learning_rate": 3.692827208832165e-06,
+      "loss": 0.4427,
+      "step": 17064
+    },
+    {
+      "epoch": 2.8089314369573866,
+      "grad_norm": 0.2882741202352136,
+      "learning_rate": 3.6923647732884143e-06,
+      "loss": 0.4429,
+      "step": 17065
+    },
+    {
+      "epoch": 2.8090960211765026,
+      "grad_norm": 0.28096012817176935,
+      "learning_rate": 3.6919023465572558e-06,
+      "loss": 0.4275,
+      "step": 17066
+    },
+    {
+      "epoch": 2.809260605395619,
+      "grad_norm": 0.3340488275443412,
+      "learning_rate": 3.6914399286437373e-06,
+      "loss": 0.4357,
+      "step": 17067
+    },
+    {
+      "epoch": 2.8094251896147355,
+      "grad_norm": 0.3125108279274556,
+      "learning_rate": 3.6909775195529026e-06,
+      "loss": 0.4279,
+      "step": 17068
+    },
+    {
+      "epoch": 2.809589773833852,
+      "grad_norm": 0.2881721468616627,
+      "learning_rate": 3.6905151192898004e-06,
+      "loss": 0.4254,
+      "step": 17069
+    },
+    {
+      "epoch": 2.8097543580529685,
+      "grad_norm": 0.33957698439504513,
+      "learning_rate": 3.690052727859473e-06,
+      "loss": 0.4403,
+      "step": 17070
+    },
+    {
+      "epoch": 2.809918942272085,
+      "grad_norm": 0.3042511204962381,
+      "learning_rate": 3.6895903452669655e-06,
+      "loss": 0.4062,
+      "step": 17071
+    },
+    {
+      "epoch": 2.8100835264912014,
+      "grad_norm": 0.3131632403271567,
+      "learning_rate": 3.689127971517327e-06,
+      "loss": 0.4254,
+      "step": 17072
+    },
+    {
+      "epoch": 2.810248110710318,
+      "grad_norm": 0.32292931932722746,
+      "learning_rate": 3.6886656066155985e-06,
+      "loss": 0.4297,
+      "step": 17073
+    },
+    {
+      "epoch": 2.8104126949294344,
+      "grad_norm": 0.3258889677088053,
+      "learning_rate": 3.6882032505668283e-06,
+      "loss": 0.4245,
+      "step": 17074
+    },
+    {
+      "epoch": 2.810577279148551,
+      "grad_norm": 0.3213719613226214,
+      "learning_rate": 3.687740903376059e-06,
+      "loss": 0.4184,
+      "step": 17075
+    },
+    {
+      "epoch": 2.8107418633676673,
+      "grad_norm": 0.28843989368418793,
+      "learning_rate": 3.687278565048338e-06,
+      "loss": 0.423,
+      "step": 17076
+    },
+    {
+      "epoch": 2.8109064475867838,
+      "grad_norm": 0.382218246035676,
+      "learning_rate": 3.6868162355887096e-06,
+      "loss": 0.4173,
+      "step": 17077
+    },
+    {
+      "epoch": 2.8110710318059002,
+      "grad_norm": 0.3187230594758411,
+      "learning_rate": 3.686353915002216e-06,
+      "loss": 0.4292,
+      "step": 17078
+    },
+    {
+      "epoch": 2.8112356160250167,
+      "grad_norm": 0.28468997295293,
+      "learning_rate": 3.685891603293906e-06,
+      "loss": 0.4286,
+      "step": 17079
+    },
+    {
+      "epoch": 2.811400200244133,
+      "grad_norm": 0.325067665311539,
+      "learning_rate": 3.685429300468819e-06,
+      "loss": 0.4266,
+      "step": 17080
+    },
+    {
+      "epoch": 2.8115647844632496,
+      "grad_norm": 0.5363543466171763,
+      "learning_rate": 3.6849670065320047e-06,
+      "loss": 0.4149,
+      "step": 17081
+    },
+    {
+      "epoch": 2.811729368682366,
+      "grad_norm": 0.36713578219493537,
+      "learning_rate": 3.684504721488504e-06,
+      "loss": 0.4214,
+      "step": 17082
+    },
+    {
+      "epoch": 2.8118939529014826,
+      "grad_norm": 0.3556618817554257,
+      "learning_rate": 3.684042445343363e-06,
+      "loss": 0.4279,
+      "step": 17083
+    },
+    {
+      "epoch": 2.812058537120599,
+      "grad_norm": 0.3790316644542553,
+      "learning_rate": 3.683580178101626e-06,
+      "loss": 0.4204,
+      "step": 17084
+    },
+    {
+      "epoch": 2.8122231213397155,
+      "grad_norm": 0.44054579314572406,
+      "learning_rate": 3.6831179197683345e-06,
+      "loss": 0.4296,
+      "step": 17085
+    },
+    {
+      "epoch": 2.812387705558832,
+      "grad_norm": 0.3110667703692109,
+      "learning_rate": 3.682655670348535e-06,
+      "loss": 0.4293,
+      "step": 17086
+    },
+    {
+      "epoch": 2.8125522897779485,
+      "grad_norm": 0.33293568691450487,
+      "learning_rate": 3.68219342984727e-06,
+      "loss": 0.4437,
+      "step": 17087
+    },
+    {
+      "epoch": 2.812716873997065,
+      "grad_norm": 0.47825615572547775,
+      "learning_rate": 3.6817311982695857e-06,
+      "loss": 0.4353,
+      "step": 17088
+    },
+    {
+      "epoch": 2.8128814582161814,
+      "grad_norm": 0.36050258840546906,
+      "learning_rate": 3.6812689756205207e-06,
+      "loss": 0.4256,
+      "step": 17089
+    },
+    {
+      "epoch": 2.813046042435298,
+      "grad_norm": 0.2679257712308787,
+      "learning_rate": 3.6808067619051253e-06,
+      "loss": 0.4215,
+      "step": 17090
+    },
+    {
+      "epoch": 2.8132106266544143,
+      "grad_norm": 0.3498276050713189,
+      "learning_rate": 3.6803445571284385e-06,
+      "loss": 0.4165,
+      "step": 17091
+    },
+    {
+      "epoch": 2.813375210873531,
+      "grad_norm": 0.29415549838126503,
+      "learning_rate": 3.679882361295504e-06,
+      "loss": 0.425,
+      "step": 17092
+    },
+    {
+      "epoch": 2.8135397950926473,
+      "grad_norm": 0.3274825930326004,
+      "learning_rate": 3.679420174411367e-06,
+      "loss": 0.4117,
+      "step": 17093
+    },
+    {
+      "epoch": 2.8137043793117638,
+      "grad_norm": 0.380254456127662,
+      "learning_rate": 3.678957996481068e-06,
+      "loss": 0.4653,
+      "step": 17094
+    },
+    {
+      "epoch": 2.8138689635308802,
+      "grad_norm": 0.38134174916262903,
+      "learning_rate": 3.6784958275096528e-06,
+      "loss": 0.414,
+      "step": 17095
+    },
+    {
+      "epoch": 2.8140335477499967,
+      "grad_norm": 0.2777538832839576,
+      "learning_rate": 3.6780336675021627e-06,
+      "loss": 0.4113,
+      "step": 17096
+    },
+    {
+      "epoch": 2.814198131969113,
+      "grad_norm": 0.39757878901837895,
+      "learning_rate": 3.6775715164636414e-06,
+      "loss": 0.4386,
+      "step": 17097
+    },
+    {
+      "epoch": 2.8143627161882296,
+      "grad_norm": 0.2996085061697317,
+      "learning_rate": 3.677109374399131e-06,
+      "loss": 0.4326,
+      "step": 17098
+    },
+    {
+      "epoch": 2.814527300407346,
+      "grad_norm": 0.3132830922240271,
+      "learning_rate": 3.676647241313676e-06,
+      "loss": 0.4243,
+      "step": 17099
+    },
+    {
+      "epoch": 2.8146918846264626,
+      "grad_norm": 0.308012152473725,
+      "learning_rate": 3.6761851172123183e-06,
+      "loss": 0.4346,
+      "step": 17100
+    },
+    {
+      "epoch": 2.814856468845579,
+      "grad_norm": 0.28324542469615294,
+      "learning_rate": 3.6757230021000977e-06,
+      "loss": 0.4333,
+      "step": 17101
+    },
+    {
+      "epoch": 2.8150210530646955,
+      "grad_norm": 0.4152341168327462,
+      "learning_rate": 3.6752608959820596e-06,
+      "loss": 0.4325,
+      "step": 17102
+    },
+    {
+      "epoch": 2.815185637283812,
+      "grad_norm": 0.30608123523114217,
+      "learning_rate": 3.674798798863245e-06,
+      "loss": 0.4336,
+      "step": 17103
+    },
+    {
+      "epoch": 2.8153502215029285,
+      "grad_norm": 0.3394719081879419,
+      "learning_rate": 3.674336710748697e-06,
+      "loss": 0.4335,
+      "step": 17104
+    },
+    {
+      "epoch": 2.815514805722045,
+      "grad_norm": 0.3288775694862542,
+      "learning_rate": 3.6738746316434567e-06,
+      "loss": 0.4478,
+      "step": 17105
+    },
+    {
+      "epoch": 2.8156793899411614,
+      "grad_norm": 0.4548371543200483,
+      "learning_rate": 3.673412561552568e-06,
+      "loss": 0.4359,
+      "step": 17106
+    },
+    {
+      "epoch": 2.815843974160278,
+      "grad_norm": 0.3064065707396901,
+      "learning_rate": 3.672950500481072e-06,
+      "loss": 0.4131,
+      "step": 17107
+    },
+    {
+      "epoch": 2.816008558379394,
+      "grad_norm": 0.37563674605906305,
+      "learning_rate": 3.6724884484340074e-06,
+      "loss": 0.4072,
+      "step": 17108
+    },
+    {
+      "epoch": 2.8161731425985104,
+      "grad_norm": 0.31215206312051025,
+      "learning_rate": 3.6720264054164214e-06,
+      "loss": 0.4345,
+      "step": 17109
+    },
+    {
+      "epoch": 2.816337726817627,
+      "grad_norm": 0.29020191266348566,
+      "learning_rate": 3.67156437143335e-06,
+      "loss": 0.4291,
+      "step": 17110
+    },
+    {
+      "epoch": 2.8165023110367433,
+      "grad_norm": 0.3353379042670644,
+      "learning_rate": 3.6711023464898397e-06,
+      "loss": 0.4314,
+      "step": 17111
+    },
+    {
+      "epoch": 2.81666689525586,
+      "grad_norm": 0.31147570248059336,
+      "learning_rate": 3.670640330590929e-06,
+      "loss": 0.4281,
+      "step": 17112
+    },
+    {
+      "epoch": 2.8168314794749763,
+      "grad_norm": 0.3117734222192966,
+      "learning_rate": 3.67017832374166e-06,
+      "loss": 0.4448,
+      "step": 17113
+    },
+    {
+      "epoch": 2.8169960636940927,
+      "grad_norm": 0.4132043734099047,
+      "learning_rate": 3.6697163259470743e-06,
+      "loss": 0.4474,
+      "step": 17114
+    },
+    {
+      "epoch": 2.817160647913209,
+      "grad_norm": 0.4119011796010191,
+      "learning_rate": 3.6692543372122123e-06,
+      "loss": 0.4319,
+      "step": 17115
+    },
+    {
+      "epoch": 2.8173252321323257,
+      "grad_norm": 0.3155796783904074,
+      "learning_rate": 3.668792357542116e-06,
+      "loss": 0.4182,
+      "step": 17116
+    },
+    {
+      "epoch": 2.817489816351442,
+      "grad_norm": 0.30699109023803706,
+      "learning_rate": 3.6683303869418244e-06,
+      "loss": 0.4266,
+      "step": 17117
+    },
+    {
+      "epoch": 2.8176544005705586,
+      "grad_norm": 0.6887270606990578,
+      "learning_rate": 3.667868425416381e-06,
+      "loss": 0.4262,
+      "step": 17118
+    },
+    {
+      "epoch": 2.817818984789675,
+      "grad_norm": 0.2751247994882229,
+      "learning_rate": 3.6674064729708247e-06,
+      "loss": 0.4238,
+      "step": 17119
+    },
+    {
+      "epoch": 2.8179835690087915,
+      "grad_norm": 0.459838495793609,
+      "learning_rate": 3.666944529610198e-06,
+      "loss": 0.4364,
+      "step": 17120
+    },
+    {
+      "epoch": 2.818148153227908,
+      "grad_norm": 0.31715489732174246,
+      "learning_rate": 3.666482595339539e-06,
+      "loss": 0.4115,
+      "step": 17121
+    },
+    {
+      "epoch": 2.8183127374470245,
+      "grad_norm": 0.3810854234866949,
+      "learning_rate": 3.6660206701638885e-06,
+      "loss": 0.435,
+      "step": 17122
+    },
+    {
+      "epoch": 2.818477321666141,
+      "grad_norm": 0.2980712904841457,
+      "learning_rate": 3.665558754088289e-06,
+      "loss": 0.4293,
+      "step": 17123
+    },
+    {
+      "epoch": 2.8186419058852574,
+      "grad_norm": 0.31554529600333936,
+      "learning_rate": 3.6650968471177778e-06,
+      "loss": 0.4396,
+      "step": 17124
+    },
+    {
+      "epoch": 2.818806490104374,
+      "grad_norm": 0.3526745457275579,
+      "learning_rate": 3.664634949257398e-06,
+      "loss": 0.4386,
+      "step": 17125
+    },
+    {
+      "epoch": 2.8189710743234904,
+      "grad_norm": 0.26955385001081505,
+      "learning_rate": 3.6641730605121874e-06,
+      "loss": 0.4217,
+      "step": 17126
+    },
+    {
+      "epoch": 2.819135658542607,
+      "grad_norm": 0.3302344739823321,
+      "learning_rate": 3.663711180887187e-06,
+      "loss": 0.4473,
+      "step": 17127
+    },
+    {
+      "epoch": 2.8193002427617233,
+      "grad_norm": 0.38799946904704896,
+      "learning_rate": 3.663249310387438e-06,
+      "loss": 0.4147,
+      "step": 17128
+    },
+    {
+      "epoch": 2.8194648269808398,
+      "grad_norm": 0.36345893830202847,
+      "learning_rate": 3.6627874490179753e-06,
+      "loss": 0.4215,
+      "step": 17129
+    },
+    {
+      "epoch": 2.8196294111999562,
+      "grad_norm": 0.3921152616720411,
+      "learning_rate": 3.6623255967838453e-06,
+      "loss": 0.4397,
+      "step": 17130
+    },
+    {
+      "epoch": 2.8197939954190727,
+      "grad_norm": 0.3370515743146148,
+      "learning_rate": 3.6618637536900817e-06,
+      "loss": 0.423,
+      "step": 17131
+    },
+    {
+      "epoch": 2.819958579638189,
+      "grad_norm": 0.298672711834632,
+      "learning_rate": 3.6614019197417274e-06,
+      "loss": 0.4148,
+      "step": 17132
+    },
+    {
+      "epoch": 2.820123163857305,
+      "grad_norm": 0.3674852867505034,
+      "learning_rate": 3.6609400949438195e-06,
+      "loss": 0.45,
+      "step": 17133
+    },
+    {
+      "epoch": 2.8202877480764217,
+      "grad_norm": 0.4480024381728986,
+      "learning_rate": 3.6604782793013998e-06,
+      "loss": 0.422,
+      "step": 17134
+    },
+    {
+      "epoch": 2.820452332295538,
+      "grad_norm": 0.43238289475601743,
+      "learning_rate": 3.660016472819506e-06,
+      "loss": 0.4339,
+      "step": 17135
+    },
+    {
+      "epoch": 2.8206169165146546,
+      "grad_norm": 0.26018647672172557,
+      "learning_rate": 3.6595546755031758e-06,
+      "loss": 0.4051,
+      "step": 17136
+    },
+    {
+      "epoch": 2.820781500733771,
+      "grad_norm": 0.27947140164193335,
+      "learning_rate": 3.659092887357451e-06,
+      "loss": 0.4224,
+      "step": 17137
+    },
+    {
+      "epoch": 2.8209460849528876,
+      "grad_norm": 0.2779553942761288,
+      "learning_rate": 3.6586311083873674e-06,
+      "loss": 0.4294,
+      "step": 17138
+    },
+    {
+      "epoch": 2.821110669172004,
+      "grad_norm": 0.3123836058428518,
+      "learning_rate": 3.658169338597968e-06,
+      "loss": 0.3996,
+      "step": 17139
+    },
+    {
+      "epoch": 2.8212752533911205,
+      "grad_norm": 0.29982736494768003,
+      "learning_rate": 3.657707577994286e-06,
+      "loss": 0.435,
+      "step": 17140
+    },
+    {
+      "epoch": 2.821439837610237,
+      "grad_norm": 0.2998633135173355,
+      "learning_rate": 3.6572458265813648e-06,
+      "loss": 0.4326,
+      "step": 17141
+    },
+    {
+      "epoch": 2.8216044218293534,
+      "grad_norm": 1.2875260936074302,
+      "learning_rate": 3.6567840843642383e-06,
+      "loss": 0.4414,
+      "step": 17142
+    },
+    {
+      "epoch": 2.82176900604847,
+      "grad_norm": 0.36380125041897954,
+      "learning_rate": 3.6563223513479505e-06,
+      "loss": 0.4419,
+      "step": 17143
+    },
+    {
+      "epoch": 2.8219335902675864,
+      "grad_norm": 0.3031862883250551,
+      "learning_rate": 3.6558606275375356e-06,
+      "loss": 0.4327,
+      "step": 17144
+    },
+    {
+      "epoch": 2.822098174486703,
+      "grad_norm": 0.31900799897958043,
+      "learning_rate": 3.655398912938031e-06,
+      "loss": 0.4384,
+      "step": 17145
+    },
+    {
+      "epoch": 2.8222627587058193,
+      "grad_norm": 0.3025173566880341,
+      "learning_rate": 3.654937207554478e-06,
+      "loss": 0.4343,
+      "step": 17146
+    },
+    {
+      "epoch": 2.822427342924936,
+      "grad_norm": 0.2663358722032339,
+      "learning_rate": 3.6544755113919113e-06,
+      "loss": 0.4463,
+      "step": 17147
+    },
+    {
+      "epoch": 2.8225919271440523,
+      "grad_norm": 0.3513917982076639,
+      "learning_rate": 3.654013824455372e-06,
+      "loss": 0.4307,
+      "step": 17148
+    },
+    {
+      "epoch": 2.8227565113631687,
+      "grad_norm": 0.5643742825543426,
+      "learning_rate": 3.653552146749895e-06,
+      "loss": 0.433,
+      "step": 17149
+    },
+    {
+      "epoch": 2.822921095582285,
+      "grad_norm": 0.3137193911706826,
+      "learning_rate": 3.65309047828052e-06,
+      "loss": 0.4303,
+      "step": 17150
+    },
+    {
+      "epoch": 2.8230856798014017,
+      "grad_norm": 0.846721645930326,
+      "learning_rate": 3.652628819052285e-06,
+      "loss": 0.4398,
+      "step": 17151
+    },
+    {
+      "epoch": 2.823250264020518,
+      "grad_norm": 0.35462356857183863,
+      "learning_rate": 3.652167169070224e-06,
+      "loss": 0.4333,
+      "step": 17152
+    },
+    {
+      "epoch": 2.8234148482396346,
+      "grad_norm": 0.2740207894520864,
+      "learning_rate": 3.6517055283393776e-06,
+      "loss": 0.4125,
+      "step": 17153
+    },
+    {
+      "epoch": 2.823579432458751,
+      "grad_norm": 0.3036748323797673,
+      "learning_rate": 3.6512438968647807e-06,
+      "loss": 0.4514,
+      "step": 17154
+    },
+    {
+      "epoch": 2.8237440166778676,
+      "grad_norm": 0.29306246669216246,
+      "learning_rate": 3.6507822746514722e-06,
+      "loss": 0.4343,
+      "step": 17155
+    },
+    {
+      "epoch": 2.823908600896984,
+      "grad_norm": 0.2752629222420503,
+      "learning_rate": 3.650320661704488e-06,
+      "loss": 0.4429,
+      "step": 17156
+    },
+    {
+      "epoch": 2.8240731851161005,
+      "grad_norm": 0.30854742164160615,
+      "learning_rate": 3.6498590580288672e-06,
+      "loss": 0.4468,
+      "step": 17157
+    },
+    {
+      "epoch": 2.824237769335217,
+      "grad_norm": 0.31442516825330213,
+      "learning_rate": 3.649397463629646e-06,
+      "loss": 0.4469,
+      "step": 17158
+    },
+    {
+      "epoch": 2.8244023535543334,
+      "grad_norm": 0.31816277090241346,
+      "learning_rate": 3.6489358785118563e-06,
+      "loss": 0.4317,
+      "step": 17159
+    },
+    {
+      "epoch": 2.82456693777345,
+      "grad_norm": 0.3820945237326531,
+      "learning_rate": 3.6484743026805423e-06,
+      "loss": 0.4237,
+      "step": 17160
+    },
+    {
+      "epoch": 2.8247315219925664,
+      "grad_norm": 0.3340528223521202,
+      "learning_rate": 3.648012736140734e-06,
+      "loss": 0.429,
+      "step": 17161
+    },
+    {
+      "epoch": 2.824896106211683,
+      "grad_norm": 0.2914111218959242,
+      "learning_rate": 3.6475511788974735e-06,
+      "loss": 0.4571,
+      "step": 17162
+    },
+    {
+      "epoch": 2.8250606904307993,
+      "grad_norm": 0.31147738717265716,
+      "learning_rate": 3.6470896309557925e-06,
+      "loss": 0.4219,
+      "step": 17163
+    },
+    {
+      "epoch": 2.825225274649916,
+      "grad_norm": 0.31456881040370577,
+      "learning_rate": 3.6466280923207297e-06,
+      "loss": 0.4485,
+      "step": 17164
+    },
+    {
+      "epoch": 2.8253898588690323,
+      "grad_norm": 0.2519660907558552,
+      "learning_rate": 3.6461665629973207e-06,
+      "loss": 0.43,
+      "step": 17165
+    },
+    {
+      "epoch": 2.8255544430881487,
+      "grad_norm": 0.32504354549709374,
+      "learning_rate": 3.6457050429906e-06,
+      "loss": 0.4183,
+      "step": 17166
+    },
+    {
+      "epoch": 2.825719027307265,
+      "grad_norm": 0.3063441324912973,
+      "learning_rate": 3.6452435323056063e-06,
+      "loss": 0.4233,
+      "step": 17167
+    },
+    {
+      "epoch": 2.8258836115263817,
+      "grad_norm": 0.2825098525755312,
+      "learning_rate": 3.644782030947373e-06,
+      "loss": 0.4453,
+      "step": 17168
+    },
+    {
+      "epoch": 2.826048195745498,
+      "grad_norm": 0.3196720940852849,
+      "learning_rate": 3.6443205389209384e-06,
+      "loss": 0.4341,
+      "step": 17169
+    },
+    {
+      "epoch": 2.8262127799646146,
+      "grad_norm": 0.30787067603334795,
+      "learning_rate": 3.6438590562313345e-06,
+      "loss": 0.4246,
+      "step": 17170
+    },
+    {
+      "epoch": 2.826377364183731,
+      "grad_norm": 0.3398306678988993,
+      "learning_rate": 3.643397582883602e-06,
+      "loss": 0.428,
+      "step": 17171
+    },
+    {
+      "epoch": 2.8265419484028476,
+      "grad_norm": 0.6485214365786198,
+      "learning_rate": 3.642936118882771e-06,
+      "loss": 0.424,
+      "step": 17172
+    },
+    {
+      "epoch": 2.826706532621964,
+      "grad_norm": 0.287575292203491,
+      "learning_rate": 3.642474664233879e-06,
+      "loss": 0.4514,
+      "step": 17173
+    },
+    {
+      "epoch": 2.82687111684108,
+      "grad_norm": 0.3992671162860481,
+      "learning_rate": 3.6420132189419624e-06,
+      "loss": 0.4402,
+      "step": 17174
+    },
+    {
+      "epoch": 2.8270357010601965,
+      "grad_norm": 0.29128519496400274,
+      "learning_rate": 3.641551783012054e-06,
+      "loss": 0.4127,
+      "step": 17175
+    },
+    {
+      "epoch": 2.827200285279313,
+      "grad_norm": 0.33820307623461265,
+      "learning_rate": 3.64109035644919e-06,
+      "loss": 0.4298,
+      "step": 17176
+    },
+    {
+      "epoch": 2.8273648694984295,
+      "grad_norm": 0.5258143016939788,
+      "learning_rate": 3.640628939258405e-06,
+      "loss": 0.4313,
+      "step": 17177
+    },
+    {
+      "epoch": 2.827529453717546,
+      "grad_norm": 0.33421939150531105,
+      "learning_rate": 3.640167531444735e-06,
+      "loss": 0.4433,
+      "step": 17178
+    },
+    {
+      "epoch": 2.8276940379366624,
+      "grad_norm": 0.3489542375657524,
+      "learning_rate": 3.639706133013212e-06,
+      "loss": 0.421,
+      "step": 17179
+    },
+    {
+      "epoch": 2.827858622155779,
+      "grad_norm": 0.3365482386357486,
+      "learning_rate": 3.6392447439688744e-06,
+      "loss": 0.4159,
+      "step": 17180
+    },
+    {
+      "epoch": 2.8280232063748953,
+      "grad_norm": 0.33791730236571293,
+      "learning_rate": 3.638783364316755e-06,
+      "loss": 0.4452,
+      "step": 17181
+    },
+    {
+      "epoch": 2.828187790594012,
+      "grad_norm": 0.3057041855992184,
+      "learning_rate": 3.6383219940618863e-06,
+      "loss": 0.4048,
+      "step": 17182
+    },
+    {
+      "epoch": 2.8283523748131283,
+      "grad_norm": 0.33764982152904666,
+      "learning_rate": 3.6378606332093046e-06,
+      "loss": 0.4338,
+      "step": 17183
+    },
+    {
+      "epoch": 2.8285169590322448,
+      "grad_norm": 0.3030336775119331,
+      "learning_rate": 3.6373992817640427e-06,
+      "loss": 0.4195,
+      "step": 17184
+    },
+    {
+      "epoch": 2.8286815432513612,
+      "grad_norm": 0.3335775815830645,
+      "learning_rate": 3.636937939731137e-06,
+      "loss": 0.4098,
+      "step": 17185
+    },
+    {
+      "epoch": 2.8288461274704777,
+      "grad_norm": 0.2719457502171315,
+      "learning_rate": 3.6364766071156185e-06,
+      "loss": 0.4247,
+      "step": 17186
+    },
+    {
+      "epoch": 2.829010711689594,
+      "grad_norm": 0.2997921615260896,
+      "learning_rate": 3.636015283922524e-06,
+      "loss": 0.444,
+      "step": 17187
+    },
+    {
+      "epoch": 2.8291752959087106,
+      "grad_norm": 0.3194612110051689,
+      "learning_rate": 3.635553970156886e-06,
+      "loss": 0.445,
+      "step": 17188
+    },
+    {
+      "epoch": 2.829339880127827,
+      "grad_norm": 0.313158008599718,
+      "learning_rate": 3.6350926658237367e-06,
+      "loss": 0.4327,
+      "step": 17189
+    },
+    {
+      "epoch": 2.8295044643469436,
+      "grad_norm": 0.28656355456574967,
+      "learning_rate": 3.634631370928113e-06,
+      "loss": 0.4306,
+      "step": 17190
+    },
+    {
+      "epoch": 2.82966904856606,
+      "grad_norm": 0.49093999604571625,
+      "learning_rate": 3.6341700854750433e-06,
+      "loss": 0.4238,
+      "step": 17191
+    },
+    {
+      "epoch": 2.8298336327851765,
+      "grad_norm": 0.312583077459305,
+      "learning_rate": 3.6337088094695677e-06,
+      "loss": 0.4267,
+      "step": 17192
+    },
+    {
+      "epoch": 2.829998217004293,
+      "grad_norm": 0.433786744476954,
+      "learning_rate": 3.6332475429167125e-06,
+      "loss": 0.4285,
+      "step": 17193
+    },
+    {
+      "epoch": 2.8301628012234095,
+      "grad_norm": 0.3632006308997577,
+      "learning_rate": 3.632786285821517e-06,
+      "loss": 0.418,
+      "step": 17194
+    },
+    {
+      "epoch": 2.830327385442526,
+      "grad_norm": 0.30953167349296806,
+      "learning_rate": 3.6323250381890107e-06,
+      "loss": 0.4368,
+      "step": 17195
+    },
+    {
+      "epoch": 2.8304919696616424,
+      "grad_norm": 0.31743954151520987,
+      "learning_rate": 3.6318638000242263e-06,
+      "loss": 0.4372,
+      "step": 17196
+    },
+    {
+      "epoch": 2.830656553880759,
+      "grad_norm": 0.30182454303424505,
+      "learning_rate": 3.6314025713321985e-06,
+      "loss": 0.4286,
+      "step": 17197
+    },
+    {
+      "epoch": 2.8308211380998753,
+      "grad_norm": 0.38935321783769483,
+      "learning_rate": 3.630941352117958e-06,
+      "loss": 0.4297,
+      "step": 17198
+    },
+    {
+      "epoch": 2.8309857223189914,
+      "grad_norm": 0.30084220698938574,
+      "learning_rate": 3.63048014238654e-06,
+      "loss": 0.4244,
+      "step": 17199
+    },
+    {
+      "epoch": 2.831150306538108,
+      "grad_norm": 0.3576438574713621,
+      "learning_rate": 3.6300189421429752e-06,
+      "loss": 0.414,
+      "step": 17200
+    },
+    {
+      "epoch": 2.8313148907572243,
+      "grad_norm": 0.8043356118524135,
+      "learning_rate": 3.6295577513922975e-06,
+      "loss": 0.4545,
+      "step": 17201
+    },
+    {
+      "epoch": 2.831479474976341,
+      "grad_norm": 0.4242885589241204,
+      "learning_rate": 3.6290965701395374e-06,
+      "loss": 0.4416,
+      "step": 17202
+    },
+    {
+      "epoch": 2.8316440591954573,
+      "grad_norm": 0.2974338529965238,
+      "learning_rate": 3.6286353983897272e-06,
+      "loss": 0.4296,
+      "step": 17203
+    },
+    {
+      "epoch": 2.8318086434145737,
+      "grad_norm": 0.3484590649079067,
+      "learning_rate": 3.6281742361479004e-06,
+      "loss": 0.429,
+      "step": 17204
+    },
+    {
+      "epoch": 2.83197322763369,
+      "grad_norm": 0.3206797393939164,
+      "learning_rate": 3.627713083419088e-06,
+      "loss": 0.4444,
+      "step": 17205
+    },
+    {
+      "epoch": 2.8321378118528067,
+      "grad_norm": 0.32626885839653796,
+      "learning_rate": 3.627251940208323e-06,
+      "loss": 0.4112,
+      "step": 17206
+    },
+    {
+      "epoch": 2.832302396071923,
+      "grad_norm": 0.31874147694777155,
+      "learning_rate": 3.626790806520635e-06,
+      "loss": 0.4244,
+      "step": 17207
+    },
+    {
+      "epoch": 2.8324669802910396,
+      "grad_norm": 0.35748771708537413,
+      "learning_rate": 3.6263296823610594e-06,
+      "loss": 0.4601,
+      "step": 17208
+    },
+    {
+      "epoch": 2.832631564510156,
+      "grad_norm": 0.2769410845748084,
+      "learning_rate": 3.6258685677346256e-06,
+      "loss": 0.4284,
+      "step": 17209
+    },
+    {
+      "epoch": 2.8327961487292725,
+      "grad_norm": 0.34461088576620236,
+      "learning_rate": 3.6254074626463626e-06,
+      "loss": 0.4311,
+      "step": 17210
+    },
+    {
+      "epoch": 2.832960732948389,
+      "grad_norm": 0.286261312823275,
+      "learning_rate": 3.624946367101307e-06,
+      "loss": 0.4372,
+      "step": 17211
+    },
+    {
+      "epoch": 2.8331253171675055,
+      "grad_norm": 0.3474396602887607,
+      "learning_rate": 3.624485281104485e-06,
+      "loss": 0.4219,
+      "step": 17212
+    },
+    {
+      "epoch": 2.833289901386622,
+      "grad_norm": 0.2918359905236362,
+      "learning_rate": 3.6240242046609327e-06,
+      "loss": 0.4382,
+      "step": 17213
+    },
+    {
+      "epoch": 2.8334544856057384,
+      "grad_norm": 0.29510176438923397,
+      "learning_rate": 3.6235631377756767e-06,
+      "loss": 0.4456,
+      "step": 17214
+    },
+    {
+      "epoch": 2.833619069824855,
+      "grad_norm": 0.3666142725303603,
+      "learning_rate": 3.623102080453751e-06,
+      "loss": 0.4147,
+      "step": 17215
+    },
+    {
+      "epoch": 2.8337836540439714,
+      "grad_norm": 0.3449592834790887,
+      "learning_rate": 3.6226410327001854e-06,
+      "loss": 0.4254,
+      "step": 17216
+    },
+    {
+      "epoch": 2.833948238263088,
+      "grad_norm": 0.474561524529995,
+      "learning_rate": 3.62217999452001e-06,
+      "loss": 0.4259,
+      "step": 17217
+    },
+    {
+      "epoch": 2.8341128224822043,
+      "grad_norm": 0.3081507109268821,
+      "learning_rate": 3.6217189659182566e-06,
+      "loss": 0.4151,
+      "step": 17218
+    },
+    {
+      "epoch": 2.834277406701321,
+      "grad_norm": 0.3440200959668479,
+      "learning_rate": 3.621257946899955e-06,
+      "loss": 0.4122,
+      "step": 17219
+    },
+    {
+      "epoch": 2.8344419909204372,
+      "grad_norm": 0.292391216530612,
+      "learning_rate": 3.6207969374701375e-06,
+      "loss": 0.448,
+      "step": 17220
+    },
+    {
+      "epoch": 2.8346065751395537,
+      "grad_norm": 0.26545459000328425,
+      "learning_rate": 3.6203359376338298e-06,
+      "loss": 0.4372,
+      "step": 17221
+    },
+    {
+      "epoch": 2.83477115935867,
+      "grad_norm": 0.33534217289780477,
+      "learning_rate": 3.6198749473960685e-06,
+      "loss": 0.4374,
+      "step": 17222
+    },
+    {
+      "epoch": 2.8349357435777867,
+      "grad_norm": 0.35976061171612134,
+      "learning_rate": 3.6194139667618774e-06,
+      "loss": 0.4318,
+      "step": 17223
+    },
+    {
+      "epoch": 2.835100327796903,
+      "grad_norm": 0.5249155215893581,
+      "learning_rate": 3.618952995736293e-06,
+      "loss": 0.4324,
+      "step": 17224
+    },
+    {
+      "epoch": 2.8352649120160196,
+      "grad_norm": 0.3894814558327808,
+      "learning_rate": 3.6184920343243406e-06,
+      "loss": 0.4212,
+      "step": 17225
+    },
+    {
+      "epoch": 2.835429496235136,
+      "grad_norm": 0.2899878228648192,
+      "learning_rate": 3.6180310825310505e-06,
+      "loss": 0.4332,
+      "step": 17226
+    },
+    {
+      "epoch": 2.8355940804542525,
+      "grad_norm": 0.2955486946134394,
+      "learning_rate": 3.617570140361454e-06,
+      "loss": 0.4232,
+      "step": 17227
+    },
+    {
+      "epoch": 2.835758664673369,
+      "grad_norm": 0.4856743837521025,
+      "learning_rate": 3.6171092078205794e-06,
+      "loss": 0.4292,
+      "step": 17228
+    },
+    {
+      "epoch": 2.8359232488924855,
+      "grad_norm": 0.2852401565044446,
+      "learning_rate": 3.6166482849134573e-06,
+      "loss": 0.4355,
+      "step": 17229
+    },
+    {
+      "epoch": 2.836087833111602,
+      "grad_norm": 0.4321899265913684,
+      "learning_rate": 3.6161873716451158e-06,
+      "loss": 0.4199,
+      "step": 17230
+    },
+    {
+      "epoch": 2.8362524173307184,
+      "grad_norm": 0.33996527413233607,
+      "learning_rate": 3.6157264680205853e-06,
+      "loss": 0.4311,
+      "step": 17231
+    },
+    {
+      "epoch": 2.836417001549835,
+      "grad_norm": 0.3379685630381169,
+      "learning_rate": 3.6152655740448963e-06,
+      "loss": 0.437,
+      "step": 17232
+    },
+    {
+      "epoch": 2.8365815857689514,
+      "grad_norm": 0.36981142619782814,
+      "learning_rate": 3.6148046897230744e-06,
+      "loss": 0.4246,
+      "step": 17233
+    },
+    {
+      "epoch": 2.836746169988068,
+      "grad_norm": 0.3729699046682807,
+      "learning_rate": 3.6143438150601513e-06,
+      "loss": 0.4486,
+      "step": 17234
+    },
+    {
+      "epoch": 2.8369107542071843,
+      "grad_norm": 0.4633154252788908,
+      "learning_rate": 3.6138829500611537e-06,
+      "loss": 0.4265,
+      "step": 17235
+    },
+    {
+      "epoch": 2.8370753384263008,
+      "grad_norm": 0.44586719350424964,
+      "learning_rate": 3.6134220947311134e-06,
+      "loss": 0.4336,
+      "step": 17236
+    },
+    {
+      "epoch": 2.8372399226454172,
+      "grad_norm": 0.2940424118350281,
+      "learning_rate": 3.6129612490750555e-06,
+      "loss": 0.4263,
+      "step": 17237
+    },
+    {
+      "epoch": 2.8374045068645337,
+      "grad_norm": 0.25917279093973444,
+      "learning_rate": 3.612500413098012e-06,
+      "loss": 0.4268,
+      "step": 17238
+    },
+    {
+      "epoch": 2.83756909108365,
+      "grad_norm": 0.31873135212679354,
+      "learning_rate": 3.612039586805011e-06,
+      "loss": 0.4382,
+      "step": 17239
+    },
+    {
+      "epoch": 2.8377336753027667,
+      "grad_norm": 0.30085694031050964,
+      "learning_rate": 3.611578770201076e-06,
+      "loss": 0.4152,
+      "step": 17240
+    },
+    {
+      "epoch": 2.8378982595218827,
+      "grad_norm": 0.3951723660102914,
+      "learning_rate": 3.611117963291242e-06,
+      "loss": 0.4378,
+      "step": 17241
+    },
+    {
+      "epoch": 2.838062843740999,
+      "grad_norm": 0.36034714559106945,
+      "learning_rate": 3.6106571660805315e-06,
+      "loss": 0.4363,
+      "step": 17242
+    },
+    {
+      "epoch": 2.8382274279601156,
+      "grad_norm": 0.34931684788064654,
+      "learning_rate": 3.6101963785739774e-06,
+      "loss": 0.417,
+      "step": 17243
+    },
+    {
+      "epoch": 2.838392012179232,
+      "grad_norm": 0.36458261655129637,
+      "learning_rate": 3.609735600776604e-06,
+      "loss": 0.4313,
+      "step": 17244
+    },
+    {
+      "epoch": 2.8385565963983486,
+      "grad_norm": 0.30291655119692484,
+      "learning_rate": 3.609274832693441e-06,
+      "loss": 0.432,
+      "step": 17245
+    },
+    {
+      "epoch": 2.838721180617465,
+      "grad_norm": 0.25942457772664856,
+      "learning_rate": 3.608814074329516e-06,
+      "loss": 0.4488,
+      "step": 17246
+    },
+    {
+      "epoch": 2.8388857648365815,
+      "grad_norm": 0.3107357539755343,
+      "learning_rate": 3.6083533256898546e-06,
+      "loss": 0.4363,
+      "step": 17247
+    },
+    {
+      "epoch": 2.839050349055698,
+      "grad_norm": 0.38995430767614275,
+      "learning_rate": 3.607892586779487e-06,
+      "loss": 0.4233,
+      "step": 17248
+    },
+    {
+      "epoch": 2.8392149332748144,
+      "grad_norm": 0.35307916284665136,
+      "learning_rate": 3.6074318576034388e-06,
+      "loss": 0.4271,
+      "step": 17249
+    },
+    {
+      "epoch": 2.839379517493931,
+      "grad_norm": 0.3120446790305785,
+      "learning_rate": 3.6069711381667385e-06,
+      "loss": 0.4256,
+      "step": 17250
+    },
+    {
+      "epoch": 2.8395441017130474,
+      "grad_norm": 0.4791097839923694,
+      "learning_rate": 3.606510428474412e-06,
+      "loss": 0.4365,
+      "step": 17251
+    },
+    {
+      "epoch": 2.839708685932164,
+      "grad_norm": 0.3135030321580087,
+      "learning_rate": 3.606049728531489e-06,
+      "loss": 0.4154,
+      "step": 17252
+    },
+    {
+      "epoch": 2.8398732701512803,
+      "grad_norm": 0.3228397935016283,
+      "learning_rate": 3.6055890383429934e-06,
+      "loss": 0.4498,
+      "step": 17253
+    },
+    {
+      "epoch": 2.840037854370397,
+      "grad_norm": 0.5532350580857291,
+      "learning_rate": 3.6051283579139522e-06,
+      "loss": 0.4389,
+      "step": 17254
+    },
+    {
+      "epoch": 2.8402024385895133,
+      "grad_norm": 0.43181483714512964,
+      "learning_rate": 3.604667687249395e-06,
+      "loss": 0.4346,
+      "step": 17255
+    },
+    {
+      "epoch": 2.8403670228086297,
+      "grad_norm": 0.33784989362316176,
+      "learning_rate": 3.604207026354345e-06,
+      "loss": 0.4429,
+      "step": 17256
+    },
+    {
+      "epoch": 2.840531607027746,
+      "grad_norm": 0.31093719233205325,
+      "learning_rate": 3.603746375233831e-06,
+      "loss": 0.4227,
+      "step": 17257
+    },
+    {
+      "epoch": 2.8406961912468627,
+      "grad_norm": 0.25628383000837457,
+      "learning_rate": 3.6032857338928787e-06,
+      "loss": 0.4347,
+      "step": 17258
+    },
+    {
+      "epoch": 2.840860775465979,
+      "grad_norm": 0.3291742395980252,
+      "learning_rate": 3.602825102336515e-06,
+      "loss": 0.4139,
+      "step": 17259
+    },
+    {
+      "epoch": 2.8410253596850956,
+      "grad_norm": 0.286703019845901,
+      "learning_rate": 3.602364480569767e-06,
+      "loss": 0.4358,
+      "step": 17260
+    },
+    {
+      "epoch": 2.841189943904212,
+      "grad_norm": 0.3262119607648993,
+      "learning_rate": 3.601903868597657e-06,
+      "loss": 0.4492,
+      "step": 17261
+    },
+    {
+      "epoch": 2.8413545281233286,
+      "grad_norm": 0.2642906042273244,
+      "learning_rate": 3.601443266425216e-06,
+      "loss": 0.4361,
+      "step": 17262
+    },
+    {
+      "epoch": 2.841519112342445,
+      "grad_norm": 0.30197138080336206,
+      "learning_rate": 3.6009826740574656e-06,
+      "loss": 0.4273,
+      "step": 17263
+    },
+    {
+      "epoch": 2.8416836965615615,
+      "grad_norm": 0.3278386663356908,
+      "learning_rate": 3.600522091499434e-06,
+      "loss": 0.4491,
+      "step": 17264
+    },
+    {
+      "epoch": 2.841848280780678,
+      "grad_norm": 0.366249872998856,
+      "learning_rate": 3.600061518756146e-06,
+      "loss": 0.4365,
+      "step": 17265
+    },
+    {
+      "epoch": 2.842012864999794,
+      "grad_norm": 0.32644035623795087,
+      "learning_rate": 3.5996009558326287e-06,
+      "loss": 0.4083,
+      "step": 17266
+    },
+    {
+      "epoch": 2.8421774492189105,
+      "grad_norm": 0.33766069885696454,
+      "learning_rate": 3.5991404027339047e-06,
+      "loss": 0.4114,
+      "step": 17267
+    },
+    {
+      "epoch": 2.842342033438027,
+      "grad_norm": 0.34999581741863056,
+      "learning_rate": 3.5986798594650023e-06,
+      "loss": 0.4398,
+      "step": 17268
+    },
+    {
+      "epoch": 2.8425066176571434,
+      "grad_norm": 0.29690638842098677,
+      "learning_rate": 3.598219326030946e-06,
+      "loss": 0.4379,
+      "step": 17269
+    },
+    {
+      "epoch": 2.84267120187626,
+      "grad_norm": 0.5821848353060289,
+      "learning_rate": 3.597758802436759e-06,
+      "loss": 0.4272,
+      "step": 17270
+    },
+    {
+      "epoch": 2.8428357860953763,
+      "grad_norm": 0.27318836340541974,
+      "learning_rate": 3.5972982886874706e-06,
+      "loss": 0.4364,
+      "step": 17271
+    },
+    {
+      "epoch": 2.843000370314493,
+      "grad_norm": 0.2907001035804837,
+      "learning_rate": 3.5968377847880992e-06,
+      "loss": 0.4385,
+      "step": 17272
+    },
+    {
+      "epoch": 2.8431649545336093,
+      "grad_norm": 0.35475419183949747,
+      "learning_rate": 3.5963772907436767e-06,
+      "loss": 0.4404,
+      "step": 17273
+    },
+    {
+      "epoch": 2.8433295387527258,
+      "grad_norm": 0.2929670971585379,
+      "learning_rate": 3.5959168065592217e-06,
+      "loss": 0.4269,
+      "step": 17274
+    },
+    {
+      "epoch": 2.8434941229718422,
+      "grad_norm": 0.31006705955902614,
+      "learning_rate": 3.5954563322397647e-06,
+      "loss": 0.419,
+      "step": 17275
+    },
+    {
+      "epoch": 2.8436587071909587,
+      "grad_norm": 0.4804142954059058,
+      "learning_rate": 3.594995867790326e-06,
+      "loss": 0.431,
+      "step": 17276
+    },
+    {
+      "epoch": 2.843823291410075,
+      "grad_norm": 0.38870693622529257,
+      "learning_rate": 3.59453541321593e-06,
+      "loss": 0.4359,
+      "step": 17277
+    },
+    {
+      "epoch": 2.8439878756291916,
+      "grad_norm": 0.3286508395949563,
+      "learning_rate": 3.5940749685216036e-06,
+      "loss": 0.4262,
+      "step": 17278
+    },
+    {
+      "epoch": 2.844152459848308,
+      "grad_norm": 0.3152080061760849,
+      "learning_rate": 3.5936145337123686e-06,
+      "loss": 0.4265,
+      "step": 17279
+    },
+    {
+      "epoch": 2.8443170440674246,
+      "grad_norm": 0.3129194774671554,
+      "learning_rate": 3.593154108793251e-06,
+      "loss": 0.4297,
+      "step": 17280
+    },
+    {
+      "epoch": 2.844481628286541,
+      "grad_norm": 0.38874431455021785,
+      "learning_rate": 3.592693693769273e-06,
+      "loss": 0.4383,
+      "step": 17281
+    },
+    {
+      "epoch": 2.8446462125056575,
+      "grad_norm": 0.37803432983638346,
+      "learning_rate": 3.592233288645461e-06,
+      "loss": 0.4305,
+      "step": 17282
+    },
+    {
+      "epoch": 2.844810796724774,
+      "grad_norm": 0.45048005972917665,
+      "learning_rate": 3.591772893426836e-06,
+      "loss": 0.426,
+      "step": 17283
+    },
+    {
+      "epoch": 2.8449753809438905,
+      "grad_norm": 0.28005271787070796,
+      "learning_rate": 3.591312508118422e-06,
+      "loss": 0.4315,
+      "step": 17284
+    },
+    {
+      "epoch": 2.845139965163007,
+      "grad_norm": 0.33550972302654886,
+      "learning_rate": 3.590852132725244e-06,
+      "loss": 0.4234,
+      "step": 17285
+    },
+    {
+      "epoch": 2.8453045493821234,
+      "grad_norm": 0.32102823504171063,
+      "learning_rate": 3.5903917672523238e-06,
+      "loss": 0.442,
+      "step": 17286
+    },
+    {
+      "epoch": 2.84546913360124,
+      "grad_norm": 0.35483966039125975,
+      "learning_rate": 3.5899314117046864e-06,
+      "loss": 0.4241,
+      "step": 17287
+    },
+    {
+      "epoch": 2.8456337178203563,
+      "grad_norm": 0.31500137223962354,
+      "learning_rate": 3.5894710660873536e-06,
+      "loss": 0.4432,
+      "step": 17288
+    },
+    {
+      "epoch": 2.845798302039473,
+      "grad_norm": 0.29601726038616466,
+      "learning_rate": 3.5890107304053498e-06,
+      "loss": 0.4144,
+      "step": 17289
+    },
+    {
+      "epoch": 2.8459628862585893,
+      "grad_norm": 0.2838766819295395,
+      "learning_rate": 3.588550404663699e-06,
+      "loss": 0.4503,
+      "step": 17290
+    },
+    {
+      "epoch": 2.8461274704777058,
+      "grad_norm": 0.32551607430476087,
+      "learning_rate": 3.5880900888674185e-06,
+      "loss": 0.4515,
+      "step": 17291
+    },
+    {
+      "epoch": 2.8462920546968222,
+      "grad_norm": 0.303023119862285,
+      "learning_rate": 3.5876297830215387e-06,
+      "loss": 0.4282,
+      "step": 17292
+    },
+    {
+      "epoch": 2.8464566389159387,
+      "grad_norm": 0.3677300499257935,
+      "learning_rate": 3.5871694871310755e-06,
+      "loss": 0.4279,
+      "step": 17293
+    },
+    {
+      "epoch": 2.846621223135055,
+      "grad_norm": 0.2903798105398287,
+      "learning_rate": 3.586709201201057e-06,
+      "loss": 0.426,
+      "step": 17294
+    },
+    {
+      "epoch": 2.8467858073541716,
+      "grad_norm": 0.2793437932527116,
+      "learning_rate": 3.5862489252365017e-06,
+      "loss": 0.4373,
+      "step": 17295
+    },
+    {
+      "epoch": 2.846950391573288,
+      "grad_norm": 0.3094071114496996,
+      "learning_rate": 3.585788659242434e-06,
+      "loss": 0.4581,
+      "step": 17296
+    },
+    {
+      "epoch": 2.8471149757924046,
+      "grad_norm": 0.3039670945668618,
+      "learning_rate": 3.5853284032238763e-06,
+      "loss": 0.4145,
+      "step": 17297
+    },
+    {
+      "epoch": 2.847279560011521,
+      "grad_norm": 0.6203273417092992,
+      "learning_rate": 3.584868157185849e-06,
+      "loss": 0.4232,
+      "step": 17298
+    },
+    {
+      "epoch": 2.8474441442306375,
+      "grad_norm": 0.43621089193038654,
+      "learning_rate": 3.5844079211333753e-06,
+      "loss": 0.4257,
+      "step": 17299
+    },
+    {
+      "epoch": 2.847608728449754,
+      "grad_norm": 0.3902462502959947,
+      "learning_rate": 3.5839476950714765e-06,
+      "loss": 0.4344,
+      "step": 17300
+    },
+    {
+      "epoch": 2.8477733126688705,
+      "grad_norm": 0.4501508527387159,
+      "learning_rate": 3.5834874790051763e-06,
+      "loss": 0.4309,
+      "step": 17301
+    },
+    {
+      "epoch": 2.847937896887987,
+      "grad_norm": 0.37557686051108913,
+      "learning_rate": 3.583027272939493e-06,
+      "loss": 0.4415,
+      "step": 17302
+    },
+    {
+      "epoch": 2.8481024811071034,
+      "grad_norm": 0.3013870558351485,
+      "learning_rate": 3.5825670768794525e-06,
+      "loss": 0.4305,
+      "step": 17303
+    },
+    {
+      "epoch": 2.84826706532622,
+      "grad_norm": 0.3542117988239237,
+      "learning_rate": 3.5821068908300713e-06,
+      "loss": 0.4372,
+      "step": 17304
+    },
+    {
+      "epoch": 2.8484316495453363,
+      "grad_norm": 0.31145913931400526,
+      "learning_rate": 3.5816467147963757e-06,
+      "loss": 0.4244,
+      "step": 17305
+    },
+    {
+      "epoch": 2.848596233764453,
+      "grad_norm": 0.3143075586074146,
+      "learning_rate": 3.5811865487833844e-06,
+      "loss": 0.4287,
+      "step": 17306
+    },
+    {
+      "epoch": 2.8487608179835693,
+      "grad_norm": 0.38184845853847327,
+      "learning_rate": 3.5807263927961173e-06,
+      "loss": 0.4578,
+      "step": 17307
+    },
+    {
+      "epoch": 2.8489254022026853,
+      "grad_norm": 0.32585932374428855,
+      "learning_rate": 3.580266246839598e-06,
+      "loss": 0.4245,
+      "step": 17308
+    },
+    {
+      "epoch": 2.849089986421802,
+      "grad_norm": 0.3651766852599073,
+      "learning_rate": 3.579806110918845e-06,
+      "loss": 0.4211,
+      "step": 17309
+    },
+    {
+      "epoch": 2.8492545706409182,
+      "grad_norm": 0.4191280874392949,
+      "learning_rate": 3.5793459850388822e-06,
+      "loss": 0.4079,
+      "step": 17310
+    },
+    {
+      "epoch": 2.8494191548600347,
+      "grad_norm": 0.32954077567316326,
+      "learning_rate": 3.578885869204727e-06,
+      "loss": 0.4324,
+      "step": 17311
+    },
+    {
+      "epoch": 2.849583739079151,
+      "grad_norm": 0.3743297123453578,
+      "learning_rate": 3.578425763421403e-06,
+      "loss": 0.4265,
+      "step": 17312
+    },
+    {
+      "epoch": 2.8497483232982677,
+      "grad_norm": 0.2850966383078337,
+      "learning_rate": 3.57796566769393e-06,
+      "loss": 0.4212,
+      "step": 17313
+    },
+    {
+      "epoch": 2.849912907517384,
+      "grad_norm": 0.41129961107418034,
+      "learning_rate": 3.5775055820273255e-06,
+      "loss": 0.4365,
+      "step": 17314
+    },
+    {
+      "epoch": 2.8500774917365006,
+      "grad_norm": 0.3414907552588458,
+      "learning_rate": 3.5770455064266136e-06,
+      "loss": 0.4338,
+      "step": 17315
+    },
+    {
+      "epoch": 2.850242075955617,
+      "grad_norm": 0.29009696112826655,
+      "learning_rate": 3.576585440896811e-06,
+      "loss": 0.4174,
+      "step": 17316
+    },
+    {
+      "epoch": 2.8504066601747335,
+      "grad_norm": 0.4310254529123287,
+      "learning_rate": 3.5761253854429407e-06,
+      "loss": 0.4204,
+      "step": 17317
+    },
+    {
+      "epoch": 2.85057124439385,
+      "grad_norm": 0.31112114797040397,
+      "learning_rate": 3.575665340070021e-06,
+      "loss": 0.4341,
+      "step": 17318
+    },
+    {
+      "epoch": 2.8507358286129665,
+      "grad_norm": 0.3279739330863808,
+      "learning_rate": 3.5752053047830725e-06,
+      "loss": 0.4322,
+      "step": 17319
+    },
+    {
+      "epoch": 2.850900412832083,
+      "grad_norm": 0.3301961644714612,
+      "learning_rate": 3.5747452795871153e-06,
+      "loss": 0.4349,
+      "step": 17320
+    },
+    {
+      "epoch": 2.8510649970511994,
+      "grad_norm": 0.277429509503672,
+      "learning_rate": 3.574285264487167e-06,
+      "loss": 0.4504,
+      "step": 17321
+    },
+    {
+      "epoch": 2.851229581270316,
+      "grad_norm": 0.4076592298941959,
+      "learning_rate": 3.57382525948825e-06,
+      "loss": 0.4261,
+      "step": 17322
+    },
+    {
+      "epoch": 2.8513941654894324,
+      "grad_norm": 0.2610069194071665,
+      "learning_rate": 3.5733652645953796e-06,
+      "loss": 0.4286,
+      "step": 17323
+    },
+    {
+      "epoch": 2.851558749708549,
+      "grad_norm": 0.3170634069177754,
+      "learning_rate": 3.5729052798135806e-06,
+      "loss": 0.4292,
+      "step": 17324
+    },
+    {
+      "epoch": 2.8517233339276653,
+      "grad_norm": 0.37389489628158157,
+      "learning_rate": 3.572445305147866e-06,
+      "loss": 0.4369,
+      "step": 17325
+    },
+    {
+      "epoch": 2.8518879181467818,
+      "grad_norm": 0.30498221629286726,
+      "learning_rate": 3.571985340603261e-06,
+      "loss": 0.4343,
+      "step": 17326
+    },
+    {
+      "epoch": 2.8520525023658982,
+      "grad_norm": 0.32470011381080516,
+      "learning_rate": 3.571525386184781e-06,
+      "loss": 0.4253,
+      "step": 17327
+    },
+    {
+      "epoch": 2.8522170865850147,
+      "grad_norm": 0.3199015472926689,
+      "learning_rate": 3.5710654418974444e-06,
+      "loss": 0.4437,
+      "step": 17328
+    },
+    {
+      "epoch": 2.852381670804131,
+      "grad_norm": 0.3304374071937174,
+      "learning_rate": 3.5706055077462715e-06,
+      "loss": 0.4349,
+      "step": 17329
+    },
+    {
+      "epoch": 2.8525462550232477,
+      "grad_norm": 0.5648046194401402,
+      "learning_rate": 3.5701455837362798e-06,
+      "loss": 0.4487,
+      "step": 17330
+    },
+    {
+      "epoch": 2.852710839242364,
+      "grad_norm": 0.3312659649653881,
+      "learning_rate": 3.5696856698724896e-06,
+      "loss": 0.4234,
+      "step": 17331
+    },
+    {
+      "epoch": 2.8528754234614806,
+      "grad_norm": 0.2999730938000472,
+      "learning_rate": 3.5692257661599166e-06,
+      "loss": 0.4201,
+      "step": 17332
+    },
+    {
+      "epoch": 2.8530400076805966,
+      "grad_norm": 0.3054685782593052,
+      "learning_rate": 3.5687658726035825e-06,
+      "loss": 0.4441,
+      "step": 17333
+    },
+    {
+      "epoch": 2.853204591899713,
+      "grad_norm": 0.29541052668524603,
+      "learning_rate": 3.5683059892085025e-06,
+      "loss": 0.4456,
+      "step": 17334
+    },
+    {
+      "epoch": 2.8533691761188296,
+      "grad_norm": 0.39264174913702915,
+      "learning_rate": 3.5678461159796953e-06,
+      "loss": 0.4203,
+      "step": 17335
+    },
+    {
+      "epoch": 2.853533760337946,
+      "grad_norm": 0.2988594884479697,
+      "learning_rate": 3.56738625292218e-06,
+      "loss": 0.4346,
+      "step": 17336
+    },
+    {
+      "epoch": 2.8536983445570625,
+      "grad_norm": 0.29792359684867575,
+      "learning_rate": 3.5669264000409724e-06,
+      "loss": 0.4224,
+      "step": 17337
+    },
+    {
+      "epoch": 2.853862928776179,
+      "grad_norm": 0.31966824404577093,
+      "learning_rate": 3.566466557341093e-06,
+      "loss": 0.43,
+      "step": 17338
+    },
+    {
+      "epoch": 2.8540275129952954,
+      "grad_norm": 0.3570823411675509,
+      "learning_rate": 3.5660067248275568e-06,
+      "loss": 0.4297,
+      "step": 17339
+    },
+    {
+      "epoch": 2.854192097214412,
+      "grad_norm": 0.2703363496172826,
+      "learning_rate": 3.565546902505383e-06,
+      "loss": 0.4263,
+      "step": 17340
+    },
+    {
+      "epoch": 2.8543566814335284,
+      "grad_norm": 0.3790065141753752,
+      "learning_rate": 3.5650870903795897e-06,
+      "loss": 0.437,
+      "step": 17341
+    },
+    {
+      "epoch": 2.854521265652645,
+      "grad_norm": 0.3175430522401861,
+      "learning_rate": 3.56462728845519e-06,
+      "loss": 0.4308,
+      "step": 17342
+    },
+    {
+      "epoch": 2.8546858498717613,
+      "grad_norm": 0.3151100167548812,
+      "learning_rate": 3.5641674967372074e-06,
+      "loss": 0.4393,
+      "step": 17343
+    },
+    {
+      "epoch": 2.854850434090878,
+      "grad_norm": 0.33334304766233647,
+      "learning_rate": 3.563707715230652e-06,
+      "loss": 0.4384,
+      "step": 17344
+    },
+    {
+      "epoch": 2.8550150183099943,
+      "grad_norm": 0.28834839600363854,
+      "learning_rate": 3.5632479439405475e-06,
+      "loss": 0.4298,
+      "step": 17345
+    },
+    {
+      "epoch": 2.8551796025291107,
+      "grad_norm": 0.3013133808479478,
+      "learning_rate": 3.562788182871906e-06,
+      "loss": 0.4468,
+      "step": 17346
+    },
+    {
+      "epoch": 2.855344186748227,
+      "grad_norm": 0.2896137490725572,
+      "learning_rate": 3.5623284320297465e-06,
+      "loss": 0.4125,
+      "step": 17347
+    },
+    {
+      "epoch": 2.8555087709673437,
+      "grad_norm": 0.2920795502549955,
+      "learning_rate": 3.561868691419084e-06,
+      "loss": 0.4212,
+      "step": 17348
+    },
+    {
+      "epoch": 2.85567335518646,
+      "grad_norm": 0.3502524102384909,
+      "learning_rate": 3.561408961044937e-06,
+      "loss": 0.4219,
+      "step": 17349
+    },
+    {
+      "epoch": 2.8558379394055766,
+      "grad_norm": 0.2963465990646538,
+      "learning_rate": 3.5609492409123205e-06,
+      "loss": 0.4222,
+      "step": 17350
+    },
+    {
+      "epoch": 2.856002523624693,
+      "grad_norm": 0.46442209356470837,
+      "learning_rate": 3.5604895310262503e-06,
+      "loss": 0.4449,
+      "step": 17351
+    },
+    {
+      "epoch": 2.8561671078438096,
+      "grad_norm": 0.33772961723301825,
+      "learning_rate": 3.5600298313917452e-06,
+      "loss": 0.4438,
+      "step": 17352
+    },
+    {
+      "epoch": 2.856331692062926,
+      "grad_norm": 0.2988364502504436,
+      "learning_rate": 3.559570142013817e-06,
+      "loss": 0.4336,
+      "step": 17353
+    },
+    {
+      "epoch": 2.8564962762820425,
+      "grad_norm": 0.3835527090232755,
+      "learning_rate": 3.5591104628974865e-06,
+      "loss": 0.4489,
+      "step": 17354
+    },
+    {
+      "epoch": 2.856660860501159,
+      "grad_norm": 0.31238237100554256,
+      "learning_rate": 3.558650794047764e-06,
+      "loss": 0.4256,
+      "step": 17355
+    },
+    {
+      "epoch": 2.8568254447202754,
+      "grad_norm": 0.30854875596452236,
+      "learning_rate": 3.558191135469672e-06,
+      "loss": 0.4261,
+      "step": 17356
+    },
+    {
+      "epoch": 2.856990028939392,
+      "grad_norm": 0.3281040543302565,
+      "learning_rate": 3.5577314871682214e-06,
+      "loss": 0.4355,
+      "step": 17357
+    },
+    {
+      "epoch": 2.8571546131585084,
+      "grad_norm": 0.46861731742534457,
+      "learning_rate": 3.557271849148427e-06,
+      "loss": 0.4253,
+      "step": 17358
+    },
+    {
+      "epoch": 2.857319197377625,
+      "grad_norm": 0.31972857803363086,
+      "learning_rate": 3.556812221415307e-06,
+      "loss": 0.4595,
+      "step": 17359
+    },
+    {
+      "epoch": 2.8574837815967413,
+      "grad_norm": 0.34830650317616807,
+      "learning_rate": 3.5563526039738758e-06,
+      "loss": 0.4333,
+      "step": 17360
+    },
+    {
+      "epoch": 2.857648365815858,
+      "grad_norm": 0.26551698406335,
+      "learning_rate": 3.5558929968291484e-06,
+      "loss": 0.4248,
+      "step": 17361
+    },
+    {
+      "epoch": 2.8578129500349743,
+      "grad_norm": 0.3552027859489436,
+      "learning_rate": 3.5554333999861386e-06,
+      "loss": 0.4278,
+      "step": 17362
+    },
+    {
+      "epoch": 2.8579775342540907,
+      "grad_norm": 0.49062427891614496,
+      "learning_rate": 3.5549738134498643e-06,
+      "loss": 0.4365,
+      "step": 17363
+    },
+    {
+      "epoch": 2.858142118473207,
+      "grad_norm": 0.39124919194394747,
+      "learning_rate": 3.554514237225339e-06,
+      "loss": 0.4399,
+      "step": 17364
+    },
+    {
+      "epoch": 2.8583067026923237,
+      "grad_norm": 2.118110390489228,
+      "learning_rate": 3.554054671317576e-06,
+      "loss": 0.4454,
+      "step": 17365
+    },
+    {
+      "epoch": 2.85847128691144,
+      "grad_norm": 0.3123363002781529,
+      "learning_rate": 3.5535951157315907e-06,
+      "loss": 0.4008,
+      "step": 17366
+    },
+    {
+      "epoch": 2.8586358711305566,
+      "grad_norm": 0.3400430711403663,
+      "learning_rate": 3.5531355704723975e-06,
+      "loss": 0.4138,
+      "step": 17367
+    },
+    {
+      "epoch": 2.858800455349673,
+      "grad_norm": 0.34525379179719407,
+      "learning_rate": 3.552676035545012e-06,
+      "loss": 0.4349,
+      "step": 17368
+    },
+    {
+      "epoch": 2.8589650395687896,
+      "grad_norm": 0.2987836071432275,
+      "learning_rate": 3.552216510954446e-06,
+      "loss": 0.4202,
+      "step": 17369
+    },
+    {
+      "epoch": 2.859129623787906,
+      "grad_norm": 0.3772431570981102,
+      "learning_rate": 3.551756996705717e-06,
+      "loss": 0.4281,
+      "step": 17370
+    },
+    {
+      "epoch": 2.8592942080070225,
+      "grad_norm": 0.27907664529479576,
+      "learning_rate": 3.5512974928038372e-06,
+      "loss": 0.4228,
+      "step": 17371
+    },
+    {
+      "epoch": 2.859458792226139,
+      "grad_norm": 0.4616168933271382,
+      "learning_rate": 3.5508379992538183e-06,
+      "loss": 0.4043,
+      "step": 17372
+    },
+    {
+      "epoch": 2.8596233764452554,
+      "grad_norm": 0.3168197813319331,
+      "learning_rate": 3.5503785160606784e-06,
+      "loss": 0.4307,
+      "step": 17373
+    },
+    {
+      "epoch": 2.859787960664372,
+      "grad_norm": 0.31139488991072456,
+      "learning_rate": 3.549919043229427e-06,
+      "loss": 0.3988,
+      "step": 17374
+    },
+    {
+      "epoch": 2.859952544883488,
+      "grad_norm": 0.34256371374820876,
+      "learning_rate": 3.5494595807650833e-06,
+      "loss": 0.4285,
+      "step": 17375
+    },
+    {
+      "epoch": 2.8601171291026044,
+      "grad_norm": 0.32999590179031507,
+      "learning_rate": 3.5490001286726546e-06,
+      "loss": 0.4054,
+      "step": 17376
+    },
+    {
+      "epoch": 2.860281713321721,
+      "grad_norm": 0.3151567022626986,
+      "learning_rate": 3.5485406869571574e-06,
+      "loss": 0.424,
+      "step": 17377
+    },
+    {
+      "epoch": 2.8604462975408373,
+      "grad_norm": 0.2944405518922663,
+      "learning_rate": 3.5480812556236054e-06,
+      "loss": 0.404,
+      "step": 17378
+    },
+    {
+      "epoch": 2.860610881759954,
+      "grad_norm": 0.31649209921217125,
+      "learning_rate": 3.5476218346770096e-06,
+      "loss": 0.4265,
+      "step": 17379
+    },
+    {
+      "epoch": 2.8607754659790703,
+      "grad_norm": 0.32118599442886553,
+      "learning_rate": 3.547162424122385e-06,
+      "loss": 0.4076,
+      "step": 17380
+    },
+    {
+      "epoch": 2.8609400501981868,
+      "grad_norm": 0.34538102601435855,
+      "learning_rate": 3.546703023964743e-06,
+      "loss": 0.4455,
+      "step": 17381
+    },
+    {
+      "epoch": 2.8611046344173032,
+      "grad_norm": 0.5686932534326282,
+      "learning_rate": 3.546243634209098e-06,
+      "loss": 0.4399,
+      "step": 17382
+    },
+    {
+      "epoch": 2.8612692186364197,
+      "grad_norm": 0.35543383109799415,
+      "learning_rate": 3.545784254860461e-06,
+      "loss": 0.4455,
+      "step": 17383
+    },
+    {
+      "epoch": 2.861433802855536,
+      "grad_norm": 0.6460606174437876,
+      "learning_rate": 3.545324885923847e-06,
+      "loss": 0.4401,
+      "step": 17384
+    },
+    {
+      "epoch": 2.8615983870746526,
+      "grad_norm": 0.30017636310835727,
+      "learning_rate": 3.544865527404266e-06,
+      "loss": 0.4167,
+      "step": 17385
+    },
+    {
+      "epoch": 2.861762971293769,
+      "grad_norm": 0.31817270123624747,
+      "learning_rate": 3.5444061793067305e-06,
+      "loss": 0.4135,
+      "step": 17386
+    },
+    {
+      "epoch": 2.8619275555128856,
+      "grad_norm": 2.0942530739338068,
+      "learning_rate": 3.5439468416362544e-06,
+      "loss": 0.424,
+      "step": 17387
+    },
+    {
+      "epoch": 2.862092139732002,
+      "grad_norm": 0.3462842069028323,
+      "learning_rate": 3.543487514397848e-06,
+      "loss": 0.4154,
+      "step": 17388
+    },
+    {
+      "epoch": 2.8622567239511185,
+      "grad_norm": 0.31711819493067556,
+      "learning_rate": 3.543028197596524e-06,
+      "loss": 0.4368,
+      "step": 17389
+    },
+    {
+      "epoch": 2.862421308170235,
+      "grad_norm": 0.392098853077831,
+      "learning_rate": 3.5425688912372943e-06,
+      "loss": 0.4379,
+      "step": 17390
+    },
+    {
+      "epoch": 2.8625858923893515,
+      "grad_norm": 0.566164161477723,
+      "learning_rate": 3.5421095953251714e-06,
+      "loss": 0.4067,
+      "step": 17391
+    },
+    {
+      "epoch": 2.862750476608468,
+      "grad_norm": 0.6690822106221481,
+      "learning_rate": 3.5416503098651653e-06,
+      "loss": 0.4234,
+      "step": 17392
+    },
+    {
+      "epoch": 2.8629150608275844,
+      "grad_norm": 0.302852719194461,
+      "learning_rate": 3.5411910348622897e-06,
+      "loss": 0.4129,
+      "step": 17393
+    },
+    {
+      "epoch": 2.863079645046701,
+      "grad_norm": 0.2589054574663652,
+      "learning_rate": 3.5407317703215562e-06,
+      "loss": 0.4519,
+      "step": 17394
+    },
+    {
+      "epoch": 2.8632442292658173,
+      "grad_norm": 0.4101958079789623,
+      "learning_rate": 3.5402725162479723e-06,
+      "loss": 0.4272,
+      "step": 17395
+    },
+    {
+      "epoch": 2.863408813484934,
+      "grad_norm": 0.26525454642865537,
+      "learning_rate": 3.5398132726465524e-06,
+      "loss": 0.4278,
+      "step": 17396
+    },
+    {
+      "epoch": 2.8635733977040503,
+      "grad_norm": 0.3779916268979564,
+      "learning_rate": 3.539354039522306e-06,
+      "loss": 0.451,
+      "step": 17397
+    },
+    {
+      "epoch": 2.8637379819231668,
+      "grad_norm": 0.36607349629965114,
+      "learning_rate": 3.5388948168802463e-06,
+      "loss": 0.4443,
+      "step": 17398
+    },
+    {
+      "epoch": 2.8639025661422832,
+      "grad_norm": 0.3140599194344739,
+      "learning_rate": 3.538435604725382e-06,
+      "loss": 0.4313,
+      "step": 17399
+    },
+    {
+      "epoch": 2.8640671503613992,
+      "grad_norm": 0.2858037967176737,
+      "learning_rate": 3.5379764030627252e-06,
+      "loss": 0.4064,
+      "step": 17400
+    },
+    {
+      "epoch": 2.8642317345805157,
+      "grad_norm": 0.4345367756236418,
+      "learning_rate": 3.537517211897286e-06,
+      "loss": 0.4254,
+      "step": 17401
+    },
+    {
+      "epoch": 2.864396318799632,
+      "grad_norm": 0.5451837729821961,
+      "learning_rate": 3.5370580312340736e-06,
+      "loss": 0.4369,
+      "step": 17402
+    },
+    {
+      "epoch": 2.8645609030187487,
+      "grad_norm": 0.37529945803443476,
+      "learning_rate": 3.5365988610781017e-06,
+      "loss": 0.4083,
+      "step": 17403
+    },
+    {
+      "epoch": 2.864725487237865,
+      "grad_norm": 0.3037152828583387,
+      "learning_rate": 3.5361397014343765e-06,
+      "loss": 0.4587,
+      "step": 17404
+    },
+    {
+      "epoch": 2.8648900714569816,
+      "grad_norm": 0.35187267786260457,
+      "learning_rate": 3.5356805523079118e-06,
+      "loss": 0.4292,
+      "step": 17405
+    },
+    {
+      "epoch": 2.865054655676098,
+      "grad_norm": 0.3494197798213748,
+      "learning_rate": 3.5352214137037142e-06,
+      "loss": 0.435,
+      "step": 17406
+    },
+    {
+      "epoch": 2.8652192398952145,
+      "grad_norm": 0.4412802947995881,
+      "learning_rate": 3.534762285626798e-06,
+      "loss": 0.4333,
+      "step": 17407
+    },
+    {
+      "epoch": 2.865383824114331,
+      "grad_norm": 0.475036997514968,
+      "learning_rate": 3.5343031680821692e-06,
+      "loss": 0.4351,
+      "step": 17408
+    },
+    {
+      "epoch": 2.8655484083334475,
+      "grad_norm": 0.3499265166691491,
+      "learning_rate": 3.5338440610748383e-06,
+      "loss": 0.4361,
+      "step": 17409
+    },
+    {
+      "epoch": 2.865712992552564,
+      "grad_norm": 0.3023794348814161,
+      "learning_rate": 3.5333849646098163e-06,
+      "loss": 0.4304,
+      "step": 17410
+    },
+    {
+      "epoch": 2.8658775767716804,
+      "grad_norm": 0.366119061707016,
+      "learning_rate": 3.5329258786921108e-06,
+      "loss": 0.42,
+      "step": 17411
+    },
+    {
+      "epoch": 2.866042160990797,
+      "grad_norm": 0.3774323397580119,
+      "learning_rate": 3.532466803326733e-06,
+      "loss": 0.4231,
+      "step": 17412
+    },
+    {
+      "epoch": 2.8662067452099134,
+      "grad_norm": 0.31430262520379176,
+      "learning_rate": 3.5320077385186905e-06,
+      "loss": 0.4533,
+      "step": 17413
+    },
+    {
+      "epoch": 2.86637132942903,
+      "grad_norm": 0.3252067503048915,
+      "learning_rate": 3.5315486842729953e-06,
+      "loss": 0.4278,
+      "step": 17414
+    },
+    {
+      "epoch": 2.8665359136481463,
+      "grad_norm": 0.33254081249056655,
+      "learning_rate": 3.531089640594653e-06,
+      "loss": 0.4292,
+      "step": 17415
+    },
+    {
+      "epoch": 2.8667004978672628,
+      "grad_norm": 0.32644976125481445,
+      "learning_rate": 3.5306306074886727e-06,
+      "loss": 0.4156,
+      "step": 17416
+    },
+    {
+      "epoch": 2.8668650820863792,
+      "grad_norm": 0.36852603099967607,
+      "learning_rate": 3.5301715849600656e-06,
+      "loss": 0.4215,
+      "step": 17417
+    },
+    {
+      "epoch": 2.8670296663054957,
+      "grad_norm": 0.3235318335429592,
+      "learning_rate": 3.5297125730138384e-06,
+      "loss": 0.4174,
+      "step": 17418
+    },
+    {
+      "epoch": 2.867194250524612,
+      "grad_norm": 0.28132456297739444,
+      "learning_rate": 3.5292535716550013e-06,
+      "loss": 0.4454,
+      "step": 17419
+    },
+    {
+      "epoch": 2.8673588347437287,
+      "grad_norm": 0.3586742226160919,
+      "learning_rate": 3.5287945808885602e-06,
+      "loss": 0.4249,
+      "step": 17420
+    },
+    {
+      "epoch": 2.867523418962845,
+      "grad_norm": 0.3134723949749536,
+      "learning_rate": 3.5283356007195267e-06,
+      "loss": 0.4231,
+      "step": 17421
+    },
+    {
+      "epoch": 2.8676880031819616,
+      "grad_norm": 0.2842395584626507,
+      "learning_rate": 3.5278766311529083e-06,
+      "loss": 0.4224,
+      "step": 17422
+    },
+    {
+      "epoch": 2.867852587401078,
+      "grad_norm": 0.5224422130616279,
+      "learning_rate": 3.527417672193709e-06,
+      "loss": 0.4343,
+      "step": 17423
+    },
+    {
+      "epoch": 2.8680171716201945,
+      "grad_norm": 0.8431668915744212,
+      "learning_rate": 3.526958723846942e-06,
+      "loss": 0.4366,
+      "step": 17424
+    },
+    {
+      "epoch": 2.868181755839311,
+      "grad_norm": 0.30047647153435625,
+      "learning_rate": 3.5264997861176117e-06,
+      "loss": 0.4433,
+      "step": 17425
+    },
+    {
+      "epoch": 2.8683463400584275,
+      "grad_norm": 0.3399635986036373,
+      "learning_rate": 3.526040859010729e-06,
+      "loss": 0.426,
+      "step": 17426
+    },
+    {
+      "epoch": 2.868510924277544,
+      "grad_norm": 0.34800450958763834,
+      "learning_rate": 3.525581942531298e-06,
+      "loss": 0.4311,
+      "step": 17427
+    },
+    {
+      "epoch": 2.8686755084966604,
+      "grad_norm": 0.35944897049779106,
+      "learning_rate": 3.5251230366843294e-06,
+      "loss": 0.4309,
+      "step": 17428
+    },
+    {
+      "epoch": 2.868840092715777,
+      "grad_norm": 0.36495359198570226,
+      "learning_rate": 3.5246641414748273e-06,
+      "loss": 0.4274,
+      "step": 17429
+    },
+    {
+      "epoch": 2.8690046769348934,
+      "grad_norm": 0.34425709534357957,
+      "learning_rate": 3.524205256907803e-06,
+      "loss": 0.4272,
+      "step": 17430
+    },
+    {
+      "epoch": 2.86916926115401,
+      "grad_norm": 0.621363614941926,
+      "learning_rate": 3.52374638298826e-06,
+      "loss": 0.4213,
+      "step": 17431
+    },
+    {
+      "epoch": 2.8693338453731263,
+      "grad_norm": 0.33192276900525747,
+      "learning_rate": 3.523287519721207e-06,
+      "loss": 0.4098,
+      "step": 17432
+    },
+    {
+      "epoch": 2.8694984295922428,
+      "grad_norm": 0.3357576294843523,
+      "learning_rate": 3.522828667111653e-06,
+      "loss": 0.4359,
+      "step": 17433
+    },
+    {
+      "epoch": 2.8696630138113592,
+      "grad_norm": 0.3811141647375937,
+      "learning_rate": 3.5223698251645993e-06,
+      "loss": 0.4202,
+      "step": 17434
+    },
+    {
+      "epoch": 2.8698275980304757,
+      "grad_norm": 0.2769062285422329,
+      "learning_rate": 3.521910993885058e-06,
+      "loss": 0.4435,
+      "step": 17435
+    },
+    {
+      "epoch": 2.869992182249592,
+      "grad_norm": 0.3068595168744032,
+      "learning_rate": 3.5214521732780315e-06,
+      "loss": 0.4417,
+      "step": 17436
+    },
+    {
+      "epoch": 2.8701567664687087,
+      "grad_norm": 0.3188531411080381,
+      "learning_rate": 3.5209933633485308e-06,
+      "loss": 0.4268,
+      "step": 17437
+    },
+    {
+      "epoch": 2.870321350687825,
+      "grad_norm": 0.36896820242208617,
+      "learning_rate": 3.5205345641015593e-06,
+      "loss": 0.4448,
+      "step": 17438
+    },
+    {
+      "epoch": 2.8704859349069416,
+      "grad_norm": 0.39656766242190944,
+      "learning_rate": 3.520075775542123e-06,
+      "loss": 0.4211,
+      "step": 17439
+    },
+    {
+      "epoch": 2.870650519126058,
+      "grad_norm": 0.32019722022746766,
+      "learning_rate": 3.519616997675229e-06,
+      "loss": 0.4338,
+      "step": 17440
+    },
+    {
+      "epoch": 2.870815103345174,
+      "grad_norm": 0.2839626543679687,
+      "learning_rate": 3.519158230505882e-06,
+      "loss": 0.4261,
+      "step": 17441
+    },
+    {
+      "epoch": 2.8709796875642906,
+      "grad_norm": 0.3489103217881381,
+      "learning_rate": 3.5186994740390904e-06,
+      "loss": 0.4542,
+      "step": 17442
+    },
+    {
+      "epoch": 2.871144271783407,
+      "grad_norm": 0.38422954466838666,
+      "learning_rate": 3.5182407282798575e-06,
+      "loss": 0.4119,
+      "step": 17443
+    },
+    {
+      "epoch": 2.8713088560025235,
+      "grad_norm": 0.46146800085205075,
+      "learning_rate": 3.5177819932331905e-06,
+      "loss": 0.4263,
+      "step": 17444
+    },
+    {
+      "epoch": 2.87147344022164,
+      "grad_norm": 0.4116554893544578,
+      "learning_rate": 3.517323268904096e-06,
+      "loss": 0.4317,
+      "step": 17445
+    },
+    {
+      "epoch": 2.8716380244407564,
+      "grad_norm": 0.44247422368616773,
+      "learning_rate": 3.5168645552975754e-06,
+      "loss": 0.4339,
+      "step": 17446
+    },
+    {
+      "epoch": 2.871802608659873,
+      "grad_norm": 0.25576977603313145,
+      "learning_rate": 3.516405852418638e-06,
+      "loss": 0.4011,
+      "step": 17447
+    },
+    {
+      "epoch": 2.8719671928789894,
+      "grad_norm": 0.3409850047465722,
+      "learning_rate": 3.5159471602722856e-06,
+      "loss": 0.4195,
+      "step": 17448
+    },
+    {
+      "epoch": 2.872131777098106,
+      "grad_norm": 0.3528487025903805,
+      "learning_rate": 3.5154884788635263e-06,
+      "loss": 0.4265,
+      "step": 17449
+    },
+    {
+      "epoch": 2.8722963613172223,
+      "grad_norm": 0.9376489113719304,
+      "learning_rate": 3.515029808197363e-06,
+      "loss": 0.4388,
+      "step": 17450
+    },
+    {
+      "epoch": 2.872460945536339,
+      "grad_norm": 0.26099699964732215,
+      "learning_rate": 3.5145711482788023e-06,
+      "loss": 0.4194,
+      "step": 17451
+    },
+    {
+      "epoch": 2.8726255297554553,
+      "grad_norm": 0.38026553404402125,
+      "learning_rate": 3.5141124991128477e-06,
+      "loss": 0.4176,
+      "step": 17452
+    },
+    {
+      "epoch": 2.8727901139745717,
+      "grad_norm": 0.29526089747718154,
+      "learning_rate": 3.5136538607045034e-06,
+      "loss": 0.4332,
+      "step": 17453
+    },
+    {
+      "epoch": 2.872954698193688,
+      "grad_norm": 0.30887571600772057,
+      "learning_rate": 3.513195233058776e-06,
+      "loss": 0.4357,
+      "step": 17454
+    },
+    {
+      "epoch": 2.8731192824128047,
+      "grad_norm": 0.344099036251958,
+      "learning_rate": 3.5127366161806655e-06,
+      "loss": 0.4215,
+      "step": 17455
+    },
+    {
+      "epoch": 2.873283866631921,
+      "grad_norm": 0.4456054969025151,
+      "learning_rate": 3.512278010075182e-06,
+      "loss": 0.4396,
+      "step": 17456
+    },
+    {
+      "epoch": 2.8734484508510376,
+      "grad_norm": 0.4540551366967078,
+      "learning_rate": 3.511819414747325e-06,
+      "loss": 0.4468,
+      "step": 17457
+    },
+    {
+      "epoch": 2.873613035070154,
+      "grad_norm": 0.3225405214185915,
+      "learning_rate": 3.511360830202101e-06,
+      "loss": 0.4459,
+      "step": 17458
+    },
+    {
+      "epoch": 2.8737776192892706,
+      "grad_norm": 0.261252551838841,
+      "learning_rate": 3.5109022564445134e-06,
+      "loss": 0.428,
+      "step": 17459
+    },
+    {
+      "epoch": 2.873942203508387,
+      "grad_norm": 0.34881152820244593,
+      "learning_rate": 3.510443693479564e-06,
+      "loss": 0.4519,
+      "step": 17460
+    },
+    {
+      "epoch": 2.8741067877275035,
+      "grad_norm": 0.3918089656304943,
+      "learning_rate": 3.5099851413122596e-06,
+      "loss": 0.4387,
+      "step": 17461
+    },
+    {
+      "epoch": 2.87427137194662,
+      "grad_norm": 0.39959836359225365,
+      "learning_rate": 3.509526599947601e-06,
+      "loss": 0.4246,
+      "step": 17462
+    },
+    {
+      "epoch": 2.8744359561657364,
+      "grad_norm": 0.33673458398190126,
+      "learning_rate": 3.5090680693905937e-06,
+      "loss": 0.432,
+      "step": 17463
+    },
+    {
+      "epoch": 2.874600540384853,
+      "grad_norm": 0.2979815248415507,
+      "learning_rate": 3.50860954964624e-06,
+      "loss": 0.4131,
+      "step": 17464
+    },
+    {
+      "epoch": 2.8747651246039694,
+      "grad_norm": 0.39709378080258345,
+      "learning_rate": 3.5081510407195443e-06,
+      "loss": 0.4392,
+      "step": 17465
+    },
+    {
+      "epoch": 2.8749297088230854,
+      "grad_norm": 0.313235605852531,
+      "learning_rate": 3.5076925426155077e-06,
+      "loss": 0.4301,
+      "step": 17466
+    },
+    {
+      "epoch": 2.875094293042202,
+      "grad_norm": 0.3350202953389066,
+      "learning_rate": 3.507234055339133e-06,
+      "loss": 0.4333,
+      "step": 17467
+    },
+    {
+      "epoch": 2.8752588772613183,
+      "grad_norm": 0.3393102073674103,
+      "learning_rate": 3.506775578895426e-06,
+      "loss": 0.4146,
+      "step": 17468
+    },
+    {
+      "epoch": 2.875423461480435,
+      "grad_norm": 0.3597545770156154,
+      "learning_rate": 3.506317113289385e-06,
+      "loss": 0.4253,
+      "step": 17469
+    },
+    {
+      "epoch": 2.8755880456995513,
+      "grad_norm": 0.42835161910731306,
+      "learning_rate": 3.505858658526017e-06,
+      "loss": 0.4154,
+      "step": 17470
+    },
+    {
+      "epoch": 2.8757526299186678,
+      "grad_norm": 0.33497983250833635,
+      "learning_rate": 3.5054002146103216e-06,
+      "loss": 0.4248,
+      "step": 17471
+    },
+    {
+      "epoch": 2.8759172141377842,
+      "grad_norm": 0.32949795625342293,
+      "learning_rate": 3.5049417815473027e-06,
+      "loss": 0.4187,
+      "step": 17472
+    },
+    {
+      "epoch": 2.8760817983569007,
+      "grad_norm": 0.32898726958386537,
+      "learning_rate": 3.504483359341961e-06,
+      "loss": 0.4363,
+      "step": 17473
+    },
+    {
+      "epoch": 2.876246382576017,
+      "grad_norm": 0.4864876133253663,
+      "learning_rate": 3.504024947999301e-06,
+      "loss": 0.444,
+      "step": 17474
+    },
+    {
+      "epoch": 2.8764109667951336,
+      "grad_norm": 0.3015309938868187,
+      "learning_rate": 3.5035665475243237e-06,
+      "loss": 0.4062,
+      "step": 17475
+    },
+    {
+      "epoch": 2.87657555101425,
+      "grad_norm": 0.278539466113373,
+      "learning_rate": 3.503108157922029e-06,
+      "loss": 0.3998,
+      "step": 17476
+    },
+    {
+      "epoch": 2.8767401352333666,
+      "grad_norm": 0.3085679687824653,
+      "learning_rate": 3.502649779197421e-06,
+      "loss": 0.4325,
+      "step": 17477
+    },
+    {
+      "epoch": 2.876904719452483,
+      "grad_norm": 0.2981762361349608,
+      "learning_rate": 3.5021914113554993e-06,
+      "loss": 0.4345,
+      "step": 17478
+    },
+    {
+      "epoch": 2.8770693036715995,
+      "grad_norm": 0.42018455478878153,
+      "learning_rate": 3.5017330544012684e-06,
+      "loss": 0.4417,
+      "step": 17479
+    },
+    {
+      "epoch": 2.877233887890716,
+      "grad_norm": 0.3194290282093159,
+      "learning_rate": 3.501274708339727e-06,
+      "loss": 0.4217,
+      "step": 17480
+    },
+    {
+      "epoch": 2.8773984721098325,
+      "grad_norm": 0.2961955978280596,
+      "learning_rate": 3.500816373175878e-06,
+      "loss": 0.4146,
+      "step": 17481
+    },
+    {
+      "epoch": 2.877563056328949,
+      "grad_norm": 0.35320320042854264,
+      "learning_rate": 3.5003580489147217e-06,
+      "loss": 0.4167,
+      "step": 17482
+    },
+    {
+      "epoch": 2.8777276405480654,
+      "grad_norm": 0.2914516303858204,
+      "learning_rate": 3.499899735561259e-06,
+      "loss": 0.4225,
+      "step": 17483
+    },
+    {
+      "epoch": 2.877892224767182,
+      "grad_norm": 0.5742647961884108,
+      "learning_rate": 3.4994414331204927e-06,
+      "loss": 0.4424,
+      "step": 17484
+    },
+    {
+      "epoch": 2.8780568089862983,
+      "grad_norm": 0.47125135149148517,
+      "learning_rate": 3.4989831415974196e-06,
+      "loss": 0.4127,
+      "step": 17485
+    },
+    {
+      "epoch": 2.878221393205415,
+      "grad_norm": 0.277633921305222,
+      "learning_rate": 3.498524860997046e-06,
+      "loss": 0.4379,
+      "step": 17486
+    },
+    {
+      "epoch": 2.8783859774245313,
+      "grad_norm": 0.35444514695476276,
+      "learning_rate": 3.498066591324366e-06,
+      "loss": 0.415,
+      "step": 17487
+    },
+    {
+      "epoch": 2.8785505616436478,
+      "grad_norm": 0.3702431165102328,
+      "learning_rate": 3.4976083325843866e-06,
+      "loss": 0.4407,
+      "step": 17488
+    },
+    {
+      "epoch": 2.8787151458627642,
+      "grad_norm": 0.36745646953654393,
+      "learning_rate": 3.4971500847821044e-06,
+      "loss": 0.427,
+      "step": 17489
+    },
+    {
+      "epoch": 2.8788797300818807,
+      "grad_norm": 0.3012571879088693,
+      "learning_rate": 3.4966918479225185e-06,
+      "loss": 0.4189,
+      "step": 17490
+    },
+    {
+      "epoch": 2.879044314300997,
+      "grad_norm": 0.3522602230756812,
+      "learning_rate": 3.4962336220106322e-06,
+      "loss": 0.4279,
+      "step": 17491
+    },
+    {
+      "epoch": 2.8792088985201136,
+      "grad_norm": 0.47725249684852394,
+      "learning_rate": 3.495775407051443e-06,
+      "loss": 0.4388,
+      "step": 17492
+    },
+    {
+      "epoch": 2.87937348273923,
+      "grad_norm": 0.3846884788711646,
+      "learning_rate": 3.495317203049953e-06,
+      "loss": 0.4184,
+      "step": 17493
+    },
+    {
+      "epoch": 2.8795380669583466,
+      "grad_norm": 0.4336049163299836,
+      "learning_rate": 3.494859010011159e-06,
+      "loss": 0.441,
+      "step": 17494
+    },
+    {
+      "epoch": 2.879702651177463,
+      "grad_norm": 0.35240779292563884,
+      "learning_rate": 3.494400827940064e-06,
+      "loss": 0.4257,
+      "step": 17495
+    },
+    {
+      "epoch": 2.8798672353965795,
+      "grad_norm": 0.3353048719207727,
+      "learning_rate": 3.4939426568416664e-06,
+      "loss": 0.4356,
+      "step": 17496
+    },
+    {
+      "epoch": 2.880031819615696,
+      "grad_norm": 0.41796988582902395,
+      "learning_rate": 3.4934844967209633e-06,
+      "loss": 0.42,
+      "step": 17497
+    },
+    {
+      "epoch": 2.8801964038348125,
+      "grad_norm": 0.33536615585729473,
+      "learning_rate": 3.493026347582957e-06,
+      "loss": 0.4173,
+      "step": 17498
+    },
+    {
+      "epoch": 2.880360988053929,
+      "grad_norm": 0.3332212640094779,
+      "learning_rate": 3.4925682094326437e-06,
+      "loss": 0.416,
+      "step": 17499
+    },
+    {
+      "epoch": 2.8805255722730454,
+      "grad_norm": 0.3471955782673057,
+      "learning_rate": 3.492110082275025e-06,
+      "loss": 0.4401,
+      "step": 17500
+    },
+    {
+      "epoch": 2.880690156492162,
+      "grad_norm": 0.3282416402822312,
+      "learning_rate": 3.4916519661150986e-06,
+      "loss": 0.4458,
+      "step": 17501
+    },
+    {
+      "epoch": 2.8808547407112783,
+      "grad_norm": 0.36896467746199585,
+      "learning_rate": 3.4911938609578637e-06,
+      "loss": 0.443,
+      "step": 17502
+    },
+    {
+      "epoch": 2.881019324930395,
+      "grad_norm": 0.3332197037487723,
+      "learning_rate": 3.4907357668083204e-06,
+      "loss": 0.4565,
+      "step": 17503
+    },
+    {
+      "epoch": 2.8811839091495113,
+      "grad_norm": 0.34050078360924096,
+      "learning_rate": 3.490277683671462e-06,
+      "loss": 0.4408,
+      "step": 17504
+    },
+    {
+      "epoch": 2.8813484933686278,
+      "grad_norm": 0.3992178710229831,
+      "learning_rate": 3.489819611552294e-06,
+      "loss": 0.4526,
+      "step": 17505
+    },
+    {
+      "epoch": 2.881513077587744,
+      "grad_norm": 0.2679678739319319,
+      "learning_rate": 3.489361550455808e-06,
+      "loss": 0.4292,
+      "step": 17506
+    },
+    {
+      "epoch": 2.8816776618068607,
+      "grad_norm": 0.3198761892317233,
+      "learning_rate": 3.4889035003870087e-06,
+      "loss": 0.4267,
+      "step": 17507
+    },
+    {
+      "epoch": 2.8818422460259767,
+      "grad_norm": 0.2830674041493437,
+      "learning_rate": 3.4884454613508885e-06,
+      "loss": 0.4211,
+      "step": 17508
+    },
+    {
+      "epoch": 2.882006830245093,
+      "grad_norm": 0.416042072950869,
+      "learning_rate": 3.487987433352449e-06,
+      "loss": 0.4309,
+      "step": 17509
+    },
+    {
+      "epoch": 2.8821714144642097,
+      "grad_norm": 0.516522262405375,
+      "learning_rate": 3.4875294163966868e-06,
+      "loss": 0.4154,
+      "step": 17510
+    },
+    {
+      "epoch": 2.882335998683326,
+      "grad_norm": 0.49139703605895474,
+      "learning_rate": 3.4870714104885977e-06,
+      "loss": 0.4338,
+      "step": 17511
+    },
+    {
+      "epoch": 2.8825005829024426,
+      "grad_norm": 0.2969184023413605,
+      "learning_rate": 3.4866134156331832e-06,
+      "loss": 0.4435,
+      "step": 17512
+    },
+    {
+      "epoch": 2.882665167121559,
+      "grad_norm": 0.3249885720627407,
+      "learning_rate": 3.4861554318354375e-06,
+      "loss": 0.4418,
+      "step": 17513
+    },
+    {
+      "epoch": 2.8828297513406755,
+      "grad_norm": 0.3798585206088055,
+      "learning_rate": 3.485697459100359e-06,
+      "loss": 0.4363,
+      "step": 17514
+    },
+    {
+      "epoch": 2.882994335559792,
+      "grad_norm": 0.2624652605932739,
+      "learning_rate": 3.4852394974329453e-06,
+      "loss": 0.4459,
+      "step": 17515
+    },
+    {
+      "epoch": 2.8831589197789085,
+      "grad_norm": 0.3741199036591475,
+      "learning_rate": 3.484781546838195e-06,
+      "loss": 0.4175,
+      "step": 17516
+    },
+    {
+      "epoch": 2.883323503998025,
+      "grad_norm": 0.28807914592528444,
+      "learning_rate": 3.4843236073210996e-06,
+      "loss": 0.4435,
+      "step": 17517
+    },
+    {
+      "epoch": 2.8834880882171414,
+      "grad_norm": 0.3991709257107399,
+      "learning_rate": 3.483865678886663e-06,
+      "loss": 0.4195,
+      "step": 17518
+    },
+    {
+      "epoch": 2.883652672436258,
+      "grad_norm": 0.2976102140387971,
+      "learning_rate": 3.483407761539878e-06,
+      "loss": 0.4226,
+      "step": 17519
+    },
+    {
+      "epoch": 2.8838172566553744,
+      "grad_norm": 0.29345802221273715,
+      "learning_rate": 3.4829498552857405e-06,
+      "loss": 0.4297,
+      "step": 17520
+    },
+    {
+      "epoch": 2.883981840874491,
+      "grad_norm": 0.3035537774606567,
+      "learning_rate": 3.4824919601292494e-06,
+      "loss": 0.4397,
+      "step": 17521
+    },
+    {
+      "epoch": 2.8841464250936073,
+      "grad_norm": 0.29332745685733963,
+      "learning_rate": 3.4820340760753984e-06,
+      "loss": 0.4377,
+      "step": 17522
+    },
+    {
+      "epoch": 2.8843110093127238,
+      "grad_norm": 0.34426693441979905,
+      "learning_rate": 3.4815762031291865e-06,
+      "loss": 0.4279,
+      "step": 17523
+    },
+    {
+      "epoch": 2.8844755935318402,
+      "grad_norm": 0.3380721839942416,
+      "learning_rate": 3.481118341295608e-06,
+      "loss": 0.4203,
+      "step": 17524
+    },
+    {
+      "epoch": 2.8846401777509567,
+      "grad_norm": 0.28653978493406196,
+      "learning_rate": 3.4806604905796602e-06,
+      "loss": 0.4092,
+      "step": 17525
+    },
+    {
+      "epoch": 2.884804761970073,
+      "grad_norm": 0.4618970250864827,
+      "learning_rate": 3.480202650986339e-06,
+      "loss": 0.4352,
+      "step": 17526
+    },
+    {
+      "epoch": 2.8849693461891897,
+      "grad_norm": 0.7745580166610236,
+      "learning_rate": 3.4797448225206386e-06,
+      "loss": 0.4404,
+      "step": 17527
+    },
+    {
+      "epoch": 2.885133930408306,
+      "grad_norm": 0.3572989624425346,
+      "learning_rate": 3.479287005187556e-06,
+      "loss": 0.4239,
+      "step": 17528
+    },
+    {
+      "epoch": 2.8852985146274226,
+      "grad_norm": 0.3958605014986573,
+      "learning_rate": 3.4788291989920847e-06,
+      "loss": 0.4419,
+      "step": 17529
+    },
+    {
+      "epoch": 2.885463098846539,
+      "grad_norm": 0.29876933676602424,
+      "learning_rate": 3.478371403939223e-06,
+      "loss": 0.4596,
+      "step": 17530
+    },
+    {
+      "epoch": 2.8856276830656555,
+      "grad_norm": 0.30567302667861224,
+      "learning_rate": 3.4779136200339635e-06,
+      "loss": 0.4351,
+      "step": 17531
+    },
+    {
+      "epoch": 2.885792267284772,
+      "grad_norm": 0.7991158892795251,
+      "learning_rate": 3.477455847281304e-06,
+      "loss": 0.4336,
+      "step": 17532
+    },
+    {
+      "epoch": 2.885956851503888,
+      "grad_norm": 0.2860365966641438,
+      "learning_rate": 3.476998085686238e-06,
+      "loss": 0.4166,
+      "step": 17533
+    },
+    {
+      "epoch": 2.8861214357230045,
+      "grad_norm": 0.35813329852227593,
+      "learning_rate": 3.47654033525376e-06,
+      "loss": 0.428,
+      "step": 17534
+    },
+    {
+      "epoch": 2.886286019942121,
+      "grad_norm": 0.340048960958515,
+      "learning_rate": 3.476082595988867e-06,
+      "loss": 0.4089,
+      "step": 17535
+    },
+    {
+      "epoch": 2.8864506041612374,
+      "grad_norm": 0.4793172775125557,
+      "learning_rate": 3.475624867896549e-06,
+      "loss": 0.4241,
+      "step": 17536
+    },
+    {
+      "epoch": 2.886615188380354,
+      "grad_norm": 0.3980190982195709,
+      "learning_rate": 3.475167150981807e-06,
+      "loss": 0.4445,
+      "step": 17537
+    },
+    {
+      "epoch": 2.8867797725994704,
+      "grad_norm": 0.28018392859825864,
+      "learning_rate": 3.4747094452496293e-06,
+      "loss": 0.4342,
+      "step": 17538
+    },
+    {
+      "epoch": 2.886944356818587,
+      "grad_norm": 0.2853823124582563,
+      "learning_rate": 3.474251750705015e-06,
+      "loss": 0.4237,
+      "step": 17539
+    },
+    {
+      "epoch": 2.8871089410377033,
+      "grad_norm": 0.36079531254173375,
+      "learning_rate": 3.473794067352956e-06,
+      "loss": 0.4208,
+      "step": 17540
+    },
+    {
+      "epoch": 2.88727352525682,
+      "grad_norm": 0.37046200754921405,
+      "learning_rate": 3.4733363951984453e-06,
+      "loss": 0.4247,
+      "step": 17541
+    },
+    {
+      "epoch": 2.8874381094759363,
+      "grad_norm": 0.3782141148710571,
+      "learning_rate": 3.472878734246479e-06,
+      "loss": 0.4253,
+      "step": 17542
+    },
+    {
+      "epoch": 2.8876026936950527,
+      "grad_norm": 0.2860690892595623,
+      "learning_rate": 3.472421084502049e-06,
+      "loss": 0.4166,
+      "step": 17543
+    },
+    {
+      "epoch": 2.887767277914169,
+      "grad_norm": 0.2638562180584719,
+      "learning_rate": 3.471963445970151e-06,
+      "loss": 0.4282,
+      "step": 17544
+    },
+    {
+      "epoch": 2.8879318621332857,
+      "grad_norm": 0.3075702618199297,
+      "learning_rate": 3.471505818655777e-06,
+      "loss": 0.4205,
+      "step": 17545
+    },
+    {
+      "epoch": 2.888096446352402,
+      "grad_norm": 0.3551562646617458,
+      "learning_rate": 3.471048202563923e-06,
+      "loss": 0.4259,
+      "step": 17546
+    },
+    {
+      "epoch": 2.8882610305715186,
+      "grad_norm": 0.30878179949248125,
+      "learning_rate": 3.470590597699579e-06,
+      "loss": 0.4219,
+      "step": 17547
+    },
+    {
+      "epoch": 2.888425614790635,
+      "grad_norm": 0.3359741753003401,
+      "learning_rate": 3.4701330040677384e-06,
+      "loss": 0.4353,
+      "step": 17548
+    },
+    {
+      "epoch": 2.8885901990097516,
+      "grad_norm": 1.0444794328678997,
+      "learning_rate": 3.4696754216733972e-06,
+      "loss": 0.4264,
+      "step": 17549
+    },
+    {
+      "epoch": 2.888754783228868,
+      "grad_norm": 0.32698250234672555,
+      "learning_rate": 3.4692178505215448e-06,
+      "loss": 0.4573,
+      "step": 17550
+    },
+    {
+      "epoch": 2.8889193674479845,
+      "grad_norm": 0.27411367522197744,
+      "learning_rate": 3.4687602906171773e-06,
+      "loss": 0.4268,
+      "step": 17551
+    },
+    {
+      "epoch": 2.889083951667101,
+      "grad_norm": 0.33172186293468775,
+      "learning_rate": 3.4683027419652847e-06,
+      "loss": 0.4443,
+      "step": 17552
+    },
+    {
+      "epoch": 2.8892485358862174,
+      "grad_norm": 0.3046771536424512,
+      "learning_rate": 3.4678452045708622e-06,
+      "loss": 0.4141,
+      "step": 17553
+    },
+    {
+      "epoch": 2.889413120105334,
+      "grad_norm": 0.36778006334965835,
+      "learning_rate": 3.4673876784389016e-06,
+      "loss": 0.4505,
+      "step": 17554
+    },
+    {
+      "epoch": 2.8895777043244504,
+      "grad_norm": 0.49063857055608223,
+      "learning_rate": 3.4669301635743913e-06,
+      "loss": 0.4283,
+      "step": 17555
+    },
+    {
+      "epoch": 2.889742288543567,
+      "grad_norm": 0.394256688057511,
+      "learning_rate": 3.46647265998233e-06,
+      "loss": 0.4507,
+      "step": 17556
+    },
+    {
+      "epoch": 2.8899068727626833,
+      "grad_norm": 0.30983342194561003,
+      "learning_rate": 3.4660151676677034e-06,
+      "loss": 0.4348,
+      "step": 17557
+    },
+    {
+      "epoch": 2.8900714569818,
+      "grad_norm": 0.30241903659440256,
+      "learning_rate": 3.4655576866355097e-06,
+      "loss": 0.4417,
+      "step": 17558
+    },
+    {
+      "epoch": 2.8902360412009163,
+      "grad_norm": 0.3153832356377743,
+      "learning_rate": 3.465100216890736e-06,
+      "loss": 0.4471,
+      "step": 17559
+    },
+    {
+      "epoch": 2.8904006254200327,
+      "grad_norm": 0.40039286031331656,
+      "learning_rate": 3.4646427584383765e-06,
+      "loss": 0.4565,
+      "step": 17560
+    },
+    {
+      "epoch": 2.890565209639149,
+      "grad_norm": 0.3497720118990314,
+      "learning_rate": 3.4641853112834214e-06,
+      "loss": 0.46,
+      "step": 17561
+    },
+    {
+      "epoch": 2.8907297938582657,
+      "grad_norm": 0.3215586088161052,
+      "learning_rate": 3.463727875430863e-06,
+      "loss": 0.4388,
+      "step": 17562
+    },
+    {
+      "epoch": 2.890894378077382,
+      "grad_norm": 0.2806799404627749,
+      "learning_rate": 3.463270450885693e-06,
+      "loss": 0.4062,
+      "step": 17563
+    },
+    {
+      "epoch": 2.8910589622964986,
+      "grad_norm": 0.30590924071024106,
+      "learning_rate": 3.462813037652901e-06,
+      "loss": 0.4415,
+      "step": 17564
+    },
+    {
+      "epoch": 2.891223546515615,
+      "grad_norm": 0.39795707215124965,
+      "learning_rate": 3.462355635737481e-06,
+      "loss": 0.4133,
+      "step": 17565
+    },
+    {
+      "epoch": 2.8913881307347316,
+      "grad_norm": 0.28841486979329806,
+      "learning_rate": 3.4618982451444197e-06,
+      "loss": 0.4262,
+      "step": 17566
+    },
+    {
+      "epoch": 2.891552714953848,
+      "grad_norm": 0.30530892538511867,
+      "learning_rate": 3.4614408658787127e-06,
+      "loss": 0.4272,
+      "step": 17567
+    },
+    {
+      "epoch": 2.8917172991729645,
+      "grad_norm": 0.6587955945287981,
+      "learning_rate": 3.4609834979453457e-06,
+      "loss": 0.4235,
+      "step": 17568
+    },
+    {
+      "epoch": 2.891881883392081,
+      "grad_norm": 0.3581837116766279,
+      "learning_rate": 3.4605261413493155e-06,
+      "loss": 0.4399,
+      "step": 17569
+    },
+    {
+      "epoch": 2.8920464676111974,
+      "grad_norm": 0.30612813779363107,
+      "learning_rate": 3.4600687960956076e-06,
+      "loss": 0.4445,
+      "step": 17570
+    },
+    {
+      "epoch": 2.892211051830314,
+      "grad_norm": 0.31405506424736673,
+      "learning_rate": 3.4596114621892133e-06,
+      "loss": 0.4457,
+      "step": 17571
+    },
+    {
+      "epoch": 2.8923756360494304,
+      "grad_norm": 0.2900030281499961,
+      "learning_rate": 3.4591541396351244e-06,
+      "loss": 0.4448,
+      "step": 17572
+    },
+    {
+      "epoch": 2.892540220268547,
+      "grad_norm": 0.28487788863551433,
+      "learning_rate": 3.4586968284383298e-06,
+      "loss": 0.4158,
+      "step": 17573
+    },
+    {
+      "epoch": 2.8927048044876633,
+      "grad_norm": 0.27083017128369313,
+      "learning_rate": 3.4582395286038197e-06,
+      "loss": 0.4343,
+      "step": 17574
+    },
+    {
+      "epoch": 2.8928693887067793,
+      "grad_norm": 0.326629990273933,
+      "learning_rate": 3.4577822401365838e-06,
+      "loss": 0.4274,
+      "step": 17575
+    },
+    {
+      "epoch": 2.893033972925896,
+      "grad_norm": 0.3504863651738228,
+      "learning_rate": 3.457324963041613e-06,
+      "loss": 0.4321,
+      "step": 17576
+    },
+    {
+      "epoch": 2.8931985571450123,
+      "grad_norm": 0.4574549399371932,
+      "learning_rate": 3.4568676973238973e-06,
+      "loss": 0.4291,
+      "step": 17577
+    },
+    {
+      "epoch": 2.8933631413641288,
+      "grad_norm": 0.3123392716219587,
+      "learning_rate": 3.4564104429884223e-06,
+      "loss": 0.415,
+      "step": 17578
+    },
+    {
+      "epoch": 2.8935277255832452,
+      "grad_norm": 0.3238830641887869,
+      "learning_rate": 3.4559532000401816e-06,
+      "loss": 0.4326,
+      "step": 17579
+    },
+    {
+      "epoch": 2.8936923098023617,
+      "grad_norm": 0.2935385519404049,
+      "learning_rate": 3.455495968484162e-06,
+      "loss": 0.4331,
+      "step": 17580
+    },
+    {
+      "epoch": 2.893856894021478,
+      "grad_norm": 0.33603634738461574,
+      "learning_rate": 3.4550387483253537e-06,
+      "loss": 0.4509,
+      "step": 17581
+    },
+    {
+      "epoch": 2.8940214782405946,
+      "grad_norm": 0.27276667451662073,
+      "learning_rate": 3.4545815395687453e-06,
+      "loss": 0.4247,
+      "step": 17582
+    },
+    {
+      "epoch": 2.894186062459711,
+      "grad_norm": 0.27054568096228415,
+      "learning_rate": 3.454124342219327e-06,
+      "loss": 0.4311,
+      "step": 17583
+    },
+    {
+      "epoch": 2.8943506466788276,
+      "grad_norm": 0.3317891914521292,
+      "learning_rate": 3.453667156282087e-06,
+      "loss": 0.4182,
+      "step": 17584
+    },
+    {
+      "epoch": 2.894515230897944,
+      "grad_norm": 0.29678551621684296,
+      "learning_rate": 3.4532099817620107e-06,
+      "loss": 0.4264,
+      "step": 17585
+    },
+    {
+      "epoch": 2.8946798151170605,
+      "grad_norm": 0.40450783480095226,
+      "learning_rate": 3.452752818664092e-06,
+      "loss": 0.4304,
+      "step": 17586
+    },
+    {
+      "epoch": 2.894844399336177,
+      "grad_norm": 0.2833720786401857,
+      "learning_rate": 3.452295666993314e-06,
+      "loss": 0.435,
+      "step": 17587
+    },
+    {
+      "epoch": 2.8950089835552935,
+      "grad_norm": 0.3042480121931492,
+      "learning_rate": 3.4518385267546703e-06,
+      "loss": 0.4325,
+      "step": 17588
+    },
+    {
+      "epoch": 2.89517356777441,
+      "grad_norm": 0.27890726007985217,
+      "learning_rate": 3.451381397953144e-06,
+      "loss": 0.4472,
+      "step": 17589
+    },
+    {
+      "epoch": 2.8953381519935264,
+      "grad_norm": 0.354090550441332,
+      "learning_rate": 3.450924280593727e-06,
+      "loss": 0.4144,
+      "step": 17590
+    },
+    {
+      "epoch": 2.895502736212643,
+      "grad_norm": 0.3311840660126875,
+      "learning_rate": 3.4504671746814054e-06,
+      "loss": 0.4328,
+      "step": 17591
+    },
+    {
+      "epoch": 2.8956673204317593,
+      "grad_norm": 0.3337723131257527,
+      "learning_rate": 3.450010080221166e-06,
+      "loss": 0.4356,
+      "step": 17592
+    },
+    {
+      "epoch": 2.895831904650876,
+      "grad_norm": 0.3563506569008164,
+      "learning_rate": 3.449552997217999e-06,
+      "loss": 0.4425,
+      "step": 17593
+    },
+    {
+      "epoch": 2.8959964888699923,
+      "grad_norm": 0.29006337535950466,
+      "learning_rate": 3.4490959256768887e-06,
+      "loss": 0.4301,
+      "step": 17594
+    },
+    {
+      "epoch": 2.8961610730891088,
+      "grad_norm": 0.4306552211887624,
+      "learning_rate": 3.448638865602825e-06,
+      "loss": 0.4371,
+      "step": 17595
+    },
+    {
+      "epoch": 2.896325657308225,
+      "grad_norm": 0.37960210912353,
+      "learning_rate": 3.4481818170007944e-06,
+      "loss": 0.4201,
+      "step": 17596
+    },
+    {
+      "epoch": 2.8964902415273417,
+      "grad_norm": 0.2869460692042597,
+      "learning_rate": 3.4477247798757857e-06,
+      "loss": 0.4372,
+      "step": 17597
+    },
+    {
+      "epoch": 2.896654825746458,
+      "grad_norm": 0.332782372252181,
+      "learning_rate": 3.447267754232781e-06,
+      "loss": 0.4449,
+      "step": 17598
+    },
+    {
+      "epoch": 2.8968194099655746,
+      "grad_norm": 0.41277560059229834,
+      "learning_rate": 3.446810740076774e-06,
+      "loss": 0.436,
+      "step": 17599
+    },
+    {
+      "epoch": 2.8969839941846907,
+      "grad_norm": 0.3226705898444896,
+      "learning_rate": 3.446353737412746e-06,
+      "loss": 0.4327,
+      "step": 17600
+    },
+    {
+      "epoch": 2.897148578403807,
+      "grad_norm": 0.3131374708035312,
+      "learning_rate": 3.445896746245685e-06,
+      "loss": 0.4289,
+      "step": 17601
+    },
+    {
+      "epoch": 2.8973131626229236,
+      "grad_norm": 0.38371779238723275,
+      "learning_rate": 3.4454397665805785e-06,
+      "loss": 0.4148,
+      "step": 17602
+    },
+    {
+      "epoch": 2.89747774684204,
+      "grad_norm": 0.3992085554809812,
+      "learning_rate": 3.444982798422412e-06,
+      "loss": 0.4241,
+      "step": 17603
+    },
+    {
+      "epoch": 2.8976423310611565,
+      "grad_norm": 0.27723506544946885,
+      "learning_rate": 3.4445258417761734e-06,
+      "loss": 0.4323,
+      "step": 17604
+    },
+    {
+      "epoch": 2.897806915280273,
+      "grad_norm": 0.47666218029781204,
+      "learning_rate": 3.444068896646846e-06,
+      "loss": 0.4318,
+      "step": 17605
+    },
+    {
+      "epoch": 2.8979714994993895,
+      "grad_norm": 0.30510120682304553,
+      "learning_rate": 3.4436119630394186e-06,
+      "loss": 0.4209,
+      "step": 17606
+    },
+    {
+      "epoch": 2.898136083718506,
+      "grad_norm": 0.4314918849748407,
+      "learning_rate": 3.4431550409588767e-06,
+      "loss": 0.4047,
+      "step": 17607
+    },
+    {
+      "epoch": 2.8983006679376224,
+      "grad_norm": 0.35357462801580614,
+      "learning_rate": 3.442698130410203e-06,
+      "loss": 0.4384,
+      "step": 17608
+    },
+    {
+      "epoch": 2.898465252156739,
+      "grad_norm": 0.30285283816494646,
+      "learning_rate": 3.4422412313983867e-06,
+      "loss": 0.4402,
+      "step": 17609
+    },
+    {
+      "epoch": 2.8986298363758554,
+      "grad_norm": 0.36947669740277606,
+      "learning_rate": 3.441784343928411e-06,
+      "loss": 0.4324,
+      "step": 17610
+    },
+    {
+      "epoch": 2.898794420594972,
+      "grad_norm": 0.3775632311292628,
+      "learning_rate": 3.441327468005263e-06,
+      "loss": 0.4165,
+      "step": 17611
+    },
+    {
+      "epoch": 2.8989590048140883,
+      "grad_norm": 0.34658287083305633,
+      "learning_rate": 3.4408706036339264e-06,
+      "loss": 0.4354,
+      "step": 17612
+    },
+    {
+      "epoch": 2.8991235890332048,
+      "grad_norm": 0.35313998235139205,
+      "learning_rate": 3.440413750819388e-06,
+      "loss": 0.4429,
+      "step": 17613
+    },
+    {
+      "epoch": 2.8992881732523212,
+      "grad_norm": 0.2952482822347777,
+      "learning_rate": 3.439956909566632e-06,
+      "loss": 0.4444,
+      "step": 17614
+    },
+    {
+      "epoch": 2.8994527574714377,
+      "grad_norm": 0.42786781905398574,
+      "learning_rate": 3.4395000798806418e-06,
+      "loss": 0.4234,
+      "step": 17615
+    },
+    {
+      "epoch": 2.899617341690554,
+      "grad_norm": 0.3311169314032787,
+      "learning_rate": 3.439043261766405e-06,
+      "loss": 0.416,
+      "step": 17616
+    },
+    {
+      "epoch": 2.8997819259096707,
+      "grad_norm": 0.33214432687721374,
+      "learning_rate": 3.4385864552289023e-06,
+      "loss": 0.4295,
+      "step": 17617
+    },
+    {
+      "epoch": 2.899946510128787,
+      "grad_norm": 0.469509524189476,
+      "learning_rate": 3.438129660273124e-06,
+      "loss": 0.4452,
+      "step": 17618
+    },
+    {
+      "epoch": 2.9001110943479036,
+      "grad_norm": 0.3206636648928445,
+      "learning_rate": 3.4376728769040476e-06,
+      "loss": 0.4426,
+      "step": 17619
+    },
+    {
+      "epoch": 2.90027567856702,
+      "grad_norm": 0.37333838540874276,
+      "learning_rate": 3.437216105126664e-06,
+      "loss": 0.4281,
+      "step": 17620
+    },
+    {
+      "epoch": 2.9004402627861365,
+      "grad_norm": 0.340705525024164,
+      "learning_rate": 3.4367593449459526e-06,
+      "loss": 0.4219,
+      "step": 17621
+    },
+    {
+      "epoch": 2.900604847005253,
+      "grad_norm": 0.3381100580160605,
+      "learning_rate": 3.4363025963668984e-06,
+      "loss": 0.4294,
+      "step": 17622
+    },
+    {
+      "epoch": 2.9007694312243695,
+      "grad_norm": 0.28531483631558896,
+      "learning_rate": 3.4358458593944868e-06,
+      "loss": 0.4275,
+      "step": 17623
+    },
+    {
+      "epoch": 2.900934015443486,
+      "grad_norm": 0.27056745382096664,
+      "learning_rate": 3.4353891340336987e-06,
+      "loss": 0.4127,
+      "step": 17624
+    },
+    {
+      "epoch": 2.9010985996626024,
+      "grad_norm": 0.29242580050975925,
+      "learning_rate": 3.434932420289521e-06,
+      "loss": 0.4216,
+      "step": 17625
+    },
+    {
+      "epoch": 2.901263183881719,
+      "grad_norm": 0.39910999046507917,
+      "learning_rate": 3.434475718166935e-06,
+      "loss": 0.4253,
+      "step": 17626
+    },
+    {
+      "epoch": 2.9014277681008354,
+      "grad_norm": 0.32521720295621315,
+      "learning_rate": 3.4340190276709265e-06,
+      "loss": 0.4128,
+      "step": 17627
+    },
+    {
+      "epoch": 2.901592352319952,
+      "grad_norm": 0.36843049245490517,
+      "learning_rate": 3.433562348806475e-06,
+      "loss": 0.4299,
+      "step": 17628
+    },
+    {
+      "epoch": 2.9017569365390683,
+      "grad_norm": 0.614088766386188,
+      "learning_rate": 3.4331056815785662e-06,
+      "loss": 0.4194,
+      "step": 17629
+    },
+    {
+      "epoch": 2.9019215207581848,
+      "grad_norm": 0.34208998169652544,
+      "learning_rate": 3.432649025992183e-06,
+      "loss": 0.4416,
+      "step": 17630
+    },
+    {
+      "epoch": 2.9020861049773012,
+      "grad_norm": 0.41863096666227234,
+      "learning_rate": 3.4321923820523056e-06,
+      "loss": 0.4309,
+      "step": 17631
+    },
+    {
+      "epoch": 2.9022506891964177,
+      "grad_norm": 0.3387098617077992,
+      "learning_rate": 3.4317357497639205e-06,
+      "loss": 0.4542,
+      "step": 17632
+    },
+    {
+      "epoch": 2.902415273415534,
+      "grad_norm": 0.3880538504682831,
+      "learning_rate": 3.431279129132007e-06,
+      "loss": 0.4242,
+      "step": 17633
+    },
+    {
+      "epoch": 2.9025798576346507,
+      "grad_norm": 0.29517302631006787,
+      "learning_rate": 3.430822520161551e-06,
+      "loss": 0.4431,
+      "step": 17634
+    },
+    {
+      "epoch": 2.902744441853767,
+      "grad_norm": 0.3217759709408525,
+      "learning_rate": 3.4303659228575337e-06,
+      "loss": 0.4526,
+      "step": 17635
+    },
+    {
+      "epoch": 2.9029090260728836,
+      "grad_norm": 0.5737184089042462,
+      "learning_rate": 3.4299093372249333e-06,
+      "loss": 0.4197,
+      "step": 17636
+    },
+    {
+      "epoch": 2.903073610292,
+      "grad_norm": 0.38688542526269387,
+      "learning_rate": 3.429452763268738e-06,
+      "loss": 0.4194,
+      "step": 17637
+    },
+    {
+      "epoch": 2.9032381945111165,
+      "grad_norm": 0.3002994499825206,
+      "learning_rate": 3.4289962009939244e-06,
+      "loss": 0.4142,
+      "step": 17638
+    },
+    {
+      "epoch": 2.903402778730233,
+      "grad_norm": 0.35474702447463446,
+      "learning_rate": 3.4285396504054792e-06,
+      "loss": 0.4275,
+      "step": 17639
+    },
+    {
+      "epoch": 2.9035673629493495,
+      "grad_norm": 0.2754523365816297,
+      "learning_rate": 3.42808311150838e-06,
+      "loss": 0.4247,
+      "step": 17640
+    },
+    {
+      "epoch": 2.9037319471684655,
+      "grad_norm": 0.30124553560901396,
+      "learning_rate": 3.4276265843076113e-06,
+      "loss": 0.4152,
+      "step": 17641
+    },
+    {
+      "epoch": 2.903896531387582,
+      "grad_norm": 0.4355278989267508,
+      "learning_rate": 3.4271700688081523e-06,
+      "loss": 0.4205,
+      "step": 17642
+    },
+    {
+      "epoch": 2.9040611156066984,
+      "grad_norm": 0.2938263585504233,
+      "learning_rate": 3.4267135650149863e-06,
+      "loss": 0.4264,
+      "step": 17643
+    },
+    {
+      "epoch": 2.904225699825815,
+      "grad_norm": 0.30681761907619126,
+      "learning_rate": 3.4262570729330934e-06,
+      "loss": 0.4337,
+      "step": 17644
+    },
+    {
+      "epoch": 2.9043902840449314,
+      "grad_norm": 0.387173245592773,
+      "learning_rate": 3.425800592567454e-06,
+      "loss": 0.4488,
+      "step": 17645
+    },
+    {
+      "epoch": 2.904554868264048,
+      "grad_norm": 0.3232799699547336,
+      "learning_rate": 3.42534412392305e-06,
+      "loss": 0.4482,
+      "step": 17646
+    },
+    {
+      "epoch": 2.9047194524831643,
+      "grad_norm": 0.27505508379676474,
+      "learning_rate": 3.4248876670048623e-06,
+      "loss": 0.4503,
+      "step": 17647
+    },
+    {
+      "epoch": 2.904884036702281,
+      "grad_norm": 0.3042903382131154,
+      "learning_rate": 3.424431221817872e-06,
+      "loss": 0.4219,
+      "step": 17648
+    },
+    {
+      "epoch": 2.9050486209213973,
+      "grad_norm": 0.40167340114884303,
+      "learning_rate": 3.423974788367057e-06,
+      "loss": 0.4319,
+      "step": 17649
+    },
+    {
+      "epoch": 2.9052132051405137,
+      "grad_norm": 0.40027143285984046,
+      "learning_rate": 3.423518366657402e-06,
+      "loss": 0.4273,
+      "step": 17650
+    },
+    {
+      "epoch": 2.90537778935963,
+      "grad_norm": 0.38083677010822076,
+      "learning_rate": 3.4230619566938843e-06,
+      "loss": 0.4543,
+      "step": 17651
+    },
+    {
+      "epoch": 2.9055423735787467,
+      "grad_norm": 0.3445738566986874,
+      "learning_rate": 3.4226055584814837e-06,
+      "loss": 0.4232,
+      "step": 17652
+    },
+    {
+      "epoch": 2.905706957797863,
+      "grad_norm": 0.2830924066386983,
+      "learning_rate": 3.4221491720251826e-06,
+      "loss": 0.4173,
+      "step": 17653
+    },
+    {
+      "epoch": 2.9058715420169796,
+      "grad_norm": 0.4947656362937295,
+      "learning_rate": 3.4216927973299577e-06,
+      "loss": 0.4471,
+      "step": 17654
+    },
+    {
+      "epoch": 2.906036126236096,
+      "grad_norm": 0.2769119444352562,
+      "learning_rate": 3.4212364344007925e-06,
+      "loss": 0.4312,
+      "step": 17655
+    },
+    {
+      "epoch": 2.9062007104552126,
+      "grad_norm": 0.6499934734448956,
+      "learning_rate": 3.420780083242663e-06,
+      "loss": 0.4367,
+      "step": 17656
+    },
+    {
+      "epoch": 2.906365294674329,
+      "grad_norm": 0.370733265460432,
+      "learning_rate": 3.4203237438605524e-06,
+      "loss": 0.431,
+      "step": 17657
+    },
+    {
+      "epoch": 2.9065298788934455,
+      "grad_norm": 0.3411392926466954,
+      "learning_rate": 3.4198674162594386e-06,
+      "loss": 0.4339,
+      "step": 17658
+    },
+    {
+      "epoch": 2.906694463112562,
+      "grad_norm": 0.39875733082792414,
+      "learning_rate": 3.4194111004442983e-06,
+      "loss": 0.4408,
+      "step": 17659
+    },
+    {
+      "epoch": 2.9068590473316784,
+      "grad_norm": 0.39200771850558425,
+      "learning_rate": 3.4189547964201145e-06,
+      "loss": 0.428,
+      "step": 17660
+    },
+    {
+      "epoch": 2.907023631550795,
+      "grad_norm": 0.2871409612914079,
+      "learning_rate": 3.4184985041918627e-06,
+      "loss": 0.42,
+      "step": 17661
+    },
+    {
+      "epoch": 2.9071882157699114,
+      "grad_norm": 0.3061717105592732,
+      "learning_rate": 3.4180422237645253e-06,
+      "loss": 0.4212,
+      "step": 17662
+    },
+    {
+      "epoch": 2.907352799989028,
+      "grad_norm": 0.2843559900515243,
+      "learning_rate": 3.417585955143078e-06,
+      "loss": 0.4528,
+      "step": 17663
+    },
+    {
+      "epoch": 2.9075173842081443,
+      "grad_norm": 0.3243487704921083,
+      "learning_rate": 3.417129698332502e-06,
+      "loss": 0.4382,
+      "step": 17664
+    },
+    {
+      "epoch": 2.907681968427261,
+      "grad_norm": 0.2980024565274041,
+      "learning_rate": 3.4166734533377747e-06,
+      "loss": 0.4441,
+      "step": 17665
+    },
+    {
+      "epoch": 2.907846552646377,
+      "grad_norm": 0.34459119960972434,
+      "learning_rate": 3.416217220163873e-06,
+      "loss": 0.4292,
+      "step": 17666
+    },
+    {
+      "epoch": 2.9080111368654933,
+      "grad_norm": 1.0791551102000445,
+      "learning_rate": 3.4157609988157784e-06,
+      "loss": 0.4346,
+      "step": 17667
+    },
+    {
+      "epoch": 2.9081757210846098,
+      "grad_norm": 0.6016048549628633,
+      "learning_rate": 3.415304789298464e-06,
+      "loss": 0.4227,
+      "step": 17668
+    },
+    {
+      "epoch": 2.9083403053037262,
+      "grad_norm": 0.43372572493792233,
+      "learning_rate": 3.414848591616914e-06,
+      "loss": 0.4334,
+      "step": 17669
+    },
+    {
+      "epoch": 2.9085048895228427,
+      "grad_norm": 0.3356910552899443,
+      "learning_rate": 3.4143924057761e-06,
+      "loss": 0.4142,
+      "step": 17670
+    },
+    {
+      "epoch": 2.908669473741959,
+      "grad_norm": 0.3306805036575712,
+      "learning_rate": 3.413936231781006e-06,
+      "loss": 0.4217,
+      "step": 17671
+    },
+    {
+      "epoch": 2.9088340579610756,
+      "grad_norm": 0.30701960353201796,
+      "learning_rate": 3.4134800696366054e-06,
+      "loss": 0.4342,
+      "step": 17672
+    },
+    {
+      "epoch": 2.908998642180192,
+      "grad_norm": 0.29094351464733453,
+      "learning_rate": 3.4130239193478754e-06,
+      "loss": 0.4628,
+      "step": 17673
+    },
+    {
+      "epoch": 2.9091632263993086,
+      "grad_norm": 0.3301904863774064,
+      "learning_rate": 3.412567780919796e-06,
+      "loss": 0.4084,
+      "step": 17674
+    },
+    {
+      "epoch": 2.909327810618425,
+      "grad_norm": 0.3566565100217812,
+      "learning_rate": 3.4121116543573416e-06,
+      "loss": 0.4111,
+      "step": 17675
+    },
+    {
+      "epoch": 2.9094923948375415,
+      "grad_norm": 0.266684336790093,
+      "learning_rate": 3.411655539665492e-06,
+      "loss": 0.4072,
+      "step": 17676
+    },
+    {
+      "epoch": 2.909656979056658,
+      "grad_norm": 0.29303718820543445,
+      "learning_rate": 3.411199436849221e-06,
+      "loss": 0.4292,
+      "step": 17677
+    },
+    {
+      "epoch": 2.9098215632757745,
+      "grad_norm": 0.32859176300301063,
+      "learning_rate": 3.4107433459135096e-06,
+      "loss": 0.422,
+      "step": 17678
+    },
+    {
+      "epoch": 2.909986147494891,
+      "grad_norm": 0.302610520053293,
+      "learning_rate": 3.4102872668633315e-06,
+      "loss": 0.4267,
+      "step": 17679
+    },
+    {
+      "epoch": 2.9101507317140074,
+      "grad_norm": 0.3279636996325862,
+      "learning_rate": 3.409831199703662e-06,
+      "loss": 0.431,
+      "step": 17680
+    },
+    {
+      "epoch": 2.910315315933124,
+      "grad_norm": 0.33009465835008517,
+      "learning_rate": 3.4093751444394813e-06,
+      "loss": 0.4268,
+      "step": 17681
+    },
+    {
+      "epoch": 2.9104799001522403,
+      "grad_norm": 0.366280287216141,
+      "learning_rate": 3.4089191010757623e-06,
+      "loss": 0.4109,
+      "step": 17682
+    },
+    {
+      "epoch": 2.910644484371357,
+      "grad_norm": 0.39484965699830255,
+      "learning_rate": 3.408463069617484e-06,
+      "loss": 0.4315,
+      "step": 17683
+    },
+    {
+      "epoch": 2.9108090685904733,
+      "grad_norm": 0.3019101335583522,
+      "learning_rate": 3.4080070500696195e-06,
+      "loss": 0.4227,
+      "step": 17684
+    },
+    {
+      "epoch": 2.9109736528095898,
+      "grad_norm": 0.4397943867074195,
+      "learning_rate": 3.4075510424371476e-06,
+      "loss": 0.4305,
+      "step": 17685
+    },
+    {
+      "epoch": 2.911138237028706,
+      "grad_norm": 0.3450226086647439,
+      "learning_rate": 3.407095046725042e-06,
+      "loss": 0.4217,
+      "step": 17686
+    },
+    {
+      "epoch": 2.9113028212478227,
+      "grad_norm": 0.5353065765995956,
+      "learning_rate": 3.4066390629382793e-06,
+      "loss": 0.4498,
+      "step": 17687
+    },
+    {
+      "epoch": 2.911467405466939,
+      "grad_norm": 0.37676124154671603,
+      "learning_rate": 3.406183091081836e-06,
+      "loss": 0.4175,
+      "step": 17688
+    },
+    {
+      "epoch": 2.9116319896860556,
+      "grad_norm": 0.37203745809563854,
+      "learning_rate": 3.405727131160684e-06,
+      "loss": 0.4295,
+      "step": 17689
+    },
+    {
+      "epoch": 2.911796573905172,
+      "grad_norm": 0.34226933420973427,
+      "learning_rate": 3.405271183179803e-06,
+      "loss": 0.4425,
+      "step": 17690
+    },
+    {
+      "epoch": 2.9119611581242886,
+      "grad_norm": 0.38383361264212845,
+      "learning_rate": 3.404815247144164e-06,
+      "loss": 0.43,
+      "step": 17691
+    },
+    {
+      "epoch": 2.912125742343405,
+      "grad_norm": 0.2692537578894957,
+      "learning_rate": 3.4043593230587454e-06,
+      "loss": 0.4179,
+      "step": 17692
+    },
+    {
+      "epoch": 2.9122903265625215,
+      "grad_norm": 0.38348850251716265,
+      "learning_rate": 3.4039034109285192e-06,
+      "loss": 0.4272,
+      "step": 17693
+    },
+    {
+      "epoch": 2.912454910781638,
+      "grad_norm": 0.2814115397311186,
+      "learning_rate": 3.4034475107584626e-06,
+      "loss": 0.4167,
+      "step": 17694
+    },
+    {
+      "epoch": 2.9126194950007545,
+      "grad_norm": 0.3396851526485684,
+      "learning_rate": 3.4029916225535498e-06,
+      "loss": 0.4293,
+      "step": 17695
+    },
+    {
+      "epoch": 2.912784079219871,
+      "grad_norm": 0.3057095780168015,
+      "learning_rate": 3.4025357463187524e-06,
+      "loss": 0.4347,
+      "step": 17696
+    },
+    {
+      "epoch": 2.9129486634389874,
+      "grad_norm": 0.5856334016361633,
+      "learning_rate": 3.4020798820590497e-06,
+      "loss": 0.4328,
+      "step": 17697
+    },
+    {
+      "epoch": 2.913113247658104,
+      "grad_norm": 0.3165345726469618,
+      "learning_rate": 3.4016240297794105e-06,
+      "loss": 0.4376,
+      "step": 17698
+    },
+    {
+      "epoch": 2.9132778318772203,
+      "grad_norm": 0.26790571564430643,
+      "learning_rate": 3.401168189484814e-06,
+      "loss": 0.4139,
+      "step": 17699
+    },
+    {
+      "epoch": 2.913442416096337,
+      "grad_norm": 0.35081641582732853,
+      "learning_rate": 3.4007123611802284e-06,
+      "loss": 0.4417,
+      "step": 17700
+    },
+    {
+      "epoch": 2.9136070003154533,
+      "grad_norm": 0.4507318909432953,
+      "learning_rate": 3.400256544870634e-06,
+      "loss": 0.4251,
+      "step": 17701
+    },
+    {
+      "epoch": 2.9137715845345697,
+      "grad_norm": 0.29400161745694303,
+      "learning_rate": 3.3998007405610003e-06,
+      "loss": 0.446,
+      "step": 17702
+    },
+    {
+      "epoch": 2.913936168753686,
+      "grad_norm": 0.3430924489995087,
+      "learning_rate": 3.399344948256301e-06,
+      "loss": 0.426,
+      "step": 17703
+    },
+    {
+      "epoch": 2.9141007529728027,
+      "grad_norm": 0.3507783651897055,
+      "learning_rate": 3.398889167961511e-06,
+      "loss": 0.425,
+      "step": 17704
+    },
+    {
+      "epoch": 2.914265337191919,
+      "grad_norm": 0.43556437810017934,
+      "learning_rate": 3.3984333996816024e-06,
+      "loss": 0.4478,
+      "step": 17705
+    },
+    {
+      "epoch": 2.9144299214110356,
+      "grad_norm": 0.3252357809987994,
+      "learning_rate": 3.39797764342155e-06,
+      "loss": 0.4286,
+      "step": 17706
+    },
+    {
+      "epoch": 2.914594505630152,
+      "grad_norm": 0.3040779954874744,
+      "learning_rate": 3.397521899186324e-06,
+      "loss": 0.4188,
+      "step": 17707
+    },
+    {
+      "epoch": 2.914759089849268,
+      "grad_norm": 0.34531608497152533,
+      "learning_rate": 3.3970661669809005e-06,
+      "loss": 0.4204,
+      "step": 17708
+    },
+    {
+      "epoch": 2.9149236740683846,
+      "grad_norm": 0.38340548767782345,
+      "learning_rate": 3.396610446810252e-06,
+      "loss": 0.4526,
+      "step": 17709
+    },
+    {
+      "epoch": 2.915088258287501,
+      "grad_norm": 0.31692373966980675,
+      "learning_rate": 3.3961547386793476e-06,
+      "loss": 0.4431,
+      "step": 17710
+    },
+    {
+      "epoch": 2.9152528425066175,
+      "grad_norm": 0.2927901325514057,
+      "learning_rate": 3.3956990425931637e-06,
+      "loss": 0.4157,
+      "step": 17711
+    },
+    {
+      "epoch": 2.915417426725734,
+      "grad_norm": 0.3381982550939023,
+      "learning_rate": 3.3952433585566697e-06,
+      "loss": 0.4346,
+      "step": 17712
+    },
+    {
+      "epoch": 2.9155820109448505,
+      "grad_norm": 0.3908240460861359,
+      "learning_rate": 3.39478768657484e-06,
+      "loss": 0.4323,
+      "step": 17713
+    },
+    {
+      "epoch": 2.915746595163967,
+      "grad_norm": 0.35719831437371163,
+      "learning_rate": 3.394332026652645e-06,
+      "loss": 0.4052,
+      "step": 17714
+    },
+    {
+      "epoch": 2.9159111793830834,
+      "grad_norm": 0.30052725251285,
+      "learning_rate": 3.393876378795059e-06,
+      "loss": 0.4373,
+      "step": 17715
+    },
+    {
+      "epoch": 2.9160757636022,
+      "grad_norm": 0.3120089814016768,
+      "learning_rate": 3.393420743007053e-06,
+      "loss": 0.433,
+      "step": 17716
+    },
+    {
+      "epoch": 2.9162403478213164,
+      "grad_norm": 0.48157706619602825,
+      "learning_rate": 3.3929651192935958e-06,
+      "loss": 0.4404,
+      "step": 17717
+    },
+    {
+      "epoch": 2.916404932040433,
+      "grad_norm": 0.34109414817190065,
+      "learning_rate": 3.3925095076596646e-06,
+      "loss": 0.4327,
+      "step": 17718
+    },
+    {
+      "epoch": 2.9165695162595493,
+      "grad_norm": 0.293311088739958,
+      "learning_rate": 3.392053908110224e-06,
+      "loss": 0.4034,
+      "step": 17719
+    },
+    {
+      "epoch": 2.9167341004786658,
+      "grad_norm": 0.3200744243515672,
+      "learning_rate": 3.391598320650252e-06,
+      "loss": 0.4451,
+      "step": 17720
+    },
+    {
+      "epoch": 2.9168986846977822,
+      "grad_norm": 0.2798542360586772,
+      "learning_rate": 3.3911427452847153e-06,
+      "loss": 0.4271,
+      "step": 17721
+    },
+    {
+      "epoch": 2.9170632689168987,
+      "grad_norm": 0.4470121931155287,
+      "learning_rate": 3.390687182018587e-06,
+      "loss": 0.441,
+      "step": 17722
+    },
+    {
+      "epoch": 2.917227853136015,
+      "grad_norm": 0.3077008837417923,
+      "learning_rate": 3.3902316308568364e-06,
+      "loss": 0.4371,
+      "step": 17723
+    },
+    {
+      "epoch": 2.9173924373551317,
+      "grad_norm": 0.3355025362900558,
+      "learning_rate": 3.389776091804437e-06,
+      "loss": 0.4223,
+      "step": 17724
+    },
+    {
+      "epoch": 2.917557021574248,
+      "grad_norm": 0.30686200898920396,
+      "learning_rate": 3.389320564866357e-06,
+      "loss": 0.4231,
+      "step": 17725
+    },
+    {
+      "epoch": 2.9177216057933646,
+      "grad_norm": 0.368801143123311,
+      "learning_rate": 3.388865050047567e-06,
+      "loss": 0.4262,
+      "step": 17726
+    },
+    {
+      "epoch": 2.917886190012481,
+      "grad_norm": 0.4914583171959461,
+      "learning_rate": 3.388409547353039e-06,
+      "loss": 0.4434,
+      "step": 17727
+    },
+    {
+      "epoch": 2.9180507742315975,
+      "grad_norm": 0.3669387457957941,
+      "learning_rate": 3.3879540567877417e-06,
+      "loss": 0.4469,
+      "step": 17728
+    },
+    {
+      "epoch": 2.918215358450714,
+      "grad_norm": 0.2725498437245764,
+      "learning_rate": 3.387498578356648e-06,
+      "loss": 0.4399,
+      "step": 17729
+    },
+    {
+      "epoch": 2.9183799426698305,
+      "grad_norm": 0.3033736328383057,
+      "learning_rate": 3.3870431120647223e-06,
+      "loss": 0.4277,
+      "step": 17730
+    },
+    {
+      "epoch": 2.918544526888947,
+      "grad_norm": 0.3208317703581786,
+      "learning_rate": 3.386587657916941e-06,
+      "loss": 0.4261,
+      "step": 17731
+    },
+    {
+      "epoch": 2.9187091111080634,
+      "grad_norm": 0.3267134398822339,
+      "learning_rate": 3.38613221591827e-06,
+      "loss": 0.4305,
+      "step": 17732
+    },
+    {
+      "epoch": 2.9188736953271794,
+      "grad_norm": 0.35091926586516287,
+      "learning_rate": 3.385676786073679e-06,
+      "loss": 0.4268,
+      "step": 17733
+    },
+    {
+      "epoch": 2.919038279546296,
+      "grad_norm": 0.40270745330681973,
+      "learning_rate": 3.385221368388139e-06,
+      "loss": 0.4418,
+      "step": 17734
+    },
+    {
+      "epoch": 2.9192028637654124,
+      "grad_norm": 0.511617782510701,
+      "learning_rate": 3.3847659628666174e-06,
+      "loss": 0.4323,
+      "step": 17735
+    },
+    {
+      "epoch": 2.919367447984529,
+      "grad_norm": 0.3435292141125187,
+      "learning_rate": 3.3843105695140855e-06,
+      "loss": 0.4325,
+      "step": 17736
+    },
+    {
+      "epoch": 2.9195320322036453,
+      "grad_norm": 0.726359205620102,
+      "learning_rate": 3.38385518833551e-06,
+      "loss": 0.4008,
+      "step": 17737
+    },
+    {
+      "epoch": 2.919696616422762,
+      "grad_norm": 0.49802888669458983,
+      "learning_rate": 3.3833998193358633e-06,
+      "loss": 0.4347,
+      "step": 17738
+    },
+    {
+      "epoch": 2.9198612006418783,
+      "grad_norm": 0.30764054482203373,
+      "learning_rate": 3.3829444625201125e-06,
+      "loss": 0.4294,
+      "step": 17739
+    },
+    {
+      "epoch": 2.9200257848609947,
+      "grad_norm": 0.3067873725877841,
+      "learning_rate": 3.3824891178932242e-06,
+      "loss": 0.4325,
+      "step": 17740
+    },
+    {
+      "epoch": 2.920190369080111,
+      "grad_norm": 0.32780206845639454,
+      "learning_rate": 3.382033785460169e-06,
+      "loss": 0.4433,
+      "step": 17741
+    },
+    {
+      "epoch": 2.9203549532992277,
+      "grad_norm": 0.4208820936167372,
+      "learning_rate": 3.3815784652259143e-06,
+      "loss": 0.4156,
+      "step": 17742
+    },
+    {
+      "epoch": 2.920519537518344,
+      "grad_norm": 0.34487559592914324,
+      "learning_rate": 3.38112315719543e-06,
+      "loss": 0.4259,
+      "step": 17743
+    },
+    {
+      "epoch": 2.9206841217374606,
+      "grad_norm": 0.33124493288463996,
+      "learning_rate": 3.380667861373683e-06,
+      "loss": 0.4227,
+      "step": 17744
+    },
+    {
+      "epoch": 2.920848705956577,
+      "grad_norm": 0.5375485315494445,
+      "learning_rate": 3.3802125777656416e-06,
+      "loss": 0.4336,
+      "step": 17745
+    },
+    {
+      "epoch": 2.9210132901756936,
+      "grad_norm": 0.3042149214419165,
+      "learning_rate": 3.379757306376274e-06,
+      "loss": 0.4223,
+      "step": 17746
+    },
+    {
+      "epoch": 2.92117787439481,
+      "grad_norm": 0.4076199164068452,
+      "learning_rate": 3.3793020472105464e-06,
+      "loss": 0.4299,
+      "step": 17747
+    },
+    {
+      "epoch": 2.9213424586139265,
+      "grad_norm": 0.2753391754673677,
+      "learning_rate": 3.3788468002734295e-06,
+      "loss": 0.4278,
+      "step": 17748
+    },
+    {
+      "epoch": 2.921507042833043,
+      "grad_norm": 0.42015172301237264,
+      "learning_rate": 3.3783915655698867e-06,
+      "loss": 0.453,
+      "step": 17749
+    },
+    {
+      "epoch": 2.9216716270521594,
+      "grad_norm": 0.3835463068552829,
+      "learning_rate": 3.3779363431048895e-06,
+      "loss": 0.4461,
+      "step": 17750
+    },
+    {
+      "epoch": 2.921836211271276,
+      "grad_norm": 0.3857933309307179,
+      "learning_rate": 3.3774811328834003e-06,
+      "loss": 0.4282,
+      "step": 17751
+    },
+    {
+      "epoch": 2.9220007954903924,
+      "grad_norm": 0.4795193717887788,
+      "learning_rate": 3.3770259349103926e-06,
+      "loss": 0.4299,
+      "step": 17752
+    },
+    {
+      "epoch": 2.922165379709509,
+      "grad_norm": 0.3049139503695512,
+      "learning_rate": 3.376570749190828e-06,
+      "loss": 0.4354,
+      "step": 17753
+    },
+    {
+      "epoch": 2.9223299639286253,
+      "grad_norm": 0.3519921407722308,
+      "learning_rate": 3.3761155757296744e-06,
+      "loss": 0.4414,
+      "step": 17754
+    },
+    {
+      "epoch": 2.922494548147742,
+      "grad_norm": 0.27790101218221924,
+      "learning_rate": 3.3756604145319e-06,
+      "loss": 0.4282,
+      "step": 17755
+    },
+    {
+      "epoch": 2.9226591323668583,
+      "grad_norm": 0.44352435070616175,
+      "learning_rate": 3.3752052656024694e-06,
+      "loss": 0.4174,
+      "step": 17756
+    },
+    {
+      "epoch": 2.9228237165859747,
+      "grad_norm": 0.7634773223040751,
+      "learning_rate": 3.374750128946351e-06,
+      "loss": 0.4345,
+      "step": 17757
+    },
+    {
+      "epoch": 2.922988300805091,
+      "grad_norm": 0.2984704720453994,
+      "learning_rate": 3.3742950045685096e-06,
+      "loss": 0.4211,
+      "step": 17758
+    },
+    {
+      "epoch": 2.9231528850242077,
+      "grad_norm": 0.2873456683663334,
+      "learning_rate": 3.3738398924739133e-06,
+      "loss": 0.4269,
+      "step": 17759
+    },
+    {
+      "epoch": 2.923317469243324,
+      "grad_norm": 0.2681318304864954,
+      "learning_rate": 3.373384792667526e-06,
+      "loss": 0.4168,
+      "step": 17760
+    },
+    {
+      "epoch": 2.9234820534624406,
+      "grad_norm": 0.31182455738571124,
+      "learning_rate": 3.3729297051543125e-06,
+      "loss": 0.4427,
+      "step": 17761
+    },
+    {
+      "epoch": 2.923646637681557,
+      "grad_norm": 0.38556683725634705,
+      "learning_rate": 3.3724746299392423e-06,
+      "loss": 0.4332,
+      "step": 17762
+    },
+    {
+      "epoch": 2.9238112219006736,
+      "grad_norm": 0.3109254497590433,
+      "learning_rate": 3.3720195670272774e-06,
+      "loss": 0.4176,
+      "step": 17763
+    },
+    {
+      "epoch": 2.92397580611979,
+      "grad_norm": 0.41052865934839455,
+      "learning_rate": 3.3715645164233857e-06,
+      "loss": 0.4197,
+      "step": 17764
+    },
+    {
+      "epoch": 2.9241403903389065,
+      "grad_norm": 0.47507018610702884,
+      "learning_rate": 3.3711094781325302e-06,
+      "loss": 0.4279,
+      "step": 17765
+    },
+    {
+      "epoch": 2.924304974558023,
+      "grad_norm": 0.3087772347515225,
+      "learning_rate": 3.3706544521596794e-06,
+      "loss": 0.4373,
+      "step": 17766
+    },
+    {
+      "epoch": 2.9244695587771394,
+      "grad_norm": 0.5240269329553043,
+      "learning_rate": 3.3701994385097955e-06,
+      "loss": 0.4133,
+      "step": 17767
+    },
+    {
+      "epoch": 2.924634142996256,
+      "grad_norm": 0.3167803647593282,
+      "learning_rate": 3.3697444371878446e-06,
+      "loss": 0.4282,
+      "step": 17768
+    },
+    {
+      "epoch": 2.9247987272153724,
+      "grad_norm": 0.32589636198064426,
+      "learning_rate": 3.369289448198793e-06,
+      "loss": 0.4122,
+      "step": 17769
+    },
+    {
+      "epoch": 2.924963311434489,
+      "grad_norm": 0.436712996183922,
+      "learning_rate": 3.368834471547601e-06,
+      "loss": 0.4315,
+      "step": 17770
+    },
+    {
+      "epoch": 2.9251278956536053,
+      "grad_norm": 0.36515264087424926,
+      "learning_rate": 3.3683795072392378e-06,
+      "loss": 0.4334,
+      "step": 17771
+    },
+    {
+      "epoch": 2.925292479872722,
+      "grad_norm": 0.33728393279336444,
+      "learning_rate": 3.3679245552786643e-06,
+      "loss": 0.441,
+      "step": 17772
+    },
+    {
+      "epoch": 2.9254570640918383,
+      "grad_norm": 0.3032828860887325,
+      "learning_rate": 3.367469615670847e-06,
+      "loss": 0.4386,
+      "step": 17773
+    },
+    {
+      "epoch": 2.9256216483109547,
+      "grad_norm": 0.37763499166331943,
+      "learning_rate": 3.367014688420749e-06,
+      "loss": 0.4214,
+      "step": 17774
+    },
+    {
+      "epoch": 2.9257862325300708,
+      "grad_norm": 0.2951330033560652,
+      "learning_rate": 3.366559773533335e-06,
+      "loss": 0.4445,
+      "step": 17775
+    },
+    {
+      "epoch": 2.9259508167491872,
+      "grad_norm": 0.30043150877186625,
+      "learning_rate": 3.366104871013568e-06,
+      "loss": 0.4342,
+      "step": 17776
+    },
+    {
+      "epoch": 2.9261154009683037,
+      "grad_norm": 0.40274738954194816,
+      "learning_rate": 3.3656499808664123e-06,
+      "loss": 0.4189,
+      "step": 17777
+    },
+    {
+      "epoch": 2.92627998518742,
+      "grad_norm": 0.3009201296585576,
+      "learning_rate": 3.365195103096832e-06,
+      "loss": 0.4267,
+      "step": 17778
+    },
+    {
+      "epoch": 2.9264445694065366,
+      "grad_norm": 0.2916867542924242,
+      "learning_rate": 3.3647402377097877e-06,
+      "loss": 0.4137,
+      "step": 17779
+    },
+    {
+      "epoch": 2.926609153625653,
+      "grad_norm": 0.259953729653548,
+      "learning_rate": 3.364285384710247e-06,
+      "loss": 0.4436,
+      "step": 17780
+    },
+    {
+      "epoch": 2.9267737378447696,
+      "grad_norm": 0.24759841138821312,
+      "learning_rate": 3.3638305441031683e-06,
+      "loss": 0.4248,
+      "step": 17781
+    },
+    {
+      "epoch": 2.926938322063886,
+      "grad_norm": 0.4603191646651383,
+      "learning_rate": 3.36337571589352e-06,
+      "loss": 0.4267,
+      "step": 17782
+    },
+    {
+      "epoch": 2.9271029062830025,
+      "grad_norm": 0.33841942087743543,
+      "learning_rate": 3.3629209000862623e-06,
+      "loss": 0.427,
+      "step": 17783
+    },
+    {
+      "epoch": 2.927267490502119,
+      "grad_norm": 0.2910413296679468,
+      "learning_rate": 3.362466096686356e-06,
+      "loss": 0.4187,
+      "step": 17784
+    },
+    {
+      "epoch": 2.9274320747212355,
+      "grad_norm": 0.3569167002034256,
+      "learning_rate": 3.362011305698767e-06,
+      "loss": 0.4152,
+      "step": 17785
+    },
+    {
+      "epoch": 2.927596658940352,
+      "grad_norm": 0.30999334016630953,
+      "learning_rate": 3.3615565271284556e-06,
+      "loss": 0.4454,
+      "step": 17786
+    },
+    {
+      "epoch": 2.9277612431594684,
+      "grad_norm": 0.30668375560823136,
+      "learning_rate": 3.3611017609803855e-06,
+      "loss": 0.4255,
+      "step": 17787
+    },
+    {
+      "epoch": 2.927925827378585,
+      "grad_norm": 0.3213329643906008,
+      "learning_rate": 3.360647007259518e-06,
+      "loss": 0.4311,
+      "step": 17788
+    },
+    {
+      "epoch": 2.9280904115977013,
+      "grad_norm": 0.39838423121582006,
+      "learning_rate": 3.3601922659708165e-06,
+      "loss": 0.4196,
+      "step": 17789
+    },
+    {
+      "epoch": 2.928254995816818,
+      "grad_norm": 0.4026760553620033,
+      "learning_rate": 3.3597375371192427e-06,
+      "loss": 0.445,
+      "step": 17790
+    },
+    {
+      "epoch": 2.9284195800359343,
+      "grad_norm": 0.29864249003243665,
+      "learning_rate": 3.359282820709756e-06,
+      "loss": 0.4314,
+      "step": 17791
+    },
+    {
+      "epoch": 2.9285841642550507,
+      "grad_norm": 0.3050868808958821,
+      "learning_rate": 3.3588281167473213e-06,
+      "loss": 0.436,
+      "step": 17792
+    },
+    {
+      "epoch": 2.928748748474167,
+      "grad_norm": 0.35953047262041005,
+      "learning_rate": 3.3583734252368973e-06,
+      "loss": 0.4484,
+      "step": 17793
+    },
+    {
+      "epoch": 2.9289133326932837,
+      "grad_norm": 0.3837045335016808,
+      "learning_rate": 3.3579187461834484e-06,
+      "loss": 0.4034,
+      "step": 17794
+    },
+    {
+      "epoch": 2.9290779169124,
+      "grad_norm": 0.280143497290015,
+      "learning_rate": 3.3574640795919327e-06,
+      "loss": 0.4285,
+      "step": 17795
+    },
+    {
+      "epoch": 2.9292425011315166,
+      "grad_norm": 0.3197785437707039,
+      "learning_rate": 3.357009425467314e-06,
+      "loss": 0.433,
+      "step": 17796
+    },
+    {
+      "epoch": 2.929407085350633,
+      "grad_norm": 0.32012188869001085,
+      "learning_rate": 3.356554783814553e-06,
+      "loss": 0.4198,
+      "step": 17797
+    },
+    {
+      "epoch": 2.9295716695697496,
+      "grad_norm": 0.3085764226696516,
+      "learning_rate": 3.356100154638608e-06,
+      "loss": 0.4484,
+      "step": 17798
+    },
+    {
+      "epoch": 2.929736253788866,
+      "grad_norm": 0.3337658242361266,
+      "learning_rate": 3.3556455379444443e-06,
+      "loss": 0.43,
+      "step": 17799
+    },
+    {
+      "epoch": 2.929900838007982,
+      "grad_norm": 0.3262390653566618,
+      "learning_rate": 3.3551909337370166e-06,
+      "loss": 0.4364,
+      "step": 17800
+    },
+    {
+      "epoch": 2.9300654222270985,
+      "grad_norm": 0.2862178416575866,
+      "learning_rate": 3.354736342021291e-06,
+      "loss": 0.4307,
+      "step": 17801
+    },
+    {
+      "epoch": 2.930230006446215,
+      "grad_norm": 0.28962583779258144,
+      "learning_rate": 3.354281762802224e-06,
+      "loss": 0.4298,
+      "step": 17802
+    },
+    {
+      "epoch": 2.9303945906653315,
+      "grad_norm": 0.3390162700678234,
+      "learning_rate": 3.3538271960847778e-06,
+      "loss": 0.422,
+      "step": 17803
+    },
+    {
+      "epoch": 2.930559174884448,
+      "grad_norm": 0.3559782541224994,
+      "learning_rate": 3.353372641873912e-06,
+      "loss": 0.4394,
+      "step": 17804
+    },
+    {
+      "epoch": 2.9307237591035644,
+      "grad_norm": 0.33601042460541813,
+      "learning_rate": 3.3529181001745855e-06,
+      "loss": 0.4316,
+      "step": 17805
+    },
+    {
+      "epoch": 2.930888343322681,
+      "grad_norm": 0.29696711292317896,
+      "learning_rate": 3.352463570991759e-06,
+      "loss": 0.428,
+      "step": 17806
+    },
+    {
+      "epoch": 2.9310529275417974,
+      "grad_norm": 0.3066111993945832,
+      "learning_rate": 3.352009054330392e-06,
+      "loss": 0.4359,
+      "step": 17807
+    },
+    {
+      "epoch": 2.931217511760914,
+      "grad_norm": 0.29821728742624687,
+      "learning_rate": 3.3515545501954443e-06,
+      "loss": 0.4345,
+      "step": 17808
+    },
+    {
+      "epoch": 2.9313820959800303,
+      "grad_norm": 0.32587962170500345,
+      "learning_rate": 3.351100058591875e-06,
+      "loss": 0.4319,
+      "step": 17809
+    },
+    {
+      "epoch": 2.9315466801991468,
+      "grad_norm": 0.41296420067737905,
+      "learning_rate": 3.350645579524644e-06,
+      "loss": 0.4125,
+      "step": 17810
+    },
+    {
+      "epoch": 2.9317112644182632,
+      "grad_norm": 0.40766325984073193,
+      "learning_rate": 3.350191112998708e-06,
+      "loss": 0.4543,
+      "step": 17811
+    },
+    {
+      "epoch": 2.9318758486373797,
+      "grad_norm": 0.34426257609135424,
+      "learning_rate": 3.3497366590190294e-06,
+      "loss": 0.4411,
+      "step": 17812
+    },
+    {
+      "epoch": 2.932040432856496,
+      "grad_norm": 0.35398131134009875,
+      "learning_rate": 3.349282217590566e-06,
+      "loss": 0.4278,
+      "step": 17813
+    },
+    {
+      "epoch": 2.9322050170756127,
+      "grad_norm": 0.260521569928367,
+      "learning_rate": 3.348827788718273e-06,
+      "loss": 0.4142,
+      "step": 17814
+    },
+    {
+      "epoch": 2.932369601294729,
+      "grad_norm": 0.31993887898910756,
+      "learning_rate": 3.3483733724071136e-06,
+      "loss": 0.4372,
+      "step": 17815
+    },
+    {
+      "epoch": 2.9325341855138456,
+      "grad_norm": 0.30420162588792804,
+      "learning_rate": 3.3479189686620434e-06,
+      "loss": 0.4063,
+      "step": 17816
+    },
+    {
+      "epoch": 2.932698769732962,
+      "grad_norm": 0.3112488967412044,
+      "learning_rate": 3.347464577488022e-06,
+      "loss": 0.4084,
+      "step": 17817
+    },
+    {
+      "epoch": 2.9328633539520785,
+      "grad_norm": 0.3019402613651006,
+      "learning_rate": 3.347010198890007e-06,
+      "loss": 0.4246,
+      "step": 17818
+    },
+    {
+      "epoch": 2.933027938171195,
+      "grad_norm": 0.2904121659476341,
+      "learning_rate": 3.3465558328729574e-06,
+      "loss": 0.4282,
+      "step": 17819
+    },
+    {
+      "epoch": 2.9331925223903115,
+      "grad_norm": 0.3680894362455315,
+      "learning_rate": 3.3461014794418306e-06,
+      "loss": 0.4292,
+      "step": 17820
+    },
+    {
+      "epoch": 2.933357106609428,
+      "grad_norm": 0.8192649715094241,
+      "learning_rate": 3.3456471386015826e-06,
+      "loss": 0.4157,
+      "step": 17821
+    },
+    {
+      "epoch": 2.9335216908285444,
+      "grad_norm": 0.2984160228076957,
+      "learning_rate": 3.345192810357173e-06,
+      "loss": 0.4317,
+      "step": 17822
+    },
+    {
+      "epoch": 2.933686275047661,
+      "grad_norm": 0.2975559190054634,
+      "learning_rate": 3.3447384947135584e-06,
+      "loss": 0.4226,
+      "step": 17823
+    },
+    {
+      "epoch": 2.9338508592667774,
+      "grad_norm": 0.2975348369606467,
+      "learning_rate": 3.3442841916756962e-06,
+      "loss": 0.4478,
+      "step": 17824
+    },
+    {
+      "epoch": 2.934015443485894,
+      "grad_norm": 0.26314404116089213,
+      "learning_rate": 3.3438299012485428e-06,
+      "loss": 0.4198,
+      "step": 17825
+    },
+    {
+      "epoch": 2.9341800277050103,
+      "grad_norm": 0.5471598419819609,
+      "learning_rate": 3.3433756234370574e-06,
+      "loss": 0.4293,
+      "step": 17826
+    },
+    {
+      "epoch": 2.9343446119241268,
+      "grad_norm": 0.27370109412400817,
+      "learning_rate": 3.342921358246195e-06,
+      "loss": 0.4321,
+      "step": 17827
+    },
+    {
+      "epoch": 2.9345091961432432,
+      "grad_norm": 0.3078652883814931,
+      "learning_rate": 3.342467105680913e-06,
+      "loss": 0.4292,
+      "step": 17828
+    },
+    {
+      "epoch": 2.9346737803623597,
+      "grad_norm": 0.37786786549615187,
+      "learning_rate": 3.342012865746169e-06,
+      "loss": 0.4358,
+      "step": 17829
+    },
+    {
+      "epoch": 2.934838364581476,
+      "grad_norm": 0.2570317047160048,
+      "learning_rate": 3.341558638446916e-06,
+      "loss": 0.4209,
+      "step": 17830
+    },
+    {
+      "epoch": 2.9350029488005926,
+      "grad_norm": 0.292775737045322,
+      "learning_rate": 3.3411044237881156e-06,
+      "loss": 0.4218,
+      "step": 17831
+    },
+    {
+      "epoch": 2.935167533019709,
+      "grad_norm": 0.3469717961034144,
+      "learning_rate": 3.3406502217747186e-06,
+      "loss": 0.4323,
+      "step": 17832
+    },
+    {
+      "epoch": 2.9353321172388256,
+      "grad_norm": 0.2748724519305954,
+      "learning_rate": 3.340196032411686e-06,
+      "loss": 0.4294,
+      "step": 17833
+    },
+    {
+      "epoch": 2.935496701457942,
+      "grad_norm": 0.34156871435154457,
+      "learning_rate": 3.339741855703971e-06,
+      "loss": 0.4181,
+      "step": 17834
+    },
+    {
+      "epoch": 2.9356612856770585,
+      "grad_norm": 0.35911621554021983,
+      "learning_rate": 3.3392876916565288e-06,
+      "loss": 0.4466,
+      "step": 17835
+    },
+    {
+      "epoch": 2.935825869896175,
+      "grad_norm": 0.3112423328817136,
+      "learning_rate": 3.3388335402743165e-06,
+      "loss": 0.4508,
+      "step": 17836
+    },
+    {
+      "epoch": 2.9359904541152915,
+      "grad_norm": 0.3101049266574248,
+      "learning_rate": 3.3383794015622886e-06,
+      "loss": 0.4365,
+      "step": 17837
+    },
+    {
+      "epoch": 2.936155038334408,
+      "grad_norm": 0.40839422295382893,
+      "learning_rate": 3.337925275525402e-06,
+      "loss": 0.4697,
+      "step": 17838
+    },
+    {
+      "epoch": 2.9363196225535244,
+      "grad_norm": 0.3001045748710366,
+      "learning_rate": 3.3374711621686104e-06,
+      "loss": 0.4264,
+      "step": 17839
+    },
+    {
+      "epoch": 2.936484206772641,
+      "grad_norm": 0.36535566306293327,
+      "learning_rate": 3.33701706149687e-06,
+      "loss": 0.4172,
+      "step": 17840
+    },
+    {
+      "epoch": 2.936648790991757,
+      "grad_norm": 0.3262169272894157,
+      "learning_rate": 3.336562973515136e-06,
+      "loss": 0.4328,
+      "step": 17841
+    },
+    {
+      "epoch": 2.9368133752108734,
+      "grad_norm": 0.3334451630590715,
+      "learning_rate": 3.3361088982283614e-06,
+      "loss": 0.4247,
+      "step": 17842
+    },
+    {
+      "epoch": 2.93697795942999,
+      "grad_norm": 0.3085853407263294,
+      "learning_rate": 3.3356548356415024e-06,
+      "loss": 0.4309,
+      "step": 17843
+    },
+    {
+      "epoch": 2.9371425436491063,
+      "grad_norm": 0.38830741910708716,
+      "learning_rate": 3.335200785759512e-06,
+      "loss": 0.4231,
+      "step": 17844
+    },
+    {
+      "epoch": 2.937307127868223,
+      "grad_norm": 0.40923246564507443,
+      "learning_rate": 3.3347467485873472e-06,
+      "loss": 0.426,
+      "step": 17845
+    },
+    {
+      "epoch": 2.9374717120873393,
+      "grad_norm": 0.31888854734326744,
+      "learning_rate": 3.3342927241299597e-06,
+      "loss": 0.4291,
+      "step": 17846
+    },
+    {
+      "epoch": 2.9376362963064557,
+      "grad_norm": 0.3555392981963307,
+      "learning_rate": 3.333838712392306e-06,
+      "loss": 0.4407,
+      "step": 17847
+    },
+    {
+      "epoch": 2.937800880525572,
+      "grad_norm": 0.27002953860896445,
+      "learning_rate": 3.333384713379337e-06,
+      "loss": 0.4179,
+      "step": 17848
+    },
+    {
+      "epoch": 2.9379654647446887,
+      "grad_norm": 0.31473988399066993,
+      "learning_rate": 3.3329307270960096e-06,
+      "loss": 0.4227,
+      "step": 17849
+    },
+    {
+      "epoch": 2.938130048963805,
+      "grad_norm": 0.29185727361215336,
+      "learning_rate": 3.332476753547278e-06,
+      "loss": 0.4283,
+      "step": 17850
+    },
+    {
+      "epoch": 2.9382946331829216,
+      "grad_norm": 0.28049103874445924,
+      "learning_rate": 3.332022792738091e-06,
+      "loss": 0.4286,
+      "step": 17851
+    },
+    {
+      "epoch": 2.938459217402038,
+      "grad_norm": 0.30212525955101455,
+      "learning_rate": 3.3315688446734078e-06,
+      "loss": 0.4362,
+      "step": 17852
+    },
+    {
+      "epoch": 2.9386238016211546,
+      "grad_norm": 0.2904014180261415,
+      "learning_rate": 3.331114909358177e-06,
+      "loss": 0.4406,
+      "step": 17853
+    },
+    {
+      "epoch": 2.938788385840271,
+      "grad_norm": 0.3172390058145223,
+      "learning_rate": 3.330660986797355e-06,
+      "loss": 0.405,
+      "step": 17854
+    },
+    {
+      "epoch": 2.9389529700593875,
+      "grad_norm": 0.2798833508296669,
+      "learning_rate": 3.330207076995893e-06,
+      "loss": 0.4175,
+      "step": 17855
+    },
+    {
+      "epoch": 2.939117554278504,
+      "grad_norm": 0.3491267780694839,
+      "learning_rate": 3.3297531799587447e-06,
+      "loss": 0.4351,
+      "step": 17856
+    },
+    {
+      "epoch": 2.9392821384976204,
+      "grad_norm": 0.2964701733388763,
+      "learning_rate": 3.3292992956908635e-06,
+      "loss": 0.4277,
+      "step": 17857
+    },
+    {
+      "epoch": 2.939446722716737,
+      "grad_norm": 0.3005906746028151,
+      "learning_rate": 3.3288454241971997e-06,
+      "loss": 0.4321,
+      "step": 17858
+    },
+    {
+      "epoch": 2.9396113069358534,
+      "grad_norm": 0.33130600309368236,
+      "learning_rate": 3.328391565482708e-06,
+      "loss": 0.413,
+      "step": 17859
+    },
+    {
+      "epoch": 2.93977589115497,
+      "grad_norm": 0.26528857819174023,
+      "learning_rate": 3.3279377195523392e-06,
+      "loss": 0.4358,
+      "step": 17860
+    },
+    {
+      "epoch": 2.9399404753740863,
+      "grad_norm": 0.28999419057896214,
+      "learning_rate": 3.3274838864110487e-06,
+      "loss": 0.4465,
+      "step": 17861
+    },
+    {
+      "epoch": 2.940105059593203,
+      "grad_norm": 0.6408457564674434,
+      "learning_rate": 3.3270300660637825e-06,
+      "loss": 0.4379,
+      "step": 17862
+    },
+    {
+      "epoch": 2.9402696438123193,
+      "grad_norm": 0.4237045146833924,
+      "learning_rate": 3.3265762585154993e-06,
+      "loss": 0.4219,
+      "step": 17863
+    },
+    {
+      "epoch": 2.9404342280314357,
+      "grad_norm": 0.2729461510661597,
+      "learning_rate": 3.3261224637711465e-06,
+      "loss": 0.4524,
+      "step": 17864
+    },
+    {
+      "epoch": 2.940598812250552,
+      "grad_norm": 0.3003957467857865,
+      "learning_rate": 3.325668681835676e-06,
+      "loss": 0.424,
+      "step": 17865
+    },
+    {
+      "epoch": 2.9407633964696682,
+      "grad_norm": 0.29451358615558454,
+      "learning_rate": 3.3252149127140414e-06,
+      "loss": 0.4476,
+      "step": 17866
+    },
+    {
+      "epoch": 2.9409279806887847,
+      "grad_norm": 0.31582228502791126,
+      "learning_rate": 3.3247611564111913e-06,
+      "loss": 0.438,
+      "step": 17867
+    },
+    {
+      "epoch": 2.941092564907901,
+      "grad_norm": 0.3113893632208832,
+      "learning_rate": 3.32430741293208e-06,
+      "loss": 0.4042,
+      "step": 17868
+    },
+    {
+      "epoch": 2.9412571491270176,
+      "grad_norm": 0.4684080838883566,
+      "learning_rate": 3.3238536822816554e-06,
+      "loss": 0.4273,
+      "step": 17869
+    },
+    {
+      "epoch": 2.941421733346134,
+      "grad_norm": 0.3282530409190831,
+      "learning_rate": 3.323399964464871e-06,
+      "loss": 0.4317,
+      "step": 17870
+    },
+    {
+      "epoch": 2.9415863175652506,
+      "grad_norm": 0.387300135890649,
+      "learning_rate": 3.3229462594866778e-06,
+      "loss": 0.4472,
+      "step": 17871
+    },
+    {
+      "epoch": 2.941750901784367,
+      "grad_norm": 0.3230209122635716,
+      "learning_rate": 3.3224925673520237e-06,
+      "loss": 0.4375,
+      "step": 17872
+    },
+    {
+      "epoch": 2.9419154860034835,
+      "grad_norm": 0.2884446232718566,
+      "learning_rate": 3.322038888065861e-06,
+      "loss": 0.4448,
+      "step": 17873
+    },
+    {
+      "epoch": 2.9420800702226,
+      "grad_norm": 0.37083274672382366,
+      "learning_rate": 3.3215852216331395e-06,
+      "loss": 0.4327,
+      "step": 17874
+    },
+    {
+      "epoch": 2.9422446544417165,
+      "grad_norm": 0.29977657531127405,
+      "learning_rate": 3.3211315680588108e-06,
+      "loss": 0.4318,
+      "step": 17875
+    },
+    {
+      "epoch": 2.942409238660833,
+      "grad_norm": 0.47841463020664227,
+      "learning_rate": 3.320677927347822e-06,
+      "loss": 0.4537,
+      "step": 17876
+    },
+    {
+      "epoch": 2.9425738228799494,
+      "grad_norm": 0.3629239245166473,
+      "learning_rate": 3.320224299505127e-06,
+      "loss": 0.4355,
+      "step": 17877
+    },
+    {
+      "epoch": 2.942738407099066,
+      "grad_norm": 0.4161272794584303,
+      "learning_rate": 3.3197706845356733e-06,
+      "loss": 0.4212,
+      "step": 17878
+    },
+    {
+      "epoch": 2.9429029913181823,
+      "grad_norm": 0.291344277246969,
+      "learning_rate": 3.31931708244441e-06,
+      "loss": 0.4333,
+      "step": 17879
+    },
+    {
+      "epoch": 2.943067575537299,
+      "grad_norm": 0.40246922396376766,
+      "learning_rate": 3.318863493236289e-06,
+      "loss": 0.4366,
+      "step": 17880
+    },
+    {
+      "epoch": 2.9432321597564153,
+      "grad_norm": 0.29864377914785467,
+      "learning_rate": 3.3184099169162553e-06,
+      "loss": 0.4286,
+      "step": 17881
+    },
+    {
+      "epoch": 2.9433967439755317,
+      "grad_norm": 0.27655697850060473,
+      "learning_rate": 3.3179563534892642e-06,
+      "loss": 0.4205,
+      "step": 17882
+    },
+    {
+      "epoch": 2.943561328194648,
+      "grad_norm": 0.2935291739535782,
+      "learning_rate": 3.3175028029602586e-06,
+      "loss": 0.4354,
+      "step": 17883
+    },
+    {
+      "epoch": 2.9437259124137647,
+      "grad_norm": 0.49612396605339865,
+      "learning_rate": 3.3170492653341935e-06,
+      "loss": 0.4144,
+      "step": 17884
+    },
+    {
+      "epoch": 2.943890496632881,
+      "grad_norm": 0.41181039529890884,
+      "learning_rate": 3.3165957406160135e-06,
+      "loss": 0.4503,
+      "step": 17885
+    },
+    {
+      "epoch": 2.9440550808519976,
+      "grad_norm": 0.34263661056886074,
+      "learning_rate": 3.316142228810668e-06,
+      "loss": 0.4118,
+      "step": 17886
+    },
+    {
+      "epoch": 2.944219665071114,
+      "grad_norm": 0.28397946210819824,
+      "learning_rate": 3.3156887299231066e-06,
+      "loss": 0.4209,
+      "step": 17887
+    },
+    {
+      "epoch": 2.9443842492902306,
+      "grad_norm": 0.3446468897435312,
+      "learning_rate": 3.315235243958276e-06,
+      "loss": 0.4275,
+      "step": 17888
+    },
+    {
+      "epoch": 2.944548833509347,
+      "grad_norm": 0.34994566616221195,
+      "learning_rate": 3.314781770921127e-06,
+      "loss": 0.4281,
+      "step": 17889
+    },
+    {
+      "epoch": 2.9447134177284635,
+      "grad_norm": 0.2673280268924159,
+      "learning_rate": 3.3143283108166047e-06,
+      "loss": 0.4245,
+      "step": 17890
+    },
+    {
+      "epoch": 2.94487800194758,
+      "grad_norm": 0.3052985036142152,
+      "learning_rate": 3.3138748636496613e-06,
+      "loss": 0.4355,
+      "step": 17891
+    },
+    {
+      "epoch": 2.9450425861666965,
+      "grad_norm": 0.24296991418516659,
+      "learning_rate": 3.3134214294252385e-06,
+      "loss": 0.4266,
+      "step": 17892
+    },
+    {
+      "epoch": 2.945207170385813,
+      "grad_norm": 0.348211095834571,
+      "learning_rate": 3.3129680081482908e-06,
+      "loss": 0.4502,
+      "step": 17893
+    },
+    {
+      "epoch": 2.9453717546049294,
+      "grad_norm": 0.3066715616668804,
+      "learning_rate": 3.3125145998237613e-06,
+      "loss": 0.4228,
+      "step": 17894
+    },
+    {
+      "epoch": 2.945536338824046,
+      "grad_norm": 0.7027922140802841,
+      "learning_rate": 3.312061204456598e-06,
+      "loss": 0.4426,
+      "step": 17895
+    },
+    {
+      "epoch": 2.9457009230431623,
+      "grad_norm": 0.35514336309042926,
+      "learning_rate": 3.3116078220517486e-06,
+      "loss": 0.4192,
+      "step": 17896
+    },
+    {
+      "epoch": 2.945865507262279,
+      "grad_norm": 0.2990787120046496,
+      "learning_rate": 3.311154452614161e-06,
+      "loss": 0.4526,
+      "step": 17897
+    },
+    {
+      "epoch": 2.9460300914813953,
+      "grad_norm": 0.3404325301953905,
+      "learning_rate": 3.3107010961487814e-06,
+      "loss": 0.4163,
+      "step": 17898
+    },
+    {
+      "epoch": 2.9461946757005117,
+      "grad_norm": 0.3363829382675295,
+      "learning_rate": 3.310247752660556e-06,
+      "loss": 0.4156,
+      "step": 17899
+    },
+    {
+      "epoch": 2.946359259919628,
+      "grad_norm": 0.3528096123152522,
+      "learning_rate": 3.309794422154433e-06,
+      "loss": 0.4366,
+      "step": 17900
+    },
+    {
+      "epoch": 2.9465238441387447,
+      "grad_norm": 0.45103200585706277,
+      "learning_rate": 3.3093411046353597e-06,
+      "loss": 0.4289,
+      "step": 17901
+    },
+    {
+      "epoch": 2.946688428357861,
+      "grad_norm": 0.25641087569932003,
+      "learning_rate": 3.308887800108278e-06,
+      "loss": 0.421,
+      "step": 17902
+    },
+    {
+      "epoch": 2.9468530125769776,
+      "grad_norm": 0.36216526002760185,
+      "learning_rate": 3.3084345085781404e-06,
+      "loss": 0.433,
+      "step": 17903
+    },
+    {
+      "epoch": 2.947017596796094,
+      "grad_norm": 0.3100472023891358,
+      "learning_rate": 3.3079812300498875e-06,
+      "loss": 0.4348,
+      "step": 17904
+    },
+    {
+      "epoch": 2.9471821810152106,
+      "grad_norm": 0.28062098127009283,
+      "learning_rate": 3.3075279645284693e-06,
+      "loss": 0.4502,
+      "step": 17905
+    },
+    {
+      "epoch": 2.947346765234327,
+      "grad_norm": 0.3299692058165166,
+      "learning_rate": 3.307074712018829e-06,
+      "loss": 0.4278,
+      "step": 17906
+    },
+    {
+      "epoch": 2.9475113494534435,
+      "grad_norm": 0.42246795002332643,
+      "learning_rate": 3.3066214725259136e-06,
+      "loss": 0.4224,
+      "step": 17907
+    },
+    {
+      "epoch": 2.9476759336725595,
+      "grad_norm": 0.275676822494741,
+      "learning_rate": 3.3061682460546688e-06,
+      "loss": 0.4379,
+      "step": 17908
+    },
+    {
+      "epoch": 2.947840517891676,
+      "grad_norm": 0.29540839746440317,
+      "learning_rate": 3.3057150326100385e-06,
+      "loss": 0.4214,
+      "step": 17909
+    },
+    {
+      "epoch": 2.9480051021107925,
+      "grad_norm": 0.31534615907655184,
+      "learning_rate": 3.305261832196971e-06,
+      "loss": 0.4332,
+      "step": 17910
+    },
+    {
+      "epoch": 2.948169686329909,
+      "grad_norm": 0.3277670198611158,
+      "learning_rate": 3.3048086448204074e-06,
+      "loss": 0.4238,
+      "step": 17911
+    },
+    {
+      "epoch": 2.9483342705490254,
+      "grad_norm": 0.3162074351131492,
+      "learning_rate": 3.304355470485297e-06,
+      "loss": 0.4091,
+      "step": 17912
+    },
+    {
+      "epoch": 2.948498854768142,
+      "grad_norm": 0.4291195573102557,
+      "learning_rate": 3.30390230919658e-06,
+      "loss": 0.4411,
+      "step": 17913
+    },
+    {
+      "epoch": 2.9486634389872584,
+      "grad_norm": 0.31681130856431333,
+      "learning_rate": 3.303449160959206e-06,
+      "loss": 0.4334,
+      "step": 17914
+    },
+    {
+      "epoch": 2.948828023206375,
+      "grad_norm": 0.31191226705513325,
+      "learning_rate": 3.302996025778117e-06,
+      "loss": 0.4234,
+      "step": 17915
+    },
+    {
+      "epoch": 2.9489926074254913,
+      "grad_norm": 0.29734890492769767,
+      "learning_rate": 3.3025429036582564e-06,
+      "loss": 0.4267,
+      "step": 17916
+    },
+    {
+      "epoch": 2.9491571916446078,
+      "grad_norm": 0.2899155809468505,
+      "learning_rate": 3.3020897946045703e-06,
+      "loss": 0.4282,
+      "step": 17917
+    },
+    {
+      "epoch": 2.9493217758637242,
+      "grad_norm": 0.25672904390104206,
+      "learning_rate": 3.301636698622002e-06,
+      "loss": 0.4406,
+      "step": 17918
+    },
+    {
+      "epoch": 2.9494863600828407,
+      "grad_norm": 0.2789684768772823,
+      "learning_rate": 3.301183615715496e-06,
+      "loss": 0.442,
+      "step": 17919
+    },
+    {
+      "epoch": 2.949650944301957,
+      "grad_norm": 0.2896431189484109,
+      "learning_rate": 3.300730545889996e-06,
+      "loss": 0.4393,
+      "step": 17920
+    },
+    {
+      "epoch": 2.9498155285210736,
+      "grad_norm": 0.3326451688102741,
+      "learning_rate": 3.3002774891504457e-06,
+      "loss": 0.4504,
+      "step": 17921
+    },
+    {
+      "epoch": 2.94998011274019,
+      "grad_norm": 0.3025192215294995,
+      "learning_rate": 3.2998244455017897e-06,
+      "loss": 0.4544,
+      "step": 17922
+    },
+    {
+      "epoch": 2.9501446969593066,
+      "grad_norm": 0.36848992028546734,
+      "learning_rate": 3.299371414948969e-06,
+      "loss": 0.4385,
+      "step": 17923
+    },
+    {
+      "epoch": 2.950309281178423,
+      "grad_norm": 0.27464068844369605,
+      "learning_rate": 3.298918397496929e-06,
+      "loss": 0.4181,
+      "step": 17924
+    },
+    {
+      "epoch": 2.9504738653975395,
+      "grad_norm": 0.3508843613547432,
+      "learning_rate": 3.2984653931506105e-06,
+      "loss": 0.4378,
+      "step": 17925
+    },
+    {
+      "epoch": 2.950638449616656,
+      "grad_norm": 0.3168887900492193,
+      "learning_rate": 3.29801240191496e-06,
+      "loss": 0.4072,
+      "step": 17926
+    },
+    {
+      "epoch": 2.9508030338357725,
+      "grad_norm": 0.36933439551737113,
+      "learning_rate": 3.2975594237949166e-06,
+      "loss": 0.4455,
+      "step": 17927
+    },
+    {
+      "epoch": 2.950967618054889,
+      "grad_norm": 0.32323328773774185,
+      "learning_rate": 3.2971064587954263e-06,
+      "loss": 0.4321,
+      "step": 17928
+    },
+    {
+      "epoch": 2.9511322022740054,
+      "grad_norm": 0.2852129816505733,
+      "learning_rate": 3.2966535069214315e-06,
+      "loss": 0.4272,
+      "step": 17929
+    },
+    {
+      "epoch": 2.951296786493122,
+      "grad_norm": 0.3360420850201866,
+      "learning_rate": 3.2962005681778704e-06,
+      "loss": 0.443,
+      "step": 17930
+    },
+    {
+      "epoch": 2.9514613707122384,
+      "grad_norm": 0.28600922843412036,
+      "learning_rate": 3.2957476425696905e-06,
+      "loss": 0.4447,
+      "step": 17931
+    },
+    {
+      "epoch": 2.951625954931355,
+      "grad_norm": 0.35561012211751764,
+      "learning_rate": 3.295294730101829e-06,
+      "loss": 0.4325,
+      "step": 17932
+    },
+    {
+      "epoch": 2.951790539150471,
+      "grad_norm": 0.3208730436976375,
+      "learning_rate": 3.2948418307792336e-06,
+      "loss": 0.4253,
+      "step": 17933
+    },
+    {
+      "epoch": 2.9519551233695873,
+      "grad_norm": 0.2722359535079783,
+      "learning_rate": 3.294388944606842e-06,
+      "loss": 0.4368,
+      "step": 17934
+    },
+    {
+      "epoch": 2.952119707588704,
+      "grad_norm": 0.2885866212922849,
+      "learning_rate": 3.293936071589597e-06,
+      "loss": 0.4176,
+      "step": 17935
+    },
+    {
+      "epoch": 2.9522842918078203,
+      "grad_norm": 0.4040311456081655,
+      "learning_rate": 3.2934832117324396e-06,
+      "loss": 0.4256,
+      "step": 17936
+    },
+    {
+      "epoch": 2.9524488760269367,
+      "grad_norm": 0.33857774276731406,
+      "learning_rate": 3.2930303650403133e-06,
+      "loss": 0.4594,
+      "step": 17937
+    },
+    {
+      "epoch": 2.952613460246053,
+      "grad_norm": 0.3265745780732682,
+      "learning_rate": 3.292577531518158e-06,
+      "loss": 0.4259,
+      "step": 17938
+    },
+    {
+      "epoch": 2.9527780444651697,
+      "grad_norm": 0.29969628802037535,
+      "learning_rate": 3.292124711170913e-06,
+      "loss": 0.4242,
+      "step": 17939
+    },
+    {
+      "epoch": 2.952942628684286,
+      "grad_norm": 0.27139539898638215,
+      "learning_rate": 3.2916719040035227e-06,
+      "loss": 0.4326,
+      "step": 17940
+    },
+    {
+      "epoch": 2.9531072129034026,
+      "grad_norm": 0.33381717418697626,
+      "learning_rate": 3.2912191100209247e-06,
+      "loss": 0.4323,
+      "step": 17941
+    },
+    {
+      "epoch": 2.953271797122519,
+      "grad_norm": 0.2919655671312646,
+      "learning_rate": 3.2907663292280632e-06,
+      "loss": 0.4084,
+      "step": 17942
+    },
+    {
+      "epoch": 2.9534363813416356,
+      "grad_norm": 0.3061555882408243,
+      "learning_rate": 3.290313561629875e-06,
+      "loss": 0.4326,
+      "step": 17943
+    },
+    {
+      "epoch": 2.953600965560752,
+      "grad_norm": 0.2730678949072666,
+      "learning_rate": 3.2898608072313046e-06,
+      "loss": 0.4271,
+      "step": 17944
+    },
+    {
+      "epoch": 2.9537655497798685,
+      "grad_norm": 0.25751069019917944,
+      "learning_rate": 3.28940806603729e-06,
+      "loss": 0.4044,
+      "step": 17945
+    },
+    {
+      "epoch": 2.953930133998985,
+      "grad_norm": 0.32760159845054954,
+      "learning_rate": 3.28895533805277e-06,
+      "loss": 0.4326,
+      "step": 17946
+    },
+    {
+      "epoch": 2.9540947182181014,
+      "grad_norm": 0.3229338862618099,
+      "learning_rate": 3.2885026232826865e-06,
+      "loss": 0.4296,
+      "step": 17947
+    },
+    {
+      "epoch": 2.954259302437218,
+      "grad_norm": 0.283933379779198,
+      "learning_rate": 3.2880499217319778e-06,
+      "loss": 0.4437,
+      "step": 17948
+    },
+    {
+      "epoch": 2.9544238866563344,
+      "grad_norm": 0.33073947677974724,
+      "learning_rate": 3.2875972334055866e-06,
+      "loss": 0.4251,
+      "step": 17949
+    },
+    {
+      "epoch": 2.954588470875451,
+      "grad_norm": 0.30371658072334495,
+      "learning_rate": 3.2871445583084486e-06,
+      "loss": 0.4361,
+      "step": 17950
+    },
+    {
+      "epoch": 2.9547530550945673,
+      "grad_norm": 0.34983625322813805,
+      "learning_rate": 3.2866918964455068e-06,
+      "loss": 0.4238,
+      "step": 17951
+    },
+    {
+      "epoch": 2.954917639313684,
+      "grad_norm": 0.2555862807040976,
+      "learning_rate": 3.2862392478216996e-06,
+      "loss": 0.4112,
+      "step": 17952
+    },
+    {
+      "epoch": 2.9550822235328003,
+      "grad_norm": 0.511680191622086,
+      "learning_rate": 3.2857866124419632e-06,
+      "loss": 0.4365,
+      "step": 17953
+    },
+    {
+      "epoch": 2.9552468077519167,
+      "grad_norm": 0.3025622763140895,
+      "learning_rate": 3.2853339903112397e-06,
+      "loss": 0.4468,
+      "step": 17954
+    },
+    {
+      "epoch": 2.955411391971033,
+      "grad_norm": 0.30488710406531333,
+      "learning_rate": 3.284881381434466e-06,
+      "loss": 0.4548,
+      "step": 17955
+    },
+    {
+      "epoch": 2.9555759761901497,
+      "grad_norm": 0.29399243678847514,
+      "learning_rate": 3.284428785816583e-06,
+      "loss": 0.4458,
+      "step": 17956
+    },
+    {
+      "epoch": 2.955740560409266,
+      "grad_norm": 0.2704503173871928,
+      "learning_rate": 3.283976203462527e-06,
+      "loss": 0.4337,
+      "step": 17957
+    },
+    {
+      "epoch": 2.9559051446283826,
+      "grad_norm": 0.36732907875717224,
+      "learning_rate": 3.2835236343772375e-06,
+      "loss": 0.4158,
+      "step": 17958
+    },
+    {
+      "epoch": 2.956069728847499,
+      "grad_norm": 0.3164426199884768,
+      "learning_rate": 3.283071078565653e-06,
+      "loss": 0.4329,
+      "step": 17959
+    },
+    {
+      "epoch": 2.9562343130666155,
+      "grad_norm": 0.31369984071288404,
+      "learning_rate": 3.2826185360327103e-06,
+      "loss": 0.4209,
+      "step": 17960
+    },
+    {
+      "epoch": 2.956398897285732,
+      "grad_norm": 0.3399865849521089,
+      "learning_rate": 3.2821660067833503e-06,
+      "loss": 0.4232,
+      "step": 17961
+    },
+    {
+      "epoch": 2.9565634815048485,
+      "grad_norm": 0.3107880242346844,
+      "learning_rate": 3.281713490822506e-06,
+      "loss": 0.4189,
+      "step": 17962
+    },
+    {
+      "epoch": 2.956728065723965,
+      "grad_norm": 0.41034059979693965,
+      "learning_rate": 3.2812609881551207e-06,
+      "loss": 0.4374,
+      "step": 17963
+    },
+    {
+      "epoch": 2.9568926499430814,
+      "grad_norm": 0.3564405810029044,
+      "learning_rate": 3.2808084987861263e-06,
+      "loss": 0.4211,
+      "step": 17964
+    },
+    {
+      "epoch": 2.957057234162198,
+      "grad_norm": 0.36128352218561705,
+      "learning_rate": 3.2803560227204654e-06,
+      "loss": 0.4399,
+      "step": 17965
+    },
+    {
+      "epoch": 2.9572218183813144,
+      "grad_norm": 0.48125988482566323,
+      "learning_rate": 3.279903559963072e-06,
+      "loss": 0.4336,
+      "step": 17966
+    },
+    {
+      "epoch": 2.957386402600431,
+      "grad_norm": 0.30767701112235857,
+      "learning_rate": 3.2794511105188835e-06,
+      "loss": 0.4245,
+      "step": 17967
+    },
+    {
+      "epoch": 2.9575509868195473,
+      "grad_norm": 0.3268399282048589,
+      "learning_rate": 3.2789986743928377e-06,
+      "loss": 0.4268,
+      "step": 17968
+    },
+    {
+      "epoch": 2.957715571038664,
+      "grad_norm": 0.2830963299254264,
+      "learning_rate": 3.2785462515898706e-06,
+      "loss": 0.4249,
+      "step": 17969
+    },
+    {
+      "epoch": 2.9578801552577803,
+      "grad_norm": 0.2991147151326696,
+      "learning_rate": 3.27809384211492e-06,
+      "loss": 0.4392,
+      "step": 17970
+    },
+    {
+      "epoch": 2.9580447394768967,
+      "grad_norm": 0.37204084254622605,
+      "learning_rate": 3.277641445972921e-06,
+      "loss": 0.4376,
+      "step": 17971
+    },
+    {
+      "epoch": 2.958209323696013,
+      "grad_norm": 0.2818280247339522,
+      "learning_rate": 3.2771890631688128e-06,
+      "loss": 0.4306,
+      "step": 17972
+    },
+    {
+      "epoch": 2.9583739079151297,
+      "grad_norm": 0.2951545466599934,
+      "learning_rate": 3.276736693707526e-06,
+      "loss": 0.4288,
+      "step": 17973
+    },
+    {
+      "epoch": 2.958538492134246,
+      "grad_norm": 0.4873851836731497,
+      "learning_rate": 3.276284337594004e-06,
+      "loss": 0.4191,
+      "step": 17974
+    },
+    {
+      "epoch": 2.958703076353362,
+      "grad_norm": 0.3930468198978656,
+      "learning_rate": 3.275831994833177e-06,
+      "loss": 0.4402,
+      "step": 17975
+    },
+    {
+      "epoch": 2.9588676605724786,
+      "grad_norm": 0.2764863576168517,
+      "learning_rate": 3.275379665429982e-06,
+      "loss": 0.4427,
+      "step": 17976
+    },
+    {
+      "epoch": 2.959032244791595,
+      "grad_norm": 0.3303930389333476,
+      "learning_rate": 3.2749273493893567e-06,
+      "loss": 0.4386,
+      "step": 17977
+    },
+    {
+      "epoch": 2.9591968290107116,
+      "grad_norm": 0.29647947734247887,
+      "learning_rate": 3.2744750467162338e-06,
+      "loss": 0.4325,
+      "step": 17978
+    },
+    {
+      "epoch": 2.959361413229828,
+      "grad_norm": 0.3180004148333902,
+      "learning_rate": 3.2740227574155516e-06,
+      "loss": 0.4111,
+      "step": 17979
+    },
+    {
+      "epoch": 2.9595259974489445,
+      "grad_norm": 0.7645019242291001,
+      "learning_rate": 3.273570481492242e-06,
+      "loss": 0.4312,
+      "step": 17980
+    },
+    {
+      "epoch": 2.959690581668061,
+      "grad_norm": 0.2773599141407237,
+      "learning_rate": 3.2731182189512437e-06,
+      "loss": 0.4186,
+      "step": 17981
+    },
+    {
+      "epoch": 2.9598551658871775,
+      "grad_norm": 0.32322410607835417,
+      "learning_rate": 3.2726659697974906e-06,
+      "loss": 0.4221,
+      "step": 17982
+    },
+    {
+      "epoch": 2.960019750106294,
+      "grad_norm": 0.3347967700591786,
+      "learning_rate": 3.2722137340359128e-06,
+      "loss": 0.4509,
+      "step": 17983
+    },
+    {
+      "epoch": 2.9601843343254104,
+      "grad_norm": 0.32349798325468926,
+      "learning_rate": 3.2717615116714523e-06,
+      "loss": 0.4223,
+      "step": 17984
+    },
+    {
+      "epoch": 2.960348918544527,
+      "grad_norm": 0.2557811785916594,
+      "learning_rate": 3.271309302709038e-06,
+      "loss": 0.4437,
+      "step": 17985
+    },
+    {
+      "epoch": 2.9605135027636433,
+      "grad_norm": 0.2802734711444316,
+      "learning_rate": 3.2708571071536076e-06,
+      "loss": 0.424,
+      "step": 17986
+    },
+    {
+      "epoch": 2.96067808698276,
+      "grad_norm": 0.3835232587119877,
+      "learning_rate": 3.270404925010092e-06,
+      "loss": 0.4255,
+      "step": 17987
+    },
+    {
+      "epoch": 2.9608426712018763,
+      "grad_norm": 0.4379819983304958,
+      "learning_rate": 3.269952756283429e-06,
+      "loss": 0.4133,
+      "step": 17988
+    },
+    {
+      "epoch": 2.9610072554209927,
+      "grad_norm": 0.33758890094692523,
+      "learning_rate": 3.2695006009785503e-06,
+      "loss": 0.4357,
+      "step": 17989
+    },
+    {
+      "epoch": 2.961171839640109,
+      "grad_norm": 1.5701706025079465,
+      "learning_rate": 3.2690484591003883e-06,
+      "loss": 0.4507,
+      "step": 17990
+    },
+    {
+      "epoch": 2.9613364238592257,
+      "grad_norm": 0.34036360947169975,
+      "learning_rate": 3.26859633065388e-06,
+      "loss": 0.4198,
+      "step": 17991
+    },
+    {
+      "epoch": 2.961501008078342,
+      "grad_norm": 0.3288506108564918,
+      "learning_rate": 3.2681442156439565e-06,
+      "loss": 0.4209,
+      "step": 17992
+    },
+    {
+      "epoch": 2.9616655922974586,
+      "grad_norm": 0.4595635703723497,
+      "learning_rate": 3.2676921140755524e-06,
+      "loss": 0.4386,
+      "step": 17993
+    },
+    {
+      "epoch": 2.961830176516575,
+      "grad_norm": 0.4226522308980864,
+      "learning_rate": 3.267240025953598e-06,
+      "loss": 0.418,
+      "step": 17994
+    },
+    {
+      "epoch": 2.9619947607356916,
+      "grad_norm": 0.3557572729847082,
+      "learning_rate": 3.2667879512830316e-06,
+      "loss": 0.4484,
+      "step": 17995
+    },
+    {
+      "epoch": 2.962159344954808,
+      "grad_norm": 0.28465195785827985,
+      "learning_rate": 3.266335890068781e-06,
+      "loss": 0.4249,
+      "step": 17996
+    },
+    {
+      "epoch": 2.9623239291739245,
+      "grad_norm": 0.30735732145131517,
+      "learning_rate": 3.2658838423157806e-06,
+      "loss": 0.42,
+      "step": 17997
+    },
+    {
+      "epoch": 2.962488513393041,
+      "grad_norm": 0.3742831146688339,
+      "learning_rate": 3.265431808028964e-06,
+      "loss": 0.4296,
+      "step": 17998
+    },
+    {
+      "epoch": 2.9626530976121574,
+      "grad_norm": 0.3994875408359112,
+      "learning_rate": 3.2649797872132614e-06,
+      "loss": 0.4255,
+      "step": 17999
+    },
+    {
+      "epoch": 2.9628176818312735,
+      "grad_norm": 0.3290846626459086,
+      "learning_rate": 3.264527779873608e-06,
+      "loss": 0.4473,
+      "step": 18000
+    },
+    {
+      "epoch": 2.96298226605039,
+      "grad_norm": 0.3710954399446858,
+      "learning_rate": 3.2640757860149335e-06,
+      "loss": 0.4315,
+      "step": 18001
+    },
+    {
+      "epoch": 2.9631468502695064,
+      "grad_norm": 0.2933739663224296,
+      "learning_rate": 3.263623805642171e-06,
+      "loss": 0.4374,
+      "step": 18002
+    },
+    {
+      "epoch": 2.963311434488623,
+      "grad_norm": 0.32381176287437485,
+      "learning_rate": 3.263171838760254e-06,
+      "loss": 0.4303,
+      "step": 18003
+    },
+    {
+      "epoch": 2.9634760187077394,
+      "grad_norm": 0.33407116028706746,
+      "learning_rate": 3.26271988537411e-06,
+      "loss": 0.4459,
+      "step": 18004
+    },
+    {
+      "epoch": 2.963640602926856,
+      "grad_norm": 0.35993035731713136,
+      "learning_rate": 3.2622679454886734e-06,
+      "loss": 0.4172,
+      "step": 18005
+    },
+    {
+      "epoch": 2.9638051871459723,
+      "grad_norm": 0.2500691271741941,
+      "learning_rate": 3.2618160191088743e-06,
+      "loss": 0.4284,
+      "step": 18006
+    },
+    {
+      "epoch": 2.9639697713650888,
+      "grad_norm": 0.28256234629310606,
+      "learning_rate": 3.2613641062396454e-06,
+      "loss": 0.4517,
+      "step": 18007
+    },
+    {
+      "epoch": 2.9641343555842052,
+      "grad_norm": 0.449253121128817,
+      "learning_rate": 3.2609122068859164e-06,
+      "loss": 0.4362,
+      "step": 18008
+    },
+    {
+      "epoch": 2.9642989398033217,
+      "grad_norm": 0.2832709344301676,
+      "learning_rate": 3.2604603210526197e-06,
+      "loss": 0.4367,
+      "step": 18009
+    },
+    {
+      "epoch": 2.964463524022438,
+      "grad_norm": 0.8754438257605548,
+      "learning_rate": 3.2600084487446854e-06,
+      "loss": 0.4173,
+      "step": 18010
+    },
+    {
+      "epoch": 2.9646281082415546,
+      "grad_norm": 0.388661775026451,
+      "learning_rate": 3.259556589967043e-06,
+      "loss": 0.4363,
+      "step": 18011
+    },
+    {
+      "epoch": 2.964792692460671,
+      "grad_norm": 0.317062220781525,
+      "learning_rate": 3.259104744724626e-06,
+      "loss": 0.4336,
+      "step": 18012
+    },
+    {
+      "epoch": 2.9649572766797876,
+      "grad_norm": 0.3325180440188855,
+      "learning_rate": 3.2586529130223596e-06,
+      "loss": 0.4218,
+      "step": 18013
+    },
+    {
+      "epoch": 2.965121860898904,
+      "grad_norm": 0.42725282445783,
+      "learning_rate": 3.25820109486518e-06,
+      "loss": 0.4465,
+      "step": 18014
+    },
+    {
+      "epoch": 2.9652864451180205,
+      "grad_norm": 0.2973333427472893,
+      "learning_rate": 3.257749290258012e-06,
+      "loss": 0.4088,
+      "step": 18015
+    },
+    {
+      "epoch": 2.965451029337137,
+      "grad_norm": 0.29434542873997904,
+      "learning_rate": 3.2572974992057905e-06,
+      "loss": 0.4318,
+      "step": 18016
+    },
+    {
+      "epoch": 2.9656156135562535,
+      "grad_norm": 0.3345839461462118,
+      "learning_rate": 3.2568457217134415e-06,
+      "loss": 0.4231,
+      "step": 18017
+    },
+    {
+      "epoch": 2.96578019777537,
+      "grad_norm": 0.851610140823981,
+      "learning_rate": 3.2563939577858965e-06,
+      "loss": 0.4497,
+      "step": 18018
+    },
+    {
+      "epoch": 2.9659447819944864,
+      "grad_norm": 0.3371965846051764,
+      "learning_rate": 3.2559422074280836e-06,
+      "loss": 0.4251,
+      "step": 18019
+    },
+    {
+      "epoch": 2.966109366213603,
+      "grad_norm": 0.33373839748298434,
+      "learning_rate": 3.2554904706449325e-06,
+      "loss": 0.428,
+      "step": 18020
+    },
+    {
+      "epoch": 2.9662739504327194,
+      "grad_norm": 0.3017815513825485,
+      "learning_rate": 3.255038747441374e-06,
+      "loss": 0.4271,
+      "step": 18021
+    },
+    {
+      "epoch": 2.966438534651836,
+      "grad_norm": 0.44310968910509685,
+      "learning_rate": 3.254587037822334e-06,
+      "loss": 0.4318,
+      "step": 18022
+    },
+    {
+      "epoch": 2.9666031188709523,
+      "grad_norm": 0.3164345259894699,
+      "learning_rate": 3.254135341792746e-06,
+      "loss": 0.4194,
+      "step": 18023
+    },
+    {
+      "epoch": 2.9667677030900688,
+      "grad_norm": 0.3420479482428773,
+      "learning_rate": 3.253683659357532e-06,
+      "loss": 0.4386,
+      "step": 18024
+    },
+    {
+      "epoch": 2.9669322873091852,
+      "grad_norm": 0.2682684443609395,
+      "learning_rate": 3.253231990521628e-06,
+      "loss": 0.4296,
+      "step": 18025
+    },
+    {
+      "epoch": 2.9670968715283017,
+      "grad_norm": 0.2814989143324126,
+      "learning_rate": 3.252780335289958e-06,
+      "loss": 0.4304,
+      "step": 18026
+    },
+    {
+      "epoch": 2.967261455747418,
+      "grad_norm": 0.33286025813902886,
+      "learning_rate": 3.25232869366745e-06,
+      "loss": 0.4282,
+      "step": 18027
+    },
+    {
+      "epoch": 2.9674260399665346,
+      "grad_norm": 0.2880496410360446,
+      "learning_rate": 3.2518770656590347e-06,
+      "loss": 0.437,
+      "step": 18028
+    },
+    {
+      "epoch": 2.967590624185651,
+      "grad_norm": 0.3780400505630464,
+      "learning_rate": 3.2514254512696377e-06,
+      "loss": 0.4435,
+      "step": 18029
+    },
+    {
+      "epoch": 2.9677552084047676,
+      "grad_norm": 0.33580026727556184,
+      "learning_rate": 3.2509738505041884e-06,
+      "loss": 0.4404,
+      "step": 18030
+    },
+    {
+      "epoch": 2.967919792623884,
+      "grad_norm": 0.3335263603656979,
+      "learning_rate": 3.250522263367613e-06,
+      "loss": 0.4098,
+      "step": 18031
+    },
+    {
+      "epoch": 2.9680843768430005,
+      "grad_norm": 0.32127404286546213,
+      "learning_rate": 3.2500706898648417e-06,
+      "loss": 0.4358,
+      "step": 18032
+    },
+    {
+      "epoch": 2.968248961062117,
+      "grad_norm": 0.3090836506465173,
+      "learning_rate": 3.2496191300008004e-06,
+      "loss": 0.4251,
+      "step": 18033
+    },
+    {
+      "epoch": 2.9684135452812335,
+      "grad_norm": 0.3899512882733559,
+      "learning_rate": 3.249167583780414e-06,
+      "loss": 0.4194,
+      "step": 18034
+    },
+    {
+      "epoch": 2.96857812950035,
+      "grad_norm": 0.29893313967120105,
+      "learning_rate": 3.2487160512086143e-06,
+      "loss": 0.4246,
+      "step": 18035
+    },
+    {
+      "epoch": 2.9687427137194664,
+      "grad_norm": 0.6207934829357318,
+      "learning_rate": 3.248264532290323e-06,
+      "loss": 0.4358,
+      "step": 18036
+    },
+    {
+      "epoch": 2.968907297938583,
+      "grad_norm": 0.53899296820925,
+      "learning_rate": 3.2478130270304716e-06,
+      "loss": 0.4318,
+      "step": 18037
+    },
+    {
+      "epoch": 2.9690718821576993,
+      "grad_norm": 0.27980429461566236,
+      "learning_rate": 3.247361535433983e-06,
+      "loss": 0.4299,
+      "step": 18038
+    },
+    {
+      "epoch": 2.969236466376816,
+      "grad_norm": 0.3590153093987848,
+      "learning_rate": 3.2469100575057864e-06,
+      "loss": 0.4175,
+      "step": 18039
+    },
+    {
+      "epoch": 2.9694010505959323,
+      "grad_norm": 0.3317868256936113,
+      "learning_rate": 3.2464585932508075e-06,
+      "loss": 0.4384,
+      "step": 18040
+    },
+    {
+      "epoch": 2.9695656348150488,
+      "grad_norm": 1.0971192220495734,
+      "learning_rate": 3.246007142673971e-06,
+      "loss": 0.4354,
+      "step": 18041
+    },
+    {
+      "epoch": 2.969730219034165,
+      "grad_norm": 0.35863929218581314,
+      "learning_rate": 3.2455557057802063e-06,
+      "loss": 0.438,
+      "step": 18042
+    },
+    {
+      "epoch": 2.9698948032532813,
+      "grad_norm": 0.3274544277696532,
+      "learning_rate": 3.2451042825744333e-06,
+      "loss": 0.4466,
+      "step": 18043
+    },
+    {
+      "epoch": 2.9700593874723977,
+      "grad_norm": 0.3313034653954024,
+      "learning_rate": 3.2446528730615857e-06,
+      "loss": 0.4321,
+      "step": 18044
+    },
+    {
+      "epoch": 2.970223971691514,
+      "grad_norm": 0.30389316188656185,
+      "learning_rate": 3.2442014772465814e-06,
+      "loss": 0.4462,
+      "step": 18045
+    },
+    {
+      "epoch": 2.9703885559106307,
+      "grad_norm": 0.2934094414626938,
+      "learning_rate": 3.243750095134352e-06,
+      "loss": 0.4343,
+      "step": 18046
+    },
+    {
+      "epoch": 2.970553140129747,
+      "grad_norm": 0.34425230910755483,
+      "learning_rate": 3.24329872672982e-06,
+      "loss": 0.4098,
+      "step": 18047
+    },
+    {
+      "epoch": 2.9707177243488636,
+      "grad_norm": 0.30450840666561735,
+      "learning_rate": 3.2428473720379085e-06,
+      "loss": 0.4053,
+      "step": 18048
+    },
+    {
+      "epoch": 2.97088230856798,
+      "grad_norm": 0.41018405150240905,
+      "learning_rate": 3.242396031063547e-06,
+      "loss": 0.4099,
+      "step": 18049
+    },
+    {
+      "epoch": 2.9710468927870965,
+      "grad_norm": 0.29096578844948423,
+      "learning_rate": 3.241944703811657e-06,
+      "loss": 0.4463,
+      "step": 18050
+    },
+    {
+      "epoch": 2.971211477006213,
+      "grad_norm": 0.31110995772814504,
+      "learning_rate": 3.2414933902871647e-06,
+      "loss": 0.4294,
+      "step": 18051
+    },
+    {
+      "epoch": 2.9713760612253295,
+      "grad_norm": 0.28542042993102296,
+      "learning_rate": 3.2410420904949933e-06,
+      "loss": 0.4252,
+      "step": 18052
+    },
+    {
+      "epoch": 2.971540645444446,
+      "grad_norm": 0.33471298146038364,
+      "learning_rate": 3.2405908044400692e-06,
+      "loss": 0.434,
+      "step": 18053
+    },
+    {
+      "epoch": 2.9717052296635624,
+      "grad_norm": 0.2776863662421888,
+      "learning_rate": 3.2401395321273175e-06,
+      "loss": 0.4107,
+      "step": 18054
+    },
+    {
+      "epoch": 2.971869813882679,
+      "grad_norm": 0.31484487462523253,
+      "learning_rate": 3.2396882735616576e-06,
+      "loss": 0.4328,
+      "step": 18055
+    },
+    {
+      "epoch": 2.9720343981017954,
+      "grad_norm": 0.325430290067415,
+      "learning_rate": 3.2392370287480172e-06,
+      "loss": 0.4164,
+      "step": 18056
+    },
+    {
+      "epoch": 2.972198982320912,
+      "grad_norm": 0.28709782951035795,
+      "learning_rate": 3.238785797691318e-06,
+      "loss": 0.429,
+      "step": 18057
+    },
+    {
+      "epoch": 2.9723635665400283,
+      "grad_norm": 0.3961152654109752,
+      "learning_rate": 3.238334580396486e-06,
+      "loss": 0.4247,
+      "step": 18058
+    },
+    {
+      "epoch": 2.972528150759145,
+      "grad_norm": 0.473927739825349,
+      "learning_rate": 3.2378833768684428e-06,
+      "loss": 0.432,
+      "step": 18059
+    },
+    {
+      "epoch": 2.9726927349782613,
+      "grad_norm": 0.2904381855630161,
+      "learning_rate": 3.2374321871121134e-06,
+      "loss": 0.4234,
+      "step": 18060
+    },
+    {
+      "epoch": 2.9728573191973777,
+      "grad_norm": 0.33940041419371153,
+      "learning_rate": 3.236981011132419e-06,
+      "loss": 0.4241,
+      "step": 18061
+    },
+    {
+      "epoch": 2.973021903416494,
+      "grad_norm": 0.39408484606284216,
+      "learning_rate": 3.236529848934285e-06,
+      "loss": 0.4076,
+      "step": 18062
+    },
+    {
+      "epoch": 2.9731864876356107,
+      "grad_norm": 0.33323138770900007,
+      "learning_rate": 3.2360787005226334e-06,
+      "loss": 0.4065,
+      "step": 18063
+    },
+    {
+      "epoch": 2.973351071854727,
+      "grad_norm": 0.4534485413101231,
+      "learning_rate": 3.2356275659023844e-06,
+      "loss": 0.4268,
+      "step": 18064
+    },
+    {
+      "epoch": 2.9735156560738436,
+      "grad_norm": 0.30028391006421495,
+      "learning_rate": 3.2351764450784653e-06,
+      "loss": 0.4139,
+      "step": 18065
+    },
+    {
+      "epoch": 2.97368024029296,
+      "grad_norm": 0.3178120219809389,
+      "learning_rate": 3.2347253380557944e-06,
+      "loss": 0.4408,
+      "step": 18066
+    },
+    {
+      "epoch": 2.973844824512076,
+      "grad_norm": 0.39134603889554914,
+      "learning_rate": 3.2342742448392964e-06,
+      "loss": 0.4269,
+      "step": 18067
+    },
+    {
+      "epoch": 2.9740094087311926,
+      "grad_norm": 0.2806020550149464,
+      "learning_rate": 3.2338231654338913e-06,
+      "loss": 0.4288,
+      "step": 18068
+    },
+    {
+      "epoch": 2.974173992950309,
+      "grad_norm": 0.34966794647504806,
+      "learning_rate": 3.2333720998445036e-06,
+      "loss": 0.4295,
+      "step": 18069
+    },
+    {
+      "epoch": 2.9743385771694255,
+      "grad_norm": 0.30994245389032565,
+      "learning_rate": 3.2329210480760544e-06,
+      "loss": 0.4247,
+      "step": 18070
+    },
+    {
+      "epoch": 2.974503161388542,
+      "grad_norm": 0.3643813783368933,
+      "learning_rate": 3.2324700101334647e-06,
+      "loss": 0.4254,
+      "step": 18071
+    },
+    {
+      "epoch": 2.9746677456076585,
+      "grad_norm": 0.4606551678610015,
+      "learning_rate": 3.2320189860216563e-06,
+      "loss": 0.4257,
+      "step": 18072
+    },
+    {
+      "epoch": 2.974832329826775,
+      "grad_norm": 0.31573605145434186,
+      "learning_rate": 3.2315679757455496e-06,
+      "loss": 0.4299,
+      "step": 18073
+    },
+    {
+      "epoch": 2.9749969140458914,
+      "grad_norm": 0.33672582474808266,
+      "learning_rate": 3.2311169793100697e-06,
+      "loss": 0.4219,
+      "step": 18074
+    },
+    {
+      "epoch": 2.975161498265008,
+      "grad_norm": 0.3283598533366402,
+      "learning_rate": 3.230665996720131e-06,
+      "loss": 0.4217,
+      "step": 18075
+    },
+    {
+      "epoch": 2.9753260824841243,
+      "grad_norm": 0.27162495539309256,
+      "learning_rate": 3.2302150279806625e-06,
+      "loss": 0.4196,
+      "step": 18076
+    },
+    {
+      "epoch": 2.975490666703241,
+      "grad_norm": 0.4643539845857057,
+      "learning_rate": 3.2297640730965787e-06,
+      "loss": 0.4439,
+      "step": 18077
+    },
+    {
+      "epoch": 2.9756552509223573,
+      "grad_norm": 0.368683998694101,
+      "learning_rate": 3.229313132072802e-06,
+      "loss": 0.4451,
+      "step": 18078
+    },
+    {
+      "epoch": 2.9758198351414737,
+      "grad_norm": 0.3074675786557951,
+      "learning_rate": 3.228862204914254e-06,
+      "loss": 0.4114,
+      "step": 18079
+    },
+    {
+      "epoch": 2.97598441936059,
+      "grad_norm": 0.34895046305914723,
+      "learning_rate": 3.228411291625854e-06,
+      "loss": 0.4456,
+      "step": 18080
+    },
+    {
+      "epoch": 2.9761490035797067,
+      "grad_norm": 0.37398568795790305,
+      "learning_rate": 3.2279603922125224e-06,
+      "loss": 0.4327,
+      "step": 18081
+    },
+    {
+      "epoch": 2.976313587798823,
+      "grad_norm": 0.3974813545686356,
+      "learning_rate": 3.2275095066791794e-06,
+      "loss": 0.4314,
+      "step": 18082
+    },
+    {
+      "epoch": 2.9764781720179396,
+      "grad_norm": 0.4649457248662648,
+      "learning_rate": 3.2270586350307456e-06,
+      "loss": 0.4399,
+      "step": 18083
+    },
+    {
+      "epoch": 2.976642756237056,
+      "grad_norm": 0.3309637999187168,
+      "learning_rate": 3.2266077772721406e-06,
+      "loss": 0.4189,
+      "step": 18084
+    },
+    {
+      "epoch": 2.9768073404561726,
+      "grad_norm": 0.30971886634736673,
+      "learning_rate": 3.2261569334082805e-06,
+      "loss": 0.4194,
+      "step": 18085
+    },
+    {
+      "epoch": 2.976971924675289,
+      "grad_norm": 0.281886081834552,
+      "learning_rate": 3.2257061034440897e-06,
+      "loss": 0.4336,
+      "step": 18086
+    },
+    {
+      "epoch": 2.9771365088944055,
+      "grad_norm": 0.3450071004495589,
+      "learning_rate": 3.2252552873844843e-06,
+      "loss": 0.4452,
+      "step": 18087
+    },
+    {
+      "epoch": 2.977301093113522,
+      "grad_norm": 0.35518661833333515,
+      "learning_rate": 3.224804485234385e-06,
+      "loss": 0.4484,
+      "step": 18088
+    },
+    {
+      "epoch": 2.9774656773326384,
+      "grad_norm": 0.28753201503866177,
+      "learning_rate": 3.2243536969987096e-06,
+      "loss": 0.4397,
+      "step": 18089
+    },
+    {
+      "epoch": 2.977630261551755,
+      "grad_norm": 0.3600946230149472,
+      "learning_rate": 3.2239029226823786e-06,
+      "loss": 0.435,
+      "step": 18090
+    },
+    {
+      "epoch": 2.9777948457708714,
+      "grad_norm": 0.2928849477916003,
+      "learning_rate": 3.223452162290309e-06,
+      "loss": 0.4306,
+      "step": 18091
+    },
+    {
+      "epoch": 2.977959429989988,
+      "grad_norm": 0.33409587075416086,
+      "learning_rate": 3.22300141582742e-06,
+      "loss": 0.4141,
+      "step": 18092
+    },
+    {
+      "epoch": 2.9781240142091043,
+      "grad_norm": 0.36276975155674085,
+      "learning_rate": 3.2225506832986306e-06,
+      "loss": 0.433,
+      "step": 18093
+    },
+    {
+      "epoch": 2.978288598428221,
+      "grad_norm": 0.3506459464381036,
+      "learning_rate": 3.222099964708856e-06,
+      "loss": 0.4429,
+      "step": 18094
+    },
+    {
+      "epoch": 2.9784531826473373,
+      "grad_norm": 0.2919003799206049,
+      "learning_rate": 3.22164926006302e-06,
+      "loss": 0.4616,
+      "step": 18095
+    },
+    {
+      "epoch": 2.9786177668664537,
+      "grad_norm": 0.2964169570888546,
+      "learning_rate": 3.2211985693660334e-06,
+      "loss": 0.4249,
+      "step": 18096
+    },
+    {
+      "epoch": 2.97878235108557,
+      "grad_norm": 0.6718651471413585,
+      "learning_rate": 3.220747892622821e-06,
+      "loss": 0.4115,
+      "step": 18097
+    },
+    {
+      "epoch": 2.9789469353046867,
+      "grad_norm": 0.3263407815241886,
+      "learning_rate": 3.2202972298382957e-06,
+      "loss": 0.4261,
+      "step": 18098
+    },
+    {
+      "epoch": 2.979111519523803,
+      "grad_norm": 0.2992958976602193,
+      "learning_rate": 3.219846581017375e-06,
+      "loss": 0.4321,
+      "step": 18099
+    },
+    {
+      "epoch": 2.9792761037429196,
+      "grad_norm": 0.3037050417576219,
+      "learning_rate": 3.219395946164979e-06,
+      "loss": 0.4311,
+      "step": 18100
+    },
+    {
+      "epoch": 2.979440687962036,
+      "grad_norm": 0.33330487575134055,
+      "learning_rate": 3.218945325286023e-06,
+      "loss": 0.4412,
+      "step": 18101
+    },
+    {
+      "epoch": 2.9796052721811526,
+      "grad_norm": 0.39981885883986784,
+      "learning_rate": 3.2184947183854244e-06,
+      "loss": 0.4415,
+      "step": 18102
+    },
+    {
+      "epoch": 2.979769856400269,
+      "grad_norm": 0.3282863712935466,
+      "learning_rate": 3.218044125468099e-06,
+      "loss": 0.4143,
+      "step": 18103
+    },
+    {
+      "epoch": 2.9799344406193855,
+      "grad_norm": 0.37273799725123524,
+      "learning_rate": 3.217593546538967e-06,
+      "loss": 0.4079,
+      "step": 18104
+    },
+    {
+      "epoch": 2.980099024838502,
+      "grad_norm": 0.3147167169465058,
+      "learning_rate": 3.2171429816029397e-06,
+      "loss": 0.4171,
+      "step": 18105
+    },
+    {
+      "epoch": 2.9802636090576184,
+      "grad_norm": 0.3325195855156046,
+      "learning_rate": 3.2166924306649386e-06,
+      "loss": 0.4365,
+      "step": 18106
+    },
+    {
+      "epoch": 2.980428193276735,
+      "grad_norm": 0.32653489583676465,
+      "learning_rate": 3.216241893729876e-06,
+      "loss": 0.4246,
+      "step": 18107
+    },
+    {
+      "epoch": 2.980592777495851,
+      "grad_norm": 0.3390616252801089,
+      "learning_rate": 3.2157913708026696e-06,
+      "loss": 0.4163,
+      "step": 18108
+    },
+    {
+      "epoch": 2.9807573617149674,
+      "grad_norm": 0.27059271621322145,
+      "learning_rate": 3.215340861888235e-06,
+      "loss": 0.4383,
+      "step": 18109
+    },
+    {
+      "epoch": 2.980921945934084,
+      "grad_norm": 0.30289374970872623,
+      "learning_rate": 3.2148903669914884e-06,
+      "loss": 0.4012,
+      "step": 18110
+    },
+    {
+      "epoch": 2.9810865301532004,
+      "grad_norm": 0.41009121277019245,
+      "learning_rate": 3.214439886117346e-06,
+      "loss": 0.4446,
+      "step": 18111
+    },
+    {
+      "epoch": 2.981251114372317,
+      "grad_norm": 0.3620401172185879,
+      "learning_rate": 3.213989419270721e-06,
+      "loss": 0.4313,
+      "step": 18112
+    },
+    {
+      "epoch": 2.9814156985914333,
+      "grad_norm": 0.3128518314458058,
+      "learning_rate": 3.2135389664565316e-06,
+      "loss": 0.4273,
+      "step": 18113
+    },
+    {
+      "epoch": 2.9815802828105498,
+      "grad_norm": 0.5072903575923572,
+      "learning_rate": 3.2130885276796925e-06,
+      "loss": 0.4363,
+      "step": 18114
+    },
+    {
+      "epoch": 2.9817448670296662,
+      "grad_norm": 0.3272357301115453,
+      "learning_rate": 3.2126381029451146e-06,
+      "loss": 0.4381,
+      "step": 18115
+    },
+    {
+      "epoch": 2.9819094512487827,
+      "grad_norm": 0.2878470990612326,
+      "learning_rate": 3.2121876922577193e-06,
+      "loss": 0.4424,
+      "step": 18116
+    },
+    {
+      "epoch": 2.982074035467899,
+      "grad_norm": 0.3381122306172689,
+      "learning_rate": 3.2117372956224154e-06,
+      "loss": 0.427,
+      "step": 18117
+    },
+    {
+      "epoch": 2.9822386196870156,
+      "grad_norm": 0.36845413810216926,
+      "learning_rate": 3.2112869130441214e-06,
+      "loss": 0.4313,
+      "step": 18118
+    },
+    {
+      "epoch": 2.982403203906132,
+      "grad_norm": 0.3289819483762642,
+      "learning_rate": 3.2108365445277496e-06,
+      "loss": 0.4278,
+      "step": 18119
+    },
+    {
+      "epoch": 2.9825677881252486,
+      "grad_norm": 0.35928315823892576,
+      "learning_rate": 3.210386190078215e-06,
+      "loss": 0.4373,
+      "step": 18120
+    },
+    {
+      "epoch": 2.982732372344365,
+      "grad_norm": 0.6564359798749756,
+      "learning_rate": 3.2099358497004318e-06,
+      "loss": 0.4324,
+      "step": 18121
+    },
+    {
+      "epoch": 2.9828969565634815,
+      "grad_norm": 0.31058811923763524,
+      "learning_rate": 3.2094855233993134e-06,
+      "loss": 0.4344,
+      "step": 18122
+    },
+    {
+      "epoch": 2.983061540782598,
+      "grad_norm": 0.2963960899838568,
+      "learning_rate": 3.2090352111797756e-06,
+      "loss": 0.4362,
+      "step": 18123
+    },
+    {
+      "epoch": 2.9832261250017145,
+      "grad_norm": 0.32776813298408725,
+      "learning_rate": 3.2085849130467273e-06,
+      "loss": 0.4204,
+      "step": 18124
+    },
+    {
+      "epoch": 2.983390709220831,
+      "grad_norm": 0.5110102874344136,
+      "learning_rate": 3.2081346290050877e-06,
+      "loss": 0.4203,
+      "step": 18125
+    },
+    {
+      "epoch": 2.9835552934399474,
+      "grad_norm": 0.30187833255750773,
+      "learning_rate": 3.2076843590597653e-06,
+      "loss": 0.4406,
+      "step": 18126
+    },
+    {
+      "epoch": 2.983719877659064,
+      "grad_norm": 0.2855605888514474,
+      "learning_rate": 3.207234103215677e-06,
+      "loss": 0.4084,
+      "step": 18127
+    },
+    {
+      "epoch": 2.9838844618781803,
+      "grad_norm": 0.32957430467355686,
+      "learning_rate": 3.206783861477735e-06,
+      "loss": 0.4372,
+      "step": 18128
+    },
+    {
+      "epoch": 2.984049046097297,
+      "grad_norm": 0.38975332516752953,
+      "learning_rate": 3.2063336338508493e-06,
+      "loss": 0.421,
+      "step": 18129
+    },
+    {
+      "epoch": 2.9842136303164133,
+      "grad_norm": 0.2980585596483328,
+      "learning_rate": 3.2058834203399356e-06,
+      "loss": 0.4303,
+      "step": 18130
+    },
+    {
+      "epoch": 2.9843782145355298,
+      "grad_norm": 0.3398236694925661,
+      "learning_rate": 3.205433220949905e-06,
+      "loss": 0.4294,
+      "step": 18131
+    },
+    {
+      "epoch": 2.9845427987546462,
+      "grad_norm": 0.26269951204573494,
+      "learning_rate": 3.2049830356856713e-06,
+      "loss": 0.4531,
+      "step": 18132
+    },
+    {
+      "epoch": 2.9847073829737623,
+      "grad_norm": 0.36395240545705293,
+      "learning_rate": 3.204532864552145e-06,
+      "loss": 0.4328,
+      "step": 18133
+    },
+    {
+      "epoch": 2.9848719671928787,
+      "grad_norm": 0.4336705652596672,
+      "learning_rate": 3.20408270755424e-06,
+      "loss": 0.4283,
+      "step": 18134
+    },
+    {
+      "epoch": 2.985036551411995,
+      "grad_norm": 0.3006524868559794,
+      "learning_rate": 3.203632564696868e-06,
+      "loss": 0.4013,
+      "step": 18135
+    },
+    {
+      "epoch": 2.9852011356311117,
+      "grad_norm": 0.322686194402558,
+      "learning_rate": 3.2031824359849384e-06,
+      "loss": 0.3927,
+      "step": 18136
+    },
+    {
+      "epoch": 2.985365719850228,
+      "grad_norm": 0.3043040631434431,
+      "learning_rate": 3.202732321423365e-06,
+      "loss": 0.4276,
+      "step": 18137
+    },
+    {
+      "epoch": 2.9855303040693446,
+      "grad_norm": 0.3084164121969129,
+      "learning_rate": 3.202282221017058e-06,
+      "loss": 0.4297,
+      "step": 18138
+    },
+    {
+      "epoch": 2.985694888288461,
+      "grad_norm": 0.35498579787356155,
+      "learning_rate": 3.2018321347709307e-06,
+      "loss": 0.4414,
+      "step": 18139
+    },
+    {
+      "epoch": 2.9858594725075776,
+      "grad_norm": 0.27871479742328575,
+      "learning_rate": 3.201382062689892e-06,
+      "loss": 0.429,
+      "step": 18140
+    },
+    {
+      "epoch": 2.986024056726694,
+      "grad_norm": 0.3174931162113501,
+      "learning_rate": 3.2009320047788546e-06,
+      "loss": 0.4191,
+      "step": 18141
+    },
+    {
+      "epoch": 2.9861886409458105,
+      "grad_norm": 0.6164069800922533,
+      "learning_rate": 3.2004819610427284e-06,
+      "loss": 0.4144,
+      "step": 18142
+    },
+    {
+      "epoch": 2.986353225164927,
+      "grad_norm": 0.5379883698054649,
+      "learning_rate": 3.2000319314864245e-06,
+      "loss": 0.4237,
+      "step": 18143
+    },
+    {
+      "epoch": 2.9865178093840434,
+      "grad_norm": 0.28452726517508903,
+      "learning_rate": 3.1995819161148554e-06,
+      "loss": 0.4254,
+      "step": 18144
+    },
+    {
+      "epoch": 2.98668239360316,
+      "grad_norm": 0.3020895643523373,
+      "learning_rate": 3.1991319149329263e-06,
+      "loss": 0.4359,
+      "step": 18145
+    },
+    {
+      "epoch": 2.9868469778222764,
+      "grad_norm": 0.35551300631374627,
+      "learning_rate": 3.1986819279455528e-06,
+      "loss": 0.4462,
+      "step": 18146
+    },
+    {
+      "epoch": 2.987011562041393,
+      "grad_norm": 0.2904016493322226,
+      "learning_rate": 3.198231955157641e-06,
+      "loss": 0.414,
+      "step": 18147
+    },
+    {
+      "epoch": 2.9871761462605093,
+      "grad_norm": 0.35154727598182106,
+      "learning_rate": 3.1977819965741037e-06,
+      "loss": 0.4332,
+      "step": 18148
+    },
+    {
+      "epoch": 2.987340730479626,
+      "grad_norm": 0.33008611051504305,
+      "learning_rate": 3.1973320521998486e-06,
+      "loss": 0.4413,
+      "step": 18149
+    },
+    {
+      "epoch": 2.9875053146987423,
+      "grad_norm": 0.2835558509576164,
+      "learning_rate": 3.1968821220397872e-06,
+      "loss": 0.4272,
+      "step": 18150
+    },
+    {
+      "epoch": 2.9876698989178587,
+      "grad_norm": 0.3609501104969594,
+      "learning_rate": 3.1964322060988277e-06,
+      "loss": 0.4495,
+      "step": 18151
+    },
+    {
+      "epoch": 2.987834483136975,
+      "grad_norm": 0.2888185844920322,
+      "learning_rate": 3.1959823043818795e-06,
+      "loss": 0.4171,
+      "step": 18152
+    },
+    {
+      "epoch": 2.9879990673560917,
+      "grad_norm": 0.2936427475665115,
+      "learning_rate": 3.1955324168938523e-06,
+      "loss": 0.422,
+      "step": 18153
+    },
+    {
+      "epoch": 2.988163651575208,
+      "grad_norm": 0.3308612101987003,
+      "learning_rate": 3.195082543639654e-06,
+      "loss": 0.4351,
+      "step": 18154
+    },
+    {
+      "epoch": 2.9883282357943246,
+      "grad_norm": 0.30759910008070246,
+      "learning_rate": 3.194632684624196e-06,
+      "loss": 0.4394,
+      "step": 18155
+    },
+    {
+      "epoch": 2.988492820013441,
+      "grad_norm": 0.5856876981238213,
+      "learning_rate": 3.1941828398523823e-06,
+      "loss": 0.4548,
+      "step": 18156
+    },
+    {
+      "epoch": 2.9886574042325575,
+      "grad_norm": 0.3194599288209232,
+      "learning_rate": 3.1937330093291276e-06,
+      "loss": 0.4259,
+      "step": 18157
+    },
+    {
+      "epoch": 2.988821988451674,
+      "grad_norm": 0.36234103840466797,
+      "learning_rate": 3.1932831930593354e-06,
+      "loss": 0.4367,
+      "step": 18158
+    },
+    {
+      "epoch": 2.9889865726707905,
+      "grad_norm": 0.30308141678484296,
+      "learning_rate": 3.192833391047915e-06,
+      "loss": 0.4476,
+      "step": 18159
+    },
+    {
+      "epoch": 2.989151156889907,
+      "grad_norm": 0.30146952433299834,
+      "learning_rate": 3.192383603299775e-06,
+      "loss": 0.4143,
+      "step": 18160
+    },
+    {
+      "epoch": 2.9893157411090234,
+      "grad_norm": 0.5577042790559326,
+      "learning_rate": 3.191933829819823e-06,
+      "loss": 0.451,
+      "step": 18161
+    },
+    {
+      "epoch": 2.98948032532814,
+      "grad_norm": 0.35589513547973517,
+      "learning_rate": 3.1914840706129684e-06,
+      "loss": 0.444,
+      "step": 18162
+    },
+    {
+      "epoch": 2.9896449095472564,
+      "grad_norm": 0.3002996637792063,
+      "learning_rate": 3.191034325684116e-06,
+      "loss": 0.4246,
+      "step": 18163
+    },
+    {
+      "epoch": 2.989809493766373,
+      "grad_norm": 0.328292620084996,
+      "learning_rate": 3.1905845950381755e-06,
+      "loss": 0.4256,
+      "step": 18164
+    },
+    {
+      "epoch": 2.9899740779854893,
+      "grad_norm": 0.6209902002277193,
+      "learning_rate": 3.190134878680054e-06,
+      "loss": 0.4308,
+      "step": 18165
+    },
+    {
+      "epoch": 2.990138662204606,
+      "grad_norm": 0.3843180872343453,
+      "learning_rate": 3.1896851766146555e-06,
+      "loss": 0.4423,
+      "step": 18166
+    },
+    {
+      "epoch": 2.9903032464237222,
+      "grad_norm": 0.4509998126064767,
+      "learning_rate": 3.1892354888468922e-06,
+      "loss": 0.4425,
+      "step": 18167
+    },
+    {
+      "epoch": 2.9904678306428387,
+      "grad_norm": 0.29625525143325837,
+      "learning_rate": 3.1887858153816664e-06,
+      "loss": 0.4219,
+      "step": 18168
+    },
+    {
+      "epoch": 2.990632414861955,
+      "grad_norm": 0.45827014828035467,
+      "learning_rate": 3.188336156223887e-06,
+      "loss": 0.4435,
+      "step": 18169
+    },
+    {
+      "epoch": 2.9907969990810717,
+      "grad_norm": 0.37408544422334067,
+      "learning_rate": 3.1878865113784593e-06,
+      "loss": 0.4229,
+      "step": 18170
+    },
+    {
+      "epoch": 2.990961583300188,
+      "grad_norm": 0.36305505750525263,
+      "learning_rate": 3.1874368808502908e-06,
+      "loss": 0.4315,
+      "step": 18171
+    },
+    {
+      "epoch": 2.9911261675193046,
+      "grad_norm": 0.326156630131236,
+      "learning_rate": 3.186987264644288e-06,
+      "loss": 0.4212,
+      "step": 18172
+    },
+    {
+      "epoch": 2.991290751738421,
+      "grad_norm": 0.2828807883216947,
+      "learning_rate": 3.1865376627653548e-06,
+      "loss": 0.4239,
+      "step": 18173
+    },
+    {
+      "epoch": 2.9914553359575375,
+      "grad_norm": 0.3545049958634287,
+      "learning_rate": 3.1860880752184006e-06,
+      "loss": 0.43,
+      "step": 18174
+    },
+    {
+      "epoch": 2.9916199201766536,
+      "grad_norm": 0.33878685908841927,
+      "learning_rate": 3.1856385020083256e-06,
+      "loss": 0.4352,
+      "step": 18175
+    },
+    {
+      "epoch": 2.99178450439577,
+      "grad_norm": 0.39217285035326416,
+      "learning_rate": 3.185188943140042e-06,
+      "loss": 0.4303,
+      "step": 18176
+    },
+    {
+      "epoch": 2.9919490886148865,
+      "grad_norm": 0.34070472410426383,
+      "learning_rate": 3.1847393986184486e-06,
+      "loss": 0.4486,
+      "step": 18177
+    },
+    {
+      "epoch": 2.992113672834003,
+      "grad_norm": 0.654226350541991,
+      "learning_rate": 3.184289868448457e-06,
+      "loss": 0.42,
+      "step": 18178
+    },
+    {
+      "epoch": 2.9922782570531195,
+      "grad_norm": 0.35024427558591065,
+      "learning_rate": 3.1838403526349687e-06,
+      "loss": 0.4253,
+      "step": 18179
+    },
+    {
+      "epoch": 2.992442841272236,
+      "grad_norm": 0.2887993308092518,
+      "learning_rate": 3.183390851182888e-06,
+      "loss": 0.419,
+      "step": 18180
+    },
+    {
+      "epoch": 2.9926074254913524,
+      "grad_norm": 0.2852910550694381,
+      "learning_rate": 3.182941364097121e-06,
+      "loss": 0.4248,
+      "step": 18181
+    },
+    {
+      "epoch": 2.992772009710469,
+      "grad_norm": 0.31844235323231934,
+      "learning_rate": 3.1824918913825725e-06,
+      "loss": 0.4137,
+      "step": 18182
+    },
+    {
+      "epoch": 2.9929365939295853,
+      "grad_norm": 0.3990404528599743,
+      "learning_rate": 3.1820424330441472e-06,
+      "loss": 0.4466,
+      "step": 18183
+    },
+    {
+      "epoch": 2.993101178148702,
+      "grad_norm": 0.3081020974323921,
+      "learning_rate": 3.1815929890867476e-06,
+      "loss": 0.4373,
+      "step": 18184
+    },
+    {
+      "epoch": 2.9932657623678183,
+      "grad_norm": 0.29220689300256936,
+      "learning_rate": 3.1811435595152807e-06,
+      "loss": 0.4105,
+      "step": 18185
+    },
+    {
+      "epoch": 2.9934303465869347,
+      "grad_norm": 0.38694100904722095,
+      "learning_rate": 3.180694144334647e-06,
+      "loss": 0.4339,
+      "step": 18186
+    },
+    {
+      "epoch": 2.993594930806051,
+      "grad_norm": 0.38125835788420304,
+      "learning_rate": 3.180244743549755e-06,
+      "loss": 0.4457,
+      "step": 18187
+    },
+    {
+      "epoch": 2.9937595150251677,
+      "grad_norm": 1.1024776134142784,
+      "learning_rate": 3.1797953571655043e-06,
+      "loss": 0.4411,
+      "step": 18188
+    },
+    {
+      "epoch": 2.993924099244284,
+      "grad_norm": 0.45183397917621165,
+      "learning_rate": 3.1793459851867988e-06,
+      "loss": 0.4411,
+      "step": 18189
+    },
+    {
+      "epoch": 2.9940886834634006,
+      "grad_norm": 0.32011488526140774,
+      "learning_rate": 3.1788966276185444e-06,
+      "loss": 0.4292,
+      "step": 18190
+    },
+    {
+      "epoch": 2.994253267682517,
+      "grad_norm": 0.5045650725636368,
+      "learning_rate": 3.178447284465641e-06,
+      "loss": 0.4179,
+      "step": 18191
+    },
+    {
+      "epoch": 2.9944178519016336,
+      "grad_norm": 0.5554772919589578,
+      "learning_rate": 3.177997955732995e-06,
+      "loss": 0.4136,
+      "step": 18192
+    },
+    {
+      "epoch": 2.99458243612075,
+      "grad_norm": 0.475211106424764,
+      "learning_rate": 3.177548641425506e-06,
+      "loss": 0.4173,
+      "step": 18193
+    },
+    {
+      "epoch": 2.9947470203398665,
+      "grad_norm": 0.3771283509591505,
+      "learning_rate": 3.1770993415480795e-06,
+      "loss": 0.4452,
+      "step": 18194
+    },
+    {
+      "epoch": 2.994911604558983,
+      "grad_norm": 0.42594227266935014,
+      "learning_rate": 3.1766500561056188e-06,
+      "loss": 0.4117,
+      "step": 18195
+    },
+    {
+      "epoch": 2.9950761887780994,
+      "grad_norm": 0.3180131807086205,
+      "learning_rate": 3.176200785103021e-06,
+      "loss": 0.4285,
+      "step": 18196
+    },
+    {
+      "epoch": 2.995240772997216,
+      "grad_norm": 0.29612484007700896,
+      "learning_rate": 3.175751528545195e-06,
+      "loss": 0.4318,
+      "step": 18197
+    },
+    {
+      "epoch": 2.9954053572163324,
+      "grad_norm": 0.28254929449057237,
+      "learning_rate": 3.1753022864370376e-06,
+      "loss": 0.4345,
+      "step": 18198
+    },
+    {
+      "epoch": 2.995569941435449,
+      "grad_norm": 0.3183231439188339,
+      "learning_rate": 3.1748530587834543e-06,
+      "loss": 0.4349,
+      "step": 18199
+    },
+    {
+      "epoch": 2.995734525654565,
+      "grad_norm": 0.46343260585566537,
+      "learning_rate": 3.1744038455893443e-06,
+      "loss": 0.4357,
+      "step": 18200
+    },
+    {
+      "epoch": 2.9958991098736814,
+      "grad_norm": 0.29529337466275296,
+      "learning_rate": 3.173954646859612e-06,
+      "loss": 0.4184,
+      "step": 18201
+    },
+    {
+      "epoch": 2.996063694092798,
+      "grad_norm": 0.31809070614092383,
+      "learning_rate": 3.173505462599157e-06,
+      "loss": 0.43,
+      "step": 18202
+    },
+    {
+      "epoch": 2.9962282783119143,
+      "grad_norm": 0.30457104675831437,
+      "learning_rate": 3.1730562928128797e-06,
+      "loss": 0.4354,
+      "step": 18203
+    },
+    {
+      "epoch": 2.9963928625310308,
+      "grad_norm": 0.3062917076160698,
+      "learning_rate": 3.1726071375056836e-06,
+      "loss": 0.4087,
+      "step": 18204
+    },
+    {
+      "epoch": 2.9965574467501472,
+      "grad_norm": 0.3124888663990827,
+      "learning_rate": 3.1721579966824684e-06,
+      "loss": 0.4394,
+      "step": 18205
+    },
+    {
+      "epoch": 2.9967220309692637,
+      "grad_norm": 0.33867425297635956,
+      "learning_rate": 3.1717088703481363e-06,
+      "loss": 0.4486,
+      "step": 18206
+    },
+    {
+      "epoch": 2.99688661518838,
+      "grad_norm": 0.273558008512084,
+      "learning_rate": 3.171259758507584e-06,
+      "loss": 0.4071,
+      "step": 18207
+    },
+    {
+      "epoch": 2.9970511994074966,
+      "grad_norm": 0.5285508319253733,
+      "learning_rate": 3.1708106611657184e-06,
+      "loss": 0.3861,
+      "step": 18208
+    },
+    {
+      "epoch": 2.997215783626613,
+      "grad_norm": 0.3128874485399929,
+      "learning_rate": 3.170361578327435e-06,
+      "loss": 0.4164,
+      "step": 18209
+    },
+    {
+      "epoch": 2.9973803678457296,
+      "grad_norm": 0.37256940242292064,
+      "learning_rate": 3.1699125099976353e-06,
+      "loss": 0.4131,
+      "step": 18210
+    },
+    {
+      "epoch": 2.997544952064846,
+      "grad_norm": 0.35543677230722404,
+      "learning_rate": 3.169463456181219e-06,
+      "loss": 0.4285,
+      "step": 18211
+    },
+    {
+      "epoch": 2.9977095362839625,
+      "grad_norm": 0.5258426866304687,
+      "learning_rate": 3.1690144168830863e-06,
+      "loss": 0.4382,
+      "step": 18212
+    },
+    {
+      "epoch": 2.997874120503079,
+      "grad_norm": 0.3970173787688329,
+      "learning_rate": 3.168565392108138e-06,
+      "loss": 0.4528,
+      "step": 18213
+    },
+    {
+      "epoch": 2.9980387047221955,
+      "grad_norm": 0.38818627062660305,
+      "learning_rate": 3.1681163818612722e-06,
+      "loss": 0.4337,
+      "step": 18214
+    },
+    {
+      "epoch": 2.998203288941312,
+      "grad_norm": 0.32909854511156844,
+      "learning_rate": 3.167667386147389e-06,
+      "loss": 0.4244,
+      "step": 18215
+    },
+    {
+      "epoch": 2.9983678731604284,
+      "grad_norm": 0.31824541264496803,
+      "learning_rate": 3.1672184049713886e-06,
+      "loss": 0.4243,
+      "step": 18216
+    },
+    {
+      "epoch": 2.998532457379545,
+      "grad_norm": 0.2944951297577863,
+      "learning_rate": 3.166769438338168e-06,
+      "loss": 0.45,
+      "step": 18217
+    },
+    {
+      "epoch": 2.9986970415986614,
+      "grad_norm": 0.31699040399767714,
+      "learning_rate": 3.1663204862526272e-06,
+      "loss": 0.4246,
+      "step": 18218
+    },
+    {
+      "epoch": 2.998861625817778,
+      "grad_norm": 0.28009816206800014,
+      "learning_rate": 3.1658715487196644e-06,
+      "loss": 0.4535,
+      "step": 18219
+    },
+    {
+      "epoch": 2.9990262100368943,
+      "grad_norm": 0.29611811965362095,
+      "learning_rate": 3.16542262574418e-06,
+      "loss": 0.409,
+      "step": 18220
+    },
+    {
+      "epoch": 2.9991907942560108,
+      "grad_norm": 0.32801166259446557,
+      "learning_rate": 3.16497371733107e-06,
+      "loss": 0.4215,
+      "step": 18221
+    },
+    {
+      "epoch": 2.9993553784751272,
+      "grad_norm": 0.29321722798470556,
+      "learning_rate": 3.1645248234852354e-06,
+      "loss": 0.4182,
+      "step": 18222
+    },
+    {
+      "epoch": 2.9995199626942437,
+      "grad_norm": 0.40426280821449323,
+      "learning_rate": 3.164075944211572e-06,
+      "loss": 0.4137,
+      "step": 18223
+    },
+    {
+      "epoch": 2.99968454691336,
+      "grad_norm": 0.5970825070217444,
+      "learning_rate": 3.1636270795149784e-06,
+      "loss": 0.4313,
+      "step": 18224
+    },
+    {
+      "epoch": 2.9998491311324766,
+      "grad_norm": 0.2644700809917794,
+      "learning_rate": 3.1631782294003545e-06,
+      "loss": 0.4145,
+      "step": 18225
+    },
+    {
+      "epoch": 3.000013715351593,
+      "grad_norm": 0.2995399869875604,
+      "learning_rate": 3.162729393872594e-06,
+      "loss": 0.4123,
+      "step": 18226
+    },
+    {
+      "epoch": 3.0001782995707096,
+      "grad_norm": 0.27034051788182417,
+      "learning_rate": 3.1622805729365987e-06,
+      "loss": 0.4177,
+      "step": 18227
+    },
+    {
+      "epoch": 3.000342883789826,
+      "grad_norm": 0.3971305869673882,
+      "learning_rate": 3.1618317665972613e-06,
+      "loss": 0.4242,
+      "step": 18228
+    },
+    {
+      "epoch": 3.0005074680089425,
+      "grad_norm": 0.3414507516646599,
+      "learning_rate": 3.1613829748594845e-06,
+      "loss": 0.4184,
+      "step": 18229
+    },
+    {
+      "epoch": 3.000672052228059,
+      "grad_norm": 0.3465525713100099,
+      "learning_rate": 3.1609341977281607e-06,
+      "loss": 0.4125,
+      "step": 18230
+    },
+    {
+      "epoch": 3.0008366364471755,
+      "grad_norm": 0.33514919669558996,
+      "learning_rate": 3.160485435208189e-06,
+      "loss": 0.4305,
+      "step": 18231
+    },
+    {
+      "epoch": 3.001001220666292,
+      "grad_norm": 0.33085851251192505,
+      "learning_rate": 3.1600366873044664e-06,
+      "loss": 0.4442,
+      "step": 18232
+    },
+    {
+      "epoch": 3.0011658048854084,
+      "grad_norm": 0.27738310096517826,
+      "learning_rate": 3.159587954021887e-06,
+      "loss": 0.44,
+      "step": 18233
+    },
+    {
+      "epoch": 3.001330389104525,
+      "grad_norm": 0.46496425947877273,
+      "learning_rate": 3.1591392353653504e-06,
+      "loss": 0.4317,
+      "step": 18234
+    },
+    {
+      "epoch": 3.0014949733236413,
+      "grad_norm": 0.5780192641756802,
+      "learning_rate": 3.15869053133975e-06,
+      "loss": 0.4333,
+      "step": 18235
+    },
+    {
+      "epoch": 3.001659557542758,
+      "grad_norm": 0.4204830312536245,
+      "learning_rate": 3.1582418419499853e-06,
+      "loss": 0.4236,
+      "step": 18236
+    },
+    {
+      "epoch": 3.0018241417618743,
+      "grad_norm": 0.31813304116451246,
+      "learning_rate": 3.157793167200947e-06,
+      "loss": 0.4434,
+      "step": 18237
+    },
+    {
+      "epoch": 3.0019887259809903,
+      "grad_norm": 0.3340994695144974,
+      "learning_rate": 3.157344507097537e-06,
+      "loss": 0.4131,
+      "step": 18238
+    },
+    {
+      "epoch": 3.002153310200107,
+      "grad_norm": 0.32359193355644744,
+      "learning_rate": 3.1568958616446466e-06,
+      "loss": 0.4416,
+      "step": 18239
+    },
+    {
+      "epoch": 3.0023178944192233,
+      "grad_norm": 0.3914003964350955,
+      "learning_rate": 3.1564472308471723e-06,
+      "loss": 0.4414,
+      "step": 18240
+    },
+    {
+      "epoch": 3.0024824786383397,
+      "grad_norm": 1.5983605977406183,
+      "learning_rate": 3.1559986147100103e-06,
+      "loss": 0.4182,
+      "step": 18241
+    },
+    {
+      "epoch": 3.002647062857456,
+      "grad_norm": 0.2656196346512441,
+      "learning_rate": 3.155550013238054e-06,
+      "loss": 0.4238,
+      "step": 18242
+    },
+    {
+      "epoch": 3.0028116470765727,
+      "grad_norm": 0.3435030234191996,
+      "learning_rate": 3.1551014264362003e-06,
+      "loss": 0.4419,
+      "step": 18243
+    },
+    {
+      "epoch": 3.002976231295689,
+      "grad_norm": 0.2774345818391581,
+      "learning_rate": 3.1546528543093423e-06,
+      "loss": 0.4331,
+      "step": 18244
+    },
+    {
+      "epoch": 3.0031408155148056,
+      "grad_norm": 0.2795919065919006,
+      "learning_rate": 3.154204296862376e-06,
+      "loss": 0.4321,
+      "step": 18245
+    },
+    {
+      "epoch": 3.003305399733922,
+      "grad_norm": 0.3060950103109964,
+      "learning_rate": 3.1537557541001964e-06,
+      "loss": 0.4324,
+      "step": 18246
+    },
+    {
+      "epoch": 3.0034699839530385,
+      "grad_norm": 0.38995358074078834,
+      "learning_rate": 3.153307226027694e-06,
+      "loss": 0.4382,
+      "step": 18247
+    },
+    {
+      "epoch": 3.003634568172155,
+      "grad_norm": 0.352334026299767,
+      "learning_rate": 3.1528587126497684e-06,
+      "loss": 0.4104,
+      "step": 18248
+    },
+    {
+      "epoch": 3.0037991523912715,
+      "grad_norm": 0.714688159733198,
+      "learning_rate": 3.1524102139713094e-06,
+      "loss": 0.4323,
+      "step": 18249
+    },
+    {
+      "epoch": 3.003963736610388,
+      "grad_norm": 0.3182506796247081,
+      "learning_rate": 3.1519617299972128e-06,
+      "loss": 0.4201,
+      "step": 18250
+    },
+    {
+      "epoch": 3.0041283208295044,
+      "grad_norm": 0.39036458320277834,
+      "learning_rate": 3.151513260732371e-06,
+      "loss": 0.4338,
+      "step": 18251
+    },
+    {
+      "epoch": 3.004292905048621,
+      "grad_norm": 0.2987276384376063,
+      "learning_rate": 3.151064806181679e-06,
+      "loss": 0.4213,
+      "step": 18252
+    },
+    {
+      "epoch": 3.0044574892677374,
+      "grad_norm": 0.3210273705242925,
+      "learning_rate": 3.15061636635003e-06,
+      "loss": 0.4198,
+      "step": 18253
+    },
+    {
+      "epoch": 3.004622073486854,
+      "grad_norm": 0.2906514077196313,
+      "learning_rate": 3.1501679412423154e-06,
+      "loss": 0.4283,
+      "step": 18254
+    },
+    {
+      "epoch": 3.0047866577059703,
+      "grad_norm": 0.325529310036599,
+      "learning_rate": 3.1497195308634315e-06,
+      "loss": 0.4142,
+      "step": 18255
+    },
+    {
+      "epoch": 3.004951241925087,
+      "grad_norm": 0.2876029969384978,
+      "learning_rate": 3.149271135218266e-06,
+      "loss": 0.4263,
+      "step": 18256
+    },
+    {
+      "epoch": 3.0051158261442033,
+      "grad_norm": 0.3009431027312247,
+      "learning_rate": 3.148822754311718e-06,
+      "loss": 0.4323,
+      "step": 18257
+    },
+    {
+      "epoch": 3.0052804103633197,
+      "grad_norm": 1.0456543690184665,
+      "learning_rate": 3.148374388148674e-06,
+      "loss": 0.4425,
+      "step": 18258
+    },
+    {
+      "epoch": 3.005444994582436,
+      "grad_norm": 0.41242992191256006,
+      "learning_rate": 3.147926036734032e-06,
+      "loss": 0.4133,
+      "step": 18259
+    },
+    {
+      "epoch": 3.0056095788015527,
+      "grad_norm": 0.3018174934163969,
+      "learning_rate": 3.147477700072681e-06,
+      "loss": 0.4475,
+      "step": 18260
+    },
+    {
+      "epoch": 3.005774163020669,
+      "grad_norm": 0.30378887389327897,
+      "learning_rate": 3.1470293781695114e-06,
+      "loss": 0.4343,
+      "step": 18261
+    },
+    {
+      "epoch": 3.0059387472397856,
+      "grad_norm": 0.3030320265528302,
+      "learning_rate": 3.1465810710294193e-06,
+      "loss": 0.4224,
+      "step": 18262
+    },
+    {
+      "epoch": 3.006103331458902,
+      "grad_norm": 0.32468997777886277,
+      "learning_rate": 3.146132778657293e-06,
+      "loss": 0.4206,
+      "step": 18263
+    },
+    {
+      "epoch": 3.0062679156780185,
+      "grad_norm": 0.3073859418492667,
+      "learning_rate": 3.145684501058027e-06,
+      "loss": 0.4397,
+      "step": 18264
+    },
+    {
+      "epoch": 3.006432499897135,
+      "grad_norm": 0.3535746497464689,
+      "learning_rate": 3.1452362382365093e-06,
+      "loss": 0.428,
+      "step": 18265
+    },
+    {
+      "epoch": 3.0065970841162515,
+      "grad_norm": 0.3509794888005568,
+      "learning_rate": 3.1447879901976346e-06,
+      "loss": 0.4145,
+      "step": 18266
+    },
+    {
+      "epoch": 3.006761668335368,
+      "grad_norm": 0.37055426427310695,
+      "learning_rate": 3.1443397569462916e-06,
+      "loss": 0.4202,
+      "step": 18267
+    },
+    {
+      "epoch": 3.0069262525544844,
+      "grad_norm": 0.3983158517239644,
+      "learning_rate": 3.143891538487374e-06,
+      "loss": 0.4378,
+      "step": 18268
+    },
+    {
+      "epoch": 3.007090836773601,
+      "grad_norm": 0.8847329921051714,
+      "learning_rate": 3.1434433348257696e-06,
+      "loss": 0.4332,
+      "step": 18269
+    },
+    {
+      "epoch": 3.0072554209927174,
+      "grad_norm": 0.40040346566986273,
+      "learning_rate": 3.1429951459663688e-06,
+      "loss": 0.4313,
+      "step": 18270
+    },
+    {
+      "epoch": 3.0074200052118334,
+      "grad_norm": 0.30438473774909447,
+      "learning_rate": 3.1425469719140652e-06,
+      "loss": 0.4317,
+      "step": 18271
+    },
+    {
+      "epoch": 3.00758458943095,
+      "grad_norm": 0.26880189766547447,
+      "learning_rate": 3.142098812673746e-06,
+      "loss": 0.4123,
+      "step": 18272
+    },
+    {
+      "epoch": 3.0077491736500663,
+      "grad_norm": 0.43929111116823344,
+      "learning_rate": 3.1416506682503034e-06,
+      "loss": 0.4256,
+      "step": 18273
+    },
+    {
+      "epoch": 3.007913757869183,
+      "grad_norm": 0.31012222245409504,
+      "learning_rate": 3.1412025386486254e-06,
+      "loss": 0.4152,
+      "step": 18274
+    },
+    {
+      "epoch": 3.0080783420882993,
+      "grad_norm": 0.3281837024857674,
+      "learning_rate": 3.140754423873605e-06,
+      "loss": 0.4276,
+      "step": 18275
+    },
+    {
+      "epoch": 3.0082429263074157,
+      "grad_norm": 0.359541434317751,
+      "learning_rate": 3.14030632393013e-06,
+      "loss": 0.4311,
+      "step": 18276
+    },
+    {
+      "epoch": 3.008407510526532,
+      "grad_norm": 0.381024987330811,
+      "learning_rate": 3.1398582388230866e-06,
+      "loss": 0.4349,
+      "step": 18277
+    },
+    {
+      "epoch": 3.0085720947456487,
+      "grad_norm": 0.2725960858202017,
+      "learning_rate": 3.1394101685573705e-06,
+      "loss": 0.4377,
+      "step": 18278
+    },
+    {
+      "epoch": 3.008736678964765,
+      "grad_norm": 0.29303313243267587,
+      "learning_rate": 3.1389621131378657e-06,
+      "loss": 0.4281,
+      "step": 18279
+    },
+    {
+      "epoch": 3.0089012631838816,
+      "grad_norm": 0.35231873351827986,
+      "learning_rate": 3.1385140725694644e-06,
+      "loss": 0.4281,
+      "step": 18280
+    },
+    {
+      "epoch": 3.009065847402998,
+      "grad_norm": 0.32507177675599463,
+      "learning_rate": 3.138066046857053e-06,
+      "loss": 0.4141,
+      "step": 18281
+    },
+    {
+      "epoch": 3.0092304316221146,
+      "grad_norm": 0.3090038291024942,
+      "learning_rate": 3.137618036005522e-06,
+      "loss": 0.4278,
+      "step": 18282
+    },
+    {
+      "epoch": 3.009395015841231,
+      "grad_norm": 0.34411886423257904,
+      "learning_rate": 3.137170040019759e-06,
+      "loss": 0.435,
+      "step": 18283
+    },
+    {
+      "epoch": 3.0095596000603475,
+      "grad_norm": 0.3207335840825621,
+      "learning_rate": 3.1367220589046525e-06,
+      "loss": 0.4271,
+      "step": 18284
+    },
+    {
+      "epoch": 3.009724184279464,
+      "grad_norm": 0.394758107332526,
+      "learning_rate": 3.1362740926650915e-06,
+      "loss": 0.423,
+      "step": 18285
+    },
+    {
+      "epoch": 3.0098887684985804,
+      "grad_norm": 0.3173756255656595,
+      "learning_rate": 3.1358261413059622e-06,
+      "loss": 0.4243,
+      "step": 18286
+    },
+    {
+      "epoch": 3.010053352717697,
+      "grad_norm": 0.36759515811796317,
+      "learning_rate": 3.1353782048321555e-06,
+      "loss": 0.4229,
+      "step": 18287
+    },
+    {
+      "epoch": 3.0102179369368134,
+      "grad_norm": 0.3570919253203036,
+      "learning_rate": 3.1349302832485547e-06,
+      "loss": 0.4363,
+      "step": 18288
+    },
+    {
+      "epoch": 3.01038252115593,
+      "grad_norm": 0.44230721314041077,
+      "learning_rate": 3.1344823765600525e-06,
+      "loss": 0.4322,
+      "step": 18289
+    },
+    {
+      "epoch": 3.0105471053750463,
+      "grad_norm": 0.41149678706373655,
+      "learning_rate": 3.1340344847715324e-06,
+      "loss": 0.4149,
+      "step": 18290
+    },
+    {
+      "epoch": 3.010711689594163,
+      "grad_norm": 0.3896154491901678,
+      "learning_rate": 3.133586607887882e-06,
+      "loss": 0.4451,
+      "step": 18291
+    },
+    {
+      "epoch": 3.0108762738132793,
+      "grad_norm": 0.4611229387144769,
+      "learning_rate": 3.1331387459139904e-06,
+      "loss": 0.4311,
+      "step": 18292
+    },
+    {
+      "epoch": 3.0110408580323957,
+      "grad_norm": 0.41381418119138585,
+      "learning_rate": 3.1326908988547425e-06,
+      "loss": 0.4272,
+      "step": 18293
+    },
+    {
+      "epoch": 3.011205442251512,
+      "grad_norm": 0.33006048790236775,
+      "learning_rate": 3.1322430667150263e-06,
+      "loss": 0.4307,
+      "step": 18294
+    },
+    {
+      "epoch": 3.0113700264706287,
+      "grad_norm": 0.2806686050008286,
+      "learning_rate": 3.1317952494997273e-06,
+      "loss": 0.4334,
+      "step": 18295
+    },
+    {
+      "epoch": 3.011534610689745,
+      "grad_norm": 0.303403131668887,
+      "learning_rate": 3.1313474472137334e-06,
+      "loss": 0.4349,
+      "step": 18296
+    },
+    {
+      "epoch": 3.0116991949088616,
+      "grad_norm": 0.36015904202072035,
+      "learning_rate": 3.130899659861931e-06,
+      "loss": 0.4269,
+      "step": 18297
+    },
+    {
+      "epoch": 3.011863779127978,
+      "grad_norm": 0.39213784370577165,
+      "learning_rate": 3.1304518874492027e-06,
+      "loss": 0.4165,
+      "step": 18298
+    },
+    {
+      "epoch": 3.0120283633470946,
+      "grad_norm": 0.4320711827383614,
+      "learning_rate": 3.130004129980438e-06,
+      "loss": 0.4225,
+      "step": 18299
+    },
+    {
+      "epoch": 3.012192947566211,
+      "grad_norm": 0.3721589267015674,
+      "learning_rate": 3.1295563874605203e-06,
+      "loss": 0.4128,
+      "step": 18300
+    },
+    {
+      "epoch": 3.0123575317853275,
+      "grad_norm": 0.38413959309953516,
+      "learning_rate": 3.1291086598943376e-06,
+      "loss": 0.4374,
+      "step": 18301
+    },
+    {
+      "epoch": 3.012522116004444,
+      "grad_norm": 0.41329479123233775,
+      "learning_rate": 3.128660947286773e-06,
+      "loss": 0.4546,
+      "step": 18302
+    },
+    {
+      "epoch": 3.0126867002235604,
+      "grad_norm": 0.2678832781833614,
+      "learning_rate": 3.128213249642714e-06,
+      "loss": 0.4175,
+      "step": 18303
+    },
+    {
+      "epoch": 3.0128512844426765,
+      "grad_norm": 0.35521270889326156,
+      "learning_rate": 3.1277655669670446e-06,
+      "loss": 0.4205,
+      "step": 18304
+    },
+    {
+      "epoch": 3.013015868661793,
+      "grad_norm": 0.3104405779774533,
+      "learning_rate": 3.1273178992646486e-06,
+      "loss": 0.4259,
+      "step": 18305
+    },
+    {
+      "epoch": 3.0131804528809094,
+      "grad_norm": 0.2858893085394464,
+      "learning_rate": 3.1268702465404133e-06,
+      "loss": 0.4198,
+      "step": 18306
+    },
+    {
+      "epoch": 3.013345037100026,
+      "grad_norm": 0.37408477887754,
+      "learning_rate": 3.1264226087992197e-06,
+      "loss": 0.4288,
+      "step": 18307
+    },
+    {
+      "epoch": 3.0135096213191424,
+      "grad_norm": 0.4217822609844656,
+      "learning_rate": 3.1259749860459565e-06,
+      "loss": 0.4422,
+      "step": 18308
+    },
+    {
+      "epoch": 3.013674205538259,
+      "grad_norm": 0.26756539192671847,
+      "learning_rate": 3.125527378285504e-06,
+      "loss": 0.41,
+      "step": 18309
+    },
+    {
+      "epoch": 3.0138387897573753,
+      "grad_norm": 0.47541946698410265,
+      "learning_rate": 3.1250797855227504e-06,
+      "loss": 0.4163,
+      "step": 18310
+    },
+    {
+      "epoch": 3.0140033739764918,
+      "grad_norm": 0.384741342753798,
+      "learning_rate": 3.1246322077625755e-06,
+      "loss": 0.438,
+      "step": 18311
+    },
+    {
+      "epoch": 3.0141679581956082,
+      "grad_norm": 0.28876904645439844,
+      "learning_rate": 3.1241846450098664e-06,
+      "loss": 0.4262,
+      "step": 18312
+    },
+    {
+      "epoch": 3.0143325424147247,
+      "grad_norm": 0.28806536770785424,
+      "learning_rate": 3.1237370972695048e-06,
+      "loss": 0.4164,
+      "step": 18313
+    },
+    {
+      "epoch": 3.014497126633841,
+      "grad_norm": 0.3084635180596369,
+      "learning_rate": 3.1232895645463744e-06,
+      "loss": 0.4196,
+      "step": 18314
+    },
+    {
+      "epoch": 3.0146617108529576,
+      "grad_norm": 0.3054116371458221,
+      "learning_rate": 3.122842046845359e-06,
+      "loss": 0.4368,
+      "step": 18315
+    },
+    {
+      "epoch": 3.014826295072074,
+      "grad_norm": 0.3836994030013055,
+      "learning_rate": 3.1223945441713413e-06,
+      "loss": 0.4339,
+      "step": 18316
+    },
+    {
+      "epoch": 3.0149908792911906,
+      "grad_norm": 0.373344678246291,
+      "learning_rate": 3.1219470565292064e-06,
+      "loss": 0.4271,
+      "step": 18317
+    },
+    {
+      "epoch": 3.015155463510307,
+      "grad_norm": 0.3233689476672488,
+      "learning_rate": 3.121499583923832e-06,
+      "loss": 0.4241,
+      "step": 18318
+    },
+    {
+      "epoch": 3.0153200477294235,
+      "grad_norm": 0.49564569197291564,
+      "learning_rate": 3.121052126360107e-06,
+      "loss": 0.4406,
+      "step": 18319
+    },
+    {
+      "epoch": 3.01548463194854,
+      "grad_norm": 0.46988181103347443,
+      "learning_rate": 3.12060468384291e-06,
+      "loss": 0.4386,
+      "step": 18320
+    },
+    {
+      "epoch": 3.0156492161676565,
+      "grad_norm": 0.4603952940576739,
+      "learning_rate": 3.1201572563771234e-06,
+      "loss": 0.4121,
+      "step": 18321
+    },
+    {
+      "epoch": 3.015813800386773,
+      "grad_norm": 0.3319710266670054,
+      "learning_rate": 3.1197098439676307e-06,
+      "loss": 0.4435,
+      "step": 18322
+    },
+    {
+      "epoch": 3.0159783846058894,
+      "grad_norm": 0.28959240783036866,
+      "learning_rate": 3.1192624466193127e-06,
+      "loss": 0.415,
+      "step": 18323
+    },
+    {
+      "epoch": 3.016142968825006,
+      "grad_norm": 0.2921197590228239,
+      "learning_rate": 3.118815064337052e-06,
+      "loss": 0.4336,
+      "step": 18324
+    },
+    {
+      "epoch": 3.0163075530441223,
+      "grad_norm": 0.29168013057506503,
+      "learning_rate": 3.1183676971257303e-06,
+      "loss": 0.4458,
+      "step": 18325
+    },
+    {
+      "epoch": 3.016472137263239,
+      "grad_norm": 0.3304046783916581,
+      "learning_rate": 3.117920344990229e-06,
+      "loss": 0.4069,
+      "step": 18326
+    },
+    {
+      "epoch": 3.0166367214823553,
+      "grad_norm": 0.3093798804530362,
+      "learning_rate": 3.1174730079354304e-06,
+      "loss": 0.4122,
+      "step": 18327
+    },
+    {
+      "epoch": 3.0168013057014718,
+      "grad_norm": 0.26771936900248855,
+      "learning_rate": 3.117025685966212e-06,
+      "loss": 0.4209,
+      "step": 18328
+    },
+    {
+      "epoch": 3.0169658899205882,
+      "grad_norm": 0.40780476703473845,
+      "learning_rate": 3.11657837908746e-06,
+      "loss": 0.4304,
+      "step": 18329
+    },
+    {
+      "epoch": 3.0171304741397047,
+      "grad_norm": 0.32813383839828936,
+      "learning_rate": 3.1161310873040504e-06,
+      "loss": 0.4279,
+      "step": 18330
+    },
+    {
+      "epoch": 3.017295058358821,
+      "grad_norm": 0.2798937789325171,
+      "learning_rate": 3.1156838106208675e-06,
+      "loss": 0.423,
+      "step": 18331
+    },
+    {
+      "epoch": 3.0174596425779376,
+      "grad_norm": 0.32055734462181357,
+      "learning_rate": 3.11523654904279e-06,
+      "loss": 0.4241,
+      "step": 18332
+    },
+    {
+      "epoch": 3.017624226797054,
+      "grad_norm": 0.3172206049676667,
+      "learning_rate": 3.1147893025746985e-06,
+      "loss": 0.4325,
+      "step": 18333
+    },
+    {
+      "epoch": 3.0177888110161706,
+      "grad_norm": 0.3434309037151508,
+      "learning_rate": 3.1143420712214735e-06,
+      "loss": 0.4463,
+      "step": 18334
+    },
+    {
+      "epoch": 3.017953395235287,
+      "grad_norm": 0.3742062244031275,
+      "learning_rate": 3.113894854987994e-06,
+      "loss": 0.434,
+      "step": 18335
+    },
+    {
+      "epoch": 3.0181179794544035,
+      "grad_norm": 0.25831774897452026,
+      "learning_rate": 3.1134476538791415e-06,
+      "loss": 0.4124,
+      "step": 18336
+    },
+    {
+      "epoch": 3.0182825636735195,
+      "grad_norm": 0.3204505571282885,
+      "learning_rate": 3.1130004678997944e-06,
+      "loss": 0.4201,
+      "step": 18337
+    },
+    {
+      "epoch": 3.018447147892636,
+      "grad_norm": 0.3682921298167499,
+      "learning_rate": 3.112553297054834e-06,
+      "loss": 0.4264,
+      "step": 18338
+    },
+    {
+      "epoch": 3.0186117321117525,
+      "grad_norm": 0.23909293619699393,
+      "learning_rate": 3.1121061413491355e-06,
+      "loss": 0.4198,
+      "step": 18339
+    },
+    {
+      "epoch": 3.018776316330869,
+      "grad_norm": 0.29361493096236757,
+      "learning_rate": 3.1116590007875837e-06,
+      "loss": 0.4076,
+      "step": 18340
+    },
+    {
+      "epoch": 3.0189409005499854,
+      "grad_norm": 0.4612104279180189,
+      "learning_rate": 3.1112118753750543e-06,
+      "loss": 0.4377,
+      "step": 18341
+    },
+    {
+      "epoch": 3.019105484769102,
+      "grad_norm": 0.281475975355378,
+      "learning_rate": 3.110764765116425e-06,
+      "loss": 0.438,
+      "step": 18342
+    },
+    {
+      "epoch": 3.0192700689882184,
+      "grad_norm": 0.3623311275412427,
+      "learning_rate": 3.1103176700165773e-06,
+      "loss": 0.4306,
+      "step": 18343
+    },
+    {
+      "epoch": 3.019434653207335,
+      "grad_norm": 0.37699299963368665,
+      "learning_rate": 3.1098705900803884e-06,
+      "loss": 0.4141,
+      "step": 18344
+    },
+    {
+      "epoch": 3.0195992374264513,
+      "grad_norm": 0.33703887500358937,
+      "learning_rate": 3.109423525312737e-06,
+      "loss": 0.4311,
+      "step": 18345
+    },
+    {
+      "epoch": 3.019763821645568,
+      "grad_norm": 0.3045504218476557,
+      "learning_rate": 3.1089764757185005e-06,
+      "loss": 0.4219,
+      "step": 18346
+    },
+    {
+      "epoch": 3.0199284058646843,
+      "grad_norm": 0.3574769206992018,
+      "learning_rate": 3.108529441302558e-06,
+      "loss": 0.4169,
+      "step": 18347
+    },
+    {
+      "epoch": 3.0200929900838007,
+      "grad_norm": 0.31503034784782696,
+      "learning_rate": 3.1080824220697882e-06,
+      "loss": 0.4291,
+      "step": 18348
+    },
+    {
+      "epoch": 3.020257574302917,
+      "grad_norm": 0.38223697652308825,
+      "learning_rate": 3.107635418025066e-06,
+      "loss": 0.423,
+      "step": 18349
+    },
+    {
+      "epoch": 3.0204221585220337,
+      "grad_norm": 0.3895900387968357,
+      "learning_rate": 3.1071884291732707e-06,
+      "loss": 0.4337,
+      "step": 18350
+    },
+    {
+      "epoch": 3.02058674274115,
+      "grad_norm": 0.38589754748451754,
+      "learning_rate": 3.106741455519278e-06,
+      "loss": 0.4202,
+      "step": 18351
+    },
+    {
+      "epoch": 3.0207513269602666,
+      "grad_norm": 0.474879256779032,
+      "learning_rate": 3.106294497067968e-06,
+      "loss": 0.4267,
+      "step": 18352
+    },
+    {
+      "epoch": 3.020915911179383,
+      "grad_norm": 0.3209696342392641,
+      "learning_rate": 3.105847553824216e-06,
+      "loss": 0.4245,
+      "step": 18353
+    },
+    {
+      "epoch": 3.0210804953984995,
+      "grad_norm": 0.34778376983732656,
+      "learning_rate": 3.105400625792899e-06,
+      "loss": 0.4101,
+      "step": 18354
+    },
+    {
+      "epoch": 3.021245079617616,
+      "grad_norm": 0.33399183177152747,
+      "learning_rate": 3.1049537129788937e-06,
+      "loss": 0.4097,
+      "step": 18355
+    },
+    {
+      "epoch": 3.0214096638367325,
+      "grad_norm": 0.2988899506318333,
+      "learning_rate": 3.104506815387077e-06,
+      "loss": 0.4483,
+      "step": 18356
+    },
+    {
+      "epoch": 3.021574248055849,
+      "grad_norm": 0.3500392847200933,
+      "learning_rate": 3.1040599330223264e-06,
+      "loss": 0.447,
+      "step": 18357
+    },
+    {
+      "epoch": 3.0217388322749654,
+      "grad_norm": 0.48930436377342484,
+      "learning_rate": 3.1036130658895134e-06,
+      "loss": 0.4277,
+      "step": 18358
+    },
+    {
+      "epoch": 3.021903416494082,
+      "grad_norm": 0.3234127630400757,
+      "learning_rate": 3.1031662139935205e-06,
+      "loss": 0.4391,
+      "step": 18359
+    },
+    {
+      "epoch": 3.0220680007131984,
+      "grad_norm": 0.27320307065801214,
+      "learning_rate": 3.102719377339217e-06,
+      "loss": 0.4229,
+      "step": 18360
+    },
+    {
+      "epoch": 3.022232584932315,
+      "grad_norm": 0.5085343141406332,
+      "learning_rate": 3.1022725559314855e-06,
+      "loss": 0.4194,
+      "step": 18361
+    },
+    {
+      "epoch": 3.0223971691514313,
+      "grad_norm": 0.2773280004945635,
+      "learning_rate": 3.101825749775196e-06,
+      "loss": 0.4516,
+      "step": 18362
+    },
+    {
+      "epoch": 3.0225617533705478,
+      "grad_norm": 0.382392052409632,
+      "learning_rate": 3.1013789588752266e-06,
+      "loss": 0.444,
+      "step": 18363
+    },
+    {
+      "epoch": 3.0227263375896642,
+      "grad_norm": 0.36171418520805143,
+      "learning_rate": 3.1009321832364516e-06,
+      "loss": 0.4243,
+      "step": 18364
+    },
+    {
+      "epoch": 3.0228909218087807,
+      "grad_norm": 0.3450737581253108,
+      "learning_rate": 3.1004854228637462e-06,
+      "loss": 0.442,
+      "step": 18365
+    },
+    {
+      "epoch": 3.023055506027897,
+      "grad_norm": 0.49415695404695414,
+      "learning_rate": 3.1000386777619857e-06,
+      "loss": 0.4384,
+      "step": 18366
+    },
+    {
+      "epoch": 3.0232200902470137,
+      "grad_norm": 0.3017298209376994,
+      "learning_rate": 3.099591947936044e-06,
+      "loss": 0.4414,
+      "step": 18367
+    },
+    {
+      "epoch": 3.02338467446613,
+      "grad_norm": 0.4087842470656403,
+      "learning_rate": 3.099145233390797e-06,
+      "loss": 0.4327,
+      "step": 18368
+    },
+    {
+      "epoch": 3.0235492586852466,
+      "grad_norm": 0.40924406967671867,
+      "learning_rate": 3.0986985341311157e-06,
+      "loss": 0.4331,
+      "step": 18369
+    },
+    {
+      "epoch": 3.023713842904363,
+      "grad_norm": 0.2763617320415296,
+      "learning_rate": 3.098251850161879e-06,
+      "loss": 0.4223,
+      "step": 18370
+    },
+    {
+      "epoch": 3.023878427123479,
+      "grad_norm": 0.9895023088998105,
+      "learning_rate": 3.0978051814879594e-06,
+      "loss": 0.4436,
+      "step": 18371
+    },
+    {
+      "epoch": 3.0240430113425956,
+      "grad_norm": 0.3230702156672306,
+      "learning_rate": 3.0973585281142276e-06,
+      "loss": 0.4243,
+      "step": 18372
+    },
+    {
+      "epoch": 3.024207595561712,
+      "grad_norm": 0.334355242509786,
+      "learning_rate": 3.0969118900455616e-06,
+      "loss": 0.4267,
+      "step": 18373
+    },
+    {
+      "epoch": 3.0243721797808285,
+      "grad_norm": 0.30770411266118664,
+      "learning_rate": 3.0964652672868314e-06,
+      "loss": 0.4324,
+      "step": 18374
+    },
+    {
+      "epoch": 3.024536763999945,
+      "grad_norm": 0.2885193574945699,
+      "learning_rate": 3.096018659842913e-06,
+      "loss": 0.4276,
+      "step": 18375
+    },
+    {
+      "epoch": 3.0247013482190614,
+      "grad_norm": 0.34113089461111257,
+      "learning_rate": 3.0955720677186783e-06,
+      "loss": 0.4392,
+      "step": 18376
+    },
+    {
+      "epoch": 3.024865932438178,
+      "grad_norm": 0.452973237345632,
+      "learning_rate": 3.0951254909190017e-06,
+      "loss": 0.4398,
+      "step": 18377
+    },
+    {
+      "epoch": 3.0250305166572944,
+      "grad_norm": 0.34959258473425364,
+      "learning_rate": 3.0946789294487554e-06,
+      "loss": 0.4286,
+      "step": 18378
+    },
+    {
+      "epoch": 3.025195100876411,
+      "grad_norm": 0.3275405935360879,
+      "learning_rate": 3.0942323833128095e-06,
+      "loss": 0.4168,
+      "step": 18379
+    },
+    {
+      "epoch": 3.0253596850955273,
+      "grad_norm": 0.3018346913482798,
+      "learning_rate": 3.093785852516041e-06,
+      "loss": 0.4273,
+      "step": 18380
+    },
+    {
+      "epoch": 3.025524269314644,
+      "grad_norm": 0.36688729357237526,
+      "learning_rate": 3.0933393370633182e-06,
+      "loss": 0.42,
+      "step": 18381
+    },
+    {
+      "epoch": 3.0256888535337603,
+      "grad_norm": 0.4861306614021659,
+      "learning_rate": 3.0928928369595167e-06,
+      "loss": 0.3911,
+      "step": 18382
+    },
+    {
+      "epoch": 3.0258534377528767,
+      "grad_norm": 0.3492407070341485,
+      "learning_rate": 3.0924463522095054e-06,
+      "loss": 0.4361,
+      "step": 18383
+    },
+    {
+      "epoch": 3.026018021971993,
+      "grad_norm": 0.3980551315439607,
+      "learning_rate": 3.0919998828181593e-06,
+      "loss": 0.4216,
+      "step": 18384
+    },
+    {
+      "epoch": 3.0261826061911097,
+      "grad_norm": 0.38357388279611354,
+      "learning_rate": 3.0915534287903485e-06,
+      "loss": 0.419,
+      "step": 18385
+    },
+    {
+      "epoch": 3.026347190410226,
+      "grad_norm": 0.4891027846917855,
+      "learning_rate": 3.091106990130943e-06,
+      "loss": 0.4204,
+      "step": 18386
+    },
+    {
+      "epoch": 3.0265117746293426,
+      "grad_norm": 0.3391882237310598,
+      "learning_rate": 3.0906605668448177e-06,
+      "loss": 0.4179,
+      "step": 18387
+    },
+    {
+      "epoch": 3.026676358848459,
+      "grad_norm": 0.33533144601984904,
+      "learning_rate": 3.0902141589368396e-06,
+      "loss": 0.4356,
+      "step": 18388
+    },
+    {
+      "epoch": 3.0268409430675756,
+      "grad_norm": 0.3100015251700531,
+      "learning_rate": 3.0897677664118844e-06,
+      "loss": 0.4194,
+      "step": 18389
+    },
+    {
+      "epoch": 3.027005527286692,
+      "grad_norm": 0.3667714832533851,
+      "learning_rate": 3.089321389274818e-06,
+      "loss": 0.4085,
+      "step": 18390
+    },
+    {
+      "epoch": 3.0271701115058085,
+      "grad_norm": 0.32901637661456695,
+      "learning_rate": 3.088875027530516e-06,
+      "loss": 0.4409,
+      "step": 18391
+    },
+    {
+      "epoch": 3.027334695724925,
+      "grad_norm": 0.3926995176310053,
+      "learning_rate": 3.088428681183845e-06,
+      "loss": 0.425,
+      "step": 18392
+    },
+    {
+      "epoch": 3.0274992799440414,
+      "grad_norm": 0.33404719587309123,
+      "learning_rate": 3.087982350239677e-06,
+      "loss": 0.4376,
+      "step": 18393
+    },
+    {
+      "epoch": 3.027663864163158,
+      "grad_norm": 0.27830564539727404,
+      "learning_rate": 3.087536034702883e-06,
+      "loss": 0.4219,
+      "step": 18394
+    },
+    {
+      "epoch": 3.0278284483822744,
+      "grad_norm": 0.411080967064806,
+      "learning_rate": 3.0870897345783307e-06,
+      "loss": 0.4101,
+      "step": 18395
+    },
+    {
+      "epoch": 3.027993032601391,
+      "grad_norm": 0.337151936162758,
+      "learning_rate": 3.0866434498708924e-06,
+      "loss": 0.4232,
+      "step": 18396
+    },
+    {
+      "epoch": 3.0281576168205073,
+      "grad_norm": 0.28520005890645866,
+      "learning_rate": 3.086197180585436e-06,
+      "loss": 0.4461,
+      "step": 18397
+    },
+    {
+      "epoch": 3.028322201039624,
+      "grad_norm": 0.33666331248057874,
+      "learning_rate": 3.0857509267268324e-06,
+      "loss": 0.4296,
+      "step": 18398
+    },
+    {
+      "epoch": 3.0284867852587403,
+      "grad_norm": 0.30655635466216224,
+      "learning_rate": 3.0853046882999488e-06,
+      "loss": 0.4181,
+      "step": 18399
+    },
+    {
+      "epoch": 3.0286513694778567,
+      "grad_norm": 0.37866615924108316,
+      "learning_rate": 3.084858465309659e-06,
+      "loss": 0.4393,
+      "step": 18400
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 30375,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 5,
+  "save_steps": 400,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 3.8309052386849587e+18,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}